Earlier this year, the CDC announced that the forecast already had come true: new HIV diagnoses increased by 14% among US MSM between 1999 and 2001.2 New York State and California—where the largest US MSM communities reside, although they are excluded from these data—provide consistent and equally alarming indicators: multiple behavioral measures coincide with unprecedented outbreaks of syphilis3 and increasing rates of rectal gonorrhea.4

The first wave of HIV gathered force and carried off thousands of MSM before the rising crest became evident. This time, there was ample warning. What is going wrong?

• Condom use, which is simple in principle, is difficult in practice.
• Changes in the milieus in which HIV spreads—cultural, attitudinal, technological—elude behavioral surveillance and outpace behavioral interventions.
• Development of biomedical interventions languishes through lack of resources, capacity, and strategy.
• Prevention and treatment research have been dichotomized instead of interwoven, to the detriment of both.
• Moralism—homophobia in particular—exacerbates rather than alleviates the threat.

When 2 men of differing HIV status engage in anal sex, an infected insertive partner puts an uninfected receptive partner at risk unless he puts on a latex condom correctly before his penis comes in contact with5 or enters his partner’s rectum. To avoid weakening the rubber, lubricants used with latex condoms must be water based. These lubricants must not contain the spermicidal detergent nonoxynol-9 (N9), which erodes patches of epithelial tissue from the walls of the rectum6 and exposes highly susceptible cells underneath.7,8 The condom must be withdrawn to ensure that no semen is spilled into the rectum. It’s that simple—and that difficult—to prevent almost all9 HIV transmission between men. What makes it so difficult?

News—often tentative, sometimes confusing, and occasionally contradictory—may be indigestible and typically travels slowly. For example, a recent study5 suggests that pre-ejaculatory fluid, which many MSM assume carries far less risk than ejaculate, may equally be able to transmit infection. Something serious to worry about—provided that further research corroborates the concern. News about the harmful effects of N9 several years ago remained so poorly disseminated that, a year later, one fifth of HIV-positive men believed N9 was more effective than condoms in preventing HIV transmission and almost 15% had used N9-containing lubricants instead of condoms.10

For MSM who pay attention to new research, it is very confusing to hear that treatment with highly active antiretroviral therapy suppresses HIV so effectively that treating the positive partner in a HIV-serodiscordant couple will greatly reduce the risk of sexual transmission during unprotected sex,11 but that men with excellent virological and immunological responses to antiretroviral therapy may intermittently shed high levels of HIV in semen.12–15 If there’s so little HIV in your plasma that it cannot be found, even by “ultra-sensitive” tests, should you or shouldn’t you?

Leaving aside knowledge—which is comparatively easy to impart and assimilate—there is the practical question of means. Do condom, correct lubricant (not spit, nor the complimentary container of hand cream in the motel washroom), and an enthusiastic partner always coincide? If not, if it’s a “one-night stand” who has only an hour or so to spare, what are you going to do? Latex barriers make sex less stimulating or pleasurable. When sex occurs with someone you deeply care for, using a barrier may feel less intimate.16

Life is short and then you die: so many men you’ve known have had their lives cut short by HIV, or suicide, or an overdose, or a gunshot, or heart failure. Or an accident: the driver opens his door just as a cyclist speeds by; a friend is left comatose for the rest of his short life. Commercial jets smash into a World Trade Center tower; another friend is crushed in the rubble. So enjoy the “good years” as long as they last. With better and better drugs, surely you can expect 10 or 15 good years even if you become infected tomorrow. Only a few years ago, people had to gag down handfuls of pills several times a day; now it’s just a few pills once a day. Why wouldn’t the drugs keep getting better and better, less and less toxic? Finally the drug companies are on it.

Besides, what’s there to live for? Do you want to be stuck in the same boring job, or maybe a worse one, or out of a job, or still in debt, or back in jail? Are there any gay men 15 or so years older whom you admire? Just once, you “blow it,” you “slip.” What’s the point anyway? Now you’re probably infected. Even if you’re not, how are you ever going to stick to a steady diet of condoms with all partners, all of the time until the elusive or illusory state of “mutual monogamy” somehow coincides with an HIV-negative partner you can love, desire, and trust?

This is an enunciation—not an explanation, certainly not a justification—of the most superficial obstacles that make consistent condom use challenging even for the best educated and informed, least impoverished MSM in the United States. Just how steady a diet of condoms really is necessary? Who knows? Estimates—mostly using data collected either before combination antiretroviral therapies came into widespread use or as HIV-infected men began to receive these regimens—agree on an approximate risk of transmission during unprotected anal sex with an infected partner of about 1 event per hundred episodes, although a single unprotected contact can suffice.17 For populations of MSM, what’s the target? What level of improved condom use is required to reduce the aggregate incidence rate by half, or by tenfold? Who knows?

Figure 2 Image courtesy of Better World Marketing (http://www.socialmarketing.com)

Behavioral interventions to promote condom use—the only strategy currently available to stem the MSM epidemic—are failing. Under the best of circumstances, when proven effective, they are not promptly and easily disseminated, and certainly not straightforwardly translatable across the age spectrum, ethnic groups, and subcultures. Without sustained but costly quality assurance procedures, they drift. They may be prone to unintended consequences. What if counteradvertising designed to persuade HIV-negative men that side effects from antiretroviral drug cocktails are no party18 inadvertently discourages adherence among men already being treated with the drugs,19 and thereby facilitates the emergence of drug resistance?20 As drugresistant HIV spreads through a community, treatments lose their potency.

Figure 3 Images courtesy of Project Neon/Program of Seattle Counseling Service.

Even when optimally implemented, behavioral interventions may be overwhelmed by technological change. The efficiency of the Internet as a means by which men form sexual liaisons21 greatly surpasses the efficiency of our outreach methods. Facilitation of negotiated “barebacking”—the professed preference for sex without condoms—has created a visible subculture22 that would not have dared speak its name when men ravaged by HIV disease hobbled down the streets of every “gay ghetto” in the country. The Internet’s potential to transform patterns of MSM mating in rural areas has yet to be analyzed. Internet sites may facilitate sex among groups as well as dyads, adding innumerable private venues for sex to the proliferation of slightly more public after-hours sex parties and rejuvenated bathhouses. Unless disclosure of HIV status accompanies the organization of such groups, group settings are much less likely to encourage disclosure of HIV status than more discreet “pillow talk” between members of a dyad.

The unprecedented and poorly understood effects of Internet-mediated sexual encounters need to be considered in the context of other changes that affect MSM sexual behavior. Behavioral surveillance systems were slow to apprehend the spread and significance of methamphetamines23 and other “party drugs.”24 Their expanded production, distribution, and use, together with easy access to licensed pharmaceuticals such as Viagra and steroids, may energize sex binges lasting days. Put these technologies together, and the potential for the incendiary spread of HIV and other sexually transmitted diseases seems unstoppable.

We cannot abandon behavioral strategies, such as they are, because they are all we have. But we need new biomedical approaches that appeal to MSM and that can be woven into daily life without making superhuman demands on adaptability, adherence, or rectitude.

Disappointing results announced last February 25th from the first efficacy trial of a preventive HIV vaccine, AIDSVAX B/B, were not unexpected, nor are they disheartening. This first trial successfully enrolled nearly 5000 US MSM at high risk; almost all complied with the rigorous requirements of participation in a 2-year experiment. The trial sponsor, VaxGen, never expected AIDSVAX to be highly protective and anticipated that counseling to encourage safer sex would be indispensable, since this would ensure that an expectation of protection would not increase risk behavior so much that it offset the benefits of immunization.26 No other HIV vaccines in advanced development are expected to be fully protective.27 Thus, while continuing to pursue preventive HIV vaccine research, we need the following other biomedical approaches as well:

• Continued research on barrier methods—improved male condoms and the equivalent of the “female condom” for protection during anal sex.
• Extending the concept of vaginal microbicides—topical products such as gels, creams, or suppositories that can afford women protection when a male partner does not use a condom—to anal intercourse, through concurrent research on and development of rectal microbicides.
• Exploring the utility of antiviral drugs to effect preventive as well as therapeutic benefits.

These objectives will require resources, strategic planning, effective management, and organizational capacity, all of which now are woefully inadequate (box on this page).

Consider rectal microbicides, a concept that MSM have again33 and again34 and again35 and again36 told researchers they want and that they would use. Although the CDC and the NIH37 recently began to fund small-scale studies to identify appropriate criteria for assessing the safety of rectal microbicides, the intent is only to guide product labeling should vaginal products, misapplied rectally, prove harmful. Strategic plans from the principal agencies in the Department of Health and Human Services whose missions encompass rectal microbicide development—the CDC and the NIH Office of AIDS Research—propose an uncoordinated array of research topics or projects—the antithesis of a coherent, strategic product development plan.

The CDC’s planning document for HIV prevention through 2005 adds rectal microbicides to preventive vaccines and postexposure prophylaxis in one strategy, but it lists development of these technologies as the lowest in priority of 9 strategies to stem the spread of HIV among MSM: one of 223 total strategies listed among 34 objectives for addressing the worldwide HIV pandemic.38 The CDC has not updated its Guide to Microbicide Research and Development—a collection of research abstracts—since 1996.39

The Office of AIDS Research, responsible for NIH goal-setting and budget allocation for HIV research, prides itself on having devised “a unique and effective model for developing a consensus on scientific priorities.” Microbicides recently joined the plan as one of 7 “cross-cutting areas” successively grafted onto 5 standing “areas of emphasis.” The “strategic plan” for microbicide research alone lists 65 strategies—the word “rectal” appears in 6—among 6 “objectives” that have some unspecified relationship to 7 “priorities.”40 As an example of the specificity of these “strategies,” one reads: “Facilitate every aspect of product development in preparation for clinical trials.”

However robust the federal fiscal budget for biomedical research and development is, industry monopolizes the intellectual capital. Established approaches to planning and program management, organizational capacity, and economies of scale all make industry an efficient mechanism for developing and testing drugs, diagnostics, devices, and vaccines; making the case for their approval to regulators; and manufacturing and distributing products once licensed or registered. Industry has been a reluctant partner in prevention research and product development.

A Fail-Safe Strategy: When in Doubt, Conduct the Trial One consideration in the 1994 decision to abandon a federally funded trial of AIDSVAX was the anticipated performance of a different kind of vaccine construct that lurked “just around the corner.”43 Ironically, last year the NIH opted not to proceed with phase III trials among US MSM of a vaccine based on that same construct. A protocol developed by the Department of Defense to test a similarly designed candidate in a large trial in Thailand remains on the drawing board, with no outcomes expected before late 2008, and no other vaccine expected to perform better than AIDSVAX is expected to enter efficacy trials before 2004.44 Even if the usual delays associated with such challenges as manufacturing scale-up, regulatory approval, and qualification of clinical research sites are minimized, no new vaccine could be licensed and available for US distribution before the end of the decade. While the “next best thing” animates the enthusiasm of basic scientists, product developers need to be supported to test what already is on the shelf and ready to enter clinical trials. In the popular mind, success has been equated with an affirmative answer—it worked—and failure with the reverse. In fact, failure would mean a trial so flawed in its design or conduct that it provides no answer. Success means that we obtained important new knowledge to advance development, even when the primary outcomes are disappointing. Well-designed trials contain stopping rules to minimize harm to volunteers. Regulators and ethics boards that approve trial designs and monitor studies while in progress consider no aspect of preclinical and early clinical data, study design, and interim findings more important than safety. Large-scale trials often contain planned interim analyses that help minimize time and the investment of resources, including the precious human resource of volunteers. These analyses make it possible to declare success early if an intervention is much more effective than expected, and to terminate trials that cannot answer the question—for example because the host community has no interest in participating, or the event rate is so low that an answer will never be obtained. Program officials who give the green light do not need to second-guess the wisdom of those multiple checks and balances and add still more advisory panels and other hurdles to the approval process: red tape already strangles us. Data from well-conducted field trials advance the field of research. Results from VaxGen’s trial of AIDSVAX B/B demonstrated no measurable benefit for MSM in the trial overall. However, initial reports of protective effects among African Americans were suggestive, no matter what the full explanation. These data raise new points that may illuminate vaccine development and evaluation: If vaccine efficacy differed among demographic groups, and if such differences correlate with differential responses to immunization or differences in exposure to HIV, those findings might guide efforts to improve the product’s protective efficacy. Subtle variations in the distribution of HIV subspecies circulating in distinct US populations, if correlated with the differential potency of vaccine-induced responses, would be informative for both vaccine development and the epidemiology of HIV transmission. Unlike other participants, African Americans who received the vaccine were 3 times more likely than African Americans who received placebo to fail to complete enough of the study to be included in the outcome analysis, a puzzling finding in a double-blind study. Understanding the reason would help in designing future efficacy trials. Even when a product lacks efficacy, vaccine developers benefit from knowing that the immune response elicited by a vaccine confers no protection. Specimens and clinical data from this trial will markedly advance our understanding of the virological and immunologic characteristics of acute and early HIV infection. Our current understanding of the critical early weeks of HIV infection has been based mainly on studies of patients seeking medical care for nonspecific flulike symptoms in settings in which clinicians were sufficiently astute to suspect acute HIV infection as a possible cause. From the large prospective VaxGen cohort, among whom periodic laboratory testing identified incident infection, we gain a unique glimpse into the prevalence and characteristics of asymptomatic acute and early HIV infection. Future vaccine trials—designed to evaluate products for their impact on minimizing disease after “breakthrough” infections, as well as their ability to forestall infection altogether—will gain essential population-based data on the profile of acute and early disease. The CDC-sponsored Vision Sub-Studies Network45 recruited MSM into a parallel nontrial cohort using methods similar to those employed in the VaxGen trial; the comparison will provide data on the impact of vaccine trial participation on risk behavior, addressing yet another hypothetical that derailed the 1994 trial. As with AIDSVAX, even if the results fall short of the benefits it hoped to find, another, concurrent large-scale randomized prevention trial—EXPLORE, whose baseline findings are presented in this issue of the Journal46,47—will contribute to improved behavioral intervention strategies. Estimating from empirical data how much behavior change and of what type (e.g., reduction in number of episodes, increase in the proportion of contacts where condoms were used) translates into modified rates of incident infection may clarify which behavioral outcomes are the most salient targets for further intervention development and which aspects of EXPLORE counseling were most strongly associated with modifying them. EXPLORE data can help characterize attitudinal and psychological profiles of MSM who do and do not continue to participate in counseling and for whom counseling does and does not coincide with behavioral change. More refined profiles of responders and nonresponders to different amounts of counseling or specific counseling content may in turn support optimally efficient delivery of the most effective amounts and influential components of such interventions to identifiable subgroups of men at risk. For industry sponsors of biomedical products, reliable government partnership in supporting field trials provides an incentive for sustained research and development. In developing countries the benefits are even more profound, because expanded capacity to conduct clinical research strengthens the overall health care system. More health professionals learn, in the context of preparing for and conducting research, how to deliver clinical care and administer antiretroviral therapy to widening numbers of patients. Secondary community benefits include expanded facilities for HIV testing, access to trained health professionals and improved infrastructure, increased integrity of medical record keeping and management systems, and economic investment in a “clean industry,” with its multiple byproducts. Figure 4 Photo by Steven Underhill.

When the NIH abandoned plans to test preventive HIV vaccine candidates from Chiron and Genentech in 1994, Chiron curtailed its HIV vaccine program, while Genentech spun off VaxGen to pursue venture capital to pay for a trial the government was expected to underwrite. Other pharmaceutical companies all but stampeded away from HIV vaccine development,41 until lured back by federal programs that assumed much of the financial risk of development and by multinational efforts to establish funds to buy vaccines for the poorest, but most afflicted, nations.

In comparison with HIV vaccine development where industry has been skittish, “big pharma” has abstained from microbicide research. Federal programs for topical microbicide research and development, similar to those that reinvigorated HIV vaccine research, have barely begun.42 Some smaller biotech companies have become engaged, but will they remain in the game if the federal government and other sponsors waver in their commitment to underwrite the large-scale trials needed to evaluate effectiveness (box on page 876)?

Perhaps most discouraging, lack of a coherent strategy for rectal microbicide development among government and nongovernmental organizations engaged in prevention research is matched by a complete absence of industry interest. An assumption that the market for products specifically designed to confer protection during anal sex is restricted to MSM very likely contributes to a lack of attention to these technologies. But anal sex does not equal “gay sex”; it can be an equal opportunity pleasure: in fact, the few data sets available suggest that many more women worldwide are at risk through unprotected anal sex than MSM.48,49

Treatment and prevention remain unduly bifurcated in the very structure of federal programs. The CDC supports prevention. The Health Resources and Services Administration and the Health Care Financing Administration underwrite care and treatment. The NIH supports research in both areas, but through different administrative units, and its multicenter networks conducting the lion’s share of clinical research are defined by distinct agendas: treatment or prevention. Only as concerted efforts focus on the global pandemic are there signs of integration of these synergistic agendas—HIV prevention research has been a principal engine of expanded HIV diagnostic and treatment capacity—but in the United States they remain largely segregated.

For example, strikingly missing from many studies that have assayed the amount of virus in the semen of men receiving highly active antiretroviral therapy12–15 is any attempt to identify optimal drug combinations to suppress seminal viral load. With 16 potent antiretroviral drugs licensed at the beginning of 2003 and several more queued up for regulatory approval, many different combinations might offer similar clinical benefits for patients. Why have multiple possible combinations not been evaluated for their concurrent preventive potential (box on page 879)?

When the CDC warned 2 years ago of the impending second wave of epidemic HIV among US MSM, recommendations included efforts at the “community, structural, and policy level” to “address the effects” of homophobia and stigmatization as underlying causes.1 Amplifying on this theme, the 2005 CDC HIV Prevention Plan notes that homophobia inhibits prevention “at all levels,” not least “the broader culture, which delivers anti-gay messages, institutionalizes homophobia through structural mechanisms, such as laws that regulate intimate sexual behavior, and lags in its support of sensitive and honest prevention for gay and bisexual youth, young adults and older men.”38

Despite such warnings from the CDC and the Institute of Medicine,59 prevention efforts fall prey to political opportunism, misplaced moralism, stigmatization, and homophobia. The popularity of the sitcom Will and Grace has no more undone discrimination against and stigmatization of homosexuals than America’s embrace of The Cosby Show eradicated racial disparities or tensions. Many distraught parents still evict gay youths, propelling them into lives of inner-city scavenging and vulnerability to exploitation. Most schools continue to refrain from even the meekest adaptation to gay adolescents’ needs for safety and mutual affiliation, much less the authentic respect that might nourish self-respect. In turn, whatever normalization school and after-school settings might provide for concurrent emotional and sexual maturation among heterosexual adolescents is denied to most of their gay counterparts, who instead are apt to retreat into furtiveness, shame, or precocious pairings with older partners.

How is the CDC to counter homophobia in Congress,60 which sets its budget, or in the executive branch, where the whims of a presidential or departmental advisor can wreak havoc? Public health officials find themselves engaged repeatedly in rearguard actions to defend what already has been achieved. The CDC, the Food and Drug Administration, and the NIH had to scurry to assemble an expert panel61,62 to counter a war on condoms led by Tom Coburn (former Republican representative from Oklahoma, subsequently selected to chair the President’s Commission on HIV and AIDS).63 Even the director of the CDC had to step up to the plate to exonerate the San Francisco Department of Health from repeated declarations by Rep Mark E. Souder (Republican from Indiana) that it had violated federal law by funding “obscene” prevention programs like Stop AIDS (delivered in the privacy of gay men’s homes).64,65 “Defend” marriage against same-sex couples and demand—despite evidence that such policies are counterproductive66—that schools substitute “abstinence only-until-marriage” programs for condom education and availability67: a perfect formula for tuning out gay students at greatest risk.68 As many heterosexuals have discovered, maintaining rewarding sexual and emotional relationships can be challenging, even with the legal, contractual, familial, social, economic, religious, and innumerable cultural supports that exist for marriage. Subtract those incentives and what do you have? “Higamous hogamus, woman’s monogamous. Hogamous, higamous, man is polygamous”69 pretty much sums it up.

On the same day that 7 astronauts and fragments of the vehicle that failed them plummeted to the fields and woods of east Texas, 6 times that many US MSM became infected. Maybe the number was higher, since it occurred on a weekend; perhaps lower, if news of catastrophe interrupted libidinous pursuits. The calculus of safety investment in Columbia measured the cost of improvements against the risk of equipment failure.70 Lives that hung in the balance fell through the equations. On the basis of CDC estimates of the lifetime expenditure for treating a single case of HIV infection,38 MSM infections acquired that single day will cost $6.5 million. The cost in human potential need not enter the calculus even for a voodoo economist, unless so muddled by moral outrage that he thinks sex between men is indeed something “to die for.”

Theodore M. Brown, PhD, Jeffrey Levi, PhD, and Gloria Weissman provided important recommendations; however, errors of omission or commission are solely those of the author.