BACKGROUND

The AIDS Malignancies Working Group (AMWG) was convened in response to emerging opportunities in research on HIV/AIDS-related malignancies and the rapid development of tools for pursuing crucial questions in this area. AMWG represents a spectrum of disciplines with an interest in AIDS malignancies and those conducting research relevant to these cancers. During this fifth meeting of AMWG, participants heard presentations on the epidemiology of AIDS malignancies, issues in immune reconstitution, and the question of HIV reservoirs and gene expression. They also heard an update on the Developmental Therapeutics Program, which aims to facilitate AIDS drug discovery and development. Participants then focused on three major issues: (1) epidemiology and databases; (2) virology, immunology, molecular biology, and drug development; and (3) prevention and treatment. After discussing these issues in subcommittees, the group reconvened to share conclusions and formulate recommendations.

MEETING OBJECTIVES

Dr. Ellen Feigal welcomed participants by pointing to the opportunities presented by AIDS malignancy research for studying the correlates between viral infection, immune dysregulation, and cancer pathogenesis. Recent events have provided new tools for this research, including the ability to quantitate HIV itself, as well as several new and potent antiretroviral therapies. Increased survival has given new impetus to research on underlying malignancies, but we may also be seeing changes in the incidence and natural history of several AIDS-related malignancies. As a result, this meeting is designed to focus attention on several related topics: the changing epidemiology of AIDS malignancies, issues in immune reconstitution, and HIV reservoirs. Participants were urged to consider the opportunities for research, the need for personnel and resources, and steps that the National Cancer Institute (NCI) can take to facilitate progress in these areas.

Participants were also asked to review and comment on the opportunities for the development of therapeutic interventions, and on the forum that is being developed for the exchange of information in AIDS malignancies (the Third National AIDS Malignancy Conference is currently scheduled for May 1999). NCI has responded to previous AMWG recommendations by developing new funding and initiatives for research, by instituting an AIDS oncology clinical scientist training program, and by organizing the national conferences. Extramural researchers now receive about 50 percent of NCI funding for research on AIDS or AIDS-related malignancy from the research grant pool, up from 17 percent in 1995.

PRESENTATIONS

Prior to breaking into subcommittees, there was additional discussion and three presentations relevant to all of the subcommittee deliberations.

Key Points

• It is probably too soon to know what impact HAART will have on the epidemiology of AIDS malignancy. Approximately two-thirds of HIV-positive persons are aware of their HIV status. While the majority are on some form of therapy, only a third of that number are on HAART, and perhaps less than 5 percent of all HIV-positive persons are on optimal HAART. There are many barriers to increasing this number C the drugs are expensive, may have intolerable side effects, and are complicated to take according to the regimens required.
• Data show that the total AIDS population in the United States has leveled off at about 700,000, and the number of deaths has declined. As a consequence, however, more patients are living with AIDS and thus at risk for its complications, including malignancies. In addition, the demographic profile of AIDS has shifted away from white male homosexuals to include greater numbers of minorities and women, as well as increasing numbers ofyounger patients. This, too, has consequences for the number of AIDS-related malignancies. For example, Kaposi=s sarcoma (KS) is less commonly reported in black males than in white males of the same exposure group, and younger patients are at risk for different cancers (e.g., more Burkitt=s and less non-Hodgkin=s lymphoma [NHL]). At the same time, however, the incidence of NHL has been rising in the general population since the 1970s, and this too has an effect on the incidence of AIDS-related disease.
• Five studies have produced actual data about cancer risk in the HAART era. The consensus appears to be that KS is declining, and that primary brain lymphomas (PBLs) may be declining, but that NHLs are not declining (see Table 1). Retrospective analysis of death certificates also shows a long-term decline in the number of AIDS deaths with KS, but a slight increase in the number with NHL.
• Analysis of data from 180,000 AIDS patients linked to cancer registries reveals that the number of AIDS patients with KS at the time of diagnosis has been declining sharply for years, long before the advent of HAART, from 17 percent in 1985 to only 2 percent in 1996, while the number with NHL has not declined. The incidence of PBL, immunoblastic lymphoma, or Burkitt=s lymphoma has also remained stable. Twelve-month followup data on these patients shows much the same patterns. The incidence of cervical cancer in women with AIDS has increased, but this may be the result of changing definitions of AIDS. Since trends in these cancers pre-date widespread use of HAART, it is difficult to attribute the changes in cancer incidence to HAART therapy. At the most, the effect of HAART therapy would have been limited since the most recent data were available only from the time HAART was just being introduced.

Key Points

• Issues surrounding host defense against malignant transformation are even less well understood than the mechanisms of immune deficiency in HIV disease. From natural history studies, the inexorable decline in the number of circulating CD4+ cells is a reasonable predictor of the short-term risk of opportunistic infection. However, it would be presumptuous to assume that this is also true for the risk of malignant transformation.
• HIV infection occurs more efficiently once the cell is activated. Evidence of immune activation in HIV disease includes heightened production of proinflammatory cytokines, tumor necrosis factor (TNF), IL-6, increased CD38, and HLA-DR expression on CD8+ cells. HIV replication appears to be the driving force for this immune activation.
• HAART not only reduces viral replication but also, at least in advanced cases, decreases immune activation. CD3 cells, B cells, and CD4+ cell populations all show a biphasic increase, peaking at about 12 weeks; evidence suggests that the slow increase in circulating naive CD4+ cell counts may represent thymic production. TNF levels fall, although they still remain above normal. Other markers of immune activation also decline, although not to normal levels. This suggests that, although viral replication is suppressed, it has not fallen to zero.
• Functionally, HAART appears to enhance lymphocyte proliferative responses with candida, but not to tetanus, over 48 weeks of therapy. This differential response may be related to the availability of antigen with restoration of responses to ubiquitous antigens and no return of responses to antigens that are not present in the patient. Importantly however, HAART does not enhance specific CD4 responses to HIV proteins such as GP-120, P-24, or P-66. It is unknown whether this is because the responsive cells are anergic or because viral replication has destroyed all HIV-reactive CD4 cells. Immunization studies are pending. It remains an intriguing question whether it would be possible to boost immune response to HIV peptides in the absence of viral replication. Patients who are not placed on HAART tend to lose CD4-dependent responses fairly early, while those who receive HAART tend to preserve the response.
• After immunization, patients with high numbers of memory cells at baseline tended to have stronger responses to recall antigens, while those with higher levels of naive cells (those recently produced by the thymus) tended to have stronger responses to neoantigens. The risk and predisposition to the development of malignancies in persons with HIV-1 infection are not well understood. B cell activation, the magnitude and duration of immune cell deficiency, and other factors may be more important than the number of circulating CD4 cells.

Key points

• The lymphoid system is the prime target of HIV infection. Within lymphoid tissues there are two principal reservoirs or cellular compartments of HIV RNA: (1) productively infected mononuclear cells and (2) intact virions contained as antibody-antigen complexes on the surface of follicular dendritic cells (FDCs) in germinal centers. Quantitation of in situ hybridization signals is achieved by affixing radiolabeled riboprobes to HIV RNA, depositing silver grains through autoradiography and then counting them; the number of grains will correspond to the viral load according to known formulae.
• Using this assay, investigators have determined that FDCs contain 10 to 50 times the amount of HIV RNA that is found in mononuclear cells, and 100 to 1000 times the amount found in peripheral blood. Cellular viral load is lower in vivo than in vitro systems, possible because of suppressive functions (cells with more copies become stronger targets for immune response) or because of differences in viral gene expression in vivo versus in vitro.
• Evaluation of serial tonsillar biopsies from a Dutch study of patients receiving HAART found there was a two-phase decline in tissue viral load. Surprisingly, the rate of decline was similar in FDCs and productive mononuclear cells, suggesting that HIV does not act as a conventional antigen in vivo. In addition, the first phase of the decline seems to correspond to the clearance of those productively infected mononuclear cells with the greatest number of RNA copies, and the second phase to the elimination of less productively infected cells (still predominantly CD3+).
• After more than 12 months of HAART, other investigators have found residual latent reactivable HIV from CD45RO memory cells. Even after 24 months, application of in situ hybridization in lymphoid biopsies occasionally reveals cells with less than 5 HIV-RNA copies. While patients appear to have cleared their FDC compartment, the persistence of p24 antigen by immunohistochemistry of some FDCs suggests that there may be continuing low-level transcription, below the limits of nucleic acid detection assays. Immune reconstitution may be measured by quantitative immunohistochemistry in lymphoid tissues of individuals treated with HAART. Cell counts reveal that the initial rise in CD4 cells are memory cells, while later increases are naive cells. CD35 staining shows that HAART restores the overall integrity of the follicular dendritic cell network.
• Tissue sanctuaries for HIV infection include the central nervous system, genital organs, and perhaps the lung, eye, thymus, and bone marrow. Some evidence suggests different levels and types of transcription at different stages of infection. To investigate this possibility, investigators are developing probes to detect class-specific levels or programs of transcription in tissues.

Key Points

• Her research focuses on the role of KS-associated herpesvirus (HHV-8) in the pathogenesis of three diseases to which it has been linked: KS, primary effusion lymphoma (PEL), and multicentric Castleman=s disease (MCD). The latter is not a malignancy but rather an atypical lymphoproliferative disease seen in AIDS patients. The goal is to localize and determine the extent of infection within representative lesions, characterize the programs of viral gene expression at the cellular level, and identify cellular and tissue reservoirs of infection. Tools include infected cell lines and tissue specimens, riboprobes specific for cellular and viral transcripts that allowed investigators to differentiate between latent and productive infections, and immunohistochemical techniques to phenotype cells within a specimen.
• Immunohistochemistry combined with in situ hybridization for viral transcripts (T0.7, T1.1, MCP) made in latent and/or productively infected cells indicated that most, if not all, of the KS spindle tumor cells are infected in both AIDS and non-AIDS patients. Similarly, most cells of the PEL are also infected. In both diseases the majority of these cells are latently infected and a minor subset contain T1.1 and MCP transcripts characteristic of productive infection. In MCD, however, relatively few cells of the lesion are infected and all of these appear to be productively infected. If these few cells contribute to, or are responsible for, the pathogenesis of MCD, they do so in a very powerful paracrine fashion. The investigators recognize that in these studies their ability to identify infected cells is dependent upon detectable levels of RNA transcripts and relies heavily on the selection of probes as well as assumptions about the programs of viral gene expression, in vivo. Thus, they are in the process of developing a sensitive method of detecting single copies of viral DNA.
• Investigators are now screening these lesions with a battery of probes for other viral transcripts. HHV-8 carries several genes that are homologues of human genes, some of which normally function in proliferation and differentiation and therefore may be involved in tumorigenesis. An example is the viral homologue of human IL-6 that drives B cell proliferation and differentiation. In situ hybridization showed high levels of viral IL-6 transcripts in both PEL (a B cell lymphoma) and MCD (characterized by plasmacytosis), suggesting that it may be involved in the pathogenesis of these diseases. KS specimens did not show significant hybridization to viral IL-6 probes, suggesting that it does not play a major role in the development of this lesion. Differential expression of viral IL-6 among the three diseases potentially relates to the phenotype (endothelial vs. lymphocytic) of the predominantly infected cell in each case.
• In general, productively infected cells contained an abundant and complex array of viral transcripts including those for late genes while latently infected cells exhibited a highly restricted profile of viral transcription. These data are derived from infected cells within, or isolated from, disease lesions of immunocompromised individuals and may not reflect the programs of latency in infected cells in otherwise healthy individuals.
• Investigators found that one of the HHV-8 probes (T0.7) hybridized in a focal manner to glandular epithelium of a significant number of prostate specimens, but not to tumor cells in those that had adenocarcinoma. Further screening of HIV-negative Italian and U.S. prostate samples obtained from the NCI revealed that roughly one third were T0.7-positive. Poor correlation with PCR and serological data generated at the NCI from the same samples may be due to insufficient sensitivity of current serological assays and/or small size and artifacts of sampling of the tissue specimens.
• The strategy for investigating reservoirs and transmission of HHV-8 is based in part on the epidemiology of KS. Target tissues include male and female genitourinary tract, gastrointestinal tract, oropharyngeal tissue, and breast. To increase the likelihood of finding viral reservoirs, samples will initially come from populations with an increased incidence of KS; HIV-positive and negative homosexual men and endemic areas such as Italy and Africa.

• Dr. Staskus explained that prostate studies are complicated by the lack of a good serological assay for HHV-8 and access to representative surgical specimens. To accentuate this problem, the era of HAART, with a decreased incidence of KS, PEL and MCD, makes it increasingly difficult to get new specimens and the research has relied primarily on pathology archives. Samples collected in previous and ongoing clinical trials will be useful, but thus far few collaborations or cross- validations have been attempted. A collaboration has been initiated to take advantage of a study of HIV-discordant couples currently underway in Minnesota.

AIDS DRUG DISCOVERY AND DEVELOPMENT: UPDATE ON REVIEW PROCESS DR. ELLEN FEIGAL AND DR. EDWARD SAUSVILLE

Key Points

• The 1995 AIDS review made several recommendations with regard to AIDS drug discovery and development. NCI has convened an outside advisory panel which is currently examining three areas:
• What are the gaps, opportunities, and targets for drug discovery and development?
• What should our national strategy be, given the research currently underway at NIH, universities, and other centers?
• What role should NCI play in supporting and promoting this strategy?

• As a complement to this effort, Dr. Feigal asked AMWG participants to consider what opportunities there might be for drug discovery and development in the area of AIDS malignancies.
• The cancer portion of the AIDS Developmental Therapeutics Program Review has been completed and the report accepted by the NCI Board of Scientific Advisors. That group reached consensus that there is a role for the NCI in screening activity directed at molecular targets that might be relevant to the pathogenesis of cancer. The implementation of that recommendation will heavily depend on extramural collaborations with the NCI.
• NCI would therefore be interested in hearing from AMWG about screening opportunities for collaborative research in finding high-affinity interactors with targets relevant to the development of virus-related malignancies. Results would be returned to the extramural community for further workup, possibly involving Centers of Excellence for in vivo screening or pharmacology.
• In the discussion that followed, it was suggested that the DTP was trying to do a better job of outreach to the extramural community. Extramural researchers have already teamed with small and large biotech companies to do high-throughput in vitro screening. The question is how best to facilitate the transition from basic to applied research. The DTP already has the capabilities to do this kind of screening, but academic centers are also setting up relevant centers; the issue is how best to coordinate and integrate these efforts in order to achieve greatest efficiency. There may also be opportunities for internal collaboration between NCI, National Institute of AIDS and Infectious Diseases, and OAR.
• Funding mechanisms have not yet been determined but might include RO1s, cooperative agreements, and small Phase I studies overseas. Participants suggested that one important opportunity would be to develop preventive interventions, e.g. screening for drugs to treat individuals who are at risk for viral lymphomas. A second opportunity would be to invest in early pathogenesis studies that might identify potential target genes; it was thought that this approach had been successful in HIV studies. A third opportunity would be to determine the role of HIV as a cofactor in promoting KS.

During and after lunch, participants met in breakout sessions to address particular issues and opportunities in three areas: (1) epidemiology and databases; (2) virology, immunology, molecular biology, and drug development; and (3) prevention and treatment. At the end of the day they reconvened to hear the reports and recommendations of these three subcommittees, which are summarized below.

EPIDEMIOLOGY AND DATABASES DR. WILLIAM BLATTNER

• The role of NCI should be to stimulate relevant research with initiatives backed by a financial commitment. The subject of these initiatives should reflect the recommendations of the AMWG.
• There is an overarching need for interdisciplinary approaches and a critical need to share tissue samples and reagents. Action should be taken to ensure access to data relevant to AIDS malignancies C for example, getting the AIDS Clinical Trial Groups to provide access to their data bearing on the impact of AIDS therapies on the incidence of viral malignancies. Members of the audience noted that some institutions are actually disposing of tissue samples; the suggestion was made for NCI to establish a central repository or clearinghouse for both samples and data.
• Specific recommendations included a focus on several specific areas:
• Research on the pathogenesis of NHL, including classification of pathogens at the molecular and virologic levels; trends in virologic failure and CD4 dissociation; screening for potential interventions; and trials of those interventions in the form of vaccine or therapy.
• Research on the domestic and international epidemiology and natural history of HHV-8, both molecular and serological. Funding is needed to develop and cross-validate HHV-8 assays. Efforts are needed to provide access to both samples and reagents, and to capture the results of outside studies. Specific attention should also be given to understanding HHV-8 gene expression over time and other potential risk factors for the development of KS.
• Research on other cancers occurring in the setting of AIDS. An example is cervical cancer, with an incidence modestly higher in HIV-positive women. This presents an opportunity to study virus-host interactions with regard to HPV, as well as the impact of HAART on HPV. Similar research opportunities exist for anal cancers, lung, and other cancers in AIDS patients. The carcinogenic effects of therapeutic drugs used to treat the underlying HIV disease might also be another area of investigation.

• In addition to the basic research outlined above, there is also a need for long-term clinical monitoring and research over the next five to ten years, both to track the evolution of AIDS-related malignancies, and to develop and test potential therapies. Three specific examples were offered:
• Antigen- and epitope-specific studies in a clinical setting to determine when patients on HAART or modified HAART regimens are at risk for malignancies including cervical and anogenital cancer, EBV and non-EBV associated lymphoma, and KS, and to test interventions such as in vitro expansion and readministration of CD4 and CD8 cells, or using cytokine and other immunoadjuvant strategies to restore immune response in HIV-positive patients. Small studies of the effect of DNA-based vaccines would also be desirable. Such studies are being done in a few centers today; it would be desirable to have them ongoing in numerous centers. It would be important to determine the role of HAART in attenuating the immune deficits.
• Further HHV-8 studies are needed to improve on our current state of relative ignorance with regard to how this virus persists in human populations and in the tissues of specific individuals. Clinical cellular immune and viral parameters (other than antibody titer) that define the population at risk for KS need to be identified. Therapeutic approaches for the Aat risk@ population need to be developed. Program officers should examine whether the current research portfolio is sufficient to produce the needed knowledge in a timely manner. Because this topic falls between the areas of virology and immunology, careful coordination and collaboration will be needed.
• In the area of therapeutics, the current bottleneck is at the transition from basic research to preliminary screening in tissue and animal models to identify potential targets at the molecular and cellular levels. The emphasis should be on identifying targets and developing leads that will be of interest to pharmaceutical companies.

• A number of specific actions could help to expand current drug development processes to include AIDS-related malignancies:
• Include tumors related to HIV (including those associated with HHV-8, Epstein Barr Virus, and HPV) in ongoing cell screening programs and in efforts to examine genetic differences between normal and malignant tissues.
• Develop novel and more useful in vitro assays and animal models for addressing questions about the virology and immunology of these tumors.
• Develop a standard sequence for testing potential drugs, in order to create a uniform Apipeline@ that will help pharmaceutical companies move drugs into clinical testing.
• Institute more aggressive programs for acquiring blood, tissue, and tumor samples, along with information on the clinical outcome of candidate therapeutics.
• Bring proposal review for viral oncology back into NCI study sections, where they will be better received.
• Encourage AIDS Associated Malignancies Clinical Trials Consortium (AMC) to include tumors beyond KS, lymphoma, and other immune deficiency states.
• Pursue gene-based stem cell transplant therapies for cancer patients.
• Encourage greater interaction and collaboration between the basic virology and oncology communities, both within and beyond the context of AIDS-related malignancies via sponsored meetings.


• Consider alternating the National AIDS Malignancy meeting with the International AIDS meeting.

There was a consensus that AIDS-related malignancies are a significant and growing problem, and that specific initiatives will be needed to accelerate the pace of discovery in this research area.

• A specific problem that needs to be solved is providing access to samples and data from existing studies, particularly longitudinal studies in patients before they acquire cancer, such as in the ACTG trials. There should be an open and competitive mechanism for gaining access to tissue samples for meritorious research proposals. Similarly, there are many opportunities to gather additional information, or at least follow patient cohorts, in current and proposed clinical trials.
• Priorities for research include the screening of more targets, to decide which might be important to develop into drugs.
• Development of animal models that are relevant to viral malignancies. At the basic level, there are unique opportunities to study the interaction between viral infection and immune response.
• There is a widespread perception in the extramural community that these research areas have a low priority at NIH. One action that might correct this perception would be to move viral lymphomas from Opportunistic Infections Special Interest Group (SIG) to the Clinical Oncology SIG, with ad hoc expertise as needed in viral oncology. However, there may not be enough proposals at present C 40 per year are needed to justify a separate study section, and at present there are only 5 to 10 per year.
• Dr. Feigal summarized the potential action items for the NCI to consider, and suggested that AMWG participants send her any further comments by email or fax. The NCI would welcome their ideas on ways to ensure that the work of the AMWG remains a dynamic ongoing process.