1. Home>
2. Specialty Care>
3. HIV/AIDS>
4. Meeting Report

Report from the 13th Retrovirus Conference

Many results presented at the 2006 Retrovirus Conference could have an immediate and direct effect on HIV clinical care.

This year’s Retrovirus Conference was held in downtown Denver at the sparkling new Colorado Convention Center. The sheer size of the site was a far cry from the original meetings 13 years ago, which took place in a much more intimate Washington, DC, hotel, but the increased scale of the meeting has only solidified it as the leading venue for presenting HIV-related research to both clinical and basic-science audiences.

As always, the focus of our meeting report is on new study results that could have an immediate and direct effect on HIV clinical care. All abstracts, as well as webcasts of the symposia, are available at the meeting website: www.retroconference.org. — Paul E. Sax, MD

Integrase Inhibitors on the Way

Both Merck and Gilead have investigational integrase inhibitors in clinical development. With the Merck drug, MK-0518, about to enter phase III testing, researchers presented interim 16-week results from an ongoing phase II study [Abstract 159LB]. A total of 167 triple-class–experienced patients were randomized to receive an optimized background regimen plus MK-0518 (200 mg, 400 mg, or 600 mg orally, twice daily) or placebo. At enrollment, the median duration of prior antiretroviral experience was about 10 years, mean viral loads ranged from 4.6 log copies/mL to 4.8 log copies/mL, and mean CD4 counts were between 220 cells/mm³ and 283 cells/mm³. At week 16, all three doses of MK-0518 were superior to placebo; 56% to 72% of MK-0518 recipients achieved viral loads <50 copies/mL, compared with 19% of placebo recipients. Overall, regimens containing MK-0518 were well tolerated, and adverse events were comparable to those observed with placebo-containing regimens. The most common side effects in the study (occurring in at least 5% or a minimum of 2 patients in any treatment group) were diarrhea, nausea, fatigue, injection-site reaction, headache, and itching. The upcoming phase III trials (BENCHMRK-1 and -2) are designed to evaluate MK-0518 in both treatment-experienced and treatment-naive patients.

Initial data were also presented on the new Gilead integrase inhibitor, GS-9137 [Abstract 160LB]. These data came from a phase I/II dose-escalation study conducted in 40 HIV-infected patients, 15 of whom were treatment-naive. Subjects were randomized to receive GS-9137 (200 mg, 400 mg, or 800 mg twice daily; 800 mg once daily; or 50 mg boosted with 100 mg of ritonavir once daily) or placebo with food for 10 days. Mean baseline viral load was 4.75 log copies/mL, and mean CD4 count was 442 cells/mm³. At all doses, GS-9137 monotherapy demonstrated significant antiviral activity compared with placebo; the greatest reduction (–2.03 log copies/mL) was seen with the ritonavir-boosted dose. Overall, 25 of 30 patients in the GS-9137 arms had a viral load-decline 1.0 log copies/mL; no placebo recipient had such a decline. Adverse events were mostly mild (none above grade 2); only fatigue was more common with GS-9137 than with placebo. A longer phase II study is being developed with more patients.

Both MK-0518 and GS-9137 appear to be highly promising for the treatment of HIV infection. The Merck drug will likely be developed as a twice-daily agent given without ritonavir (because MK-0518 is metabolized through glucuronidation, and ritonavir does not affect its pharmacokinetics), whereas the Gilead drug will probably be a once-daily agent requiring ritonavir-boosting. We eagerly await further studies on both agents. — Charles B. Hicks, MD

New Data on TMC114 and TMC125

Patients with triple-class antiretroviral experience have had promising treatment responses to the investigational PI darunavir (TMC114). However, data have been sparse on resistance and correlates of virologic response. To address this, researchers evaluated pooled data from the previously described POWER-1 and -2 studies (ACC Feb 1 2006), as well as additional data from the as-yet-unpresented POWER-3 study [Abstract 157]. Diminished virologic response to darunavir was seen in patients with at least 10 primary PI mutations. The mutations that particularly affected response in these patients included V32I, L33F, I47V, I54L, and L89V. Notably, viruses that were susceptible to tipranavir at baseline remained so, even after darunavir failure.

One major issue with the clinical use of darunavir is the availability of additional active antiretrovirals to pair with it. To evaluate whether potential drug-drug interactions preclude using TMC125 (a next-generation NNRTI) as a second active agent, investigators assigned 11 patients with triple-class resistant infection to receive darunavir + TMC125 + an optimized background regimen [Abstract 575c]. Although TMC125 levels were about 30% lower when the two drugs were given together, the concurrent use of TMC125 did not affect darunavir levels. The combination was associated with a median viral-load reduction of 2.76 log copies/mL, and all patients experienced a decline of at least 2 log copies/mL.

Additional information about TMC125 came from the C223 study, which assessed the influence of baseline resistance on virologic response [Abstract 154]. In contrast to the "all or none" pattern of resistance seen with currently licensed NNRTIs, TMC125 exhibited gradations of virologic response. The greatest reductions were observed among patients with no baseline NNRTI mutations (–1.82 log copies/mL), followed by those with one mutation (–1.65 log copies/mL), two mutations (–1.00 log copies/mL), or at least three mutations (–0.66 log copies/mL). These data suggest that TMC125 might be most effective when used before multiple NNRTI mutations accumulate.

Finally, investigators studied the effects of adding TMC125 to a regimen containing lopinavir/ritonavir, saquinavir, and at least two NRTIs [Abstract 575b]. TMC125 was associated with significantly reduced lopinavir AUC_12h and C_max but did not appear to affect lopinavir C_min. No significant change was seen for saquinavir or ritonavir, and the use of these PIs did not affect TMC125 concentrations. — Charles B. Hicks, MD

Vicriviroc Less Effective than Efavirenz in Treatment-Naive Patients

In the fall of 2005, Schering-Plough terminated its phase II comparison of vicriviroc and efavirenz in treatment-naive patients because the vicriviroc arm had a higher rate of virologic failure (ACC Nov 8 2005). Now, the company has provided additional details on the study, which included 92 treatment-naive subjects with CCR5-tropic virus and baseline CD4 counts >150 cells/mm³ [Abstract 161LB]. Patients were randomized to receive vicriviroc (25 mg, 50 mg, or 75 mg once daily) or placebo for 14 days. At day 15, AZT/3TC was added to all regimens, and the placebo was replaced with efavirenz. At that time, the vicriviroc-treated patients showed an approximate 1-log copies/mL decline in viral load, consistent with earlier phase I studies. However, by study closure, a much higher proportion of patients in the vicriviroc arm (26 of 68) than in the efavirenz arm (1 of 24) had experienced virologic failure, especially at the two lower doses of vicriviroc. Furthermore, all 22 patients with virologic failure and obtainable genotypes in the vicriviroc arm showed the M184V mutation. The reason for this disappointing result is not yet clear; no consistent pattern of mutations was seen in envelope sequences, and tropism switches — seen in three placebo recipients and five vicriviroc recipients — did not explain the virologic rebounds. Overall, these results raise important questions about the use of CCR5 antagonists in treatment-naive subjects — a concern further enhanced by the closing of the once-daily treatment arm of a maraviroc study in a similar patient population. Notably, the twice-daily arm of that study is ongoing, as are vicriviroc and maraviroc studies in treatment-experienced patients.

Further information on vicriviroc was released on March 3, 2006, detailing the occurrence of five cases of malignancy (4 lymphoma, 1 adenocarcinoma of the stomach) in ACTG 5211. This study involves 118 treatment-experienced patients, randomized to receive one of three doses of vicriviroc or placebo in addition to an optimized background regimen. All cases occurred in the vicriviroc-treated group. The study has been unblinded, and further investigation of the cases is ongoing. — Paul E. Sax, MD

Treatment Interruption, Interrupted

Six studies presented at the conference advanced our knowledge of treatment interruption (TI) for the management of chronic HIV infection in patients who have good viral control and preserved CD4-cell counts while on antiretroviral therapy. As shown in Table 1, the studies fall into three categories: natural history of TI, TI based on fixed intervals, and TI driven by CD4-cell count.

The natural-history study, ACTG 5170, demonstrated that TI was generally safe in terms of HIV disease progression among patients with a nadir CD4 count >350 cells/mm³; it also showed that patients could remain off therapy for an extended duration (median interruption, 96 weeks) [Abstract 101]. Although the CD4-cell count fell rapidly during the first 8 weeks of interruption (20 cells/mm³/week), the rate of decline subsequently slowed to 2 cells/mm³/week. There were five deaths in the study, one case of acute retroviral syndrome, and four cases of severe thrombocytopenia. Nadir CD4-cell count was the best predictor of CD4-cell decline or adverse events during TI.

In the two studies of fixed-interval TI, the TI and continuous-therapy groups had similar results [Abstracts 103 and 104]. Elevated rates of resistance were noted in the TI arm of the ISS PART study (which suggests a need for caution when using NNRTI-based regimens), and 5% of TI patients had a low platelet count in ANRS 106.

Results from the three studies of CD4-driven TI are particularly illuminating: They strongly suggest that the CD4 count should not be allowed to fall below 350 cells/mm³ because of an increased risk for serious morbidity. In ANRS 1269, which was conducted in Africa, most of the conditions seen in the CD4-based TI arm were serious bacterial infections (non-AIDS–defining) [Abstract 105LB]. Although this arm of the trial was halted, a comparison of continuous therapy and fixed-interval TI is ongoing. In the SMART study, 47 deaths and 46 confirmed AIDS-defining events occurred in the TI arm versus 29 deaths and 15 AIDS-defining events in the continuous-therapy arm [Abstract 106LB]. Most of the deaths and many of the serious disease events were not due to standard AIDS-defining conditions. This observation, along with data from the other trials, suggests that stopping effective antiretroviral therapy might trigger HIV-related immune deficiency or inflammation that could lead to a range of serious conditions affecting vital organs (such as the heart, liver, kidneys, and bone marrow). In the STACCATO trial, the threshold for reinitiation (CD4 count <350 cells/mm³) was higher than in the other studies, and most participants used ritonavir-boosted saquinavir-based therapy (80%) [Abstract 102]. After patients in the TI arm reinitiated treatment, the two arms had similar proportions of subjects with a viral load <50 copies/mL and similar proportions with CD4 counts >350 cells/mm³.

One wonders what the long-term outcome of the SMART study (the largest HIV randomized clinical trial ever) might have been if the study had been allowed to continue with longer follow-up. One also wonders about the level of high-risk behavior and HIV transmission to sexual partners during TI; none of the presentations provided this important public-health information. Overall, the results of the CD4-based TI trials suggest that the door is essentially closed to TI studies that enroll patients with AIDS and that allow the CD4 count to fall below 350 cells/mm³. Further, the findings make the CD4-based TI strategy undesirable in resource-poor countries. For patients with high CD4 nadirs, TI remains a useful experimental strategy for evaluating therapeutic vaccination or immune-based therapies that might be antiretroviral-sparing at higher CD4-cell counts. — Keith Henry, MD

Should We Be Starting Treatment Earlier?

Several analyses from well-known cohort studies suggested that it might be beneficial to start HIV treatment at higher CD4-cell counts. In the ART Cohort Collaboration, researchers followed 10,855 patients for a median of 2.7 years after treatment initiation [Abstract 525]. Not surprisingly, patients who started therapy at CD4 counts 200 cells/mm³ were significantly more likely to progress to AIDS or death than those who started at 201–350 cells/mm³ (hazard ratio, 2.93). More importantly from a policy standpoint, a strong trend was seen across even higher CD4-cell counts, favoring treatment initiation at 351–500 cells/mm³ versus 201–350 cells/mm³ (HR, 1.26). Two particular strengths of this analysis were the relatively large study population and the investigators’ avoidance of lead-time bias by accounting for the time before treatment initiation in patients who delayed therapy. Investigators with the HIV Outpatient Study evaluated the risk for antiretroviral toxicity among treatment-naive patients and found that the incidence rates of renal insufficiency, neuropathy, and lipoatrophy decreased as CD4-cell counts at treatment initiation increased. [Abstract 769]. Finally, in the Hopkins HIV Cohort and Athena National Cohort, researchers found that the absolute increase in CD4-cell count over time was greatest in patients who started therapy at lower values, but the likelihood of achieving a normal CD4-cell count was greatest in those who started at higher values [Abstracts 529 and 530]. The results of these studies, along with implicit messages from the SMART study, cumulatively move the pendulum toward an earlier optimal time to start therapy in asymptomatic patients. — Paul E. Sax, MD

Metabolic Complications and Cardiovascular Disease

Although many antiretrovirals can cause dyslipidemia and insulin resistance, research on metabolic complications is increasingly focused on how chronic HIV infection and host factors contribute to these complications. Several studies at this year’s Retrovirus Conference demonstrated a lower prevalence of diabetes among HIV-infected adults than among population-based controls. In each of these studies, traditional risk factors (e.g., body-mass index and age) overshadowed the effect of antiretroviral therapy on diabetes risk [Abstracts 759, 760, and 761]. During the past several years, many clinicians have focused on minimizing cardiovascular risk by addressing these risk factors, and that focus seems to be paying off. Several large cohort studies suggested that rates of myocardial infarction (MI) and coronary heart disease in HIV-infected adults are stabilizing or even declining [Abstracts 735, 737, and 144]. Although the shifting rates have been attributed to changing patterns of antiretroviral use and increased attention to lipid-lowering and antihypertensive therapy, there is still plenty of room for improvement: In the Swiss HIV Cohort Study, only one third of patients with dyslipidemia or hypertension were receiving lipid-lowering or antihypertensive therapy [Abstract 740]. Randomized trials in HIV-infected patients demonstrated the efficacy of two new options for managing hypertriglyceridemia: salmon oil (3 g daily) [Abstract 756] and fish oil with or without fenofibrate [Abstract 146].

Despite the clear role of traditional risk factors in metabolic complications, antiretrovirals are not completely off the hook. In the Multicenter AIDS Cohort Study, the metabolic syndrome was more common in HIV-positive than HIV-negative men, and the use of potent combination antiretroviral therapy (especially PI use) seemed to be an important contributor [Abstract 747]. Researchers with the Data Collection on Adverse Events of Anti-HIV Drugs (DAD) Study previously reported an association between longer exposure to combination antiretroviral therapy and increased risk for MI (N Engl J Med 2003; 349:1993). Now, observations from that study have been extended to an analysis of specific drug classes and MI risk [Abstract 144]. During nearly 95,000 person-years of follow-up, each additional year of PI exposure increased the risk for MI by 16%. Although a longer duration of NNRTI exposure also increased the risk for MI slightly, the effect was not statistically significant after adjustment for NRTI use. In a matched cohort study, the 3-year rate of progression of carotid intima-media thickness (IMT) was not significantly greater in PI-treated patients than in non-PI–treated patients, but there was some suggestion of a PI effect [Abstract 145]. In another study, cytomegalovirus-specific T-cell responses and high-sensitivity C-reactive protein levels (but not levels of T-cell activation) were significantly correlated with cross-sectional measures of carotid IMT [Abstract 741].

Pulmonary hypertension was a recognized complication of HIV infection before the advent of potent combination antiretroviral therapy, and the problem does not seem to have diminished. In one prospective study, HIV-infected patients had an increased risk for pulmonary hypertension [Abstract 743], and in another, the prevalence of pulmonary hypertension (0.21%) remained similar to that reported in the 1990s [Abstract 744]. Treatment with the oral endothelian antagonist bosentan appears to be well tolerated in HIV-infected patients [Abstract 745], and clinicians are reminded to consider pulmonary hypertension in HIV-infected patients with a differential diagnosis of dyspnea.

The management of lipoatrophy and fat accumulation continues to be a challenging clinical issue. News from this year’s conference confirms that antiretroviral substitutions (switching from AZT or d4T to either abacavir or a nucleoside-sparing regimen) are the best currently available options for managing lipoatrophy [Abstract 755]. Placebo-controlled studies of pioglitazone [Abstract 151LB) and rosiglitazone [Abstract 147] showed small but statistically significant increases in limb fat; in both studies, the effect of the glitazone was diminished in the presence of d4T. Neither metformin nor replacement doses of testosterone significantly reduced visceral fat in controlled trials [Abstracts 147, 148, and 149]. — Judith Currier, MD, MSc

Renal Tenofovir Debate Continues

Given the frequent development of nephrotoxicity with adefovir and cidofovir, concerns about the potential for nephrotoxicity with tenofovir are well founded. Several new studies highlighted the frequency of this problem in tenofovir-treated patients and the potential importance of drug-drug interactions.

In the ESS40006 study, patients receiving tenofovir experienced significant declines in glomerular filtration rate (GFR) at 24 and 48 weeks, whereas patients receiving efavirenz did not [Abstract 777]. In another study of patients initiating tenofovir, 4% developed renal insufficiency within one year, and 13% developed hypophosphatemia [Abstract 778]. Renal insufficiency seemed to occur with cumulative exposure to tenofovir, whereas hypophosphatemia developed acutely. Risk factors for nephrotoxicity included being treatment-naive and having previously received amphotericin B. In a much larger study, CDC investigators analyzed longitudinal data from 11,362 HIV-infected patients, all of whom had GFR >90 mL/min at baseline [Abstract 779]. Mild renal impairment (GFR, 60–89 mL/min) was seen in 35.1% of patients, moderate impairment (GFR, 30–59 mL/min) was seen in 6.4% of patients, and severe impairment (GFR, <30 mL/min) was seen in 2.6% of patients. Tenofovir treatment was significantly associated with mild and moderate renal insufficiency, but not with more severe renal impairment. In an observational study of 497 patients initiating tenofovir, 87 patients (17.5%) developed renal dysfunction [Abstract 780]. Most experienced moderate declines in GFR (60 mL/min), but eight had more severe declines (>80 mL/min) that were associated with concomitant treatment with ritonavir-boosted regimens containing either amprenavir or lopinavir. In stark contrast to these reports were combined data from the tenofovir expanded-access program and postmarketing safety reports, which did not suggest a high incidence of serious renal adverse events in tenofovir-treated patients [Abstract 781].

Taken collectively, these findings suggest that clinicians should be concerned about the development of renal dysfunction in patients receiving tenofovir. Routine laboratory monitoring of patients on this drug should include calculation of GFR and assessment of serum phosphate. For the most part, renal dysfunction appears to be reversible when the agent is discontinued expediently. The incidence and risk factors for tenofovir-induced nephrotoxicity still need to be defined in large studies that control for antiretroviral regimen and associated comorbidities. In addition, more data are needed on drug-drug interactions and the possible role of boosted PIs in potentiating tenofovir-associated renal toxicity. — G. Sonia Nagy, MD

Extending the Role of Atazanavir

The PI atazanavir is increasingly popular for both treatment-naive and treatment-experienced patients because of its low pill count, tolerability, and favorable effects on lipid levels. Although many clinicians routinely offer boosted atazanavir to their treatment-naive patients, there have been no prospective data on this practice. In the first study to address this, researchers randomized 200 treatment-naive individuals to receive a nucleoside backbone of 3TC and extended-release d4T with either unboosted atazanavir (400 mg) or boosted atazanavir (300 mg plus 100 mg of ritonavir) once daily [Abstract 107LB]. The 48-week intent-to-treat analysis showed no significant differences in the proportion of subjects achieving a viral load <400 copies/mL (86% in the boosted arm and 85% in the unboosted arm) or <50 copies/mL (75% and 70%, respectively). Although study-drug discontinuation rates were similar in both arms, the boosted arm had a higher rate of discontinuation for adverse events; it also had a greater proportion of subjects (59%, compared with 20% in the unboosted arm) experiencing a rise in total bilirubin to greater than 2.5 times the upper limit of normal. As expected, changes in total cholesterol, LDL cholesterol, and triglyceride levels were more pronounced in the boosted arm. More virologic failures occurred in the unboosted arm (10 vs. 3), and three patients in this arm developed the signature mutation for atazanavir resistance, I50L. No PI resistance was seen in the boosted group.

In a small pilot study, lopinavir/ritonavir was effective as sole maintenance therapy in patients who were previously fully suppressed after triple-drug therapy that included lopinavir/r (J Acquir Immune Defic Syndr 2005; 40:280). In ACTG 5201, a single-arm pilot study, researchers evaluated whether boosted atazanavir alone might also be an effective maintenance therapy [Abstract 108LB]. The study involved 36 subjects with a CD4 count 250 cells/mm³ who had achieved a viral load <50 copies/mL on PI-based combination therapy for at least 48 weeks and had no history of virologic failure. After study entry, participants were all switched to ritonavir-boosted atazanavir plus two NRTIs for 6 weeks. If the viral load remained fully suppressed, the two NRTIs were discontinued, and subjects continued on boosted atazanavir alone for 48 weeks. At 24 weeks, 91% of subjects achieved a viral load <200 copies/mL. The eight subjects who provided semen samples for viral-load testing all had values <150 copies/mL. None of the three participants with virologic failure had any resistance, and two of the three had no detectable plasma atazanavir levels. The results of this pilot study suggest that ritonavir-boosted atazanavir should now be thoroughly evaluated in a larger, randomized trial. — Judith Feinberg, MD

Antiretroviral Therapy in Resource-Limited Settings

This year’s Retrovirus Conference featured a marked and gratifying increase in presentations related to HIV in resource-limited settings. Two summary talks were of particular importance and are available as webcasts. The first, by Dr. Tony Harries, was a comprehensive and thoughtful review of the epidemiologic convergence of HIV and tuberculosis (TB) in sub-Saharan Africa [Abstract 9]. The second, by Dr. Tom Quinn, was a tour-de-force presentation of the epidemiologic, biologic, research, and policy issues surrounding male circumcision as a strategy for HIV prevention [Abstract 120].

Antiretroviral therapy has now been available long enough in resource-limited settings to explore and document its effectiveness. Initial concerns about adherence have decreased, examples of biologic and clinical benefit have been documented, and now reports are emerging about the cumulative side effects and toxicities in large-scale national rollout programs. In rural Uganda, researchers evaluated early clinical toxicity among more than 1000 adults receiving an initial regimen, most commonly d4T + 3TC + nevirapine [Abstract 142]. The overall rate of toxicity was 4.5 per 100 person-months, and the rate of severe toxicity was 1.3 per 100 person-months. Most adverse events were grade 1 or 2. Neuropathy and rash were the most common events and were manageable by single drug substitutions. A smaller but more detailed study from Nairobi, using the same regimen, yielded similar conclusions [Abstract 143]. In this study, most adverse events occurred within 6 months of follow-up, were of grade 1 or 2 in severity, and were resolved with regimen switches. In Cape Town, South Africa, approximately 11% of patients on an initial regimen of d4T + 3TC + nevirapine had a substitution for toxicity in the first 36 months of treatment [Abstract 66]. Most substitutions were for d4T and were related to the appearance of peripheral neuropathy and, more disturbingly, lactic acidosis. The latter appeared at a higher rate than anticipated based on research in developed countries and is of particular concern because of its vague clinical symptoms and the difficulty of documenting lactate levels. Women were at particularly high risk for developing lactic acidosis on d4T, as were patients weighing more than 75 kg.

These caveats aside, the emerging data indicate that, overall, adverse events related to commonly used first-line regimens are not a significant barrier to provision of antiretroviral therapy in government rollouts. — Gerald H. Friedland, MD

The New ABCs: Antiretrovirals, Barriers, and Circumcision

During the keynote lecture, Dr. James Curran told us that "prevention is not as simple as ABC," referring to Abstinence, Being faithful, and using Condoms. However, at this meeting, we became familiar with the "second-generation ABCs": Antiretrovirals, Barriers (condoms and microbicides), and Circumcision. Will this new generation of ABCs gain more respect than the last one as a viable set of HIV prevention tools?

Antiretrovirals are now used routinely for nonoccupational postexposure prophylaxis (nPEP) [Abstract 54]. However, there is no clinical evidence of benefit, and often the therapy is not initiated within 72 hours of exposure as recommended by CDC guidelines. In a review of 880 calls placed to a national PEP hotline following sexual exposures, researchers found that 28% of the calls came in more than 72 hours after exposure [Abstract 906]. Although many countries now have guidelines for PEP [Abstract 904], it is still uncertain which regimen is best. Data from a prospective study in France suggest that AZT/3TC + LPV/ritonavir might be the combination with the best tolerability [Abstract 905]. Pre-exposure prophylaxis is also being evaluated, with five trials under way in humans. To date, the animal data seem promising: In a study of nonhuman primates, tenofovir/FTC offered 100% protection against repeated rectal SHIV challenge [Abstract 32LB]. Ultimately, the drugs of choice for pre- and post-exposure prophylaxis may be determined by the penetration of antiretrovirals into genital secretions [Abstracts 129, 396, and 569].

Barrier methods beyond condoms are necessary to protect women from HIV. The growing consensus is that a highly effective female-controlled barrier requires a combination of products or methods; one compelling approach is a cervical barrier with microbicides [Abstract 55].

The potential value of circumcision in HIV prevention received widespread attention last year, when a randomized trial in South Africa demonstrated a 60% reduction in HIV acquisition among circumcised men compared with uncircumcised men (PLoS Med 2005; 2:e298). Now, data from the Rakai Community Cohort Study suggest that male circumcision also reduces the incidence of HIV infection among female partners [Abstract 128]. Incidence rates were 6.6 per 100 person-years among the wives of infected circumcised men and 10.3 per 100 person-years among the wives of infected uncircumcised men. An ongoing controlled trial in Rakai will yield more information later this year on how male circumcision affects HIV acquisition in female partners. — Carlos del Rio, MD

More on False-Positive Results in Rapid HIV Testing

In late 2005, counselors at several HIV testing programs (mainly in New York City and San Francisco) reported an unusually high number of false-positive results from the OraQuick Advance Rapid HIV-1/2 Antibody Test with oral fluid (ACC Jan 4 2006). Since then, CDC investigators have analyzed data from four prospective studies that involved simultaneous testing of whole blood and oral fluid between 2000 and 2005 [Abstract 34LBb]. They found that the specificity of OraQuick was 99.9% with whole blood and 99.6% with oral fluid; in comparison, the specificity of serum EIA was 99.7%. These data suggest that the specificity of HIV rapid testing is slightly lower with oral fluid than with whole blood, but is still well above the FDA minimum threshold (98%) for both specimen types. At three testing sites, the excess number of false-positive results appeared to be related to unidentified host- or site-specific factors; CDC investigators found no evidence of a lot- or device-related problem.

Investigators also evaluated an algorithm that was adopted in New York and San Francisco in late 2005. According to this algorithm, a positive result on an oral-fluid rapid test should be followed up with a whole-blood HIV rapid test using a fingerstick. When this algorithm was used, positive results from both oral fluid and fingerstick were almost always followed by a positive Western blot result, whereas a positive result from oral fluid followed by a negative fingerstick was almost always followed by a negative Western blot result. These data suggest that performing a fingerstick test after a reactive oral-fluid test might reduce the number of people who receive false-positive results. However, confirmatory testing is still required for all reactive rapid tests. — Carlos del Rio, MD

Expanding HIV Testing

The incidence of HIV in the U.S. is holding steady at about 40,000 new cases per year, with the epidemic being fueled largely by the 250,000 HIV-infected people who are unaware of their serostatus. Studies show that these individuals are more likely than those who know their serostatus to practice unprotected sex with HIV-negative partners (ACC Sep 14 2005). To address this problem, the CDC developed an initiative in 2003 called "Advancing HIV Prevention: New Strategies for a Changing Epidemic." One strategy, "Prevention for Positives," focuses on changing the behavior of those already infected. Another is designed to expand HIV testing by making it a routine, voluntary part of care for all individuals aged 13 to 64 [Abstract 164]. This approach is feasible because 22% of all testing is currently done in hospitals, emergency rooms, and outpatient clinics, and 27% of all positive tests occur in these settings. The CDC is currently revising its testing recommendations and may propose that HIV testing be included in an individual’s overall consent to receive care. Under such a policy, a person who did not want to be tested for HIV would have to opt out — a distinct departure from the current policy of opting in, which requires that patients give explicit, signed consent for HIV testing when it is offered. Hopefully, this approach will result in more HIV-infected people being identified, referred for care, and placed on antiretroviral therapy as appropriate; this should, in turn, reduce transmission through both serostatus awareness and controlled viral replication.

Another strategy by the CDC is to accept only name-based HIV test results from state and local health departments, so that classic public health strategies, such as contact tracing, can be used to identify more infected individuals. If health departments do not implement name-based reporting, they risk losing federal HIV/AIDS funds. Apparently, the threat is working: Shortly after the Retrovirus Conference, California health officials announced that they would move to confidential name-based testing. — Judith Feinberg, MD

Conclusion

As the year unfolds, we’ll eagerly await follow-up information from many of the studies presented at the Retrovirus Conference. On the treatment front, we hope to see accelerated approval of TMC114 (darunavir), additional data on integrase inhibitors, and more information on the risks and benefits of CCR5 antagonists in treatment-naive and treatment-experienced patients. On the prevention front, we’ll look to Rakai for more data on male circumcision and to the CDC for new recommendations on HIV testing. — Paul E. Sax, MD

Published in AIDS Clinical Care March 15, 2006