Introduction

And when Pasteur Institute workers correctly classified the AIDS pathogen as a lentiviral retrovirus, the only agents desperate physicians could offer even more desperate patients were suramin, DNCB, isoprinosine, and the like (see note 2). Speaking at the Decade of HAART symposium that Fauci keynoted, Martin Delaney (Project Inform, San Francisco) recalled risking imprisonment to smuggle another nostrum, thymic humoral factor, into the States.

For Fauci and research colleagues, the humiliation lay in their inability to offer AIDS patients a true antiretroviral until zidovudine (AZT), a long-abandoned nucleoside, showed that the stampeding retrovirus could indeed be slowed. Of course it's easy to cite reasons for the pharmacologic futility. "We were coming in on the fifth act of a five-act play," Fauci recalled, "and we had no idea what happened in acts one to four."

At the time everyone who understood retroviruses considered AZT a triumph. The drug dashed from its first test in humans to approval in 21 months, still a record, Delaney observed. But few recall that another four years passed before a second antiretroviral, didanosine (ddI), reached the market, Fauci reminded attendees. And nearly everyone saw ddI not as a companion to AZT, but as second-line monotherapy. Indeed, sequential monotherapy -- a strategy of expedience rather than ratiocination -- almost brought antiretroviral therapy to its knees.

Miraculously, fewer than two decades since AZT's debut, US clinicians can now choose from seven nucleoside(-tide) reverse transcriptase inhibitors (NRTIs), two nonnucleoside reverse transcriptase inhibitors (NNRTIs), 10 protease inhibitors (PIs), and one fusion inhibitor (if one excludes zalcitabine [ddC] from the NRTI list and delavirdine [DLV] from the NNRTI roster). By Fauci's reckoning, anti-HIV therapy has advanced from the Dark Ages (1981 to 1986), when nothing slowed HIV, to the Pre-HAART Decade of monotherapy and dual therapy (1987 to 1995), to the Decade of HAART (1996 to 2006), when AIDS and death rates tumbled as fast as T cells once did in people taking suramin.

Despite this virtual cure for infection with a randily replicative virus, the epidemic remains out of control. And not just in Africa, India, or Costa Rica. Every year, Fauci noted, another 40,000 Americans -- and the same number of Western Europeans -- get infected with HIV. While everyone knows how to handcuff HIV in a newly infected person, no one seems to know how to stop that person from passing the virus immediately to others.

The International Association of Physicians in AIDS Care (IAPAC) assembled an all-star cast of world experts to examine why prevention keeps failing and other still unsolved conundrums. At the same symposium, four fearless HIV visionaries cited mathematical models, empiric evidence, and financial spreadsheets to project the future of HIV prevention, antiretroviral resistance, morbidity and mortality, and the cost of care. Finally, top treatment analysts cast a collective eye to the future of HIV research and management.

Three Million Years of Life Saved

Walensky spelled out her findings in print³ and at IAPAC's Decade of HAART symposium. She divided HIV treatment into six eras:

• 1989 to present: Pneumocystis jiroveci pneumonia (PCP) prophylaxis

• 1993 to present: PCP and Mycobacterium avium complex (MAC) prophylaxis

• 1996 to 1997: ART 1 -- potent HAART including PIs

• 1998 to 1999: ART 2 -- sequential NNRTI and PI regimens

• 2000 to 2002: ART 3 -- three effective regimens available including ritonavir-boosted PIs for salvage

• 2003: ART 4 -- improved ART efficacy and tolerability plus enfuvirtide for salvage

The model also included two eras of mother-to-child transmission (MTCT) prevention -- opportunistic infection prophylaxis and AZT alone, and prophylaxis plus combination ART (Table 1).

Patient numbers came from national samples of people with US Centers for Disease Control and Prevention (CDC)-defined AIDS and thus eligible for treatment in each era. The model used to estimate per-person survival factored in CD4 counts, viral load, antiretroviral efficacy, opportunistic infection incidence, treatment, and prophylaxis, and other variables. Walensky stressed that the analysis rests on conservative assumptions that would underestimate survival benefit. For example, limiting the population to people with AIDS excludes many who started treatment earlier in their disease course and benefited as a result.

Compared with people who never got appropriate therapy -- all those who died in the pre-ART "Dark Ages," for example -- people who got PCP prophylaxis gained 3.1 months of life, those who got prophylaxis for both PCP and MAC plus ART 1 gained 93.7 months, and those who got prophylaxis plus ART 4 gained 159.9 months of life (Table 1). Fauci noted that those 160 months exceed the 92 months added by bone marrow transplantation for relapsed non-Hodgkin lymphoma, the extra 50 months afforded by comprehensive care after myocardial infarction, the 29 months gained by adjuvant chemotherapy for node-positive breast cancer, or the seven months appended by chemotherapy for non-small cell lung cancer.

Cumulative benefit for all these interventions toted up to 2,951,371 years of life. If one assumes that 76% of people with AIDS got treated, instead of the 57% reckoned in the primary analysis, years of life saved vaulted by another 710,000. When Walensky relaxed all conservative assumptions of the primary analysis, life years saved hit 5.2 million.

Successful ART saves health care dollars in wealthy countries, at least when figured on a per-annum basis. Fauci cited a University of Alabama study that calculated yearly costs as US$36,532 for a person with fewer than 50 CD4 cells/mm³ versus US$13,885 for a person with at least 350 cells/mm³.⁴ Of course lifetime expenditures for HIV care go up because people survive longer when they get effective ART. (See "Economic history of HIV care," below.)

Delaney noted one small but poignant milestone passed in an early year of the HAART era: On August 13, 1998, the Bay Area Reporter printed its "No obits" issue, the first time in well over a decade that this chronicle of San Francisco's HIV epidemic did not receive an obituary notice.

Eight years later plentiful research confirms that potent ART has a similarly profound impact on survival in countries with the worst HIV epidemics. Indeed, the impact seems even more dramatic than in the United States or Europe because the arrival of antiretrovirals hoisted most people from the no-treatment era straight into Walensky's ART 4 bracket. Brian Gazzard (Chelsea and Westminster Hospital, London) outlined data putting 15-month survival at about 95% in people from Uganda and Zimbabwe enrolled in the DART trial compared with less than 50% in an Entebbe, Uganda cohort of people taking other AIDS remedies.

Acute Shortage of New HIV Docs

Jens Lundgren proposed an international clinician-exchange to improve AIDS care in poor nations while motivating young clinicians to take up HIV medicine when they come home.

But an HIV clinician crisis also looms in Europe and North America, as fewer medical graduates pick HIV as the field they want to cultivate for the next 40 years. Stefan Mauss (Center for HIV and Hepatogastroenterology, Düsseldorf, Germany), who became embroiled in the epidemic as an activist before getting his medical degree, notes that the HIV clinician cohort is aging in pace with the HIV patient cohort. But young people continue to get infected with HIV, while young physicians have stopped getting excited about it. Jens Lundgren (University of Copenhagen) was working in intensive care when HIV medicine hooked him during a six-month stint at the US National Institutes of Health (NIH) in 1984.

Now both Mauss and Lundgren wonder where new HIV physicians -- and nurses -- will come from. Lundgren has worked with the same nurse for a decade. Diane Havlir (University of California, San Francisco) recalled that the collective enduring of despair in saving people with AIDS, followed by fervid HAART-inspired hope, created a tight-knit professional community that carries the load of HIV care to this day. But that about-face of fate will never recur, so those who pursue HIV medicine will need another font of inspiration.

Practical issues -- like income -- now propel many medical graduates into more remunerative fields, noted Huldrych Günthard (University Hospital, Zurich). Motivating young physicians to take on HIV research is tough because heart or prostate research pays so much more. When his generation graduated from medical school, HIV had the febrile allure of a hot, new topic. But no more.

Canada does not recognize HIV as a specialty, noted Marianne Harris (University of British Columbia, Vancouver), so compensation lags the rewards of sanctioned specialties. Most Canadian clinicians are those stamped with an emotional commitment to the disease, often gay men.

Germany led the world in infectious disease research before World War II, Mauss said. But the infectious disease community remained moribund after the war, so when HIV arrived, Germany had no professional contingent primed to manage the disease. Gastroenterologists, like Mauss, oncologists, and others filled the breach. But today the country has no natural wellspring from which to draw HIV clinicians.

In one sense today's much-tested cadre of HIV clinicians had it easy, suggested Graeme Moyle (Chelsea and Westminster Hospital, London): They learned the ins and outs of ART one drug at a time. Today's recruits must master two dozen drugs all at once. The only way they can cope with the resulting complexities, Moyle proposed, is to work for at least a year in a high-volume HIV clinic.

Lundgren suggested pushing that model one step farther, endorsing an international exchange of young clinicians so that today's generation of Western Europeans and others can see HIV infection at its worst in the ravaged countries only now introducing ART. These young physicians could teach native colleagues what they've learned about infectious disease -- and probably learn even more than they teach.

Havlir believes that approach may have merit because she sees plenty of idealism in residents she trains. If that idealism can be focused on underserved populations around the world, a second generation of HIV clinicians may become ready to replace the first.

Praphan Phanuphak (Thai Red Cross Research Center, Bangkok) confirmed that training HIV clinicians in low- and middle-income countries remains a huge obstacle needing immediate attention. Even in a clinically sophisticated country like Thailand, he pointed out, only one of 15 internists in a hospital may be willing to treat people with HIV infection.

Late Treatment, No Treatment and Adherence

Harris reported that only 50% to 70% of eligible HIV-infected British Columbians start free ART. She believes a community-centered approach to antiretroviral uptake -- rather than patient-by-patient ministration -- could expand treatment.

Renato Maserati (University of Pavia, Italy) noted that 130,000 Italians have HIV infection, but only 60,000 take antiretrovirals. He proposed that persisting anti-HIV stigma -- among the public and physicians -- at least partly accounts for this treatment gap. Even committed physicians, Maserati suspects, tend to watch and monitor HIV-infected patients too long. AIDS has dropped from the national consciousness, and one result is the perception that treatment is a less urgent priority. On top of that, slightly more than half of all HIV diagnoses in 2005 came in people who already had AIDS.

Late diagnosis is also a problem in Denmark, where Lundgren reported that 30% to 40% of HIV-infected people get their diagnosis when they have AIDS, even though access to treatment and care is also free there. But Lundgren reminded colleagues that frustration at beginning ART in hard-to-reach Western European populations pales in comparison with frustration in getting antiretrovirals to Eastern Europeans. If he boards a plane and flies 50 minutes eastward, Lundgren said, he can debark in any number of countries with "tremendous out-of-control epidemics."

Once the diagnosis is made, getting people to take their drugs on time may be the greatest enduring problem of HIV care. Maserati thinks the single most important step is convincing patients that you care whether they take their pills. Lundgren believes treatment works best if prescribed by the clinician then reinforced by the nurse who knows the patient best. He recommends a three-way meeting involving patient, clinician, and nurse.

David Cooper (St. Vincent's Hospital, Sydney) maintains that once-daily regimens have made "a huge difference" in easing adherence. He thinks most seasoned HIV clinicians do a decent job in stressing the importance of timely pill taking. An adherence clinic involving the patient, physician, nurse, and pharmacist may marginally improve adherence, Cooper suggested. But even with tolerable once-daily regimens and ongoing support, 100% adherence is not possible in everyone.

Harris stressed that what works to improve adherence in one population, gays, for example, won't necessarily work in others. She said Vancouver clinicians essentially "pay" injecting drug users to take their antiretrovirals in exchange for methadone, food vouchers, or whatever currency makes sense.

Why Prevention Fails in High-Income Countries

Diagnosing people just infected with HIV and not trying to follow up with their sex partners, Rochelle Walensky said, is like "finding a needle in a haystack and throwing it back."

Indeed, Denmark's Lundgren noted, the acute attention belatedly paid to headlong epidemics in Africa, Asia, Eastern Europe, and parts of Latin America helps shape the mirage that AIDS is a problem of "other countries." The HIV prevention message must be brought with new vigor to the general population of developed countries, he argued, and not with abstinence as the primary tenet. Germany's Mauss agreed, adding that "safe sex" campaigns must be realistic. "Unsafe sex is normal sex," he observed. "Safe sex is not normal."

The disappearance of AIDS as a medical priority spawned a rising incidence of sexually transmitted diseases (STDs) in Italy, according to Maserati, and "safe sex" evangelism seems to have a waning impact. HIV incidence numbers indicate that an influx of immigrants does not explain persistently high rates of new infections in Italy. Rather, Maserati suspects young gay men have become a new HIV reservoir.

In Australia, Cooper reported, prevention messages have proved unsustainable among gays -- and he doesn't think anyone has a good handle on how to change that in Australia or elsewhere. Most new infections are showing up not in young men, however, but in 30- or 40-year-old gays. Sexuality is "slightly fluid" in aboriginals, Cooper noted, so men infected by other men may pass the virus on to women. Even though Australia is an island continent, he added, it faces a steady seep of illegal immigrants from a "northern tier of instability" that includes Indonesia, Papua New Guinea, the Solomon Islands, and East Timor.

Cooper pointed to evidence that recently infected people who still have high HIV loads account for a disproportionate quota of new infections.^6,7 Targeting just-infected people to forestall rapid transmission to sex partners is a testable strategy, he argued, because we have the tools to identify newly infected individuals.

Carlos del Rio (Emory University, Atlanta) seconded this point, noting that Christopher Pilcher at the University of North Carolina devised a method to identify infected people in a population soon after they pick up the virus.⁸ Spotting these potential "hyperspreaders" and tracing their social networks, del Rio proposed, could become a public health paradigm to truncate transmission. Finding and educating uninfected people in such social networks may protect many from HIV.

Walensky offered preliminary data suggesting that proactive early diagnosis programs in Massachusetts nearly doubled the diagnosis rate (from 1.3 to 2.3 per month). The estimated cost came to US$4,850 per case identified -- a bargain when one considers the lifetime cost of HIV care. Diagnosing just-infected people and not following up with their sex partners, Walensky said, is like "finding a needle in a haystack and throwing it back."

Harris cited three reasons prevention of new HIV infections faltered in Canada, and those reasons probably apply to other countries with similar demographics:

• A "ceiling effect" on behavioral interventions

• A "huge repository" of undiagnosed HIV-infected people

• Incomplete penetration of ART into groups known to be infected, partly because the epidemic has drifted farther into disenfranchised populations

Mauss reported that, unlike Australia, Germany has seen an upswing in new infections among the youngest gay men. Immigrants, including Africans and Eastern Europeans, often prove hard to reach with prevention advice and care because they lack peer leaders who can facilitate entrée to such groups. Mauss also thinks HIV incidence may be climbing in Germany because of antiretroviral treatment interruptions. Although he supports treatment breaks as a management strategy, Mauss has seen people come back from drug holidays with syphilis and other STDs.

Fauci labeled the 40,000 new infections in the United States every year for the past 15 years "a classic example of an overwhelming failure at prevention." He favors a "harm-reduction" approach that takes high-risk behavior as a given and focuses on cutting the risk of HIV transmission, for example, through pre- and postexposure prophylaxis.

Preventing mother-to-child transmission has become the "best success story" in British HIV prevention, according to Moyle. He attributes that success to a switch to opt-out HIV testing, which resulted in close to 100% testing of pregnant women and virtual eradication of vertical transmission. But the government gives little support to other prevention efforts -- thanks to political cold feet resulting from fumbled prevention efforts of the past. Whitehall has de facto ceded the prevention campaign to nongovernmental organizations.

If high-income countries can't stanch the stream of new HIV infections at home, Cooper wondered, can they really help elsewhere? He suggested that biomedical interventions may be the truest route to successful prevention.

But Richard Feachem, outgoing Executive Director of the Global Fund to Fight AIDS, TB & Malaria (Global Fund), urged colleagues to be wary of placing too much trust in globally feasible "technofixes." Even when an effective vaginal microbicide appears, he argued, most women in the world won't be able to go into town to buy it, won't have the money to buy it, will have no place to hide it from their husbands and mothers-in-law, and would be severely beaten if caught using it -- because husbands would take it as an accusation that they have HIV.

Gazzard thinks microbicides would have the advantage of low cost (compared with some other strategies) and could give autonomy to women (at least to those living in more equable societies). But he wonders how big trials must be to prove a microbicide effective. He suggested it could take 10,000 people per trial for each agent, and then you must figure that some of the same people may be enrolled in HIV vaccine trials, and some male partners will be circumcised while some won't.

Circumcision leapt into AIDS headlines again shortly after the meeting, when the NIH stopped two randomized trials of the procedure in sexually active Ugandan and Kenyan men because the procedure halved their risk of HIV infection.⁹ The results confirmed an earlier randomized trial in South African men.¹⁰

Experts were quick to caution that circumcisions must be done by trained professionals under sterile conditions, that the procedure does not afford complete protection or give men a license to have indiscriminate sex, and that the effect of circumcision on transmission of HIV to women -- or between gay men -- remains unknown. But Feachem said the Global Fund would probably support well-planned proposals for circumcision programs. And few can doubt that these three trials confirm the procedure's great potential in preventing HIV infection.

In Senegal, a largely Muslim country, circumcision is routine. Papa Salif Sow (Fann Hospital, Dakar) believes a nearly 100% circumcision rate, plus a good public education program, explain Senegal's 1.5% HIV prevalence, the lowest in West Africa. He noted, though, that circumcision apparently is not slowing the spread of HIV among men who have sex with men (MSM). The latest estimates put HIV prevalence at an unprepossessing 25% among Senegal's MSM.

Why Prevention Fails (or Works) in Low and Middle-Income Countries

Free antiretrovirals are essential to prevention, argued Celso Ramos-Filho. You can't tell people, "you're going to die of AIDS, but before you die, use condoms."

The biggest problem South Africa faces now in stemming HIV's spread, Wood maintained, is an environment of network sex, transactional sex, and transgenerational sex. He believes that serial monogamy -- rather than network sex -- is what keeps HIV from exploding into heterosexual populations in Western countries. South Africa's main challenge today, Wood proposed, is identifying risk factors in the young and stopping a second epidemic from igniting.

Elly Katabira (Mulaga Hospital, Kampala) thinks no one understands exactly why Uganda sliced HIV prevalence from 30% to 7%, but he argued that abstinence was not the primary reason "because we know that abstinence was impossible." Rather he partly credits President Yoweri Museveni for encouraging an open discussion of AIDS and an education program that targets not only adults but also school students. There is only one major "risk group" in a country with a largely sexually transmitted epidemic, Katabira said, sexually active people. Despite Uganda's early success, recent evidence suggests HIV may be on the rebound.¹¹

Thailand is another country that won good marks for early efforts to stall HIV. But Phanuphak believes the country is not sustaining its much-acclaimed drop in HIV incidence. STDs are on the rise, and 20% of MSM have wives and patronize female sex workers. Married women in Thailand now have a higher risk of HIV infection than single women.

Phanuphak sees two primary reasons for Thailand's wavering defense against HIV: (1) Commitment to fight AIDS waned among government ministers. Thaksin Shinawatra, the prime minister deposed in a 2006 coup, refused to speak on World AIDS Day three years in a row, for example, for fear that talk of HIV and drug users would tarnish Thailand's luster as a tourist destination. (2) Stigma and discrimination persist in Thailand, limiting the number of people who get tested and start antiretroviral therapy.

Mexico's AIDS plight can be partly traced to the vicissitudes of global economics, according to Jorge Saavedra (National AIDS Program, Mexico City). Because Mexico is a member of the Organization for Economic Cooperation and Development (OECD), it is not eligible for support from bodies like the Global Fund. And the country can't make its own generic antiretrovirals because of the North American Free Trade Agreement (NAFTA). Universal access to antiretrovirals began two years ago. As a result most of the national AIDS budget goes to buy antiretrovirals and little is left for prevention.

HIV prevalence stands at only 0.3% in Mexico, half that of the United States. But with 105,000 registered AIDS cases, Mexico ranks second to Brazil in AIDS prevalence. Most HIV-infected Mexicans, 83%, are men, and most of them MSM. Diagnosing HIV infection in gays and getting them to start antiretrovirals remain a steep challenge, Saavedra said, because stigma prevents many from getting tested. Homophobia is rampant in Mexico.

Although the World Bank prediction of a runaway epidemic in South Africa proved true, the World Bank's forecast that Brazil would have one million people with HIV infection missed the mark by 400,000. The country's vaunted antiretroviral program partly explains this success, according to Celso Ramos-Filho (University of Rio de Janeiro). You can't tell people, "you're going to die of AIDS, but before you die, use condoms," Ramos said. To prevent HIV's spread, you must give people a reason to get tested.

Brazil's anti-AIDS program "is not something that grew suddenly after a heavy tropical rain" when the government made ART free, Ramos explained. The country has been training HIV physicians since 1987 and offering treatment for opportunistic infections since 1991. But perhaps the most important ingredient in bridling the epidemic in Brazil, he suggested, is openness in discussing sex. "I've never been to a country where it is so easy to talk about sex, AIDS, or condoms," Ramos maintained. Camisinha, or "little shirt," the common word for condom, "is an absolutely guiltless word -- children use it and no one is shocked."

A Call for a Prevention Cocktail

Hopes for an effective HIV vaccine remain long range, at best, Grant said. Speaking in 2004, vaccine expert Ronald Desrosiers likened immediate prospects for hitting on a worthy HIV vaccine to hitting a baseball blindfolded. Although Desrosiers' hometown baseball team, the Boston Red Sox, seemingly defied similar odds by winning the World Series that year, few expect the same fortuity in the HIV vaccine arena.

Several trials of PEP (event-driven postexposure prophylaxis) and PREP (regular dosing for pre-exposure prophylaxis) are under way, including studies of tenofovir (TDF) in 900 high-risk African women, 400 US MSM, and 1,600 Thai injecting drug users. Trials of Truvada, the fixed-dose pill mingling TDF and emtricitabine (FTC), are about to start in 1,200 heterosexual men and women in Botswana and 1,400 MSM in Andean countries.

Preliminary results of a placebo-controlled PREP trial of TDF at a dose of 300 mg daily showed two seroconversions in the TDF arm and six in the placebo arm, but this early analysis did not have the power to demonstrate the efficacy of TDF PREP (P=0.24).¹³ Trial participants did report fewer sexual partners and more condom use during the study, and most side effects were not drug related in this ongoing study of 936 high-risk people in Cameroon, Ghana, and Nigeria.

Grant cautioned that antiretroviral PREP is costly; the drugs, their delivery, and monitoring all have a price tag. But a modeling study he presented at the XVI International AIDS Conference earlier this year determined that PREP would be more cost effective than treatment.¹⁴

A PEP-in-the-pocket plan did work in a study of 200 MSM in Rio de Janeiro given a 4-day supply of AZT and lamivudine (3TC) and told how to use it after likely exposure to HIV.¹⁵ After a median two years of follow-up, researchers logged 10 seroconversions in men who, for one reason or another, did not start PEP and only one in men who unpocketed their PEP.

Although some argue that PEP or PREP may encourage people to have more sex, Grant theorized that the strategies could foster a safer sex life by (1) serving as a daily reminder of risk, (2) promoting a vision of an AIDS-free life, and (3) stabilizing sexual relations. At the same time PEP or PREP programs could help destigmatize people at risk of HIV infection. Even small surges in risk behavior could offset the protective effect of prevention tactics if those tactics have only a modest impact.¹⁴ But Grant believes a highly effective intervention such as circumcision will not be greatly affected by small spurts in risk behavior.

Grant reminded colleagues of a seemingly obvious but often unstated dictum of HIV prevention -- effective strategies work only if applied. And so far behavioral and technologic advances in HIV have not seen wide use. Hopes for new evidence of feasible interventions must be seen in that context, Grant cautioned. Ultimately, just as with ART, he believes a cocktail of effective prophylactic components may prove the best way to stymie HIV.

Forecasting HIV Incidence and Mortality in the Developing World

Distrust of modeling persists among some members of the medical community, Blower conceded. Yet experts and laypeople alike embrace modeling predictions of weather and the economy. And when predicting clinical trends, she maintained, the only alternative to modeling is guessing.

The model Blower and colleagues devised to foretell the epidemiologic impact of antiretroviral allocation in this South African province focused on HIV-susceptible individuals living in Durban or one of five rural communities who become infected with drug-resistant HIV or a mixture of susceptible and resistant virus and who do or do not begin ART (Figure 2). The formulas Blower fashioned aimed to predict results of three antiretroviral allocation strategies on (1) HIV incidence, (2) transmission of resistant virus, and (3) AIDS-related deaths:

• Strategy 1. Allocation to both urban and rural clinics based on population size

• Strategy 2. Allocation to both urban and rural clinics based on HIV prevalence within each community weighted by distance from the urban center

• Strategy 3. Allocation only to urban clinics

The model calculated the impact of treating 500,000 people through 2008. Right now 56% of KwaZulu-Natal's 9.4 million people live in rural areas, where overall HIV prevalence stands at 9%.

Both urban-rural strategies (1 and 2) would lower the new infection rate 21%, whereas sending all the drugs to Durban (strategy 3) would lower incidence 25%. In raw numbers treating HIV-infected people only in Durban would avert 15,000 more new infections than treating people in Durban and the surrounding countryside, including 5,400 fewer new infections with antiretroviral-resistant virus. Limiting ART to Durban would prevent 1,400 more AIDS-related deaths than distributing drugs both to Durban and rural clinics.

No one, Blower trusts, would ban antiretrovirals from rural communities because a mathematical model says treating only city dwellers would do more to curtail the epidemic -- or for any other reason. But the model does underscore "the difficult choices between epidemiological and ethical objectives that will have to be made by the governments and health policy officials of resource-constrained countries to decide how to divide the scarce drugs between the urban and rural areas in their countries."¹⁶

Indeed, Blower maintains, policy makers can use modeling to help decide how antiretrovirals may be divided among available clinics to approach treatment equity. Another study she recently published had that goal in mind.¹⁷ In this work Blower used another spatial mathematical model that accounted for heterogeneity in HIV prevalence and access to treatment at various clinics in KwaZulu-Natal. Specifically, the model rated equitable antiretroviral distribution if the province has 17, 27, or 54 facilities dispensing drugs, and if the catchment area for each of those centers has a 20-, 40-, or 60-km radius.

Modeling results showed that the fairest antiretroviral distribution plan in KwaZulu-Natal would treat HIV-infected people at 54 clinics serving a catchment of 40 to 60 km. Blower proposed three conclusions:

1. Complicated drug-allocation strategies are necessary if we wish to achieve treatment equity.

2. Using a fixed amount of antiretrovirals but different drug-allocation strategies will have substantially different effects on HIV incidence, transmitted resistance, and AIDS mortality.

3. We need to know the specific drug-allocation strategy that will be used to make accurate predictions.

Trends and Predictions in HIV Drug Resistance

Work in the UK CHIC cohort confirms that the risk that at least one resistance mutation will emerge parallels virologic failure rates over six years of follow-up.¹⁸ In this cohort one in 10 people had one or more resistance mutations after two years of follow-up, 20% had at least one mutation after four years, and 27% had at least one mutation after six years. Over the same years failure rates climbed from about 20% to 40%.

Whether a cohort member began a regimen based on abacavir (ABC), an unboosted PI, a boosted PI, or an NNRTI correlated directly with the risk of double- or triple-class resistance upon failure (Table 2). Compared with NNRTI therapy, boosted PI therapy more than halved the risk of multiclass resistance, while unboosted PI and ABC regimens moderately raised that risk. In this cohort estimated probability of survival with multidrug-resistant (MDR) virus measured 3% one year after diagnosis of MDR virus, 8% after two years, and 13% after three years.

In the EuroSIDA cohort prevalence of triple-class failure rose inexorably from 1998 to 2003 in people beginning HAART with NRTI experience.¹⁹ In January 2003, the rate of triple-class failure topped 15% in 2,028 cohort members with pre-HAART experience. But among 58 people who began HAART as their first antiretroviral regimen, the triple-class failure rate remained below 5% through January 2003.

Does infection with resistant virus affect initial response to HAART? Yes, at least in North Americans starting treatment before 2003.²⁰ In this landmark resistance transmission study, Susan Little (University of California, San Diego) and colleagues determined that people who began therapy with a 50% inhibitory concentration ratio below 2.5 had a significantly slower time to viral suppression below 500 copies/mL than did people starting treatment with a ratio above 10 through 200 days of followup. But transmitted drug resistance appears not to influence the natural history of HIV infection over a longer term, Pillay himself found in a CASCADE collaboration study.²¹ CD4 counts dropped faster in the first year of infection among people with drug-resistant virus than among those with drug-susceptible HIV. But five years after seroconversion, CD4 trends in the two groups virtually overlapped.

In nearly 1,000 untreated Europeans, people infected while injecting drugs or during heterosexual sex ran a similar risk of picking up resistant virus as MSM.²² People originally infected in high HIV prevalence countries had about half the risk of infection with resistant virus as MSM, probably because antiretroviral use remains so limited in these countries.

Once transmitted, Pillay noted, antiretroviral-resistant virus may persist for years. Two recent studies charted persistence of thymidine analog mutations (TAMs), the K103N nonnucleoside mutation, PI mutations, and MDR virus in nine of 11 US residents after a median follow-up of nine months²³ and in 14 of 16 UK residents with follow-up extending to three years.²⁴ Transmitted TAMs at reverse transcriptase positions 41 and 215 proved the longest lived in a study of 440 CASCADE cohort members, surviving 30 and 40 months of follow-up, respectively, after the first resistance test.²⁵ The 3TC- and FTC-induced M184V mutation, on the other hand, reverts rapidly to wild type in the absence of drug pressure.²⁶

Overall "drug resistance burden" may be dropping in developed countries, according to work by Deenan Pillay. And that decline may lead to stable resistance transmission rates.

The researchers divided each person's follow-up into one-month periods and identified the viral load nearest to the start of each month. When viral load exceeded 1,000 copies/mL, monthly "viral load burden" equaled viral load times numbers of days in the month. When viral load lay below 1,000 copies/mL, monthly "viral load burden" equaled zero and transmission risk was assumed to be zero.

Pillay and colleagues stratified "resistance burden" according to the presence or absence of six key mutations, M41L, T69D/L, T215(any), M184V, and K103N in reverse transcriptase and L90M in protease. They classified each patient-month by the presence of each mutation and assumed the patient had no mutations if none of the key mutations were detectable in that month.

The primary analysis focused on the presence or absence of each key mutation. In the first of two sensitivity analyses, the researchers assumed that detected mutations persisted even if they were not detected in later resistance tests. In the second sensitivity analysis, they assumed that, if mutations become undetectable six months after being spotted, the virus remained wild type until the next genotype.

In the primary analysis the proportion of resistance burden contributed by antiretroviral-naive people rose from 17% in 2000 to 41% in 2001 but then receded to 26% in 2002 and 34% in 2003. If primary resistance mutations persist and acquired mutations do not, however, the proportion of resistance burden contributed by treatment-naive people climbed from 20% in 2000 to 54% in 2001, and to 60% in 2002 and 2003. Mutation prevalence differed greatly in genotypes of people with chronic infection (M184V > T215 (any) > K103N > M41L > T69D/L = L90M) versus people with primary infection (K103N = T215 (any) > M41L > L90M; no M184V or T69D/L).

Pillay proposed that declining "resistance burden" after 2001 in this cohort probably explains at least some of the recently charted drop in British transmission of resistant virus.

Research in Uganda's Rakai cohort established the highest risk of sexual HIV transmission during the first few months of infection.²⁸ Calculated rates of transmission per coital act measured 0.0082 during the first five months of HIV infection, compared with 0.0015 in months six to 15 and 0.0010 in months 16 to 35. Transmission rates also climbed marginally in the last months before death, when viral loads flare again. But those late transmission rates never returned to the 0.0082 rate seen just after infection.

For the resource-rich world, Pillay proposed five conclusions on resistance trends:

• Overall drug resistance burden may be declining in developed countries.

• That decline may lead to stable resistance transmission rates.

• Antiretroviral-naive populations may become a growing source of transmitted resistance.

• Migration must be considered when trying to project resistance trends.

• Prevention remains a priority in resource-rich countries.

Economic History of HIV Care

• Chaos and cost: start of epidemic to Ryan White CARE Act

• Integration under subsidy: from Ryan White CARE Act to HAART

• Dominance of chemotherapy: the HAART era

In the first era most HIV-infected people sought care when they already had an opportunistic disease. People who did not die soon after diagnosis found themselves in the hospital an average three times yearly. Severe neurologic disease was common, and many people needed medical and nonmedical home care. Per-patient costs were lofty in this era, hitting an estimated US$147,000 per patient over 13 months around 1985. That estimate dropped by 1990 but still stood at a prodigious US$80,000 yearly. Estimated costs for 1985 totaled US$630 million. Although per-patient bills fell over the years, higher HIV prevalence in later years meant total costs for 1991 came to US$8.5 billion.

Politicoeconomic hallmarks of this bleak era included lack of financial commitment and deficient physical and human infrastructure. Underinvestment in AIDS care was the rule rather than the exception, and health care coverage for HIV-infected people varied greatly from region to region. Infrastructure deficits ran the gamut from clinics, hospital beds, and laboratories to providers and drug delivery.

The Ryan White CARE Act became law in the United States in 1990 "to provide emergency assistance" for people with HIV infection. Total federal spending on HIV and AIDS climbed to US$5.1 billion in 1995, including US$2.5 billion in Medicare and Medicaid. All care expenditures totaled about US$7.2 billion. The clinical situation improved markedly with the introduction of antiretrovirals, and with improved prophylaxis, treatment, and maintenance therapy for opportunists. This era also saw an increasing impact of greater clinical expertise on outcomes.

A small band of clinicians and researchers focused on the disease because of its exotic, at first poorly understood natural history and the resulting fear. Even though care remained underfunded, a comprehensive care model evolved to incorporate:

• Inpatient-outpatient continuity

• Access to specialists

• Non-physician medical providers

• Case managers and other medical providers

Chemotherapy dominated the clinical landscape with the HAART epiphany. Highly effective therapies diffused rapidly, but pharmacy costs stayed high. The intricacies of antiretroviral therapy saw a drive toward specialization in the clinical arena. Per patient per month costs fell because of cost offsets and because lifetime costs (along with life expectancy) rose. Research pegged lifetime costs at US$13,000 to US$23,000 per quality adjusted life year. Graphing HIV care expenditures from the onset of the HAART era through January 1999 showed overall drops in inpatient, outpatient, and drug costs, though drug costs displaced inpatient costs to claim the greatest part of the budget.

In the US Veterans Administration (VA), use of NRTIs and PIs soared after 1995 (Figure 3). Prescription of NNRTIs climbed more gradually after their introduction around 1996. But as PI use declined after 1998 with the discovery of longer-term antiretroviral side effects, NNRTI uptake continued to rise and pulled even with PI prescriptions around 2003. The advent of HAART in 1995 also saw precipitous plummets in all-cause mortality and inpatient stays for cardiovascular or cerebrovascular disease in the VA. Rates for both all-cause mortality and cardio- or cerebrovascular disease stood above 20 events per 100 patient-years in 1995 and approached five events per 100 patient-years in 2003.

US federal funding for HIV and AIDS -- through the Ryan White CARE Act, Medicare, Medicaid, and other programs -- expanded throughout the HAART era, from US$3.7 billion in 1995, to US$7.4 billion in 2000, and to US$11.0 billion in 2004. Over that period Ryan White CARE Act funding nearly quadrupled from US$0.6 billion in 1995 to US$2.0 billion in 2004. In fiscal year 2004 the lion's share of HIV/AIDS funding, 59%, went to care, while 16% supported research, 10% international programs, 9% cash and housing assistance, and 5% prevention programs.

Annual HIV/AIDS care cost estimates presented in 2006 show marked differences depending on the method used and the country concerned. Ray Chen (University of Alabama at Birmingham) and colleagues figured an annual cost of US$13,885 for people with a CD4 count above 350 cells/mm³ versus US$36,553 for those with fewer than 50 cells/mm³ in 2001.⁴ Nonantiretroviral medication costs drove overall costs for people with advanced disease. Kelly Gebo (Johns Hopkins University, Baltimore) and colleagues figured a US$21,869 annual cost for people with a CD4 count above 500 cells/mm³ in 2003, compared with US$57,565 annually for those with fewer than 50 cells/mm³.²⁹ In contrast, a 2006 modeling study involving Côte d'Ivoire calculated an overall US$500 per patient annual cost when antiretrovirals cost US$300 yearly.³⁰

Bozzette predicted that outlays for HIV care will rise faster in the United States and that drug prices will increasingly dominate costs. Numerous economic issues involving HIV care remain unexplored, including indirect costs, returns to primary and secondary prevention, and returns to comprehensive quality improvement, continuity, and recruitment efforts.

Future of HIV Research and Management

NNRTI regimens largely replaced PI combinations in first-line planning, Gazzard believes, not only because of their generally more genial side effect profile, but also because they yield good viral load suppression at moderate levels of adherence.^32,33 Such findings and the availability of new antiretrovirals lead Gazzard to believe that a new treatment paradigm is emerging (Figure 4).

First-line regimens will rely on a fixed-dose nucleoside couplet (TDF/FTC or ABC/3TC) plus an NNRTI. Already that option may mean the one-pill once-daily melding of TDF, FTC, and efavirenz (EFV). Depending on resistance patterns that emerge with first-line failure, second-line therapy would involve a boosted PI such as darunavir (DRV) or possibly an integrase inhibitor plus two other untried drugs.

Experimental integrase inhibitors have already impressed several Decade of HAART symposium speakers, including Gazzard, who labeled research results so far "extremely impressive." Although he surmised that integrase blockers may find routine use as early as the second round of antiretroviral sequencing (Figure 4), these agents clearly stand to play a critical role in salvage therapy. They may make sub-50-copy viral loads a realistic target for most people with virus resistant to the first three antiretroviral classes.

Anthony Fauci suggested that, by preventing HIV integration into host cell genes, integrase inhibitors could greatly limit this untouchable reservoir of latent virus.

Like most antiretroviral mavens, Decade of HAART symposium delegates take a wait-and-see attitude on how CCR5 antagonists will fit into the treatment scheme. Gazzard sees three hurdles in continuing development and potential use of this class. First, how readily CCR5-tropic virus will shift allegiance to the CXCR4 coreceptor during treatment remains unknown, as does how -- or whether -- such a shift will toggle progression of HIV infection. Second, tracking coreceptor shifts remains complicated because of coreceptor assay limitations. Finally, toxic surprises marred smooth development of this class. GlaxoSmithKline abandoned work on aplaviroc because of liver toxicity, and new cancers that emerged during a vicriviroc trial -- though not certainly ascribed to the drug -- "make you sit back and think," Gazzard allowed. He suggested side effect findings may point to CCR5 antagonist interference with immune surveillance.

Citing research on genetic clues to abacavir hypersensitivity,^34-36 Gazzard speculated that Simon Mallal (Royal Perth Hospital, Australia) and colleagues may have hit on a general mechanism of toxicity with this work. At the very least, he proposed, pharmacogenetics are here to stay in antiretroviral therapy.

Despite clamorous advances in anti-HIV therapy, a daunting list of unknowns will challenge researchers for the next decade of HAART, and probably well beyond that, Gazzard and others cautioned. And new antiretroviral research won't be easy to launch, warned Moyle, because HIV-infected people display a growing reluctance to sign up for clinical trials. In the first decades of ART, trials often offered patients the only avenue to potent experimental therapies. With so many strong and tolerable antiretrovirals on the market today, that enticement disappeared.

Raising money for clinical trials keeps getting tougher in Britain, according to Moyle, and the same applies to Spain, said Gatell. Not long after the Decade of HAART symposium wound up, US officials announced a cutback in AIDS Clinical Trials Group funding, putting some sites out of business unless they find emergency cash.

And with HIV-infected volunteers and government funding growing scarcer, it still takes big trials to answer the biggest strategic questions, Lundgren observed. Only large populations ensure the statistical power needed to unveil subtle differences between regimens or treatment schemes. Although Lundgren did not mention specifics, one could point to the large and decisive SMART trial of treatment interruptions,³⁷ which may settle several disputes about this controversial tactic that smaller studies could not.

Returning to the theme of humility raised by Fauci in his keynote talk, Havlir ruefully reminded colleagues that -- for all we have learned about HIV infection in the decade of HAART -- physicians still don't even know the best time to start therapy. Benchmark three-drug regimens may not be the best option for every patient, she proposed. Research shows that triple therapy works best on the population level, but it may not work best for every member of that population. And to this day no one knows what "remission" of HIV infection means and whether HAART can bring it about.

Mark Mascolini writes about HIV infection (markmascolini@earthlink.net).

References and Notes

1. Rozenbaum W, Dormont D, Spire B, et al. Antimoniotungstate (HPA 23) treatment of three patients with AIDS and one with prodrome. Lancet 1985;23;1:450-451.

2. Anthony Fauci's list of frenziedly proffered AIDS remedies includes suramin, AL 721, HPA-23, ribavirin, isoprinosine, thymosin, alpha interferon, interferon gamma, interleukin 2, thymic transplantation, lymphocyte transfer, bone marrow transplantation, and DNCB.

3. Walensky RP, Paltiel AD, Losina E, et al. The survival benefits of AIDS treatment in the United States. J Infect Dis 2006;194:11-19.

4. Chen RY, Accortt NA, Westfall AO, et al. Distribution of health care expenditures for HIV-infected patients. Clin Infect Dis 2006;42:1003-1010.

5. UNAIDS. Report on the global HIV epidemic 2006. May 2006.

6. Pilcher CD, Tien HC, Eron JJ Jr, et al. Brief but efficient: acute HIV infection and the sexual transmission of HIV. J Infect Dis 2004;189:1785-1792.

7. Pilcher CD, Chilongozi D, Martinson F, et al. Comparison of the concentration of HIV in semen during acute, early and late HIV infection: evaluation of the risk of the sexual transmission of HIV. 1st International Workshop on HIV Transmission. August 11-12, 2006. Toronto. Abstract 3.

8. Pilcher CD, Fiscus SA, Nguyen TQ, et al. Detection of acute infections during HIV testing in North Carolina. N Engl J Med. 2005;352:1873-1883.

9. National Institutes of Health. Adult male circumcision significantly reduces risk of acquiring HIV: trials in Kenya and Uganda stopped early. December 13, 2006.

10. Auvert B, Taljaard D, Lagarde E, et al. Randomized, controlled intervention trial of male circumcision for reduction of HIV infection risk: the ANRS 1265 trial. PLoS Med 2005;2(11):e298. Epub 2005 Oct 25. Erratum in: PLoS Med 2006;3(5):e298.

11. Shafer LA, Biraro S, Kamali A, et al. HIV prevalence and incidence are no longer falling in Uganda -- a case for renewed prevention efforts: evidence from a rural population cohort 1989-2005, and from ANC surveillance. XVI International AIDS Conference. August 13-18, 2006. Toronto. Abstract THLB0108.

12. Koblin B, Chesney M, Coates T. Effects of a behavioural intervention to reduce acquisition of HIV infection among men who have sex with men: the EXPLORE randomised controlled study. Lancet 2004;364:41-50.

13. Peterson L, Taylor D, Clarke EEK, et al. Findings from a double-blind, randomized, placebo-controlled trial of tenofovir disoproxil fumarate (TDF) for prevention of HIV infection in women. XVI International AIDS Conference. August 13-18, 2006. Toronto. Abstract THLB0103.

14. Grant R, Lama J, Goicochea P, et al. Cost-effectiveness analysis of HIV chemoprophylaxis. XVI International AIDS Conference. August 13-18, 2006. Toronto. Abstract THLB0102.

15. Schechter M, do Lago RF, Mendelsohn AB, et al. Behavioral impact, acceptability, and HIV incidence among homosexual men with access to postexposure chemoprophylaxis for HIV. JAIDS 2004;35:519-525.

16. Wilson DP, Kahn J, Blower SM. Predicting the epidemiological impact of antiretroviral allocation strategies in KwaZulu-Natal: the effect of the urban-rural divide. Proc Natl Acad Sci USA 2006;103:14228-14233.

17. Wilson DP, Blower SM. Designing equitable antiretroviral allocation strategies in resource-constrained countries. PLoS Med 2002; 2(2):e50.

18. The UK Collaborative Group on HIV Drug Resistance, UK CHIC Study Group. Long term probability of detection of HIV-1 drug resistance after starting antiretroviral therapy in routine clinical practice. AIDS 2005;19:487-494.

19. Mocroft A, Ledergerber B, Viard JP, et al. Time to virological failure of 3 classes of antiretrovirals after initiation of highly active antiretroviral therapy: results from the EuroSIDA study group. J Infect Dis 2004; 190:1947-1956.

20. Little SJ, Holte S, Routy JP, et al. Antiretroviral-drug resistance among patients recently infected with HIV. N Engl J Med 2003;2002; 347:385-394.

21. CASCADE Virology Collaboration. The impact of transmitted drug resistance on the natural history of HIV infection and response to first-line therapy. AIDS 2006;20:21-28.

22. Wensing AMJ, Vercauteren J, van de Vijver DA, et al. Transmission of drug-resistance in Europe is characterized by single mutations and revertants. Antivir Ther 2006;11:S111.

23. Little SJ, Koelsch KK, Ignacioet CC, al. Persistence of transmitted drug-resistant virus among subjects with primary HIV infection deferring antiretroviral therapy. 11th Conference on Retroviruses and Opportunistic Infections. February 8-11, 2004. San Francisco. Abstract 36LB.

24. Pao D, Andrady U, Clarke J, et al. Long-term persistence of primary genotypic resistance after HIV-1 seroconversion. JAIDS 2004;37: 1570-1573.

25. Bezemer D, de Ronde A, Prins M, et al. Evolution of transmitted HIV-1 with drug-resistance mutations in the absence of therapy: effects on CD4+ T-cell count and HIV-1 RNA load. Antivir Ther 2006;11:173-178.

26. Metzner KJ, Rauch P, Walter H, et al. Detection of minor populations of drug-resistant HIV-1 in acute seroconverters. AIDS 2005;19:1819-1825.

27. Fisher M, Sabin C, Pao D, et al. What is the drug resistance mutational infectious burden in an HIV-1 prevalent cohort and the relationship to incidence of transmitted resistance? 1st International Workshop on HIV Transmission. August 11-12, 2006. Toronto. Abstract 48.

28. Wawer MJ, Gray RH, Sewankambo NK, et al. Rates of HIV-1 transmission per coital act, by stage of HIV-1 infection, in Rakai, Uganda. J Infect Dis 2005;19:1403-1409.

29. Gebo K, J Fleishman J, Conviser R, et al. Contemporary costs of HIV health care in the HAART era. 13th Conference on Retroviruses and Opportunistic Infections. February 5-8, 2006. Denver. Abstract 537.

30. Goldie SJ, Yazdanpanah Y, Losina A, et al. Cost-effectiveness of HIV treatment in resource-poor settings -- the case of Cote d'Ivoire. N Engl J Med 2006;355:1141-1153.

31. Bartlett JA, Fath MJ, Demasi R, et al. An updated systematic overview of triple combination therapy in antiretroviral-naive HIV-infected adults. AIDS 2006;20:2051-2064.

32. Bangsberg DR. Less than 95% adherence to nonnucleoside reverse transcriptase inhibitor therapy can lead to viral suppression. Clin Infect Dis 2006;43:939-941.

33. Weiser SD, Guzman D, Riley ED. Higher rates of viral suppression with nonnucleoside reverse transcriptase inhibitors compared to single protease inhibitors are not explained by better adherence. HIV Clin Trials. 2004;5:278-287.

34. Mallal S, Nolan D, Witt C, et al. Association between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002;359:727-732.

35. Hetherington S, Hughes AR, Mostellar M, et al. Genetic variations in HLA-B region and hypersensitivity reactions to abacavir. Lancet 2002;359:1121-1122.

36. Rauch A, Nolan D, Martin A, et al. Prospective genetic screening decreases the incidence of abacavir hypersensitivity reactions in the Western Australian HIV cohort study. Clin Infect Dis 2006;43:99-102.

37. Strategies for Management of Antiretroviral Therapy (SMART) Study Group. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med 2006;355:2283-2296.