Emerging Infectious Diseases
[Volume 6 No.2 / March - April 2000]

Vol. 6, No. 2 | March–April 2000 

Perspectives

Rumors of Disease in the Global Village: Outbreak Verification, 
T.G.Grein

Malaria on the Move: Human Population Movement and Malaria 
Transmission, P. Martens

Synopses 

The bdr Gene Families of the Lyme Disease and Relapsing Fever 
Spirochetes: Potential Influence on Biology, Pathogenesis, and 
Evolution, D.M. Roberts

Attenuated Viral Vaccines for Mucosal Immunity to Emerging 
Infectious Diseases, F. van Ginkel

Research 

American Robins as Reservoir Hosts for Lyme Disease, D. Richter
Vibrio cholerae O139 in Calcutta, 1992-1998: Incidence, 
Antibiograms, and Genotypes, A. Basu

Multivariate Markovian Modeling of Tuberculosis, S.M. Debanne

Serologic Response to Culture Filtrate Antigens of Mycobacterium 
ulcerans during Buruli Ulcer Disease, K.M. Dobos

Role of H. influenzae Type B and S. pneumoniae in Pneumonia in 
China, O.S. Levine

Dispatches 

Bacteroides fragilis Enterotoxin Gene Sequences in Patients with 
Inflammatory Bowel Disease, T.P. Prindiville

Outbreak among Drug Users Caused by a Clonal Strain of Group A 
Streptococcus, L.M. Böhlen

Erythromycin Resistance in S. pyogenes in Italy, M. Bassetti

SNV Antibody Response during Two Phases of HPS, P. Bostik

Bovine TB Threatens the Endangered Iberian Lynx, V. Briones

Haff Disease: From the Baltic Sea to the U.S. Shore, U. Buchholz

An Epizootic of C. pneumoniae in African Clawed Frogs, K.D. Reed

The Impact of Health Communication and Enhanced Laboratory-Based 
Surveillance on Detection of Cyclosporiasis Outbreaks in California, 
J.C. Mohle-Boetani

Norwalk-Like Viral Gastroenteritis Outbreak in U.S. Army Trainees, 
M.K. Arness

Letters 

Preventing Zoonotic Diseases in Immunocompromised Persons: The Role 
of Physicians and Veterinarians, N. Nowotny

Reply to Drs. Nowotny and Deutz, C. W. Olsen

Book Reviews 

International Law and Infectious Diseases (D.P. Fidler), D. 
Patterson

Oxford Handbook of Tropical Medicine (M. Eddleston and S. Pierini), 
R. Beharry

News and Notes 

Conference Summary

Upcoming Events

ICEID 2000—Call for Abstracts       



Perspectives

Rumors of Disease in the Global Village: Outbreak Verification

Thomas W. Grein, Kande-Bure O. Kamara, Guénaël Rodier,
Aileen J. Plant, Patrick Bovier, Michael J. Ryan, Takaaki Ohyama,
and David L. Heymann
World Health Organization, Geneva, Switzerland

Emerging infectious diseases and the growth of information
technology have produced new demands and possibilities for
disease surveillance and response. Increasing numbers of outbreak
reports must be assessed rapidly so that control efforts can be
initiated and unsubstantiated reports can be identified to protect
countries from unnecessary economic damage. The World Health
Organization has set up a process for timely outbreak verification
to convert large amounts of data into accurate information for
suitable action. We describe the context and processes of outbreak
verification and information dissemination.

Globalization presents new challenges and opportunities in
combating diseases likely to cause epidemics. As a result of
increased international travel and trade, local events acquire
international importance. At the same time, the rapid global
expansion of telecommunications and broadened access to news
media and the Internet have changed the way society treats
information. Reports of disease outbreaks are more widely
disseminated and more easily accessible than ever before.
However, the quality of information is no longer controlled and
may be provided out of context, often causing unnecessary public
anxiety and confusion. Rumors that later prove to be
unsubstantiated may lead to inappropriate response, causing
disruption in travel and trade and economic loss to affected
countries. The World Health Organization (WHO), speaking for
191 member countries, is uniquely positioned to coordinate
infectious disease surveillance and response at the global level.
WHO receives reports of disease outbreaks around the world from
various sources. While some of these reports are warnings of
genuine epidemics, others may reflect endemic disease or may be
mere rumors.


To investigate and follow up outbreak reports, WHO established
an innovative mecha nism—outbreak verification—in early 1997.
Outbreak verification is a new approach to global disease
surveillance (1). Its aim is to improve epidemic disease control by
informing key public health professionals about confirmed and
unconfirmed outbreaks of international public health importance.

The Outbreak Verification System

The outbreak verification system follows the general principles of
surveillance: systematic collection, collation, analysis, and
interpretation of data and dissemination to those who need the
information for action (Figure 1). Data derived from an extensive
network of information sources are transformed by the outbreak
verification team into timely, accurate information about important
disease outbreaks. When the outbreak verification team receives an
unconfirmed outbreak report, the relevance to international public
health is assessed, and, if appropriate, further information is
sought. Once an outbreak is substantiated and considered of public
health importance, information is rapidly disseminated to a
network of international partners.

Figure 1. Outbreak verification at the World Health Organization.

Sources of Information

Outbreak verification is based on a broad range of information
sources, including national institutes of public health, WHO offices
at regional and national level, the United Nations system,
nongovernmental organizations, WHO collaborating centers,
newspapers, television, and radio (2). With the advent of modern
communication technologies, many initial outbreak reports now
originate in the electronic media and electronic discussion groups.
Indeed, the abundance of outbreak-related documents on the World
Wide Web presents a challenge: identifying reports of global
public health importance.

The tasks of identifying and extracting outbreak reports from the
electronic media is mainly performed by the Global Public Health
Information Network (GPHIN), an electronic surveillance system
developed by Health Canada. GPHIN continuously monitors some
600 sources, including all major news wires, newspapers, and
biomedical journals. The system focuses its search on
communicable diseases but will soon also cover noncommunicable
diseases, food and water safety, environmental health risks, and the
health impact of natural disasters (3). The quality of reports
retrieved by GPHIN varies considerably, and information may be
presented out of context (4).

Other information providers are the Internet and electronic-mail-
based discussion groups. Their scope and readership may be
worldwide (e.g., ProMed), regional (e.g., PACNET in the Pacific
region), or specific (e.g., TravelMed). These groups can be
accessed through free and unrestricted subscription. Because they
receive outbreak information from many sources, including sources
other than the electronic media, they are valuable information
providers (5).

Selection of Outbreak Reports for Verification

The verification team first determines if an event is of potential
international public health importance. International public health
importance has been defined as serious health impact or
unexpectedly high rates of illness and death; potential for spread
beyond national borders; interference with international travel or
trade; or likely need for international assistance in disease control.

Each event is assessed individually on the basis of these criteria.
While some diseases will almost always be regarded important for
international public health (e.g., Ebola hemorrhagic fever, cholera),
others may not, depending on the circumstances.

Process of Verification

Once an event has been assessed as of potential international
importance, the process of verification is initiated. The outbreak
verification team establishes the potential importance of the event,
on the basis of available background information, endemicity
levels, and details of previous outbreaks. This information is then
shared by e-mail with designated contacts in WHO regional
offices, who seek confirmation of details from health authorities in
the countries concerned, usually through the WHO representative.
The outbreak verification team may seek additional information
from other organizations in the field, such as the International Red
Cross, Médecins sans Frontières, and Medical Emergency Relief
International.

Upon receipt of feedback, the outbreak verification team
determines if the event meets the definition of an outbreak
(observed number of cases exceeds expected number of cases in a
given population for a given period) and the criteria for
international public health importance. Reaching a final decision
may require further consultation with the WHO regional office or
the country representative or health authorities in-country.

Dissemination of Information

Timely dissemination of outbreak information to those who need to
know is a key aspect of the outbreak verification process, and
details of outbreaks with potential for international public health
importance are disseminated through various channels. Information
is shared directly with partners for immediate action (epidemic
response) but also routinely with a wider audience through the
Outbreak Verification List, the WHO Disease Outbreak News on
the World Wide Web, and the Weekly Epidemiological Record
(WER). The Outbreak Verification List is distributed weekly by e-
mail to approximately 800 subscribers. The distribution list
includes WHO staff worldwide, other UN agencies, national health
authorities, field epidemiology training programs, and
nongovernmental organizations. Because the Outbreak Verification
List is not an official WHO publication, its distribution is limited
to subscribers.


The WHO Disease Outbreak News is on the WHO web page and
provides the public with information about outbreaks of
international importance. Often events that initially appeared in the
Outbreak Verification List are subsequently reported in Outbreak
News. Because Outbreak News is in the public domain, only
information about officially confirmed outbreaks is disseminated.
Outbreak News (http://
www.who.int/emc/outbreak_news/index.html) is one of the most
frequently accessed sites on the WHO home page.

The third mechanism for communicating outbreak-related
information is the WER. This report is published in French and
English and issued in print and electronically (http://
www.who.int/wer/index.html). It covers epidemiologic
information on cases and outbreaks of diseases under the
International Health Regulations (yellow fever, plague, cholera)
and also on other communicable diseases of public health
importance. Recently, an Outbreak News section mirroring the
Outbreak News on the web page has been added to the WER.

Outbreak Response

Coordination of timely and effective epidemic response is
intrinsically linked to dissemination of information about important
disease outbreaks. During the verification process, WHO routinely
offers technical assistance for the investigation and control of the
event. Such assistance may range from advice (e.g., identifying
appropriate laboratory facilities) to deployment of field teams.
WHO coordinates the deployment of field teams, drawing from
within WHO and among collaborating centers and other
international partners.

Effectiveness of Outbreak Verification

From July 1, 1997, to July 1, 1999, the outbreak verification team
identified 246 outbreak reports of potential importance for world
health and disseminated them in the Outbreak Verification List. Of
the 246 outbreaks, 43% occurred in the African region of WHO;
12% each in the regions of the Americas, eastern Mediterranean,
and Europe; 11% in the Southeast Asian region; and 9% in the
Western Pacific region. Countries subject to complex emergencies
were involved in 121 (49%) outbreaks and industrialized countries
in 6 (2%) events. The most common diseases or syndromes
disseminated in the Outbreak Verification List were cholera (n =
78), acute hemorrhagic fevers (n = 24), and acute diarrheal diseases
(n = 22). In two (0.8%) cases, the Outbreak Verification List
disseminated information about events that could not be
substantiated later (Figure 2). Seventy-one percent of the initial
reports were retrieved from informal or unofficial sources (e.g., the
media, electronic discussion groups, nongovernmental
organizations), and 29% were provided by official sources (e.g.,
WHO network, Ministries of Health). Unofficial sources were the
most frequent providers of initial information in all WHO regions
and for all diseases, including those subject to the International
Health Regulations (cholera, plague, yellow fever). Information
about the date of onset of an outbreak was available in 134 (55%)
cases. The median time between reported onset of an outbreak and
the outbreak verification team's receipt of the first report was 18
days (from 1 to 215 days). This interval was similar for official and
unofficial sources but varied considerably for different diseases: 13
to 15 days (median) for acute hemorrhagic fevers, anthrax, and
cholera; 20 to 35 days (median) for yellow fever and plague; and
>50 days (median) for acute respiratory syndrome and
meningococcal disease. Most reports were verified within a few
days and important events usually within <48 hours. The median
time between receipt of a first report and appearance of the event in
the weekly Outbreak Verification List was 3 days (0 to 69 days). In
addition to the 246 disseminated outbreak reports, 69 events were
verified from July 1, 1997, to July 1, 1999, but were not reported in
the Outbreak Verification List. Follow up was undertaken because
initial reports suggested international public health importance. Of
the 69 events, 58 (84%) were excluded from the Outbreak
Verification List because they did not meet the criteria for
outbreaks or for international public health importance. Four (6%)
reports were unsubstantiated, including two reports of smallpox,
one of yellow fever, and one of viral hemorrhagic fever. In seven
(10%) events, follow up could not be completed, and the
verification process remained inconclusive. The 69 excluded
events did not differ from the 246 disseminated outbreaks with
regard to their distribution by WHO region, initial source of
information, or type of disease or syndrome. A reassessment of the
62 verified events did not identify any outbreaks that should have
been classified retrospectively as of international importance.

Figure 2. Reports of outbreaks disseminated in Outbreak
Verification List, July 1, 1997, to July 1, 1999 (n = 246).

Whenever the outbreak verification team invokes a verification
process, assistance to the country in which the event takes place is
offered directly by WHO headquarters or through the WHO
regional and country offices. Past examples of such assistance
include supply of essential materials to outbreak sites, transport of
laboratory specimens from the field to appropriate diagnostic
facilities, organization of vaccination programs, training of field
staff as part of outbreak control measures, or deployment of field
teams for disease control. Recent examples of direct assistance by
WHO and its partners in field investigations include support for
Rift Valley fever in Kenya and Somalia (6), monkeypox in the
Democratic Republic of the Congo (7), avian influenza (H5N1) in
Hong Kong, Special Administrative Region of China, Ebola
hemorrhagic fever in Gabon (8), relapsing fever and acute
respiratory infections in southern Sudan, influenza in Afghanistan,
and Marburg virus infection in the Democratic Republic of the
Congo.

Conclusions

Outbreak verification is a new approach to global disease
surveillance. Its aim is to improve epidemic disease control by
providing accurate and timely information about important disease
outbreaks. While the outbreak verification concept has remained
unchanged since its start in early 1997, its daily application
continues to evolve as more data are gathered and more experience
is gained.

Currently, most outbreak reports are received from the media, and
field personnel are mainly contacted for assistance with verifying
reported events. This approach is subject to information bias,
which results from the uneven dispersal and use of modern
technology throughout the world. Also, different languages are not
equally represented in the news media or addressed by electronic
search engines. While these shortcomings are partly offset by the
information received directly from the WHO network, a more
active dialogue should be established with field personnel.
Receiving primary information directly from the field will lead to
earlier detection of important events and events that escape
identification. Although thought to be small, the number of
important outbreaks recognized only locally is unknown. The
number of outbreak reports selected for verification is small
compared with the number of reports received by the outbreak
verification team. While the criteria for selecting outbreak reports
for verification have been established, their application requires an
individual assessment of each single event. Some see in this
selection process a lack of transparency and argue that the reader is
the best judge of what to believe. This may be the case for those
who have time, good information networks, and access to
advanced communication technology. However, most international
public health workers have none of these and are poorly informed
about such events. WHO therefore considers that sharing filtered
information is valuable. In a recent survey among the Outbreak
Verification List recipients, 72% percent of the respondents stated
that the list was very useful or indispensable to their work, and
70% cited the list as their first source of information about a
particular event. Applying the selection criteria is also difficult if
available information is insufficient to determine if an event should
be classified as an outbreak (number of cases in excess of expected
numbers). This problem arises particularly when dealing with
endemic diseases in the absence of established epidemic
thresholds. The Outbreak Verification List addresses the issue by
mentioning events with clear implications for international public
health that are not regarded as outbreaks in a separate Notes
section. The Outbreak Verification List shares relevant and often
sensitive information with public health professionals while the
verification process is still under way. Although this has led on rare
occasions (<1%) to the dissemination of information about
unsubstantiated events, the Outbreak Verification List usually
provides timely and accurate information about important disease
outbreaks.

Because of its confidential nature, the Outbreak Verification List is
not in the public domain, and some argue that WHO is not timely
in addressing the information needs of the public about epidemics
(4). However, WHO communicates information as soon as it is
verified. In some instances, this takes time, but the delay prevents
release of inaccurate information. Industrialized countries feature
infrequently in the Outbreak Verification List because it is assumed
that they can deal with outbreak situations. This is, of course, not
always true and leads to an overrepresentation of developing
countries in the Outbreak Verification List. However, most
outbreaks in developing countries are contributed by nations with
complex emergencies. While the reporting may accurately reflect
the breakdown of the public health and social infrastructures, it
may also contain an element of overreporting due to heightened
media attention associated with complex emergencies. As a new
concept, early outbreak verification efforts focused mainly on the
development of process indicators (information gathering,
verification, information dissemination). More outcome- oriented
indicators need to be addressed to assess the outbreak verification
impact at country level and within WHO. While providing public
health professionals with timely and accurate information about
important disease outbreaks improves epidemic preparedness and
response, this has not been quantified. Possible outcome indicators
could include the time interval between first report and the
commencement of investigation and control efforts or the
proportion of outbreaks with laboratory confirmation. Additional
tasks to be addressed in the future are more detailed analyses,
including electronic and print mapping to provide both baseline
(endemic) and outbreak information, and standardized reports to
regions and countries.

Dr. Grein is a medical officer in the Department of Communicable
Disease Surveillance and Response at the World Health
Organization in Geneva, Switzerland. His activities at WHO
include the investigation and control of epidemics and training in
field epidemiology.

Address for correspondence: Thomas Grein, Department of
Communicable Disease Surveillance and Response, World Health
Organization, 20 Avenue Appia, CH-1211 Geneva 27,
Switzerland; fax: 41-22-791-4198; e-mail: greint@who.int

References

1. Heymann DL, Rodier GR. Global surveillance of communicable
diseases. Emerg Infect Dis 1998;4:362- 5.
2. World Health Organization. Summary records and reports of
committees of the 48th World Health Assembly. Geneva: the
Organization; 1995 Report No. WHA48/1995/REC/3.
3. Cripp R. Docs using net as disease detector. Wired News [serial
online] 1998 Apr. Available from URL:
http://www.wired.com/news/news/technology/story/ 11466.htm
4. Eysenbach G, Diepgen TL. Towards quality management of
medical information on the internet: evaluation, labelling, and
filtering of information. BMJ 1998;317:1496-502.
5. Taubes G. Virus hunting on the web. Technology review [serial
online] 1998 Nov-Dec. Available from URL:
http://www.techreview.com/articles/nov98/ taubes.htm
6. An outbreak of Rift Valley fever, Eastern Africa, 1997- 98.
Wkly Epidemiol Rec 1998;73:105-12. 7. Human monkeypox in
Kasai Oriental, Democratic Republic of the Congo (former Zaire)-
preliminary report of October 1997 investigation. Wkly Epidemiol
Rec 1997;72:365-72.
8. Ebola haemorrhagic fever. A summary of the outbreak in Gabon.
Wkly Epidemiol Rec 1997;72:7-8.


Perspectives

Malaria on the Move: Human Population Movement and Malaria
Transmission

Pim Martens and Lisbeth Hall
Maastricht University, Maastricht, the Netherlands

Reports of malaria are increasing in many countries and in areas
thought free of the disease. One of the factors contributing to the
reemergence of malaria is human migration. People move for a
number of reasons, including environmental deterioration,
economic necessity, conflicts, and natural disasters. These factors
are most likely to affect the poor, many of whom live in or near
malarious areas. Identifying and understanding the influence of
these population movements can improve prevention measures and
malaria control programs.

Malaria, the world's most prevalent vector-borne disease, is
endemic in 92 countries, with pockets of transmission in an
additional eight countries (1). Approximately 41% of the world's
population is at risk, and each year 300 million to 500 million
clinical cases of malaria, >90% of them in Africa, are reported.
Worldwide, approximately 2 million deaths per year can be
attributed to malaria, half of these in children under 5 years of age.

Historically, population movement has contributed to the spread of
disease (2). Failure to consider this factor contributed to failure of
malaria eradication campaigns in the 1950s and 1960s (3). The
movement of infected people from areas where malaria was still
endemic to areas where the disease had been eradicated led to
resurgence of the disease. However, population movement can
precipitate or increase malaria transmission in other ways as well.
As people move, they can increase their risk for acquiring the
disease through the ways in which they change the environment
and through the technology they introduce, for example, through
deforestation and irrigation systems (4). Such activities can create
more favorable habitats for Anopheles mosquitoes; at the same
time, workers may have increased exposure to the vector.
Furthermore, people can inadvertently transport infectious
mosquitoes to malaria-free areas, reintroducing disease. Population
movement is also increasingly implicated in the spread of drug
resistance in malaria (5).

The unprecedented increase in mobility in the last few decades has
led to greater concern about the relationship between mobility and
malaria. There are a number of reasons for increased mobility.
First, sophisticated forms of transport now permit the swift
movement of people over huge distances. Air travel has increased
by almost 7% a year in the last 20 years and is predicted to increase
by >5% a year during the next 20 years (6). Second, in the
developing world a rapidly increasing population is putting
pressure on scarce resources, leading to major population
redistribution. This particularly involves the movement from rural
to urban areas. Third, natural disasters such as droughts and floods
have created approximately 25 million environmental refugees (7).
Finally, conflict, often a result of population pressures and
environmental degradation, displaces vast numbers of people. We
examine the impact of population movement on malaria
transmission.

A Typology of Population Movement

The decision-making process leading to population movement can
best be understood in the light of "push and pull" forces (8). When
their needs can no longer be met in a particular environment,
people move elsewhere. The "push factor" could be environmental
degradation, population pressure on land, droughts, famines,
conflict, or loss or lack of employment. When people are satisfied
with their situation but believe that a move elsewhere will provide
new and attractive opportunities, a "pull factor" is involved. This
pull factor could be better political, economic, or social
opportunities or improved living conditions. Push and pull factors
can operate simultaneously; for example, people can be pushed by
environmental deterioration and scarce resources and pulled by the
economic opportunities offered by development projects.
Population movements can be differentiated by their temporal and
spatial dimensions. Temporal dimensions include circulation and
migration. Circulation encompasses a variety of movements,
usually short-term and cyclical and involving no longstanding
change of residence. Migration involves a permanent change of
residence (9). Circulation can be subdivided into daily, periodic,
seasonal, and long term (Table) (9). Daily circulation involves
leaving a place of residence for up to 24 hours. Periodic circulation
may vary from 1 night to 1 year, although it is usually shorter than
seasonal circulation. Seasonal circulation is a type of periodic
circulation in which the period is defined by marked seasonality in
the physical or economic environment. This type of circulation
involves persons or groups who are absent from their permanent
homes during a season or seasons of the year. Long-term
circulation, defined as absence from home for longer than a year,
affects groups such as wage laborers and traders, who maintain
close social and economic ties with their home area and intend to
return. In terms of spatial dimensions, the movements to and from
malarious areas are of epidemiologic importance. People who
move can be categorized as either active transmitters or passive
acquirers (2). Active transmitters harbor the parasite and transmit
the disease when they move to areas of low or sporadic
transmission. Passive acquirers are exposed to the disease through
movement from one environment to another; they may have low-
level immunity or may be nonimmune, which increases their risk
for disease.


Based on the above definitions, a typology can be devised that
identifies categories of population movement. Different activities
can be associated with these categories, and these activities, in
turn, can be associated with differing risks for malaria
transmission. All types of population movement can be
accommodated in this typology, and people may exhibit more than
one type of mobility.  Table. A typology of population movement
(adapted from [2])

Population Movement and Malaria

In developing countries, activities involving population movement
include urbanization, colonization, labor related to agriculture and
mining, and conflict. In industrialized countries, the impact of
population movements on malaria risk is mainly related to
intercontinental travel. In some regions, malaria risk may increase
as a result of a combination of different forms of mobility, as well
as other factors unrelated to population movements. For example,
in the African highlands, many of the issues described below act
concurrently (10). The categorization of the various examples
below simplifies a complex issue but is useful for indicating major
processes related to mobility and malaria risk.

Urbanization

The world's urban population is growing at four times the rate of
the rural population (11). Urban pull is prevalent throughout the
developing world, with rural-to-urban migration taking place faster
than ever before (12). Sub-Saharan Africa is the most rapidly
urbanizing region in the world (13), and the urban population in
India has doubled in the last 2 decades (14). When accompanied by
adequate housing and sanitation, urbanization can lead to a
decrease in malaria through reductions in human-vector contact
and vector breeding sites. However, in developing countries, rapid,
unregulated urbanization often leads to an increase in or
resumption of malaria transmission because of poor housing and
sanitation, lack of proper drainage of surface water, and use of
unprotected water reservoirs that increase human-vector contact
and vector breeding. Although water pollution in urban areas
usually leads to decreases in vector populations, some vectors,
such as An. arabiensis in the forest belt of West Africa, may adapt
to breeding in polluted waters (15). In Asia, An. stephensi is
proving adaptable to urban conditions, and in India it is a well-
established vector of urban malaria. In urban areas of India, water
is not supplied regularly and is stored in houses, providing
extensive breeding places for An. stephensi in overhead tanks and
cisterns. In periurban areas, where 25% to 40% of the urban
population live in poor housing without proper water supply and
drainage, another vector, An. culcifacies, also transmits malaria
(14). In India, the fact that the National Malaria Eradication
Program concentrated only on rural areas, ignoring the problem of
urban malaria, was one of the factors leading to a resurgence of
malaria in the 1970s (14). Several types of population movement
contributed to malaria transmission in India. First, circulation from
stable rural malaria areas to unstable urban areas had firmly
established malaria transmission in urban areas. Then, after the
National Malaria Eradication Program, rural areas became free of
endemic malaria but were receptive, so circulation from urban
areas back to rural areas reintroduced malaria transmission.
Changes in vector behavior (exophilic and exophagic behavior
limiting the effectiveness of spraying), vector resistance to
insecticides, and increasing drug resistance, especially in
Plasmodium falciparum (14), also played a role. Population
movement also contributed to drug resistance, with people of
different immune status moving from endemic- to nonendemic-
disease areas, accelerating transmission of resistant strains.

Colonization of New Territory

Through the interaction of a number of factors, the colonization of
unpopulated or sparsely populated areas may be accompanied by
an increase in malaria. Settlers, who have low-level immunity or
are nonimmune, may migrate into a disease-endemic area,
spreading the disease. Initially, housing tends to be basic, leading
to close human-vector contact. Moreover, housing is often near
rivers or lakes to facilitate water collection, increasing the exposure
of humans to mosquitoes. Activities to develop an area, such as
deforestation and irrigation, can increase the number of vector
breeding sites, contributing to an increase in malaria. Colonization
may be accompanied by major building projects, such as dams,
canals, highways, or mining activities—referred to as the tropical
aggregation of labor—which can further enhance malaria
transmission.

Reemergence of malaria through mobility related to colonization
occurred in Brazil. Malaria had been practically eradicated from
most areas of the Amazon region by the national malaria campaign
in the 1950s and 1960s (16). Since the 1960s, however, the
incidence of malaria has increased dramatically because of massive
population movements to colonize new territory. New highways
were built in the 1960s, linking the Amazon region to the rest of
the country and attracting laborers to work on road construction. In
the 1970s, many more people were attracted to the region by
agricultural settlements and hydroelectric projects. Finally, in the
1980s, the discovery of gold led to a greater influx of people, along
with the establishment of hundreds of mines throughout the region.
The population of Rondônia State, which received the greatest
number of migrants, increased from 113,000 in 1970 to 1,200,000
in 1990. Malaria cases in Rondônia increased from 20,000 to
174,000 in the same period. In Brazil as a whole, approximately
50,000 cases of malaria were reported in 1970; by 1990, reports
had increased to 577,520, representing 10% of the world's reported
cases outside Africa (17). Of this total, >98% were recorded in the
Amazon region.

The types of population movement involved in the colonization of
the Amazons are migration and long-term circulation from malaria-
free areas of Brazil to the malaria-endemic Amazon region. The
people involved are nonimmune passive acquirers who on
becoming infected can become active transmitters. If these active
transmitters return to their initial place of residence in a malaria-
free but highly receptive area, they can reintroduce the parasite and
initiate an outbreak of malaria. For example, in 1985, 26 new
active foci of malaria were recorded in Brazilian states outside the
Amazon region (16). Settlers in the Amazon region are highly
mobile, moving with daily, periodic, and seasonal circulation from
settlements in unstable disease-endemic regions to hyperendemic-
disease regions of the rainforests. This mobility keeps settlements
unstable and at high risk for epidemics through the constant flow
of parasitemic laborers (17,18).

Agricultural Labor

Swaziland provides an example of how agricultural labor has
changed the spread of malaria. In the 1950s, control measures
(DDT spraying) were successfully implemented in the lowveld so
that agricultural development could take place, and by 1959,
malaria had been all but eradicated from Swaziland (19). However,
agricultural developments in the 1950s involving an irrigation
project for the cultivation of sugar cane created conditions
favorable for malaria. Vector density increased, along with a high
frequency of feeding on humans, as no domestic or wild animals
were around the project area to serve as alternative hosts. This
resurgence of malaria was catalyzed by the reintroduction of
parasite carriers in the form of migrant workers from disease-
endemic areas of Mozambique, who were involved in migration or
long-term circulation to work on the sugar estates in the 1960s and
early 1970s (19).

In Colombia, the annual parasite index (defined as the ratio
between the number of cases reported and the population at risk)
has increased threefold since the 1960s (20). This increase seems
to be related to the migration of nonimmune people to areas such
as the Naya basin, where malaria is endemic, and to the circulation
of groups within the Naya basin. The circulation is predominantly
seasonal, related to agriculture. People descend from hills and
terraces, where malaria risk is minimal, to the malarious delta zone
to cultivate and harvest their crops. In doing so, they are exposed to
the anopheline population of the area and are at high risk for
malaria. A large number of people are involved in this circulation
(approximately 60% of the area's population is mobile for
approximately 4 months of the year), and this population density,
combined with the large vector population, maintains transmission
at high levels (20).

Refugees

The number of officially recognized refugees has steadily
increased, from approximately 5 million in 1980 to >20 million in
late 1994 (21). In addition, an estimated 25 million people have
fled their homes but remain internally displaced in their countries
of origin (22). The displacement of large numbers of people and
their circulation can favor malaria transmission. If refugees are
nonimmune, they could travel through or to malarious regions and
acquire the infection, and if they are infectious, they could
disseminate the disease to other areas.

Malaria is one of the most commonly reported causes of death
among refugees and has caused high rates of both illness and death
among refugees and displaced persons in disease-endemic
countries, such as Thailand, Sudan, Somalia, Burundi, Rwanda,
and the Democratic Republic of Congo (23). In recent outbreak
among Burundian refugees at a refugee camp in northwestern
Tanzania, deaths from malaria and anemia in children under 5
years of age have increased 10-fold since the outbreak, reflecting
the lack of immunity in this age group (24). In the Sahel region of
Africa, where civil wars and conflicts have occurred for many
decades and large numbers of displaced people live in resettlement
or refugee camps often located in lowland disease-endemic areas,
epidemics are common (25).

As a result of 15 years of continuous war, which displaced
hundreds of thousands of people, Luanda, the capital of Angola,
underwent an unprecedented population increase in the 1980s. This
population movement resulted in a shift in malaria endemicity in
Luanda from hypoendemic to mesoendemic level within 5 years
(26). As a cause of child deaths, malaria moved from sixth to first
place. Increasing parasite resistance to chloroquine also became a
major problem. This situation arose because of the enormous
influx of displaced people of low socioeconomic status into an
environment with stagnant water reservoirs. The population
movements that increased malaria transmission in Luanda were
long-term circulation and migration from stable rural areas to an
unstable urban area.

Besides movements of large numbers of people, wars and civil
unrest tend to favor malaria transmission. The disruptive effect of
war on agriculture and water management can increase vector
breeding sites; the destruction of housing can increase human-
vector contact; the destruction of cattle can prompt zoophilic
vectors to become anthropophilic if their usual food supply is
disrupted (27); and control measures can be seriously diminished if
health-care facilities are reduced or unavailable.

Intercontinental Travel

The intercontinental transfer of malaria can occur through the
introduction of an infective vector into a nonendemic-disease area,
as in so-called airport malaria, or through the movement of a
parasitemic person to a nonendemic-disease area, as in imported
malaria. Although the incidence of these cases is low, they account
for most malaria transmission in industrialized countries.

Airport Malaria

Airport malaria is defined as malaria acquired through the bite of
an infected tropical anopheline mosquito by persons whose
geographic history excludes exposure to this vector in its natural
habitat (28). The vector is usually introduced into a nonendemic-
disease country on an international flight. For example, random
searches of airplanes at Gatwick Airport (London) found that 12 of
67 airplanes from tropical countries contained mosquitoes (29).
After a mosquito leaves the aircraft, it may survive long enough to
take a blood meal and transmit the disease, usually in the vicinity
of an airport. In temperate climates, temperature and humidity can
be favorable in the summer for the mosquito not only to survive
but also to move around and perhaps lay eggs. With the enormous
and continuing increase in air traffic, cases of airport malaria may
increase. Several such cases are described below. During a hot
summer in 1994, six cases of airport malaria were identified in and
around Roissy-Charles-de-Gaulle Airport (30). Four of the patients
were airport workers, and the others lived in Villeparisis,
approximately 7.5 km away. Anopheline mosquitoes were thought
to have traveled in the cars of airport workers who lived next door
to two of the patients. In 1989, two cases of P. falciparum malaria
were identified in Italy in two persons who lived in Geneva (31).
Another five cases of airport malaria were reported in Geneva in
the summer of 1989 (32). High minimum temperatures were
thought to have allowed the survival of infected anophelines
introduced by aircraft. In Britain, two cases of P. falciparum
malaria were observed in persons living 10 km and 15 km from
Gatwick Airport (33). Hot, humid weather in Britain may have
facilitated the survival of an imported mosquito.

Imported Malaria

Throughout the world, many countries are reporting an increasing
number of cases of imported malaria (Figure) because of the great
increase in long-distance travel in recent decades. For example,
cases imported from Africa to the United Kingdom rose from 803
in 1987 to 1,165 in 1993, and the ratio of all imported cases of
falciparum to vivax malaria rose from 0.76 in 1984 to 1.52 in 1993
(36).

Recently, a woman in Italy was infected with malaria through a
bite from a local species, An. labranchiae (37). This species was a
common malaria vector in Italy until the country was declared
malaria free in 1970. Local breeding sites, including isolated pools
in dried-up irrigation channels, were identified, and the mosquito
responsible is thought to have acquired the parasite after biting a
parasitemic girl who had acquired malaria in India. Airport malaria
was ruled out because of the distance from the nearest airport. This
may be the first case of malaria introduced to Europe in 20 years
and demonstrates the hazards of population movement (the parasite
had been introduced from India) combined with human activities
(providing vector breeding sites).

In the United States, recent outbreaks of presumed local mosquito-
borne transmission have been reported in California, with
migration from a disease-endemic area (38). The people involved
were migrant workers from malaria-endemic areas. An outbreak in
1986 involved 28 cases (26 in Mexican migrant workers) of P.
vivax during a 3-month period (39). The epidemic curve indicated
secondary spread, which confirmed local mosquito-borne
transmission. In the early 1990s, outbreaks were identified in
neighborhoods of Houston with many immigrants from countries
with malaria transmission. These outbreaks occurred when the
weather was hot and humid and thus conducive to the completion
of the sporogonic cycle and the survival of female anophelines
(38). Given that climate change could lead to more favorable
conditions for vector survival in Europe and the United States (40),
the increase in incidence in both airport and imported malaria is
cause for concern. If temperatures increase, uninfected introduced
or local mosquitoes could survive long enough after taking a blood
meal from a parasitemic person for the completion of the
sporogonic cycle of the parasite, thus enabling transmission.

Figure. Imported cases of malaria in Europe (34,35).

Conclusions

Population movements that either place people at risk for malaria
or cause them to pose a risk to others cannot be stopped. However,
prevention measures can address the causes of these movements
which, in developing countries, are often prompted by need rather
than choice; if living conditions and opportunities improved in
their place of residence, people would not be forced to move. In
cases where movements are unavoidable, people should be made
aware of the risks and have adequate access to treatment.

Epidemics are more likely to take place in areas where health care
may be poor or lacking. Areas at risk for epidemics through the
influx of infected people should be identified to avoid or control
epidemics. Particular attention should be paid to urban areas, given
the increasing number of cases of urban malaria and the ongoing
trend toward uncontrolled urbanization.

The risk for increased malaria transmission through economic
development of an area should be analyzed thoroughly before
development begins. For example, sound environmental
management can restrict vector breeding and reduce human-vector
contact. Such management could include proper maintenance of
irrigation systems and adequate water supply and sanitation
facilities for workers. Personal vector-control measures, such as
bed nets, and antimalarial drugs should be used. Regarding the
risks that airport and imported malaria pose for industrialized
countries, measures should be taken to ensure that cases of malaria
are promptly diagnosed and treated. Strengthened surveillance
would help prevent the reintroduction of malaria transmission by
local mosquitoes, which could acquire the infection by biting
persons with airport or imported malaria. Given the economic
resources and quality of health care in industrialized countries, the
likelihood is low that isolated outbreaks of malaria will lead to
reestablishment of transmission.

The relationship between malaria transmission and population
movement is complex, but future attempts to eradicate or control
malaria will be futile if they are not based on understanding of this
link. Because the current magnitude and diversity of population
movements are unprecedented, this issue is worthy of attention.

Acknowledgments

Comments on earlier versions of this paper from Bas Amelung,
Steve Lindsay, Saskia Nijhof, Mansell Prothero, Piebe de Vries,
and two anonymous referees are gratefully appreciated.

This work was supported by the United Nations Environment
Programme (Project Number FP/3210-96-01- 2207), the Dutch
National Research Program on Global Air Pollution and Climate
Change (Project Number 952257), and the Netherlands Foundation
for the Advancement of Tropical Research (Project Number WAA
93-312/313).

Dr. Martens is director of the Global Assessment Centre,
International Centre for Integrative Studies, Maastricht University.
His area of expertise includes assessment of the health impact of
globalization and global environmental changes. He is editor-in-
chief of the journal Global Change and Human Health.

Ms. Hall studied environmental health sciences at Maastricht
University and toxicology and pharmacology at the University of
London. The focus of her research is behavioral change in relation
to environmental impact.

Address for correspondence: Pim Martens, International Centre for
Integrative Studies, Maastricht University, P.O. Box 616, 6200
MD Maastricht, the Netherlands; fax: 31-43- 388-4916; e-mail:
p.martens@icis.unimaas.nl.

References
1. World Health Organization. World malaria situation in 1994.
Wkly Epidemiol Rec 1997;72:269-76.
2. Prothero RM. Disease and mobility: a neglected factor in
epidemiology. Int J Epidemiol 1977;6:259-67.
3. Bruce-Chwatt LJ. Movements of populations in relation to
communicable disease in Africa. East Afr Med J 1968;45:266-75.
4. Service MW. Agricultural development and arthropod-borne
diseases: a review. Rev Saude Publica 1991;25:165-8.
5. Rajagopalan PK, Jambulingam P, Sabesan S, Krishnamoorthy
K, Rajendran S, Gunasekaran K, et al. Population movement and
malaria persistence in Rameswaram island. Soc Sci Med
1986;22:879-86.
6. World Health Organization. The World Health Report 1996:
fighting disease, fostering development. Geneva: The
Organization; 1996.
7. Ramlogan R. Environmental refugees: a review. Environmental
Conservation 1996;23:81-8.
8. Kosinski LA, Prothero RM. The study of migration. In: Kosinski
LA, Prothero RM. People on the move: studies on internal
migration. London: Methuen; 1975. p. 1-38.
9. Gould WTS, Prothero RM. Space and time in African
population mobility. In: Kosinski LA, Prothero RM. People on the
move: studies on internal migration. London: Methuen; 1975. p.
39-49.
10. Lindsay SW, Martens P. Malaria in the African highlands: past,
present and future. Bull World Health Organ 1998;76:33-45.
11. United Nations. World urbanisation prospects: the 1996
revision. New York: United Nation; 1997.
12. Knudsen AB, Slooff R. Vector-borne disease problems in rapid
urbanisation: new approaches to vector control. Bull World Health
Organ 1992;70:1-6.
13. World Bank. Toward environmentally sustainable development
in sub-Saharan Africa. Washington: The Bank; 1996.
14. Sharma VP. Re-emergence of malaria in India. Indian J Med
Res 1996;103:26-45.
15. Molineaux L. The epidemiology of human malaria as an
explanation of its distribution, including some implications for its
control. In: Wernsdorfer WH, McGregor I, editors. Malaria:
principles and practice of malariology. Vol 2. Edinburgh: Churchill
Livingstone; 1988. p. 913-98.
16. Marques AC. Migrations and the dissemination of malaria in
Brazil. Mem Ins Oswaldo Cruz 1986;81 Suppl II:17-30.
17. Camargo LMA, Ferreira MU, Krieger H, De Camargo EP, Da
Silva LP. Unstable hypoendemic malaria in Rondônia (Western
Amazon region, Brazil): epidemic outbreaks and work-associated
incidence in an agro-industrial rural settlement. Am J Trop Med
Hyg 1994;51:16-25.
18. McGreevy PB, Dietze R, Prata A, Hembree SC. Effects of
immigration on the prevalence of malaria in rural areas of the
Amazon basin of Brazil. Instituto Oswaldo Cruz 1989;84:485-91.
19. Packard RM. Agricultural development, migrant labour and the
resurgence of malaria in Swaziland. Soc Sci Med 1986;22:861-7.
20. Sevilla-Casas E. Human mobility and malaria risk in the Naya
River basin in Colombia. Soc Sci Med 1993;37:1155-67.
21. U.S. Commission on Refugees. World refugee survey.
Washington: GPO; 1994.
22. Toole MJ, Waldman RJ. The public health aspects of complex
emergencies and refugee situations. Annu Rev Public Health
1997;18:283-312.
23. Centres for Disease Control and Prevention. Famine affected,
refugee and displaced populations: recommendations for public
health issues. MMWR Morb Mortal Wkly Rep 1992;41:RR-13.
24. Crowe S. Malaria outbreak hits refugees in Tanzania. Lancet
1997;350:41.
25. Bennett O, editor. Greenwar-environment and conflict.
London: The Panos Institute; 1991.
26. Kanji N, Harpham T. From chronic emergency to development:
an analysis of the health of the urban poor in Luanda, Angola. Int J
Health Serv 1992;22:349-63.
27. Onori E, Grab B. Indicators for the forecasting of malaria
epidemics. Bull World Health Organ 1980;58:91-8.
28. Isaäcson M. Airport malaria: a review. Bull World Health
Organ 1989;67:737-43.
29. Curtis CF, White GB. Plasmodium falciparum transmission in
England: entomological and epidemiological data relative to cases
in 1983. Journal of Tropical Medicine and Hygiene 1984;87:101-
14.
30. Giacomini T, Mouchet J, Mathieu P, Petithory JC. Study of 6
cases of malaria acquired near Roissy-Charles- de-Gaulle in 1994.
Bull Acad Natl Med 1995;179:335-51.
31. Majori G, Gradoni L, Gianzi FP, Carboni P, Cioppi A, Aureli
G. Two imported malaria cases from Switzerland. Tropical
Medicine and Parasitology 1990;41:439-40.
32. Bouvier M, Pittet D, Loutan L, Starobinski M. Airport malaria:
mini epidemic in Switzerland. Schweiz Med Wochenschr
1990;120:1217-22.
33. Whitfield D, Curtis CF, White GD, Targett GA, Warhurst DC,
Bradley DJ. Two cases of falciparum malaria acquired in Britain.
BMJ 1984;289:1607-9.
34. Bruce-Chwatt LJ, De Zulueta J. The rise and fall of malaria in
Europe. London: Butler and Tanner Ltd.; 1980.
35. Zahar AR. Vector bionomics in the epidemiology and control
of malaria. Part II: The WHO European and the WHO
Mediterranean region. Vol II. Applied field studies. Geneva: World
Health Organization; 1990.
36. Bradley D, Warhurst D, Blaze M, Smith V. Malaria imported
into the United Kingdom in 1992 and 1993. Communicable
Disease Report 1994;4:R169-71.
37. Baldari M, Tamburro A, Sabatinelli G, Romi R, Severini C,
Cuccagna G, et al. Malaria in Maremma, Italy. Lancet
1998;351:1246-7.
38. Zucker JR. Changing patterns of autochthonous malaria
transmission in the United States: a review of recent outbreaks.
Emerg Infect Dis 1996;2:39-43.
39. Maldonado YA, Nahlen BL, Roberto RR, Ginsberg M,
Orellana E, Mizrahi M, et al. Transmission of Plasmodium vivax
malaria in San Diego County, California, 1986. Am J Trop Med
Hyg 1990;42:3-9.
40. Martens P, Kovats RS, Nijhof S, De Vries P, Livermore MTJ,
Bradley DJ, et al. Climate change and future populations at risk of
malaria. Global Environmental Change 1999;S9:89-107.


Synopses

The bdr Gene Families of the Lyme Disease and Relapsing Fever
Spirochetes: Potential Influence on Biology, Pathogenesis, and
Evolution

David M. Roberts,* Jason A Carlyon,† Michael Theisen,‡ and
Richard T. Marconi*
*Medical College of Virginia at Virginia Commonwealth
University, School of Medicine, Richmond, Virginia, USA; †Yale
School of Medicine, Yale University, New Haven, Connecticut ,
USA; and ‡Statens Serum Institute, Copenhagen, Denmark

Species of the genus Borrelia cause human and animal infections,
including Lyme disease, relapsing fever, and epizootic bovine
abortion. The borrelial genome is unique among bacterial genomes
in that it is composed of a linear chromosome and a series of linear
and circular plasmids. The plasmids exhibit significant genetic
redundancy and carry 175 paralogous gene families, most of
unknown function. Homologous alleles on different plasmids could
influence the organization and evolution of the Borrelia genome by
serving as foci for interplasmid homologous recombination. The
plasmid-carried Borrelia direct repeat (bdr) gene family encodes
polymorphic, acidic proteins with putative phosphorylation sites
and transmembrane domains. These proteins may play regulatory
roles in Borrelia. We describe recent progress in the
characterization of the Borrelia bdr genes and discuss the possible
influence of this gene family on the biology, pathogenesis, and
evolution of the Borrelia genome.

Species of the genus Borrelia cause human and animal infections
(1). In North America, Lyme disease and endemic relapsing fever
pose the greatest threat to human health and have received the most
attention of the borrelial diseases. Approximately 14,000 cases of
Lyme disease are reported in the United States each year; however,
the actual number of cases may be 10-fold higher (2). Lyme
disease was not recognized as a distinct clinical entity in North
America until the 1970s (3). The causative agent, a previously
uncharacterized spirochete transmitted through the bite of infected
ticks of the Ixodes ricinus complex (I. scapularis in the Northeast
and Midwest and I. pacificus on the West Coast) (4,5), was
classified in the genus Borrelia and named B. burgdorferi. With the
emergence of Lyme disease and the identification of its etiologic
agent, Borrelia research focused on the development of reliable
Lyme disease diagnostic assays and vaccines, and the phenotypic
and genotypic diversity of Borrelia was thoroughly analyzed.
Through modern molecular taxonomic techniques, several newly
described species of Borrelia have emerged as possible causative
agents of Lyme disease or at least as agents genetically related to
B. burgdorferi (6-15).

The B. burgdorferi sensu lato complex is composed of the
following species: B. turdae, B. tanukii, B. bissettii, B. valaisiana,
B. lusitaniae, B. bissettii, B. andersonii, B. japonica, B. garinii, and
B. afzelii. Of these, B. burgdorferi, B. garinii, and B. afzelii are the
dominant species associated with infection in humans.

Relapsing fever has been studied not only for its impact on human
health but also as a model system for antigenic variation. There are
two general forms of relapsing fever, epidemic (louse
borne—Pediculus humanus) and endemic (tick
borne—Ornithodoros spp.) (1). Epidemic relapsing fever tends to
be associated with poor living conditions and social disruption
(famine and war) and is rare in the United States. Endemic
relapsing fever is more prevalent, predominantly in the western
regions. Three closely related Borrelia species, B. hermsii, B.
turicatae, and B. parkeri, are associated with this disease.

Hallmark features of relapsing fever include cyclic fever and
spirochetemia. The molecular basis for these features can be
attributed to the differential production of dominant variable
surface antigens of the Vmp protein families (16). The 40 or so
plasmid-carried vmp related genes in the B. hermsii genome are
expressed only one at a time. A single expression locus exists, and
genes not at this site lack a promoter element and are therefore not
transcribed (17). The expressed Vmp becomes a primary target of a
vigorous humoral immune response that kills most of the
spirochetal population. However, at a frequency of approximately
of 1 x 10 -3 to 1 x 10 -4 per generation, the identity of the
expressed Vmp changes (18) through gene conversion (19). The
net effect of this nonreciprocal event is to replace the gene located
in the expression locus with one that was previously silent.
Production of a new antigenically distinct Vmp allows evasion of
the humoral immune response. This ongoing change in Vmp
synthesis allows the relapsing fever spirochete population to
reestablish itself in the host, thus leading to spirochetemia and the
relapse of fever. Antigenic variation systems have also been
identified in the Lyme disease spirochetes; however, they appear to
exert a more subtle effect (20).

While clinical relapsing fever and Lyme disease differ from each
other in many ways, their causative agents share many similarities
at both the biologic and genetic levels. At the biologic level, they
are host associated and undergo similar environmental transitions
in the course of cycling between mammals and arthropods. In view
of the distinctly different characteristics of these environments, the
spirochetes must be able to adapt rapidly. Evidence suggests that
the relapsing fever and Lyme disease spirochetes use related
proteins to adapt to or carry out similar functions in changing
environments. For example, homologs of the plasmid-carried ospC
gene of the Lyme disease spirochetes are carried by several other
Borrelia species, including the relapsing fever spirochetes (21).
Both ospC and its relapsing fever spirochete homolog (vmp33) are
selectively expressed during the early stages of infection, which
suggests that they play a common functional role (22,23). The B.
burgdorferi Rep or Bdr protein family is also distributed
genuswide. Members of this polymorphic protein family possess
highly conserved putative functional motifs and structural
properties, which suggests that they may also carry out an
important genuswide role (24,25).

The Borrelia Genome

At the molecular level, a unique feature of Borrelia is the unusual
organization and structure of their genome. Unlike most bacteria,
which carry their genetic material in the form of a single, circular
DNA molecule, Borrelia have a segmented genome (26-28). Most
genetic elements carried by these bacteria are linear with covalently
closed termini or telomeres (27). The telomeres are characterized
by short hairpin loops of DNA (29). If heat denatured, these linear
molecules relax to form a single-stranded circular molecule. If
reannealed, they base-pair upon themselves to form a double-
stranded linear molecule that by physical necessity possesses a
short single-stranded hairpin loop at each telomere. Genetic
elements of this structure are rare in bacteria and are reminiscent of
certain viral genomes. In B. burgdorferi (isolate B31), the largest of
the linear genomic elements is the 911- kb chromosome (30). The
chromosome carries 853 putative ORFs, most of which are thought
to encode housekeeping functions. The remaining 12 linear and 8
circular genetic elements are plasmids. The plasmids might best be
thought of as mini-chromosomes, since as a group they are
indispensable in situ and may carry genes encoding proteins
involved in housekeeping functions (31). In addition, they may
further deviate from the true definition of a plasmid in that their
replication may not be independent and may instead be tightly
coordinated with the replication of the chromosome (32,33).
Nearly 50% of the plasmid-carried ORFs lack homology with
known sequences, which suggests that their encoded proteins may
define the unique biologic and pathogenetic aspects of Borrelia
(30). Several of the proteins derived from these plasmid-carried
genes of unknown function are antigenic or selectively expressed
during infection, which indicates that they function in the
mammalian environment (20,34-37). A striking feature of the
plasmid-carried ORFs is that they are organized into 175
paralogous gene families of two or more members (30). Hence, the
DNA content of the plasmids is highly redundant. Since the
maintenance of DNA is energetically expensive, it is likely that this
redundant DNA is of biologic importance to Borrelia. The
paralogous gene families of Borrelia have been the focus of
intensive research as they are thought to play important roles in
pathogenesis and to influence genome organization and evolution
(20,30,35,38-40).

Table 1. Borrelia species carrying bdr-related genes or expressing
proteins immunoreactive with anti-Bdr antisera

Identification of Borrelia Direct Repeat (bdr) Related Genes

The bdr gene family is a large, polymorphic, plasmid-carried,
paralogous gene family of unknown function that was originally
identified in B. burgdorferi (41,42). Members of this gene family
have been characterized in several Borrelia species and isolates
(Table 1) and have been assigned various gene names (25,41-44)
(Table 2). We have adopted the bdr designation in the context of a
nomenclature system (25), summarized below. Genes belonging to
the bdr gene family were first identified through the analysis of
repeated DNA sequences in B. burgdorferi sensu lato complex
isolates (41,42). Seven nonidentical but closely related copies of a
plasmid-carried repeated element were identified in B. burgdorferi
297 (42). Three additional copies of this repeated sequence were
further identified in B. burgdorferi 297 (45). These loci carry
several ORFs that were designated as rep+, rep-, LPA, LPB (the LP
genes have recently been redesignated as mlp for multicopy
lipoprotein [45]), rev, and the orfABCD operon (note: ORFs A and
B have been redesignated as blyA and blyB). Some of these genes,
particularly rep and mlp, exhibit allelic variation and encode
polymorphic proteins, the functions of which are under
investigation. Focusing specifically on the rep or bdr genes, the rep
designation was originally.

Table 2. Borrelia Bdr homology groups and gene nomenclature

Figure 1. Restriction fragment length polymorphism pattern
analysis of the rep or bdr genes of the Lyme disease spirochetes.
Total DNA, isolated from Borrelia cultures, was digested with
Xba1, fractionated by electrophoresis, and transferred onto
membranes for hybridization. Hybridization was performed by the
bdrAB-R1 oligonucleotide (46). The species and isolates analyzed
are indicated above each lane. MW markers in kb are indicated.
chosen to reflect a central repeat motif carrying domains in the
deduced amino acid sequences. The + and - designations were
assigned to indicate that the overlapping rep+ and rep- genes are
located on opposing DNA strands. Plasmid-carried repeated DNA
sequences were also identified in B. burgdorferi B31 and found to
carry either all or a subset of seven ORFs, designated A through G
(41). Of relevance to this discussion are the ORF-E sequences that
are rep or bdr homologs. A bdr-related gene was also identified in
B. afzelii DK1 and designated as p21 (43). B. afzelii causes Lyme
disease in Europe and Asia. The rep+, ORF-E, and p21
designations have recently been replaced with bdr gene
designations (24,25,44).

To assess and compare the composition and complexity of the bdr
gene family among species and isolates of the B. burgdorferi sensu
lato complex, restriction fragment length polymorphism (RFLP)
patterns were determined (Appendix). Genomic DNA digested
with Xba1 was Southern blotted and probed with an
oligonucleotide targeting the bdr genes (Figure 1). A variable
number of hybridizing bands of different size were detected. These
analyses demonstrate that extensive bdr gene families are carried
by B. burgdorferi sensu lato complex isolates and that the RFLP
patterns vary at the inter- and intraspecies level. Hybridization
analyses of other Borrelia species showed that they also carry bdr-
related gene families (24,25,46). bdr-related genes have been
detected by hybridization in B. turicatae, B. hermsii, B. parkeri, B.
coriaceae, and B. anserina (25,46). Isolates of these species also
exhibit substantial variation in their bdr RFLP patterns at the
intraspecies level. Table 1 lists the Borrelia species that carry bdr-
related genes and indicates the methods by which these genes or
proteins were detected.

Sequences flanking some bdr alleles also appear to be distributed
genus wide. Some bdr alleles of B. turicatae, B. parkeri, and B.
hermsii are flanked by genes that are homologs of genes carried by
the Lyme disease spirochetes (24,25). As a specific example, the B.
turicatae bdrA 1 gene is flanked by ORFs that are homologs of the
BBG34 and BBG30 genes of B. burgdorferi (24,25). In the Lyme
disease spirochetes, BBG34 is part of a three-member paralogous
gene family, while BBG30 is a single-copy gene (30). Located
between BBG30 and BBG34 is BBG33, a member of the bdr gene
family (recently redesignated as bdrF 2 ) (25). Although these
divergent Borrelia species carry related genes, their organization
differs (24), which indicates that rearrangement has taken place in
the ancestral plasmid that carried these homologs. Figure 2
compares the organization of two bdr loci from B. turicatae and B.
burgdorferi.

Figure 2. General organization of two bdr loci in Borrelia turicatae
and B. burgdorferi. The gene arrangement depicted for B. turicatae
was determined through cloning and sequence analysis of a 2,217
base-pair XbaI restriction fragment. The arrangement for the bdr-
carrying locus of B. burgdorferi was previously determined through
the sequencing of the B. burgdorferi B31 genome (30). The arrows
indicate the direction of transcription. Genes exhibiting homology
are indicated by similar shading or hatch marks. Genes indicated
by unfilled arrows are not homologous. The numbering is indicated
for scale and is not indicative of the positioning of these genes on
the plasmids that carry them.

Evolutionary Analyses of bdr-Related Sequences: Revised
Nomenclature for the Bdr-Related Proteins

To simplify the complicated nomenclature of bdr-related genes, a
bdr nomenclature system has been developed that assigns gene
names on the basis of phylogenetic relationships inferred from
comparative analysis of genetically stable regions of the bdr genes
(25). This system, which is applicable genuswide, allows for a
ready assessment of relationships among bdr paralogs and
orthologs. The rationale for this system stemmed from the results
of a comprehensive evolutionary analysis of >50 bdr-related
sequences from five Borrelia species that demonstrated that bdr
sequences are organized into six distinct subfamilies, designated A
through F (25). Subfamilies are not necessarily species specific;
some contain bdr alleles from different Borrelia species (25). Since
members of a given subfamily are closely related to one another
with identity values for the N terminal domain being >95%, each
member is assigned the same gene name designation, and paralogs
are distinguished by a numerical subscript. In B. turicatae OZ-1,
two bdr subfamilies, bdrA and bdrB, contain at least four and five
members, respectively (24). Members of the bdrA subfamily are
designated bdrA 1 , bdrA 2 , bdrA 3, and bdrA 4, while members
of the bdrB family are designated bdrB 1 through bdrB 5 . This
revised Bdr nomenclature scheme was modeled after that proposed
for bacterial polysaccharide synthesis genes (47) and is in
accordance with the nomenclature guidelines established by
Demerec (48).

The subfamily affiliation of bdr genes can be readily determined
through comparative sequence analyses of the amino acid segment
preceding the polymorphic repeat motif region of these proteins
(described in detail below) (25). Relationship assessments based on
the genetically stable N terminal domain (vs. complete sequences)
are preferable because the calculated evolutionary distances and
clustering relationships are not artificially skewed by the variable
number of repeat motifs present in the repeat motif domain. Since
the genetically unstable repeat motif domain comprises as much as
50% of the total coding sequence in some alleles, it can have a
substatial impact on inferred relationships. In addition, extensive
sequence variation in the carboxyl termini of the Bdr proteins at
the interspecies level makes it difficult to align this domain with
confidence, which further influences the inferred relationships.

Figure 3. Key features and putative functional domains of the Bdr
proteins. The schematic depicts a prototype Bdr protein with the
characteristics of each domain indicated. The abbreviation, ID%, is
for percentage amino acid identity at either the inter- or intra-
family level as indicated in the figure. Standard amino acid
abbreviations are used in the figure to denote the conserved C-
terminal lysine (K) or asparagine (N) residues, which are thought
to be exposed in the periplasm and the cytoplasmically located core
tripeptide of the repeat motif (lysine-isoleucine-aspartic acid; KID).


bdr evolutionary analyses show that Borrelia species carry
members of at least two bdr subfamilies (25,44). In fact, B.
burgdorferi carries three distinct subfamilies. Multiple Bdr
subfamilies in diverse Borrelia species suggest that there has been
selective pressure to maintain multiple bdr alleles and bdr genetic
diversity. This genetic diversity may increase the functional
diversity of the Bdr proteins.

Molecular Features and Physical Properties of the Bdr Proteins

While early analyses of Borrelia bdr genes demonstrated their
multicopy nature (41,42,46), the full extent of the complexity of
the bdr gene family in the Lyme disease spirochetes was not fully
recognized until the B. burgdorferi genome sequence was
determined (30). B. burgdorferi B31 was found to carry 17 distinct
bdr-related genes (and one truncated variant) distributed among
different linear and circular plasmids. B. turicatae, which carries at
least nine different bdr alleles, carries these genes exclusively on
linear plasmids (24,25,46). Other relapsing fever spirochete species
(B. parkeri and B. hermsii) are similar to the Lyme disease bacteria
in that they carry bdr genes on both linear and circular plasmids
(25). In the Lyme disease spirochetes each of the 32-kb circular
plasmids, with the exception of plasmids M and P, carry two
different bdr genes separated by seven or eight ORFs. Each of
these circular plasmids carries one bdrD subfamily member and
one bdrE subfamily member. The maintenance of genes belonging
to different subfamilies on a single plasmid is consistent with the
possibility that each carries out a different function. In contrast, in
the Lyme disease spirochetes, the bdrF subfamily members are
localized to linear plasmids with only a single bdr gene per
plasmid. These observations suggest that there has been selective
pressure to maintain the association of specific subfamilies with
specific types of plasmids. Less is known about the bdr-carrying
plasmids and the organization of the bdr genes and subfamilies in
the relapsing fever borreliae. However, as in the Lyme disease
spirochetes, in B. turicatae most bdr-carrying plasmids carry two
bdr genes, one from subfamily bdrA and one from subfamily bdrB
(24). The sequence of more than 50 bdr alleles from five different
Borrelia species has been determined (Table 2) (24,25,41-43,46).
These extensive comparative sequence analyses led to the
identification of conserved features that provide insight into the
possible biologic roles of the Bdr proteins. For example, all bdr
alleles carry centrally located repeat motif domains (Figure 3).
Although conserved in sequence, these domains vary in length
among alleles as a result of varying numbers of the repeat motif.
The core tripeptide the repeat is the sequence KID. The repeat
motifs encode consensus casein kinase 2 phosphorylation (CK2P)
motifs of the sequence T/SKID/E (43). While it may appear
somewhat paradoxical for bacteria to carry casein kinases, casein
kinase is a descriptive term broadly applied to at least two classes
of ubiquitous protein kinases for which the substrates may include
various enzymes and noncatalytic proteins involved in important
cellular regulatory functions (49). Most proteins phosphorylated by
CK2-like kinases are highly acidic, as are the Borrelia Bdr proteins
(isoelectric points between 5 and 6). The phosphorylation site in
CK2P motifs is either the Ser or Thr residue of the motif. Although
histidine kinases have been known to exist in some bacteria, it has
been widely held that bacteria lack Ser - Thr kinases. However, Ser
-Thr kinases have recently been identified in several bacterial
species, including Myxococcus, Anabeana, Freymella, Yersinia,
and Streptomyces (50). Most importantly, analysis of the B.
burgdorferi genome sequence identified a putative Ser - Thr kinase
designated BB0648 (30,50). This ORF carries a domain that
exhibits homology with the active site of Ser - Thr kinases. B.
burgdorferi also carries a homolog of the PPM family of eucaryotic
protein Ser - Thr phosphatases (30,50). The presence of these
genes in B. burgdorferi suggests that the Borrelia possess the
machinery necessary for Ser - Thr phosphorylation and
dephosphorylation. Another important conserved feature identified
through sequence analyses is the hydrophobic carboxyl terminal
domain of approximately 20 amino acids. Computer analyses
conducted with the TMpred program indicate that this domain has
a high propensity to form a transmembrane helix (24,25). The
Tmpred values for the 20 aa C-terminal domains are 2,000 to
2,600. A value of 500 or greater is considered significant (24,25).
Comparison of the Bdr putative transmembrane domain sequences
from the Lyme disease spirochetes with those from the relapsing
fever spirochetes indicates that, while there is conservation in
physical properties, there is essentially no conservation of primary
sequence. However, sequence conservation does exist at the
subfamily level (24,25). Since the Bdr proteins lack an obvious
export signal, membrane association would most likely be with the
spirochetal inner membrane, with the rest of the protein, which is
hydrophilic, extending into the cytoplasm. The terminal residue of
the protein is in almost all cases a positively charged amino acid
(lysine or asparagine). This residue could extend into the periplasm
and serve to anchor the Bdr proteins to other cellular components,
such as the peptidoglycan.

Immunologic Analyses of the Bdr Proteins

The presence of multiple bdr alleles and bdr subfamilies within
isogeneic populations has prompted speculation that there may be
differential expression at either the subfamily or individual allele
level, possibly in response to environmental stimuli (46). Limited
studies of bdr expression and production, based on either mRNA
detection or immunoblot analyses, have been performed. Porcella
et al. (42) used Northern hybridization to determine if expression
of B. burgdorferi bdr-related genes occurs during cultivation in the
laboratory under standard culture conditions (33°C in BSK media).
Bdr transcripts were not detected by this approach. Similarly, in an
earlier analysis, we also conducted Northern hybridization
experiments to assess bdr expression (46). We detected expression
of B. turicatae OZ1 bdrA subfamily members in bacteria cultivated
under standard laboratory growth conditions (46). However, when
reverse transcriptase (RT)-PCR methods were applied,
transcription of a single bdrA allele was detected (46). B. turicatae
OZ-1 was later demonstrated to carry at least nine bdr alleles, four
of which belong to the bdrA subfamily. Analysis of the sequence of
these alleles showed that all four should have been readily
amplified by the RT-PCR primer set because of the conservation of
the primer binding sites (24). The lack of detection of transcript
derived from these alleles suggested that only a subset of the bdr A
subfamily alleles is expressed. This raised the possibility that other
bdr alleles are either nonfunctional genes or their expression
requires different environmental stimuli. The transcriptional
expression of the bdrB subfamily has not been specifically
assessed. Thorough transcriptional analyses using allele-specific
probes and primers are an important step, since they allow specific
assessment of the expression of individual bdr alleles under
differing environmental conditions. In addition, analyses of the
upstream DNA sequences of individual bdr alleles and their
genomic location may elucidate the molecular basis for bdr
transcriptional regulation.

Figure 4. Immunoblot analyses demonstrating the variation in Bdr
protein expression in Borrelia species and isolates. Bacteria were
cultivated and prepared for analysis as described in the methods.
Proteins were fractionated by SDS-PAGE, immunoblotted and
screened with anti-BdrF 1-B.afzelii DK1 antisera. The species and
isolates analyzed are indicated above each lane in panels A, B and
C. The migration positions of the protein standards are indicated in
each panel.  Immunologic analyses have provided a somewhat
different overall picture regarding Bdr production. Immunologic
analyses described in this report and elsewhere (44) demonstrate
that several members of the bdr gene family are expressed during
in vitro cultivation. We conducted a comprehensive analysis of the
expression of Bdr proteins among Borrelia species. When antisera
raised against recombinant B. afzelii BdrF 1 (24) were used in
immunoblot analyses, several immunoreactive proteins were
detected in cell lysates of all Borrelia species tested (Figure 4). The
only exception was B. anserina, a causative agent of avian
spirochetosis. Although bdr-related sequences have been detected
in B. anserina by hybridization techniques (46), immunoreactive
proteins were not detected in immunoblot analyses. Additional
analyses are required to determine if this indicates absence of
translational expression or the lack of epitope conservation in this
species. In any event, the fact that immunoreactive bands were not
detected in this species attests to the specificity of the anti-Bdr
antisera. As a further demonstration of the specificity of the
antisera and to highlight the fact that the Bdr proteins are unique to
Borrelia, a cell lysate of Leptospira interrogans was included in the
immunoblot analyses. Immunoreactivity with proteins in the Bdr
size range was not observed with the anti-Bdr antisera in this
spirochete species. Borrelia species that expressed immunoreactive
proteins included B. garinii, B. burgdorferi, B. turdae, B. tanukii,
B. japonica, B. valaisiana, B. afzelii, B. coriaceae, B. bissettii, B.
miyamotoi, B. parkeri, B. hermsii, and B. turicatae (Table 1).
Particularly striking was the extensive variation in the number and
molecular weight of the immunoreactive proteins expressed, with
up to 12 distinct Bdr proteins detected. Variation in expression
patterns was observed at both the inter- and intraspecies level.
Analysis of three B. burgdorferi isolates (B31G, cN40, and CA12)
demonstrated variability in both the size and number of expressed
Bdr proteins. Isolate B31G has been demonstrated by genomic
sequencing to carry 18 distinct bdr alleles. Immunoblot analyses
show that not all alleles are expressed during in vitro cultivation;
therefore, some alleles may be differentially regulated.

The broad immunoreactivity of the antisera with diverse Borrelia
species indicates that some epitopes are conserved genuswide. In
view of the sequence divergence in the N and C terminal domains
of the Bdr proteins derived from different subfamilies, it is likely
that the cross-reactive epitopes reside in the conserved repeat motif
region. Consistent with this, computer analyses of the repeat
domain of all determined Bdr protein sequences predict them to be
alpha helical and to have a surface exposed on the protein and a
positive Jameson-Wolf antigenic index (24,25,44). The
conservation and synthesis of these polymorphic proteins in such a
diverse group of Borrelia species suggest that they play an
important role in Borrelia biology genuswide.

The Bdr Proteins and Borrelia Biology: An Overview

Bdr genes and extensive bdr gene families have now been
identified and characterized in several diverse Borrelia species
(24,25,42-44,46). Comparative sequence analyses, which have
identified conserved putative functional domains, have provided
the basis for the development of hypotheses regarding Bdr function
and cellular location. The Bdr proteins, which lack known
consensus export signals, are likely anchored to the cytoplasmic
membrane through their conserved, hydrophobic, putative
transmembrane spanning domain. The C-terminal positively
charged amino acid may be exposed to the periplasm, where it may
interact with other cellular components that may include the
peptidoglycan. The repeat motif domain, which is predicted by
computer analyses to be hydrophilic and surface exposed on the
protein, likely extends into the cytoplasm. The conserved repeat
motif domain that carries the putative Ser - Thr phosphorylation
motifs may then be accessible for phosphorylation or to interact
with other cytoplasmic proteins or DNA to form a membrane
anchored complex. As with numerous other proteins,
phosphorylation and dephosphorylation could play a regulatory
role, perhaps in signaling or sensing.

Multiple polymorphic bdr alleles may increase the functional range
and diversity of the Bdr proteins. Functional partitioning among
Bdr proteins could offer a possible explanation of why Borrelia
expend such biologic energy to maintain these genes in large gene
families and express variants of these proteins. The homology
among bdr alleles may also allow or lead to the continual
modification of these genes through homologous recombination. In
fact, the variable nature of the repeat motif region, which is clearly
not evolutionarily stable, has likely arisen from slipped-strand
mispairing, recombination, or rearrangement. In view of the
extensive genetic redundancy of the plasmid component of the
Borrelia genome, recombination in and among related sequences
on different plasmids could affect the organization and evolution of
the genome and ultimately host-pathogen interaction. Inter- or
intra-plasmid exchange of DNA sequences could provide a
mechanistic basis for the extensive genetic variability that has been
widely described for Borrelia plasmids (28,29,51- 59). In spite of
the apparent necessity for at least most of the plasmids for survival,
as inferred from their ubiquitous distribution among Borrelia
isolates, these bacteria are able to tolerate remarkable genomic
variability. Diversity in the plasmids and the genes they carry may
actually be exploited as a tool for phenotypic diversity and rapid
environmental adaptation.

Acknowledgments

We thank Michael Norgard, Steve Porcella, Justin Radolf, and the
Molecular Pathogenesis research group at Virginia Commonwealth
University for helpful discussions.

This work was supported in part by a grant from the Jeffress
Memorial Trust.

Mr. Roberts is a 3rd-year Ph.D. student, Department of
Microbiology and Immunology, Medical College of Virginia at
Virginia Commonwealth University. His scientific interests are
centered on the study of the molecular mechanisms of pathogenesis
and adaptive responses of arthropod-carried spirochetes,
specifically on the role of plasmid-carried gene families in Borrelia
biology and pathogenesis.


Address for correspondence: Richard T. Marconi, Department of
Microbiology and Immunology, Medical College of Virginia at
VCU, Richmond, VA 23298-0678, USA; fax: 804-828-9946; e-
mail: Rmarconi@hsc.vcu.edu.

References

1. Barbour AG, Hayes SF. Biology of Borrelia species.
Microbiological Reviews 1986;50:381-400.
2. Young JD. Underreporting of Lyme disease. N Engl J Med
1998;338:1629.
3. Steere AC, Malawista SE, Snydman DR, Shope RE, Andiman
WA, Ross MR, et al. Lyme arthritis: an epidemic of oligoarticular
arthritis in children and adults in three Connecticut communities.
Arthritis Rheum 1977;20:7-17.
4. Burgdorfer W, Barbour AG, Hayes SF, Benach JL, Grunwaldt E,
Davis JP. Lyme disease—a tick-borne spirochetosis? Science
1982;216:1317-9.
5. Benach JL, Bosler EM, Hanrahan JP, Coleman JL, Bast TF,
Habicht GS, et al. Spirochetes isolated from the blood of two
patients with Lyme disease. N Engl J Med 1983;308:740-2.
6. Baranton G, Postic D, Saint Girons I, Boerlin P, Piffaretti J-C,
Assous M, et al. Delineation of Borrelia burgdorferi sensu stricto,
Borrelia garinii sp. nov., and group VS461 associated with Lyme
borreliosis. Int J Syst Bacteriol 1992;42:378-83.
7. Marconi RT, Garon CF. Identification of a third genomic group
of Borrelia burgdorferi through signature nucleotide analysis and
16S rRNA sequence determination. Journal of General
Microbiology 1992;138:533-6.
8. Marconi RT, Garon CF. Phylogenetic analysis of the genus
Borrelia: a comparison of North American and European isolates
of B. burgdorferi. J Bacteriol 1992;174:241-4.
9. Marconi RT, Liveris D, Schwartz I. Identification of novel
insertion elements, restriction fragment length polymorphism
patterns, and discontinuous 23S rRNA in Lyme Disease
spirochetes: phylogenetic analyses of rRNA genes and their
intergenic spacers in Borrelia japonica sp. nov. and genomic group
21038 (Borrelia andersonii sp. nov.) isolates. J Clin Microbiol
1995;33:2427-34.
10. Fukunaga M, Hamase A, Okada K, Inoue H, Tsuruta Y,
Miyamoto K, et al. Characterization of spirochetes isolated from
ticks (Ixodes tanukin, I. turdus and Ixodes columnae) and
comparision of the sequences with those of Borrelia burgdorferi
sensu lato strains. Appl Environ Microbiol 1996;62:2338-44.
11. Fukunaga M, Hamase A, Okada K, Nakao M. Borrelia tanuki
sp. nov. and Borrelia turdae sp. nov. found from Ixodid ticks in
Japan: rapid species identification by 16S rRNA gene—targeted
PCR analysis. Microbiol Immunol 1996;40:877-81.
12. Fukunaga M, Takahashi Y, Tsuruta Y, Matsushita O, Ralph D,
McClelland M, et al. Genetic and phenotypic analysis of Borrelia
miyamotoi sp. nov., isolates from the Ixodid tick Ixodes
persulcatus, the vector for Lyme disease in Japan. Int J Syst
Bacteriol 1995;45:804-10.
13. Wang G, van Dam AP, Le Flecha A, Postic D, Peter O,
Baranton G, et al. Genetic and phenotypic analysis of Borrelia
valaisiana sp. nov. (Borrelia genomic groups VS116 and M19). Int
J Syst Bacteriol 1997;47:926-32.
14. Postic D, Belfazia J, Isogai E, Saint Girons I, Grimont PAD,
Baranton G. A new genomic species in Borrelia burgdorferi sensu
lato isolated from Japanese ticks. Res Microbiol 1993;144:467-73.
15. Postic D, Edlinger C, Richaud C, Grimont F, Dufresne Y,
Perolat P, et al. Two genomic species in Borrelia burgdorferi. Res
Microbiol 1990;141:465-75.
16. Barbour AG. Antigenic variation of a relapsing fever Borrelia
species. Annu Rev Microbiol 1990;44:155-71. 17. Barbour AG,
Byrman N, Carter CJ, Kitten T, Berstrom S. Variable antigen genes
of the replasing fever agent Borrelia hermsii are activated by
promoter addition. Mol Microbiol 1991;5:489-93.
18. Stoenner HG, Dodd T, Larsen C. Antigenic variation of B.
hermsii. J Exp Med 1982;156:1297-311.
19. Plasterk RHA, Simon MI, Barbour AG. Transposition of
structural genes to an expression sequence on a linear plasmid
causes antigenic variation in the bacterium B. hermsii. Nature
1985;318:257-63.
20. Zhang J-R, Hardham JM, Barbour AG, Norris AG. Antigenic
variation in Lyme disease Borreliae by promiscuous recombination
of vmp like sequence cassettes. Cell 1997;89:275-85.
21. Marconi RT, Samuels DS, Schwan TG, Garon CF.
Identification of a protein in several Borrelia species which is
related to OspC of the Lyme disease spirochetes. J Clin Microbiol
1993;31:2577-83.
22. Schwan TG, Piesman J, Golde WT, Dolan MC, Rosa PA.
Induction of an outer surface protein on Borrelia burgdorferi during
tick feeding. Proc Natl Acad Sci U S A 1995;92:2909-13.
23. Schwan TG, Hinnebusch BJ. Bloodstream- versus tick-
associated variants of a relapsing fever bacterium. Science
1998;280:1938-40.
24. Carlyon JA, Roberts DM, Theisen M, Marconi RT. Molecular
analyses of the B. turicatae bdr genes: a polymorphic, linear
plasmid carried, paralogous gene family. In press 1999.
25. Carlyon JA, Roberts DM, Marconi RT. Evolutionary and
molecular analyses of the Borrelia rep super gene family:
delineation of six distinct sub-families and demonstration of the
genus wide conservation of putative functional domains, structural
properties and repeat motifs. Microb Pathog. In press 2000.
26. Ferdows MS, Barbour AG. Megabase-sized linear DNA in the
bacterium Borrelia burgdorferi, the Lyme disease agent. Proc Natl
Acad Sci U S A 1989;86:5969-73.
27. Barbour AG, Garon CF. Linear plasmids of the bacterium
Borrelia burgdorferi have covalently closed ends. Science
1987;237:409-11.
28. Barbour AG. Plasmid analysis of Borrelia burgdorferi, the
Lyme disease agent. J Clin Microbiol 1988;26:475-8.
29. Hinnebusch J, Barbour AG. Linear plasmids of Borrelia
burgdorferi have a telomeric structure and sequence similar to
those of a eukaryotic virus. J Bacteriol 1991;173:7233-9.
30. Fraser C, Casjens S, Huang WM, Sutton GG, Clayton R,
Lathigra R, et al. Genomic sequence of a Lyme disease spirochaete,
Borrelia burgdorferi. Nature 1997;390:580-6.
31. Barbour AG. Linear DNA of Borrelia species and antigenic
variation. Trends Microbiol 1993;1.
32. Hinnebusch J, Tilly K. Linear plasmids and chromosomes in
bacteria. Mol Microbiol 1993;10:917-22.
33. Hinnebusch J, Barbour AG. Linear- and circular-plasmid copy
numbers in Borrelia burgdorferi. J Bacteriol 1992;174:5251-7.
34. Champion CI, Blanco DR, Skare JT, Haake DA, Giladi M,
Foley D, et al. A 9.0 kilobase-pair circular plasmid of Borrelia
burgdorferi encodes an exported protein: evidence for expression
only during infection. Infect Immun 1994;62:2653-61.
35. Akins D, Porcella SF, Popova TG, Shevchenko D, Baker SI, Li
M, et al. Evidence for in vivo but not in vitro expression of a
Borrelia burgdorferi outer surface protein F (OspF) homologue.
Mol Microbiol 1995;18:507-20.
36. Suk K, Das S, Sun W, Jwang B, Barthold SW, Flavell RA, et
al. Borrelia burgdorferi genes selectively expressed in the infected
host. Proc Natl Acad Sci U S A 1995;92:4269-73.
37. Skare JT, Foley DM, Hernandez SR, Moore DC, Blanco DR,
Miller JN, et al. Cloning and molecular characterization of
plasmid-encoded antigens of Borrelia burgdorferi. Infect Immun
1999;67:4407-17.
38. Marconi RT, Sung SY, Hughes CN, Carlyon JA. Molecular
and evolutionary analyses of a variable series of genes in Borrelia
burgdorferi that are related to ospE and ospF, constitute a gene
family, and share a common upstream homology box. J Bacteriol
1996;178:5615-26.
39. Sung S-Y, LaVoie C, Carlyon JA, Marconi RT. Evolutionary
instability of ospE related members of the UHB gene family in
Borrelia burgdorferi sensu lato complex isolates. Infect Immun
1998;66:4656-68.
40. Zhang JR, Norris SJ. Genetic variation of the Borrelia
burgdorferi gene vlsE involves cassette-specific, segmental gene
conversion. Infect Immun 1998;66:3698-704.
41. Zuckert WR, Meyer J. Circular and linear plasmids of Lyme
Disease spirochetes have extensive homology: characterization of a
repeated DNA element. J Bacteriol 1996;178:2287-98.
42. Porcella SF, Popova TG, Akins DR, Li M, Radolf JR, Norgard
MV. Borrelia burgdorferi supercoiled plasmids encode multicopy
open reading frames and a lipoprotein gene family. J Bacteriol
1996;178:3293-307.
43. Theisen M. Molecular cloning and characterization of nlpH,
encoding a novel surface exposed, polymorphic, plasmid-encoded
33kD lipoprotein of Borrelia afzelii. J Bacteriol 1996;178:6435-42.
44. Zuckert WR, Meyer J, Barbour AG. Comparative analysis and
immunological characterization of the Borrelia Bdr protein family.
Infect Immun 1999;67:3257-66.
45. Yang X, Popova TG, Hagman KE, Wikel SK, Schoeler GB,
Caimano MJ, et al. Identification, characterization, and expression
of three new members of the Borrelia burgdorferi Mlp (2.9)
lipoprotein gene family. Infect Immun 1999;67:6008-18.
46. Carlyon JA, Marconi RT. Cloning and molecular
characterization of a multi-copy, linear plasmid-carried, repeat
motif-containing gene from Borrelia turicatae, a causative agent of
relapsing fever. J Bacteriol 1998;180:4974-81.
47. Reeves PR, Hobbs M, Valvano MA, Kido N, Klena J, Maskell
D, et al. Bacterial polysaccharide synthesis and gene nomenclature.
Trends Microbiol 1996;4:498-503.
48. Demerec M. A proposal for a uniform nomenclature in
bacterial genetics. Genetics 1966;54:61-76.
49. Pinna LA. Casein kinase 2: an "eminence grise" in cellular
regulation. Biochim Biophys Acta
1990;1054:267-84.
50. Shi L, Potts M, Kennelly PJ. The serine, threonine and/ or
tyrosine protein kinases and protein phosphatases of procaryotic
organisms: a family portrait. FEMS Microbiol Rev 1998;22:229-
53.
51. Xu Y, Kodner C, Coleman L, Johnson RC. Correlation of
plasmids with infectivity of Borrelia burgdorferi sensu stricto type
strain B31. Infect Immun 1996;64:3870-6.
52. Xu Y, Johnson RC. Analysis and comparision of plasmid
profiles of Borrelia burgdorferi sensu lato strains. J Clin Microbiol
1995;33:2679-85.
53. Stålhammar-Carlemalm M, Jenny E, Gern L, Aeschlimann A,
Meyer J. Plasmid analysis and restriction fragment length
polymorphisms of chromosomal DNA allow a distinction between
Borrelia burgdorferi strains. Zentralbl Bakteriol 1990;274:28-39.
54. Simpson WJ, Garon CF, Schwan TG. Analysis of supercoiled
circular plasmids in infectious and non-infectious Borrelia
burgdorferi. Microb Pathog 1990;8:109-18.
55. Schwan TG, Schrumpf ME, Karstens RH, Clover JR, Wong J,
Daugherty M, et al. Distribution and molecular analysis of Lyme
disease spirochetes, Borrelia burgdorferi, isolated from ticks
throughout California. J Clin Microbiol 1993;31:3096-108.
56. Marconi RT, Casjens S, Munderloh UG, Samuels DS. Analysis
of linear plasmid dimers in Borrelia burgdorferi sensu lato isolates:
implications concerning the potential mechanism of linear plasmid
replication. J Bacteriol 1996;178:3357-61.
57. Marconi RT, Samuels DS, Landry RK, Garon CF. Analysis of
the distribution and molecular heterogeneity of the ospD gene
among the Lyme disease spirochetes: evidence for lateral gene
exchange. J Bacteriol 1994;176:4572-82.
58. Ferdows MS, Serwer P, Griess GA, Norris SJ, Barbour AG.
Conversion of a linear to a circular plasmid in the relapsing fever
agent Borrelia hermsii. J Bacteriol 1996;178:793-800.
59. Carlyon JA, LaVoie C, Sung SY, Marconi RT. Analysis of the
organization of multi-copy linear and circular plasmid carried open
reading frames in Borrelia burgdorferi sensu lato isolates. Infect
Immun 1998;66:1149-58.

Appendix

Bacterial Cultivation, DNA Isolation, and Southern Hybridization
Analyses

Isolates belonging to the Borrelia burgdorferi sensu lato complex
were cultivated in complete BSK-H media (Sigma) at 33°C. To
cultivate the relapsing fever borreliae and other Borrelia species,
the complete BSK-H media were supplemented with additional
rabbit sera (Sigma) to a final concentration of 12% (vol/vol).
Bacteria were harvested by centrifugation and washed with
phosphate buffered saline (pH 7.0), and DNA was extracted (25).
For Southern hybridization analyses, 5 ˜g of DNA from each
isolate was digested under standard conditions with Xba1 and
fractionated by electrophoresis in 0.8% GTG agarose gels. (The
DNA was transferred onto membranes for hybridization by vacuum
blotting using the VacuGene system as described by the
manufacturer (Pharmacia). All other Southern hybridization
methods were as previously described (39)

Immunoblot Analyses

Bacterial cultures were grown and harvested as described above.
One OD600 equivalent of cells was pelleted and resuspended in
100 ˜l of standard SDS-sample buffer with reducing agents. The
cell lysates (7 ˜l) were fractionated by electrophoresis in 15% SDS-
PAGE gels and electroblotted onto Immobilon P membranes (38).
The immunoblots were blocked overnight in blocking buffer (1X
PBS, 0.2% Tween, 0.002% NaCl, and 5% nonfat dry milk) and
then incubated with a 1:1,000 antisera dilutions. ImmunoPure Goat
antimouse IgG (H+L) peroxidase conjugate served as the
secondary antibody. The secondary antibody was incubated with
the blots for 1 hour at room temperature at a 1:40,000- fold dilution
and then the blots were washed three times with wash buffer. For
chemiluminescent detection, the Supersignal West Pico Stable
Peroxide solution and the Supersignal West Pico
Luminol/Enhancer solution were used. Both reagents were from
Pierce Chemical Company, Rockford, IL and were used as
described by the manufacturer. The immunoblots were exposed to
film for time frames of 5 to 30 seconds.


Synopses

Vaccines for Mucosal Immunity to Combat Emerging Infectious
Diseases
Frederik W. van Ginkel, Huan H. Nguyen, and Jerry R. McGhee
The University of Alabama at Birmingham, Birmingham,
Alabama, USA

The mucosal immune system consists of molecules, cells, and
organized lymphoid structures intended to provide immunity to
pathogens that impinge upon mucosal surfaces. Mucosal infection
by intracellular pathogens results in the induction of cell-mediated
immunity, as manifested by CD4-positive (CD4 + ) T helper-type 1
cells, as well as CD8 + cytotoxic T-lymphocytes. These responses
are normally accompanied by the synthesis of secretory
immunoglobulin A (S-IgA) antibodies, which provide an important
first line of defense against invasion of deeper tissues by these
pathogens. New-generation live, attenuated viral vaccines, such as
the cold-adapted, recombinant nasal influenza and oral rotavirus
vaccines, optimize this form of mucosal immune protection.
Despite these advances, new and reemerging infectious diseases
are tipping the balance in favor of the parasite; continued mucosal
vaccine development will be needed to effectively combat these
new threats.
 
Behavioral changes in the human host may be contributing to the
emergence of new diseases. Perhaps emerging pathogens become
resistant to antibiotics or (through genetic recombination) become
more resistant to host defenses. Recombination events or lack of
exposure can result in loss of immunity of the population to the
pathogen, as has been well documented with influenza virus.
Recombination events increase the infection rate by the emerging
pathogen and, in the case of influenza virus, occasionally result in
pandemics. A large number of emerging pathogens are mucosally
transmitted and must cross mucosal barriers to infect the host.
Thus, our ultimate defense against new and reemerging infectious
disease will require effective mucosal vaccines. Mucosal surfaces
are prominent in the gastrointestinal, urogenital, and respiratory
tracts and provide portals of entry for pathogens. Increased
sanitation and hygiene, the use of antibiotics, and childhood
vaccination have enormously decreased the death rate from
infectious diseases over the last century (1). Thus, infectious agents
not controlled by antibiotics and improved sanitation and hygiene
measures would most likely be prevalent under current conditions.
Pathogens in this category include respiratory viruses, for example,
influenza virus. Another group includes sexually transmitted
disease (STD) pathogens, for example, HIV. The fact that a)
pneumonia and influenza virus and b) HIV infection were listed as
the leading causes of death (ranked number 6 and 8, respectively)
by infectious agents in the United States in 1997 confirms this
notion (1). However, other viral and bacterial STDs are of major
concern. The best defense against these predominantly mucosal
pathogens would be vaccines, preferably mucosal vaccines capable
of inducing both systemic and mucosal immunity. Although
numerous strategies exist for the induction of mucosal immune
responses, we will focus on the use of live attenuated vectors, such
as Salmonella typhi and adenovirus. In a recent National Institutes
of Health news release (April 8, 1999) from the Institute of
Medicine report on domestic vaccine priorities for the future,
which compared costs and health benefits, influenza was listed as
one of seven diseases requiring an effective vaccine. The induction
of protective mucosal immunity to influenza virus has made
considerable progress with the development of cold-adapted
influenza strains, which are in phase-3 clinical trials. Mucosal
immunity, which is important for long-term protection, forms a
first

line of defense against mucosally transmitted pathogens such as
influenza. Mucosal defense against pathogens consists of both
innate barriers, such as mucous, epithelium, and innate immune
mechanisms, and adaptive host immunity, which at mucosal
surfaces consists predominantly of CD4 + T cells, secretory
immunoglobulin A (S-IgA), and antigen-specific cytotoxic T-
lymphocytes (CTLs). This review will focus on the antigen-
specific mucosal immune system.

 Figure 1. M cells and the induction of mucosal immunity. M cells
are present in mucosal inductive sites in both the intestinal and
upper respiratory tract, specifically in Peyer's patches and the
nasal-associated lymphoid tissue, the tonsils and adenoids. M cells
are thought to play an important role in antigen processing and
possibly the induction of antigen-specific mucosal immunity in
mucosal effector sites. Tissues followed by question marks are
presumed sites since limited data are available on these tissues.

Emerging Pathogens

The major obstacle in combating emerging infectious diseases is
the lack of more effective antibiotics and vaccines. Misuse of
antibiotics has led to antibiotic-resistant pathogens, further
intensifying the need for mucosal vaccine development—a cost-
effective disease-prevention tool. The Department of Vaccines and
Other Biologicals within the World Health Organization (WHO)
has defined several priority vaccines for accelerated introduction;
they include pneumococcal, Haemophilus influenzae type b,
rotavirus, and hepatitis B. Other vaccine goals within WHO are
eradication of polio, reduction of measles cases by 90%, and
elimination of neonatal tetanus (2). For most pathogens targeted by
WHO for vaccines, induction of mucosal immunity appears most
appropriate based on the routes of infection. Thus, a better
understanding of the mucosal immune system is needed before
effective mucosal vaccines can be developed.

The Mucosal Immune System

Mucosal inductive sites in humans, such as the Peyers patches in
the intestinal tract and the nasal-associated lymphoreticular tissue
in the oropharyngeal cavity, stand as sentinels to the intestinal and
respiratory tracts and represent the major sites where mucosal
immune responses are initiated (Figure 1). Common features of
these inductive sites are microfold or M cells.     

Although the precise function of M cells has not yet been
established, recent studies indicate that they are involved in uptake,
transport, processing, and possibly presentation of microbial
antigens (3,4). The interaction of epithelial cells with T and B
lymphocytes induces epithelial cells to differentiate into M cells in
vitro (5), indicating the importance of lymphocyte-epithelial cell
interactions for maintaining M cells in the follicle-associated
epithelium of the Peyer's patches. These lymphocyte- M cell
interactions can occur in the pocket of the M cells and are mediated
through thin cellular extensions, indicating that cell-cell
interactions are an intricate part of the M-cell function and that
they may facilitate transfer of luminal antigens, viruses, bacteria,
and other protein components (3). The ability of the M cell to
transport particulates from the lumen across the epithelial barrier
has been exploited by some pathogens to facilitate entry into the
host, as has been demonstrated for invasive strains of Salmonella
(6) and reovirus (7) (Figure 1). Identification of bacterial and viral
virulence factors associated with targeting M cells, such as the
sigma protein from reovirus, may allow development of mucosal
vaccines and vectors to deliver vaccine components directly into
mucosal inductive sites.

Figure 2. Differentiation and regulation of T-helper subsets and the
immune response in the mucosal compartments. Encounter of
pathogen- derived antigen or vaccine antigen will stimulate T-
helper cells to secrete cytokines. Depending on the stimulus, a Th1
or Th2 cell response is induced. For example, intracellular
pathogens will induce production of IL-12/IL-18 by macrophages,
activating IFN- ð production by NK cells and inducing
differentiation to a Th1- mediated immune response, which
supports CMI and production of complement-fixing antibodies,
presumably by production of cytokines such as IFN-  ð, IL-2, and
TNF-   ð. A Th2 response can be observed upon infection with
parasites or upon vaccine administration; this response is
characterized by production of cytokines such as IL-4, IL-5, IL-6,
IL-10, IL-13 which support humoral immunity. However, for
induction of a S-IgA, TGF-  ð1 is required to enable B cells to
switch to IgA. TGF-  ð1 production is associated with inhibition of
IL-4 production by Th2 cells inhibiting IgE production.

T-Cell and Cytokine Involvement in B-Cell Isotype Commitment
to IgA

Although the variables involved in the switching of B cells to
polymeric IgA (pIgA)-producing plasma cells have been studied,
many questions remain. In recent years, gene-deleted or knockout
mice have contributed to a better understanding of the role of
specific cells, cytokines, and surface molecules involved in IgA
isotype switching. Presumably, isotype switching occurs in
mucosal inductive sites, while IgA production by plasma cells
occurs in mucosal effector sites, separating the IgA switching and
IgA secretion by B cells into different immune compartments (8).
Each of these stages requires specific signals, such as
costimulatory molecules, cytokines, and T-helper cells, to give rise
to antigen-specific S-IgA Abs in mucosal effector sites. Neither
Th1- nor Th2-type cytokines contributed significantly to the
switching of B cells to surface IgA positive (sIgA + ) B cells
(Figure 2). This process required the presence of transforming
growth factor   ð1 (TGF   ð1), which can activate the switch of B
cells to the IgA isotype (9). TGF-  ð1 induces a small proportion
(<2%) of B cells to switch to IgA in activated B-cell cultures
(9,10). However, TGF-  ð1, when used in combination with
additional signals, increased TGF-  ð1-induced switching in 10% to
20% of B cells and approached IgA + B-cell levels observed in
Peyer's patches (11). Thus, multiple activation signals contribute to
the switch to IgA, i.e., B-cell activation by cross-linking the B-cell
antigen receptor, CD40-CD40L interactions to promote switching,
TGF- â ð1 by directing the switch to IgA, and Th2-type cytokines
by increasing the number of post-switch IgA + B cells and their
differentiation into IgA-secreting plasma cells. In addition,
activated T cells and dendritic cells from the Peyer's patches were
more effective in switching sIgM + sIgA - B cells to IgA-producing
cells than were T cells and dendritic cells derived from the spleen
(12). This suggests that mucosal inductive sites contain specialized
T cells or dendritic cells beneficial for B cells to differentiate into
IgA-producing cells. T-cell helper functions play important roles in
generating antigen-specific humoral and cell-mediated immunity in
both systemic and mucosal compartments. The importance of CD4
+ T cells for generating protective immunity is illustrated by the
lack of these cells in AIDS patients. The differentiation of Th0
cells into either Th1 or Th2 is driven by cytokines such as
interleukin 12 (IL-12), interferon  ð (IFN- ð), and IL-4,
respectively. For example, intracellular pathogens, such as viruses
and intracellular bacteria, induce production of IL-12 or IL-18 by
activated macrophages, presumably after ingestion of the
particulate pathogen, inducing IFN-  ð production in natural killer
(NK) cells, which in turn drives the differentiation of Th0 cells
toward a Th1 phenotype producing IFN- ð, IL-2, and tumor
necrosis factor  ð (TNF- ð)ð (Figure 2). Murine Th1-type responses
are associated with cell-mediated immunity, such as delayed-type
hypersensitivity and IgG2a antibody responses (8). Th0 cells are
differentiated into Th2-type cells when soluble exogenous antigen
is administered, triggering CD4 + , NK1.1 + T cells to produce IL-
4. The Th2 cell produces more IL-4, expanding Th2- cells, which
support the associated immune response. Th2 cells secrete
cytokines such as IL-4, IL-5, IL-6, IL-9, IL-10, and IL-13. The
production of IL-4 supports IgG1 subclass and IgE production, but
other antibodies such as IgG2b and IgA, are also produced during a
Th2- dominated response (8).

It is not known whether Th1 or Th2 cells are beneficial for optimal
S-IgA production. Historically, Th2-type cytokines were
considered major helpers for antibody responses. For example, S-
IgA Ab responses were supported by mucosal adjuvants such as
cholera toxin, which induced polarized Th2 cell responses (13).
However, S-IgA Ab responses may also be induced through Th1-
dominated responses, as observed with intracellular pathogens such
as Salmonella in the gastrointestinal tract (14) or influenza virus in
the upper respiratory tract (15). Thus, either Th1 or Th2 cells or a
combination of these cell types can support antigen-specific S-IgA
Ab responses. In this respect, Th2-type cytokines play a role in
terminal differentiation of B cells, that are already committed to
IgA (16-18), while the Th1- type cytokine IFN- ð has been
implicated in the induction of the polymeric Ig receptor (pIgR)
needed for transport of S-IgA (19). Cross-inhibition of Th1 and
Th2 cell-directed IgG2a and IgE production was mediated through
IFN-  ð and IL-4, respectively (Figure 2) (20,21). Mucosal S-IgA
Antibody Responses The hallmark of the mucosal immune system
is the production of S-IgA. S-IgA results from transcytosis of pIgA
across the epithelium through binding to the pIgR. S-IgA is
released from the pIgR by cleavage of the receptor, resulting in
pIgA covalently associated with a substantial part of the pIgR, i.e.,
the secretory component (22). This complex, referred to as S-IgA,
seems to be more resistant to proteolysis in external secretions.
Additional roles for S-IgA in protection are suggested by its
reduction of influenza virus attachment and its prevention of
internalization of virus into baby hamster kidney cells. In contrast,
the action of monomeric IgA is indistinguishable from that
reported for IgG and is less efficient than S-IgA for inhibition of
influenza virus entry (23). In addition, pIgA, as opposed to IgG or
monomeric IgA, neutralizes virus intracellularly, as first was
shown with Sendai virus (24). Furthermore, transport of pIgA
containing immune complexes across epithelial cells expressing
the pIgR is another defense mechanism of the mucosal immune
system against pathogen entry (25).

Figure 3. Pathways of intracellular pathogen clearance from
infected cells by cytotoxic cells. Intracellular pathogen-derived
antigens complexed to MHC class I molecules are recognized by
CTLs, while NK cells recognize the absence or suppressed levels
of MHC class I molecules on infected cells. Activated cytotoxic
cells deliver apoptotic signals through Fas ligand and perforin to
infected cells. They also secrete cytokines (IFN- ð, TNF- ð) and
chemokines (Rantes, MIP-1 ð, MIP- ð) to inhibit or suppress
intracellular pathogen replication.

These characteristics of pIgA are beneficial in preventing infection
and inflammation at epithelial surfaces. The transport of pIgA
across epithelial cells allows active elimination of immune
complexes at mucosal sites and even virus inside epithelial cells.
Evidence that these observations were not an artifact of the in vitro
system was provided by the murine backpack model of rotavirus-
specific monoclonal antibodies (mAb) to VP4 and VP6 proteins. In
this model, nonneutralizing VP6-specific IgA mAb were
protective, but not when administered directly to the
gastrointestinal tract, indicating that IgA transcytosis played a
prominent role in effective immune exclusion (26). Thus, virus-
specific, intra-epithelial IgA can inhibit viral entry and replication.

Cell-Mediated Immunity at Mucosal Surfaces

Although S-IgA has been shown to be an important effector
molecule to protect mucosal surfaces, the contribution to mucosal
protection by the cellular immune system should not be
underestimated. The strategic advantage of cell-mediated versus
antibody-mediated immune responses is that T cells can recognize
peptides derived from core proteins of the pathogen, such as
influenza virus. Core proteins are usually expressed and presented
much earlier during infection than proteins targeted for neutralizing
antibodies, such as HA and NA of influenza virus. Subsequently,
cell-mediated immunity (CMI) occurs before the induction of
antibodies and forms an early line of defense. Although antibodies
to core proteins are also formed later in the immune response, their
exact role in protective immunity is not clear. Besides supporting
humoral immunity, CD4 + T-helper cells function in CMI as
producers of cytokines, which mediate delayed-type
hypersensitivity and support CTLs and which as such are critical
components of the CMI responses to intracellular pathogens. For
example, major histocompatibility complex (MHC)-restricted CTL
responses are supported by Th1 cells. Cytotoxic cells can be
classified based on antigen specificity and MHC restriction, i.e.,
nonspecific cytotoxic cells and antigen-specific, MHC-restricted
CTLs. The first kind is composed of various cell types, including
NK cells and antibody-dependent cytotoxicity, and functions very
early in the immune response (day 1 to 3), and these cells are
detected throughout the mucosal immune system. Presumably, they
decrease pathogen load in the early stage of the immune responses,
while antigen-specific responses are still being established. The
second type, antigen-specific CTL, achieved optimal activity a
little later than nonspecific CTL, i.e., at day 3 to 5 of the immune
response before antibody production. Both antigen-specific and
nonspecific cytotoxic cell types can control growth of intracellular
pathogens by two distinct mechanisms (Figure 3). First, they can
respond to the infection by secreting a number of cytokines such as
IFN-  ð and TNF-  ð (27) or chemokines such as Rantes,
macrophage-inflammatory protein-  ð (MIP-1 ð)ð, and MIP- ð
(28,29). These soluble factors inhibit growth of intracellular
pathogens such as viruses without destroying the host cell. Second,
cytotoxic cells can effectively and efficiently recognize and lyse
infected cells and prevent multiplication of viruses. Little is known
about the induction, compartmentalization, and homing pattern of
cytotoxic cells. Their presence in the mucosal compartment upon
infection reflects their importance for protection against pathogens
at mucosal surfaces.

Antigen-Specific CTLs

CTLs play an important role in the elimination of cells infected
with various intracellular pathogens by recognizing pathogen-
specific antigen/MHC complexes. Antigen-specific CTLs inhibit
further spread of pathogens and help to terminate infections.
Compartmentalization of pathogen-specific CTL responses has
been reported and located at the site of initial infection. For
example, CTLs preferentially compartmentalize in mucosa-
associated lymphoreticular tissues after pulmonary or intestinal
infection.

CTLs in the Intestinal Tract

Presentation of rotavirus by the intestinal mucosal surface was not
required for induction of virus-specific cytotoxic intra-epithelial
lymphocytes in the intestinal tract (30). In addition, the site at
which rotavirus is first presented to the immune system will
determine the site where rotavirus-specific CTL precursors (pCTL)
first appear; however, regardless of the route of inoculation,
rotavirus-specific pCTL can be found throughout the lymphoid
system 21 days after the initial infection (31). Adoptive transfer of
splenic lymphocytes from immunized animals protected suckling
mice against murine rotavirus-induced gastroenteritis in the
absence of rotavirus-specific neutralizing antibodies, indicating
that antigen-specific CTLs protect against mucosal pathogens in
the intestinal tract (32). Thus, thymus-derived  ð T cells can
migrate to the intestinal epithelium after antigen-specific activation
and protect the host against subsequent challenge. This notion is
supported by findings that systemic immunization with attenuated
macaque-specific Simian immunodeficiency virus induced virus-
specific CTL responses in gut-associated lymph nodes and limited
superinfection following mucosal challenge (33).

CTLs in the Respiratory and Urogenital Tracts

The distribution of CTLs following influenza virus infection in
different mucosal compartments indicates that lymph nodes
draining mucosal surfaces function as reservoirs for memory T
cells. Mediastinal lymph nodes, the draining lymph nodes of the
lungs, are considered the site where antigen presentation to T cells
initially occurs before clonal expansion. Subsequently, T cells
migrate to effector sites (mesenchyma of the lungs and airways) to
interact with infected cells (34,35). Thus, compartmentalization of
memory CTL responses to mucosa-associated lymphoreticular
tissue may be related to the initial site of virus infection. This
notion was confirmed by the observation that induction of
protective antiviral memory CTL in mucosal tissues depends on
region-specific mucosal immunization (36). HIV-specific CTL
have been detected in the cervix (37) or semen (38) of HIV-
infected persons. These studies indicate that the initial site of
antigen exposure and induction of antigen-specific CTL responses
in the urogenital tract are associated. Further evidence for this
notion comes from the use of MHC class 1 tetramer technology, by
which antigen-specific quantitation of CD8 + T cells can be
performed. Upon intranasal influenza administration, most antigen-
specific, IFN- ð-producing effector CD8 + T cells were located in
bronchial lavages and both effector (eCTL) and memory CTL
(mCTL) occurred at a much higher frequency than initially thought
based on limiting dilution assays (39). Since respiratory virus
infection induces enlargement of the mediastinal lymph nodes early
in the immune response and since these lymph nodes contain a
relative small number of mCTL after initial exposure, a strong
recruitment from circulating T cells occurs, or alternatively, clonal
expansion of the resident mCTL takes place (39).

The presence of CTLs in mucosal compartments may contribute to
the control of, and recovery from, infection by intracellular
pathogens at mucosal surfaces. Since different pathogens have
distinct infection routes or different localization in the host,
compartmentalization of protective, antigen-specific CTLs may
vary, based on the specific pathogen. In general, mucosal infection
induces primarily antigen-specific CTLs in the mucosal
compartment and mucosa-associated lymphoid organs to control
pathogens at the portal of entry, i.e., the mucosal surfaces.

The Common Mucosal Immune System

Antigenic exposure at mucosal sites activates mucosal B and T-
lymphocytes to emigrate from the inductive site and home to
various mucosal effector sites. The common mucosal immune
system involves homing of antigen-specific lymphocytes to
mucosal effector sites other than the site where initial antigen
exposure occurred. This pathway has almost exclusively been
documented for S-IgA antibody responses at mucosal surfaces
mediated by B cells, but similar events are assumed to take place
with T cells. Different immunization routes, such as oral, rectal,
and intranasal, can induce generalized mucosal immune responses.
However, oral immunization induced a more restricted mucosal
response, as reflected by a more restricted homing receptor profile
than nasal immunization. Specifically, after systemic immunization
the predominant homing receptor on antibody secreting cells is the
L selectin, after oral immunization the ð ð7 integrin, and after nasal
immunization a large portion expressed both the L selectin and the 
ð ð7 integrin. The fact that nasal immunization induced antibodies
in a broader range of tissues, such as saliva and the urogenital tract,
than oral immunization reflects the more restricted nature of oral
immunization (8). Circumstantial evidence indicates the existence
of a common mucosal immune system for cell-mediated immunity
(44). The data available indicate that antigen-specific CTL
responses at mucosal surfaces are dictated by induction of CTL
locally and are not due to migration from distant sites. CTL do
normally migrate to the systemic compartment. It could be
hypothesized that the presence of antigen-specific CTL in the
systemic compartment would allow for quick, protective responses
at any mucosal site, but more research is needed to confirm this
hypothesis. Such research is crucial, since limited CTL activity at
mucosal surfaces could be a built-in mechanism to protect the
mucosal epithelium from damage, a notion supported by the
observation that pCTL in immunologically privileged sites fail to
differenti ate into fully functional CTL, unless exposed to antigen
(40). This concept could have a major influence on future vaccine
development. If mucosal antigen-specific memmory CTL
responses are observed only after mucosal immunization, optimal
protection against pathogens would require the use of mucosal
vaccine. However, systemically induced CTL can generate an
antigen-specific mucosal CTL response; in addition, systemic
immunization can be used for cell-mediated protection at mucosal
surfaces.

Mucosal Vaccines

Although mucosal application of vaccines is attractive for many
reasons, only a few mucosal vaccines, mostly oral, have been
approved for human use. These vaccines include poliovirus,
Salmonella typhi, and the recently approved tetravalent rotavirus
vaccine, RotaShield, consisting of reassorted rhesus-human
rotaviruses. The latter vaccine has recently been associated with
intussusception, a type of bowel obstruction, and its use has been
suspended to await a more detailed analysis of this major problem.
Human approved mucosal vaccines so far involve live attenuated
pathogens, and for this reason oral poliovirus vaccine is
recommended after receiving the injected inactivated virus, since a
limited number of polio cases occur after immunization with the
live attenuated virus. Since the live virus induces better, longer-
lasting protection, it is given after some level of systemic immunity
has been achieved, to limit possible problems. Another oral
vaccine is the typhoid fever vaccine, which consists of attenuated
S. typhi strain Ty21a. The cold-adapted influenza virus (CAIV),
which is in advanced clinical trials, is the first mucosal vaccine
given nasally to humans and has been shown to generate protective
immune responses (41). Thus, CAIV are promising vehicles for
generating protective immunity to influenza in children. The use of
CAIV may also resolve some of the problems observed during the
recent outbreak of the Hong Kong virus. Due to its relatedness with
A/Chicken/Hong Kong/258/ 97 (H5N1) virus, the production of
this virus for vaccine purposes was severely hampered because of
its lethal effect on chicken eggs. The use of CAIV, which is readily
produced by reassortment, might overcome this problem and allow
production of high-titer virus for vaccine purposes. An alternative
approach is the use of DNA vaccines. Plasmid DNA was used in
clinical trials to induce protection against several pathogens,
including hepatitis B virus, herpes simplex virus, HIV, malaria,
and influenza (42). However, in all cases induction of antibodies
and CTL in the systemic but not the mucosal compartment was
reported. Although some progress has been made in inducing
mucosal immunity in laboratory animals with DNA vaccines by
using cationic lipids or other delivery vehicles, as well as
immunostimulatory CpG dinucleotide motifs, no reports exist on
the induction of mucosal immunity by DNA vaccines in humans.
Unmethylated CpG dinucleotides are immunostimulatory,
especially when presented in a 6 base-pair motif in which the
central CpG is flanked by two 5' purines and two 3' pyrimidines.

Another promising avenue for mucosal vaccines is the use of
bacterial adhesins. Mucosal antibodies to these proteins block the
pathogen's ability to penetrate mucosal barriers. Adhesins are very
attractive options because of the highly conserved nature of these
proteins due to their association with conserved host receptor
proteins. The pilus-associated adhesin FimH from uropathogenic
E. coli binding to mannose-oligosaccharides is a vaccine target.
Mucosally administered vaccines containing FimH are in clinical
trials that will assess their efficacy compared with parenterally
administered vaccines. Furthermore, the recently approved
acellular pertussis vaccine also contains adhesins, i.e., the
filamentous hemagglutinin and pertactin, which recognize
sulphated sugars on glycoconjugates and the integrin-binding
protein motif Arg-Gly-Asp, respectively. This indicates that
adhesin-specific immunity might be a successful approach for
generating mucosal protection against pathogens (43). The
importance of blocking the initial attachment and entry into the
host cell has been recognized for some time for viruses such as
influenza, but the use of this approach for bacteria, still in its
infancy, has enormous potential for mucosal vaccines.

Future Directions

The mucosal immune system is a complex and redundant system
that generates large amounts S-IgA as well as cell-mediated
immunity at mucosal surfaces to prevent pathogen infiltration and
inflammation. The mucosal immune system should be most
efficient in providing protection against pathogens and generating
longer-lasting protection through using attenuated pathogens for
vaccines purposes. The only mucosal vaccines approved for
humans are attenuated pathogens. Future mucosal vaccines will
also involve vaccine strategies other than attenuated pathogens. For
example, DNA vaccines or subunit vaccines, such as bacterial
adhesins, in combination with potent mucosal adjuvants (such as
QS-21, a saponin derived from the bark of the South American tree
Quillaja saponia Molina, mutant enterotoxins, unmethylated CpG
motifs, or cytokines such as IL-12) or mucosal delivery systems,
such as microspheres, will have the potential to be the next
generation of vaccines inducing mucosal protection to pathogens in
humans.

Acknowledgment

We thank Kimberly McGhee for editorial assistance on this
manuscript.

The work was supported by National Institutes of Health grants
P30 DK 54781, AI 18958, AI 43197, DK 44240, T32 AI 07150
and contracts N01 65298 and N01 65299.

Dr. van Ginkel is assistant professor of microbiology at the
University of Alabama at Birmingham. His research interests focus
on the mucosal immune response in the respiratory tract, with an
emphasis on viral immunity.

Address for correspondence: Frederik W. van Ginkel, Department
of Microbiology, University of Alabama at Birmingham, BBRB
Room 775, 845 19th Street South, Birmingham, AL 35294, USA;
fax: 205-975-4431; e-mail: fritsv@micro.microbio.uab.edu.


References

1. Control of infectious diseases. MMWR Morb Mortal Wkly Rep
1999;48:621-8.
2. V&B annual report, 1998. Department of vaccines and other
biologicals. Geneva: World Health Organization; 1999.
3. Neutra MR, Frey A, Kraehenbuhl J-P. Epithelial M cells:
gateways for mucosal infection and immunization. Cell
1996;86:345-8.
4. Allan CH, Mendrick DL, Trier JS. Rat intestinal M cells contain
acidic endosomal-lysosomal compartments and express class II
major histocompatibility complex determinants. Gastroenterology
1993;104:698-708.
5. Kerneis S, Bogdanova A, Kraehenbuhl J-P, Pringault E.
Conversion by Peyer's patch lymphocytes of human enterocytes
into M cells that transport bacteria. Science 1998;277:949-52.
6. Jones BD, Ghori N, Falkow S. Salmonella typhimurium initiates
murine infection by penetrating and destroying the specialized
epithelial M cells of the Peyer's patches. J Exp Med 1994;180:7-9.
7. Wolf JL, Rubin DH, Finberg R, Kauffman RS, Sharpe AH, Trier
JS, et al. BN Intestinal M cells: a pathway for entry of reovirus into
the host. Science 1981;212:471-2.
8. McGhee JR, Lamm ME, Strober W. Mucosal immune
responses: an overview. In: Mucosal immunology. Ogra PL,
Mestecky J, Lamm ME, Strober W, Bienenstock J, McGhee JR,
eds. San Diego: Academic Press; 1999. p. 485-506.
9. Ehrhardt RO, Strober W, Harriman GR. Effect of transforming
growth factor (TGF)-  ð1 on IgA isotype expression. TGF- ð1
induces a small increase in sIgA + B cells regardless of the method
of B cell activation. J Immunol 1992;148:3830-6.
10. Coffman RL, Lebman DA, Shrader B. Transforming growth
factor- ð specifically enhances IgA production by
lipopolysaccharide-stimulated murine B lymphocytes. J Exp Med
1989:170:1039-44.
11. McIntyre TM, Kehry MR, Snapper CM. Novel in vitro model
for high-rate IgA class switching. J Immunol 1995;154:3156-61.
12. Spalding DM, Williamson SI, Koopman WJ, McGhee JR.
Preferential induction of polyclonal IgA secretion by murine
Peyer's patch dendritic cell T cell mixtures. J Exp Med
1984;160:941-6.
13. Xu-Amano J, Kiyono H, Jackson RJ, Staats HF, Fujihashi K,
Burrows PD, et al. Helper T cell subsets for immunoglobulin A
responses: oral immunization with tetanus toxoid and cholera toxin
as adjuvant selectively induces Th2 cells in mucosal associated
tissues. J Exp Med 1993;178:1309-20.
14. VanCott JL, Staat HF, Pascual DW, Roberts M, Chatfield SN,
Yamamoto M, et al. Regulation of mucosal and systemic antibody
responses by T-helper subsets, macrophages, and derived cytokines
following oral immunization with live recombinant Salmonella. J
Immunol 1996;156:1504-14.
15. Novak M, Yamamoto M, Fujihashi K, Moldoveanu Z, Kiyono
H, McGhee JR, et al. Ig-secreting and interferon-gamma-producing
cells in mice mucosally immunized with influenza virus. Adv Exp
Med Biol 1995;371B:1587-90.
16. Briere F, Bridon JM, Chevet D, Souillet G, Bienvenu F, Guvet
C, et al. Interleukin 10 induces B lymphocytes from IgA-deficient
patients to secrete IgA. J Clin Invest 1994;94:97-104.
17. Fujihashi K, McGhee JR, Lue C, Beagley KW, Taga T, Hirano
T, et al. Human appendix B cells naturally express receptors for
and respond to interleukin 6 with selective IgA1 and IgA2
synthesis. J Clin Invest 1991;88:248-52.
18. Lebman DA, Nomura DY, Coffman RL, Lee FD. Molecular
characterization of germline immunoglobulin A transcripts
produced during transforming growth factor  ð-induced isotype
switching. Proc Natl Acad Sci U S A 1990;87:3962-6.
19. Wira CR, Richardson J, Prabhala R. Endocrine regulation of
mucosal immunity: effect of sex hormones and cytokines on the
afferent and efferent arms of the immune system in the female
reproductive tract. In: Ogra PL, Mestecky J, Lamm ME, McGhee
JR, Strober W, Bienenstock J, editors. Handbook of mucosal
immunology. San Diego: Academic Press; 1994. p. 705-16.
20. Rizzo LV, DeKruyff RH, Umetsu DT, Caspi RR. Regulation of
the interaction between Th1 and Th2 T cell clones to provide help
for antibody production in vivo. Eur J Immunol 1995;25:708-16.
21. DeKruyff RH, Rizzo LV, Umetsu DT. Induction of
immunoglobulin synthesis by CD4 + T cell clones. Semin
Immunol 1993;5:421-30.
22. Mostov KE. Transepithelial transport of immunoglobulins.
Annu Rev Immunol 1994;12:63-84. 23. Taylor HP, Dimmock NJ.
Mechanism of neutralization of influenza virus by secretory IgA is
different from that of monomeric IgA or IgG. J Exp Med
1985;161:198-209.
24. Mazanec MB, Kaetzel CS, Lamm ME, Fletcher D, Nedrud JG.
Intracellular neutralization of virus by immunoglobulin A
antibodies. Proc Natl Acad Sci U S A 1992;89:6901-5.
25. Kaetzel CS, Robinson JK, Lamm ME. Epithelial transcytosis of
monomeric IgA and IgG cross-linking through antigen to
polymeric IgA. A role for monomeric antibodies in the mucosal
immune response. J Immunol 1994;152:72-6.
26. Burns JW, Siadat-Pajouh M, Krishnaney AA, Greenberg HB.
Protective effects of rotavirus VP6- specific IgA monoclonal
antibodies that lack neutralizing activity. Science 1996;272:46-8.
27. Chisari FV. Cytotoxic T cells and viral hepatitis. J Clin Invest
1997;99:1472-7.
28. Fehniger TA, Herbein G, Yu H, Para MI, Bernstein ZP,
O'Brien WA, et al. Natural killer cells from HIV-1 + patients
produce C-C chemokines and inhibit HIV-1 infection. J Immunol
1998;161:6433-8.
29. Yang OO, Kalams SA, Trocha A, Cao H, Luster A, Johnson
RP, et al. Suppression of human immunodeficiency virus type 1
replication by CD8 + cells: evidence for HLA class I-restricted
triggering of cytolytic and noncytolytic mechanisms. J Virol
1997;71:3120-8.
30. Offit PA, Dudzik KI. Rotavirus-specific cytotoxic T
lymphocytes appear at the intestinal mucosal surface after rotavirus
infection. J Virol 1989;63:3507-12.
31. Offit PA, Cunningham SL, Dudzik KI. Memory and
distribution of virus-specific cytotoxic T lymphocytes (CTLs) and
CTL precursors after rotavirus infection. J Virol 1991;65:1318-24.
32. Offit PA, Dudzik KI. Rotavirus-specific cytotoxic T
lymphocytes passively protect against gastroenteritis in suckling
mice. J Virol 1990;64:6325-8.
33. Cranage MP, Whatmore AM, Sharpe SA, Cook N,
Polyanskaya N, Leech S, et al. Macaques infected with live
attenuated SIVmac are protected against superinfection via the
rectal mucosa. Virology 1997;229:143-54.
34. Hou S, Doherty PC. Partitioning of responder CD8 + T cells in
lymph node and lung of mice with Sendai virus pneumonia by
LECAM-1 and CD45RB phenotype. J Immunol 1993;150:5494-
500.
35. Carding SR, Allan W, McMickle A, Doherty PC. Activation of
cytokine genes in T cells during primary and secondary murine
influenza pneumonia. J Exp Med 1993;177:475-582.
36. Gallichan WS, Rosenthal KL. Long-lived cytotoxic T
lymphocyte memory in mucosal tissues after mucosal but not
systemic immunization. J Exp Med 1996;184:1879-90.
37. Musey L, Hu Y, Eckert L, Christensen M, Karchmer T,
McElrath MJ. HIV-1 induces cytotoxic T lymphocytes in the
cervix of infected women. J Exp Med 1997;185:293-303.
38. Quayle AJ, Coston WMP, Trocha AK, Kalams SA, Mayer KH,
Anderson DJ. Detection of HIV-1-specific CTLs in the semen of
HIV-infected individuals. J Immunol 1998;161:4406-10.
39. Flyn KJ, Belz GT, Altman JD, Ahmed R, Woodland DL,
Doherty PC. Virus-specific CD8 + T cells in primary and
secondary influenza pneumonia. Immunity 1998;8:683-91.
40. Ksander BR, Streilein JW. Failure of infiltrating precursor
cytotoxic T cells to acquire direct cytotoxic function in
immunologically privileged sites. J Immunol 1990;145:2057-63.
41. Gruber WC, Belshe RB, King JC, Treanor JJ, Piedra PA,
Wright PA, et al. Evaluation of live attenuated influenza vaccines
in children 6-18 months of age: safety, immunogenicity, and
efficacy. J Infect Dis 1996;173:1313-9.
42. Weiner DB, Kennedy RC. Genetic vaccines. Sci Am
1999;281:50-7.
43. Wizemann TM, Adamou JE, Langermann S. Adhesins as
targets for vaccine development. Emerg Infect Dis 1999;5:395-403.
44. Gallichan WS, Rosenthal KL. Long-lived cytotoxic T
lymphocyte memory in mucosal tissues after mucosal but not
systemic immunization. J Exp Med 1996;184:1879-90.


Research

Competence of American Robins as Reservoir Hosts for Lyme
Disease Spirochetes

Dania Richter,*† Andrew Spielman,* Nicholas Komar,* and
Franz-Rainer Matuschka*†
*Harvard School of Public Health, Boston, Massachusetts, USA;
and †Charité, Medizinische Fakultät der Humboldt-Universität zu
Berlin, Germany

To explore the competence of American robins as a reservoir for
Lyme disease spirochetes, we determined the susceptibility of these
birds to tickborne spirochetes and their subsequent infectivity for
larval vector ticks. Robins acquired infection and became
infectious to almost all xenodiagnostic ticks soon after exposure to
infected nymphal ticks. Although infectivity waned after 2 months,
the robins remained susceptible to reinfection, became infectious
again, and permitted repeated feeding by vector ticks. In addition,
spirochetes passaged through birds retained infectivity for
mammalian hosts. American robins become as infectious for vector
ticks as do reservoir mice, but infectivity in robins wanes more
rapidly.

Lyme disease spirochetes (Borrelia burgdorferi s.l.) perpetuate in
cycles involving rodent reservoir hosts, such as white-footed mice (
Peromyscus leucopus) in eastern North America (1) and Apodemus
spp. mice in Eurasia (2). Such hosts readily become infectious to
vector ticks and appear to remain so lifelong. They serve as natural
reservoirs of infection because numerous subadult vector ticks
parasitize them, they are abundant, and they remain in constant
residence in tick-infested sites. Rats (Rattus norvegicus and R.
rattus) (3,4) and European dormice (Glis glis and Eliomys
quercinus) (5,6) are similarly competent and locally important
hosts to these pathogens. Although hares (Lepidus timidus) support
infection with Lyme disease spirochetes in Europe (7), the
competence of rabbits for these spirochetes has not been proven. A
related spirochete, B. andersonii, appears to be specific to
American rabbits (8,9). Ungulates, however, are not competent as
hosts for Lyme disease spirochetes. Deer in intensely enzootic sites
fail to infect vector ticks (10,11), and cattle and sheep exposed to
infected vector ticks never become infectious (12). Lyme disease
spirochetes infect diverse mammals, but not all of them serve as
competent hosts.

Certain birds may also contribute to the transmission of Lyme
disease spirochetes. Numerous vector ticks parasitize birds in
nature, including spirochete-infected larval ticks that presumably
acquired infection from these birds (13-19). Seabirds appear to
maintain such a cycle on Arctic islands (20). In an enzootic site,
American robins (Turdus migratorius) are considered to be an
avian candidate as a reservoir of infection because they are locally
abundant, forage in ground and brush vegetation, and are
frequently infested by vector ticks (21). Catbirds (Dumetella
carolinensis) in enzootic sites, however, appear not to infect vector
ticks (22). Although spirochetes have been isolated from naturally
infected European blackbirds (Turdus merula) (15), a laboratory
study failed to demonstrate reservoir competence of these birds
(23); the reason for this discrepancy remains unclear. Nymphal
vector ticks occasionally become infected after feeding on
spirochete-exposed pheasants (Phasianus colchicus); these birds,
however, cannot contribute to transmission because larval ticks
seem not to feed on them, either in the laboratory or in nature
(24,25).

Figure 1. American robin with attached nymphal deer ticks (Ixodes
dammini).

Chickens (Gallus gallus) become infectious, but only transiently
and only when they are about a week old (26). Spirochetes
inoculated by syringe can be detected in various tissues of canaries
(Serinus canaria), bobwhite quail (Colinus virginianus), and
Japanese quail (Coturnix coturnix) (27-29). The competence of
candidate reservoir birds as hosts for tickborne spirochetes has not
been thoroughly evaluated. Tickborne Lyme disease spirochetes
may readily infect certain birds, and these birds may subsequently
infect numerous vector ticks. Therefore, we determined the
susceptibility of American robins to tickborne spirochetes and their
subsequent infectivity for larval deer ticks (Ixodes dammini, which
differ from I. scapularis [30]). American robins are suitable
candidate hosts because they are infested naturally by numerous
vector ticks. We evaluated the reservoir competence of captive
robins and determined whether they could subsequently be
reinfested and reinfected by vector ticks and whether spirochetes
passaged through birds remain infective to rodents.

Materials and Methods

American robins were captured in mist nets in Brookline, MA
(U.S. Fish and Wildlife Scientific Collecting Permit# MB719506-
0), and maintained at 20°C with a photoperiod of 16:8 hrs (L:D)
(Animal Welfare Assurance# A-3431-01). To establish that these
birds had not already been infected by Lyme disease spirochetes,
nymphs derived from xenodiagnostic larval deer ticks that had
engorged on them were examined by dark-field microscopy. No
spirochete-infected ticks were found. The birds were held in
captivity for 6 months before they were exposed to infected ticks.
The colony of laboratory-raised deer ticks was originally isolated
from ticks captured in Ipswich, MA, and was maintained by
feeding adults on rabbits and subadults on outbred laboratory mice.
To infect ticks, larvae were fed on outbred laboratory mice infected
with the N40 strain of B. burgdorferi s.s. The resulting infected
nymphal ticks were used to infect the robins. Xenodiagnosis was
used to determine whether host animals had become infected by
spirochetes; larval ticks were permitted to feed on them and the
resulting nymphs were examined for spirochetes by dark-field
microscopy. The larval ticks used for the xenodiagnosis were in
their third generation of continuous laboratory breeding and had
not previously been exposed to spirochete-infected hosts. Engorged
ticks were held at 20±2°C over supersaturated MgSO 4 in sealed
desiccator jars with 16 hours of light per day. Hatched or molted
ticks remained in screened vials until they were placed on hosts 2-4
months later.

To infest the robins with subadult ticks, each bird was restrained
while ticks were brushed onto its head and neck. Each bird was
then placed in a cotton bag suspended over water and kept in the
dark for approximately 3 hours to limit their activity. The birds
were then caged individually over pans of water in cages with
wire-mesh floors and fronts (Safeguard Products, New Holland,
PA). The contents of the pans were inspected twice a day, and
detached ticks were removed promptly. The birds were exposed
three times to nymphal ticks to test their susceptibility to repeated
feeding (Figure 1). The first and last of these infestations also
served to infect them. Xenodiagnostic larval ticks were permitted
to feed on these robins nine times during the study.

Results

To evaluate the reservoir competence of birds for Lyme disease
spirochetes, we exposed each of four American robins to the bites
of 12 spirochete-infected nymphal deer ticks and determined the
duration of their subsequent infectivity for larval deer ticks. The
presence of spirochetes in these infecting nymphs was verified by
dark-field microscopic examination of the gut contents of the
resulting adults; virtually all (97%) contained spirochetes. Birds
were exposed initially to


xenodiagnostic larvae at the same time they were exposed to the
infecting nymphs. Xenodiagnosis was repeated after two days and
at intervals for six months. Although larval ticks feeding
simultaneously with the infected nymphs failed to acquire
spirochetes, two birds infected xenodiagnostic larvae within 6 days
after exposure to infected nymphs (Figure 2). All four birds were
infectious to xenodiagnostic ticks that detached 12 days after
exposure and infected 88% (±6.6) of ticks for at least another 3
weeks (Figure). Infectivity waned by 2 months and disappeared by
6 months. Almost all the ticks became infected that fed 2 and 4
weeks after the birds were infected (Table 1). American robins are
fully but transiently competent as hosts for Lyme disease
spirochetes.

Figure 2. Infectivity to larval vector ticks of four American robins
exposed to nymphal deer ticks infected with Lyme disease
spirochetes on days 0 and 186. Each observation was recorded on
the day on which each xenodiagnostic test was complete.

Table 1. Infectivity to larval deer ticks of four American robins
infected with tickborne Lyme disease spirochetes

 We then determined whether American robins tolerate reinfection
by tickborne spirochetes. Six months after they were initially
infected, when vector ticks no longer acquired spirochetes from
them, eight infected nymphs were permitted to engorge on each
bird. All four robins regained infectivity for ticks within the next 2
weeks (Figure 2). Virtually all xenodiagnostic larval ticks that fed
on three of these reinfected birds acquired spirochetes, and a third
of them did so on the fourth bird. Although spirochete-infected
robins only transiently infect vector ticks, the birds remain tolerant
to reinfection and become infectious again.

The ability of nymphal deer ticks to reinfest American robins was
evaluated by recording their feeding success. On initial exposure,
96% of 48 nymphs feeding on the four birds engorged successfully;
98% of 40 nymphs engorged 3 weeks later; and 82% of 32 nymphs
engorged 6 months later, when the birds were exposed for the third
time. The robins appear to freely tolerate repeated feeding by
nymphal deer ticks. To determine whether spirochetes passaged
through birds retain infectivity for mice, we permitted nymphs that
had acquired infection as larvae from birds to feed on mice. The
ticks that were used had acquired spirochetes 3 weeks after their
avian hosts had been infected. Each of four outbred laboratory
mice was exposed to six of the resulting nymphs. Two weeks later,
10 xenodiagnostic larval ticks were permitted to engorge on each
of these mice. All but one of the 40 resulting nymphal ticks
acquired spirochetes (Table 2). Spirochetes, therefore, remain
infectious to mammals after a tickborne passage through birds.

Discussion

A competent reservoir host for the agent of Lyme disease readily
acquires infection from vector ticks, permits spirochetes to
proliferate, and readily infects vector ticks. In white-footed mice,
infection generally becomes established after a single feeding by an
infectious tick, and more than half retain infection for
approximately 6 months (31). As many as three-quarters of the
larval ticks that feed on such mice acquire infection, which
subsequently wanes. Hamsters are similarly infectious (32). The
various murids, glirids, voles, and sciurids that have been tested
appear as reservoir-competent for spirochetes as are cricetids
(2,5,6,33). Although certain genospecies of the Lyme disease
spirochetes are said to be more mouse-adapted than others (34), no
experimental evidence is available to support this concept.
Rodents, in general, are readily infected by Lyme disease
spirochetes, and most remain infectious to vector ticks for at least 6
months.

Table 2. Infectivity for laboratory mice of Lyme disease
spirochetes that had previously infected American robins

 The standard of proof that has been applied to rodent reservoirs of
spirochetal infection has not previously been applied to candidate
avian reservoirs. Certain birds, including quail and canaries, can
readily be infected by syringe-inoculated spirochetes. Spirochetal
DNA generally becomes detectable in the viscera (27-29), and
viable spirochetes can be cultured from these tissues. Tickborne
spirochetes infect pheasants (24) and week-old chickens (26).
About a third of these chicks infect ticks, but they remain
infectious for only a week. Adult pheasants infect about a quarter
of vector ticks that engorge on them. European blackbirds appear
not to be susceptible to tickborne infection (23). American robins,
however, appear far more reservoir-competent. Robins bitten by
spirochete-infected nymphal ticks subsequently infect larval ticks,
and virtually all larvae become infected after feeding on these birds
throughout the following month. Infectivity declines, however, and
few ticks acquire infection from these birds after approximately 3
months. Certain, but not all, bird species can readily be infected by
Lyme disease spirochetes and subsequently become highly but
transiently infectious.

A competent reservoir host tolerates feeding by both subadult
stages of vector ticks. Although nymphs feed readily on pheasants,
larvae seldom attach to these birds, either in nature or in the
laboratory (24). About as many nymphal as larval deer ticks attach
naturally to American robins (21). Several times as many larvae as
nymphs, however, feed on white-footed mice in nature (35).
Nymphal vector ticks attach more readily to rats, dormice, or the
American reservoir host (P. leucopus) than to European mice
(Apodemus spp.) or voles (2,3,5,31). Perhaps particular subadult
stages of vector ticks are more readily attracted to certain
vertebrate animals than to others.

To be competent as reservoir hosts for Lyme disease spirochetes,
an animal must tolerate repeated feedings by vector ticks. White-
footed or laboratory mice serve as hosts to deer ticks almost as
readily during their fifth exposure as their first (36). Voles are
much less tolerant (37), and rabbits (unpub. obs.) and guinea pigs
(38) seldom tolerate more than one episode of attachment. Vector
ticks feed repeatedly on American robins as successfully as on the
natural rodent reservoirs of the Lyme disease spirochete.
Spirochete-infected larval ticks frequently are taken from birds in
Lyme disease-enzootic areas. Of 20 species of North American
passerine birds that were parasitized by larval vector ticks, 45%
hosted spirochete-infected larvae (13,18). Spirochetes infected
10% to 40% of these batches of infected ticks. In Europe, 81% of
16 bird species harbored infected ticks; of these cohorts, 11% to
75% of the larval ticks were infected (17). Because Lyme disease
spirochetes infect <1% of questing larval vector ticks captured in
nature and because inherited infection is exceedingly rare (39,40),
these larval ticks most likely acquired spirochetes from their avian
hosts. Nevertheless, the avian host-range of Lyme disease
spirochetes remains to be defined. An effective reservoir of
infection for a pathogen would generate for each primary infection
at least as many secondary infections. The contribution of various
alternative reservoir hosts for Lyme disease spirochetes, however,
can be compared by estimating the proportion of infections in
nymphal vector ticks that derive from a particular vertebrate
population. This calculation includes estimates of feeding density
of vector ticks on the candidate reservoir host, its local density, and
its prevalence of infection (2,5). At least twice as many vector ticks
appear to acquire infection from a population of edible dormice
(Glis glis) as from other rodents locally abundant in Central Europe
(5). Similar evidence implicates white-footed mice in eastern North
America (1). Information that would permit comparable estimates
of the reservoir significance of local populations of birds, however,
remains unavailable. Thus, it is still uncertain whether a bird,
although reservoir-competent for the spirochete, may serve as an
important reservoir in nature.

Because the prevalence of Lyme disease spirochetes in American
robins has not yet been determined for a representative enzootic
site, the importance of these birds as reservoir hosts remains
speculative. Robins may contribute to the force of transmission of
the agent of Lyme disease, because these locally abundant birds are
reservoir-competent and may be infested by numerous larval ticks.
Of the larvae that feed on peridomestic birds in North American
enzootic sites, approximately three-quarters appear to engorge on
robins (21). Although they may be about as abundant as white-
footed mice, robins appear to forage more frequently on lawns than
do mice (21). The contribution of robins to the peridomestic risk
for Lyme disease may depend on the ability of engorged larval
ticks detaching from the birds to develop in such habitat. Questing
nymphal ticks appear less prevalent on lawns than on other
vegetation in residential enzootic sites (41). Birds are more vagile
than mice, which might dilute risk by diffusing infected ticks into
sites in which transmission is unlikely. Alternatively, their vagility
might seed new foci of transmission, if ticks detach in sites that are
suitable for their development. Certain migratory passerines have
been associated with long-distance dispersal of vector ticks (17-
19). Robins may contribute to the emergence of Lyme disease in
previously unaffected sites to the extent that the season of their
migration overlaps with that of the activity of subadult vector ticks.
Our finding that American robins are reservoir-competent for
Lyme disease spirochetes warrants further epizootiologic studies,
including estimates of the prevalence of spirochetes in these birds
in nature.

This study was supported by grant Ma 942/10-1 from the Deutsche
Forschungsgemeinschaft and by grant AI 42402-01 from the
National Institutes of Health.

Dr. Richter is a postdoctoral fellow conducting joint research in the
laboratories of Dr. Matuschka at the Charité Medical School,
Humboldt-Universität zu Berlin, and Dr. Spielman, at the Harvard
School of Public Health. Her research interests focus on the
immunologic and molecular interface of the host-vector-pathogen
relationship in the epizootiology of Lyme disease.

Address for correspondence: Dania Richter, Institut für Pathologie,
Charité, Medizinische Fakultät der Humboldt-Universität zu
Berlin, Malteserstraße 74-100, 12249 Berlin, Germany; fax: 49-30-
776-2085; e-mail: dania.richter@charite.de.

References

1. Spielman A. Lyme disease and human babesiosis: evidence
incriminating vector and reservoir hosts. In: Englund PT, Sher A,
editors. The biology of parasitism. New York: Alan R. Liss; 1988.
p. 147-65.
2. Matuschka F-R, Fischer P, Heiler M, Richter D, Spielman A.
Capacity of European animals as reservoir hosts for the Lyme
disease spirochete. J Infect Dis 1992;165:479-83.
3. Matuschka F-R, Endepols S, Richter D, Ohlenbusch A, Eiffert
H, Spielman A. Risk of urban Lyme disease enhanced by the
presence of rats. J Infect Dis 1996;174:1108-11.
4. Smith RP, Rand PW, Lacombe EH, Telford SR, Rich SM,
Piesman J, et al. Norway rats as reservoir hosts for Lyme disease
spirochetes on Monhegan Island, Maine. J Infect Dis
1993;168:687-91.
5. Matuschka F-R, Eiffert H, Ohlenbusch A, Spielman A.
Amplifying role of edible dormice in Lyme disease transmission in
central Europe. J Infect Dis 1994;170:122-7.
6. Matuschka F-R, Allgöwer R, Spielman A, Richter D.
Characteristics of garden dormice that contribute to their capacity
as reservoirs for Lyme disease spirochetes. Appl Environ
Microbiol 1999;65:707-11.
7. Jaensen TGT, Tälleklint L. Lyme borreliosis spirochetes in
Ixodes ricinus (Acari: Ixodidae) and the varying hare on isolated
islands in the Baltic Sea. J Med Entomol 1996;33:339-43.
8. Anderson JF, Magnarelli LA, LeFebvre RB, Andreadis TG,
McAninch JB, Perng G-C, et al. Antigenically variable Borrelia
burgdorferi isolated from cottontail rabbits and Ixodes dentatus in
rural and urban areas. J Clin Microbiol 1989;27:13-20.
9. Marconi RT, Liveris D, Schwartz I. Identification of novel
insertion elements, restriction fragment length polymorphism
patterns, and discontinuous 23S rRNA in Lyme disease
spirochetes—phylogenetic analyses of rRNA genes and their
intergenic spacers in Borrelia japonica sp. nov. and genomic group
21038 (Borrelia andersonii sp. nov.) isolates. J Clin Microbiol
1995;33:2427-34.
10. Matuschka F-R, Heiler M, Eiffert H, Fischer P, Lotter H,
Spielman A. Diversionary role of hooded game in the transmission
of Lyme disease spirochetes. Am J Trop Med Hyg 1993;48:693-9.
11. Telford SR, Mather TN, Moore SI, Wilson ML, Spielman A.
Incompetence of deer as reservoirs of the Lyme disease spirochete.
Am J Trop Med Hyg 1988;39:105-9.
12. Matuschka F-R, Eiffert H, Ohlenbusch A, Richter D, Schein E,
Spielman A. Transmission of the agent of Lyme disease on a
subtropical island. Tropical Medicine and Parasitology
1994;45:39-44.
13. Anderson JF, Johnson RC, Magnarelli LA, Hyde FW.
Involvement of birds in the epidemiology of the Lyme disease
agent Borrelia burgdorferi. Infect Immun 1986;51:394-6.
14. Anderson JF, Magnarelli LA, Stafford KC. Bird-feeding ticks
transstadially transmit Borrelia burgdorferi that infect Syrian
hamsters. J Wildl Dis 1990;26:1-10.
15. Humair P-F, Postic D, Wallich R, Gern L. An avian reservoir
(Turdus merula) of the Lyme borreliosis spirochetes. Zentralbl
Bakteriol 1998;287:521-38.
16. Nakao M, Miyamoto K, Fukunaga M. Lyme disease
spirochetes in Japan: enzootic transmission cycles in birds, rodents,
and Ixodes persulcatus ticks. J Infect Dis 1994;170:878-82.
17. Olsen B, Jaenson TGT, Bergström S. Prevalence of Borrelia
burgdorferi sensu lato-infected ticks on migrating birds. Appl
Environ Microbiol 1995;61:3082-7.
18. Rand PW, Lacombe EH, Smith RP, Ficker J. Participation of
birds in the emergence of Lyme disease in southern Maine. J Med
Entomol 1998;35:270-6.
19. Smith RP, Rand PW, Lacombe EH, Morris SR, Holmes DW,
Caporale DA. Role of bird migration in the long-distance dispersal
of Ixodes dammini, the vector of Lyme disease. J Infect Dis
1996;174:221-4.
20. Olsen B, Jaenson TGT, Noppa L, Bunikis J, Bergström S. A
Lyme borreliosis cycle in seabirds and Ixodes uriae ticks. Nature
1993;362:340-2.
21. Battaly GR, Fish D. Relative importance of bird species as
hosts for immature Ixodes dammini (Acari: Ixodidae) in a
suburban residential landscape of southern New York State. J Med
Entomol 1993;30:740-7.
22. Mather TN, Telford SR, MacLachlan AB, Spielman A.
Incompetence of catbirds as reservoirs for the Lyme disease
spirochete (Borrelia burgdorferi). J Parasitol 1989;75:66-9.
23. Matuschka F-R, Spielman A. Loss of Lyme disease spirochetes
from Ixodes ricinus ticks feeding on European blackbirds. Exp
Parasitol 1992;74:151-8.
24. Kurtenbach K, Carey D, Hoodless AN, Nuttall PA, Randolph
SE. Competence of pheasants as reservoirs for Lyme disease
spirochetes. J Med Entomol 1998;35:77-81.
25. Kurtenbach K, Peacey M, Rijpkema SGT, Hoodless AN,
Nuttall PA, Randolph SE. Differential transmission of the
genospecies of Borrelia burgdorferi sensu lato by game birds and
small rodents in England. Appl Environ Microbiol 1998;64:1169-
74.
26. Piesman J, Dolan MC, Schriefer ME, Burkot TR. Ability of
experimentally infected chickens to infect ticks with the Lyme
disease spirochete, Borrelia burgdorferi. Am J Trop Med Hyg
1996;54:294-8.
27. Bishop KL, Khan MI, Nielsen SW. Experimental infection of
northern bobwhite quail with Borrelia burgdorferi. J Wildl Dis
1994;30:506-13.
28. Isogai E, Tanaka S, Braga IS, Itakura C, Isogai H, Kimura K, et
al. Experimental Borrelia garinii infection of Japanese quail. Infect
Immun 1994;62:3580-2.
29. Olsen B, Gylfe A, Bergström S. Canary finches (Serinus
canaria) as an avian infection model for Lyme borreliosis.
Microbial Pathogenesis 1996;20:319-24.
30. Telford SR. The name Ixodes dammini epidemiologically
justified. Emerg Infect Dis 1998;4:132-4.
31. Donahue JG, Piesman J, Spielman A. Reservoir competence of
white-footed mice for Lyme disease spirochetes. Am J Trop Med
Hyg 1987;36:92-6.
32. Piesman J. Intensity and duration of Borrelia burgdorferi and
Babesia microti infectivity in rodent hosts. Int J Parasitol
1988;18:687-9.
33. Matuschka F-R, Endepols S, Richter D, Spielman A.
Competence of urban rats as reservoir hosts for Lyme disease
spirochetes. J Med Entomol 1997;34:489-93.
34. Humair P-I, Peter O, Wallich R, Gern L. Strain variation of
Lyme disease spirochetes isolated from Ixodes ricinus ticks and
rodents collected in two epidemic areas in Switzerland. J Med
Entomol 1995;32:433-8.
35. Wilson ML, Spielman A. Seasonal activity of immature Ixodes
dammini (Acari: Ixodidae). J Med Entomol 1985;22:408-14.
36. Richter D, Spielman A, Matuschka F-R. Effect of prior
exposure to noninfected ticks on susceptibility of mice to Lyme
disease spirochetes. Appl Environ Microbiol 1998;64:4596-9.
37. Davidar P, Wilson M, Ribeiro JMC. Differential distribution of
immature Ixodes dammini (Acari: Ixodidae) on rodent hosts. J
Parasitol 1989;75:898-904.
38. Nazario S, Das S, de Silva AM, Deponte K, Marcantonio N,
Anderson JF, et al. Prevention of Borrelia burgdorferi transmission
in guinea pigs by tick immunity. Am J Trop Med Hyg
1998;58:780-5.
39. Matuschka F-R, Schinkel TW, Klug B, Spielman A, Richter D.
Failure of Ixodes ticks to inherit Borrelia afzelii infection. Appl
Environ Microbiol 1998;64:3089-91.
40. Piesman J, Donahue JG, Mather TN, Spielman A.
Transovarially acquired Lyme disease spirochetes (Borrelia
burgdorferi) in field-collected larval Ixodes dammini (Acari:
Ixodidae). J Med Entomol 1986;23:219.
41. Frank DH, Fish D, Moy FH. Landscape features associated
with Lyme disease risk in a suburban residential environment.
Landscape Ecol 1998;13:27-36.


Research

Vibrio cholerae O139 in Calcutta, 1992-1998: Incidence,
Antibiograms, and Genotypes

Arnab Basu,* Pallavi Garg,* Simanti Datta,* Soumen
Chakraborty,* Tanuja Bhattacharya,* Asis Khan,* T.
Ramamurthy,* S.K. Bhattacharya,* Shinji Yamasaki,† Yoshifumi
Takeda,‡ and G. Balakrish Nair*
*National Institute of Cholera and Enteric Diseases, Calcutta,
India; †International Medical Center of Japan, Tokyo, Japan; and
‡National Institute of Infectious Diseases, Tokyo, Japan

We report results of surveillance for cholera caused by Vibrio
cholerae O139 from September 1992, when it was first identified,
to December 1998. V. cholerae O139 dominated as the causative
agent of cholera in Calcutta during 1992-93 and 1996-97, while the
O1 strains dominated during the rest of the period. Dramatic shifts
in patterns of resistance to cotrimoxazole, neomycin, and
streptomycin were observed. Molecular epidemiologic studies
showed clonal diversity among the O139 strains and continuous
emergence of new epidemic clones, reflected by changes in the
structure, organization, and location of the CTX prophages in the
V. cholerae O139 chromosome.

Vibrio cholerae, the gram-negative organism that causes cholera, is
well defined on the basis of biochemical tests and DNA homology
studies (1), but the serogroups of the species differ in their
pathogenic potential. Of the 193 recognized "O" serogroups of V.
cholerae (2), only O1 and O139 cause epidemic and pandemic
cholera. V. cholerae O139 was first identified in September 1992
in southern India (3) and rapidly spread to all cholera-endemic
areas in India (4) and neighboring countries (5). In February 1994,
a new clone of V. cholerae O1 El Tor biotype (6) replaced the
O139 serogroup as the dominant serogroup causing cholera in
Calcutta (7). After a 33-month quiescent period, a new clone of V.
cholerae O139 (8,9) appeared in August 1996 in Calcutta (10) and
was the dominant serogroup until September 1997. This new clone
has also spread to other parts of India (11). The National Institute
of Cholera and Enteric Diseases, Calcutta, conducts continuous
surveillance for cholera in Calcutta in particular, and in India in
general. During September-October 1998, we observed an increase
in the incidence of O139 cholera, prompting study of these O139
strains. We report the findings of surveillance performed from
September 1992 to December 1998 in India, which enabled us to
track and catalog changes in the O139 strains since its
identification.

The Study

Hospital Surveillance and Bacteriology

Stool specimens were obtained from patients admitted to the
Infectious Diseases Hospital, Calcutta, the only hospital that admits
cholera patients from the city and its suburbs. Since 1995, on two
randomly selected days per week, every fifth patient has been
enrolled from all patients with diarrhea or dysentery, with or
without other complaints, visiting the emergency department of the
Infectious Diseases Hospital, Calcutta. Before 1995, only patients
with diarrhea admitted to the Infectious Diseases Hospital,
Calcutta, from Monday to Friday between 9 a.m. and 1 p.m. were
enrolled. Methods of collection, transport, and bacteriologic
examination of stool samples and identification and serotyping of
V. cholerae have been described (7). The National Institute of
Cholera and Enteric Diseases, Calcutta, is India's national
reference laboratory for cholera, and strains from throughout the
country are sent here for confirmation and phage typing. Strains
received are characterized by an array of tests (4). In this study,
representative strains of V. cholerae O139 isolated from
hospitalized patients in Calcutta from 1992 to 1998 and 15 V.
cholerae O139 strains from other parts of India (eastern, central,
northern, and southern India) isolated in 1998 were randomly
selected for molecular characterization.

Table 1. Oligonucleotide primer sequences used in polymerase
chain reaction assays of Vibrio cholerae O139

Antimicrobial Susceptibility

The V. cholerae O139 strains were examined for resistance to
ampicillin (10 ˜g), chloramphenicol (30 ˜g), cotrimoxazole (25 ˜g),
ciprofloxacin (5 ˜g), furazolidone (100 ˜g), gentamycin (10 ˜g),
neomycin (30 ˜g), nalidixic acid (30 ˜g), norfloxacin (10 ˜g),
streptomycin (10 ˜g), and tetracycline (30 ˜g) by using antibiotic-
impregnated commercial disks (Hi Media, Mumbai, India).
Characterization of the strains as susceptible or resistant was based
on the size of inhibition zones around each disk, according to the
manufacturer's instructions, which followed World Health
Organization recommendations (12). Strains showing an
intermediate zone of inhibition were interpreted as resistant on the
basis of previous minimum inhibitory concentration studies
conducted with V. cholerae (13).

Polymerase Chain Reaction (PCR) Assay

PCR assays (both multiple and single) were used to screen the
O139 strains for five important virulence genes. The first PCR
assay used three pairs of oligonucleotide primers for virulence
genes ctxA (301 bp), tcpA (classical; 617 bp), and tcpA (El Tor;
471 bp), and the second and third PCR assays used primer pairs for
the virulence genes zot and ace, respectively (Table 1) (15). The
composition of the reaction mix and cycling conditions for
amplification have been described (16). Amplification was done by
using an automated thermal cycler (Biometra Gottingen, Germany).
V. cholerae strain VC20 (El Tor, Ogawa), 569B (classical, Inaba)
and Escherichia coli DH5-   strains were used as positive and
negative controls. Amplified products were electrophoresed in a
2% agarose gel (SRL, Bombay, India) in 1x TAE buffer along with
standard molecular weight markers (MWM), stained with ethidium
bromide (Sigma), and recorded with a video documentation system
(Pharmacia Biotech, San Francisco, CA).

Pulsed-Field Gel Electrophoresis (PFGE)

Genomic DNA of various strains of V. cholerae was prepared in
agarose plugs (17). For complete digestion of the DNA, 50U of
NotI enzyme was used. PFGE of inserts was done by the contour-
clamped homogeneous electric field method on a CHEF-mapper
(Bio-Rad) with 0.5x TBE Buffer [44.5 mM Tris, 44.5 mM boric
acid, 1 mM EDTA (pH 8.0)] for 40.24 hours. A DNA size standard
( -ladder; Bio-Rad) was used as molecular weight standard. A
model 1,000 minichiller (Bio-Rad) was used to maintain the
temperature of buffer at 14°C. Run conditions were generated by
the auto-algorithm mode of a CHEF Mapper PFGE system with a
size range of 20-300 kb. Gels were stained in distilled water
containing 1.0 ˜g ethidium bromide per ml for 30 minutes, rinsed
several times, and photographed. After electrophoresis, ethidium
bromide staining and photography, transfer of DNA from gel to
Hybond N + membrane (Amersham International PLC,
Buckinghamshire, England) and Southern blot hybridization with
an O139-specific probe were done as described (17). The DNA
adjacent to Tn5lac insertion in MO10 (an O139 strain from the
Madras outbreak) rendered the strain O139 negative in
agglutinability with O139-specific antiserum. The O139 probe
used in this study was prepared as described earlier (18).

Restriction Fragment Length Polymorphism (RFLP) of ctxA Gene
and O139 Specific Gene

A ctxA probe consisting of a 540-bp XbaI-ClaI fragment of ctxA
cloned in pKTN901 with EcoRI 

linkers was used (19). A modification of the method of Murray and
Thompson (20) was used for DNA extraction for ctx RFLP (16).
The transfer of DNA from gel to Hybond N + membrane
(Amersham) and hybridization with probes was done (16) with the
ECL Nucleic Acid Detection System (Amersham). The membranes
were then washed and exposed to Kodak Biomax film (Eastman
Kodak Co., Rochester, NY) and developed according to
manufacturer's instructions.

Figure 1. Monthly isolation profile of the Vibrio cholerae O1 and
O139 serogroups from patients hospitalized with acute secretory
diarrhea at the Infectious Diseases Hospital, Calcutta, India, from
March 1992 to December 1998. Arrows denote the month in which
changes in V. cholerae strains were noted: A, appearance of V.
cholerae O139 in Calcutta in November 1992; B, displacement of
V. cholerae O1 by V. cholerae O139 in January 1993; C,
reappearance of V. cholerae O1; D, domination of V. cholerae O1
over V. cholerae O139; E, isolation of nontoxigenic V. cholerae
O139 in April 1995; F, appearance for the first time of
cotrimoxazole-susceptible strains of V. cholerae O139; G,
appearance of cotrimoxazole-susceptible and neomycin-resistant
strains of V. cholerae O139; H, domination of V. cholerae O139
over V. cholerae O1; I, replacement of V. cholerae O139 by V.
cholerae O1 as the major cholera causing serogroup; J, appearance
of cotrimoxazole- and streptomycin-susceptible strains of V.
cholerae O139. % of V. chollerae isolations

Figure 2. Monthly isolation profile of the percentage of V. cholerae
O1 and O139 serogroups from patients hospitalized with acute
secretory diarrhea at the Infectious Diseases Hospital, Calcutta,
India, from March 1992 to December 1998.


Results

The O139 serogroup dominated as the major cholera-causing
serogroup in 1993 and from September 1996 to September 1997,
while the O1 serogroup dominated during the remaining part of the
study period (Figures 1, 2). The rate of isolation of O139 serogroup
was lower in 1998 than in any other year since 1992, the year it
was identified (Table 2). The seasonality in incidence of the O139
serogroup showed an interesting trend: apart from 1993 and 1994,
more O139 cholera cases were recorded during the latter half of the
year, coinciding with the second peak of cholera cases (Figures 1,
2). However, the peak incidence of cholera caused by the O1
serogroup was generally observed during June-July, except for
1993 and 1997, when the O1 peak occurred during October and
September, respectively (Figures 1, 2). The O139 strains were the
dominant cholera-causing serogroup during these years (Table 2;
Figures 1, 2).

The antibiotic resistance patterns of randomly selected O139
strains isolated from Calcutta during 1992-1998 (Figure 3) showed
dramatic changes in resistance to cotrimoxazole, with all the 1992,
1993, and 1994 O139 strains being resistant, while most of the
strains isolated during 1997 and 1998 were sensitive. Likewise,
resistance to neomycin and streptomycin also increased until 1996,
then declined. O139 strains isolated throughout the study were
resistant to furazolidone and most were resistant to ampicillin. The
dominant drug-resistance patterns among O139 strains isolated in
Calcutta from 1993 to 1998 were A C Co Fz S (30%), A C Co Fz S
(48.1%), A Co Fz S (49.1%), A Fz N S (96%), A Fz N S (32%),
and A Fz (98.3%). PCR studies with 106 representative O139
strains showed that all but one strain (CO853). were positive for
the 471-bp tcpA (El Tor) amplicon, indicating that almost all have
the organelle required for intestinal colonization. The presence of
301-bp ctxA, 843-bp zot, and 282- bp ace amplicons in all strains
except CO853 indicates that all the tcpA-positive strains have an
intact CTX prophage.

Figure 3. Antibiotic resistance pattern of the V. cholerae O139
strains. Abbreviations: A, ampicillin; C, chloramphenicol; Cf,
ciprofloxacin; Co, cotrimoxazole; Fz, furazolidone; G, gentamycin;
N, neomycin; Na, nalidixic acid; Nx, norfloxacin; S, streptomycin;
T, tetracycline. % resistance

The PFGE profile of three randomly selected O139 strains, sharing
the unique CTX ImmCalcutta prophage and isolated from Calcutta
during 1996 and 1997, has a banding pattern similar to that of the
reference strain MO45 (ATCC 51394) isolated during 1992 in
Madras (Figure 4A). The band patterns exhibited by all the O139
strains differed from that of the classical and El Tor biotype
representative strains of V. cholerae O1, 569B, 2164-88, and
CO840. However, the pattern of the reference O139 strain MO45
was identical to that of the O1 strain MO1, which was described as
the progenitor strain of the O139 serogroup (21). The Southern blot
of the PFGE DNA fragments with O139-specific probe showed
that the probe hybridized at the same position in all the O139
strains. However, the control O1 strains and the progenitor strain
MO1 did not hybridize with the probe (Figure 4B).

RFLP was then conducted to determine any differences in number,
location, and arrangement of the CTX prophage in the genome of
O139 strains. As the CTX prophage of the 1992-93 and 1996 O139
strains is well characterized (8,11,16,22), we took only one
representative strain from 1992 (SG24) and 1996 (AM258)
(8,11,22). Five O139 strains (PG265, PG269, PG316, PG337, and
PG351) isolated during September 1998 (when an upsurge in O139
strains was observed) were included in the study. RFLP of the
CTX prophage with HindIII showed one band with SG24, PG265,
PG269, PG316, and PG351; AM258 and PG337 had three bands of
14, 9.2, and 7.0 kb (Table 3). As HindIII has no site in the CTX
prophage (23), four of the five 1998 O139 strains (PG265, PG269,
PG316, and PG351), like the 1992 O139 strain (SG24), have the
CTX prophage at only one site in the chromosome. RFLP with PstI
and BglII showed only one band each of 5.6 kb and 8 kb,
respectively, in four of the 1998 O139 strains (PG265, PG269,
PG316, and PG351), while the 1992 O139 strain (SG24) exhibited
two bands of sizes 7.0 and 5.6 and 8.0 and 7.0 kb, respectively
(Table 3). As PstI and BglII have only a single site in the CTX
prophage but not within the ctxA gene, these results indicate that
most of the 1998 O139 strains have only one CTX prophage,
unlike the 1992 O139 strain, which has two CTX prophages
arranged in tandem (Figure 5). Further, as both the 8- and 5.6-kb
bands are shared by the 1992 and 1998 V. cholerae O139 strains,
the downstream regions of the CTX prophage in the 1992 and 1998
V. cholerae O139 strains are 97% similar, suggesting that they may
share the identical location in the V. cholerae genome. Southern
blot of HindIII-digested genomic DNA, using a ctxA probe of the
1996 O139 strain (AM258), showed three bands, and the Southern
blots of the PstI-and BglII-digested genomic DNA, also using the
ctxA probe, showed only two bands. As discussed (8,16), the strain
has the same organization of the CTX prophage as the 1996-97
O139 strains isolated from Calcutta. It has three tandemly
duplicated CTX prophages, with the second and third CTX
prophages being new and differing from the first in having an
altered restriction endonuclease site (8,9) (Figure 5). The Southern
blots of the 1998 O139 strain PG337 exactly match those of the
1996-97 representative strain AM258 (Table 3), indicating that its
structure resembles that of the 1996-O139 strains.

Figure 5. Schematic diagram of CTX prophage (not to scale) as
deduced from Southern blot hybridization data. Arrows and boxes
correspond to the restriction sites and core regions of CTX
prophage, respectively, with hatched portions in the box
representing ctxAB genes. Restriction site abbreviations: A, AvaI;
BI, BglI; B, BglII; H, HindIII; P, PstI.

Table 2. Isolation rates of Vibrio cholerae O139 from patients with
acute secretory diarrhea admitted to the Infectious Diseases
Hospital, Calcutta, 1992–1998

Figure 4. Pulsed-field gel electrophoresis profiles of the V.
cholerae O139 strains (A) and Southern blot of the same gel with
O139-specific probe. (B). The numbers at the top indicate the
strain numbers. Lanes 1, MWM (bacteriophage  ladder); 2, 569B
(O1 classical Inaba); 3, 2164-88 (O1 El Tor Ogawa); 4, CO870
(O1 El Tor Ogawa); 5, MO1 (progenitor of the O139 strain); 6,
MO45 (O139 strain isolated during 1992-93); 7, AM231 (O139
strain isolated during 1996-97); 8, AM233 (O139 strain isolated
during 1996-97); 9, AS258 (O139 strain isolated during 1996-97);
10, MWM (bacteriophage   ladder). 

Table 3. Fragment sizes of restriction enzyme-digested genomic
DNA of Vibrio cholerae O139 strains, isolated by ctxA probe,
Calcutta, 1992-1998

Southern blot analysis of HindIII-digested genomic DNA of the 15
O139 strains isolated from different parts of India in 1998 showed
four variations when probed with ctxA. Of the 15 strains
examined, eight (designated group I strains) showed three bands of
size 14, 9.2, and 7.0 kb (the pattern displayed by the 1996-97 O139
strains); two strains had three fragments of sizes 14, 12, and 7.0 kb
(designated group II strains); and four had two fragments of sizes
23 and 7.0 kb (designated group III strains); only one had two
bands of 14 and 7.0 kb (designated group IV strains) (Table 4). All
group I and group II strains showed a single band of about 23 kb in
the blot when digested with BglII and probed with ctxA (Table 4),
indicating that in these strains the BglII site resides outside the
CTX prophage, as described (8). Thus, group I and group II strains
have the same unique organization of the CTX prophage as that of
the 1996-97 O139 strains (8). The group III and group IV strains
also showed a single band when digested with BglII and probed
with ctxA, indicating that in these strains the BglII site may also be
outside the CTX prophage (Table 4). Representative strains of the
four types showed a common band of approximately 7 kb when
digested with BglI, AvaI, and PstI and probed with the ctxA probe
(Table 4). Since all these enzymes have a single restriction site in
the CTX prophage but not within the ctxA gene (23) and the size
of an intact CTX prophage is approximately 7 kb, the 7-kb band in
all these enzymes indicates a tandem duplication in all strains. All
the group I strains have three CTX prophages in tandem (8,16),
with the second and third prophages having a HindIII site instead
of the BglII site in the RS region of the CTX prophage, like the
1996-97 O139 strains (Figure 5). The group II strains differ from
the group I strains by the position of the second band in the blots
(Table 4). Thus, the group II strains have exactly the same
arrangement of the CTX prophage in their genome as the group I
strains; the difference in the size of the second band could be due
to the integration of the prophage at a different site in the genome
(Figure 5). The group III and group IV strains showed two bands in
HindIII, BglI, and AvaI blots when probed with ctxA. There is no
HindIII restriction site in the CTX prophage of El Tor and classical
strains (23), while AvaI, BglII and BglI have only one site within
the CTX prophage (23). In the CTX ImmCalcutta prophage,
present in the 1996-97 O139 strains, the positions of the HindIII
and the BglII enzymes have been interchanged (8,9). The presence
of two bands in the blots with a common 7.0 kb in HindIII, AvaI,
and BglI blots with ctxA and one band in the BglII blot (Table 4)
with same probe indicates that group III and group IV have two
copies of the CTX prophage in tandem, with both the CTX
prophages in these strains being the new type of the O139 CTX
prophage, CTX ImmCalcutta prophage (Figure 5). Thus, these
O139 strains (all isolated in southern India) may derive from the
1996-97 O139 strains but lack the El Tor CTX prophage of the
three CTX prophages that they have in tandem. Alternatively, these
group III and group IV strains may be a new type of O139 that
lacks the El Tor CTX prophage. The only difference between the
group III and group IV strains is in the HindIII blot, which may be
due to polymorphism in the HindIII site (Table 4; Figure 5).

Table 4. Restriction enzyme-digested fragments of genomic DNA
of Vibrio cholerae O139 strains isolated with the ctxA probe, India,
1998

Discussion

The emergence of V. cholerae O139 is a puzzling event in the
history of cholera. The sudden appearance of the O139 serogroup
in late 1992, its rapid spread through Southeast Asia in 1993
followed by a quiescent period during 1994- 95 and its subsequent
emergence in 1996 and 1997 are inadequately understood, but
characterize the unpredictable nature of the epidemiology of
cholera. However, the reemergence of O139 in 1996 indicates that
its appearance was not a one-time event and the serogroup has the
potential to persist and spread to other continents. Comparison of
the O139 strains isolated during the 7-year study period revealed
interesting patterns of antibiotic resistance to various common
antibiotics. While the strains remained largely susceptible to
ciprofloxacin, tetracycline, and gentamycin, resistance to
ampicillin and susceptibility to cotrimoxazole (sulphamethoxazole,
trimethoprim), chloramphenicol, and streptomycin increased
during the period of the study. Resistance to cotrimoxazole
(sulphamethoxazole and trimethoprim) and streptomycin is
encoded by a 62-kb self- transmissible chromosomally integrated
transposon termed the SXT element, which is not apparently linked
to genes encoding O139 antigen (24). The SXT element was
present in 1992-93 O139 strains and is probably silent or absent in
1998 O139 strains, which were sensitive to cotrimoxazole and
streptomycin. The presence of this novel region in 1996 O139
strains is difficult to predict because these strains are susceptible to
cotrimoxazole (sulphamethoxazole and trimethoprim) but resistant
to streptomycin. However, this pattern of rapid shift is consistent
with recent reports indicating an enhanced mobility in genetic
elements, which confers resistance to antibiotics (25) and has also
been observed in V. cholerae O1 strains (7). A multiple-antibiotic
resistance plasmid belonging to incompatibility group C has been
associated with drug-resistance plasmid of V. cholerae O139 (17).
Since the multidrug-resistance plasmid is self-transmissible by
conjugation, the incidence of plasmid-carrying strains and hence
drug resistance may change, depending on the presence of these
plasmids in O139 strains. Molecular studies show continuous
change in the structure and organization of CTX prophage during
the study period, along with evolution of a new type of CTX
prophage. The 1992-93 strains have two CTX prophages connected
by an RS1 element, while the 1996 O139 strains have three CTX
prophages arranged in tandem (8,22). Most of the 1998 O139
strains from Calcutta have only one CTX prophage, while those
isolated from other parts of India have either the arrangement of
the 1996-97 O139 strains (group I and II strains) or have two CTX
prophages arranged in tandem (group III and IV strains). However,
as reported elsewhere (9), the 1996 O139 strains have two types of
CTX prophages, with the first of the three an El Tor type CTX
prophage and the second and third CTX prophages being a new
type of CTX prophage that differs primarily in the rstR gene, the
gene that codes for the repressor protein of CTX . In 1998, we
observed two new clones of O139 at two epicenters, Calcutta and
Alleppey. From the restriction profile data of the 1998 O139
strains, it can be predicted that most of the O139 strains from
Calcutta have only the El Tor type CTX prophage and not the
unique O139 CTX prophage of the 1996 O139 strains, while the
reverse is the case with the south Indian (Alleppey) strains.
Therefore, two different clones of O139 are circulating at two
locations with different types of CTX prophages, indicating that
reassortment in the genome is taking place in the O139 strains. Our
study indicates a continuous emergence of new clones of toxigenic
V. cholerae, possibly through natural selection involving
unidentified factors and immunity of the host population (25-28).
Another possibility is that the genetic reassortments observed here
are random changes in the organism. The strains that gained
advantage as a result of this rearrangement may have infected
humans and become enriched inside the gastrointestinal tract so
that they became detectable as new toxigenic strains. Molecular
analysis of V. cholerae strains isolated during epidemics from 1961
to 1996 in Bangladesh revealed similar clonal diversity among
strains isolated during different epidemics (25-29). These studies
demonstrated three different ribotypes among the V. cholerae O139
isolated from Bangladesh, with different ribotypes often showing
different CTX prophage genotypes (26,28).
This work was supported in part by the Japan International
Cooperation Agency (JICA/NICED Project 054-1061-E-0).

Mr. Basu holds a Masters in Biochemistry and is completing his
doctoral degree in Microbiology under the supervision of G.
Balakrish Nair, Deputy Director of the National Institute of
Cholera and Enteric Diseases, the National Reference Centre for
Cholera in India. Mr. Basu's doctoral thesis is an in-depth
molecular analysis of the CTX genetic element in strains of Vibrio
cholerae isolated in the last 3 decades, including recent isolates of
V. cholerae O139.

Address for correspondence: G. Balakrish Nair, National Institute
of Cholera and Enteric Diseases, P-33, CIT Road, Scheme XM,
Beliaghata, Calcutta, 700 010, India; fax: 91-33- 350-5066; e-mail:
gbnair@vsnl.com.


References
1. Baumann P, Furniss AL, Lee JV. Bergey's manual of systematic
bacteriology. Vol 1. In: Genus 1, Vibrio P. Klieg NR, Holt JG,
editors. Baltimore: Williams and Wilkins; 1984. p. 518-38.
2. Yamai S, Okitsu T, Shimada T, Katsube Y. Distribution of
serogroups of Vibrio cholerae non-O1 non-O139 with specific
reference to their ability to produce cholera toxin and addition of
novel serogroups. Journal of Japanese Association of Infectious
Diseases 1997;71:1037-45.
3. Ramamurthy T, Garg S, Sharma R, Bhattacharya SK, Nair GB,
Shimada T, et al. Emergence of a novel strain of Vibrio cholerae
with epidemic potential in southern and eastern India. Lancet
1993;341:703-4.
4. Nair GB, Ramamurthy T, Bhattacharya SK, Mukhopadhyay AK,
Garg S, Bhattacharya MK, et al. Spread of Vibrio cholerae O139
Bengal in India. J Infect Dis 1994;169:1029-34.
5. Nair GB, Albert MJ, Shimada T, Takeda Y. Vibrio cholerae
O139 Bengal: the new serogroup causing cholera. Medical
Microbiolical Reviews 1996;7:43-51.
6. Sharma C, Nair GB, Mukhopadhyay AK, Bhattacharya SK,
Ghosh RK, Ghosh A. Molecular characterization of V. cholerae O1
biotype El Tor strains isolated between 1992 and 1995 in Calcutta,
India: evidence for the emergence of a new clone of the El Tor
biotype. J Infect Dis 1997;175:1134-41.
7. Mukhopadhyay AK, Garg S, Mitra R, Basu A, Dutta D,
Bhattacharya SK, et al. Temporal shifts in traits of Vibrio cholerae
strains isolated from hospitalized patients in Calcutta: a 3-year
(1993-1995) analysis. J Clin Microbiol 1996;34:2537-43.
8. Sharma C, Maiti S, Mukhopadhyay AK, Basu A, Basu I, Nair
GB, et al. Unique organization of the CTX genetic element in
Vibrio cholerae O139 strains which reemerged in Calcutta, India,
in September, 1996. J Clin Microbiol 1997;35:3348-50.
9. Kimsey H, Nair GB, Ghosh A, Waldor MK. Diverse CTXphis
and evolution of new pathogenic Vibrio cholerae. Lancet
1998;352:457-8.
10. Mitra R, Basu A, Dutta D, Nair GB, Takeda Y. Resurgence of
Vibrio cholerae O139 Bengal with altered antibiogram in Calcutta,
India. Lancet 1996;348:1181.
11. Mukhopadhyay AK, Basu A, Garg P, Bag PK, Ghosh A,
Bhattacharya SK, et al. Molecular epidemiology of reemergent
Vibrio cholerae O139 Bengal in India. J Clin Microbiol
1998;36:2149-52.
12. World Health Organization. Guidelines for cholera control.
Geneva: The Organization; 1993.
13. Yamamoto T, Nair GB, Parodi CC, Takeda Y. In vitro
susceptibilities to antimicrobial agents of V. cholerae O1 and
O139. Antimicrob Agents Chemother 1993;39:241-4.
14. Keasler SP, Hall RH. Detecting and biotyping V. cholerae O1
with multiplex polymerase chain reaction. Lancet 1993;341:1661.
15. Colombo MM, Mastrandrea S, Santona A, de Amdrade AP,
Uzzau S, Rabino S. et al. Distribution of the ace, zot, and ctxA
toxin genes in the clinical and environmental Vibrio cholerae. J
Infect Dis 1994;170:750-1.
16. Basu A, Mukhopadhyay AK, Sharma C, Jyot J, Gupta N,
Ghosh A, et al. Heterogeneity in the organization of the CTX
genetic element in strains of Vibrio cholerae O139 Bengal isolated
from Calcutta, India and Dhaka, Bangladesh and its plausible link
to the dissimilar incidence of O139 cholera in the two locales.
Microb Pathog 1998;24:175-83.
17. Yamasaki S, Nair GB, Bhattacharya SK, Yamamoto S,
Kurazono H, Takeda Y. Cryptic appearance of a new clone of
Vibrio cholerae O1 biotype El Tor in Calcutta, India. Microbiol
Immunol 1997;41:1-6.
18. Waldor MK, Mekalanos JJ. Vibrio cholerae O139 specific gene
sequence. Lancet 1994;343:1366.
19. Kaper JB, Morris JG Jr, and Nishibuchi M. DNA probes for
pathogenic Vibrio species. In: Tenover FC, editor. DNA probes for
infectious disease. Boca Raton (FL): CRC Press, Inc.; 1992. p. 65-
77.
20. Murray MG, Thompson WF. Rapid isolation of high molecular
weight plant DNA. Nucleic Acids Res 1980;8:4321-5.
21. Pajni S, Sharma C, Bhasin N, Ghosh A, Ramamurthy T, Nair
GB, et al. Studies on the genesis of Vibrio cholerae O139:
identification of probable progenitor strains. J Med Microbiol
1994;42:20-5.
22. Bhadra RK, Roychoudhury S, Banerjee RK, Kar S, Majumdar
R, Sengupta S, et al. Cholera toxin (CTX) genetic element in
Vibrio cholerae O139. Microbiology 1995;141:1977-83.
23. Mekalanos JJ. Duplication and amplification of cholera toxin
genes in Vibrio cholerae. Cell 1983;35:253-63.
24. Waldor MK, Tschape H, Mekalanos JJ. A new type of
conjugative transposon encodes resistance to sulfamethoxazole,
trimethoprim and streptomycin in Vibrio cholerae O139. J
Bacteriol 1996;178:4157-65.
25. Faruque SM, Alim ARMA, Rahman MM, Siddique AK, Sack
RB, Albert MJ. Clonal relationships assay classical Vibrio cholerae
O1 strains isolated between 1961 and 1992 in Bangladesh. J Clin
Microbiol 1993;31:2513-6.
26. Faruque SM, Ahmed KM, Siddique AK, Zoman K, Alim
ARMA, Albert MJ. Molecular analysis of toxigenic Vibrio
cholerae in Bangladesh studied by numerical analysis of rRNA
gene restriction patterns. J Clin Microbiol 1995;33:2833-8.
27. Faruque SM, Roy SK, Alim ARMA, Siddique AK, Albert MJ.
Molecular epidemiology of toxigenic Vibrio cholerae in
Bangladesh studied by numerical analysis of rRNA gene restriction
patterns. J Clin Microbiol 1995;33:2833-8.
28. Faruque SM, Ahmed KM, Alim ARMA, Qadri F, Siddique
AK, Albert MJ. Emergence of a new clone of toxigenic Vibrio
cholerae O1 biotype El Tor displacing V. cholerae O139 Bengal in
Bangladesh. J Clin Microbiol 1997;35:624-30.
29. Faruque SM, Alim ARMA, Roy SK, Khan F, Nair GB, Sack
RB, et al. Molecular analysis of rRNA and cholera toxin genes
carried by the new epidemic strain of toxigenic Vibrio cholerae
O139 synonym Bengal. J Clin Microbiol 1994;33:1050-3.


Research

Multivariate Markovian Modeling of Tuberculosis: Forecast for the
United States

Sara M. Debanne,* Roger A. Bielefeld,* George M. Cauthen,†
Thomas M. Daniel,* and Douglas Y. Rowland‡
*Case Western Reserve University, Cleveland, Ohio, USA;
†Centers for Disease Control and Prevention, Atlanta, Georgia,
USA; and ‡D.Y. Rowland Associates, Cleveland, Ohio, USA

We have developed a computer-implemented, multivariate Markov
chain model to project tuberculosis (TB) incidence in the United
States from 1980 to 2010 in disaggregated demographic groups.
Uncertainty in model parameters and in the projections is
represented by fuzzy numbers. Projections are made under the
assumption that current TB control measures will remain
unchanged for the projection period. The projections of the model
demonstrate an intermediate increase in national TB incidence
(similar to that which actually occurred) followed by continuing
decline. The rate of decline depends strongly on geographic, racial,
and ethnic characteristics. The model predicts that the rate of
decline in the number of cases among Hispanics will be slower
than among white non-Hispanics and black non-Hispanics—a
prediction supported by the most recent data.

After many years of decline, tuberculosis (TB) cases began to
resurge in the United States in 1984, and continued to climb for a
decade before declining again (1-4). Reemergent TB did not affect
all segments of the population equally. While incidence continued
to decline in some geographic areas, it increased in others, notably
in the greater New York City area, southern Florida, the border
areas of Texas, and urban California. Disadvantaged populations
were disproportionately affected, with peak incidences in blacks
and Hispanics now seen in young adults. TB in foreign-born
persons accounts for nearly 40% of the total annual cases in the
United States (5). Finally, HIV infection has played a large part in
the recent increase in TB incidence. Effective TB control requires
reliable estimates of future secular trends. We have constructed a
population dynamics, mechanistic Markov chain model for the
epidemiology of TB in the United States by separate modeling for
disaggregated groups defined by race, ethnicity, age, and
geography. With this model, we have projected TB incidence for
the U.S. population at large, as well as for racial and ethnic and
geographically defined groups within the population.

The Study

In support of this work, we obtained datasets from the U.S. Bureau
of the Census and the Centers for Disease Control and Prevention
(6-19). The epidemiology of TB in the United States has an
underlying Markovian nature (20-22). Except in unknown or
difficult-to-predict influxes of immigrants, future infections and
cases of TB may be probabilistically estimated by knowing
sufficiently well the present situation of susceptible, infected, and
ill persons; rates of transition from healthy to infected and infected
to ill; patterns of contact; and rates of fertility and death. Indeed, a
Markov chain is completely specified by an initial state and
transition probabilities. To model the future course of TB in the
United States, we used the structure depicted in Figure 1.

The U.S. population was disaggregated into states multivariately
defined by sociodemographic, health, and disease descriptors: age
(single-year

age groups); sex; race (white, black, Asian or Pacific islander,
Native American or Alaskan native, or other race); ethnicity
(Hispanic or non-Hispanic); TB infection or disease status; and
location. One hundred sixty locations were defined: 60
corresponded to the largest standard metropolitan statistical areas
(SMSAs) in the United States, 50 to groupings of smaller SMSAs
in the 50 states, and 50 to groupings of non-SMSA counties in the
50 states. The impact of the HIV epidemic on the TB epidemic was
addressed by dealing with certain location-specific populations in a
way that recognizes that a fraction of them are dually affected. A
state in the Markov chain was completely defined by the joint
specification of all these descriptors.

Figure 1. Tuberculosis model states and possible transitions.
Thicker lines depict transitions considered in sensitivity analysis.

In the absence of longitudinal data, the population database for the
model was constructed by assembling population profiles for each
of the individual years from historical surveys and population
projections (5-17,23). These profiles consisted of the sizes of
population groups defined multivariately by age, race, sex,
ethnicity, and geographic location. The popula tion projections of
the Bureau of the Census incorporate expected immigration
patterns. Transition probabilities for the model were based on rates
cited in the literature to describe how these transitions occur for
individuals. The probability of transition from healthy to infected
has been termed the "annual risk for infection." To estimate this
probability from tuberculin skin test surveys, we used the catalytic
method of Muench (19,24,25). This method relates incidence to
prevalence, given several data points of the latter, under the
assumption of an exponential model, the exponential curve being
called "catalytic" at the time of Muench, who used graphic
methods for estimation. The average number of susceptible persons
infected by each infectious (i.e., sputum-positive) case in a
geographic location in 1 year was termed the "contagious
parameter" by Styblo, who noted the correspondence between the
prevalence rate of the disease and the annual risk for infection
(25,26). Since we assume that new active cases are treated during
their year of occurrence, our use of the contagious parameter is
consistent with that of Styblo.

An important assumption made in the model about the spread of
infection is that each ill person infects susceptible persons entirely
within race ethnicity location-specific subgroups, with the number
of infections determined by the contagious parameter. The age-sex
distribution of newly infected persons in the model was based on
national TB contact investigation data (27). Time trends for
contagious parameters for the various race ethnicity location-
specific subgroups were based on quadratic regression fits of the
total cases of TB for those subgroups over time from 1980 through
1993. For illustration, we present in Table 1 baseline values for the
contagious parameter for 17- to 25-year-old white men, by state,
derived from the U.S. naval recruits data (28-35).  The probability
of transition from infected to ill is a product of host-parasite
interactions and is determined by the biologic features of these
interactions. A person with a new Mycobacterium tuberculosis
infection (i.e., an infection of <1 year) is at much higher risk for
clinical disease than someone with a mature infection, who has
been at an ongoing risk in later years (36). These risks vary with
age. Although not explicitly modeled, reinfection of cured persons
is allowed at the same rate as infections with the same
characteristics in the population. The case rates among persons
with new and mature infections in the next year are based on the
upper and lower bounds of rates reported in earlier studies, and are
adjusted to calibrate the model to the historical time series (27,37-
42) (Table 2). The effects of the HIV epidemic and the emergence
of multidrug-resistant TB are modeled through time-dependent
exogenous inputs for appropriate geographic locations and, within
these locations, for specific demographic subgroups, as dictated by
trends in prevalence and dual-infection rates (43-45). These
calculations result in time-dependent changes to the probabilities
of transition from new or mature infection to clinical TB.

Table 1. Baseline values of contagious parameters, for 17- to 25-
year-old white men, by state

We established 1980 as the baseline year for all variables used in
the model. The initial values of these variables were based on the
Reports of Verified Cases of Tuberculosis (RVCT) files and
published reports of TB incidence from Centers for Disease
Control and Prevention. Because the RVCT program was not
implemented nationwide until 1985, reporting for the years 1980 to
1984 was incomplete. To estimate the 1980 baseline values, we
inferred values of missing data according to the multivariate
structure observed in 1985. Baseline information also included
counts of mature infections prevalent in 1980 and counts of new
infections in 1980. These values were estimated by applying
difference equations relating the numbers of cases, new infections,
and mature infections in a given year to the numbers new
infections, and mature infections in the previous year, where the
relationships between the figures for the 2 years were dependent on
model parameters such as survival rates, rates of transition from
mature infection to disease, rates of transition from new infection
to disease, and contagious parameters, and where each difference
equation pertained to a specific multivariately defined
subpopulation. Through repeated application of the difference
equations, it was possible to express the number of mature
infections in 1980 in terms of TB case counts for an arbitrary
number of subsequent years for which actual data were available.
Estimates thus obtained were examined for consistency and the
most stable was selected. Because of the lack of reliable
information, we did not attempt to address the underreporting of
TB cases or underestimates of some populations.

Table 2. Literature-based lower, best, and upper estimates of the
risk of developing tuberculosis for infected persons, by age and sex

Figure 2. Total number of new cases of tuberculosis in white non-
Hispanics, blacks, and Hispanics in the United States, 1980–2010.

Computer implementation of the model was carried out on a Sun
SPARCstation 5 with 80 megabytes of main memory and network
access to a SPARCserver 1000E with approximately 100 gigabytes
of online disk storage. All software was written in C++, an object-
oriented programming language. A library of C++ objects and
functions (46,47) was used as the basis for the representation of
fuzzy numbers in the software. Because of the extremely large
sizes of the multidimensional arrays containing cross-sectional
population data and year-to-year transition data, specialized data
structures and algorithms were developed to take advantage of
characteristics of these data that allowed them to be stored in a
more compact form (48). For cross-sectional population data, a
height-balanced binary search tree structure (49,50) was used.
Within each tree node, both cell information and array subscript
information were stored. Because the subscripts would occupy at
least 9 bytes within each node of the tree if stored conventionally, a
4-byte compressed representation of the subscripts was stored
within each node. For transition data, a simple linked-list structure
was used in which each node contained both cell information and
compressed subscript information. This approach resulted in a 96%
reduction of storage space required for population data and a 99%
reduction of storage space required for transition data (48).

Results

Figure 2 presents model projections of new cases of TB for the
United States for 1980 through 2010 as well as actual data for
white non-Hispanics, blacks, and Hispanics, the groupings used in
TB case reporting in the United States. For these three groups,
differences between actual counts and model predictions were
generally <1,000 for blacks and Hispanics and <1,500 for white
non-Hispanics. Table 3 presents actual and projected TB case rates
per 100,000 for each 5 years from 1980 to 2010 with results
presented for racial, ethnic, and geographic groups. Generally, the
model performs best, within the 1980 to 1995 validation period, for
large subgroups with relatively high case rates. For example, it
projects TB case rates to within 12% of actual rates for the U.S.
population at large. Its performance is not as good in subgroups
having lower case rates (e.g., there is a 23% projection error for
non-Hispanic whites in 1995). In addition, projection errors are
relatively high for smaller subgroups, especially those in which
case reporting and population reporting are subject to inaccuracy
(e.g., projection errors are consistently above 20% for Hispanics
and nearly as high for American Indians and Asian/Pacific
Islanders). The model's best performance tends to be among the
older population: the model predicts case rates to within 1%
accuracy for the U.S. population aged 65 and over for 1990 and
1995. This is of significance in that the highest TB case rates are
among the elderly.

Table 3. Actual and projected tuberculosis rates per 100,000 for
selected population groups, 1980 to 2010

Figure 4. Total number of new cases of tuberculosis in the United
States, 1980–2010 and sensitivity of model projections.


Figure 3 presents model projections of the TB case rate quartiles,
by state, for the year 2010. Note that the projected case rate for
Wyoming spuriously falls into the highest quartile, as a result of its
small population size. TB is projected to decrease in all parts of the
United States and remain largely a problem of states receiving
large numbers of immigrants.

In Figure 4, the model projection for the United States for 1980
through 2010 is presented together with actual numbers of newly
reported TB cases for 1980 through 1997. Also depicted is the
sensitivity of the model to changes in key parameters, seen in
thicker lines in Figure 1. Specifically, changes were allowed for the
contagious parameters and for rates of transition from healthy to
infected and infected to ill. Specifically, above and below the
central projection are lines representing projections from
simultaneous 10% increases and decreases in these key parameters
for each year in the projection period. A simultaneous decrease of
10% in the parameters for each year of the projection period results
in a decrease of 27% (from 13,458 to 9,861) in projected TB cases
in 2010. A simultaneous increase of 10% yields an increase of 44%
(from 13,458 to 19,381) in projected cases in 2010. Not shown in
the figure, a simultaneous decrease of 5% in the same parameters
yields a decrease of 15% (from 13,458 to 11,459) for projected
cases in 2010, and a simultaneous increase of 5% yields an
increase of 19% (from 13,458 to 16,063) in 2010.

Conclusions

The studies of Waaler and colleagues, who carefully noted and
credited other early workers in this field, mark the beginning of a
modern approach to modeling and the thoughtful consideration of
transition probabilities in modeling (51,52). ReVelle, Lynn, and
Feldman, using many of the assumptions of Waaler and his
colleagues, projected declining TB case rates in developing
countries as a result of BCG vaccination (53). Azuma modeled TB
incidence in Japan (54). Estimates of AIDS prevalence were used
by Styblo (55), Schulzer and colleagues (56), and Heymann (57) to
project TB incidence in sub-Saharan Africa. Modeling was used by
Murray, Styblo, and Rouillon to examine the economic impact of
TB in developing countries (58). Blower and colleagues used a
differential equation-based model to examine the decline of TB
epidemics and emphasize the importance of effective case
treatment in TB control (59,60). Murray and colleagues used
differential equation-based models to compare the global impact of
different control strategies (61,62). Brewer and colleagues used
simulation techniques to compare TB control policies (63).

Mathematical modeling of TB in the United States probably began
in 1939 with the early work of Frost, who used age cohort analyses
of secular trends to predict the future decline of TB (64). Ferebee
developed a model for TB in the United States based on data from
1963 and used it to emphasize the potentially important
contribution of chemoprophylaxis to TB control (65). The model
we present is substantially more complex than Ferebee's and
differs from her early work by disaggregating the American
population. Given the marked heterogeneity of the population of
the United States and the corresponding diversity of TB case rates
in various American population groups, our approach offers major
advantages over aggregate modeling both in improved accuracy of
estimation of input data and in output that can be disaggregated to
relate to segments of the population with very different TB control
problems.

A highly disaggregated model incorporating population dynamics
is appropriate. When applied to small groups, population dynamics
are aptly described by a linear model, particularly when natural
constraints are stated as totals. The model we constructed is
Markovian in that it is independent of its history prior to its last
known state. Our computer implementation of the theoretical
model is clearly Markovian because its operation parallels the
nature of a Markov chain, in the sense that the initial state was the
state of our system in 1980, and the state in each subsequent year
was generated from the previous year's state and 1-year transition
rates. Since no information other than the previous year's state and
1-year transition rates are available to the computer software to
generate the state for a year, the Markovian nature is established.
Our model for the intermediate years predicted (on the basis of
1980 baseline data) that TB would increase late in the ensuing
decade and decline again. This prediction reasonably approximated
the subsequent events. While the model does include changes in
TB incidence related to the HIV epidemic and to immigration, it
does not specifically incorporate changes in TB control measures,
whose impacts are only addressed indirectly through net changes of
transition probabilities in the model. The latter could be addressed
by a submodel for appropriate population subgroups, and could
affect projections. The projections presented here suggest that the
number of new cases and incidence rates of TB will now continue
to decrease to a level approximately half the present level
nationally. Measuring the extent to which increased control efforts
contributed to the return of the TB epidemic to its formerly
declining phase will require more detailed analysis.

The handling of geographic location necessarily relies on
politically defined entities which determine how data are collected.
Since disease patterns do not necessarily follow these boundaries,
the model is not uniformly accurate by geographic location. This
may be seen in Table 3 where the TB case rate in 1995 is
overestimated for Los Angeles and underestimated for Miami.
However, the very fine age structure imposed has resulted in
projections by age group within 20% accuracy for age 25 and older
(Table 3). Refinement of the geographic characterization could
result in corresponding increases in accuracy. The model
projections are smoother over time than the actual counts, as
should be expected because random fluctuations are not built into
the model (Figures 2,3). In Figure 4 the time series of actual cases
is moving toward the lower 10% bound. This may be a temporary
fluctuation (as in the mid-1980s) but could also indicate changes in
transitions in the real-world disease process. Errors of a projection
model may arise from shortcomings of model structure, from
inaccuracies in initial conditions, or from errors in the values of
key model parameters. The structure of the Markov chain model is
mechanistic, driven by highly detailed population dynamics. Each
part of the mechanistic process may prompt suggestions for model
refinements and discussions of variability; notably, occurrence of
infection (66,67), time from infection to disease (68,69), and
involvement of HIV (70) and immigrant status (71). The greatest
potential for model inaccuracy due to initial conditions arises from
estimates of the 1980 pool of prevalent infections. Errors thus
introduced become less consequential with passing time, as
persons infected before 1980 die. So, model projections in the
latter part of the projection period are least affected by errors in the
numbers of those infected before 1980. Model parameters not only
have the obvious role of governing transitions for various
population subgroups, but also, as reflected by their time
dependency, are used to depict exogenous phenomena such as the
HIV epidemic, implementation of TB control efforts, and changes
in immigration patterns. Geographically focused changes of TB
control efforts implemented after 1993 related to directly observed
therapy could result in overestimation in projections in those areas.

Although the model does not allow mixing of groups from one
location to another, this limitation is in large measure adjusted in
the calculations of transition probabilities since these are based on
actual data that implicitly reflect the movements of healthy,
infected, and ill groups. The other type of mixing not accounted for
in the model occurs in a location among different race-ethnicity
subgroups. The most serious projection errors would then occur
when a subgroup of uninfected persons lives in the same location
as a subgroup of persons with active TB, for then any cases arising
in the former subgroup would not be predicted by our model. This
type of situation is likely to arise only in locations with very few
active TB cases. Otherwise, the projection errors arise only from
the extent to which the cross-spread of infection fails to cancel out
(i.e., the net extent to which a subgroup infects others, over and
above the number of the latter being infected by the former).

Although HIV status is addressed in this model, adding structure
that more precisely reflects changes in transition rates of HIV-
infected persons might also improve model projections. The model
predicts that an increasing portion of cases will occur in Hispanics,
and this prediction is supported by the most recent data, which
show that the rate of decline in the number of cases in Hispanics is
slower than for white non-Hispanics and for black non-Hispanics.

The foreign born account for a large fraction of the new cases seen
in the United States (71). The model assumptions regarding the
future contribution of the foreign born to the U.S. population are
from the U.S. Bureau of the Census projections; therefore, this
component of the population may be thought of as forming part of
the backdrop against which the model projections are made.
Departure from the patterns of immigration projected by the
Census Bureau could cause two kinds of errors in model
projections. A surge of immigration from countries with high rates
of TB would soon be followed by increases in TB cases and case
rates in the United States that would not be projected by the model.
A different effect would be seen as a consequence of unexpected
large increases in immigrants. Regardless of the TB status of the
countries of origin, an increase in the number of immigrants would
be reflected first in an artificial lowering of case rates because of
the increase in denominators, followed by a longer-term slow
increase in numbers of cases. The model could not recognize these
changes within its current structure. These potential errors may be
avoided as additional structure better reflecting immigrant status
becomes incorporated into the model so that unexpected influxes
may be modeled as exogenous inputs.

Acknowledgment

The authors thank Philip J. Smith for helpful comments.

This work was supported in part by a contract with the Centers for
Disease Control and Prevention under its cooperative agreement
with the Association of Teachers of Preventive Medicine.

Dr. Debanne is an associate professor of epidemiology and
biostatistics at the School of Medicine, Case Western Reserve
University. She specializes in the construction of mathematical
models of infectious diseases and has worked extensively in the
area of tuberculosis.

Address for correspondence: Sara M. Debanne, Department of
Epidemiology and Biostatistics, School of Medicine, Case Western
Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106-
4945, USA; fax: 216-368-3970; e-mail: smd@hal.cwru.edu.

References

1. Bloch AB, Rieder HL, Kelly GD, Cauthen GM, Hayden CH,
Snider DE Jr. The epidemiology of tuberculosis in the United
States. Clin Chest Med 1989;10:297-313.
2. Ellner JJ, Hinman AR, Dooley SW, Fischl MA, Sepkowitz KA,
Goldberger MJ, et al. Tuberculosis symposium: emerging
problems and promise. J Infect Dis 1993;168:537-51.
3. Cantwell MF, Snider DE Jr, Cauthen GM, Onorato IM.
Epidemiology of tuberculosis in the United States, 1985 through
1992. JAMA 1994;272:535-9.
4. Comstock GW. Variability of tuberculosis trends in a time of
resurgence. Clin Infect Dis 1994;19:1015-22.
5. Centers for Disease Control and Prevention. Tuberculosis in the
United States, 1997. Atlanta: July 1998:10-11.
6. Census of Population and Housing, 1990: Public use microdata
sample U.S. technical documentation and machine-readable data
file/prepared by the Bureau of the Census. Washington: The
Bureau; 1992.
7. 1990 Census of Population and Housing, 5% and 1% public use
microdata samples-U.S. geographic equivalency files machine-
readable data files / prepared by the Bureau of the Census.
Washington: The Bureau; 1994.
8. Census of Population and Housing, 1980; summary tape file 3A
machine-readable data file / prepared by the Bureau of the Census.
Washington: The Bureau; 1982.
9. Census of Population and Housing, 1980; summary tape file 3B
machine-readable data file / prepared by the Bureau of the Census.
Washington: The Bureau; 1982.
10. Census of Population and Housing, 1980; summary tape file
3C machine-readable data file / prepared by the Bureau of the
Census. Washington: The Bureau; 1982.
11. Census of Population and Housing, 1980; summary tape file 3
technical documentation / prepared by the Data User Services
Division, Bureau of the Census. Washington: The Bureau; 1982.
12. Census of Population and Housing, 1990: summary tape file 3
on CD-ROM machine-readable data files / prepared by the Bureau
of the Census. Washington: The Bureau; 1992.
13. Intercensal estimates of the population of counties by age, sex,
and race: 1980-1990 machine-readable data files / prepared by the
Bureau of the Census. Washington: The Bureau; 1994.
14. Population projections for states, by age, sex, race, and
hispanic origin: 1993 to 2020 machine-readable data file / prepared
by the Bureau of the Census Population Division. Washington: The
Bureau; 1994.
15. U.S. Bureau of the Census. Current population reports, P25-
1111, population projections for states by age, sex, race, and
hispanic origin: 1993 to 2020, by Paul R. Campbell. Washington:
Government Printing Office; 1994.
16. Tuberculosis in the United States, for years 1980 through 1997,
Atlanta (GA): Centers for Disease Control and Prevention,
Division of Tuberculosis Control, 1982-1997.
17. AIDS Public Information Data Set, Statistics and Data
Management Branch, Division of HIV/AIDS, CDC, 1993, Atlanta,
GA.
18. Reports of Verified Cases of Tuberculosis (RVCT), CDC,
Surveillance System for TB (SURVS-TB), Division of
Tuberculosis Control, 1980-1993, Atlanta, GA.
19. Daniel TM, Debanne SM. Estimation of the annual risk of
tuberculous infection for white men in the United States. J Infect
Dis 1997;175:1535-7.
20. Markov AA. Investigation of a noteworthy case of dependent
trials. Izv Akad Nauk Ser Biol 1907;1.
21. Markov AA. The calculus of probabilities. 4th ed. Moscow:
GIZ; 1924.
22. Takács LF. Stochastic processes. London: Methuen, 1960.
23. Liaw K-L, Frey WH. Destination choices of the 1985- 90
young adult immigrants to the United States: importance of race,
educational attainment and labor market forces. International
Journal of Population Geography 1998;4:49-61.
24. Muench H. Derivation of rates from summation data by the
catalytique curve. Journal of the American Statistical Association
1934;29:25-38.
25. Styblo K. The relationship between the risk of tuberculous
infection and the risk of developing infectious tuberculosis.
Bulletin of the International Union against Tuberculosis
1985;60:117-9.
26. Styblo K. Epidemiology of tuberculosis. Selected papers. Vol
24. The Hague (The Netherlands): Royal Netherlands Tuberculosis
Association (KNCV); 1991. p. 49-50.
27. Tuberculosis Program Management in the United States, 1984.
Atlanta (GA): Centers for Disease Control and Prevention, Center
for Prevention Services, Division of Tuberculosis Control; 1986.
28. Edwards LB, Acquaviva FA, Livesay VT, Cross FW, Palmer
CE. An atlas of sensitivity to tuberculin, PPD-B, and histoplasmin
in the United States. Am Rev Respir Dis 1969;99 Suppl:1-132.
29. Edwards LB, Smith DT. Community-wide tuberculin testing in
Pamlico County, North Carolina. Am Rev Respir Dis 1965;92:43-
54.
30. Rosenthal SR, Afremow ML, Nikurs L, Loewinsohn E,
Leppmann M, Katele E, et al. BCG vaccination and tuberculosis in
students of nursing. Am J Nurs 1963;63:88-93.
31. Hanzel GD, Rogers KD. Multiple-puncture and Mantoux
tuberculin tests in high school students. A comparative study.
JAMA 1964;190:1038-42.
32. Rhoades ER, Alexander CP. Reactions to the tuberculin time
test in air force recruits. An analysis of 193,856 tests. Am Rev
Respir Dis 1968;98:837-41.
33. Comstock GW. Frost revisited: the modern epidemiology of
tuberculosis. Am J Epidemiol 1975;101:363-82.
34. Reichman LB, O'Day R. Tuberculous infection in a large urban
population. Am Rev Respir Dis 1978;117:705-12.
35. Engel A, Roberts J. Tuberculin skin test reaction among adults
25-74 years, United States, 1971-1972. DHEW Publication No.
(HRA) 77-1649.
36. Comstock GW, Cauthen GM. In: Reichman LB, Hershfield ES,
editors. Tuberculosis: a comprehensive international approach.
New York: Marcel Dekker, Inc.; 1993. p. 23-48.
37. Zeidberg LD, Gass RS, Dillon A, Hutcheson RH. The
Williamson County tuberculosis study. A twenty-four-year
epidemiologic study. Am Rev Respir Dis 1963;87 Suppl:1-88.
38. Ferebee SH. Controlled chemoprophylaxis trials in
tuberculosis. A general review. Advances in Tuberculosis Research
1970;17:28-106.
39. Devadatta S, Dawson JJY, Fox W, Janardhanam B,
Radhakrishna S, Ramakrishna CVL, et al. Attack rate of
tuberculosis in a 5-year period among close family contacts of
tuberculous patients under domiciliary treatment with isoniazid
plus PAS or isoniazid alone. Bull World Health Organ
1970;42:337-51.
40. Morris SI. Tuberculosis as an occupational hazard during
medical training: report of case-finding and follow-up study, with
effective control program for tuberculosis in women medical
students. American Review of Tuberculosis 1946;54:140-58.
41. Myers JA. Tuberculosis. A half century of study and conquest.
St. Louis (MO): Warren H. Green, Inc.; 1970. p. 166-99.
42. Ferebee SH, Mount FW. Tuberculosis morbidity in a controlled
trial of the prophylactic use of isoniazid among household contacts.
Am Rev Respir Dis 1962;85:490-521.
43. Daley CL, Small PM, Schecter GF, Schoolnik GK, McAdam
RA, Jacobs WR Jr, et al. An outbreak of tuberculosis with
accelerated progression among persons infected with the human
immunodeficiency virus. N Engl J Med 1992;326:231-5.
44. Di Perri G, Cruciani M, Danzi MD, Luzatti R, De Checchi G,
Malena M, et al. Nosocomial epidemic of active tuberculosis
among HIV-infected patients. Lancet 1989;2:1502-4.
45. Pitchenik AE, Burr J, Laufer M, Miller G, Cacciatore R, Bigler
WJ, et al. Outbreaks of drug-resistant tuberculosis in an AIDS
center. Lancet 1990;336:440-1.
46. Bielefeld RA. Fuzzy representation of uncertainty in disease
progression [Ph.D. dissertation]. Department of Epidemiology and
Biostatistics. Cleveland (OH): Case Western Reserve University;
1992.
47. Bielefeld RA. Estimation of HIV infection and AIDS onset
times using fuzzy arithmetic techniques. In: Proceedings of the
Seventh World Congress on Medical Informatics. Amsterdam
(North-Holland); 1992. p. 921-7.
48. Bielefeld RA, Debanne SM, Rowland DY. Storing sparse and
repeated data in multivariate Markovian models of tuberculosis
spread. Computers and Biomedical Research 1996;29:85-92.
49. Adel'son-Vel'skii GM, Landis EM. An algorithm for the
organization of information. Dokl Acad Nauk SSSR Math
1962;146:263-6.
50. Standish TA. Data structure techniques. Reading (MA):
Addison-Wesley; 1980. p. 108-18.
51. Waaler H, Geser A, Anderson S. The use of mathematical
models in the study of the epidemiology of tuberculosis. Am J
Public Health 1962;52:1002-13.
52. Waaler HT, Piot MA. The use of an epidemiologic model for
estimating the effectiveness of tuberculosis control measures:
sensitivity of the effectiveness of tuberculosis control measures to
the coverage of the population. Bull World Health Organ
1969;41:75-93.
53. ReVelle CS, Lynn WR, Feldman F. Mathematical models for
the economic allocation of tuberculosis control activities in
developing countries. Am Rev Respir Dis 1967;96:893-909.
54. Azuma Y. A simple simulation model of tuberculosis
epidemiology for use without large-scale computers. Bull World
Health Organ 1975;52:313-22.
55. Styblo K. The global aspects of tuberculosis and HIV infection.
Bulletin of the International Union against Tuberculosis and Lung
Disease 1990;65:28-32.
56. Schulzer M, Fitzgerald JM, Enarson DA, Grzybowski S. An
estimate of the future size of the tuberculosis problem in sub-
Saharan Africa resulting from HIV infection. Tuber Lung Dis
1992;73:52-8.
57. Heymann SJ. Modeling the efficacy of prophylactic and
curative therapies for preventing the spread of tuberculosis in
Africa. Trans R Soc Trop Med Hyg 1993;87:406-11.
58. Murray CJL, Styblo K, Rouillon A. Tuberculosis in developing
countries: burden, intervention and cost. Bull Internat Union
against Tuberc 1990;65:2-20.
59. Blower SM, McLean AR, Porco TC, Small PM, Hopewell PC,
Sanchez MA, Moss AR. The intrinsic transmission dynamics of
tuberculosis epidemics. Nature Medicine 1995;1:815-21.
60. Blower SM, Small PM, Hopewell PC. Control strategies for
tuberculosis epidemics: New models for old problems. Science
1996;273:497-500.
61. Murray CJL, Salomon JA. Expanding the WHO tuberculosis
control strategy: rethinking the role of active case-finding.
International Journal of Tuberculosis and Lung Disease 1998;2:S9-
15.
62. Murray CJL, Salomon JA. Modeling the impact of global
tuberculosis control strategies. Proc Natl Acad Sci U S A
1998;95:13881-6.
63. Brewer TF, Heymann SJ, Colditz GA, Wilson ME, Auerbach
K, Kane D, Fineberg HV. Evaluation of tuberculosis control
policies using computer simulation.
JAMA 1996;276:1898-903. 64. Frost WH. The age selection of
mortality from tuberculosis in successive decades. American
Journal of Hygiene 1939;30:91-6.
65. Ferebee SH. An epidemiological model of tuberculosis in the
United States. Bulletin of the National Tuberculosis Association
1967 (January):4-7.
66. Rieder HL. Methodological issues in the estimation of the
tuberculosis problem from tuberculin surveys. Tuber Lung Dis
1995;76:114-21.
67. Gammaitoni L, Nucci MC. Using a mathematical model to
evaluate the efficacy of TB control measures. Emerg Infect Dis
1997;3:335-42.
68. MacIntyre CR, Plant AJ. Preventability of incident cases of
tuberculosis in recently exposed contacts. International Journal of
Tuberculosis and Lung Disease 1998;2:56-61.
69. Salpeter EE, Salpeter SR. Mathematical model for the
epidemiology of tuberculosis, with estimates of the reproductive
number and infection-delay function. Am J Epidemiol
1998;142:398-406.
70. West RW, Thompson JR. Modeling the impact of HIV on the
spread of tuberculosis in the United States. Math Biosci
1997;143:35-60.
71. Liu Z, Shilkret KL, Tranotti J, Freund CG, Finelli L. Distinct
trends in tuberculosis morbidity among foreign-born and U.S.-born
persons in New Jersey, 1986 through 1995. Am J Public Health
1998;88:1064-7.


Research

Serologic Response to Culture Filtrate Antigens of Mycobacterium
ulcerans during Buruli Ulcer Disease

Karen M. Dobos,*† Ellen A. Spotts,*† Barbara J. Marston,* C.
Robert Horsburgh Jr.,* and C. Harold King*
*Emory University School of Medicine, Atlanta, Georgia, USA;
and †Centers for Disease Control and Prevention, Atlanta, Georgia,
USA

Buruli ulcer (BU) is an emerging necrotic skin disease caused by
Mycobacterium ulcerans. To assess the potential for a
serodiagnostic test, we measured the humoral immune response of
BU patients to M. ulcerans antigens and compared this response
with delayed-type hypersensitivity responses to both Burulin and
PPD. The delayed-type hypersensitivity response generally
supported the diagnosis of BU, with overall reactivity to Burulin in
28 (71.8%) of 39 patients tested, compared with 3 (14%) of 21
healthy controls. However, this positive skin test response was
observed primarily in patients with healed or active disease, and
rarely in patients with early disease (p=0.009). When tested for a
serologic response to M. ulcerans culture filtrate, 43 (70.5%) of 61
BU patients had antibodies to these antigens, compared with 10
(37.0%) of 27 controls and 4 (30.8%) of 13 tuberculosis patients.
There was no correlation between disease stage and the onset of
this serum antibody response. Our findings suggest that serologic
testing may be useful in the diagnosis and surveillance of BU.

Buruli ulcer (BU) disease, caused by Mycobacterium ulcerans, is
characterized by severe necrotizing ulcers. The disease, which is
found worldwide but primarily in tropical climates (1,2), occurs
predominantly in children ages 5 to 14 (3) and is associated with
severe illness and permanent disabilities in >26% of patients (1,3).
Over the past decade, the incidence of BU disease has dramatically
increased in several West African countries, notably Ghana, Côte
d'Ivoire, Benin, and Togo (1,2). In some West African
communities, BU has replaced tuberculosis (TB) and leprosy as the
most prevalent mycobacterial disease, affecting up to 22% of the
population (4). However, the true global impact of the disease is
unknown because reliable tools for its rapid diagnosis and
surveillance are lacking. In addition, neither the mode of
transmission nor the environmental reservoir for M. ulcerans is
known, although observational studies suggest that M. ulcerans
infection is transmitted through the skin after contact with
contaminated water, vegetation (5), or insects (6).

Current treatment for cutaneous disease is primarily surgical.
Ulcerative lesions require wide debridement and skin grafting in
specialized health-care facilities, long hospital stays (7), and
increased cost to the health-care system and the community (7).
Antibiotics are ineffective once ulcers are present, possibly because
M. ulcerans is absent from lesions or subcutaneous tissues have
already been destroyed, preventing penetration of
antimycobacterial drugs (3). Antimicrobial therapy has not been
studied in preulcerative lesions because apparently patients rarely
visit primary health-care providers with painless nodular lesions
and diagnostic tests are not available to clearly identify M. ulcerans
infection before it progresses to clinical disease. Early
identification and surgical excision of these preulcerative nodular
lesions can be curative and does not require inpatient care (8). We
investigated the serologic response to the M. ulcerans culture
filtrate (MUCF) of BU patients at different stages of disease to
determine whether serodiagnosis is feasible as a tool for early
diagnosis and intervention of BU disease.

Table 1. Burulin skin test responses of Buruli ulcer patients and
healthy controls from the same villages


Persons were scored Burulin-only skin-test positive if the diameter
of induration at the site of Burulin injection was 10 mm at 48 or 72
hours and the diameter of induration at the site of PPD injection
was at least 3 mm smaller than in the Burulin response. Persons
were scored as only PPD positive if the diameter of induration at
the site of PPD injection was 10 mm at 48 or 72 hours and the
diameter of induration at the site of Burulin injection was at least 3
mm smaller than that in the PPD response. Persons were scored as
both Burulin and PPD skin-test positive if the diameters of
induration of both the Burulin and PPD injection sites were 10 mm
at 48 or 72 hours and the differences in the diameters of the zone of
induration between the two was <3 mm. Persons were scored as
skin test negative for both if the induration at both injection sites
was <10 mm at 48 or 72 hours.

Table 2. Serologic and skin test responses of Buruli ulcer patients
and healthy controls from the same villages

The Study

BU patients were identified through a case-control study conducted
in 1991 in the Daloa region of Côte d'Ivoire (3). Patients with early
ulcerative BU were identified by small, painless, or minimally
painful nodules or edema, or ulcerative lesions for <2 weeks.
Patients with ulcerative BU were identified by lesions that were
chronic (>2 weeks duration), were painless or minimally painful,
or had undermined edges. Patients with healed BU were identified
by stellate scars with retraction in areas of prior ulceration.
Descriptive and clinical data were obtained from all BU patients
and controls (3). Blood specimens were collected from 82 BU
patients and 164 healthy controls in the Daloa region.

Because TB is endemic in many of the areas where BU occurs, 13
TB patients from the United States were used as controls for cross-
reactivity to MUCF. Use of sera from these patients ensured the
lowest probability for exposure to M. ulcerans, so that only cross-
reactivity to antigens common to both M. tuberculosis and M.
ulcerans would be measured. TB patients were identified through
an ongoing study in Atlanta by clinical presentation, a positive
PPD response, and isolation of M. tuberculosis from sputum. All
specimens tested negative for HIV and were strongly reactive to
many of the culture filtrate antigens from M. tuberculosis (data not
shown). Informed consent was obtained for all participants, and
protocols were approved by the Institutional Review Boards of the
Centers for Disease Control and Prevention and the Ministry of
Health of Côte d'Ivoire for the BU study and the Institutional
Review Board at Emory University for the TB study.

Skin Testing

During sera collection, Burulin (9) and PPD (purified protein
derivative of M. tuberculosis, Connaught Laboratories, U.K.) were
administered intradermally into the flexor surface of the forearms
of 39 BU patients and 21 controls, and the diameter of induration
was recorded at 48 and 72 hours. The criterion for using Burulin
and PPD dual testing to confirm clinical cases of BU has been
reported (3,9), and the cumulative delayed-type hypersensitivity
(DTH) findings for this study have been described (Table 1) (3).
Sera were also tested for reactivity to MUCF antigen by Western
blot (Table 2).


Preparation of MUCF Antigen for Serologic Analysis

M. ulcerans strain S-WT (from the Centers for Disease Control and
Prevention culture collection; Atlanta) was recovered from a 1-ml
lyophilisate in 10% glycerol by incubation in 5 ml Middlebrook
7H9 with OADC supplement (Remel, Lenexa, KS) at 32.5°C for 7
days. This 5-ml inoculum was passaged into 25 ml Middlebrook
7H9 broth supplemented with tryptose and glucose (7H9TG; 0.1%
wt:vol and 2% wt:vol, respectively) and incubated at 32.5°C for 10
days, followed by passage into 100 ml of 7H9TG and growth at
32.5°C for 10 days. Cultivation in this protein- and serum-free
media allowed isolation and analysis of specific secreted
mycobacterial proteins. The 100-ml culture was passaged into 1L
of 7H9TG, incubated at 32.5°C in 850-cm 2 roller bottles, and
rotated at 90 rotations per hour. Bacilli were harvested during mid-
log phase growth by centrifugation. The 1L of MUCF, containing
mycobacterial proteins of interest, was concentrated at 4°C under
N 2 by using a stirred-cell apparatus (Amicon Inc., Beverly, MA)
with a 10-kDa molecular weight membrane. MUCF was
exchanged into 0.1 M NH 4 HCO 3 by dialysis, lyophilized, and
purified by separation of residual small molecular weight
contaminants from antigens of interest by size exclusion
chromatography. MUCF was loaded onto a fast-performance liquid
chromatography system (FPLC) (Pharmacia, Piscataway, NJ)
equipped with a G-25 superfine Sephadex desalting column
(Pharmacia, Piscataway, NJ), UV monitor, and conductivity
monitor for separation and detection of proteins and salts,
respectively. MUCF was eluted with an isocratic gradient of 0.1 M
NH 4 HCO 3 . MUCF protein was quantitated by bicinchoninic
acid assay (Pierce Chemical Co., Rockford, IL) and lyophilized
until needed. Approximately 5 mg of MUCF protein was obtained
per liter of culture.

Aliquots of 8 g of MUCF protein were resolved by discontinuous
sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-
PAGE) (10) by using 10% to 20% gradient 10-well gels (Novex,
San Diego, CA). MUCF proteins were visualized by staining with
silver (11), and this profile was reexamined throughout these
experiments to ensure that the MUCF proteins did not degrade
upon storage; no antigen variability was seen during our studies.
7H9TG media alone processed in the same way contained no
proteins by silver stain analysis after SDS-PAGE and bicinchoninic
acid assay.

Amino Acid Sequence Analyses

The N-terminal amino acid sequences of MUCF proteins were
determined and compared with known mycobacterial sequences.
One hundred micrograms of MUCF protein was resolved by two-
dimensional gel electrophoresis (2D-GE) by the method of
O'Farrell (12) with the modifications of Sonnenberg and Belisle
for resolution of mycobacterial proteins (13). 2D-GE resolved
proteins were transferred to 0.1 mm polyvinyl difluoride (PVDF)
membrane (14) and visualized by staining with 0.5% Coomassie
blue R250 in 40% methanol/10% acetic acid for 2 minutes. The N-
terminal amino acid sequence was obtained by loading membranes
containing the excised spot of interest onto a Procise-cLC
automated Edman sequencer (Applied Biosystems, Foster City,
CA) and reading the generated chromatographs against amino acid
standards.

Statistical Analyses

Statistical analyses were done by the Mantel-Haenszel chi-square
method. Statistical significance for the chi-square p values and
Fisher's exact p values was evaluated at alpha = 0.05. Fisher's
exact p value was used when the expected cell count was <5. Odds
ratios (OR) were determined by calculating the ratio of the odds of
exposure among cases to that among controls. The test for
specificity was determined by dividing the true negatives (n = 24;
disease-negative and 70 kDa-negative) by the total number of
controls (n = 27). All analyses were done on BU patients and
healthy controls from the Daloa region of Côte d'Ivoire, where the
disease is endemic, unless otherwise specified. All data were
analyzed by using Exact 2.0b software (D. Martin and H. Austin,
Atlanta, GA).

Results

Seventeen (43.6%) of the 39 skin-tested BU patients had
induration of > 10 mm within 72 hours of Burulin administration,
with PPD induration at least 3 mm smaller than that with Burulin
(Table 1). Only 3 (14.3%) of the 21 skin-tested healthy controls
from the area had a similar DTH pattern (Table 1). In addition, 11
(28.2%) of the 39 BU patients had an equally strong response to
Burulin and PPD, which suggests that Burulin skin testing may not
be reliable in regions where BU and TB are endemic and BCG
vaccination is widely used (Table 1). Dual-positive responses
included, 28 (71.8%) of the 39 BU patients were Burulin positive.
However, when this total Burulin-positive population was
delineated by disease stage, positive responses were observed in
most of the patients with healed (93.8%) and ulcerative disease
(64.7%), but in few (33.4%) patients with early ulcerative disease
(Table 2). When healed and active Burulin responses were
compared with the DTH response seen in early ulcerative BU
patients, a significant difference between the disease stages was
found (Fisher's exact p = 0.009). These skin-tested BU patient and
control populations from the Daloa region were then tested for
reactivity of their sera to MUCF by Western blot, and Burulin
induration was compared with serum antibody response (Table 2).
Twenty-six (66.7%) of the 39 skin-tested BU patients had an
antibody response to the MUCF, versus 7 (33.3%) of the 21 skin-
tested controls (chi-square p = 0.014). BU patient populations had
approximately the same positive antibody responses to the MUCF,
regardless of disease stage. Specifically, 68.8% of the healed
patients were antibody positive, with active ulcerative disease at
64.7% and early disease at 66.7% (Table 2). When the antibody
responses of healed and active patients were compared with those
seen in early BU patients, the serologic response did not differ
significantly by disease stage (Fisher's exact p = 0.999). Therefore,
serologic testing may be useful in the early diagnosis and
surveillance of BU.

Figure. Western blot reactivity to and Silver stain analysis of
Mycobacterium ulcerans culture filtrates (MUCF). A)
Representative antibody responses to MUCF. M; molecular weight
markers with annotations corresponding to molecular weight on the
left; lanes 1-6, representative BU patient sera with reactivity to
MUCF; lanes 7-10, representative antibody reactivity in healthy
persons from the disease-endemic area; lanes 11-12, serologic
reactivities to MUCF of two representative tuberculosis (TB)
patients. TB patient sera that were serologically reactive to MUCF
were included as a control for cross-reactivity. Regions of antibody
reactivity corresponding to M. ulcerans antigens of 70, 38/36, and
5 kDa are noted (arrows). B) Silver-stained SDS-PAGE gel of
MUCF. The identification of putative proteins corresponding to the
serologically reactive 70, 38/36, and 5 kDa MUCF antigens are
noted (stars). For Western blot analyses, aliquots (50 ˜g protein) of
MUCF were resolved by discontinuous SDS-PAGE (10) with
preparative 10% to 20% gradient 1-well gels (Novex, San Diego,
CA) and then transferred to nitrocellulose (14). Nitrocellulose
sheets were cut into 2-mm strips and stored in 5% skim milk in
Tris-buffered saline, pH 7.6, until used. Antibody to MUCF was
detected by probing nitrocellulose strips with sera at a 1:50 (vol:vol
dilution). Bound serum antibodies were detected with alkaline
phosphatase-conjugated anti-human antibody (Sigma Chemical
Co., St. Louis, MO), and the substrate 4-Bromo-3- Chloro-2-
Indoyl-1-Phosphate/Nitro blue tetrazoleum (Sigma Chemical Co.,
St. Louis, MO). All sera were tested in duplicate for confirmation
of the antibody responses.

The antibody response to the MUCF was then determined for all
61 clinically diagnosed BU patients and 27 healthy controls, and
antibody reactivity to individual antigens was recorded. Three M.
ulcerans antigens of 70, 38/36, and 5 kDa were commonly
recognized by BU patient sera antibodies (Figure). The 36-kDa
antigen was found to be a degradation product of the 38-kDa
antigen (unpub. results). Forty-three (70.5%) of 61 BU patient sera
had antibodies to at least 1 of 3 antigens, compared with 10
(37.0%) of 27 sera from healthy controls from the area and 4
(30.8%) of 13 sera from TB patients (Table 3). The antibody
response to the 70-kDa M. ulcerans antigen was associated with
BU disease (OR = 12.33; chi-square p = 0.00002), and the
response to this protein was consistent throughout early to late
disease stage (OR = 12.4; chi-square p = 0.0000001; Table 3).
Results were similar for the 38/36-kDa antigen (OR = 5.94; chi-
square p = 0.0016), regardless of disease stage (OR = 3.00; chi-
square p = 0.0026).

Table 3. Antibody responses to specific Mycobacterium ulcerans
culture filtrate proteins in all Buruli ulcer patients and controls


N-terminal amino acid sequence of the 38 kDa protein was
DPAPAPAPRD, and had mycobacterial sequences that precede
glycosylation sites (15). The serologic response to the 5-kDa
antigen, although common in the seroreactive specimens tested, did
not differ significantly between BU patients and healthy controls
from disease-endemic areas or TB patients (chi-square p = 0.764
and Fisher's exact p = 0.999, respectively); nor were antibodies to
this antigen observed in BU patients at the early ulcerative stage
(Table 3). A cross-reactive serologic response to MUCF was
observed in some TB patients who had not been exposed to M.
ulcerans and in some healthy controls from disease-endemic areas
(Figure; representative sample and Table 3). These positive
responses may be due to exposure to other mycobacterial
pathogens with similar antigens as M. ulcerans or BU disease
undetectable by clinical diagnosis, respectively.

Conclusions

Diagnosis of BU currently relies on the clinical presentation of the
ulcerative disease stage, by which time the infection has already
caused damage, therapeutic options are limited, and outcome is
associated with severe disease (3,7,8). Preulcerative nodules and
edemous lesions, when identified, frequently contain numerous
extracellular mycobacteria (1,2). Thus, diagnosis of BU disease at
this early stage may lead to more effective treatment. Indeed,
diagnosis of preulcerative BU disease is associated with a better
therapeutic outcome and reduced illness (8), and accurate
surveillance at the preulcerative stage would enable assessment of
the prevalence and global impact of BU disease. Surveillance may
also identify risk factors as well as prevention strategies for BU.
Consistent with previous reports (3,9), DTH to Burulin, a crude
preparation of M. ulcerans lysate (9), was observed in many
(71.8%) of the clinically diagnosed BU patients. This overall
response was similar to the well-defined cell-mediated DTH
responses reported for other mycobacterial infections (16-18),
including M. tuberculosis (19,20). In contrast, in BU patients, the
Burulin response was associated with ulcerative and healed
disease, but not with early ulcerative disease. Thus, unlike other
DTH-based monitoring programs that indicate mycobacterial
infection before and during disease in immunocompetent persons,
Burulin testing may be more useful for confirming ulcerative
disease or monitoring past disease with M. ulcerans in conjunction
with PPD results. The marked shared reactivity to PPD and Burulin
antigens by BU patients, as measured by similar induration upon
intradermal injection, was expected to some degree, considering
the commonality of antigenic lipids on the mycobacterial surface,
genetic relatedness of mycobacterial pathogens, opportunity of
exposure to both mycobacterial pathogens and environmental
mycobacteria, and rate of BCG vaccination in the region. When
Burulin-positive results were analyzed in conjunction with either
the presence of a BCG vaccination scar or recollection of BCG
vaccination, 12 of 14 BCG-vaccinated BU patients were Burulin
positive, and 6 of these 12 patients were also PPD positive.
Additionally, all the Burulin- and PPD-positive controls either
recalled BCG vaccination or had a BCG scar (data not shown).

We found that a specific response to the MUCF was observed in
BU patient sera, but unlike the Burulin DTH response, the antibody
response did not correlate with disease stage  (Table 2). The
antibody reactivity in the sera from some of the controls from
disease-endemic areas may reflect preclinical infection with M.
ulcerans or exposure to other mycobacteria and cross-reactivity to
antigens common to mycobacterial species. Interestingly, 10
(55.6%) of the 18 antibody-negative BU patients, versus 15 (33%)
of the 43 antibody-positive BU patients, recalled or had a BCG
vaccination scar, which suggests that BCG vaccination did not
contribute to the antibody response of BU patients to MUCF (data
not shown).

Thirty percent of the TB patient sera were reactive to MUCF, with
reactivity principally towards lower molecular weight proteins and
not the 70- or 38/36-kDa proteins (Figure; Table 3). N-terminal
amino acid sequence of one of these lower molecular weight
proteins, a 31 kDa protein, was found to be FSRPGLPVEY and
demonstrated homology with the mycolyl transferases found in M.
tuberculosis and other mycobacteria (21). Thus, reactivity of the
TB patient sera based on antibodies generated against common
mycobacterial antigens is the likely explanation for the few TB
patient sera reactive to MUCF.

Our preliminary results indicate the usefulness of the MUCF in
detecting BU; specifically, individual antigens in the MUCF may
be used for the development of a sensitive and specific test for BU
(Table 3). Both the 70- and 38/36-kDa proteins were indicators of
BU disease, and antibody responses to both of these proteins were
similarly present in BU-reactive patient serum samples, regardless
of disease stage. In addition, reactivity to the 70-kDa protein was
specific for M. ulcerans infection (88.8%).

The 70- and 38/36-kDa proteins may be useful for the development
of a serologic test for BU in areas where TB is endemic. Based on
these results, we are isolating and characterizing these two M.
ulcerans antigens for additional testing. The N-terminal sequence
for the 38/36-kDa protein has already been determined as part of
the purification process for this protein, and identification and
purification of the 70-kDa protein are under way. If a serologic test
proves to be diagnostic, early excision of nodules and
antimycobacterial chemotherapy trials of patients could reduce the
public health and socioeconomic impact of this emerging disease.

Acknowledgments

We thank John Stanford for providing the Burulin reagent, Jennifer
Ekmark for discussion and critical reading of the manuscript,
Merywin Wigley for laboratory assistance, Carolyn Carter for N-
terminal sequence analysis, the Emory Medical Care Foundation
for collection and use of the TB patient and control sera (HIC #
455-96), and the National Center for Infectious Diseases for the
initial case study of BU disease in Côte d'Ivoire.

This study was supported in part by Centers for Disease Control
and Prevention Cooperative agreement No. U50/ CCU416560-01-
1, Novel Diagnostic Tests for Infections of Public Health
Significance, and by an Oak Ridge Institute for Science and
Education Fellowship awarded to Karen M. Dobos.

Dr. Dobos is a senior postdoctoral fellow in the laboratory of Dr.
C. Harold King at the Emory University School of Medicine. Her
current research interests focus on bacterial proteomic analysis.
She will continue studying the proteins of Mycobacterium ulcerans
for development of future vaccine and diagnostic candidates. 

Address for correspondence: C. Harold King, Emory University
School of Medicine, 69 Butler St. S.E., Atlanta, GA 30303, USA;
fax: 404-880-9305; e-mail: cking01@emory.edu.

References

1. Horsburgh CR Jr, Meyers WM. Buruli ulcer. In: Pathology of
emerging infections. Horsburgh CR Jr, Nelson AM, editors.
Washington: American Society for Mibrobiology Press; 1997. p.
119-26.
2. Dobos KM, Horsburgh CR Jr, Ashford DA, Quinn FD, King
CH. Emergence of a unique group of necrotizing mycobacterial
diseases. Emerg Infect Dis 1999;5:367-78.
3. Marston BJ, Diallo MO, Horsburgh CR Jr, Diomande I, Saki
MZ, Kanga J-M, et al. Emergence of Buruli ulcer disease in the
Daloa region of Côte d'Ivoire. Am J Trop Med Hyg 1995;52:219-
24.
4. Amofah GK, Sagoe-Moses C, Adjei-Acquah C, Frimpong EH.
Epidemiology of Buruli ulcer in Amansie West district, Ghana.
Trans R Soc Trop Med Hyg 1993;87:644-5.
5. Hayman J. Postulated epidemiology of Mycobacterium ulcerans
infection. Int J Epidemiol 1991;20:1093-8.
6. Portaels F, Elsen P, Guimaraes-Peres A, Fonteyne PA, Meyers
WM. Insects in the transmission of Mycobacterium ulcerans
infection. Lancet 1999;353:986.
7. Asiedu K, Etuaful S. Socioeconomic implications of Buruli
ulcer in Ghana: a three-year review. Am J Trop Med Hyg
1998;59:1015-22.
8. Amofah G, Asamoah S, Afram-Gyening G. Effectiveness of
excision of pre-ulcerative Buruli lesions in field situations in a rual
district in Ghana. Trop Doct 1998;28:81-3.
9. Stanford TL, Revill WDL, Gunthorpe WJ, Grange JM. The
production and preliminary investigation of Burulin, a new skin
test for Mycobacterium ulcerans infection. Journal of Hygiene
(London) 1975;74:7-16.
10. Laemmli UK. Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 1970;227:680-5.
11. Morrissey JH. Silver stain for proteins in polyacrylamide gels:
a modified procedure with enhanced, uniform sensitivity. Anal
Biochem 1981;117:307-10.
12. O'Farrell PH. High resolution two dimensional analysis of
proteins. J Biol Chem 1975;250:4007-21.
13. Sonnenberg MG, Belisle JT. Definition of Mycobacterium
tuberculosis culture filtrate proteins by two dimensional
polyacrylamide gel electrophoresis, N-terminal amino acid
sequencing, and electrospray mass spectrometry. Infect Immun
1997;65:4515-24.
14. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of
protein from polyacrylamide gels to nitrocellulose sheets:
procedure and some applications. Proc Natl Acad Sci U S A
1979;76:4350-4.
15. Dobos KM, Khoo KH, Swiderek KM, Brennan PJ, Belisle JT.
Definition of the full extent of glycosylation of the 45-kilodalton
glycoprotein of Mycobacterium tuberculosis. J Bacteriol
1996;178:2498-506.
16. De Vries RR. An immunogenetic view of delayed type
hypersensitivity. Tubercle 1991;72:161-7.
17. Sengupta U. Studies on lepromin and soluble antigens of M.
leprae: their classification standardization and use. Indian J Lepr
1991;63:457-65.
18. Shield MJ, Stanford JL, Paul RC, Carswell JW. Multiple skin
testing of tuberculosis patients with a range of new tuberculins, and
a comparison with leprosy and Mycobacterium ulcerans infection.
Journal of Hygiene (London) 1977;78:331-48.
19. Dannenberg AM Jr. Delayed-type hypersensitivity and cell-
mediated immunity in the pathogenesis of tuberculosis. Immunol.
Today 1991;7:228-33.
20. Orme IM, Andersen P, Boom WH. T cell response to
Mycobacterium tuberculosis. J Infect Dis 1993;167:1481-97.
21. Belisle JT, Vissa VD, Sievert T, Takayama K, Brennan PJ,
Besra GS. Role of the major antigen of Mycobacterium
tuberculosis in cell wall biogenesis. Science 1997;276:1420-2.


Research

Haemophilus influenzae Type b and Streptococcus pneumoniae as
Causes of Pneumonia among Children in Beijing, China

Orin S. Levine,*† Gang Liu,† Robert L. Garman,*‡ Scott F.
Dowell,* Sangie Yu,† and Yong-Hong Yang†
*Centers for Disease Control and Prevention, Atlanta, Georgia,
USA; †Beijing Children's Hospital Affiliated with Capital
University of Medical Sciences, Beijing, China; and ‡Rollins
School of Public Health, Emory University, Atlanta, Georgia, USA

To determine if Haemophilus influenzae type b (Hib) and
Streptococcus pneumoniae could be identified more often from the
nasopharynx of patients with pneumonia than from control
patients, we obtained nasopharyngeal swab specimens from 96
patients with chest x-ray-confirmed pneumonia and 214 age-
matched control patients with diarrhea or dermatitis from the
outpatient department at Beijing Children's Hospital. Pneumonia
patients were more likely to be colonized with Hib and S.
pneumoniae than control patients, even after the data were adjusted
for possible confounding factors such as day-care attendance, the
presence of other children in the household, and recent antibiotic
use. In China, where blood cultures from pneumonia patients are
rarely positive, the results of these nasopharyngeal cultures provide
supporting evidence for the role of Hib and S. pneumoniae as
causes of childhood pneumonia.

New vaccines for Haemophilus influenzae type b (Hib) and
Streptococcus pneumoniae effectively prevent disease caused by
these pathogens (1,2). However, the relatively high cost of these
vaccines inhibits their widespread use in developing countries.
Therefore, for most developing countries, surveillance for disease
caused by Hib and S. pneumoniae is needed to demonstrate that the
investment in these highly effective vaccines is warranted (3). Data
on the incidence of Hib and pneumococcal disease in China are
limited to a few meningitis surveillance studies, which suggest a
surprisingly low incidence of confirmed Hib meningitis. In most
cases the rate is 5 to 25 times lower than those observed in areas of
North America, Europe, Africa, South America, and Oceania
where careful surveillance studies have been carried out (4-7).
However, the apparent low incidence in these studies is difficult to
interpret because of concerns that physicians may not routinely
perform lumbar punctures on all patients with suspected
meningitis, that the laboratory methods may have been inadequate,
and that widespread use of oral antibiotics for outpatients may have
artificially reduced the yield of cultures (8). In part because of the
uncertainty surrounding the local rate of Hib disease, China has not
yet made Hib vaccination a routine infant vaccination. Many
authors and organizations have called for further studies of the
epidemiology in China, but few data are currently available (9-11).

Pneumonia is a leading cause of illness and death among children
worldwide (12). Demonstrating the role of bacterial agents such as
H. influenzae and S. pneumoniae in pneumonia, however, is
difficult. The most widely accepted method for demonstrating that
a bacterial agent causes pneumonia is isolation of the bacterium
from cultures of blood or, in some cases, lung aspirates. However,
blood cultures are rarely positive in pediatric pneumonia even
when microbiologic techniques are optimal, and fastidious
organisms such as Hib will not grow unless the appropriate culture
medium is used. In China, parents and physicians often resist the
collection of blood for cultures, and given the high rate of prior
antibiotic use, the yield from these specimens is likely to be low.
Because alternatives are needed to traditional methods of
documenting Hib and S. pneumoniae as causes of pneumonia in
China, we designed a study to investigate the use of a readily
available source of specimens that is likely to be positive more
often than a blood culture. We compared the rates of isolation of
Hib and S. pneumoniae from nasopharyngeal swabs and blood
cultures of patients with radiographically confirmed pneumonia
and a group of control patients without pneumonia.

The Study

The People's Republic of China is the most populous nation in the
world, with an estimated population of >1.2 billion and >20
million births each year (13). Beijing, the capital, has a population
of approximately 11 million people. Beijing Children's Hospital is
one of the largest children's hospitals in Asia, with >700 beds in 24
wards. The outpatient department typically has 2,500 to 4,000
patient visits per day. Before patients were enrolled, a study nurse
or physician explained the study to the parent or guardian,
answered questions, and obtained written informed consent. The
protocol for this study was approved by the human subjects review
committees at Beijing Children's Hospital and the Centers for
Disease Control and Prevention. Pneumonia cases were defined as
illness in patients 2 to 60 months of age with radiographic evidence
of pulmonary infiltrates and at least three of the following: fever
(temperature 38.0°C), tachypnea (>50 breaths per min for infants
<12 months old, and >40 breaths per min for children 12 to 60
months old), cough, auscultation findings indicative of lower
respiratory disease (including rhonchi, crackles, or bronchial breath
sounds), or chest indrawing. Specimens for blood culture were
collected from all 96 pneumonia patients. Urine specimens were
obtained from 89 (93%) pneumonia patients and 199 (93%)
controls.

Chest radiographs from all pneumonia patients were interpreted
initially by a radiologist at Beijing Children's Hospital, who
characterized the pattern of infiltrates as evidence of alveolar
consolidation only, interstitial infiltrates only, or a mixed pattern
with evidence of interstitial infiltrates and consolidation. The
radiographs were then recorded by a digital camera and read later
by one of the investigators who is a pediatrician (SFD). The
pediatrician characterized the radiographs as having evidence of
obvious or not obvious pneumonia. Both the radiologists and the
pediatricians were blinded to the colonization status of the patients.

Control patients were children with a diagnosis of diarrhea or
dermatitis who had no indication of respiratory tract disease. All
subjects were recruited from patients attending the Outpatient
Department. None of the participating children had received Hib
vaccine. Of 121 eligible pneumonia patients identified in the
Radiology department, 96 (79%) were enrolled; 214 controls were
also enrolled, 169 with diarrhea and 45 with dermatitis. Cases and
controls were frequency-matched by age in a 1-to-2 ratio in the
following age categories: 2 to 5 months, 6 to 11 months, 12 to 35
months, and 36 to 60 months.

Laboratory Methods

After measuring the distance from the nares to the earlobe, a
researcher inserted a fine flexible Dacron swab past the point of
mild resistance to the posterior pharynx. The swab was rotated
twice, withdrawn, and immediately used to make a uniform
suspension in a skim milk/glucose/ glycerol solution. The
suspension was then plated onto a blood agar plate with gentamicin
(for isolation of pneumococci) and a chocolate agar plate
supplemented with X and V factors and bacitracin (for isolation of
H. influenzae). Plates were refrigerated before inoculation and
warmed to room temperature before use.

Within 3 hours of inoculation, blood agar plates were placed in an
incubator at 35 o C to 37 o C and 5% to 10% CO 2 , and individual
colonies were isolated in the Microbiology and Immunology
Laboratory, Beijing Children's Hospital. After 18 to 36 hours of
incubation, the plates were assessed for the appearance of   -
hemolytic colonies resembling streptococci. A second plate as then
streaked for confluent growth and an optochin disk placed on it.
Strains inhibited by optochin were confirmed as pneumococci by
bile solubility testing. For isolation of H. influenzae, X and V
factor-supplemented chocolate agar plates were placed in an
incubator at 35 o C to 37 o C and 5% to 10% CO 2 , and single
colonies were isolated. After overnight incubation, the plates were
assessed each morning for 2 days for the appearance of large, flat,
colorless-to-gray opaque colonies, without hemolysis or
discoloration of the medium. H. influenzae were determined by
dependency for X and V factors and absence of hemolysis on blood
agar plates. All isolates were serotyped by using type-specific
antiserum. For blood cultures, which were done in the Clinical
Microbiology Laboratory at Beijing Children's Hospital, 2 to 5 ml
of blood was inoculated into blood culture bottles that contained
agents to inhibit antibiotic activity. The bottles were incubated by
the BacTec 9120 system (Organon Teknika, Durham, NC). All
positive blood cultures were subcultured onto blood agar and
chocolate agar plates. These plates were prepared by the same
techniques and quality control used in the research laboratory at
Beijing Children's Hospital. For colonies with appropriate
morphologic features on subcultures, the same protocol for
identification of pneumococcus and H. influenzae was used as
described earlier. Reagents and media were checked weekly for
their ability to support the growth of control strains of H.
influenzae and S. pneumoniae. Urine specimens were tested for
antigens to Legionella pneumophila serogroup 1 by an optical
immunoassay and for evidence of prior antimicrobial activity by a
Micrococcus luteus inhibition assay. To determine if Chlamydia
pneumoniae were present in the nasopharynx, nasopharyngeal
swabs were cultured on HEp-2 cells by using a modification of
standard techniques involving repeat centrifugation steps (14). C.
pneumoniae cultures were performed only on a subset of 30
patients >2 years old.

Statistical Analysis

Recent day-care attendance was defined as attendance in the last
month before enrollment. Exposure to other children was defined
as residing in the same household with at least one other child <18
years old. All analyses were conducted by SAS (SAS for Windows
v.6.12; SAS Institute, Cary, NC). Logistic regression was used to
assess the independent association between colonization and
pneumonia and to adjust for potential confounders. Odds ratios
(OR) were calculated by comparing pneumonia patients with
control patients without pneumonia. Separate ORs were calculated
for both the clinical and radiologic definitions of pneumonia. ORs
with 95% confidence intervals that do not include 1.00 and p
values <0.05 were considered statistically significant.

Results

Four patients who were initially enrolled as pneumonia patients
were later determined to have myocarditis and were thus excluded.
One pneumonia patient had a blood culture positive for S.
pneumoniae; no patients had a blood culture positive for H.
influenzae. All urine specimens were negative for L. pneumophila
serogroup 1. C. pneumoniae was not cultured from any of the 30
specimens tested. Eleven pneumonia patients were hospitalized.
Among the pneumonia patients, 45 had a pattern of alveolar
consolidation, 23 had an interstitial pattern, and 28 had a mixed
pattern (both consolidation and interstitial).

Comparison of the patients with radiographic pneumonia and
control patients showed that the distribution of ages was not
consistently one pneumonia to two nonpneumonia patients in all
age groups (Table 1). Although the differences in the distributions
are not statistically significant, we adjusted all subsequent analyses
by age to account for any residual confounding by this factor.
Pneumonia patients were significantly more likely to have received
antibiotics before attending the outpatient clinic, to attend day care,
and to have other children in the household (Table 1; p < 0.01, for
each comparison). Of the 96 patients who had radiographic
evidence of pneumonia, 32 (33%) were considered to have obvious
pneumonia by the pediatrician's reading. The prevalence of Hib
colonization was significantly greater among pneumonia patients
than among patients without pneumonia (Table 2). The association
was stronger, however, when the patients meeting the clinical
definition of pneumonia were compared with the controls (OR =
5.8) than when the radiologist's definition of pneumonia was used
(OR = 4.2). Multivariate analysis by logistic regression showed
that the association remained after the data were. adjusted for
potential confounders (recent antibiotic use, recent day-care
attendance, and other children in the household), and in the case of
the clinical definition of pneumonia, increased the strength of the
association. In univariate analysis, colonization with S.
pneumoniae was not significantly higher among pneumonia
patients than among patients without pneumonia (Table 2). This
association, however, was confounded by the higher proportion of
pneumonia patients with prior antibiotic use, an exposure that
decreases colonization substantially. After the data were adjusted
for prior antibiotic use, recent day-care attendance, and exposure to
other children in the household, colonization with S. pneumoniae
was significantly associated with pneumonia by either the
radiographic or clinical definition of pneumonia.

Table 1. Characteristics of pneumonia cases and diarrhea and
dermatitis control patients, Beijing, China

Table 2. Association of nasopharyngeal colonization with
Haemophilus influenzae type b (Hib) and Streptococcus
pneumoniae and radiographically confirmed pneumonia, Beijing,
China

Conclusions

We found that Hib and S. pneumoniae can be isolated significantly
more often from the nasopharynx of children with radiographically
confirmed pneumonia in China than from comparable children
without pneumonia, demonstrating the potential role of these
bacteria as a cause of pneumonia in Chinese children. These results
suggest that further studies should be undertaken to quantify the
role of these agents in China, the prevalent serotypes, and the
potential impact of vaccines for their prevention. The prevalence
rates of Hib and S. pneumoniae colonization observed among
controls are similar to those we reported in a study of outpatients at
Beijing Children's Hospital, as well as by other investigators
(15,16). The prevalence of Hib colonization among nonpneumonia
patients (~2%) is consistent with that observed in other studies of
young children in China, but lower than commonly seen in other
developing countries (5%-15%) (17-19). Even in industrialized
countries such as the United States, Finland, and the United
Kingdom, prevaccination colonization rates of 3% to 6% were
generally observed (20). Host and environmental factors may play
a role in this apparently low prevalence of Hib colonization in
nonpneumonia patients. For example, widespread overuse of
antibiotics in children and the relatively small family size as the
result of the "one family, one child" policy may contribute to this
observation.

The rate of positive blood cultures among pneumonia patients was
lower (1%) than that generally observed in studies from other
countries, but consistent with reports of a low rate of bacterial
meningitis and invasive disease from mainland China and Taiwan
(5,6,10). In rural Papua New Guinea, S. pneumoniae and H.
influenzae were isolated from the blood of 11% and 12%,
respectively, of children >6 years old with moderate or severe
acute lower respiratory tract illness (21). However, a recent study
from Dallas found no positive blood cultures among 106
ambulatory outpatients <5 years old with radiographically
confirmed pneumonia (22). Thus, the rate of positive blood
cultures was lower than that seen in hospital-based studies in
developing countries but consistent with recent data from studies in
industrialized countries of children with pneumonia who attend
outpatient clinics.

Several steps were taken to ensure the quality of the blood
specimens collected and their processing. For example, all blood
specimens were collected before intravenous antibiotics were
administered. The protocol for the study was to inoculate bottles
with 2 and 5 ml of blood, a volume considered optimal for the 15-
ml blood culture bottles used in the BacTec system, and the blood
culture bottles contained resin-coated beads designed to neutralize
the antibacterial activity of antecedent antibiotic use.

The observation that colonization with Hib and S. pneumoniae is
more likely among pneumonia patients, especially those with
radiographically obvious pneumonia, strengthens the contention
that pneumonia is associated with these agents. This finding is
consistent with the observation from studies with Hib conjugate
vaccine showing that the protection afforded by these vaccines is
most apparent in children with obvious radiographic pneumonia or
alveolar consolidation and not apparent in children with milder
forms of pneumonia (23,24). However, an inherent limitation of
case-control studies is that they cannot establish that the exposure,
in this case colonization with Hib or S. pneumoniae, preceded the
onset of illness.

The potential role of other agents of pneumonia besides Hib and S.
pneumoniae remains unclear. In this study, H. influenzae
colonization (regardless of serotype) was also associated with
pneumonia (data not shown). The observation that no patients had
antigens to L. pneumophila serogroup 1 in urine is consistent with
findings from the United States and elsewhere indicating that this
agent rarely causes pneumonia among young children (25). The
finding that none of 30 patients >2 years of age were positive for
C. pneumoniae by culture suggests that if this organism causes
pneumonia in this age group, alternative methods for detection are
needed to define its role. The potential role of other agents such as
Mycoplasma pneumoniae and viruses such as respiratory syncytial
virus warrants further investigation. Characterizing the
epidemiology of these agents may have important implications for
the choice of empiric therapy in pneumonia patients and for the
development of vaccines for pneumonia prevention.

Acknowledgments

We thank the following persons for assistance with laboratory
training and specimen processing: Barry Fields, Jan Pruckler, Steve
Bosshardt, Valerie Stevens, John Elliott, and Sophie Newland at
CDC, and Jie Li, Yue Juan Tong, Wei Gao, Yan Li, and Gui Rong
Zhang at Beijing Children's Hospital. We also thank Sun Guo
Qiang, Jin Jin Zeng, and Sai Ying Xu, the radiologists who
systematically interpreted the radiographs for us in Beijing, and
Zaifang Jiang and Dexiu Yao for their support and assistance.

This work was supported by funding from the Emerging Infectious
Diseases Program, National Center for Infectious Diseases, Centers
for Disease Control and Prevention.

Dr. Levine is an epidemiologist in the Division of Microbiology
and Infectious Diseases, National Institute of Allergy and
Infectious Diseases. His research interests include bacterial
pneumonia and meningitis and their prevention by vaccination. He
collaborates on related projects in South America, Africa, and the
Middle East.

Address for correspondence: Orin S. Levine, Respiratory Diseases
Branch, Division of Microbiology and Infectious Diseases,
National Institute of Allergy and Infectious Diseases, 6700-B
Rockledge Drive, Room 3133, Bethesda, MD 20892-7630, USA;
fax: 301-496-8030; e-mail: Olevine@niaid.nih.gov.

References

1. Steinhoff MC. Haemophilus influenzae type b infections are
preventable. Lancet 1997;349:1186-7.
2. Eskola J, Anttila M. Pneumococcal conjugate vaccines. Pediatr
Infect Dis J 1999;18:543-51.
3. Levine MM, Levine OS. Disease burden, public perception and
other factors that influence new vaccine development,
implementation and continued use. Lancet 1997;350:1386-92.
4. Yang Y, Leng Z, Shen X, Lu D, Jiang Z, Rao J, et al. Acute
bacterial meningitis in children in Hefei, China 1990-1992.
Chinese Medical Journal 1996;109:385-8.
5. Wang CH, Lin TY. Invasive Haemophilus influenzae diseases
and purulent meningitis in Taiwan. J Formos Med Assoc
1996;95:599-604.
6. Lau YL, Low LC, Yung R, Ng KW, Leung CW, Lee WH, et al.
Invasive Haemophilus influenzae type b infections in children
hospitalized in Hong Kong, 1986- 1990. Hong Kong Hib Study
Group. Acta Paediatrica 1995;84:173-6.
7. Levine OS, Schwartz B, Pierce NF, Kane M. Development,
evaluation and implementation of Hib vaccines for young children
in developing countries: current status and priority actions. Pediatr
Infect Dis J 1998;17:S95-113.
8. Salisbury DM. Summary statement: The First International
Conference on Haemophilus influenzae type b infection in Asia.
Pediatr Infect Dis J 1998;17:S93-5.
9. Lau YL, Yung R, Low L, Sung R, Leung CW, Lee WH.
Haemophilus influenzae type b infections in Hong Kong. Pediatr
Infect Dis J 1998;17:S165-9.
10. Yang Y, Shen X, Jiang Z, Liu X, Leng Z, Lu D, et al. Study on
Haemophilus influenzae type b diseases in China: the past, present
and future. Pediatr Infect Dis J 1998;17:S159-65.
11. Lee HJ. Epidemiology of systemic Haemophilus influenzae
disease in Korean children. Pediatr Infect Dis J 1998;17:S185-9.
12. World Health Organization. Acute respiratory infections: a
forgotten pandemic. Bull WHO 1998;76:101-3.
13. UNICEF. The state of the world's children 1998. New York:
Oxford University Press; 1999. p. 1-131.
14. Pruckler J, Masse N, Stevens V, Gang L, Yang Y, Zell ER, et
al. Optimizing culture of Chlamydia pneumoniae using multiple
centrifugations. J Clin Microbiol. In press 1999.
15. Whitney CG, Wang JF, Elliott JA, Levine OS, Dowell SF,
Yang YH. Nasopharyngeal carriage of Streptococcus pneumoniae
among children in Beijing: drug resistance and risk factors for
carriage. 63-63. 1998. Washington: World Health Organization.
Pneumococcal vaccines for the world 1998 conference. 10-12-
1998.
16. Wang H, Huebner R, Chen M, Klugman K. Antibiotic
susceptibility patterns of Streptococcus pneumoniae in China and
comparison of MICs by agar dilution and E-test methods.
Antimicrob Agents Chemother 1998;42:2633-6.
17. Sung RY, Ling JM, Fung SM, Oppenheimer SJ, Crook DW,
Lau JT, et al. Carriage of Haemophilus influenzae and
Streptococcus pneumoniae in healthy Chinese and Vietnamese
children in Hong Kong. Acta Paediatrica 1995;84:1262-7.
18. Bijlmer HA, Evans NL, Campbell H, van Alphen L,
Greenwood BM, Valkenburg HA, et al. Carriage of Haemophilus
influenzae in healthy Gambian children. Trans R Soc Trop Med
Hyg 1989;83:831-5.
19. Gomez E, Moore A, Sanchez J, Kool J, Castellanos PL, Feris
JM, et al. The epidemiology of Haemophilus influenzae type b
carriage among infants and young children in Santo Domingo,
Dominican Republic. Pediatr Infect Dis J 1998;17:782-6.
20. Barbour ML. Conjugate vaccines and the carriage of
Haemophilus influenzae type b. Emerg Infect Dis 1996;2:176-82.
21. Barker J, Gratten M, Riley I, Lehmann D, Montgomery J, Kajoi
M, et al. Pneumonia in children in the Eastern Highlands of Papua
New Guinea: a bacteriologic study of patients selected by standard
clinical criteria. J Infect Dis 1989;159:348-52.
22. Wubbel L, Muniz L, Ahmed A, Trujillo M, Carubelli C,
McCoig C, et al. Etiology and treatment of community-acquired
pneumonia in ambulatory children. Pediatr Infect Dis J
1999;18:98-104.
23. Mulholland K, Hilton S, Adegbola R, Usen S, Oparaugo A,
Omosigho C, et al. Randomized trial of Haemophilus influenzae
type b tetanus protein conjugate for prevention of pneumonia and
meningitis in Gambian infants. Lancet 1997;349:1191-7.
24. Levine OS, Lagos R, Munoz A, Villaroel J, Alvarez AM,
Abrego P, et al. Defining the burden of pneumonia in children
preventable by vaccination against Haemophilus influenzae type b.
Pediatr Infect Dis J 1999;18:1060-4.
25. Orenstein WA, Overturf G, Leedom JM, Alvarado R, Geffner
M, Fryer A, et al. The frequency of Legionella infection
prospectively determined in children hospitalized with pneumonia.
J Pediatr 1981;99:403-6.


Dispatches

Bacteroides fragilis Enterotoxin Gene Sequences in Patients with
Inflammatory Bowel Disease

Thomas P. Prindiville, Rafik A. Sheikh, Stuart H. Cohen,
Yajarayma J. Tang, Mary C. Cantrell, and Joseph Silva, Jr.
University of California, Davis Medical Center, Sacramento,
California, USA

We identified enterotoxigenic Bacteroides fragilis in stool
specimens of patients with inflammatory bowel disease and other
gastrointestinal disorders. The organism was detected in 11
(13.2%) of 83 patients with inflammatory bowel disease. Of 57
patients with active disease, 19.3% were toxin positive; none of
those with inactive disease had specimens positive for
enterotoxigenic Bacteroides fragilis gene sequences.


Bacteroides fragilis, a gram-negative rod, constitutes 1% to 2% of
the normal colonic bacterial microflora in humans (1,2). It is
frequently associated with extraintestinal infections such as
abscesses and soft tissue infections, as well as diarrheal diseases in
animals and humans (3-6). Enterotoxigenic B. fragilis (ETBF) is an
emerging enteric pathogen associated with diarrheal diseases in
children, adults, and animals (4,7-9). The pathogen is unusual in
children during the first year of life, but diarrhea associated with it
is common in children 1 to 5 years of age, which suggests early
acquisition of the organism and maternal protection. The
pathogenicity of B. fragilis is related to its production of a potent
enterotoxin, a zinc-containing metalloprotease with a molecular
weight of 20,000 (10). The enterotoxin is tight-junction specific;
causes rounding, swelling, and pyknosis of cultured enterocytes;
and induces a fluid response in ligated intestinal loops and a
cytotoxic response in the HT-29 colon cell line (7). Ulcerative
colitis and Crohn disease are inflammatory diseases of the
gastrointestinal tract characterized by spontaneous remissions and
relapses. Many microbial pathogens, particularly Mycobacterium
paratuberculosis, paramyxoviruses, and Listeria monocytogenes
have been implicated in the etiology of inflammatory bowel
disease (IBD) (11). In addition, enteric pathogens such as
Campylobacter jejuni, Salmonella, Shigella, Yersinia, and
Escherichia coli have also been associated with relapses of IBD
(12). We investigated the prevalence of ETBF in 83 patients with
idiopathic IBD, in 18 patients with routine culture-negative
diarrhea, and in a control population of 69 outpatients (Table 1).

Table 1. Demographic characteristics of patients with
inflammatory bowel disease (IBD)

The Study

The study protocol was approved by the Human Subjects Research
Review Committee of the University of California, Davis. All
samples were collected after informed consent was obtained. Of
the 83 patients in the IBD group, 60 had Crohn disease, and 23 had
ulcerative colitis (Table 2). Active disease was present in 68.6% of
these patients on endoscopy (20 ulcerative colitis and 37 Crohn
disease patients) in the form of mucosal erythema and ulceration.
In the miscellaneous diarrhea group, 10 patients had irritable bowel
syndrome; however, no mucosal erythema, edema, or ulceration
was observed. None of the control patients had a history of
diarrhea, and no erythema or ulceration was observed. Fecal
specimens from the IBD group were collected endoscopically from
patients undergoing colonoscopy or flexible sigmoidoscopy for
evaluation of symptoms of diarrhea or abdominal pain. The fecal
specimens from the 18 diarrhea patients with negative routine stool
cultures and the 69 controls were collected by endoscopy. Fecal
specimens were cultured for B. fragilis in the selective medium
Bacteroides Bile Esculin agar. Positive cultures were identified by
using the RapID ANA II Panel (REMEL, Inc, Lenexa, KS). Plates
were incubated anaerobically at 37°C for 48 hours. The presence of
B. fragilis enterotoxin in the isolates was detected in the HT-29
colon cell line (13,14). HT-29 cells were grown and maintained in
RPMI medium with glutamine (Gibco, Life Technologies, Inc.,
Grand Island, NY) supplemented with penicillin (100 IU/ml),
streptomycin (100 ˜g/ml) (Sigma, Saint Louis, MO), and heat-
inactivated fetal bovine serum (Hyclone Laboratories, Inc., Logan,
UT) at 12% in 25-ml flasks at 37°C in 5% CO 2 . For the
cytotoxicity assay, HT-29 cells were suspended in 20 ml of
medium for each plate, and 180 ˜l/well were placed into 96-well
tissue culture plates (Corning Glass Works, Corning, NY). The
cells were allowed to attach and grow for 2 to 3 days. Supernatants
from B. fragilis isolates grown on Brain Heart Infusion Broth were
filtered through a 0.45-˜m Acrodisk syringe filter (Gelman
Sciences, Ann Arbor, MI), and 20-˜l serial dilutions were placed
into the wells in duplicate. The plates were incubated at 37°C for 3
to 4 hours under 5% CO 2 and examined for typical cytopathic
changes. Cultures were considered positive for B. fragilis
enterotoxin if a visible cytopathic effect was neutralized by specific
antiserum. The highest dilution of the culture supernatant
producing cytopathic changes in at least 50% of the cells after 3 to
4 hours of incubation was considered the cytotoxic titer. For the
neutralization assay, dilutions of 1:25 anti-enterotoxin rabbit
antiserum in phosphate-buffered saline were mixed with culture
supernatants positive for enteroxin. After incubation for 30 minutes
at 37°C, 20 ˜l of each mixture was inoculated into HT-29 cells as
in the cytotoxicity assay. Neutralization was indicated by the lack
of cytotoxic effect.

Table 2. Bacteroides fragilis enterotoxin gene amplification
products in patients with inflammatory bowel disease (IBD)

DNA was extracted from the fecal specimens and amplified by
using specific primers (14) to detect B. fragilis enterotoxin gene
sequences.1 Chi-square was used to determine statistical
significance

ETBF was cultured from four patients with IBD and two patients
with diarrhea. However, B. fragilis enterotoxin gene sequences
were detected in the stools of 11 (13.25%) of 83 patients with
idiopathic IBD (five with Crohn disease and six with ulcerative
colitis). All 11 patients positive for ETBF had active disease by
endoscopic and histologic tests (Table 2). No ETBF was found in
patients with inactive disease. The Crohn disease patients who
were positive for enterotoxin gene sequences had superficial
mucosal disease. In the control group, 2 (2.9%) of 69 patients were
positive for B. fragilis enterotoxin gene sequences. Enterotoxin
amplification products were also detected in the specimens of 5
(27.7%) of 18 patients with diarrhea due to miscellaneous causes
(Table 2). All the specimens were positive when amplified with
primers specific for the 16S rRNA gene of enteric bacteria.

Conclusions

The normal colonic microflora of humans is a complex ecosystem
of approximately 500 species of aerobic and anaerobic
microorganisms. Although the gut of the newborn infant is sterile,
Bacteroides species—the predominant anaerobic constituent of the
colonic flora—appear at approximately 10 days, are established by
2 weeks, and usually remain constant lifelong (1,2). In breast-fed
infants, Bifidobacterium are the predominant population, and
Bacteroides group organisms remain undetectable. However, after
weaning, the Bacteroides group organisms increase and
Bifidobacterium organisms decrease substantially (16).

The exact etiology of idiopathic IBD is still unknown, although a
potential role for infectious agents or toxins that may stimulate an
inflammatory response has been suggested (16,17). For example,
Peptostreptococcus, Coprococcus, and Bacteroides sp. have been
reported in patients with Crohn disease (17). A role for these
microorganisms in the disease process is also suggested by the
clinical responses of some patients to antibiotics (18). Observations
of B. thetaiotaomicron in patients with ulcerative colitis (16) and
B. vulgatus in guinea pigs with experimentally induced ulcerative
colitis (19) suggest that microorganisms may influence the
development or maintenance of intestinal inflammation in IBD.

In this study, B. fragilis enterotoxin gene sequences were detected
by nested polymerase chain reaction (PCR) in the stools of 13.2%
of patients with inflammatory bowel disease and 2.9% controls.
The low recovery of ETBF in culture may be due to the length of
time from specimen collection to processing in the laboratory
(most specimens were kept frozen for at least 2 weeks before
culturing). Similar results have been reported by Sack et al. (9),
who found a marked reduction in the recovery of B. fragilis with
time. Amplification of enterotoxin gene directly in the stools of
these patients appears to be a more sensitive detection method. In a
previous study (14), we found 100% correlation between PCR and
enterotoxin production in isolates; we therefore did not routinely
perform the HT-29 cell assay in specimens of these patients.

In the IBD group, all the patients positive for ETBF had active
disease, which suggests an association with disease activation or
flare-up. ETBF was also found in patients with ulcerative proctitis,
collagenous colitis, and microscopic colitis. In patients with Crohn
disease, ETBF was usually seen in the colonic superficial
inflammatory disease type. The presence of ETBF in IBD patients
may represent alterations of endogenous bacterial flora, which may
be related to either the etiology or flare-up of the disease or both.
Colonization with ETBF may be acquired early in life or may be a
de novo infection related to flare-ups of the disease.

Acknowledgment

The authors thank David M. Lyerly for providing the anti-
enterotoxin rabbit antiserum.

Dr. Prindiville is an associate professor of medicine in the Division
of Gastroenterology at the University of California, Davis,
California, USA.

Address for correspondence: Stuart H. Cohen, Department of
Internal Medicine, Division of Infectious Diseases, 4150 V Street,
Patient Services and Support Building, Suite 500, Sacramento, CA
95817, USA; fax: 916-734-7766; e-mail:
stcohen@ucdavis.edu.


1 A 100-mg sample of stool was suspended in 400 ˜l of TES buffer
(50 mM Tris [pH 8], 5 mM EDTA, 50 mM NaCl) and centrifuged
at 1,000 x g for 3 min to remove large particles. The supernatant
was then centrifuged at 5,000 x g for 7 min. The pellet was washed
once in 200 ˜l of TES, and centrifuged at 5,000 x g for 3 min, and
the supernatant was discarded. The pellet was suspended in 100 ˜l
of sterile H 2 O and boiled for 10 min, then centrifuged at 1,000 x
g for 2 min and the supernatant containing the DNA extracted
twice with phenol: chloroform: isoamyl alcohol (25:24:1) and
precipitated with ethanol. The sequences of the primers and probes
used and PCR conditions were as described (14). Amplification
with the outer primers RS-3 (5'
TGAAGTTAGTGCCCAGATGCAGG 3') and RS-4 (5'
GCTCAGCGCCCAGTATA TGACC 3') yielded a 367-bp product.
Amplification of this product with the inner primers RS-1 (5'
TGCGGCGAACTCGGTTAATGC 3') and RS-2
(5'AGCTGGGTTGTAGACATCCCACTGG' 3') amplified a 290-
bp product. The reaction mixtures were prepared in 1X PCR buffer
(50 mM KCl, 20 mM Tris HCl, 2.5 mM MgCl 2 , 100 ˜g bovine
serum albumin per ml [pH 8.4]) and contained per reaction 20
pmol of the respective primers, 0.1 mM concentrations each of 2'-
deoxynucleoside 5'-triphosphate, 2 U of recombinant DNA
polymerase (rTaq) (Perkin Elmer, Norwalk, CT), and 10 ˜l purified
fecal DNA. The final reaction volume was adjusted to 100 ˜l with
sterile deonized water. The PCR profile included a denaturing step
at 95°C for 30 sec, followed by a 60°C annealing step for 30 sec,
with extension at 72°C for 30 sec. The outer PCR was performed
for 35 cycles in a thermal cycler (MJ Research). Amplification
with the inner primers was done for 30 cycles. Negative controls
included a blank containing all PCR reagents with no DNA. As
control for amplifiable DNA in the stool specimens, primers
targeting the 16S rRNA gene of enteric bacteria were used as
described by Kato et al. (15).
  
References

1. Moore WE, Cato EP, Holdeman LV. Some current concepts in
intestinal bacteriology. Am J Clin Nutr 1978;31:S33-42.
2. Simon GL, Gorbach SL. Intestinal flora in health and disease.
Gastroenterology 1984;86:174-93.
3. Gorbach SL, Barlett JG. Anaerobic Infections. N Engl J Med
1974;290:1177-84.
4. Myers LL, Firehammer BD, Shoop DS, Border MM.
Bacteroides fragilis: a possible cause of acute diarrheal diseases in
newborn lambs. Infect Immun 1984;44:241-4.
5. Sack RB, Myers LL, Almeido-Hill J, Shoop DS, Bradbury WC,
Reid R, et al. Enterotoxigenic Bacteroides fragilis: epidemiologic
studies of its role as a human diarrhoeal pathogen. J Diarrhoeal Dis
Res 1992;10:4-9.
6. Pantosti A, Menozzi MG, Frate A, Sanfilippo L, D'Ambrosio F,
Malpeli M. Detection of enterotoxigenic Bacteroides fragilis and
its toxin in stool samples from adults and children in Italy. Clin
Infect Dis 1997;24:12-6.
7. Weikel CS, Grieco FD, Reuben J, Myers LL, Sack RB. Human
colonic epithelial cells, HT-29/C1, treated with crude Bacteroides
fragilis enterotoxin dramatically alter their morphology. Infect
Immun 1992;60:321-7.
8. SanJoaquin VH, Griffis JC, Lee C, Sears CL. Association of
Bacteroides fragilis with childhood diarrhea. Scand J Infect Dis
1995;27:211-5.
9. Sack RB, Albert MJ, Alam K, Neogi PK, Akbar MS. Isolation
of enterotoxigenic Bacteroides fragilis from Bangladeshi children
with diarrhea: a controlled study. J Clin Microbiol 1994;32:960-3.
10. VanTassell RL, Lyerly DM, Wilkins TD. Purification and
characterization of an enterotoxin from Bacteroides fragilis. Infect
Immun 1992;60:1343-50.
11. Sartor RB. Microbial factors in the pathogenesis of Crohn's
disease, ulcerative colitis, and experimental intestinal
inflammation. In: Kirsner JB, Shorter RJ, editors. Inflammatory
bowel disease. 4th ed. Baltimore: Williams and Wilkins; 1995. p.
96-124.
12. Hermens DJ, Miner PB Jr. Exacerbation of ulcerative colitis
[clinical reference]. Gastroenterol 1991;101:254-62.
13. Pantosti A, Cerquetti M, Colangeli R, D'Ambrosio F.
Detection of intestinal and extraintestinal strains of enterotoxigenic
Bacteroides fragilis by the HT-29 cytotoxicity assay. J Med
Microbiol 1994;41:191-6.
14. Shetab R, Cohen SH, Prindiville TP, Tang YJ, Cantrell M,
Rahmani D, et al. Detection of Bacteroides fragilis enterotoxin
gene by PCR. J Clin Microbiol 1998;36:1729-32.
15. Kato N, Ou CY, Kato H, Bartley SL, Luo CC, Kilgore E, et al.
Detection of toxigenic Clostridium difficile in stool specimens by
the polymerase chain reaction. J Infect Dis 1993;167:455-8.
16. Dore J, Sghir A, Hannequart-Gramet G, Corther G, Pochart P.
Design and evaluation of a 16S rRNA-targeted oligonucleotide
probe for specific detection and quantitation of human faecal
Bacteroides populations. System Appl Microbiol 1998;21:65-71.
17. Van de Merwe JP, Schroder AM, Wensinck F, Hazenberg MP.
The obligate anaerobic faecal flora of patients with Crohn's disease
and their first degree relatives. Scand J Gastroenterol
1988;23:1125-31.
18. Sartor RB. Current concepts of the etiology and pathogenesis
of ulcerative colitis and Crohn's disease. Gastroenterol Clin North
Am 1995;24:475-507.
19. Onderdonk AB, Cisneros RL, Bronson RT. Enhancement of
experimental ulcerative colitis by immunization with Bacteroides
vulgatus. Infect Immun 1983;42:783-8.

Dispatches

Outbreak among Drug Users Caused by a Clonal Strain of Group A
Streptococcus

Lorenz M. Böhlen, Kathrin Mühlemann, Olivier Dubuis, Christoph
Aebi, and Martin G. Täuber
University of Berne, Berne, Switzerland

We describe an outbreak among drug users of severe soft-tissue
infections caused by a clonal strain of group A streptococcus of M-
type 25. Cases (n = 19) in drug users were defined as infections
(mainly needle abscesses) due to the outbreak strain. Comparison
with controls showed that infected drug users bought drugs more
often at a specific place. Drug purchase and use habits may have
contributed to this outbreak.

In the 1980s, reemergence of severe group A streptococcus (GAS)
infections, especially toxic shock syndrome, necrotizing fasciitis,
and bacteremia associated with high death rates, was observed
(1,2). Temporal and geographic clustering of cases with severe
GAS infection has been described (3,4). Outbreaks caused by
clonal strains have been reported in households, schools, and
hospitals (5).

In September 1997, a sudden increase was observed in needle
abscesses due to GAS among drug users hospitalized in Berne,
Switzerland. Analysis of GAS isolates suggested that the outbreak
was caused by clonal strains. An outbreak investigation included a
case-control study of potential sources and risk factors for
acquisition of GAS infection among drug users in Berne. To our
knowledge, this is the first report of an outbreak among drug users
of invasive GAS infections caused by clonal strains.

The Study

Cases of GAS infection were identified through culture records of
the Institute of Medical Microbiology, University of Berne,
Switzerland. During September to December 1997, all GAS
isolates from any site were prospectively stored and included in the
study. Cases were defined as GAS infection in drug users from
September 22 to November 20, 1997, due to the outbreak strain, as
determined by pulsed-field gel electrophoresis (PFGE). During
December 15-20, 1997, controls were randomly chosen from drug
users visiting a government shelter for drug use and needle
exchange in Berne. An exclusion criterion was a skin infection at
an injection site since September. Drug users with cases, as well as
controls, were interviewed by a standardized questionnaire
including age, employment, recent infections, and drug use habits
(Table). Crude odds ratios and 95% confidence intervals were
calculated with EpiInfo software, version 6.02 (CDC, Atlanta,
GA). Categorical data were compared with Fisher's exact test or
the chi-square test, and continuous data with Student's t-test. To
identify independent risk factors, logistic regression analysis was
performed in EGRET (Seattle, WA).

GAS was isolated on sheep blood agar and identified by gram
staining and the bacitracin test (6). Throat swabs were obtained at
the time of the interview from the 55 controls. Culture specimens
were taken with moistened (sterile saline) swabs from spoons (n =
10) and filters (n = 2) used for drug preparation.

Table. Sociodemographic and drug use habits of 19 drug users with
group A streptococcal infection and 55 healthy controls, Berne,
Switzerland, 1997

Samples of cocaine confiscated by police during September 15-28
and October 12-20 were cultured for GAS. The drug specimens
had been stored for 1 to 6 weeks in a dry place before culture. For
each cocaine specimen, three 0.2-g samples were dissolved in
sterile saline. One sample was inoculated into tryptic soy broth
(TSB); the other two were filtered through sterile 0.45-˜m
membranes. One filter was placed on a sheep blood agar plate, and
the other was inoculated into 500 ml TSB. Cultures were incubated
at 37°C. Subcultures from TSB were placed on sheep blood agar at
12-hour intervals until the broth became turbid. PFGE was
performed on GAS isolates as described (7). Briefly, whole
genomic DNA was restricted with SmaI, and fragments were
separated in a CHEF DRIII unit (BioRad, Glattbrugg, Switzerland)
under the following conditions: 0.5xTRIS-borate-EDTA running
buffer, 6 volts/cm, 14°C, 120°C angle, and a 1.2- to 54-s ramped
switch time for 18 hours. Gels were stained with bromide, and
banding patterns were compared. Only isolates with identical
banding patterns were considered to belong to the same clone.

Figure. Pulsed-field gel electrophoresis of group A streptococci
causing clinical infection in 19 drug users, Berne, Switzerland.

The M-type of the outbreak strain was determined by PCR
amplification and sequencing of a region of the emm gene (3,8).
The presence of the pyrogenic exotoxin A gene was evaluated as
described (3).

PFGE analysis of isolates showed that 19 of the 21 infections were
caused by the same clone (Figure). Most of these infections (16 of
19) were needle abscesses at the injection site; two were
complicated by erysipelas and one by osteomyelitis distant from
the needle abscess. None of the patients had streptococcal toxic
shock syndrome (1). Seventeen cases required inpatient treatment,
including surgery; all patients recovered. All patients lived (n = 16)
or purchased drugs in Berne (n = 3).

Cases and controls did not differ by mean age, current employment
status, sharing of paraphernalia (e.g., needles, spoons, and filters)
for drug use, and place of shelter (Table). No GAS could be
isolated from the eight confiscated cocaine samples. Five (9%) of
55 controls were colonized by GAS at the time of the interview;
three of them carried the outbreak strain. Unfortunately, four of the
five colonized controls refused to disclose the location and
nationality of their drug dealer(s). One admitted buying drugs in
various places, including place X, and from dealers of several
nationalities, including nationality A. He was colonized by the
outbreak strain.

The outbreak strain was M-type 25, which did not carry the gene
encoding for the pyrogenic exotoxin A. The strain was susceptible
to penicillin, clindamycin, and erythromycin, as determined by E-
testing.

Conclusions

During the past decade, GAS infections have attracted increased
attention because of their worldwide reemergence. Molecular
typing studies suggested that this might have been due to the
intercontinental spread of a virulent clone of M-type 1 (3). Isolates
of other M-types have since been associated with clusters of severe
infections (9). Person-to-person spread by respiratory droplets from
colonized patients or asymptomatic carriers has been thought to be
the main mode of transmission. Vaginal carriage of GAS in health-
care workers has been associated with nosocomial spread of GAS
(10). Acquisition of GAS by contaminated food and foodborne
GAS outbreaks has also been described (11).

The epidemic of GAS infections among drug users described here
is to our knowledge the first report of a clonal epidemic with GAS
in this patient population. GAS is commonly isolated from soft-
tissue infections in intravenous drug users (12,13). High rates of
pharyngeal carriage with epidemic clones may play a role in this
high frequency of GAS infections, although no clonality of such
GAS isolates was found in two studies (14,15).

The case-control study showed a strong association between
infection and purchase of drugs at X, a place commonly used for
drug dealing in Berne. The association with place X was strong
enough that bias due to the high refusal rate among controls is
unlikely.

Outbreaks of infections with other pathogens were previously
described in drug users (e.g., C. tetani, C. botulinum, Candida
albicans, and hepatitis A virus) (16-20). These outbreaks could be
due to contamination of the drug or drug paraphernalia. In our
population, sharing paraphernalia such as needles, spoons, and
filters was reported infrequently and did not differ between cases
and controls. We believe that this strain of GAS was spread
through cocaine or its containers. Drug dealers and users often hide
cocaine in their mouths during police raids. Therefore, GAS may
be spread to drug users by contamination of the plastic bags
containing the cocaine or the cocaine itself from persons with GAS
colonization of the mouth and throat. The strong association of the
outbreak with a common place of drug purchase suggests that one
or several drug dealers there were colonized by the outbreak strain
and spread it by contaminated drug containers or respiratory
droplets. Infected drug users may also have become colonized by
hiding the drug containers in their mouths. Alternatively, GAS may
have entered the subcutaneous tissue directly by contaminated drug
or handling paraphernalia with contaminated hands. Although we
were not able to culture GAS from cocaine samples, the drug
cannot be ruled out as a potential vehicle of GAS, since these
samples had been stored for some weeks and GAS may not survive
in sufficient numbers in dry (cocaine) powder.

We obtained a throat culture from only one of the infected drug
users, because they had already received antibiotics for at least 1-2
days when they were identified by the microbiologic studies. The
single throat culture showed carriage of the outbreak strain in the
throat as well as at the site of infection (erysipelas of the leg).
Oropharyngeal carriage of the outbreak strain was also found in
5.4% of our controls; the strain was not found in randomly selected
isolates from nondrug users, indicating that the outbreak strain was
probably circulating mainly in the drug user population. Needle
abscesses in drug users have been associated with cocaine use. The
local vasoconstriction induced by cocaine may predispose to
abscess formation (13,21). Cocaine, unlike heroin, is usually not
heated before injection, since it is thought to lose its activity when
heated. Failure to heat the drug likely increases the risk of
inoculating pathogens. In our study, none of the 19 cases heated the
dissolved cocaine, while six of the controls did. Some reported that
they started to heat the drug after hearing about the outbreak.

From the end of the outbreak in November 1997 to May 1998, we
observed three sporadic cases of infection due to the outbreak
strain among drug users, but not the general population. A
retrospective analysis by PFGE of clinical GAS isolates cultured in
our institution demonstrated that the outbreak strain has been
circulating among drug users in Berne since at least February 1997.
This study also revealed two previous clonal GAS outbreaks
among drug users in 1993 (unpubl. obs.). These findings suggest
that GAS outbreaks may be observed among drug users more
frequently than previously appreciated and that their propagation
may involve transmission of the outbreak clones by mechanisms
related to drug purchase and use. 

Acknowledgments

The authors thank Susanne Aebi for performing the PFGE and the
staff of the Contact organization, Berne, Switzerland, for
supporting the case-control study.

Dr. Böhlen is a specialist in internal medicine with clinical
expertise in HIV and infectious diseases. At the time of this study,
he was a clinical fellow in the division of infectious diseases of the
university hospital of Berne, Switzerland. He is now specializing in
dermatology, with a focus on skin infections.

Address for correspondence: Kathrin Mühlemann, Institute of
Medical Microbiology, Friedbühlstrasse 51, CH-3010 Berne,
Switzerland; fax: 41-31-632-3550; e-mail: muehlemann@
imm.unibe.ch.


References

1. Stevens DL. Invasive group A streptococcus infections. Clin
Infect Dis 1992;14:2-11.
2. Cleary PP, Kaplan EL, Handley JP, Wlazlo A, Kim MH, Hauser
AR, et al. Clonal basis for resurgence of serious Streptococcus
pyogenes disease in the 1980s. Lancet 1992;339:518-21.
3. Musser JM, Kapur V, Szeto J, Pan X, Swanson DS, Martin DR.
Genetic diversity and relationships among Streptococcus pyogenes
strains expressing serotype M1 protein: recent intercontinental
spread of a subclone causing episodes of invasive disease. Infect
Immun 1995;63:994-1003.
4. Johnson DR, Stevens DL, Kaplan EL. Epidemiologic analysis of
group A streptococcal serotypes associated with severe systemic
infections, rheumatic fever, or uncomplicated pharyngitis. J Infect
Dis 1992;166:374-82.
5. Schwartz B, Elliott JA, Butler JC, Simon PA, Jameson BL,
Welch GE, et al. Clusters of invasive group A streptococcal
infections in family, hospital, and nursing home settings. Clin
Infect Dis 1992;15:277-84.
6. Kurzynski TA, Van Holten CM. Evaluation of techniques for
isolation of group A streptococci from throat cultures. J Clin
Microbiol 1981;13:891-4.
7. Burki D, Bernasconi C, Bodmer T, Telenti A. Evaluation of the
relatedness of strains of Mycobacterium avium using pulsed-field
gel electrophoresis. Eur J Clin Microbiol Infect Dis 1995;14:212-7.
8. Beall B, Facklam R, Hoenes T, Schwartz B. Survey of emm
gene sequences and T-antigen types from systemic Streptococcus
pyogenes infection isolates collected in San Francisco, California;
Atlanta, Georgia; and Connecticut in 1994 and 1995. J Clin
Microbiol 1997;35:1231-5.
9. Schwartz B, Facklam RR, Breiman RF. Changing epidemiology
of group A streptococcal infection in the USA. Lancet
1990;336:1167-71.
10. Berkelman RL, Martin D, Graham DR, Mowry J, Freisem R,
Weber JA, et al. Streptococcal wound infections caused by a
vaginal carrier. JAMA 1982;247:2680-2.
11. Rammelkamp CH Jr. Food-borne streptococcal epidemics. N
Engl J Med 1969;280:953-4.
12. Summanen PH, Talan DA, Strong C, Mc Teague M, Bennion
R, Thompson JE Jr, et al. Bacteriology of skin and soft-tissue
infections: comparison of infections in intravenous drug users and
individuals with no history of intravenous drug use. Clin Infect Dis
1995;20 Suppl 2:S279-82.
13. Bergstein JM, Baker EJt, Aprahamian C, Schein M, Wittmann
DH. Soft tissue abscesses associated with parenteral drug abuse:
presentation, microbiology, and treatment. Am Surg 1995;61:1105-
8.
14. Navarro VJ, Axelrod PI, Pinover W, Hockfield HS, Kostman
JR. A comparison of Streptococcus pyogenes (group A
streptococcal) bacteremia at an urban and a suburban hospital. The
importance of intravenous drug use. Arch Intern Med
1993;153:2679-84.
15. Lentnek AL, Giger O, E OR. Group A beta-hemolytic
streptococcal bacteremia and intravenous substance abuse. A
growing clinical problem? Arch Intern Med 1990;150:89-93.
16. Passaro DJ, Werner SB, McGee J, Mac Kenzie WR, Vugia DJ.
Wound botulism associated with black tar heroin among injecting
drug users. JAMA 1998;279:859-63.
17. Clemons KV, Shankland GS, Richardson MD, Stevens DA.
Epidemiologic study by DNA typing of a Candida albicans
outbreak in heroin addicts. J Clin Microbiol 1991;29:205-7.
18. Servant JB, Dutton GN, Ong-Tone L, Barrie T, Davey C.
Candidal endophthalmitis in Glaswegian heroin addicts: report of
an epidemic. Transactions of the Ophthalmological Society of the
United Kingdom 1985;104:297-308.
19. Sundkvist T, Johansson B, Widell A. Rectum carried drugs
may spread hepatitis A among drug addicts. Scand J Infect Dis
1985;17:1-4.
20. Sun KO. Outbreak of tetanus among heroin addicts in Hong
Kong. J R Soc Med 1994;87:494-5.
21. Hoeger PH, Haupt G, Hoelzle E. Acute multifocal skin
necrosis: synergism between invasive streptococcal infection and
cocaine-induced tissue ischaemia? Acta Derm Venereol
1996;76:239-41.

Dispatches

Erythromycin Resistance in Streptococcus pyogenes in Italy

Matteo Bassetti,* Graziana Manno,* Andrea Collidà,* Alberto
Ferrando,* Giorgio Gatti,* Elisabetta Ugolotti,* Mario Cruciani,†
and Dante Bassetti*
*University of Genoa, G. Gaslini Children's Hospital, Genoa, Italy;
and †Department of Infectious Diseases, Verona, Italy

In a prospective study of acute pharyngitis in Italian children, 69
(38.3%) of 180 isolates of Streptococcus pyogenes were resistant
to macrolides. S. pyogenes was eradicated in 12 (63.1%) of 19
patients with erythromycin-resistant S. pyogenes treated with
clarithromycin and in 22 (88%) of 25 patients with erythromycin-
susceptible strains. The constitutive-resistant phenotype was
correlated with failure of macrolide treatment.

Over the last decade, severe infections due to Streptococcus
pyogenes and its complications have reemerged in several parts of
the world (1-3). S. pyogenes is uniformly susceptible to penicillin,
which remains the drug of choice for treating infections by this
organism. Erythromycin and other macrolides have been
recommended as alternative treatment for patients allergic to
penicillin (2,4); however, resistance to erythromycin and related
drugs in S. pyogenes has become widespread (5). Resistance to
erythromycin was first described in 1955 in the United Kingdom
(6) and, more recently, has been reported in Japan (7), Finland (8),
Taiwan (9), Australia (10), the United States (11), Spain (12,13),
and Italy (14-16).

From 1991 to 1996 in Genoa, the percentage of S. pyogenes
resistant or with intermediate resistance to erythromycin increased
from 0% to 50% (17). This abrupt increase in the rate of
erythromycin-resistant strains is of concern, since erythromycin has
been effective against most S. pyogenes isolates.

We investigated the prevalence and distribution of macrolide
resistance phenotypes among S. pyogenes and carried out a clinical
study in patients with S. pyogenes pharyngitis to correlate clinical
and microbiologic outcomes with in vitro susceptibility patterns.

The Study

Ten pediatricians in Genoa (population 700,000) participated in
this study. Children included in the study had to have two or more
of the following signs and symptoms: oropharyngeal erythema,
fever and sore throat, tonsillar exudate or cervical lymphadenitis,
and strawberry tongue. S. pyogenes was confirmed by culture of
throat swabs in agar blood; ß-hemolytic colonies were identified as
S. pyogenes by the bacitracin disk (Difco Laboratories, Detroit,
MI) and latex-agglutination test (Streptex, Wellcome, U.K.).

Minimum inhibitory concentrations (MICs) for penicillin,
cefixime, ceftriaxone, chloramphenicol, rifampin, tetracycline,
trimethoprim/sulfamethoxazole, and vancomycin were determined
by using the PASCO MIC gram-positive panel (Difco
Laboratories, Detroit, MI), supplemented with horse blood. MICs
for clindamycin, erythromycin, azithromycin, and clarithromycin
were determined by using E-test strips (AB Biodisk, Solna,
Sweden) on Mueller-Hinton agar supplemented with 5% horse
blood incubated in an atmosphere containing 5% carbon dioxide.
Phenotypes of macrolide resistance were differentiated according
to the description of Seppala et al. (18) and Suttcliffe et al. (19).
Resistance to both erythromycin and clindamycin indicated a
constitutive type of resistance (CR), blunting of the clindamycin
zone of inhibition proximal to erythromycin indicated an inducible
type of resistance (IR), and susceptibility to clindamycin without
blunting indicated the M-phenotype of resistance. For all
antibiotics tested, the breakpoints suggested by the National
Committee for Clinical Laboratory Standards were used (20,21).
At their physicians' discretion, eligible patients received a 10-day
course of one of the following drugs: amoxicillin 75 mg/kg three
times a day; amoxicillin/clavulanic acid 50 mg/kg twice a day;
cefaclor 50 mg/kg twice a day; or clarithromycin 15 mg/kg twice a
day. The attending physician was blinded to the results of
microbiologic tests. Fisher's exact test and the chi-square test were
performed by using Epi Info, version 6. For all tests, a p value of <
0.05 was considered statistically significant.

Table 1. In vitro susceptibility a of Streptococcus pyogenes from
180 pharyngitis patients

Table 2. Frequency of clarithromycin failure, by susceptibility
profile

Six hundred children ages 1-13 years (median age 7.0) with acute
pharyngitis were observed, and 180 (30%) whose throat cultures
were positive for S. pyogenes were included in the study.
Amoxicillin was prescribed to 42 patients, amoxicillin/clavulanic
acid to 56, cefaclor to 35, and clarithromycin to 44. The clinical
cure rates were 79.5% (35 of 44) in the clarithromycin group, 92%
(39 of 42) in the amoxicillin group (p = 0.14 for comparison with
clarithromycin), 100% (56 of 56) in the amoxicillin/clavulanic acid
group (p = 0.0003 for comparison with clarithromycin), and 97.1%
(34 of 35) in the cefaclor group (p = 0.03 for comparison with
clarithromycin).

Results of post-treatment throat swabs were available from 159
patients. Bacterial eradication response rates were 77.2% (34 of 44)
with clarithromycin, 88.8% (32 of 36) with amoxicillin group (p =
0.28 for comparison with clarithromycin), 95.8% (46 of 48) with
amoxicillin/clavulanic acid (p = 0.03 for comparison with
clarithromycin), and 90.3% (28 of 31) with cefaclor (p = 0.24 for
comparison with clarithromycin). All 180 strains were susceptible
to penicillin (MIC 90 <0.06 ˜g/l) and other ß-lactams tested (Table
1). Overall, 69 (38.3%) of the 180 isolates were resistant to one or
more macrolides, 7 (3.9%) to clindamycin, and 21 (11.6%) to the
16-member macrolide rokitamycin. Sixty-two percent (43 of 69
strains) of the erythromycin-resistant strains showed the M
phenotype of resistance, 11.5% (8 strains) the CR phenotype, and
26.0% (18 strains) the IR phenotype.

Among the 159 patients, 19 (43.1%) of 44 treated with
clarithromycin, 16 (44.4%) of 36 treated with amoxicillin, 13
(27.0%) of 48 treated with amoxicillin/clavulanic acid, and 8
(25.8%) of 31 treated with cefaclor had S. pyogenes resistant to
erythromycin at the first swab collected before treatment.

S. pyogenes was eradicated in 12 (63.1%) of the 19 patients with
erythromycin-resistant isolates and in 22 (88.0%) of 25 patients
with erythromycin-susceptible isolates treated with clarithromycin
(p = 0.07). As a control, the results of ß-lactam treatment were also
studied. The rates of microbiologic eradication in patients with
erythromycin-resistant isolates were 87.5% (14 of 16) for
amoxicillin, 100% (13 of 13) for amoxicillin/clavulanic acid, and
100% (8 of 8) for cefaclor. Rates of microbiologic eradication for
erythromycin-susceptible strains were 90% (18 of 20) for
amoxicillin, 94.2% (33 of 35) for amoxicillin/clavulanic acid, and
86.9% (20 of 23) for cefaclor (p = 1.0; p = 1.0; p = 0.54,
respectively, for comparison with erythromycin-resistant isolates).

In clarithromycin-treated patients, 6 of the 7 treatment failures
were related to isolates with a CR phenotype (p = 0.0002 for
comparison of percentages with other phenotypes of resistance, and
p = 0.0001 for comparison with erythromycin-susceptible isolates)
(Table 2).

Conclusions

Our results show that in Genoa, 38% of S. pyogenes isolated from
pharyngitis patients are erythromycin resistant. Sixty-three percent
of such isolates belonged to a recently reported, noninducible M
phenotype, described as having low-level resistance to
erythromycin and sensitivity to clindamycin and 16-member
macrolides (18). Twenty-six percent of resistant strains were
classified as IR phenotype, characterized by low-level resistance to
erythromycin and inducible resistance to 16-member macrolides
and clindamycin after exposure to subinhibitory concentrations of
erythromycin (5). The remaining isolates (11.5%) showed the CR
phenotype, characterized by high-level resistance to macrolides and
clindamycin.

Our study also examined whether in-vitro resistance could be a
good predictor of clinical outcome in children with pharyngitis.
Although physicians were instructed to choose the antibiotic
without regard to clinical signs and symptoms, a bias due to
selective antibiotic choice based on clinical presentation cannot be
excluded. Clarithromycin was prescribed to 44 patients, 19 with
erythromycin-resistant isolates and 25 with erythromycin-
susceptible isolates. Although the rate of microbiologic eradication
did not differ between patients with erythromycin-resistant isolates
and those with erythromycin-susceptible isolates (63.1% vs.
88.0%; p = 0.07), a clear trend was observed toward a higher rate
of eradication among erythromycin-susceptible isolates. When
results in clarithromycin-treated patients were analyzed by
phenotype of resistance, the rate of treatment failure was 100% (6
patients) for CR phenotype, compared with 7.6% (1 of 13 patients)
for other phenotypes (p = 0.0002) and 12% (3 of 25 patients) for
erythromycin-susceptible isolates (p = 0.0001). Failure of
erythromycin to eradicate group A streptococci with high levels of
resistance to erythromycin and lyncosamide has been reported
(22,23). Seppala et al. (8), in a retrospective analysis of medical
records, found that erythromycin failed in 47% of pharyngitis
patients with erythromycin-resistant isolates, a rate significantly
higher than the 4% observed in patients with erythromycin-
susceptible isolates. The susceptibility profile of these strains,
however, was consistent with phenotypes other than CR. The
eradication rate in patients with isolates belonging to phenotypes
other than CR, thus showing low levels of resistance to macrolides,
was comparable with that observed for erythromycin- susceptible
isolates. However, our findings suggest that CR phenotype will be
an accurate predictor of in-vivo failure of macrolides in the
treatment of streptococcal pharyngitis. Whether the discrepancy
between our results and those of a previous Finnish study (8)
should be attributed to differing macrolides remains to be proven
by large, well-controlled studies. Despite in-vitro cross-resistance
with other 14-member macrolides, clarithromycin is characterized
by elevated concentrations attained in different tissues (including
tonsil tissue) because of its improved pharmacokinetic profile
(5,24). Because only a few alternative antimicrobial agents can be
used to treat pharyngitis in patients allergic to ß-lactams, adequate
interventions include a controlled use of macrolides and
surveillance for the susceptibility of group A streptococci.
Determining erythromycin resistance phenotypes seems to be a
useful tool, particularly in areas where macrolides are frequently
prescribed. Should the CR phenotype, reported infrequently at
present, become prevalent, its high-level resistance may threaten
the efficacy of macrolides and clindamycin in the treatment of
streptococcal pharyngitis.

Acknowledgments

The authors thank Antonio Di Biagio, Enrico Mantero, Sandra
Ratto, and Maria Luisa Belli for their helpful comments and
Roberto Ceccarelli for his review of the manuscript.

Dr. M. Bassetti is a physician in the Department of Infectious
Diseases, University of Genoa, Italy. He works in Gaslini
Children's Hospital, the largest children's hospital in Italy, and in
San Martino teaching hospital. His research focuses on the clinical
importance of antibacterial resistance and the application of cost-
effective policies for antimicrobial use.

Address for correspondence: Matteo Bassetti, Infectious Diseases
Department, University of Genoa, G. Gaslini Children's Hospital,
Largo G. Gaslini 5, 16147 Genoa, Italy;

References

1. Bisno AL. Group A streptococcal infections and acute rheumatic
fever. N Engl J Med 1991; 325:783-93.
2. Kaplan EL. The resurgence of group A streptococcal infections
and their sequelae. Eur J Clin Microbiol Infect Dis 1991;10:55-7.
3. Moses AE, Ziv A, Harari M, Rahav G, Shapiro M, Englehard D.
Increased incidence and severity of Streptococcus pyogenes
bacteremia in young children. Pediatr Infect Dis J 1995;14:767-70.
4. Peter G. Streptococcal pharyngitis: current therapy and criteria
for evaluation of new agents. Clin Infect
5. Schito GC, Pesce A, Marchese AJ. The role of macrolides in
Streptococcus pyogenes pharyngitis. J Antimicrob Chemother
1997;39:562-5.
6. Lowbury EJL, Hurst L. The sensitivity of staphylococci and
other wound bacteria to erythromycin, oleandomycin and
spiramycin. J Clin Pathol 1959;12:163-4.
7. Maruyama S, Yoshiota H, Fujita K, Takimoto M, Satake Y.
Sensitivity of group A streptococci to antibiotics. Am J Dis Child
1979;133:151-4.
8. Seppala H, Nissinen A, Jarvinen H, Huovinen S, Henriksson T,
Herva E, et al. Resistance to erythromycin in group A streptococci.
N Engl J Med 1992;326:292-7.
9. Hsueh PR, Chen HM, Huang AH, Wu JJ. Decreased activity of
erythromycin against Streptococcus pyogenes in Taiwan.
Antimicrob Agents Chemother 1995;39:2239-42.
10. Stingemore N, Francis GRJ, Toohey M, McGechie DB. The
emergence of erythromycin resistance in Streptococcus pyogenes
in Fremantle, Western Australia. Med J Aust 1989;150:626-31.
11. Gentry JL, Burns WW. Antibiotic resistant streptococci. Am J
Dis Child 1980;133:801.
12. Garcia-Bermejo I, Cacho J, Orden B, Alos J, Gomez-Garces
JL. Emergence of erythromycin-resistant, clindamycin-susceptible
Streptococcus pyogenes isolates in Madrid, Spain. Antimicrob
Agents Chemother 1998;42:989-90.
13. Orden B, Perez-Trallero E, Montes M, Martinez R.
Erythromycin resistance of Streptococcus pyogenes in Madrid.
Pediatr Infect Dis J 1998;17:470-3.
14. Cornaglia G, Ligozzi M, Mazzariol A, Valentini M, Orefici G,
Fontana R. Rapid increase of resistance to erythromycin and
clindamycin in Streptococcus pyogenes in Italy, 1993-1995. Emerg
Infect Dis 1996;2:339-42.
15. Cocuzza C, Blandino G, Mattina R, Nicoletti F, Nicoletti G.
Antibiotic susceptibility of group A streptococci in 2 Italian cities:
Milano and Catania. Microb Drug Resist 1997;3:379-84.
16. Cornaglia G, Ligozzi M, Mazzariol A. Resistance of
Streptococcus pyogenes to erythromycin and related antibiotics in
Italy. Clin Infect Dis 1998;27 Suppl 1;S87-92.
17. Bassetti M, Mantero E, Gatti G, Di Biagio A, Bassetti D.
Streptococcus pyogenes erythromycin resistance in Italy. Emerg
Infect Dis 1999;5:302-3.
18. Seppala H, Nissinen A, Yu Q, Houvinen P. Three different
phenotypes of erythromycin-resistant Streptococcus pyogenes in
Finland. J Antimicrob Chemother 1993;32:885-91.
19. Sutcliffe J, Tait-Kamradt A, Wondrack L. Streptococcus
pneumoniae and Streptococcus pyogenes resistant to macrolides
but sensitive to clindamycin: a common resistance pattern
mediated by an efflux system. Antimicrob Agents Chemother
1996;40:1817-24.
20. National Committee for Clinical Laboratory Standard
Performance Standards for antimicrobial disk susceptibility tests.
Document M2- A6. Villanova (PA): The Committee; 1997.
21. National Committee for Clinical Laboratory Method for
dilution for antimicrobial susceptibility tests for bacteria that grow
aerobically. Document M7- A4. Villanova (PA): The Committee;
1997.
22. Sanders E, Foster MT, Scott D. Group A beta-hemolytic
streptococci resistant to erythromycin and lincomycin. N Engl J
Med 1968;278:538-40.
23. Phillips G, Parratt D, Orange GV, Harper I, McEwan H, Young
N. Erythromycin resistant Streptococcus pyogenes. J Antimicrob
Chemother 1990;25:723-4.
24. Lode H, Boeckh M, Scaberg T. Human pharmacokinetics of
macrolide antibiotics. In: Macrolides. Chemistry, pharmacology
and clinical uses. Bryskier A, Butzler JP, Neu HC, Tulkens PM,
editors. Paris: Arnette Blackwell; 1993; p. 409-20.

Dispatches

Sin Nombre Virus (SNV) Ig Isotype Antibody Response during
Acute and Convalescent Phases of Hantavirus Pulmonary
Syndrome

Pavel Bostik,* Jorn Winter,*† Thomas G. Ksiazek,† Pierre E.
Rollin,† Francois Villinger,* Sherif R. Zaki,† C.J. Peters,† and
Aftab A. Ansari*
*Emory University School of Medicine, Atlanta, Georgia, USA;
and †Centers for Disease Control and Prevention, Atlanta, Georgia,
USA

Serum samples from 22 hantavirus pulmonary syndrome (HPS)
patients were tested for Sin Nombre virus (SNV)-reactive
antibodies. In the acute phase of HPS, 100% and 67% of the
samples tested positive for SNV-specific immunoglobulin (Ig) M
and IgA, respectively. Among the virus-specific IgG antibodies, the
most prevalent were IgG3 (in 97% of samples), followed by IgG1
(70%), IgG2 (30%), and IgG4 (3%).

Hantaviruses are associated with hemorrhagic fever with renal
syndrome (HFRS) and hantavirus pulmonary syndrome (HPS) in
humans (1). Sin Nombre virus (SNV), a newly identified
hantavirus, was identified as the causative agent in a 1993 outbreak
of HPS with a >60% case-fatality rate in the southwestern United
States (2). SNV and other American hantaviruses associated with
HPS have a specific rodent reservoir; transmission was inferred to
occur primarily by inhalation of infected aerosols from the rodent
urine and excreta (3,4). However, an outbreak with human-to-
human transmission of HPS caused by Andes virus has also been
reported from South America (5). HPS is clinically characterized
as an acute febrile illness associated with headache, malaise, and
myalgia that proceeds to thrombocytopenia, pulmonary edema, and
hypotension or shock (1,2). In fatal cases, death usually occurs 1 to
2 days after onset of respiratory symptoms. Although SNV-specific
CD4 + and CD8 + T lymphocytes have been documented (6), the
virus-specific humoral immune response has not yet been well
defined. The different patterns of increase in immunoglobulin (Ig)
classes are associated with Th1- or Th2-type immune responses in
mice (7). In some parasitic diseases, increased antigen-specific
IgG4 subclass levels correspond to a variable extent with the
induction of antigen-specific T-cell anergy, sometimes associated
with serious or disseminated disease (8,9). Antibody responses to
polysaccharide antigens have been reported to be predominantly
IgG1/IgG2 (10). In viral diseases, IgG1 and IgG3 subclasses were
detected predominantly in primary infections, while IgG2 and IgG4
subclasses were more characteristic of recurrent infections (11-14).
Ig class- and subclass-specific titers of antibodies to SNV, their
kinetic appearance, and their role in preventing disease or death in
HPS patients has not been reported. To address this issue, we
developed SNV-specific, immunoglobulin class/subclass-specific
enzyme-linked immunosorbent assays and evaluated the levels of
SNV-specific IgA, IgM, and IgG subclass antibodies in the serum
samples from HPS patients. Ig class- and subclass-specific titers
were compared in the sera of patients who survived and those who
died.

The Study

Thirty-three serum samples were obtained from 22 patients
hospitalized with HPS in 1993. Patients were diagnosed as having
HPS on the basis of results from immunohistochemistry analysis,
alone or in combination with serologic studies or reverse
transcription-polymerase chain reaction (16). The samples were
divided into three groups with respect to the time of collection:
acute phase (1-8 days posthospitalization, 12 samples), early
convalescent phase (9-30 days posthospitalization, 9 samples), and
late convalescent phase (31 days to 1.5 years posthospitalization,
12 samples). Human Ig antibodies were captured by affinity-
purified, polyclonal, monospecific sheep antihuman IgM, IgG1,
IgG2, IgG3, IgG4, and IgA reactive antibodies (IgM: Biosource,
Camarillo, CA; IgG1, IgG2, IgG3, IgG4, IgA: The Binding Site,
San Diego, CA). The capture antibodies were coated onto 96-well
high-binding microtiter plates (Corning Costar Corporation,
Charlotte, NC) at 4°C for a minimum of 16 hours (0.5 mg/ml in
25mM sodium carbonate buffer pH 9.6; 0.1 ml/ well). Sera and
detection antibodies were diluted in PBS, pH 7.4 to 7.6, containing
0.05% Tween 20 supplemented with 4% milk (PBSM). Initially, all
plates were washed and blocked with PBSM for 30 minutes at
room temperature. Fourfold dilutions of the plasma samples in a
volume of 0.1 ml were incubated for 1 hour at 37°C, then the plates
were washed and SNV-infected Vero E6 cell slurries (CDC,
Atlanta, GA; SPR 268) and uninfected control slurries (CDC,
Atlanta, GA; SPR 391), diluted 1:10, were added to appropriate
wells. After incubation for 1 hour at 37°C, the plates were washed
and incubated with a polyclonal rabbit anti-SNV serum (CDC,
Atlanta, GA; diluted 1:1000, 0.1 ml/well) as described. Then the
plates were washed and an alkaline phosphatase-conjugated sheep
Ig anti-rabbit Ig (Southern Biotechnology Associates, Birmingham,
AL) diluted 1:4000 (0.1 ml/well) was added to each well. After
incubation for 1 hour at 37°C, the plates were washed and the
substrate p-nitrophenylphosphate (Bio-Rad, Hercules, CA) was
added. The reaction was stopped by the addition of 0.1 ml 0.4 M
NaOH, and the optical density was read at 405 nm.

Where applicable, antibody titers are presented as geometric mean
titers (GMT), with the range of positive titers. Antibody titers for
significant differences between survivors and patients who died in
the acute-phase group were compared by unpaired Student's t-test.
SNV-specific IgM, which had the highest titers overall (GMT
15,097, range 1,600-102,400), was detected in all acute-phase
samples and 89% of samples from the early convalescent phase
(Table). SNV-specific IgM was also detected in one sample from
the late convalescent phase (titer 25,600 at day 67), a phenomenon
occasionally observed in other HPS convalescent-phase samples
(data not shown). SNV-specific serum IgA was detected in 67% of
acute-phase samples and 78% of samples from the early
convalescent phase, but the titers were generally lower (GMT
9,406, range 100 to 409,600) than the IgM titers. Two of the 12
patients during the late convalescent phase had detectable virus-
specific IgA, although the titers were relatively low (1,600 and
100). Patients who had very low or undetectable SNV-specific IgA
generally also had lower titers of SNV-specific IgM during the
acute phase of illness.

In the IgG class, the highest SNV-specific antibody titers were
predominantly of the IgG3 subclass (GMT 4,935, range 100-
409,600), detected in 97% of samples. SNV-specific IgG1
antibodies were also detected in 76% of samples, but in much
lower titers (GMT 447, range 100- 6,400). SNV-specific IgG2
antibodies were detected in only 30% of samples (GMT 348, range
100-6,400), and only one serum tested positive for SNV-specific
IgG4, with a very low titer (1:100). The Ig class- and subclass-
specific antibody titers were further analyzed with regard to the
outcome of the disease. The 12 patients who had samples drawn
during the acute phase of infection included six who died and six
who survived. The SNV-specific antibody titers of these two
groups did not differ significantly in any Ig class or subclass tested
(p > 0.1 in each Ig class or IgG subclass group, data not shown).
When total IgG, IgA, and IgM concentrations were measured in 18
serum samples from randomly chosen HPS patients, 12% of the
IgM values (reference range 56-332 mg/dl), 39% of the IgA values
(reference range 70-312 mg/dl) and 56% of the IgG values
(reference range 639-1349 mg/dl) fell outside the normal reference
range. Interestingly, increased titers of both SNV-specific IgA and
IgM did not correlate with patients' levels of total serum IgA and
IgM. This lack of correlation suggests that the increase of SNV-
specific IgM and IgA antibodies is not due to nonspecific
polyclonal activation of the immune system. As expected, SNV-
specific IgM antibodies are observed during the initial phase of
disease; both SNV-specific serum IgA and IgM antibodies peaked
early during the acute phase. These levels were maintained during
both the acute and the early convalescent phases. SNV-specific
antibodies of IgG1 and IgG3 subclasses appeared during the first
10 days, coinciding with the appearance of symptoms. However,
high titers were maintained throughout early (IgG1) or early and
late convalescent phases (IgG3). In 5 (70%) of the 7 samples that
tested positive for SNV-specific IgG2 antibodies, these titers
appeared during the acute and early convalescent phases but were
lost thereafter.

Table. Total Ig concentrations and specific Sin Nombre virus Ig
titers in sera from patients with hantavirus pulmonary syndrome

Conclusions

We tested serum and plasma samples from HPS patients for the
presence of IgA and IgM class and the IgG subclass-specific
antibody titers against one of the SNV antigens. Although the
design of this study is cross sectional, we observed a characteristic
pattern of appearance of specific Ig classes/subclasses during the
course of infection. In addition, sequential samples from several
patients were included in the study (represented by the decimal
values in the Table) and the data from these samples show trends
similar to those observed in the composite with single samples.
SNV-specific IgM and IgA antibodies appeared in high titers
predominantly during the early phases of infection (in 95% and
71% of samples, respectively). In our IgG subclass-specific assays,
most samples contained IgG3 and IgG1 subclass-specific SNV-
reactive titers throughout illness. The detection of substantial IgG3
antibody titers early in hospitalization can be explained by the fact
that since infection occurred several days before onset of illness,
there was sufficient time to allow a switch to the IgG class to
occur. In contrast with other primary viral infections, a substantial
number of samples showed high titers of SNV antigen-reactive
antibodies belonging to the IgG3 subclass rather than the IgG1
subclass. It is possible that in SNV infection, specific sets of
cytokines may be produced that preferentially stimulate the
production of IgG3 subclass antibody. The similarity in antibody
titers between deceased and surviving patients is an important
observation. Regardless of the eventual outcome of infection, the
sera from patients during hospitalization with either HPS or HFRS
contained high levels of antibody. Abundant viral antigens were
present within endothelial cells of the pulmonary microvasculature,
as well as significant levels of CD8 + T-cell infiltrates and
evidence of circulating pro-inflammatory cytokines in serum.
These findings suggest that the disease is most likely secondary to
immunopathologic mechanisms (6,15,16). Thus, antibodies may
well play an important role in containing the initial viremic phase
of the infection, but T-cell activation may have an important role in
inducing disease. In addition, there may be epitopes of the virus
antigens that are not detected by the assays used in the studies
reported here and antibodies against these epitopes may distinguish
infected persons who do or do not become ill. Further studies to
identify differences in the viral epitopes recognized by sera from
patients who died compared with those who survived, as well as
comparison of antibody functionality, are needed to address these
issues. Finally, anti-SNV IgM antibodies were detected in 100% of
the patients in the acute phase of the infection in relatively high
titers (1,600 to 102,400). This finding confirms that the SNV-
specific IgM-capture enzyme immunoassay we describe can be
used as a valuable tool in the early diagnosis of HPS.

Acknowledgments

The authors thank the patients, health-care providers, and local and
state health departments, who have been crucial for the collection
of the samples. This work was supported by the Centers for
Disease Control and Prevention cooperative agreement # U50/
CCU411374-03-01.

Dr. Bostik is an instructor of pathology in the Department of
Pathology and Laboratory Medicine, Emory University School of
Medicine, Atlanta, GA. His interests focus on molecular
immunology and T-cell immunity in humans and nonhuman
primate models of viral infections.


Address for correspondence: A.A. Ansari, Winship Cancer Center,
Emory University School of Medicine, 1365B Clifton Rd, Atlanta,
GA 30322, USA; fax: 404-778-5016; e-mail: pathaaa@emory.edu.

References

1. Peters CJ, Simpson GL, Levy H. Spectrum of hantavirus
infection: hemorrhagic fever with renal syndrome and hantavirus
pulmonary syndrome. Annu Rev Med 1999;50:531-45.
2. Duchin JS, Koster FT, Peters CJ, Simpson GL, Tempest B, Zaki
SR, et al. Hantavirus pulmonary syndrome; clinical description of
disease caused by a newly recognized hemorrhagic fever virus in
the southwestern United States. N Engl J Med 1994;330:949-55.
3. Peters CJ. Hantavirus pulmonary syndrome in the Americas. In:
Scheld WM, Craig WA, Hughes JM, editors. Emerging Infections
II. Washington: American Society for Microbiology Press; 1998. p.
17-64.
4. Zeitz PS, Butler JC, Cheek JE, Samuel MC, Childs JE, Shands
LA, et al. A case-control study of hantavirus pulmonary syndrome
during an outbreak in the southwestern United States. J Infect Dis
1995;171:864-70.
5. Padula PJ, Edelstein A, Miguel SD, Lopez NM, Rossi CM,
Rabinovich RD. Hantavirus pulmonary syndrome outbreak in
Argentina: molecular evidence for person-to- person transmission
of Andes virus. Virology 1998;241:323-30.
6. Ennis FA, Cruz J, Siropoulou CF, Waite D, Peters CJ, Nichol
ST, et al. Hantavirus pulmonary syndrome: CD8+ and CD4+
cytotoxic T lymphocytes to epitopes on Sin Nombre virus
nucleocapsid protein isolated during acute illness. Virology
1997;238:380-90.
7. Snapper CM, Paul WE. Interferon-gamma and B cell stimulatory
factor-1 reciprocally regulate Ig isotype production. Science
1987;236:944-7.
8. Yazdanbakhsh M, Paxton WA, Kruize YCM, Sartono E,
Kurniawan A, van het Wout A, et al. T cell responsiveness
correlates differentially with antibody isotype levels in clinical and
asymptomatic filariasis. J Infect Dis 1993;167:925-31.
9. Ulrich M, Rodriguez V, Centeno M, Convit J. Differing
antibody IgG isotypes in the polar forms of leprosy and cutaneous
leishmaniasis characterized by antigen-specific T cell anergy. Clin
Exp Immunol 1995;100:54-8.
10. Rautonen N, Pelkonen J, Sipipnen S, Kayhty H, Makela O.
Isotype concentrations of human antibodies to group A
meningococcal polysaccharide. J Immunol 1986;137:2670-5.
11. Hashido M, Kawana T. Herpes simplex virus-specific IgM, IgA
and IgG subclass antibody responses in primary and nonprimary
genital herpes patients. Microbiol Immunol 1997;41:415-20.
12. Wagner DK, Muelenaer P, Henderson FW, Snyder MH,
Reimer CB, Walsh EE, et al. Serum immunoglobulin G antibody
subclass response to respiratory syncytial virus F and G
glycoproteins after first, second, and third infections. J Clin
Microbiol 1989;27:589-92.
13. Torfason EG, Reimer CB, Keyserling HL. Subclass restriction
of human enterovirus antibodies. J Clin Microbiol 1987;25:1376-9.
14. Forthal DN, Landucci G, Habis A, Zartarian M, Katz J, Tilles
JG. Measles virus-specific functional antibody responses and
viremia during acute measles. J Infect Dis 1994;169:1377-80.
15. Ksiazek TG, Peters CJ, Rollin PE, Zaki SR, Nichol S,
Spiropoulou C, et al. Identification of a new North American
hantavirus that causes acute pulmonary insufficiency. Am J Trop
Med Hyg 1995;52:117-23.
16. Zaki SR, Greer PW, Coffield LM, Goldsmith CS, Nolte KB,
Foucar K, et al. Hantavirus pulmonary syndrome: pathogenesis of
an emerging infectious disease. Am J Pathol 1995;146:552-79.


Dispatches

Bovine Tuberculosis and the Endangered Iberian Lynx

Victor Briones* Lucía de Juan,* Celia Sánchez,† Ana-Isabel
Vela,* Margarita Galka,‡ Natalia Montero,* Joaquín Goyache,*
Alicia Aranaz,* Ana Mateos,* and Lucas Domínguez*
*Facultad de Veterinaria, Universidad Complutense de Madrid,
Spain; †Parque Nacional de Doñana, Huelva, Spain; ‡Parque
Natural de Doñana, Huelva, Spain

We report the first case of bovine tuberculosis in a free-living
Iberian lynx (Lynx pardina), an extremely endangered feline, from
Doñana National Park in Spain. The isolate (Mycobacterium bovis)
correlates by molecular characterization with other isolates from
wild ungulates in the park, strongly suggesting an epidemiologic
link. Mycobacterium bovis infects many animal species, with wild
and free-ranging domestic ungulates being the main reservoirs in
nature (1).


Carnivores generally acquire the infection by eating infected food
(2,3), but reports of tuberculosis (TB) in wild carnivores are rare
(2-6). However, TB monitoring in free-living carnivores has
provided increased reports of the disease (7). TB does not pose a
serious threat to most wild carnivore populations but could have a
devastating effect in a small and divided population such as the
Iberian lynx. This species is the most endangered feline in the
world (8), with a declining number of animals living in reduced
and isolated areas in Spain and Portugal.

An adult male Iberian lynx, seen limping in August 1998, died in
October 1998. We could perform only direct and radiologic
examinations of the empty carcass. The right elbow joint was
enlarged, with fistulization. On X-ray, the lesion was diagnosed as
septic arthritis on the basis of radiolucent areas and sclerosis, along
with bony excrescences. A fragment from the right elbow joint, a
small fecal sample, and a nasal swab were collected and routinely
processed for detection of mycobacteria by auramine acid-fast
staining, culture, and polymerase chain reaction (PCR) (9-11).

Auramine staining of a smear from the elbow lesion was performed
by the method of Smithwick (9), and transmission fluorescence
microscopy at x400 showed acid-fast bacilli. Samples were
simultaneously processed for culture following standard
procedures: a suspension was obtained from a nasal swab, and the
other samples were homogenized in sterile distilled water in a
tissue homogenizer. The homogenates and suspension were then
decontaminated with hexadecylpyridinium chloride and
centrifuged, and sediments were inoculated onto Coletsos and
Löwenstein-Jensen media for mycobacteria; they were incubated at
37°C and inspected weekly for growth (10). Colonies were
examined for acid-fast bacilli by the Ziehl-Neelsen technique (9).
Identification procedures were performed as described elsewhere,
by means of direct PCR either on the processed samples or on
isolated colonies (10), and specific mycobacterial molecular
characterization was done by spoligotyping (11). The only positive
sample was the elbow joint: direct PCR was positive for M.
tuberculosis complex. An isolate consisting of acid-fast bacilli on
selective media was identified as M. bovis. Additionally, molecular
characterization by spoligotyping showed that the lynx isolate was
identical to other isolates from samples of wild ungulates living in
the park, including nine wild boars (Sus scrofa) and four fallow
deer (Dama dama) studied during the past 3 years.

Once TB has been confirmed in a free-living lynx, it is important
to determine whether a single animal has been affected or the
infection has spread within the population. From 1993, postmortem
examinations have been performed on 21 lynxes in conditions that
permitted reliable interpretation of pathologic findings. Any
granulomatous lesions found during necropsy, as well as fecal
samples taken regularly from temporarily or permanently captive
animals, were tested by acid-fast staining, direct PCR, and culture.
Radiologic screening has also been performed on captive and live-
captured animals. TB had never been diagnosed in lynxes before
this case. However, because of the insidious nature of the infection,
this case will not be the last; other lynxes could already be infected.
The prevalence of the infection in the species is extremely difficult
to determine because the number of lynxes is low and very few
animals can be monitored, since captures are limited to a
minimum. In addition, rapid, specific, and accurate in vivo
diagnosis of TB in lynxes is difficult. A reliable, rapid technique
for diagnosis is urgently needed to ensure that any captured animal
will be correctly diagnosed and treated.

Another important consideration is the role of wild ungulates and
domestic livestock as reservoirs. Populations of red deer (Cervus
elaphus), fallow deer, and wild boar live in the National Park.
Traditional farming in the area means that many domestic animals
(mainly free-ranging cattle) coexist with wild species. All these
species are susceptible to TB and could act as reservoirs (1).
Carnivore populations are generally considered spill-over hosts that
become incidentally infected and therefore are unlikely to maintain
the disease without an external source of infection (3,6,7). Because
eradication campaigns are conducted in domestic cattle, the
prevalence of bovine TB is low, but cattle are not TB free. Wild
ungulates have been monitored over the last 11 years; a few
isolated cases were detected before 1993, although the number of
cases has increased dramatically in recent years. This observation
may represent a real increase in prevalence or a higher intensity of
surveillance. Therefore, surveys of the prevalence of TB among
wild ungulates in the Park need to be enhanced to allow a more
accurate assessment of the problem. The isolate of M. bovis from
the lynx correlates exactly by spoligotyping with isolates obtained
recently from wild boars and fallow deer living in the same area,
suggesting a common source of infection. As fallow deer are
known to be part of the lynx's diet (12), these results are not
surprising and strongly suggest interspecies infection. However,
since no cultures from cattle are available, the molecular
comparison of strains isolated from wild and domestic animals is
not possible and a potential epidemiologic link among them cannot
be demonstrated. Nevertheless, TB transmission between domestic
cattle and wild ungulates sharing the same pastures has been
described (13). The different patterns of lesions in red and fallow
deer (mainly respiratory) and wild boars (digestive) suggest the
probable routes of infection, but the modes of transmission need to
be elucidated.

Infectious diseases affecting both domestic and wild animals
become a problem for both conservationists and farmers. In this
case, the concern is not only transmission between cattle and wild
ungulates, but transmission to the Iberian lynx, an endangered
species protected by strict conservation policies. Thus, health
requirements for domestic animals entering the park should be
strengthened to prevent TB, and livestock populations must be
managed and controlled, as a total ban seems unfeasible. The role
of lagomorphs in this case is speculative, as TB has not been
diagnosed in these species at the park. Although naturally acquired
TB seems to be extremely rare in rabbits, it is not uncommon in
hares (14). Since these animals are a basic item of the lynx's diet
(12), monitoring TB in lagomorphs must be included in a
prevention program so that a potential source of infection should
not be overlooked.

The detection of bovine TB in an Iberian lynx leads to another
question: what should be done with a free-living lynx with a
positive TB diagnosis? Intraspecies transmission seems quite
unlikely because of the rare contact between lynxes except during
mating season. However, basic rules of animal health dictate that
infected animals should be isolated to interrupt the transmission
cycle of many infectious diseases, including TB.

Anti-TB therapy in animals is controversial, and euthanasia is the
most commonly accepted option. However, the critical situation of
the Iberian lynx, with a declining population of under 1,000, makes
every animal essential. Multidrug, long-term, daily therapy has
been used in domestic cats (2,3), but extrapolating this treatment to
lynxes maintained in captivity would be difficult. Although therapy
must be the choice, adequate facilities and containment and
preventive measures must be in place to ensure that disease will
not spread during treatment. In addition, the likelihood of success
of treatment, which depends on the severity of the disease and the
general condition of the animal, must be evaluated in light of the
well-known collateral effects of antimycobacterial drugs (2,3).
Finally, even if treatment is successful, the animals will probably
not be able to live in the wild again because of possible physical or
behavioral changes.

In conclusion, survival of many species already depends on human
management of populations (15); in this case, every measure must
be adopted to ensure that bovine TB will not threaten the
endangered Iberian lynx. Dr. Briones is associate professor,
Departamento de Sanidad Animal, Facultad de Veterinaria,
Universidad Complutense de Madrid. His responsibilities include
clinical microbiology and infectious diseases (research and
teaching). His research interests focus on bacterial diseases of
wildlife and exotic pets in Spain.

Address for correspondence: Lucas Domínguez, Departamento de
Sanidad Animal, Facultad de Veterinaria, Universidad
Complutense de Madrid, 28040-Madrid, Spain; fax: 34-91- 394-
3908; e-mail: vbriones@eucmax.sim.ucm.es.

References

1. Hunter DL. Tuberculosis in free-ranging, semi free-ranging and
captive cervids. Rev Sci Tech 1996;15:171- 81.
2. Gunn-Moore DA, Jenkins PA, Lucke VM. Feline tuberculosis: a
literature review and discussion of 19 cases caused by an unusual
mycobacterial variant. Vet Rec 1996;138:53-8.
3. Greene CE, Gunn-Moore DA. Tuberculous mycobacterial
infections. In: Greene CE, editor. Infectious diseases of dog and
cat. 2nd ed. Philadelphia (PA): WB Saunders; 1998. p. 313-21.
4. Helman RG, Russell WC, Jenny A, Miller J, Payeur J. Diagnosis
of tuberculosis in two snow leopards using polymerase chain
reaction. J Vet Diagn Invest
1998;10:89-92.
5. Thorel MF, Karoui C, Varnerot A, Fleury C, Vincent V.
Isolation of Mycobacterium bovis from baboons, leopards and a
sea-lion. Vet Res 1998;29:207-12.
6. Artois M, Claro F, Remond M, Blancou J. Infectious pathology
of Canidae and Felidae in zoological parks. Rev Sci Tech
1996;15:115-40.
7. Keet DF, Kriek NPJ, Penrith M-L, Michel A, Huchzermeyer H.
Tuberculosis in buffaloes (Syncerus caffer) in the Kruger National
Park: spread of the disease to other species. Onderstepoort J Vet
Res 1996;63:239-44.
8. Nowell K, Jackson P, editors. Wild cats: status survey and
conservation action plan. IUCN Publications. Cambridge: The
Burlington Press; 1996.
9. Smithwick RW, editor. Laboratory Manual for acid-fast
microscopy. 2nd ed. Georgia: U.S. Department of Health,
Education and Welfare; 1976.
10. Aranaz A, Liebana E, Pickering X, Novoa C, Mateos A,
Dominguez L. Use of polymerase chain reaction in the diagnosis of
tuberculosis in cats and dogs. Vet Rec 1996;138:276-80.
11. Kamerbeek J, Schouls L, Kolk A, van Agterveld M, van
Soolingen D, Kuijper S, et al. Simultaneous detection and strain
differentiation of Mycobacterium tuberculosis for diagnosis and
epidemiology. J Clin Microbiol 1997;35:907-14.
12. Delibes M. Feeding ecology of the Spanish lynx in the Coto
Doñana. Acta Theriologica 1980;25:309-24.
13. Tweddle NE, Livingstone P. Bovine tuberculosis control and
eradication programs in Australia and New Zealand. Vet Microbiol
1994;40:23-39.
14. Gill JW, Jackson R. Tuberculosis in a rabbit: a case revised.
New Zealand Veterinary Journal 1993;41:147.
15. Kirkwood JK. Interventions for wildlife conservation, health
and welfare. Vet Rec 1993;132:235-8.

Dispatches

Haff Disease: From the Baltic Sea to the U.S. Shore

Udo Buchholz,*† Eric Mouzin,*† Robert Dickey,‡ Ron
Moolenaar,† Neil Sass,§ and Laurene Mascola*
*County of Los Angeles Department of Health Services, Los
Angeles, California, USA; †Centers for Disease Control and
Prevention, Atlanta, Georgia, USA; ‡Food and Drug
Administration, Dauphin Island, Alabama, USA; §Food and Drug
Administration, Laurel, Maryland, USA

Haff disease, identified in Europe in 1924, is unexplained
rhabdomyolysis in a person who ate fish in the 24 hours before
onset of illness. We describe a series of six U.S. patients from
1997 and report new epidemiologic and etiologic aspects.
Although Haff disease is traditionally an epidemic foodborne
illness, these six cases occurred in two clusters and as one sporadic
case.

In the summer and fall of 1924, physicians near the Königsberger
Haff shores along the Baltic coast recognized an outbreak of an
illness characterized by sudden, severe muscular rigidity (1-3). No
neurologic abnormalities, fever, splenomegaly, or hepatomegaly
were observed (1). Patients often had coffee-colored urine. The
clinical spectrum of disease varied, few patients died, and most
survivors recovered quickly. In the following 9 years, similar
outbreaks, affecting an estimated 1,000 persons, occurred
seasonally in the summer and fall along the coast of the "haff" (a
shallow lagoon). Recent ingestion of fish, usually cooked, was
common among those who became ill; species of fish included
burbot, eel, and pike. Seabirds and cats reportedly died after eating
fish in the wild. Because of the absence of fever and the fact that
fish had been cooked, known infectious causes were eliminated.
Several toxic etiologies were proposed but could not be confirmed.
Among these was arsenic poisoning (4), which is still cited in
modern medical dictionaries as the cause of Haff disease (5). From
1934 until 1984, other outbreaks resembling Haff disease were
described in Sweden and the Soviet Union (6-9). The first two
cases reported in the United States occurred in Texas in June 1984;
through 1996, only four more cases were reported: two from Los
Angeles, California, in 1985, and two from San Francisco,
California, in April 1986 (M. Tormey, pers. comm.). All patients
had eaten buffalo fish before onset of illness. Tests of the remains
of one of the fish meals suggested a neutral lipid as a causative
agent. Reports of six cases of Haff disease from California and
Missouri during a 5-month period (March-August) in 1997
prompted an investigation with the objectives of describing the
epidemiology and clinical characteristics of the 1997 U.S. cases of
Haff disease, tracing back implicated fish, and elucidating the
cause of Haff disease.

The Study

Based on the clinical description of the original cases from 1924 to
1933, we defined a case of Haff disease as illness in a person with
unexplained rhabdomyolysis who had eaten fish in the 24 hours
before onset of symptoms. The laboratory marker used to define
rhabdomyolysis was a fivefold or greater elevation in creatine
kinase (CK) levels, with a muscle/brain (MB) fraction <5% (10).
Cases were identified through county and state epidemiologists, as
well as the Food and Drug Administration (FDA) laboratories in
Bothell, Washington, and Dauphin Island, Alabama. We
interviewed all persons with cases reported to the Los Angeles
County Health Department or FDA. We reviewed medical records
for demographic information, medical history, course of illness,
and food exposures, including method of preparation and source of
food purchase. For the California cases (March 8- 9 and August 8),
we conducted active surveillance in city or community hospitals
near the case-patients' residence during the 5 days surrounding the
dates of onset. Surveillance included reviews of laboratory logs
(for cases with highly elevated CK but a low MB fraction) and
emergency room logs for all patients with a diagnosis of suspected
fish poisoning. 

Table. Symptoms of Haff disease cases, United States, 1997

State and local environmental management staff visited stores and
markets where fish was bought to trace the implicated fish lots. We
interviewed fish wholesalers regarding the number and location of
fishermen catching buffalo fish. Because fish eaten by the patients
originated from Louisiana and Missouri, we discussed with health
departments and environmental experts possible sources of fish
intoxication. Recovered leftovers and uncooked fish from the same
lot were tested for sodium channel-active biotoxins (e.g.,
ciguatoxin [the toxin of ciguatera] or saxitoxin [the toxin of
paralytic shellfish poisoning]), and cyanobacterial toxins (e.g.,
microcystin or nodularin, which are toxins of blue-green algae)
(11,12). Samples from the Bakersfield case were tested for
organophosphates, and other samples were tested for arsenic. To
characterize the physicochemical properties of the toxin, extract
from both cooked and uncooked fish was partitioned into water-
soluble, nonpolar lipid (hexane) and polar lipid (chloroform)
fractions. These fractions were then administered intraperitoneally
and orally to laboratory mice. Fractions causing toxicity to mice
were further analyzed for identification of the toxin.

From March through August 1997, two clusters of Haff disease
cases occurred in Los Angeles, California, and St. Louis, Missouri
(13); one isolated case occurred in Bakersfield, California. The Los
Angeles cluster consisted of two sisters who lived together and a
third patient who was admitted to the same hospital during the
same weekend. The St. Louis cluster was a married couple. The
four meals, all eaten at home, contained buffalo fish. Six (75%) of
the eight persons who ate buffalo fish became ill. All patients were
>30 years of age (33 to 87 years); five were Ukrainian immigrants,
and one was African American. Three of the six patients were
taking medications, including aspirin, codeine, and simvastatin,
that could have exacerbated rhabdomyolysis. Two persons who
vomited shortly after the meal had either milder symptoms or
lower laboratory values. The median incubation time to onset of
symptoms after fish ingestion was 8 hours (6 to 21 hours). All
patients were hospitalized, none died, and the median hospital stay
was 3 days. Clinically, five of six patients had rapid onset of
generalized muscular pain and rigidity, so severe that in one case
ventilation was required (Table). In one patient who had chest pain
only, the diagnosis of Haff disease was made through the exclusion
of other causes of chest pain and the epidemiologic link with two
other case-patients who had eaten buffalo fish purchased from the
same market. The predominant laboratory abnormalities were
elevated CK and myoglobin. Other muscle enzymes, such as
glutamate oxalate transaminase, glutamate pyruvate transaminase,
and lactate dehydrogenase, were also elevated. The CK was
elevated to a mean of 12,700 IU/L (normal is usually <300 IU/L)
(Figure). At the peak of the CK, the MB fraction was <5% in all
cases. The mean maximum myoglobin level, tested in three
patients, was 6,997 IU/L (upper limit of normal 100 IU/L).
Treatment consisted of intravenous fluids, in addition to mannitol
or bicarbonate. Several of the patients had variable sequelae,
including weakness and fatigue, for several months after the acute
stage. Active surveillance has identified no further cases. Buffalo
fish (Ictiobus cyprinellus) is a bottom-feeding freshwater fish
similar to carp. The fish was fried (four cases) or cooked for 3.5
hours as "gefilte fish" (two cases). Fish were purchased locally
either in a supermarket (California) or alive from a fish tank at a
market (Missouri). Fish lots were caught by approximately 25
commercial fishermen from rivers in Louisiana (three incidents) or
five fishermen in rivers and lakes near St. Louis, Missouri (one
incident). No specific bodies of water could be identified as the
origin of implicated fish. No fish or unusual animal die-off was
noted in the areas where buffalo fish were caught.

Figure. Creatine kinase elevation after the onset of symptoms in a
patient with Haff disease, California, 1997.

Tests of the fish for toxins were either negative or below toxicity
thresholds. The fact that the eaten fish was thoroughly cooked
suggests that the presumed toxin is heat-stable. Mice fed hexane-
soluble products extracted from cooked fish had behavioral
changes consistent with muscle impairment; bladders contained
red-brown urine.

Conclusions

Historically, Haff disease has been identified during seasonal
outbreaks in Europe. This report documents that it may also occur
sporadically or in small clusters. Ten of the 12 U.S. cases of Haff
disease reported since 1984 occurred during March through August
(M. Tormey, pers. comm.). All patients had eaten buffalo fish
before becoming ill; 8 of the 12 patients were California residents,
although the buffalo fish was caught in Louisiana or Missouri
waters.

Most reported cases or outbreaks of Haff disease have been
associated with freshwater fish, unlike most other seafood-related
illnesses (e.g., ciguatera, scombroid fish poisoning, or paralytic
shellfish poisoning), which are associated with saltwater fish (14).
The clinical symptoms of Haff disease also differ from those of any
other fish-related toxic or bacterial illness. Symptoms of
rhabdomyolysis predominate and neurologic features are absent, in
contrast to ciguatera or the various forms of shellfish poisoning
(14). Because of the spectrum of symptoms, the potential for
misdiagnosis is considerable. Furthermore, an increasing number
of health-conscious persons prefer diets that include fish. In 1998,
more than 4 billion pounds of fish, both domestic and imported,
were eaten in the United States (website of National Oceanic and
Atmospheric Administration: www.nmfs.org) Unusual smell or
taste does not help identify toxic fish, and normal cooking methods
cannot detoxify a fish capable of causing Haff disease. Because
Haff disease may occur not only in epidemics but also in small
clusters or sporadically, fish consumption should be included in the
history of patients with unexplained rhabdomyolysis.

Acknowledgments

We thank Steve Musser, Martin Robl, James Ahlrep, and Richard
Honkanen for laboratory support; Doug Dodson, Diane Rackers,
Louise McFarland, Al Medina, Michael Tormey, and S.B. Werner
for their advice and help; Piotr Kramarz for his translation of
Russian articles; and the staff of the Kern County Health
Department, particularly Drs. Dulan, Talbot, and Steve Terrell-
Perica, for their cooperation and tireless support in the
investigation.

Dr. Buchholz, who was with the Centers for Disease Control and
Prevention, at the time of this study, is a medical epidemiologist
with the Los Angeles County Department of Health Services. His
interests include foodborne disease epidemiology, outbreak
investigations, and surveillance for listeriosis and influenza.

Address for correspondence: Udo Buchholz, Los Angeles County
Department of Health Services, Acute Communicable Disease
Control, 313 N. Figueroa St., Room 212, Los Angeles, CA 90012,
USA; fax: 213-482-4856; e-mail: ubuchholz@dhs.co.la.ca.us.

References

1. Zu Jeddeloh B. Haffkrankheit [Haff disease]. [Ger]. Ergebnisse
in der inneren Medizin 1939;57:138-82.
2. Assmann H, Bielenstein H, Habs H, zu Jeddeloh B.
Beobachtungen und Untersuchungen bei der Haffkrankheit 1932
[Observations and investigations about Haff disease 1932]. [Ger].
Dtsch Med Wochenschr 1933;I:122-6.
3. Lentz O. Über die Haffkrankheit [About Haff disease]. [Ger].
Med Klin 1925;I:4-8.
4. Lockemann G. Chemische Untersuchungen zur Haff-krankheit
[Chemical investigations about Haff disease]. [Ger]. Biochemische
Zeitschrift 1929;207:194-216.
5. Taylor EJ, editor. Dorland's illustrated medical dictionary. 28th
ed. Philadelphia: W.B. Saunders Co.; 1994. Disease, Haff: 483.
6. Berlin R. Haff disease in Sweden. Acta Med Scandinavica
1948;129:560-72.
7. Leschenko PD, Khoroshilova NV, Slipchenko LM, Kaznachei
RY. Observation of Haff-Üchs disease cases. [Rus]. Vopr Pitan
1965;24:73-6.
8. Strusevich AV. Haff-Iuksov-Sartlan Disease. [Alimentary-toxic
paroxysmal myoglobinuria]. [Rus] Arkh Patol 1966;28:56-60.
9. Sidorova LD. Iuksovsk-Sartlansk disease. Kidney lesions in
dietary and toxic paroxysmal myoglobinuria. [Rus]. Ter Arkh
1985;57:120-3.
10. Salluzzo RF. Rhabdomyolysis. In: Rosen P, Barkin R, editors.
Emergency medicine: concepts and clinical practice. 4th ed. St.
Louis: Mosby-Year Book; 1997. p. 2478-87.
11. Manger RL, Leja LS, Lee SY, Hungerford JM, Hokama Y,
Dickey RW, et al. Detection of sodium channel toxins: directed
cytotoxicity assays of purified ciguatoxins, brevetoxins, saxitoxins,
and seafood extracts. J AOAC Int 1995;78:521-7.
12. Honkanen RE, Mowdy DE, Dickey RW. Determination of
DSP-toxins, okadaic acid and dinophysis toxin-1 in shellfish by
Serine/Threonin Protein Phosphatase Assay. J AOAC Int
1996;79:1336-43.
13. Centers for Disease Control and Prevention. Haff disease
associated with eating buffalo fish–United States, 1997. MMWR
Morb Mortal Wkly Rep 1998;47:1091-3.
14. Centers for Disease Control and Prevention. Tetrodotoxin
poisoning associated with eating puffer fish transported from
Japan—California, 1996. MMWR Morb Mortal Wkly Rep
1996;45:389-91.

Dispatches

Chlamydia pneumoniae Infection in a Breeding Colony of African
Clawed Frogs (Xenopus tropicalis)

Kurt D. Reed,* George R. Ruth,† Jeanine A. Meyer,* and Sanjay
K. Shukla*
*Marshfield Medical Research Foundation, Marshfield, Wisconsin,
USA; and †Marshfield Laboratories, Marshfield, Wisconsin, USA

More than 90% of a breeding colony of clawed frogs (Xenopus
tropicalis) imported to the United States from western Africa died
in an epizootic of chlamydiosis. Chlamydial inclusions were
observed by light and electron microscopy in the liver of an
infected frog. Chlamydia pneumoniae was isolated in cell cultures
from four frogs. A cutaneous infection by a chytridiomycete fungus
observed in two frogs could have been a cofactor in the die-off.ous
Diseases


Chlamydia infections cause disease in humans, birds, and
mammals. Of the four currently recognized Chlamydia species, C.
psittaci is the most important animal pathogen. Psittacosis, which
can manifest as severe enteric and respiratory illness in many avian
species, is highly contagious and can be transmitted to humans and
many other mammals (1). C. pecorum, the newest species to be
recognized, appears to have a highly restricted host range.
Infections due to C. pecorum have been associated with sporadic
encephalitis, polyarthritis, pneumonia, and conjunctivitis in pigs,
sheep, cattle, and koalas (2,3). C. trachomatis, an agent responsible
for millions of cases of ocular and urogenital infections worldwide,
causes most chlamydial infections in humans. C. pneumoniae, an
acute respiratory tract pathogen of cosmopolitan distribution, may
be linked with chronic diseases such as coronary atherosclerosis
and multiple sclerosis (1,4,5).

Initial reports of infections due to C. pneumoniae suggested that
the organism's host range was limited to humans. Subsequently,
infections due to C. pneumoniae have been documented in koalas,
a horse, and most recently, a giant barred frog (Mixophyes iteratus)
from Australia (6-8). We report an epizootic of chlamydiosis due
to C. pneumoniae in a commercial colony of African clawed frogs
(Xenopus tropicalis).

In July 1998, a biologic supply company located in the midwestern
United States imported 241 X. tropicalis from western Africa to
start a breeding colony. Upon arrival, the frogs were placed in
quarantine. Within 4 months, 220 (91%) of the frogs had died.
Before death, most of the frogs had sloughing of the skin and
lethargy and appeared bloated. Two frogs were submitted to our
veterinary diagnostic laboratory for necropsy. On external
examination, the frogs were 5.5 cm and 5.2 cm from snout to vent
and weighed 6.5 gm and 29.5 gm, respectively. Both frogs had
roughening and early sloughing of the posterior skin. The smaller
frog had scattered subserosal intestinal and hepatic nodules
containing tetrathrydian cestodes. The larger frog was extremely
edematous, with clear gelatinous fluid in the subcutaneous tissues
and body cavities. In hematoxylin- and eosin-stained 4-˜m sections
of liver from the larger frog, there was moderate to severe
lymphohistiocytic hepatitis with individual hepatocyte necrosis,
fibrosis, and numerous granular basophilic bodies resembling
bacteria within hepatocytes and histiocytes. Transmission electron
microscopy of liver tissue from the larger frog confirmed the
presence of bacteria replicating within vacuoles of infected cells.
The organisms had a biphasic structure consistent with a
Chlamydia spp. (Figure). Other Chlamydia was identified by
sequence analysis of the 16S rDNA and major outer membrane
protein A (ompA) genes. Bacterial DNA was extracted by lysis
(Generations, Gentra, Maryland) of liver and cell culture
specimens from each of the four frogs. The broad-range eubacterial
primers FD1(5'-AGAGTTTG ATCCTGGCTCAG-3') and RD1(5'-
AAGGAGGTG ATCCAGCC-3') were used in a polymerase chain
reaction (PCR) to amplify a 1,520-bp segment of the 16S rDNA
gene (10); primers 20CHOMP (5'-
TTAGAGGTRAGWATGAARAA-3') and CHOMP371 (5'-
TAGAAICKGAATTGIGCRTTI AYGTGIGCIGC-3') were used to
amplify a 1,200- bp segment of the ompA gene (11). Agarose gel
electrophoresis of the PCR products yielded single bands for each
specimen. Amplified products were column purified to remove
unincorporated primers and nucleotides, and sense and antisense
strands were sequenced by using nested primers in an automated
laser fluorescent sequencer (ALF Express, Amersham Pharmacia
Biotech, New Jersey). Sequence fragments were aligned with the
assembly software program DNAsis 2.5 (Hitachi, California), and
ambiguities resolved. A contiguous sequence was established from
the overlapping regions, which were used as a query sequence in
the GenBank database.

Figure. Transmission electron micrograph of chlamydial particles
in liver from a captive African clawed frog (Xenopus tropicalis).
Note the reticulate bodies (R), intermediate bodies (I), and highly
condensed elementary bodies (E). Bar, 270 nm. lesions present in
both frogs included mild to moderate lymphohistiocytic interstitial
pneumonia and mild lymphocytic interstitial nephritis. No
chlamydial inclusions were observed in other organs by light
microscopy. Both frogs had multifocal epidermal hyperplasia with
numerous intracorneal fungal organisms characterized by thick-
walled sporangia containing multiple 2-˜m to 3-˜m round spores.
Many of the sporangia had single tubular extensions (discharge
papillae) directed toward the skin surface, and transmission
electron microscopy showed the spores had flagella. These features
were considered diagnostic of a chytridiomycete fungus (9). To
further characterize the Chlamydia species, liver samples were
submitted for culture from the larger frog and three additional frogs
obtained subsequently from the same breeding colony. Tissue
homogenates were injected into shell-vials having monolayers of
either buffalo green monkey kidney (BGMK), Hep-2, McCoy, or
HeLa - 229 cells. Cultures were incubated at 35°C for 72 hours and
then the cell monolayers were stained with a fluorescein-
conjugated monoclonal antibody specific for the genus Chlamydia
(Pathfinder, Kallestad, California). Chlamydia grew from all four
liver specimens and in all the cell lines used. The organism grew
better, with more and larger inclusions, in the McCoy and BGMK
cell lines.

The 1,520-bp 16S rDNA sequences from the Chlamydia infecting
the four X. tropicalis (GenBank-AF139200) were identical to one
another and had 99.1%, 99.3%, and 99.4% homology with
published sequences for C. pneumoniae strains N16 (U68426.2),
CWL026 (AE001668), and TW-183 (L06108), respectively. In
addition, there was 100% homology with the 235-bp segment of
16S rDNA reported for the C. pneumoniae isolated recently from
the giant barred frog from Australia (AF102832). In contrast, there
was 96.0% sequence homology with C. psittaci strain 6BC
(U68447.2), 95.7% homology with C. pecorum strain E58
(D88317.1), and 93.8% homology with C. trachomatis strain
D/UW-3/CX (AE001345).

The sequence of a 279-bp segment of the variable domain IV
region of the ompA gene was determined for one of the frog
Chlamydia isolates (AF184214). There were six nucleotide
differences (97.8% homology) from the giant barred frog biovar of
C. pneumoniae (AF102830) and three differences (98.9%
homology) from C. pneumoniae strain CWL029 (AE001652). Less
than 75% homology was observed with the ompA variable domain
IV sequences reported for the three other Chlamydia species
(results not shown).

Previous reports of confirmed Chlamydia infections among captive
amphibians in North America involved X. laevis, another member
of the African clawed frog family (12-14). These reports occurred
before the discovery and molecular characterization of C.
pneumoniae, and the Chlamydia species involved were either not
known or presumed to be C. psittaci. Our study is the second
documented report of C. pneumoniae involving an amphibian and
the first describing an epizootic in a captive population of X.
tropicalis.

The spectrum of clinical disease and histopathologic lesions
associated with chlamydial infections in amphibians is unknown.
In our study, only two animals were evaluated histologically. The
more severely affected animal had active hepatitis with hepatocyte
necrosis, lymphohistiocytic infiltrates, and inclusion bodies within
hepatocytes and sinusoidal histiocytes. These lesions are similar to
those observed in a 1984 outbreak of chlamydiosis among
laboratory-bred X. laevis; in that outbreak, hepatosplenomegaly
and histologic evidence of active hepatitis were the major
pathologic findings in affected animals (12). In the recent case of
C. pneumoniae involving a free-ranging giant barred frog, the
predominant necropsy findings were chronic mononuclear
pneumonia, nonregenerative anemia, and pancytopenia (8).
Additional studies will be necessary to determine the epidemiology
and virulence of C. pneumoniae in captive and free-ranging
populations of amphibians and reptiles, as well as the pathogenicity
and transmissibility of this agent to humans and other mammals.

This breeding colony had >90% deaths during the 4-month
epizootic, and the few surviving frogs appeared ill and were
subsequently euthanized. We were unable to determine the
prevalence of C. pneumoniae infection in the colony but believe
that it was very high, since all four frogs submitted for culture were
positive. The crowding and stress associated with the captive
environment may have contributed to spread of infection within the
colony. The potential for zoonotic transmission of C. pneumoniae
from amphibians to humans has important public health
implications, especially in situations involving frequent and
prolonged human contact with amphibians in crowded conditions,
such as commercial breeding colonies or zoos. Until more is
known about the host range of this pathogen, chlamydiosis should
at least be considered in the differential diagnosis of an
unexplained respiratory or febrile illness in persons exposed to
amphibians.

Whether Chlamydia was the primary pathogen responsible for the
die-off or served as a cofactor with the chytrid fungus or other
parasites was difficult to determine. Although the Chlamydia
infection is the primary focus of this report, the presence of
cutaneous chytridiomycosis in two of the frogs is also important.
This highly pathogenic zoosporic fungus has only been found in
the American and Australian continents, where it has been
implicated as a major factor in amphibian deaths and population
declines in the rain forests (9). The origin of the chytrid fungus
found in our cases was not determined. However, the rapid onset of
illness affecting the frogs after their arrival in the United States
suggests that at least some of the frogs may have been infected
before capture.

The pipid frog X. tropicalis has a diploid genome and short
generation time, which make it an ideal model organism for
multigenerational genetic analysis. Current demand for this species
is higher than most biologic supply companies can meet.
Continued importation of X. tropicalis and other amphibians from
regions of the world experiencing poorly understood population
declines raises concern about the inadvertent spread of virulent
pathogens to naive populations of amphibians and reptiles, as well
as transmission of these agents to mammals. Until more is known
about the epidemiology and prevention of these infections, caution
must be exercised in transportation, husbandry, and human contact
with these animals.

Acknowledgments

The authors thank Lynn Ivacic for expert technical assistance with
the chlamydial cultures, Donald Stoiber for preparing the electron
micrographs, and the veterinary pathology section of the Armed
Forces Institute of Pathology for identifying the cestodes and
reviewing the histopathology.

This work was supported in part by grants from the Marshfield
Medical Research Foundation. Jeanine Meyer was an
undergraduate research fellow of Marshfield Clinic.

Dr. Reed is a clinical pathologist and director of Clinical Research
at the Marshfield Medical Research Foundation. His research
focuses on the diagnosis and molecular epidemiology of tickborne
and other zoonotic pathogens.


Address for correspondence: Kurt D. Reed, Clinical Research
Department, Marshfield Medical Research Foundation, 1000 N.
Oak Avenue, Marshfield, WI 54449, USA; fax:715-389- 4145; e-
mail: reedk@mfldclin.edu.

References

1. Schacter J. Infection and disease epidemiology. In: Chlamydia:
intracellular biology, pathogenesis, and immunity. Stephens RS,
editor. Washington: American Society for Microbiology Press;
1999. p. 139-69.
2. Fukushi H, Hirai K. Proposal of Chlamydia pecorum sp. nov. for
Chlamydia strains derived from ruminants. Int J Syst Bacteriol
1992;42:306-9.
3. Glassick TV, Giffard P, Timms P. Outer membrane protein 2
gene sequence indicates that Chlamydia pecorum and Chlamydia
pneumoniae, cause infection in koalas. Syst Appl Microbiol
1996;19:457-64.
4. Kuo C, Jackson LA, Campbell LA, Grayston JT. Chlamydia
pneumoniae (TWAR). Clin Microbiol Rev 1995;8:451-61.
5. Campbell LA, Kuo C, Grayston JT. Chlamydia pneumoniae and
cardiovascular disease. Emerg Infect Dis 1998;4:571-9.
6. Jackson M, White N, Gifford P, Timms P. Epizootic of
Chlamydia infection in two free-ranging Koala populations. Vet
Microbiol 1999;65:255-64.
7. Storey C, Lusher M, Yates P, Richmond S. Evidence for
Chlamydia pneumoniae of non-human origin. Journal of General
Microbiology 1993;139:2621-6.
8. Berger L, Volp K, Mathews S, Speare R, Timms P. Chlamydia
pneumoniae in a free-ranging giant barred frog (Mixophyes
iteratus) from Australia. J Clin Microbiol 1999;37:2378-80.
9. Berger L, Speare R, Daszak P, Green DE, Cunningham AA,
Goggin CL, et al. Chytridiomycosis causes amphibian mortality
associated with population declines in the rain forests of Australia
and Central America. Proc Nat Acad Sci U S A 1998;95:9031-6.
10. Weisburg WG, Barns SM, Pelletier DA, Lane DJ. 16S
ribosomal DNA amplification for phylogenetic study. J Bacteriol
1991;173:697-703.
11. Kaltenboeck B, Kousoulas KG, Storz J. Structures of and
allelic diversity and relationships among the major outer membrane
protein (ompA) genes of four chlamydial species. J Bacteriol
1993;175:487-502.
12. Howerth EW. Pathology of naturally occurring chlamydiosis in
African clawed frogs (Xenopus laevis). Vet Pathol 1984;21:28-32.
13. Newcomer CE, Anver MR, Simmons JL, Wilcke BW, Nace
GW. Spontaneous and experimental infections of Xenopus laevis
with Chlamydia psittaci. Lab Anim Sci 1982;32:680-6.
14. Wilcke BW, Newcomer CE, Anver MR, Simmons JL, Nace
GW. Isolation of Chlamydia psittaci from naturally infected
African clawed frogs (Xenopus laevis). Infect Immun 1983;41:789-
94.


Dispatches

The Impact of Health Communication and Enhanced Laboratory-
Based Surveillance on Detection of Cyclosporiasis Outbreaks in
California

Janet C. Mohle-Boetani,* S. Benson Werner,* Stephen H.
Waterman,* and Duc J. Vugia*†
*California Department of Health Services, Berkeley, California,
USA; and †The California Emerging Infections Program,
Berkeley, California, USA

We investigated the timing of diagnosis, influence of media
information on testing for Cyclospora, and the method used to
identify cases during eight cyclosporiasis outbreaks in California in
spring of 1997. We found that Internet information, media reports,
and enhanced laboratory surveillance improved detection of these
outbreaks.

Cyclospora cayetanensis, a recently identified coccidian parasite of
humans (1), causes explosive watery diarrhea that can persist for
several weeks (2). The oocysts cannot be detected by standard ova
and parasite testing; methods specific for Cyclospora include
modified acid-fast or other stains, autofluorescence with ultraviolet
epifluorescence microscopy, and wet mount under phase-contrast
microscopy (3). The infection can be treated with trimethoprim-
sulfamethoxazole (TMP-SMX) (4). Eight outbreaks in California
were among the 41 event-associated clusters of cyclosporiasis
associated with Guatemalan raspberries identified in the United
States and Canada in 1997 (5). We examined factors contributing
to the identification of outbreaks of this emerging infectious
disease. The detection of outbreaks was enhanced by Internet
information that patients brought to their physicians, media reports,
and enhanced laboratory-based (ELB) surveillance. ELB
surveillance, part of the California Emerging Infections Program
(EIP), is a cooperative agreement between the California State
Health Department, the University of California Berkeley School
of Public Health, selected local counties, and the Centers for
Disease Control and Prevention.


The Study

An outbreak of cyclosporiasis was defined as diarrhea (three or
more loose stools per day for at least 3 days) in two or more
persons who shared a meal and became ill within 2 weeks, and
laboratory confirmation of Cyclospora infection in at least one
person. The first person with laboratory-confirmed Cyclospora
infection identified in an outbreak was designated the index
patient.

We interviewed each index patient about the onset of symptoms
and medical care, and we reviewed laboratory records to determine
the date of testing for Cyclospora. Through information that we
gathered from patients and their physicians, we determined the
reason for testing for Cyclospora (e.g., symptoms alone, symptoms
and media attention, or symptoms and Internet information). We
interviewed all guests at a wedding reception (event 3, Table) to
determine the proportion of ill persons who sought medical
attention, submitted stool specimens tested specifically for
Cyclospora, and received recommended treatment for Cyclospora
infections, all with respect to timing of media coverage of
Cyclospora outbreaks.

To assess the impact of ELB surveillance on outbreak detection,
we determined the proportion of outbreaks identified by ELB
surveillance. All 22 laboratories in eight participating counties in
northern California submitted biweekly or monthly laboratory
reports on Cyclospora to EIP. EIP staff then forwarded case
information to the local health departments for follow-up. The
eight index patients were tested for Cyclospora infections a median
of 18.5 days after symptom onset (range 3-39 days, Table); 6
(75%) were diagnosed >14 days after onset (events 1-6, Table).
The two index cases diagnosed within 1 week of disease onset
occurred after media announcements about clusters of
cyclosporiasis in the United States were widespread in late May
1997 (events 7 and 8, Table). Testing for 6 (75%) index patients
was apparently prompted by this media coverage (events 3-8,
Table). After reading newspaper articles, three index patients
(events 3, 5, and 8) obtained information on the diagnosis and
treatment of Cyclospora from Internet searches, brought this
information to their physicians, and requested testing. One index
patient (event 4) read about other clusters in the United States in
the newspaper and requested testing from her physician. Another
index patient (event 6) suspected that she and a group of friends
had Cyclospora infections after reading a list of symptoms in a
newspaper article and reported this cluster to the local health
department, which then recommended testing for Cyclospora. One
index patient (event 7) was tested by a clinician who watched a
television report about cyclosporiasis the morning of the patient's
visit to the clinic.

Table. Event-associated outbreaks of cyclosporiasis, California,
1997

Within 2 weeks after a wedding party in May (event 3, Table), 30
(65%) of 46 guests had diarrhea; within 3 weeks, 13 sought
medical attention. Although seven patients submitted stool
specimens, none was tested for Cyclospora. Approximately 3
weeks after the wedding, several ill guests and the bride (who had
not eaten the mixed berry dessert and was not ill) read media
accounts of a Cyclospora outbreak that had occurred in Reno,
Nevada. They recognized similar symptoms in themselves or their
family members, obtained more information about Cyclospora
from the CDC Web site, and distributed it to other guests. All 16
guests who brought that Internet information to their physicians
received appropriate treatment with TMP-SMX; symptoms
resolved promptly (typically within 24 hours). Three of these 16
guests were tested for Cyclospora, and stool specimens were
positive.

In three outbreaks (events 1, 2, and 6, Table), the index patient was
tested after a suspected foodborne outbreak had been reported to
the local health department by symptomatic patients (events 1 and
6, Table) or an infectious disease specialist (event 2, Table). Index
patients from these three outbreaks were identified by the health
department within 1 day of their laboratory verification.

In six outbreaks, patients lived in counties with ELB surveillance;
5 (83%) of these outbreaks were detected by interviewing index
patients with positive specimens identified through ELB
surveillance (events 3, 4, 5, 7, and 8, Table). The health
departments were notified of 4 of the 5 index cases detected
through ELB surveillance 3 to 6 days after laboratory confirmation.
The detection of these outbreaks was delayed primarily because the
diagnosis of cases was delayed, rather than because health
departments were not notified of positive laboratory results
(Table). The delays in diagnosis were due to delays in seeking
medical care and receiving evaluation by a physician. For example,
the index patients in outbreaks 3 and 5 sought medical attention
after > 14 days of symptoms, after learning about other outbreaks
of cyclosporiasis, conducting Internet searches on diagnosis and
treatment, and bringing this information to their physicians; both
were tested promptly. In contrast, the index patient in outbreak 6
was tested 24 days after seeking medical attention, in response to a
recommendation by the local health department; this cluster of
illnesses was reported to the health department when some of the
patients learned about the cyclosporiasis outbreaks in the news.

The wedding outbreak investigation (event 3) best illustrates delays
in clinician evaluations and effects of Internet information
provided by patients on proper diagnosis and treatment (6).
Thirteen (43%) of ill persons consulted physicians, and 54% of
these submitted stool specimens; however, disease diagnosis was
delayed because none of the physicians ordered specific laboratory
testing that could identify Cyclospora (3). Patients were not tested
until media attention about other clusters of cyclosporiasis in the
United States led the bride and ill guests to conduct Internet
searches.

Conclusions

The role that the media and Internet information played in testing
for Cyclospora in these outbreaks was striking. Most index patients
prompted their physicians to test for Cyclospora by providing them
with Internet or media information. At least seven of the ill
wedding guests contacted their physicians several times over a 2-
to 3-week period but were not tested or treated properly until they
provided information on Cyclospora. The two index patients who
did not have a delay in diagnosis were prompted by widespread
media reports to seek both Internet information on Cyclospora and
medical attention. In the only instance in which appropriate testing
was ordered without prompting by the patient or the health
department, the physician had watched a television report on
cyclosporiasis that morning.

Several factors can contribute to delays in testing for Cyclospora,
and all of these factors were noted in the wedding outbreak.
Persons with diarrheal illnesses often do not seek medical
attention; 57% of ill persons in the wedding outbreak did not.
Since Cyclospora is a new infection, many physicians are not
familiar with its symptoms and treatment. Physicians often do not
request testing of stool; 46% of ill persons who sought medical
care in the wedding outbreak did not have any stool testing.

Detection of cases not only influences detection of clusters but also
prompts appropriate medical treatment with TMP-SMX. When a
clinical decision is made to assess a patient for a parasitic
gastrointestinal infection (e.g., patients with prolonged diarrhea),
clinicians should order both routine ova and parasite examination
and specific testing for Cyclospora and Cryptosporidium.
Laboratory-confirmed cases should be promptly reported so that
outbreaks can be detected and investigated, ongoing transmission
can be interrupted, and other ill persons can be provided with
effective treatment. Because symptoms can be prolonged,
investigation even 1 month after a common event can lead to
effective treatment.

Cyclospora is not yet reportable in California; however, clinicians
are required to report all unusual diseases and outbreaks to the
local health departments. Nevertheless, diseases that are not
reportable are rarely reported. Therefore, Cyclospora outbreaks
identified through ELB surveillance (83% of outbreaks that
involved residents of the counties with this surveillance system)
might not have been detected without enhanced surveillance.

Acknowledgments

We thank Gwen Bell, Amy Bellomy, Louise Gresham, Anthony
Marfin, Beth Schultz, and Sirlura Taylor for providing information
on index patients; Louise Gresham, Carol Greene, Sara Cody, Jon
Rosenberg, and Robert Murray for conducting patient interviews;
Judy Rees and Joelle Nadel for providing information from the
California Emerging Infections Program; and Barbara Herwaldt for
her thoughtful review of the manuscript.

Dr. Mohle-Boetani is a medical epidemiologist in the Disease
Investigations and Surveillance Branch of the Division of
Communicable Disease Control, California Department of Health
Services, Berkeley, California. Her areas of expertise include
outbreak investigation, cost-effectiveness analysis, and surveillance
of infectious diseases.

Address for correspondence: Janet Mohle-Boetani, Disease
Investigations Section, Disease Investigation and Surveillance
Branch, Division of Communicable Disease Control, California
Department of Health Services, 2151 Berkeley Way, Room 708,
Berkeley, CA 94704, USA; fax: 510-540-2570; e-mail:
jmohlebo@dhs.ca.gov.

References

1. Ortega YR, Sterling CR, Gilman RH, Vitaliano AC, Diaz F.
Cyclospora species—a new protozoan pathogen of humans. N Engl
J Med 1993;328:1308-12.
2. Soave R. Cyclospora: an overview. Clin Infect Dis 1996;23:429-
37.
3. Eberhard ML, Pieniazek NJ, Arrowood MJ. Laboratory
diagnosis of Cyclospora infections. Arch Pathol Lab Med
1997;121;792-7.
4. Hoge CW, Shlim DR, Ghimire M, Rabold JG, Pandey P, Walch
A, et al. Placebo-controlled trial of co-trimoxazole for Cyclospora
infections among travelers and foreign residents in Nepal. Lancet
1995;346:691-3.
5. Herwaldt BL, Beach MJ, Cyclospora Working Group. The
return of Cyclospora in 1997: another outbreak of cyclosporiasis in
North America associated with imported raspberries. Ann Intern
Med 1999;130:210-20.
6. Silberg WM, Lundberg GD. Musacchio RA. Assessing,
controlling, and assuring the quality of medical information on the
Internet. JAMA 1997;277:1244-5.

Dispatches

Norwalk-Like Viral Gastroenteritis Outbreak in U.S. Army
Trainees

Mark K. Arness,* Brian H. Feighner,† Michelle L. Canham,†
David N. Taylor,‡ Stephan S. Monroe,§ Theodore J. Cieslak¶
Edward L. Hoedebecke,† Christina S. Polyak,† Judy C. Cuthie,†
Rebecca L. Fankhauser,§ Charles D. Humphrey,§ Tamra L.
Barker,# Chris D. Jenkins,# and Donald R. Skillman#
*Uniformed Services University of the Health Sciences, Bethesda,
Maryland, USA; †U.S. Army Center for Health Promotion and
Preventive Medicine, Edgewood, Maryland, USA; ‡Walter Reed
Army Institute of Research, Washington, D.C., USA; §Centers for
Disease Control and Prevention, Atlanta, Georgia, USA; ¶U.S.
Army Medical Research Institute of Infectious Diseases, Frederick,
Maryland, USA; and #William Beaumont Army Medical Center,
El Paso, Texas, USA

An outbreak of acute gastroenteritis hospitalized 99 (12%) of 835
U.S. Army trainees at Fort Bliss, El Paso, Texas, from August 27
to September 1, 1998. Reverse transcriptase polymerase chain
reaction tests for Norwalk-like virus were positive for genogroup 2.
Gastroenteritis was associated with one post dining facility and
with soft drinks.

Norwalk viral gastroenteritis has been identified as an important
cause of illness among military troops (1-3). We report the
epidemiologic, environmental, and laboratory investigations of a
large foodborne outbreak of acute gastroenteritis in military
trainees in which a Norwalk-like virus, genogroup 2 (NLV2), was
identified by reverse transcriptase-polymerase chain reaction (RT-
PCR).

The Outbreak

From August 27 to September 1, 1998, 99 soldiers with acute
gastroenteritis, assigned to the Fort Bliss Air Defense Artillery
Training site, El Paso, Texas, were admitted to the William
Beaumont Army Medical Center for observation and hydration.
The criteria for admission were lack of medical care in barracks
and isolation of cases to contain the outbreak. All soldiers admitted
to the medical center with gastroenteritis were isolated in a
temporary medical facility. The most prominent symptoms were
acute nausea, vomiting, abdominal pain, diarrhea, and fever (Table
1). The median hospital stay was 24 hours (12 to 72 hours). Initial
clinical laboratory evaluations suggested acute viral gastroenteritis
(Table 2). All soldiers were from the same unit and lived near each
other in the training area compound, a rectangular block of 15
buildings. The hospitalization rate for gastroenteritis in this unit for
the week of August 27 was 99 (12%) of 835. The compound had
two dining facilities, DF1 and DF2, located across from each other
in the center of the compound. Preliminary interviews with
hospitalized patients implicated DF1, which was closed on August
28. Although some patients preferred one dining facility over the
other, in general they dined as directed by their drill sergeants or
where the wait was shortest. In addition to U.S. soldiers, the unit
trained small numbers of US marines and Japanese defense force
personnel, who also used these dining facilities.

Table 1. Frequency of symptoms in hospitalized soldiers

Table 2. Frequency of selected findings in hospitalized soldiers

Table 3. Odds ratios (OR) for selected foods and dining facilities

Patients, food handlers, facility engineers, post public health
officers, and training unit officers were interviewed, and the
training compound and dining facilities were inspected. Sign-in
rosters, menus, and meal preparation documents, as well as daily
personnel status reports and training schedules, were reviewed.
Cases were mapped by building of residence. Foodborne outbreak
questionnaires were administered to 86 of the hospitalized patients
available for interview; a randomly chosen (using the last digit of
Social Security numbers: 3,5,7) group of 198 soldiers who had not
sought medical care; and US marines and 15 Japanese defense
force troops (total 323 questionnaires).

Because of the time elapsed since the start of the outbreak, the
questionnaire was designed with a food preference format, based
on the foods listed on the previous week's menu. Data were
abstracted from 90 inpatient records and 323 questionnaires and
meal sign-in rosters from August 25 to August 27, 1998, and were
analyzed for trends by SPSS, a statistical software package. The
case definition for acute gastroen teritis was three or more loose
diarrheal stools (with or without vomiting) within a 24-hour period
in the week before the outbreak, with or without admission to the
hospital. To better characterize the point source of the outbreak, 98
(77.8%) of 126 cases with onset on August 27 to August 28 were
termed "first wave" cases and were compared with propagated
cases and controls for association with selected exposures (Table
3). Limited food samples and cultures of the ice cream and soft
drink dispensers were obtained and sent for analysis—the ice
cream dispensers had been turned off and left at room temperature
for 36 hours before sampling. Post water production facilities were
reviewed, and water distribution maps were obtained from the Fort
Bliss engineers. Construction in the area had necessitated closure
of a nearby water main on August 27, but the main was reopened
on the same day. Water samples were taken from multiple sites in
the training compound and elsewhere on post. Stool samples were
collected from all hospitalized patients and processed for
Salmonella, Shigella, Campylobacter, and Yersinia. The first 15
samples were processed for Aeromonas, Plesiomonas, and
Escherichia coli O157:H7, as well as for fecal leukocytes and ova
or parasites, including staining for Cyclospora and Cryptosporidia.
Twenty-four specimens were also sent to the Centers for Disease
Control and Prevention (CDC), Atlanta, Georgia, for electron
microscopy and PCR testing for NLVs (4). Of the 222 U.S.
soldiers and marines selected to complete a questionnaire as
controls, 31 (14.0%) also met the case definition for acute
gastroenteritis. Extrapolating to 736 nonhospitalized U.S. soldiers
and 24 marines, 106 (14.0%) of 760 would have been expected to
have had unreported illness. Added to the 99 hospitalized troops,
this yields a crude unit attack rate among all U.S. trainees of (106 +
99) (23.9%) of 859. Of the 15 Japanese soldiers who were
interviewed, nine (60%) had been ill and met the case definition. In
all, 40 (16.9%) of the 237 controls also met the case definition and
were reclassified as case-patients. An epidemic curve, based on
time of symptom onset in cases, was constructed from the 126
identified cases of acute gastroenteritis with completed
questionnaires (Figure).

Mapping of the outbreak cases demonstrated a discrete geographic
clustering in the training compound, with the exceptions of the
Japanese soldiers and two U.S. Army officers. Interviews with the
Japanese troops found that they all had used DF1 exclusively for
morning and evening meals during the week before illness. The
two army officers reported eating just one meal at DF1 at lunch on
August 26, with soft drinks and pudding pie as the only food in
common. Interviews with food handlers found that a confection
baker had become acutely ill while baking in the DF1 facility
between 2 a.m. and 4 a.m. on August 26. A DF1 housekeeper (not
food handler) also reported self-limited gastrointestinal illness
between August 27 and August 29.

None of the workers in DF2 reported illness. Univariate analysis of
the abstracted questionnaire data showed that soldiers who ate in
DF1 in the week before the outbreak were 9.8 times as likely to
contract acute gastroenteritis as those who did not use DF1 (95%
CI: 2.8, 40.7). Although univariate analysis indicated that illness
was also associated with eating crumb cake and cinnamon rolls
prepared by the baker, as well as ice cream and soda from DF1,
multivariate analysis did not support these associations.
Multivariate analysis indicated that the best predictors of illness
were dining in DF1 and drinking soda from the dining facility
(Table 3). All post water samples tested negative for fecal
coliforms. General sanitation in the dining facilities was
satisfactory, and no back-siphoning hazards were found. The soft
drink dispensers had antisiphoning valves. Cultures of food
specimens were negative except for nonpathogenic coliform
bacteria (Citrobacter diversus and Serratia liquefaciens) isolated
from the ice cream dispenser in DF1 and Enterobacter cloacae
coliform bacteria isolated from the soda fountain in DF2. All stool
cultures from hospitalized trainees were negative for bacterial fecal
pathogens. No ova or parasites, fecal leukocytes, or bacteriologic
pathogens were found. RT-PCR was positive for NLV2 in 17 of 24
stool specimens submitted to CDC for analysis. Electron
microscopy was performed on seven specimens, and all were
positive for 30-nm particles consistent with NLV.

Conclusions

This gastroenteritis outbreak was notable for the explosive onset of
an intense but brief illness with a short incubation period of 24 to
36 hours. The outbreak curve was characteristic of a point-source,
propagated, foodborne illness, with clinical and epidemiologic
features suggestive of NLV gastroenteritis, which was
subsequently confirmed by RT-PCR (5-14).

NLVs have been identified as the predominant cause of viral
gastroenteritis and have been implicated in 42% to 96% of
nonbacterial gastroenteritis outbreaks since 1976 (5,10-12).
Although contaminated water is often the vehicle of outbreaks, the
water distribution on Fort Bliss is within a closed loop system, and
the tight geographic distribution of cases was inconsistent with
proximal contamination of the general water supply. In addition,
ice from the dining facilities was not associated with illness (Table
3). Interviews supported the hypothesis that the outbreak was
caused by point-source contamination of food or drink in DF1.
Statistical analysis showed a strong association between illness and
dining in DF1 and weaker associations with several food items,
particularly crumb cake prepared by the ill confectioner on the
morning of August 26 and Figure. Date of onset of acute
gastroenteritis symptoms (AGE) in 126 soldiers, week of August
24, 1998 served at meals on August 26 and 27. The weakness of
the association could be due in part to recall bias, since the onset of
illness occurred fully a week before the investigation. Both pre-and
postsymptomatic contamination of foods has been documented in
outbreaks traced to food handlers, which could account for a few
early, sporadic cases before the major outbreak. The strong
statistical association of gastroenteritis with soda drinking suggests
contamination of the soft drink dispenser in DF1—a distinct
possibility, as NLVs are hardy and persistent in the environment,
resisting both disinfection and chlorination (5,8,10). Bacterial
contamination of the ice cream dispenser in DF1 and the soda
fountain in DF2, while an incidental finding, implies that
mechanical transmission of pathogenic organisms is possible in
these facilities. A confounding effect, such as a tendency of ill
soldiers to drink more soda, could also explain this association.
Secondary person-to-person transmission promotes viral
propagation in outbreaks and maintains these viruses in circulation.
The prompt closure of DF1 and use of temporary medical facilities
for quarantine likely decreased secondary propagation in this
outbreak. Limitations of this investigation include inability to
identify the viral agent from the confection baker, failure to obtain
acute- and convalescent-phase sera from ill DF1 staff and case-
patients, design limitations in using food preferences in the survey,
and inability to identify the agent in suspect foods, since NLVs
cannot be cultured (5-10).

NLVs are distributed worldwide; serum antibody to Norwalk virus
in some countries approaches 100% seroprevalence in adults. The
ubiquitous nature and persistence of these viral agents make
similar future outbreaks of NLV gastroenteritis a near certainty.
RT-PCR is a valuable diagnostic tool in the identification of NLVs
as the etiologic agent in this setting. Dr. Arness is an Air Force
senior flight surgeon and preventive medicine officer currently
assigned to the 86th Contingency Response Group, US Air Forces
Europe. He is the commander of the Environmental Medicine
Flight, a unit specialized in medical force protection, disease
surveillance and assessment of nuclear, chemical and biological
threats.

Address for correspondence: Mark K. Arness, PSC 10 Box 554,
APO AE 09142; fax: 314-496-7590; e-mail: marness@pol.net.

References

1. Sharp TW, Thornton SA, Wallace MR, Defraites RF, Sanchez
JL, Batchelor RA, et al. Diarrheal disease among military
personnel during operation Restore Hope, Somalia, 1992-1993.
Am J Trop Med Hyg 1995;52:188-93.
2. Hyams KC, Hanson RK, Wignall FS, Escamilla J, Oldfield EC.
The impact of infectious diseases on the health of U.S. troops
deployed to the Persian Gulf during operations Desert Shield and
Desert Storm. Clin Infect Dis 1995;20:1497-504.
3. Warner RD, Carr RW, McClesky FK, Johnson PC, Goldy-Elmer
LM, Davison VE. A large nontypical outbreak of Norwalk virus:
gastroenteritis associated with exposing celery to nonpotable water
and with Citrobacter freundii. Arch Intern Med 1991;151:2419-24.
4. Ando T, Monroe SS, Gentsch JR, Jin Q, Lewis DC, Glass RI.
Detection and differentiation of antigenically distinct small round-
structured viruses (Norwalk-like viruses) by reverse transcription-
PCR and Southern hybridization. J Clin Microbiol 1995;33:64-71.
5. Kapikian AZ, Estes MK, Chanock RM. Norwalk group of
viruses. In: Fields BN, Knipe DM, Howley PM, Chanock M,
Melnick JL, Monath TP, et al, editors. Fields' Virology. 3rd ed.
Philadelphia (PA): Lippincott-Raven Publishers; 1996. p. 783-810.
6. Caul EO. Viral gastroenteritis: small round structured viruses,
caliciviruses and astroviruses. Part I. The clinical and diagnostic
perspective. J Clin Pathol 1996;49:874-80.
7. Caul EO. Viral gastroenteritis: small round structured viruses,
caliciviruses and astroviruses. Part II. The epidemiological
perspective. J Clin Pathol 1996;49:959-64.
8. Hedberg CW, Osterholm MT. Outbreaks of food-borne and
waterborne viral gastroenteritis. Clin Micro Reviews 1993;6:199-
210.
9. Kapikian AZ. Viral gastroenteritis. JAMA 1993;269:627-30.
10. Centers for Disease Control. Viral agents of gastroenteritis:
public health importance and outbreak management. MMWR
Morb Mortal Wkly Rep 1990;39(RR-5):1-24.
11. Kaplan JE, Feldman R, Campbell DS, Lookabaugh C, Gary
GW. The frequency of a Norwalk-like pattern of illness in
outbreaks of acute gastroenteritis. AJPH 1982;72:1329-32.
12. Kaplan JE, Gary GW, Baron RC, Singh N, Schonberger LB,
Feldman R, et al. Epidemiology of Norwalk gastroenteritis and the
role of Norwalk virus in outbreaks of acute nonbacterial
gastroenteritis. Ann Intern Med 1982;96:756-61.
13. Adler JL, Zickl R. Winter vomiting disease. J Infect Dis
1969;119:668-73.
14. Kapikian AZ, Wyatt RG, Dolin R, Thornhill TS, Kalica AR,
Chanock RM. Visualization by immune electron microscopy of a
27-nm particle associated with acute infectious nonbacterial
gastroenteritis. J Virol 1972;10:1075-81.

Letters

Preventing Zoonotic Diseases in Immunocompromised Persons:
The Role of Physicians and Veterinarians

To the Editor: We read with great interest the article by Grant and
Olsen on the role of physicians and veterinarians in preventing
zoonotic diseases in immunocompromised persons (1) and the
letter by Barton et al. on the risk of pregnant women and young
infants for petassociated illnesses (2). Essentially all aspects and
conclusions of the study by Grant and Olsen are also valid in
Europe.

In Austria, veterinarians are well educated in zoonotic infections.
However, it is impossible for practitioners to know all zoonotic
agents in detail. In addition to immunocompromised persons,
pregnant women, and young infants, persons in certain occupations
are at higher risk for zoonotic infections. Veterinarians are one of
these groups. We have completed seroepidemiologic studies
involving veterinarians (3) and are testing other groups at high risk,
such as slaughterhouse workers, farmers, and zoo employees. We
surveyed 52% of the veterinarians in an Austrian federal state who
agreed to participate in the study. They completed a questionnaire
and provided case histories so that risk factors could be assessed.
We also obtained blood samples. The sera were tested for
antibodies to viral, bacterial, and parasitic zoonotic agents. After
correlating the serologic results with the statements in the
questionnaire, a statistical analysis, and another questionnaire of
selected participants, we found transmission of zoonotic agents
from animals to veterinarians for influenza A virus H1N1
(prevalence of the infection was much higher among veterinarians
than in the general population; a significant number of seropositive
veterinarians were swine practitioners [chi-square = p xx < .01]);
Coxiella burnetii (veterinarians who removed bovine placenta
without gloves had a higher risk [chi-square = p xxx < .001] of
acquiring Coxiella burnetii infections); Brucella sp.; Chlamydia
psittaci; Leptospira sp.; Toxoplasma gondii; and Toxocara
canis/cati (antibody prevalence was 20 times higher among
veterinarians than in the general Austrian population). As a result
of the survey, veterinarians know about their profession- specific
risk factors and take adequate measures to prevent infections; also,
they are more qualified to advise pet owners and persons of other
professions at high risk.

In addition to the pathogens listed by Grant and Olsen and the
agents listed in this letter, some other zoonotic agents present (at
least in Central Europe) are cowpox (mainly acquired from cats; 4-
6) and parapox viruses, lymphocytic choriomeningitis virus,
Newcastle disease virus, Erysipelothrix rhusiopathiae, and
Capillaria hepatica (7). On the other hand, animals are not the
source of human disease such as leukemia; there is no evidence
that feline leukemia virus and feline immunodeficiency virus, for
example, are transmitted to humans (8). The risk of acquiring a
zoonotic infection from a pet animal is definitely lower than the
emotional and health benefits of pet ownership.

Norbert Nowotny* and Armin Deutz†
*Institute of Virology, University of Veterinary Sciences, Vienna,
Austria; and †Animal Health Service, Styrian Provincial
Government, Department of Veterinary Affairs, Graz, Austria

References

1. Grant S, Olsen CW. Preventing zoonotic diseases in
immunocompromised persons: the role of physicians and
veterinarians. Emerg Infect Dis 1999;5:159-63.
2. Barton LL, Villar RG, Connick M. Pet-associated zoonoses.
Emerg Infect Dis 1999;5:598.
3. Nowotny N, Deutz A, Fuchs K, Schuller W, Hinterdorfer F,
Auer H, et al. Prevalence of swine influenza and other viral,
bacterial, and parasitic zoonoses in veterinarians. J Infect Dis
1997;176:1414- 5.
4. Nowotny N. The domestic cat: a possible transmitter of viruses
from rodents to man. Lancet 1994;343:921. 
5. Nowotny N. Serologische Untersuchungen von Hauskatzen auf
potentiell humanpathogene Virusinfektionen wildlebender
Nagetiere. Zentralblatt für Hygiene und Umweltmedizin
1996;198:452-61.
6. Nowotny N, Fischer OW, Schilcher F, Schwendenwein I, Loupal
G, Schwarzmann Th, et al. Pockenvirusin-fektionen bei
Hauskatzen: klinische, pathohistolo-gische, virologische und
epizootiologische Untersuchungen. Wiener Tierärztliche
Monatsschrift 1994;81:362-9.
7. Juncker-Voss M, Prosl H, Lussy H, Enzenberg U, Auer H,
Nowotny N. Serological detection of Capillaria hepatica by indirect
immunofluorescence assay. J Clin Microbiol 2000;38:431-3.
8. Nowotny N, Uthman A, Haas OA, Borkhardt A, Lechner K,
Egberink HF, et al. Is it possible to catch leukaemia from a cat?
Lancet 1995;346:252-3.

Letters

Reply to Drs. Nowotny and Deutz

To the Editor: We thank Drs. Nowotny and Deutz for their Letter
to the Editor regarding our letter in the January-February issue of
Emerging Infectious Diseases. We agree that increased education
and research efforts regarding zoonoses would benefit not only at-
risk patients, but also veterinary and human health professionals.
We also applaud their efforts to provide serologic evidence of
exposure to zoonotic pathogens among veterinarians in Austria.
However, readers should be aware of the zoonotic potential of two
of the pathogens in their screening. Specifically, although both
animals and humans suffer from respiratory syncytial virus
infections, evidence is minimal for interspecies transmission of the
domestic animal and human strains. Similarly, bovine viral
diarrhea virus is an important bovine pathogen, but there is little
evidence for its ability to infect humans.

Christopher W. Olsen* and Leslie L. Barton†
*University of Wisconsin-Madison, Madison, Wisconsin, USA;
and †University of Arizona Health Sciences Center, Tucson,
Arizona, USA.

Book Reviews

International Law and Infectious Diseases.
David P. Fidler. Oxford University Press, 1999.

This thoughtful book should find its readership among health
professionals, lawyers, international relations scholars, and
activists addressing issues of infectious diseases in the context of
public health, international trade, environment protection, or war.
Its chapters include the following topics: a historical overview,
International Health Regulations, international legal framework,
trade law, human rights law, war and weapons, environmental law,
international relations, and a Draft Framework Convention on
Global Infectious Disease Prevention and Control.

Early attempts at controlling the epidemic spread of diseases by
quarantining international shipping vessels were costly and
ineffective. Beginning in the mid-19th century, a series of
international sanitary conferences sought legal agreements among
states to reconcile the competing demands of commerce and public
health. As microbes and their vectors do not respect international
boundaries, David Fidler has chosen a rich subject for this study of
international legal regulation.

In democracies at least, rational trade policies may be swayed by
popular concerns; for example, bans on certain foodstuffs may be
more driven by the public's fear of disease than by good science.
The International Health Regulations, which are supposed to
address these issues, are outdated, limited in scope, and not
respected. Fidler therefore proposes a broader international
convention on global infectious disease prevention and control,
which would incorporate revised international health regulations;
he provides a draft convention for consideration.

As AIDS now kills more people globally than any other infectious
disease, Fidler's discussion of the right to health and the
confluence of human rights and public health in the context of the
AIDS pandemic is of particular interest. Yet his draft convention of
nation-states does not address either the incapacity of the states that
are worst affected or the importance of active participation by
nongovernmental organizations and transnational corporations.
Present-day international law seems incapable of addressing the
broader issue of collective international responsibility to act in the
face of the global AIDS pandemic.

The book is well paced and scholarly. Fidler's analysis is cautious,
but he does not shrink from discussing the failure of the World
Health Organization to apply the successes of international legal
regulation in fields such as international trade, aviation, labor
standards, and the environment to infectious diseases and public
health. One of the many ironies delineated in the book is the
resurgence of interest in international legal solutions to public
health challenges when science and medicine fail to provide
enduring national solutions. Interest in international public health
law waned in the 20th century, when vaccination and
improvements in public services and hygiene reduced or eliminated
the threats of smallpox, plague, cholera, and yellow fever. Yet
today, the increasing speed and scale of global trade and population
movements pose new risks from emerging and reemerging
diseases. Fidler's proposed international convention on global
infectious disease prevention and control would represent progress
toward addressing these concerns and deserves serious
consideration.

David Patterson
Geneva, Switzerland.

Book Reviews

Oxford Handbook of Tropical Medicine.
Michael Eddleston and Stephen Pierini. Oxford University Press,
New York, 1999.

In judging a new textbook, one is initially influenced by the
authors' reputation and qualifications. The authors of this Oxford
Handbook, Michael Eddleston and Stephen Pierini, are listed on
the front cover, but their names do not appear again. In the
introduction, these mystery authors tell us that the Oxford
Handbook of Tropical Medicine was written to answer a need for a
soft-cover, pocket-sized (18 x 10 x 2.5 cm), inexpensive,
lightweight (300 grams) handbook of clinical medicine in the
tropics for "junior doctors" who work in the developing world,
where few laboratory tests are available and technical and human
resources may be lacking. This book was written in collaboration
with World Health Organization (WHO) staff, as evidenced in the
acknowledgments, in which no fewer than 22 WHO consultants
appear, as well as the foreword by David Heymann, Executive
Director, Communicable Diseases, WHO.

So what's in this handbook, and how is it organized? The first 16
pages make up the introduction, which briefly covers such topics as
WHO's Essential Drugs Programme, outbreak investigations,
universal and isolation precautions, and integrated management of
childhood diseases. The next section (160 pages) covers five major
infectious disease areas (malaria, HIV and sexually transmitted
diseases, tuberculosis, diarrheal diseases, and acute respiratory
infections). Except for the section on fevers and systemic signs, the
rest of the book is systems based, covering most internal medicine
topics, including cardiology, chest medicine, renal diseases,
gastroenterology, neurology, dermatology, endocrinology,
hematology, nutrition, injuries and poisoning, and immunization.
Each section spans 30-50 pages of very small print, for which those
of us over 40 will need the assistance of a magnifying glass.

Before we tell you what we thought of the book, let us point out
two refreshing features. The authors ask that readers provide
comments and criticisms for improving future editions. (This is a
clue as to the identity of the authors.) The second is the authors'
expectation that readers would and should adapt the book for local
conditions. To this end, they have included blank pages throughout
the book for the reader to add and modify treatments and
diagnostic tests as necessary.

This book focuses on diseases, both infectious and non-infectious
that are seen in tropical developing countries and on therapies that
are available there. You will not find imipenem, moxifloxacin,
insulin pumps, or even culture and sensitivity data. Drugs
recommended may not always be ideal, but they are likely to be
available locally. Another major feature of the infectious disease
section is the liberal use of excellent WHO algorithms, which are
particularly useful for the inexperienced physician.

Disease descriptions are concisely written and organized with a
section on etiology/life cycle, clinical features, diagnosis,
management, and control. Most diseases are summarized in one
page or less, except for major diseases such as malaria,
tuberculosis, and HIV. A good part of the book is problem-based,
by symptoms, and this is by far its major strength. This problem-
oriented approach is ideal for the rural developing world, where
"medicine by intuition" is often practiced, and clinical skills,
knowledge, and judgment are all that may be available for disease
management. The major infectious disease sections and the one on
nutrition are excellent. The systems sections are very good but
somewhat lacking in perspective. Differential diagnoses are always
listed, but the inexperienced physician may have some difficulty
sorting out the top five conditions to be considered. Clearly, these
vary in different parts of the world (hence the blank pages), but
there are common problems everywhere that tend to head most
lists. We particularly loved the section on how to do a burr hole,
complete with diagrams (a worthwhile technique to practice on
your teenage children).

We were a bit disappointed that Bacille Calmette-Guérin vaccine
was covered in only seven lines, that bed nets were not mentioned
for malaria control, that typhoid was not included in the differential
diagnosis of lymphocytosis or prolonged fever, that short-course
therapy was not recommended for typhoid fever, and that
tinidazole or single-dose metronidazole were not suggested for
invasive amebiasis. In addition, the book is somewhat inconsistent
about which diagnostic tests are recommended: on the one hand,
bacterial cultures are almost never available in rural areas of the
tropics, but little emphasis is given to presumptive treatment. On
the other hand, in the nephrology section, the authors surprisingly
recommend an autoantibody screen and complements, urine pH,
and calcium levels.

Enough of nitpicking. The bottom line? This is an excellent first
edition of a handbook of tropical and internal medicine for the
rural practitioner. It is a comprehensive, concise, well written, and
(for the most part) practical handbook that provides a wealth of
information on diagnosis, treatment, and decision making. We
recommend it highly for medical students, residents, and even
infectious and tropical disease consultants planning to work in the
tropics or to care for patients from the tropics. Given their level of
training, the authors have done a remarkable job. We have decided,
on the basis of the following clues, that the authors of this
handbook are probably medical residents, or "registrars" in the
British system. First, the book is dedicated to their parents rather
than their spouses, so they are probably young and unmarried. They
acknowledge their "long-suffering mentors, David Warrell and
David Theakston," who send "fresh-faced medical students out to
remote corners of the world …" This book was almost certainly
written by "kids," recent medical students who had an international
health experience during training. In this world of academic
medicine, it is a shame that credibility is accorded only those who
have more initials after their names than in their names. We should
judge an excellent book such as this one by its contents and not by
the prestige of its authors. As Butch Cassidy said to the Sundance
Kid, "Who are those guys?" In this case, it doesn't matter.

Rani Beharry* and Jay S. Keystone†
*Royal College of Surgeons in Ireland, Dublin, Ireland; and
†University of Toronto, Toronto, Canada.

News and Notes

Conference Summary

Institute of Medicine's Forum on Emerging Infections: Workshop
on Managed-Care Systems and Emerging Infections

The Institute of Medicine's Forum on Emerging Infections was
established in 1996 in response to a request from the National
Center for Infectious Diseases, Centers for Disease Control and
Prevention, and the National Institute of Allergy and Infectious
Diseases, National Institutes of Health. The forum provides
structured opportunities for representatives of academia, industry,
professional and special-interest groups, and government agencies
to examine and discuss scientific and policy issues related to
research and prevention, detection, and management of emerging
infectious diseases. The forum, which fosters the exchange of
ideas, identifies areas for study, clarifies policy issues, and updates
decision-makers, seeks to illuminate issues rather than resolve
them; it does not provide advice or recommendations on any policy
pending before any agency or organization. Its strengths are the
diversity and dedication of its membership. The forum sponsors a
series of workshops. The first of these addressed public- and
private-sector collaboration (1). The second explored many aspects
of antimicrobial resistance, surveillance, and response (2). The
third workshop (3), which we summarize here, examined
opportunities posed for managed-care systems and challenges to
their ability to address the threat of emerging infectious diseases.
Representatives of managed-care organizations, hospitals,
government agencies, pharmaceutical companies, and academia
convened to discuss issues, suggest solutions, and highlight
impediments to be overcome in five key areas: basic and clinical
infectious disease research, clinical practice guidelines, emerging
infections surveillance and monitoring, education and outreach,
and drug formularies and product development. A common theme
of the discussions was the heterogeneity and rapid evolution of the
managed-care industry. Although a few large health maintenance
organizations (HMOs) were mentioned as effective research
partners, others have different capabilities and corporate cultures.
The participants recognized that, as the restructuring of the nation's
health-care system evolves, forging stronger partnerships with
managed-care systems will likely have a positive effect not only on
health-care delivery but on many aspects of public health. The
workshop revealed that the structure of managed-care
organizations provides few incentives for taking the broader and
longer-term public health view. Some incentives, even those
concerning quality assurance for individual patients and formulary
restrictions, may actually intensify public health problems. For
example, one presenter cited a study in which, for every condition
except otitis media, increased formulary restrictions were
associated with increased numbers of physician office visits,
emergency department visits, hospitalizations, and prescriptions,
along with an increased cost of prescriptions over a 12-month
period. The evidence regarding managed care's actual performance
and impact on emerging infections needs clarification. Some
managed-care plans have integrated services and sophisticated
research capabilities, whereas others provide little more than cost
reimbursement for conventional health-care services. The
managed-care industry by itself cannot be expected to develop and
implement solutions for those challenges. The industry could
become a more productive partner in combatting emerging
infections. Incentives might include support to cover the marginal
costs of research and demonstration activities, for example, the
gathering of drug-dispensing data and crude surveillance for
multidrug resistance among the organisms that cause tuberculosis
and sexually transmitted diseases. Some HMOs may garner a
competitive advantage in being viewed as progressive, research-
oriented organizations. Major purchasers of managed care will also
have an important role alongside the managed-care industry in
developing and implementing solutions to emerging infections
problems. Likewise, academic health centers and government
agencies can act as catalysts, as well as partners in research, to
allow greater participation by managed care in addressing
emerging infections. The need for better information to support the
provision of quality health care was also.

News and Notes

Upcoming Events


First United Kingdom Workshop on Borna Disease Virus: Borna
Disease Virus: A
Veterinary and Public Health Problem?
Rhondda, Wales, March 23-24, 2000
Borna Disease Virus (BDV) is endemic in parts of Europe, infects
a broad range of species, and causes a rare meningoencephalitis in
horses and sheep. The virus has not been clearly linked to any
human disease, but an association between infection with the virus
and certain neuropsychiatric disorders has been suggested.
However, the methods used in previous studies and the
significance of the findings are controversial. This first United
Kingdom Workshop on BDV is organized by the Public Health
Laboratory Service, the Health and Safety Executive, the Ministry
of Agriculture, and the Welsh Development Agency. The
Workshop will bring together agencies and researchers in the
United Kingdom interested in the diagnosis, pathology, and
epidemiology of BDV in human and animal populations; provide
opportunities for collaboration and sharing of expertise;
disseminate the latest findings from research on BDV in Europe
and elsewhere; and provide guidance for veterinary and public
health policymakers in developing surveillance and research
programs. The program will include sessions on the detection,
pathology, and epidemiology of BDV in animal and human
populations. Keynote speakers will be from Europe and the United
States. The registration fee of £69 includes delegate package,
meals, and proceedings. Delegates must make their own
arrangements for accommodations. For more information, visit our
web site: http://www.cdsc.wales.nhs.uk/bcon.htm. addressed by the
workshop. For example, an expert at the workshop reported on
preliminary studies indicating that formulary practices may have an
adverse impact on antibiotic resistance, prompting the need for
additional comprehensive data about these practices and their
impact. The use of outcomes information was identified as one
way to develop and implement new clinical practice guidelines.
Another promising outcome of the workshop was the identification
of the potential for integrated, electronic information systems to
assist physicians in diagnosing and treating infectious diseases;
managed-care organizations in tracking antibiotic use, costs, and
outcomes; and public health agencies in monitoring and even
preventing emerging infections and antibiotic resistance. Many of
the issues raised during the workshop, including drug formularies
and surveillance, have international as well as domestic
implications. For example, there are three health systems in Latin
America— private, public, and employee systems—but many of
the providers in the region work with all three systems. As
governments face increasing pressure to downsize, the impact of
that change on the vital public health functions of surveillance,
control, and prevention of infectious diseases concerns all systems.
In the United States, the National Institutes of Health, the Centers
for Disease Control and Prevention, and private groups that are
working on these issues should be involved in international
deliberations, and managed care should be part of foreign policy
initiatives in the area of infectious diseases. For information,
contact the Forum on Emerging Infections, Division of Health
Sciences Policy, Institute of Medicine, The National Academies,
2101 Constitution Avenue N.W., Washington, D.C. 20418, USA;
telephone: 202- 334-1888; fax: 202-334-1329; e-mail:
bugs@nas.edu. For information on the Institute of Medicine, visit
the Institute's home page at http://www.iom.edu. Jonathan R.
Davis* and Joshua Lederberg† *Institute of Medicine, The
National Academies, Washington DC, USA; and †The Rockefeller
University, New York, NY, USA

References
1. Institute of Medicine. Orphans and incentives: developing
technologies to address emerging infections. Washington: National
Academy Press; 1997.
2. Institute of Medicine. Antimicrobial resistance: issues and
options. Washington: National Academy Press; 1998.
3. Institute of Medicine. Managed care systems and emerging
infections: challenges and opportunities for strengthening
surveillance, research, and prevention. Washington: National
Academy Press; 2000.

News and Notes

10th International Symposium on Viral Hepatitis and Liver Disease
Atlanta, Georgia April 9–13, 2000

The preliminary program of the 10th International Symposium on
Viral Hepatitis and Liver Disease is now available for viewing on
the conference web site, http://www.hep2000.com For additional
information, contact the Organizing Secretariat at
info@hep2000.com or by telephone at 404-233-4490.

News and Notes


The Johns Hopkins University Graduate Summer Institute of
Epidemiology and
Biostatistics
Baltimore, Maryland
The Graduate Summer Institute, sponsored by the Johns Hopkins
University School of Public Health, Baltimore, MD, provides an
opportunity for the physician, nurse, or health professional to enroll
in courses for academic or continuing education credit. Cost is
$675 for one course; $1,000 for two courses; $175 for each
additional course (non-credit 3-week courses). Application
deadline is June 1, 2000. For information contact Ayuesha Khan,
Department of Epidemiology, School of Public Health, Johns
Hopkins University, 615 North Wolfe Street, Baltimore, MD
21205. Phone: 410/955-7158; fax: 410/955-0863; e-mail:
akhan@jhsph.edu.

News and Notes

Negative Strand Viruses 2000
Eleventh International Conference on Negative Strand Viruses
Québec City, Québec, Canada
June 24-29, 2000

For information, contact the Conference Committee at Negative
Strand Viruses 2000, P.O. Box 33799, Decatur, GA 30033-799,
USA; telephone: 404-728-0564; fax: 404-728-0032. The e-mail
address is nsv2000@aol.com and the web site address is http:
www.nsv2000.com.

News and Notes

5th International Meeting on Molecular Epidemiology and
Evolutionary Genetics in Infectious Diseases
Hyderabad, India
November 12-16, 2000

The conference is sponsored by the Centers for Disease Control
and Prevention, USA; Institut de Recherche pour le
Développement and the Centre National de la Recherche
Scientifique, France; and the Indian Council of Medical Research
and the Department of Biotechnology, India. Objectives are to
integrate epidemiologic, molecular biologic, and evolutionary
genetics approaches in the areas of diagnosis, strain typing, species
identification, pathogenesis, antigenic variation, drug and vaccine
resistance, host susceptibility, and vector specificity; to foster
interactions between epidemiologists and laboratory scientists
working on the same pathogens; and to provide a forum for health-
care providers, public health professionals, policy-makers,
epidemiologists, laboratory scientists, and program managers to
discuss the tools and methods needed for the diagnosis and
management of emerging, reemerging, and endemic infectious
diseases. The priority areas of the conference will be malaria,
Mycobacteria, leishmania, HIV, cholera, hepatitis, opportunistic
infections, enteric pathogens, antimicrobial resistance, molecular
entomology, and human genetic epidemiology. For information,
see MEEGID5@WWW.CDFD. ORG.IN or contact Altaf A. Lal,
Division of Parasitic Diseases, National Center for Infectious
Diseases, Centers for Disease Control and Prevention, MS F-22,
Atlanta, GA 30333, USA, e-mail: aal1@cdc.gov; Michel
Tibayrenc, Centre d'Etudes sur le Polymorphisme des Micro-
organismes, IRD, Montpellier, France, e-mail:
michel.Tibayrenc@cepm.mpl.orstom.fr; Hooman Momen, Instituto
Oswaldo Cruz, Av. Brasil 4365, 21045-900, Rio de Janeiro, Brazil,
e-mail: hmomen@gene.dbbm.fiocruz.br; or Seyed Hasnain, Center
for DNA Fingerprinting and Diagnostics, ECIL Road, Nacharam,
Hyderabad 500 076, India, e-mail: ehtesham@www.cdfd.org.in.
Deadline for abstract submission is April 30, 2000; 250-word
abstracts should be submitted to Dr. Hasnain at the above address.
Registration fee for abstracts submitted by April 30 is US$150.00.
Late fee for abstracts submitted by June 30 is US$250.00. Deadline
for acceptance for oral and poster presentations is July 31, 2000.

News and Notes

ICEID 2000—Register Now!
Nearly 500 abstracts in 36 categories were submitted for
consideration to ICEID 2000. Updated program information will be
posted on the conference Web site as it becomes final. Abstracts of
late-breaking research results may be submitted through the Web
site beginning in late March, until June 16. Go to the Web site at
www.asmusa.org/mtgsrc/ iceid99main.htm and follow the links to
abstract submission. ICEID 2000 is using the American Society for
Microbiology's Web-based Abstract Submission System. If you
have ever used the system before, you will be able to reenter the
system as a returning user.

Registration is limited! You can register for this important
conference through the Web site listed above, or e-mail
iceid@asmusa.org to ask for a copy of the printed preliminary
program whch contains forms for registration and housing.



Emerging Infectious Diseases
National Center for Infectious Diseases
Centers for Disease Control and Prevention
Atlanta, GA

URL: http://www.cdc.gov/ncidod/EID/vol6no2/ascii/vol6no2.txt

Please note that figures, equations, and tables are not available 
in ASCII format.  Many symbols do not convert to ASCII.