Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

14th Conference on Retroviruses and Opportunistic Infections

Los Angeles, California / February 25-28, 2007

There may be no single optimal combination in the control of HIV, but a variety of data presented here at CROI has provided new guidance for evaluating the relative strengths and weaknesses of currently available antiretroviral drugs in first- or second-line regimens. It is now well accepted that the potency of the individual agents in the currently available fixed-dose combinations is retained when combined in a single pill, but other issues, such as relative long-term safety, remain important concerns.

Examining Renal Function

Several presentations here devoted to the risk of renal toxicity with tenofovir provide an example of where the clinical science appears to be heading.

We hypothesized that tenofovir-treated patients taking a [ritonavir-boosted protease inhibitor] PI/r would have greater declines in renal function than patients taking tenofovir with a non-nucleoside reverse transcriptase inhibitor [NNRTI],” reported Dr. Miguel Goicoechea, University of California, San Diego. “This is exactly what we found. Although both groups had a decline in [glomerular filtration rate] GFR, the decline was twice as great in those tenofovir patients receiving a PI/r than those receiving a NNRTI (13.9 vs. 6.2 mL/min/year; P=0.033).”

Renal toxicity is a characteristic adverse event of tenofovir that must be considered when selecting nucleoside reverse transcriptase inhibitors (NRTIs) for backbone therapy, whether it is prescribed independently or in a fixed-dose combination with emtricitabine, but the evidence that the risk may be exacerbated by co-administration with a PI/r informs treatment choices. Other NRTIs also have characteristic adverse events, such as the 5% incidence of hypersensitivity associated with abacavir, but there is limited evidence that these are exacerbated by other drugs in the regimen. Such considerations will be important as fixed-dose combinations, including triple therapy combinations, continue to be introduced into routine therapy.

In the study evaluating renal toxicity on tenofovir in the context of a PI/r vs. a NNRTI, 147 patients were evaluated. Of these, 51 received tenofovir plus a PI/r, 29 received tenofovir plus a NNRTI, and 67 received regimens without tenofovir. The patients were drawn from a larger therapeutic drug monitoring study known as CCTG 578. GFR, as estimated by both Cockcroft-Gault and Modification of Diet in Renal Disease (MDRD) calculations, was similar in the three groups at baseline. The patients were evaluated over 48 weeks.

Over time, there was a decline in GFR in all groups, including the group that did not receive tenofovir. The degree of decline did not differ between the group receiving tenofovir plus a NNRTI and the group that did not receive tenofovir. When a variety of factors with potential relevance to renal function were evaluated, only age was associated with a significant decline in GFR (-1.2 mL/min/year; P=0.001). CD4+ cell count, viral load or previous antiretroviral therapy (vs. no previous treatment) were not associated with GFR decline.

When the groups were compared, patients receiving a PI/r did have a higher plasma exposure to tenofovir, but the authors of this multicentre study reported that pharmacokinetic parameters did not predict change in renal function over time. It is not clear from this study whether it is an interaction with ritonavir, which is well known to produce renal impairment, which exacerbates the risk of renal toxicity associated with tenofovir. However, a second study looking at the same issue did implicate ritonavir specifically.

“In unadjusted analyses, tenofovir/r was associated with increased impairment of renal function as measured with MDRD relative to no tenofovir,” reported Dr. Homan Wai, who is also affiliated with the University of California, San Diego. “In models adjusted for body mass index, baseline CD4, baseline viral load and baseline MDRD clearance, tenofovir and ritonavir remained a significant predictor of impaired renal function.”

In this retrospective study of 635 patients, three groups were assessed: those who did not receive tenofovir (380 patients); those who received tenofovir/r (184 patients); and those who received tenofovir but not ritonavir (71 patients). Two definitions of renal impairment were employed, a reduction in MDRD of 30 cc/min and a reduction in MDRD of 25% from baseline. In the adjusted analyses, the hazard ratio (HR) among patients receiving both tenofovir and ritonavir for reaching the first definition of impairment was 1.9 (P=0.001). The HR for reaching the second definition was 1.8 (P<0.0001).

However, not all of the data evaluating the risk of renal toxicity of tenofovir were consistent. In a retrospective analysis of data from 565 patients, tenofovir was associated with renal impairment as measured by GFR, but the long-term decline was characterized as modest. According to senior author Dr. Richard Moore, Associate Professor of Medicine, Johns Hopkins University, Baltimore, Maryland, the decline in GFR on tenofovir was confined to previously treatment-experienced patients and most often observed during the first six months of treatment. Other factors associated with a reduction in GFR in this series included a CD4+ cell count <50 cells/mm3, diabetes, and age greater than 45.

In the Swiss HIV Cohort Study, progressive renal impairment as measured with GFR was identified in all four treatment groups evaluated, of which two received tenofovir and two did not. However, the decrease in GFR “was significantly more frequent in patients treated with tenofovir than either of the comparators,” reported Dr. Christoph A. Fox, University Hospital, Berne, Switzerland.

In the Swiss study, the patients were divided into those who were treatment-naive initiating HAART with tenofovir (100 patients); treatment-naive initiating HAART without tenofovir (357 patients); experienced patients reinitiating HAART after one year off therapy with a regimen containing tenofovir (113 patients); and the same type of experienced patients starting a regimen without tenofovir (137 patients). Change in renal function was evaluated with both Cockcroft-Gault and MDRD equations. According to Dr. Fox, patients treated with tenofovir therapy when evaluated in a multiple Cox regression analysis “had a significant risk to reach both endpoints of renal impairment with a HR of 2.63 [95% CI, 1.87-3.69].”

Considerations for Future Strategies

The issue of how antiretroviral therapies may differ for long-term outcomes was also brought into focus by another set of data from the Swiss HIV Cohort Study. In this analysis, which also returns to the persistent question of whether it is best to start HAART containing a PI/r or a NNRTI, 1155 patients who were started on HAART containing a PI/r or a NNRTI but had at least one treatment failure were evaluated for drug class resistance. According to the results, the two types of HAART regimens were similarly potent at baseline but generated a different risk for subsequent resistance. Specifically, there was significantly less resistance after virologic failure with PI/r- than NNRTI-based HAART.

“These data emphasize that in addition to drug potency, the genetic barriers produced by virologic failure should also be considered when choosing first-line therapy,” stated Dr. Victor von Wyl, University Hospital Zurich, Switzerland.

In this study, 42 (5.9%) of 711 patients started on a NNRTI and 22 (5.0%) of 444 patients started on a PI/r experienced virologic failure. Mutations as defined by the International AIDS Society were found in four (33.3%) of the 12 patients receiving a PI/r vs. 22 (73.3%) of the 30 patients receiving a NNRTI. According to Dr. von Wyl, the difference between the PI/r and NNRTI failures “remained statistically significant when this analysis was repeated using Stanford genotypic sensitivity scores with varying cut-offs.” Overall, 67% of the NNRTI patients lost two or more classes upon first-line failure vs. only 8.3% of the PI/r group.

Summary

The definition of optimal HAART continues to evolve as new strategies, such as fixed-dose combinations, are evaluated in the context of long-term prognosis defined by safety and preservation of future options in the case of failure. Current fixed-dose combinations of NRTIs such as abacavir/lamivudine and tenofovir/emtricitabine are marked by a high degree of safety but may not be interchangeable in combination with PI/r- vs. NNRTI-based HAART. Although current HAART regimens are potent, the issue of which fixed-dose combination provides the greatest likelihood of long-term HIV control continues to be debated.