The 16th International AIDS Conference: What Did It Achieve?

By Brian Boyle, MD

Over the past decade, some have characterized the international AIDS conferences as long on political rhetoric and short on new scientific data. In the following commentary, Brian Boyle, MD, medical editor of HIV and Hepatitis.com, offers his perspective as a treating physician and a delegate to the 16th International AIDS Conference, held August 13-18, 2006 in Toronto, Canada, on the political and scientific achievements of the meeting.

Introduction

In the view of this author, the importance of the recent International AIDS Conference (IAC) in advancing the worldwide struggle against HIV/AIDS cannot be overstated. The conference theme, "Time to Deliver," brought home the message that mere promises of future aid - which so often in the past have turned out to be empty gestures - will no longer be acceptable to those involved in the struggle. [Please note that, unless otherwise stated, all references are to the Program and Abstracts of the 16th International AIDS Conference. August 13-18, 2006. Toronto, Canada].

The 16th IAC was also notable in being the first HIV/AIDS to "sell out," with the conference organizers announcing prior to its opening that no further applications were being accepted. All told, more than 25,000 researchers, health care providers, media representatives, patient advocates, and people with HIV attended the conference, and there were more than 5,000 oral and poster presentations. Many of the presentations concerned scientific and clinical information, but also served as a platform to examine the sociological and political issues surrounding HIV/AIDS worldwide, especially the disturbing status of the epidemic still facing those living in developing countries.

The Global Situation

As presentations at the 16th IAC made clear, some progress is being made in providing care to all who need it, but this progress has been slow, and we are still a long way from controlling the spread of HIV/AIDS and treating all of the persons currently living with the disease. For example, while provision of antiretroviral therapy has increased, from 7% of patients who needed treatment in 2003 to 20% in 2005, and availability of mother-to-child prevention services for pregnant women increased from 7.6% in 2003 to 9% in 2005, it is clear that there are still many people with HIV/AIDS who are unable to get the care or antiretroviral therapy they need. (UNAIDS Report on the Global AIDS Epidemic, 2006).

It is also disturbing that many of the interventions used over the past 25 years have not led to a plateau in the number of individuals living with HIV/AIDS or the number of orphans (see Figure 1). More effective strategies and more research are urgently required. Helene Gayle, one of the IAC co-chairs said it best: "Very soon, we could have new, highly effective ways to prevent many of the 4 million new HIV infections that occur every year. But these tools will have little impact in the real world unless we take immediate steps to complete current trials, mount new ones, and reach people most in need."

Figure 1: From UNAIDS Report on the Global AIDS Epidemic, 2006




Prevention was also a significant focus of this conference, and in his speech, former President Bill Clinton stated, "I just don't believe we can reverse this if we keep having more people infected every year than we are increasing the number of people on medication."

As discussed at the conference, there are numerous studies underway to find effective prevention strategies, focusing on:

Male circumcision;
Cervical barriers such as diaphragms;
Daily administration of antiretroviral drugs currently used for HIV treatment (PREP);
Suppression of herpes simplex virus (HSV) infection;
Topical microbicides;
HIV vaccines.

While there were a few promising studies regarding some of these strategies, the discussion and presentations concerning the most hoped for strategy - an HIV vaccine - were not very encouraging. As Cate Hankins, the chief HIV scientist at the United Nations, stated, "A vaccine is going to be critical in the long run to containing the epidemic. Unless there's some striking breakthrough that we're not aware of, it's likely to be another 10 years before [an effective vaccine] is commercially available." Obviously, we need to move forward with new vaccine strategies, and the sooner the better, since every day lost means thousands of new infections and AIDS deaths.

New IAS-USA Guidelines Regarding Antiretroviral Therapy

Highly active antiretroviral therapy (HAART) has remarkably changed the landscape of HIV disease in developed countries, and, in these nations, we now have the luxury of discussing which starting regimens should be "recommended" and which are "alternative."

There are now 2 sets of guidelines for developed countries: the International AIDS Society (IAS)-USA Guidelines (S Hammer and others, Treatment for Adult HIV Infection: 2006 Recommendations of the International Aids Society-USA Advisory Panel. JAMA 296(7): 827-843, August 16, 2006) and the U.S. Department of Health and Human Services (DHHS) Guidelines; (Revised May 4, 2006).

Both sets of guidelines use an evidence-based approach; however, the authors of the IAS-USA guidelines, like most clinicians, are more willing to use clinical inference - i.e., "inferences from studies with similar drugs (or in similar settings)" - as they did when they listed atazanavir/ritonavir (boosted Reyataz) as a recommended protease inhibitor (PI) in the 2004 iteration of the guidelines. (P Yeni P and others. Treatment for Adult HIV Infection: 2004 Recommendations of the International Aids Society-USA Panel. JAMA 292(2): 251-265, July 14, 2004).

Initiation of Therapy

An update of the IAS-USA guidelines was issued concurrently with the IAC, and some of the changes were based upon data presented at the conference. [LINK: http://www.hivandhepatitis.com/2006icr/16aids/081506_d.html] There were no changes to the IAS-USA guidelines with respect to the recommended time to start antiretroviral treatment. The guidelines continue to recommend antiretroviral therapy for all symptomatic patients and for asymptomatic patients with CD4 lymphocyte counts below 200 cells/mm3. Further, antiretroviral therapy should be considered and discussed with patients whose CD4 counts are 200-350 cells/mm3. Therapy should typically be deferred for asymptomatic patients with CD4 counts above 350 cells/mm3.

What Therapy to Use First?

Regarding which antiretroviral therapy to start in a treatment-naive patient, some important changes were made to the latest IAS-USA guidelines. In the nucleoside/nucleotide reverse transcriptase inhibitor (NRTI) class, the revised guidelines recommend all of the fixed-dose dual-NRTI combination pill: tenofovir/emtricitabine (Truvada), AZT/3TC; Combivir), and abacavir/3TC (Epzicom).

Efavirenz (Sustiva/Stocrin) is still listed as the preferred non-nucleoside reverse transcriptase (NNRTI), with nevirapine (Viramune) suggested for special patient populations and for those intolerant of efavirnez. Only ritonavir-boosted PIs are suggested, including lopinavir/ritonavir (Kaletra), atazanavir/ritonavir, fosamprenavir/ritonavir (boosted Lexiva), and saquinavir/ritonavir (boosted Invirase). Fosamprenavir/ritonavir was added to the list in part due to the results of the KLEAN study presented at the IAC, and ritonavir-boosted indinavir (Crixivan) has been removed from the list.

Resistance Testing and Adherence

In other important antiretroviral therapy management updates, HIV resistance testing is now recommended for all antiretroviral-naive patients, given the rising prevalence of viral resistance in patients new to treatment, including those with established infection. In the setting of virological failure on the initial combination regimen, a resistance test also should be performed. Adherence counseling should be offered as the first step in modifying the combination, and the patient's regimen should be adjusted to include at least 2 active agents.

Virological Failure and Treatment Interruption

For patients who have experience treatment failure on multiple regimens, the updated IAS-USA guidelines suggest that the goal of a new antiretroviral combination should be to achieve a viral load below 50 copies/mL, given the potency of some of the newly FDA-approved PIs, including tipranavir (Aptivus) and darunavir (Prezista), and the potential of several experimental drugs available through expanded access programs, including etravirine (TMC125, a NNRTI), MK-0518 (an integrase inhibitor), and maraviroc (a CCR5 inhibitor), all of which were discussed at the conference.

Finally, the guidelines recommend against structured treatment interruption strategies, which were also extensively discussed at IAC. IAC COVERAGE OF SMART AND DART STI TRIALS:

Pregnancy and Antiretroviral Therapy

The IAS-USA guidelines also include some updates with respect to special patient populations as well. For pregnant women, AZT (Retrovir), 3TC (Epivir), and emtricitabine (Emtriva; FTC) are recommended. Tenofovir (Viread) should be used only if indicated by resistance testing, due to concerns about potential effects on bone formation in utero, even though the data surrounding this issue is derived from studies of monkey given extremely high doses of the drug.

Providers are warned about initiating nevirapine in pregnant women with CD4 counts above 250 cells/mm3 because of the potential for hepatotoxicity. Efavirenz is contraindicated in pregnancy due to the risk of birth defects, and nelfinavir (Viracept) is no longer a suggested as a preferred agent for pregnant women because of concerns about its concerns. For patients with hepatitis B virus (HBV) coinfection, tenofovir, 3TC, and emtricitabine are recommended, since these agents are also active against HBV.

The IAS-USA guidelines, as good and thorough as they are, fail to answer some central questions with which many clinicians must struggle every day, e.g., is it better to start with a ritonavir-boosted PI or a NNRTI, and which of the recommended NRTI fixed-dose combinations is optimal? Fortunately, data presented at the 16th IAC provide some guidance regarding the answers.

Boosted Fosamprenavir vs Lopinavir/ritonavir in Antiretroviral-Naive Patients (KLEAN)

Prior head-to-head studies of boosted PIs have been flawed due to the discrepancies in the patients enrolled or the regimens used, but in general, previous data have indicated that many boosted PIs have similar potency. This is the basis of the inclusive IAS-USA guidelines, which list as "recommended" PIs lopinavir/ritonavir, atazanavir/ritonavir, saquinavir/ritonavir, and (in the 2006 iteration) fosamprenavir/ritonavir. In contrast, the DHHS guidelines, citing a lack of comparative data, list only lopinavir/ritonavir as a "preferred" PI.

The KLEAN study (Kaletra BID vs Lexiva BID, both with Epivir and Abacavir QD, in Antiretroviral-Naive Patients) was designed to evaluate the efficacy and tolerability of fosamprenavir/ritonavir as compared with lopinavir/ritonavir. In this study, 878 treatment-naive subjects were randomly assigned to receive either fosamprenavir/ritonavir or lopinavir/ritonavir (soft-gel), with the fixed-dose formulation of abacavir/3TC as the NRTI backbone.

The primary endpoint was the proportion of patients with HIV RNA below 400 copies/mL at week 48 and the proportion that needed to discontinue their randomized treatment due to adverse events. At baseline, the two study arms were well matched, with a median HIV viral load of 5.07 log10 copies/mL and a median CD4 cell count 192 cells/mm3. Overall, 54% of the study subjects had HIV RNA above 100,000 copies/mL and 17% had a CD4 cell count below 50 cells/mm3.

The results at 48 weeks, using the time to loss of virological response (TLOVR) analysis, are shown in Table 1.

Table 1: 48-week Data from KLEAN Study



None of the differences between fosamprenavir/ritonavir and lopinavir/ritonavir were statistically significant. Further, response rates for fosamprenavir/ritonavir and lopinavir/ritonavir were similar in those patients with HIV RNA greater than 100,000 copy/mL, as well as across all CD4 count levels.

Three patients in the fosamprenavir/ritonavir group and 4 in the lopinavir/ritonavir group had virus with the M184V mutation. Although 2 patients in each arm developed PI-associated mutations upon virological rebound, these were not associated with phenotypic loss of susceptibility to either drug.

No differences between the study groups were observed in terms of treatment-related adverse events (including gastrointestinal symptoms) or in the rate of grade 3-4 laboratory abnormalities. In particular, fasting lipid levels (total cholesterol, LDL cholesterol, HDL cholesterol, and triglycerides) increased similarly in both treatment arms.

The results of this large, randomized study - also published in the August 5, 2006 edition of The Lancet - show clearly that fosamprenavir/ritonavir is non-inferior to lopinavir/ritonavir, with a remarkable overlap in treatment response and adverse events. As mentioned above, this study led to the IAS-USA committee to list fosamprenavir/ritonavir as a recommended PI in their 2006 revised guidelines. It will be interesting to see if the DHHS committee follows suit.

Efavirenz vs Lopinavir/ritonavir in Antiretroviral-Naive Patients (ACTG 5142)

One of the questions nagging clinicians for many years has been, "Should one prefer a boosted PI or efavirenz for antiretroviral-naive patients?" This question was answered to some degree in ACTG 5142, an important trial presented at the IAC (S Riddler and others, abstract THLB0204).

ACTG 5142 evaluated 3 different 2-class antiretroviral regimens in 753 treatment-naive patients:

lopinavir/ritonavir + 2 NRTIs;
efavirenz + 2 NRTIs;
lopinavir/ritonavir plus efavirenz.

The primary objectives of the study were to compare these regimens regarding the time to confirmed virological failure (defined as viral load > 200 copies/mL after 32 weeks) and the time to regimen completion (defined as virological failure or a toxicity-related discontinuation of any part of the regimen).

The 7553 subjects enrolled had a median CD4 count of 182 cells/mm3 and a median HIV RNA level of 100,000 copies/mL; they were followed for a median time period of 112 weeks. All used NRTI background regimens containing AZT/3TC (42%), tenofovir (34%), or d4T (stavudine, Zerit XR; 24%).

The major results of ACTG 5142 through 96 weeks using an intent-to-treat (ITT) analysis are shown in Table 2 with statistically significant differences (P < 0.016) noted.

Table 2: 96-week Data from ACTG 5142



These data demonstrate the superiority of efavirenz to lopinavir/ritonavir in terms of time to virological failure and suppression of HIV RNA to below 50 copies/mL, and a strong trend toward efavirenz also being superior to lopinavir/ritonavir regarding time to regimen completion (P = 0.02) and suppressing HIV RNA to below 200 copies/mL (P = 0.041). The efavirenz arm did not achieve quite as robust a CD4 count increase as the other two arms. However, it is unlikely that these small differences are clinically relevant.

Development of resistance was assessed in the patients who experienced virological failure; these results are shown in Table 3.

Table 3: Resistance Data in ACTG 5142

*P < 0.05 compared to lopinavir/ritonavir + efavirenz
†P < 0.05 compared to lopinavir/ritonavir
‡ P < 0.05 compared to efavirenz

The differences in resistance profiles are in line with those observed in prior studies, except for the higher rate of NNRTI resistance in the lopinavir/ritonavir + efavirenz arm, and it is unclear why this occurred.

Based upon the data from ACTG 5142, efavirenz + 2 NRTIs appears to produce higher rates of sustained viral suppression and to have less metabolic impact than lopinavir/ritonavir + 2 NRTIs; however, in those patients who experience virological failure, the risk of resistance appears to be higher among those treated with efavirenz.
Despite the resistance data, which are not unexpected and which are in line with those of prior studies, these results should certainly give more confidence regarding the use of efavirenz-based HAART as initial therapy. These data will challenge the roughly 40% of clinicians who initiate therapy with ritonavir-boosted PI regimens to reconsider their choice.

While the NRTI-sparing arm did surprisingly well overall, it is unlikely that this study will lead to a displacement of the role of NRTIs as part of first-line therapy, given the simplicity and low toxicity associated with some NRTI-containing regimens, especially those involving tenofovir/emtricitabine or abacavir/3TC.



Efficacy of Efavirenz in Patients with High Viral Load and Low CD4 Count (ACTG 5095)

Retrospective analyses of several efavirenz trials have indicated that efavirenz-based HAART is effective in patients with high HIV RNA or low CD4 cell count. Despite these data, some clinicians prefer using a ritonavir-boosted PI in such patients, since they view it as more potent and likely to lead to successful therapy.

In large part, this preference is based on some data supporting the efficacy of ritonavir-boosted PIs in patients with more advanced disease and the lack of definitive data establishing that the same is true of efavirenz. A sub-analysis of the ACTG 5095 trial now provides what appear to be definitive data demonstrating that efavirenz is effective across the spectrum of HIV RNA and CD4 cell levels.

As many will recall, initially ACTG 5095 randomized 1,147 antiretroviral-naive patients to receive either:

co-formulated abacavir + AZT + 3TC (Trizivir);
efavirenz + co-formulated abacavir + AZT + 3TC;
efavirenz + co-formulated AZT + 3TC (Combivir).

As previously reported, the triple-NRTI arm was stopped due to underperformance compared to the 2 efavirenz-containing arms, and the 2 efavirenz-containing arms performed equally well in the overall population, indicating that adding a third NRTI does not provide any clinical benefit.

In a follow-up study presented at this conference, efficacy of the 2 efavirenz-containing arms at 3 years was assessed across different levels of HIV RNA and CD4 cell counts in the 765 patients randomized to those arms (H Ribaudo and others, abstract THLB0211). All analyses used a Cox proportional hazard model and were based on an intent-to-treat model.

As shown in Figure 2, efavirenz was effective across all viral load and CD4 count strata. Further, there were no differences in response rates between the efavirenz arms regarding viral load (P = 0.043) or CD4 count (P = 0.099). Thus, these data make relatively clear that (1) efavirenz is highly effective regardless of baseline CD4 count or viral load; and (2) there is absolutely no benefit to adding a third NRTI to efavirenz-based HAART in antiretroviral-naive patients.

Figure 2: Efficacy of Efavirenz in ACTG 5095

Lipoatrophy: Worse with AZT than Tenofovir

The GS 934 study compared 2 regimens:

efavirenz + tenofovir + emtricitabine;
efavirenz + AZT + 3TC.

The 96-week data from GS 934 presented at the IAC indicated that HIV RNA suppression and CD4 count increases remained relatively consistent with the 48-week data, i.e., better results in the tenofovir/emtricitabine arm. The data regarding the risk of fat loss and development of lipoatrophy continued to emphasize the differences between tenofovir/emtricitabine and AZT/3TC.

While baseline DEXA scanning was not performed, a subset of 100 individuals (51 randomized to tenofovir/emtricitabine and 49 randomized to AZT/3TC), had scans performed at week 48 and again at week 96. At week 48, limb fat differed significantly between the two groups, with 7.4 kg in the tenofovir/emtricitabine group and 6 kg in the AZT/3TC group (P = 0.034). The follow-up scan at week 96 found that limb fat levels in the tenofovir/emtricitabine group rose to 8.1 kg, while in the AZT/3TC group they declined to 5.5 kg (P < 0.001). A further cross-sectional evaluation of a larger proportion of the study population at week 96 also found significant differences in limb fat between the two groups.

These data strongly suggest that limb fat mass is declining in the AZT/3TC-treated patients, but is unaffected in the tenofovir/emtricitabine group. Finally, consistent with the GS 903 study at 144 weeks, long-term follow-up in the GS 934 study does not suggest any association between tenofovir use and clinically significant deterioration in renal function.

CD4-Guided Treatment Interruptions: Not So SMART

The SMART study, which was designed to assess the wisdom of CD4-guided treatment interruptions, was initially presented at the 2006 Conference on Retroviruses and Opportunistic Infections. At that conference, it was reported that the study had been stopped early due to a significantly greater risk of opportunistic infections (OIs) or death in those who interrupted therapy (W El-Sadr and others, 13th CROI, 2006, abstract 106LB).

In a follow-up analysis presented at the IAC, Jens Lundgren and colleagues attempted to determine why patients in the interruption arm, who were to restart therapy whenever their CD4 counts fell below 250 cells/mm3 and spent very little time below 200 cells/mm3, did so poorly (J Lundgren and others, abstract WEAB0203).

The updated analysis evaluated CD4 counts and viral load and found that, in large part, these explained the differences between the two arms. Remarkably, CD4 count and viral load measures had similar overall impact on this adjustment, and CD4 percentage had a similar impact to that of CD4 count. Further, combining both CD4 count and viral load was more predictive of the risk than either marker alone, demonstrating that the risk of OIs or death is reflected by both lower CD4 counts and higher viral loads.

However, it should be noted that CD4 count and viral load markers did not explain all of the risk increase and there were other factors -- yet to be identified -- that also played some role in increasing the risk of OIs or death when interrupting therapy.

In another sub-analysis, Wafaa El-Sadr and colleagues reviewed this hazard ratio over a range of populations and subgroups in order to explore the applicability of the SMART findings (W El-Sadr and others, abstract WEAB0204) In short, they found that across a range of baseline demographic characteristics, the findings were similar. The only baseline characteristic that had a different outcome was baseline viremia, i.e., the risk of treatment interruption was most pronounced in patients who entered the study with a viral load below 400 copies/mL, which suggests that viremic patients, whether on or off antiretroviral therapy, carry a similar risk of adverse events.

For now, treatment interruptions should be avoided unless motivated by some significant need, such as serious antiretroviral toxicity. Further, if an interruption occurs, patients should be closely monitored and antiretroviral therapy preferentially restarted at higher CD4 count levels (about 350 cells/mm3) than were used in the SMART study.


Conclusion

The 16th IAC provided a wealth of new and important scientific information for researchers, clinicians, government officials, patient advocates, and patients to consider. While the studies discussed above represent some of the most important clinical data presented, there were a number of other important studies that showed:

Antiretroviral therapy can be highly effective in developing countries and lead to significant improvements in morbidity and mortality (A Khera, abstract WEPE0096; P Munderi, abstract THLB0208).

In Europe, transmission of antiretroviral resistance appears to be stable, if not declining (A M Wensing, abstract TUAB0101).

Testing for the HLA-B*570 genotype is specific (95%-98%) but not sensitive (8%-57%) for risk of abacavir hypersensitivity, especially in Black patients, where sensitivity is only 8%. Thus, HLA-B*5701 may be effective at screening patients out (i.e., those who are HLA-B*5701 positive should not be treated with abacavir), but cannot be used to screen patients in (i.e., those who are HLA-B*5701 negative may still be at risk for abacavir hypersensitivity (M Mosteller, abstract WEPE0171).

Lopinavir/ritonavir monotherapy is relatively effective as an initial and maintenance strategy; however, lipid issues persist and more PI resistance may develop than when lopinavir/ritonavir is combined with 2 NRTIs (J-F Delfraissy, abstract THLB0202; W Cameron, abstract THLB0201; J Arribas, abstract THLB0203).

Darunavir (TMC114) is highly effective and safe through 48 weeks of therapy (A Lazzarin, abstract TUAB0104).

CCR5 antagonists have limited impact in antiretroviral-experienced patients who have both R5 and X4 tropic virus, but can lead to significant therapeutic benefits in those with only R5 tropic HIV (H Mayer, abstract THLB0215; R Gulick, abstract THLB0217).

MK-0518, the new Merck integrase inhibitor, may be as effective as efavirenz in suppressing HIV RNA in antiretroviral-naive patients, and may lead to a more rapid decline in viremia after therapy initiation (M Markowitz, abstract THLB0214).

In HIV-HCV coinfected patients, the chance of sustained virological response declines as fibrosis increases, indicting that earlier therapy may be beneficial in improving cure rates (J Sasadeusz, abstract WEPE0040

In summary, this conference provided a great deal of encouraging data regarding therapeutic options in developed countries and highlighted the need for us to get these options to patients who are desperately awaiting them in developing countries. Now it is truly "Time to Deliver"!

09/22/06

References
Unless otherwise noted, all references in the text are to the Program and Abstracts of the 16th International AIDS Conference. Toronto, Canada. August 13-18, 2006.