1. Home>
2. Specialties>
3. HIV/AIDS Clinical Care>
4. Top Story

Top HIV Medicine Stories of 2004

The editors of ACC offer their perspectives on the most important stories of the past year.

Bad Nukes Come in Threes

The bad news about triple-NRTI regimens in 2003 only seemed to get worse in 2004. The disappointment that started with preliminary data on Trizivir in ACTG 5095 (see separate Top Story) grew with each additional study. At the Retrovirus Conference in San Francisco, Jemsek and colleagues described a single-arm pilot study of once-daily ddI + 3TC + tenofovir (11th CROI, Abstract 51); the pattern was the same as that seen in studies of once-daily abacavir + 3TC + tenofovir: a high rate of early virologic failure and the rapid development of resistance. Viral load decreases from baseline were progressively smaller over time (0.75 log at week 4; 0.61 log at week 12; 0.49 log at week 24 or treatment discontinuation); which led to the study's early termination. As in other studies of tenofovir combined with two non-thymidine NRTIs, researchers found significant genotypic resistance: 20 subjects developed 3TC resistance (some as early as week 4); 10 patients developed primary tenofovir resistance (which also reduces susceptibility to ddI), and this also appeared to evolve very early in treatment. Landman and colleagues reported on another single-arm pilot study of abacavir + 3TC + tenofovir, which was terminated early when 12 of the 38 enrolled patients experienced virologic failure by week 24 of the planned 48-week trial (11th CROI, Abstract 52). Virologic failure was associated with higher baseline viral loads. The rate of the resistance development was breathtaking: of the 12 patients with genotypic results at the time of virologic failure, 9 developed both 3TC and tenofovir resistance (M184V and K65R), and 2 had only 3TC resistance. Based on pharmacokinetic data from these and other studies presented in the past year (e.g., Hawkins T et al. XV International AIDS Conference, Abstract TuPeB4627), early virologic failure with these triple-NRTI regimens clearly results from selective pressure for resistance rather than from untoward drug interactions.

Taken together with disappointing data regarding tenofovir + ddI + efavirenz (see Meeting Notes in this issue of ACC), these data warn us that all that glitters on the horizon of simplified therapy is not gold. Novel combinations should be tested thoroughly before becoming part of routine clinical practice, despite the seeming logic of combining drugs with similar and simpler dosing schedules. -- Judith Feinberg, MD

Perinatal Transmission Interventions: The Benefits Come with Resistance

In the past year, progress has been made toward widespread provision of antiretroviral drugs in developing countries. Legitimate concern exists, however, that suboptimal operation of antiretroviral programs could result in the emergence and spread of drug-resistant HIV.

Particularly worrisome is the development of resistance with the use of antiretroviral therapy for prevention of perinatal transmission, as previously reported with single-dose nevirapine (sdNVP) or with short-course AZT. At this year's Retrovirus Conference, Mantinson and colleagues reported that NNRTI mutations were detected in 32% to 44% of women in the early postnatal period, a figure substantially higher than the 15% to 24% reported in earlier studies (11th CROI, Abstract 38). Resistant viruses continue to be detected at 6 months' postpartum in about 14% of women and disappear entirely between 12 and 24 months.

In a study that has been widely misinterpreted, Jourdain and colleagues addressed the question of whether antiretrovirals used in perinatal interventions may be less effective when used subsequently (N Engl J Med 2004; 351:229). Thai women exposed to sdNVP were significantly less likely than unexposed women to achieve viral loads <50 copies/mL on a nevirapine-containing regimen; however, the long-term consequences of failure to achieve such maximum suppression are unknown, and the two groups did not differ significantly in clinical response, immunologic response, or virologic suppression to <400 copies/mL. The rate of virological failure (>50 copies/mL) was highest among women with the highest baseline viral loads; these women account for a minority of antenatal clinic attendees in developing countries and are the very individuals who should qualify for potent combination antiretroviral therapy. Although important, these findings are insufficient to preclude sdNVP from programs designed to reduce perinatal HIV transmission.

Several studies on perinatal prevention strategies presented in 2004 did, however, demonstrate ways to reduce resistance. A preliminary report from South Africa suggested that reducing viral replication by combining sdNVP with postnatal AZT/3TC may lower NVP resistance in women to about 10% (J McIntyre et al. XV International AIDS Conference, Abstract LBOrB09). Studies from Africa and Thailand presented at the Retrovirus Conference found that combining NVP with AZT from between 28 to 36 weeks of pregnancy resulted in NVP resistance rates of 18% to 28% (Chalermchokcharoenkit A et al. Abstract 96; Jourdain G et al. Abstract 41LB; Chaix ML et al. Abstract 657). Adding dual-NRTI regimens to NVP lowered the rate to 15% (Lyons F et al. Abstract 892). -- Hoosen Coovadia, MD

HAART and the Heart

Since the Data Collection on Adverse Events of Anti-HIV Drugs (DAD) Study Group published its research at the end of 2003 (N Engl J Med 2003; 349:1993; 350:955), evidence has continued to mount suggesting that long-term use of combination antiretroviral therapy may contribute to an increased risk for atherosclerosis. The original DAD paper documented a 26% increased risk for myocardial infarction per year of exposure to combination therapy. Recently, additional work in this cohort extended these observations to include cerebrovascular events (DAD Study Group. AIDS 2004; 18:1811). In addition, another group of investigators used carotid intima thickness - a surrogate marker for atherosclerosis - to show greater rates of progression among HIV-positive patients on long-term antiretroviral treatment than among HIV-negative controls (Hsue PY et al. Circulation 2004; 109:1603).

Although neither the magnitude of the increased risk nor the specific contribution of metabolic changes associated with antiretrovirals is well defined, the evidence continues to point towards a relation between prolonged antiretroviral use and cardiovascular and cerebrovascular events. As we await additional data in this area, one certainty is that HIV providers are developing strategies for minimizing cardiovascular risk factors (including use of lipid-lowering therapy and counseling patients on diet, exercise, and smoking cessation) and making these more prominent features of clinical care. -- Judith Currier MD, MSc

When to Start: Guidelines and Intuition

Two large observational cohort studies published in the fall of 2003 added to the body of data on the perennial question of when to start antiretroviral therapy. In the first, Johns Hopkins researchers found a trend towards more clinical events among patients starting therapy at CD4 counts of 201-350 cells/mm³ than among those starting therapy at CD4 counts >350 cells/mm³ (Sterling TR et al. J Infect Dis 2003; 188:1659). In the second study, patients with pretreatment CD4 counts of 200-350 cells/mm³ who were moderately adherent to therapy had no greater mortality than those starting treatment at higher CD4-cell counts (Wood E et al. Ann Intern Med 2003; 139:810). Patients who were less adherent, however, had higher mortality rates than those who were moderately adherent, regardless of their pretreatment CD4-cell count.

Results of a third observational study that looked at treatment decisions specifically in IV drug users were published this past year (Wang C et al. J Infect Dis 2004; 190:1046). Among patients beginning treatment with CD4 counts >350 cells/mm³, mortality was similar to that among drug users without HIV infection. Among patients with baseline counts 200-350 cells/mm³, treatment improved the mortality rate, but did not return it to the level seen among the uninfected. Adherence to medication was not addressed in this study, but presumably strongly influenced these results.

Treatment guidelines reissued this year continue to endorse delaying treatment for asymptomatic persons with CD4 counts >500 cells/mm³, treating those with counts <200 cells/mm³, and individualizing decisions for those with counts in between, taking such factors as patient preference into consideration.

The bottom line: Treatment in patients with CD4 counts <200 cells/mm³ is lifesaving. However, the data seem to point to a considerable "crystal ball" factor when it comes to decisions for those with CD4 counts between 200 and 350 cells/mm³, which requires the physician to make his or her best possible guess as to a patient's likelihood of adhering to treatment before actually putting pen to prescription pad. -- Abigail Zuger, MD

Resistance Rates: High in New Infections, Higher in Older Ones

This year, researchers continued to track the prevalence of drug resistance among patients with newly diagnosed HIV infections. A large U.S. study that compiled data on more than 1000 treatment-naive patients diagnosed at 39 clinical centers between 1997 and 2001 found that the overall prevalence of drug-resistance mutations was 8.3%; 1.9% of the cohort had virus that was resistant to two or more drug classes (Weinstock HS et al. J Infect Dis 2004; 189:2174). These figures square with international data presented at the 11th Conference on Retroviruses and Opportunistic Infections, which showed overall resistance rates of 6% to 12% for the same period in newly diagnosed patients (Bezemer D et al. Abstract 679; Yerly S et al. Abstract 680; Masquelier B et al. Abstract 683).

Resistance rates among patients with long-term infection far exceed these numbers. One large study demonstrated phenotypic resistance rates of 71% for NRTIs, 41% for PIs, and 25% for NNRTIs among patients in the U.S. with chronic infection and detectable virus. Overall, 76% had resistance to at least one drug class, 48% had resistance to two classes, and 13% had resistance to all three drug classes (Richman DD et al. AIDS 2004; 18:1393).

How drug resistance affects treatment success - and what to do about it - is still being sorted out. Resistance does not always compromise treatment efficacy: one study presented at the 11th Conference on Retroviruses and Opportunistic Infections found that in a small group of newly infected patients, time to viral suppression with triple-drug treatment was not affected by resistance to any one class (Pillay D et al. Abstract 685). These findings were supported by the results of a randomized study of the efficacy of genotype and phenotype testing in preventing persistent treatment failure: among patients without extensive drug experience, resistance testing had no significant effect on outcome. However, among highly drug-experienced patients and those with NNRTI exposure, access to resistance testing did result in more durable viral suppression (Wegner SA et al. Clin Inf Dis 2004; 38:723). -- Abigail Zuger, MD

Uncertainty About Treatment of Acute HIV Infection

Interest in aggressive identification and treatment of acute HIV infection was stimulated by the initial observation by Rosenberg and colleagues that early treatment can promote immune control of HIV when therapy is withdrawn (Nature 2000; 407:523). Adding to the interest in treating primary HIV infection were observations that acute HIV infection is associated with a potential for efficient transmission of HIV to sex partners (Pilcher CD et al. J Infect Dis 2004; 189:1785) and that early treatment might decrease the size of the cellular reservoir of HIV (Pires A et al. J Acquir Immune Def Syndr 2004; 36:783).

However, recent follow-up of the cohort in the Rosenberg study has demonstrated that the viral control observed after treatment of acute HIV infection has limited durability (Kaufmann DE et al. PLoS Med 2004; 1:e36). These disappointing results, combined with the experience of other investigators and clinicians, led Rosenberg and several of his colleagues to pen an editorial (Smith DE et al. AIDS 2004; 18:709) that called for more studies while concluding that there was no clear evidence that immediate treatment of acute HIV infection results in a significant clinical benefit. These developments are reflected in the most recent iteration of the DHHS Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents, which state that therapy for acute HIV infection is optional (updated October 29, 2004; available at www.aidsinfo.nih.gov; see page 29 in the PDF). -- Keith Henry, MD

STIs in Chronic Infection: The Jury Is Still Out

This year saw little progress towards consensus on the feasibility of structured treatment interruptions in managing patients with established HIV infection; indeed, the newest data still straddle both sides of the question.

Last year, a large U.S. study of a 16-week treatment interruption prior to salvage in late-stage patients was stopped when a high rate of AIDS-related complications emerged in patients taken off treatment (Lawrence J et al. N Engl J Med 2003; 349:837). This year, however, a similar French study was published with opposite results: patients taking an 8-week STI before starting a new regimen had superior CD4-count and viral-load responses to salvage treatment compared with those starting a new regimen immediately, although the clinical outcome in both groups was similar (Katlama C et al. AIDS 2004; 18:217-26). Commentators tentatively pointed to differences in the length of the STIs and in the salvage protocols to explain the discrepancy between these studies, but without further studies, the bottom line for clinicians remains unclear.

The XV International AIDS Conference in Bangkok hosted progress reports from several ongoing European studies of STIs. A 7-day on/7-day off treatment arm in one was prematurely curtailed by a high rate of virologic failure (Ananworanich J et al. Abstract WeOrB1283). However, two protocols using predetermined CD4-cell criteria to configure STI duration are ongoing. Another small study evaluating a 5-day on/2-day off treatment schedule in patients with virologic control showed a 90% success rate after almost a year of follow up (Cohen CJ et al. Abstract TuPeB4575). Plans to expand this study are underway. The most recent analysis of STI in chronic infection found no difference between patients who were receiving therapy at the outset of the TORO trials of T-20 and those who had discontinued therapy. (Miralles GD et al. 7th International Conference on Drug Therapy in HIV Infection, Abstract P17). Although this was not a randomized study, it adds to the confusing picture of STI in chronic infection. -- Abigail Zuger, MD

Back to the Future: Lopinavir/ritonavir as "Monotherapy"

Several studies are investigating the use of lopinavir/ritonavir as single-drug treatment, either as initial therapy or as the latter part of an induction-maintenance strategy. Two research teams reported encouraging results of single-agent lopinavir/ritonavir at the International AIDS Conference in Bankok. Joseph Gathe presented follow-up results from his pilot study of 30 treatment-naive patients - many with advanced HIV disease - receiving lopinavir/ritonavir as initial therapy (Abstract MoOrB1057). In an intent-to-treat analysis at 48 weeks, 67% and 60% of patients had achieved a viral load <400 copies/mL and <50 copies/mL, respectively. Jose Arribas presented data on an induction-maintenance study, in which 42 patients with virologic suppression were randomized to continue their regimens of lopinavir/ritonavir + 2 NRTIs or switch to lopinavir/ritonavir alone (Abstract TuPeB4486). At 24 weeks, rates of virologic failure in the two groups did not differ significantly; however, there was a trend towards a higher rate of failure in the maintenance arm.

Encouragingly, in both studies, patients who did experience virologic failure had no lopinavir resistance on phenotypic or genotypic testing and promptly experienced virologic suppression upon switching to triple-drug therapy. Larger studies of these strategies (including at least one with boosted atazanavir) are ongoing; we will have to review the results carefully before this novel approach is adopted in clinical practice. -- Paul E. Sax, MD

ACTG 5095: Efavirenz Beats AZT + 3TC + Abacavir

Concerns about the toxicity of PIs, the desire to preserve treatment options by using class-sparing regimens, and the ease of Trizivir (AZT/3TC/abacavir) administration prompted the AIDS Clinical Trials Groups to perform a direct comparison of three PI-sparing regimens: Trizivir alone, efavirenz + AZT/3TC (Combivir), and efavirenz + Trizivir (Gulick RM et al. N Engl J Med 2004; 350:1850). A total of 1147 antiretroviral-naive subjects were enrolled in the study, 382 in the triple-NRTI arm and 765 in the two efavirenz-containing arms. At baseline, approximately 43% of subjects had a viral load >100,000 copies/mL, and nearly half had a CD4 count <200 cells/mm³. Following an interim review of the results at 32 weeks, the data safety monitoring board recommended that the triple-NRTI arm be stopped and that the data be presented comparing this arm with the pooled efavirenz-containing arms. At the time of the interim analysis, 21% of the subjects in the triple-NRTI arm had experienced virologic failure, compared with 11% in the combined efavirenz arms. Time to virologic failure was also significantly shorter in the triple-NRTI arm. These differences were observed regardless of baseline viral load. As a result of this study, the DHHS guidelines now list the triple-NRTI combination of AZT/3TC/abacavir as not recommended, unless no other treatment options are available. Quite clearly, efavirenz wins again. -- Carlos del Rio, MD

To 3TC or Not To 3TC?

The M184V mutation confers high-level resistance to 3TC and FTC, but it is also associated with a significant reduction in the HIV's ability to replicate. Some researchers therefore have hypothesized that continuation of 3TC despite documented resistance could provide an overall benefit. Two studies that have tested this idea - one in which patients with 3TC resistance continued the medication as monotherapy during a drug holiday and the other in which patients continued it as part of a larger salvage regimen - reached very different conclusions.

In a pilot study presented in Bangkok, 50 patients who had the M184V mutation and were planning to interrupt antiretroviral therapy were randomized to stop all drugs or to continue on 3TC monotherapy (Castagna A et al. XV International AIDS Conference, Abstract WeOrB1286). After 24 weeks, the group receiving 3TC monotherapy had a significantly higher CD4-cell count and a significantly lower viral load than the group receiving no therapy. In the COLATE trial presented in San Francisco, 131 patients with virologic failure on a 3TC-containing regimen were randomized to continue 3TC (150 mg twice daily) or discontinue it; all patients also received a triple-drug salvage regimen (Dragsted U et al. 11th Conference on Retroviruses and Opportunistic Infections, Abstract 549). No differences between the two groups were found in primary and secondary analyses of virologic efficacy, and the study was terminated early. These results, however, may not hold for patients with fewer treatment options or after longer follow-up.

These small, but intriguing studies suggest that among patients with the M184V mutation, 3TC monotherapy may be better than no drug at all for those planning to interrupt therapy, but continuing 3TC in conjunction with a salvage regimen might not be worthwhile. -- Helmut Albrecht, MD

Nevrirapine and Efavirenz: Similar Efficacy, Different Toxicities, and Better Separately than Combined

The 2NN, a large multinational randomized trial addressed the issue of which NNRTI -nevirapine or efavirenz - was best for treatment-naive patients and whether the combination of the two was preferable to either individually (van Leth F et al. Lancet 2004; 363:1253). More than 1200 patients were randomized to receive one of four regimens: nevirapine 200 mg twice daily, nevirapine 400 mg daily, efavirenz 600 mg daily, or the combination of efavirenz 800 mg plus nevirapine 400 mg daily. All regimens were combined with d4T + 3TC. The primary outcome measure was a combination of virologic failure, disease progression, and treatment change.

The results indicated that used separately, nevirapine (at either dosage) and efavirenz did not differ significantly in efficacy, whereas the dual-NNRTI combination was associated with significantly higher rates of treatment failure. Nevirapine was associated with significantly higher rates of hepatobiliary toxicity, and efavirenz was associated with significantly higher rates of CNS/psychiatric adverse events.

For the clinician starting patients on antiretroviral therapy, nevirapine and efavirenz are associated with roughly similar rates of success but with different toxicities, and the choice between them should be driven by consideration of safety profiles on a patient-by-patient basis. This study does not support the use of dual-NNRTI therapy. -- Gerald Friedland

DNA-MVA Prime Boost - Another Vaccine Candidate Bites the Dust

The elicitation of a strong cellular response to an HIV vaccine has eluded scientists for many years. However, with the recent development of the prime boost strategy, many anticipated that higher immune response rates would be achieved, and this approach quickly came to dominate HIV vaccine research of late.

Animal studies and phase I clinical trials of the DNA prime and modified vaccinia Ankara (MVA) boost strategy showed promise. This vaccine combination was shown to be safe and to induce HIV-specific responses in eight of nine human recipients in a phase I study in the U.K. (Mwau M et al. J Gen Virol 2004; 85:911). However, interim data from the phase II trials presented at the AIDS Vaccine 2004 conference in Lausanne were very disappointing. The vaccine candidate's ability to elicit anti-HIV cell-mediated response was extremely poor, falling short of expectations established during preclinical study. Only limited immune responses were achieved - 19% in U.K. volunteers, 25% in Kenyan volunteers, and 10% in Ugandan volunteers. These results are no better than earlier canary pox vaccines that were discarded for lack of promise. Although these trials will be completed, the vaccine's sponsors have decided not to fund it further.

Unfortunately, another promising vaccine has bitten the dust, but not all prime boost vaccine candidates, nor even all of those with DNA-MVA, are the same. Promising preclinical results have been achieved with adenovirus type 5 (Ad5) vector. However, this vaccine may have limited utility because of the high prevalence of Ad5 antibodies throughout much of Africa; for populations with low prevalence of Ad5 antibodies this vaccine holds great promise.

The disappointing results from the clinical trials of the DNA-MVA vaccine candidate call into question how animal studies are interpreted; a closer look at the selection of primate models may be needed. An AIDS vaccine remains an important hope, but its realization remains distant and the design challenges formidable. -- Salim S. Abdool Karim, MD

— ACC Editors

Published in AIDS Clinical Care January 1, 2005