February 11, 2003
At the 10th Conference on Retroviruses and Opportunistic Infections, known affectionately as "The Retro Conference" by those in the business, updated results from the largest comparative study of the use of tenofovir (TDF, Viread) in antiretroviral therapy-naive patients were presented. These data indicate that a regimen consisting of TDF + 3TC (lamivudine, Epivir) + efavirenz (EFV, Sustiva) was as potent as d4T (stavudine, Zerit) + 3TC + EFV in suppressing HIV viral load almost two years after the start of the study. However, while details on metabolic toxicities were provided, something was missing, if not amiss.
Since its approval by the FDA, TDF has become an antiretroviral success story. The drug has become increasingly popular and is used in both salvage and initial antiretroviral regimens. It clearly is the little nucleotide that could. Approval of TDF was based on studies conducted in treatment-experienced patients. To determine how well the drug performed in the antiretroviral-naive patient, Gilead Sciences, the manufacturer of the drug, launched a large clinical trial comparing TDF + 3TC + EFV versus a leading initial antiretroviral regimen, d4T + 3TC + EFV in 600 HIV-1 infected patients. The study was conducted in a double-blind fashion, meaning neither subjects nor study staff knew which regimen participants were taking.
Initial data from the first 48 weeks of on-study follow-up were presented in Barcelona at the World AIDS Conference. These results demonstrated that both regimens led to extremely impressive virologic suppression, with over 80 percent of subjects in each arm achieving a viral load below 50 copies/mL a year after starting the study. The only differences observed between the TDF and d4T study regimens after a year of treatment were in toxicity. In addition to the expected excess cases of peripheral neuropathy in the d4T arm, significantly higher triglycerides, total cholesterol and LDL cholesterol were also seen in this arm compared to the TFG-assigned group.
Data presented here described the virologic and safety profiles of the study arms through 96 weeks of study participation. At baseline, the median HIV viral load was 81,300 copies/mL in both the TDF + 3TC + EFV and the d4T + 3TC + EFV arms. The median entry CD4 cell count in the TDF group was 276/mm3 and in the d4T arm was 282/mm3. Only about 15 percent of the subjects in each study arm discontinued the study prematurely and for similar reasons (toxicity, lost to follow-up, non-adherence, etc.).
The trends in viral load suppression seen at 48 weeks persisted at week 96. Of the TDF-assigned subjects, 78 percent had an HIV RNA PCR level less than 50 copies/mL, which was not significantly different from the 74 percent of the d4T-treated subjects who also achieved an undetectable viral load. From week 48 to week 96, CD4 cell counts continued to increase in both study arms up from baseline by 266/mm3 in the d4T group and by 261/mm3 in the TDF group.
In terms of toxicity, equal proportions of subjects in each arm experienced moderate to severe clinical adverse events (about 23 percent in each arm). Likewise, overall major laboratory abnormalities were also basically identical across the study regimens (34 percent for TDF, 39 percent for d4T) -- but this is where the similarities ended. As in the earlier analysis, the changes in lipids during the study were higher in the d4T + 3TC + EFV arm, only more so. Fasting triglyceride levels increased by about 100 mg/dL in the d4T arm compared to approximately 5 mg/dL in the TDF arm (p<0.001). Significant differences in elevations of total cholesterol were also seen, with d4T-assigned subjects seeing a 50 mg/dL increase in total cholesterol versus 30 mg/dL among the tenofovir subjects. LDL cholesterol (the "bad" cholesterol) rose in both study arms, but again more so among those randomized to d4T. HDL (the "good" cholesterol) also went up in both arms but again favored the TDF arm.
In an attempt to gauge the clinical significance of these lipid changes, they conducted an analysis of when a lipid-lowering statin or fibrate medication was started. It was demonstrated that 10 percent of the d4T-assigned subjects received a lipid-lowering drug by week 96 of the study compared to only 2 percent of the TDF subjects (p<0.001). These data indicate that the on-study lipid elevations seen in the d4T arm was of enough concern to clinicians that they felt compelled to initiate medication to reduce cholesterol and/or triglycerides.
At 96 weeks there remained more peripheral neuropathy and investigator-defined lipodystrophy in the d4T arm. To better evaluate body shape changes, full body DEXA scans were performed at week 96. These demonstrated a difference in limb fat between the study arms that again spelled bad news for d4T. There was approximately eight pounds less limb fat observed in the d4T versus the TDF-assigned subjects (p<0.001). Baseline full body DEXA scans were not performed. However, DEXA scans of the hip and spine to measure bone density were done at baseline and during the study as mandated by the Food and Drug Administration (FDA) due to evidence of TDF-induced bone disorders seen in animal studies. Unfortunately, these data were not included in this public presentation despite their importance and the conspicuousness of their absence. Sooner or later these bone mineralization data will be made widely available, perhaps even at this conference, although they certainly belonged in this poster presentation. When they are, an update to this report will follow.
TDF is a potent agent that is becoming increasingly popular among patients and their clinicians. The manufacturers of TDF would do well to remember our sad love affair with another popular agent that also began warmly but ended bitterly in the denials and obfuscations of the drug's toxicity by its makers. That drug? You guessed it, d4T.