|
TB Notes Newsletter
No. 2, 2007
CLINICAL AND HEALTH SYSTEMS RESEARCH BRANCH UPDATES
MDR TB and XDR TB Clinical
Trials Design Working Group Formed
In 2004, over 424,000 cases of multidrug-resistant (MDR)
TB are estimated to have occurred worldwide,
representing an increase of 55% from 20001. These cases
accounted for 4.3% of all new and previously treated TB
cases. Persons with MDR TB are 54% more likely to die or
have treatment failure2. The recent identification of
outbreaks of XDR TB among HIV-infected persons3 has
emphasized the lack of controlled studies to define
optimal treatment regimens for MDR and XDR TB. A working
group has been formed under the aegis of the TB Trials
Consortium and is actively working to identify
strategies to best evaluate new TB drugs for the
treatment of MDR and XDR TB.
The goals of the working group include-
- Developing a research agenda for improving
outcomes of treatment of MDR and XDR TB,
- Identifying the most efficient path to new drug
regimens for MDR and XDR TB,
- Integrating animal and preclinical research with
clinical trials design,
- Stimulating funding for clinical trials of MDR
and XDR TB treatment by providing a defined pathway
for the future, and
- Defining compatibility between antiretrovirals
and drugs for MDR treatment.
The working group is developing a management plan to
guide policymakers and scientists, both clinical and
preclinical, in the effort to develop new therapies for
MDR TB. The group is emphasizing the need for
development of new MDR regimens in parallel with new
drug development for drug-susceptible TB. Members of the
group include Bill Burman (Denver Public Health), Peter
Cegielski (Division of TB Elimination, CDC), Mary Ann de
Groote (Colorado State University), Mark Harrington
(Treatment Action Group), Bob Horsburgh (Boston
University School of Public Health), Dikoe Makhene (DMID,
NIAID, NIH), Carole Mitnick (Harvard Medical School),
Sonal Munsiff (Division of TB Elimination, CDC), Nesri
Padayatchi (CAPRISA: Centre for the Program for AIDS
Research in South Africa), Jussi Saukkonen (Boston
University Medical Center), Neil Schluger (Columbia
University), Barbara Seaworth (University of Texas,
Tyler), Tim Sterling (Vanderbilt University School of
Medicine), and Elsa Villarino (Division of TB
Elimination, CDC).
—Submitted by C. Robert Horsburgh, Jr.,
MD
Boston Univ. School of Public Health
References
- Zignol M, Hosseini MS, Wright A, et al. Global
incidence of Multidrug-resistant Tuberculosis. J
Infect Dis 2006;194:479-85.
- Centers for Disease Control and Prevention.
Emergence of Mycobacterium tuberculosis with
extensive resistance to second-line drugs –
worldwide, 2000-2004. MMWR 2006;55:301-5.
- Gandhi NR, Moll A, Sturm AW, et al. Extensively
drug-resistant tuberculosis as a cause of death in
patients co-infected with tuberculosis and HIV in a
rural area of South Africa. Lancet 2006;368:1575-80.
Back
to Top
The Long Road to a Shorter,
Stronger, Safer Cure for TB – How to Get There Faster
How short is shorter, how strong is stronger, how
safe is safer?
There are common worldwide goals for a TB treatment
regimen: duration of 2 months or less; high efficacy
(99% cure rate) and intermittent dosing (once or twice a
week), with a relapse rate approaching 0%; and
better-tolerated therapy (decreased common and serious
side effects; fewer drug interactions, especially with
anti-HIV medicines). The good news in
TB treatment is that there are now at least six new
TB drugs in development. The bad news: it could be
several decades before we arrive at a shorter, more
effective treatment regimen which employs these drugs.
In 1962, TB treatment lasted 24 months, requiring
1460 doses and hospitalization, and was associated with
treatment failure rates of 10%, relapse in 2%, and
acquired resistance in 7% (Tubercle 1962, 43:201-67). In
1979, less than 20 years later, it had been shown that
treatment duration could be shortened to 6 months,
requiring 96 doses and no hospitalization, and with 0%
treatment failure, 2% relapse, and 0% acquired
resistance (Am Rev Respir Dis 1979;579-85).
It has now been 28 years since “short-course” therapy
was proven effective. For most of that time, there were
no new candidate TB drugs; no movement down the road was
even possible. Now, with six new drugs in the pipeline,
can we expect to make the kind of exciting progress that
was witnessed in the 1960s and 1970s? The recent
recognition of extensively drug-resistant (XDR) TB in
South Africa and other sites adds urgency to the need
for evaluation of novel agents to treat TB (MMWR
2006;55: 301-305).
Traditionally, clinical trials for new TB drugs are
based on 6- to 9-month treatment regimens, with
follow-up efficacy evaluations lasting 1 to 2 years to
detect any cases of relapse. Accounting for the
completion of phase I, II, and III trials, it can take
6–7 years or more to register a new drug for TB. Most
experts anticipate that more than a single drug
substitution in the multidrug TB treatment regimen will
be required to achieve a regimen that is dramatically
shorter and more effective than current treatment.
Because the effect of single substitutions doesn’t
predict the effect of combining multiple new drugs in a
treatment regimen, it would theoretically be necessary
to conduct trials with all possible permutations of the
six or eight drug classes contributing to a three- or
four-drug regimen to discover the “correct” combination.
This approach would take centuries to complete, and only
dumb luck could be expected to shorten the process!
Four strategies have the potential to provide
significant shortcuts to our destination: the use of
preclinical data from mouse models; the use of surrogate
markers (biomarkers) to predict relapse risk; a clinical
trials format designed to speed the evaluation of
multiple new agents in a multidrug regimen; and clinical
evaluation of novel agents through compassionate-use
protocols for patients with XDR TB.
The first shortcut to consider is the use of mouse
models to assist in choosing candidate regimens for
study. Despite the many obvious biologic differences
between mice and people, and the fact that the overall
course of TB infection in mice does not closely follow
the course of TB infection in humans (no real correlate
of latent TB infection), the mouse model of TB disease
has proved very useful in predicting the results of
human TB treatment. In the early development of TB
treatment, mouse models predicted some important tenets
of now-proven regimens: that INH provides two phases of
killing (early and rapid, then prolonged and slow) and
that the addition of rifampin and/or PZA provides
sterilizing activity (J Exp Med 1956;104:737-62,Tubercle
1978;59:287-97). Recent data from mouse models predict
the utility of quinolones in combination with a
rifamycin; these data were used to inform the design of
Study 28 (Nuermberger et al. Am J Respir Crit Care Med
2004;169:421-6).
The second shortcut strategy is to find and employ
biomarkers that can predict the risk that TB treatment
will be unsuccessful. TB treatment can fail in two ways:
1) it can fail to effect a clinical response (such as
weight gain or defervescence) or a microbiologic
response (sputum cultures rendered sterile) while the
therapy is being given; and 2) it can lead to relapse of
disease within 2 years after completion of therapy.
Because these failures of treatment will not be directly
evident for months or years after the start of a
treatment regimen, clinical trials can take several
years to reach these important endpoints. It is possible
to identify biomarkers that can be measured early during
the treatment period and for which a threshold value can
predict later treatment failure or relapse. Once the
predictive value of such a biomarker is established, the
biomarker result could become the primary outcome of a
clinical trial. For example, the lack of 2-month culture
conversion has been associated with increased risk of
treatment failure and relapse; this endpoint is a
primary outcome for a current TBTC phase II trial, Study
28. Using a biomarker measured at 2 months after
starting therapy (instead of 2½ years), it would be
feasible to complete a clinical trial in a year or less.
The problem with using a biomarker that has a
dichotomous result (positive or negative), instead of a
continuous result for a primary study endpoint, is the
relatively large sample size required to demonstrate a
difference in study arms. Potentially useful biomarkers
that provide results on a continuum include quantitative
sputum cultures, quantitative nucleic acid amplification
tests, and therapeutic drug levels (pharmacokinetic
sampling). In the past, quantitative sputum cultures
were frequently employed to determine the early
bactericidal activity of individual TB drugs (Jindani et
al. Am Rev Respir Dis 1980; 121: 939-49), but have not
been used as a primary endpoint for multidrug treatment
trials.
The third shortcut strategy is to pursue the
identification of a very promising new multidrug TB
regimen (utilizing one or more new agents) for
evaluation in a phase III trial by first conducting
sequential phase II trials. Concurrent phase II trials
can be used to efficiently identify the optimal dose of
each new drug (e.g., for each candidate drug, a trial
with two dose arms of 150 patients each with enrollment
and follow-up completed in 8 months). The next step
would be to use a multiarm phase II study to identify
the place of a new drug (X) or drugs within a regimen,
e.g., Arm 1: HRZX, Arm 2: MRX, Arm 3: MRZX, with 150
patients per arm with enrollment and follow-up completed
in 6 to 8 months (H=INH, R=rifamycin, Z=PZA, M=moxifloxacin,
X=new drug). As mentioned above, data from the mouse
model can be very useful in choosing arms for these
trials. A phase III trial of a new TB regimen will
require enrollment of 1400 patients or more (based on
80% power to detect a decrease of relapse or toxicity
from 5% to 2%) and require a minimum of 3 years to
complete enrollment and follow-up. By using sequential
phase II trials to design optimal new regimens for
inclusion in phase III trials, it is feasible to reach
the goal of a shorter regimen in the next decade.
The design of clinical trials to evaluate the
effectiveness of novel TB drugs could also be based on
the approach used for studying therapies for
drug-resistant HIV in treatment-experienced patients. A
common approach to the evaluation of novel agents for
HIV treatment has been to use 1 (or more) novel agent in
a controlled trial in which all patients (with
drug-resistant HIV) receive optimized background therapy
(or salvage therapy) based on currently approved or
available medications and a proportion are randomized to
receive the novel agent in addition (Lazarrin et al. N
Engl J Med 2003;348:2186-95). Indeed, novel drugs for TB
will likely have the greatest impact on the treatment of
MDR TB and studying new TB agents in patients with MDR
TB is a reasonable approach to demonstrate
effectiveness. Studies could be designed as add-on
studies in which a placebo is compared against the new
agent in the context of the background regimen
containing the best options that would otherwise be
available to the patient (optimized background therapy
or OBT). In such a design, patients with differing drug
resistance patterns may be enrolled because each patient
will have individualized OBT.
The discussion above was based on presentations given
by William Burman, MD, at TBTC Semiannual meetings.
Back
to Top
Study enrollment updates:
Study 24 is a single-arm study of largely
intermittent, short-course therapy for patients with INH-resistant
TB or INH intolerance. Enrollment closed December 2004
with a total of 98 patients. By mid 2007, all patients
will have reached the end of follow-up for study
outcomes (treatment failure and relapse).
Study 26 is a trial of short-course treatment of
latent TB infection among contacts of active cases,
using a 3-month once-weekly regimen of isoniazid 900 mg
and rifapentine 900 mg, compared to standard 9-month
therapy with isoniazid 300 mg. As of March 17, 2007,
Study 26 enrollment was up to 7046, over 88% of the
intended 8000 subjects for total enrollment. Enrollment
completion is anticipated by October 2007.
Study 27 was a double-blind, placebo-controlled
comparison of 2-month culture conversion rates when
substituting moxifloxacin for ethambutol in the
initiation phase of treatment of pulmonary TB.
Enrollment began in July 2003 and was completed in March
2005, with a total of 337 patients. Over 50% of patients
were enrolled from two African study sites. The primary
study results were published in 2006 and showed there
was no difference between study arms in terms of time to
sputum conversion (Burman WJ et al. Am J Respir Crit
Care Med 2006; 174: 331-338). There were differences,
however, between North American sites and African sites,
with significantly more North American patients
converting their sputum to negative by 2 months (84%)
compared to African patients (60%). Further analyses are
ongoing.
Study 28 is a double-blind, placebo-controlled
comparison of 2-month culture conversion rates when
substituting moxifloxacin for isoniazid in the
initiation phase of treatment of pulmonary TB. This
isoniazid-sparing regimen for TB treatment is based on
data from the murine model of TB; in this model, the
substitution of moxifloxacin for isoniazid resulted in
significant reductions in the time to culture conversion
and the time to sterilization when compared to the
standard combination of rifampin, isoniazid, and
pyrazinamide. Improved sputum culture conversion after 2
months of treatment with a moxifloxacin-containing
regimen would support phase-3 clinical trials of
moxifloxacin-based treatment regimens of less than the
current 6-month standard regimens. The plan is to enroll
410 patients from both domestic and international TBTC
sites. Enrollment began in March 2006 and was completed
in March 2007 with 433 enrolled.
—Submitted by Susan M. Ray, MD
Emory Univ. School of Medicine
Member, Advocacy & External Relations Committee, TBTC
Last Reviewed: 05/18/2008 Content Source: Division of Tuberculosis Elimination
National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention
|