• Long-Term Follow-Up of Antiretroviral-Naive HIV-Infected Patients Treated with d4T and ddI (qd) and Nevirapine (bid or qd) in the VIRGO Study (Abstract 53)
  Authored by V. Reliquet, V. Ferre, M. Audrain, et al.

The use of nevirapine (NVP) once daily has never been approved by the FDA, still many clinicians dose and many patients use NVP in this fashion. The rationale for once-daily dosing of NVP is based upon the long half-life of NVP and several studies, including the Atlantic trial, which have had moderately successful results using the once-daily dosing of NVP. Still, the data are not clear that once-daily dosing of NVP is fully equivalent to twice-daily dosing.

The VIRGO trial was an open-label, multicenter study (12 centers) which evaluated the efficacy of stavudine (d4T) and didanosine (ddI) and nevirapine (NVP) for the 12 months between November 1997 to October 1998. The patients were all HIV-1 infected, antiretroviral-naive with CD4 cell counts >200 cells/mm³ and plasma HIV RNA >5,000 copies/mL. The mean CD4+ T cell count and viral load of the enrolled patients were 432 cells/mm³ and 4.70 log10 copies/mL. The antiretroviral regimen consisted of d4T 30mg bid (patients <60 kg) or 40mg bid (patients >60 kg), ddI 300mg qd (<60 kg) or 400mg qd (>60 kg) and NVP 200mg qd for the first 14 days then 200mg bid (60 patients-VIRGO I) or 400mg qd (40 patients-VIRGO II). This is a follow-up of those patients still receiving therapy at 36 months. The follow-up consisted of clinical examination every three months with CD4+ T cell counts and HIV RNA plasma viral loads.

The original data presented at 12 months was based on 45 patients from VIRGO I (of 60 patients enrolled) and 22 patients from VIRGO II (of 40 patients enrolled). There are no details as to the reason for the loss of the other patients. Of those still remaining in the trial at 12 months, 87% in VIRGO I and 77% in VIRGO II had HIV-1 RNA < 500 copies/mL; 67% and 59%, respectively, had <50 copies/mL. At 36 months, 34 patients remained in VIRGO I, although data is only provided for 30 patients (no explanation for the other four patients) and 14 remained in VIRGO II with data presented for 13 patients (missing data on one patient).

The authors report there were three serious adverse events between 12 months and 36 months of treatment: one unexplained death at 36 months (unclear if a participant in VIRGO I or II), cytolytic hepatitis at 21 months (VIRGO I) and hyperlactatemia with peripheral neuropathy and diabetes at month 17 (VIRGO I). They also report that 9 patients (15%) experienced lipoatrophy.

At 36 months, 27 of the 30 patients (90%) in VIRGO I (again no explanation as to the exclusion of the data on the other four patients still remaining in the study) and 12 of 13 patients (92%) in VIRGO II had HIV RNA <500 copies/mL. No data on number of patients <50 copies/mL were presented. There was a mean change from baseline in CD4+ T cell count of 326 cells/mm³ in VIRGO I (now based on all 34 patients) and 275 cells/mm³ in VIRGO II (13 patients) at 36 months.

The authors conclude that therapy with "d4T and ddI and NVP shows a sustained and durable activity with a gain of CD4 cell counts still increasing beyond 12 months of treatment." Furthermore, they state "virological failure (plasma viral load >5,000 copies/mL) occurred in 12 patients, half of them had a poor observance," but do not report the raw data on these patients or how they determined that they "had a poor observance." Finally, they also conclude "long-term acceptability and safety are good with a low rate of fat redistribution syndrome."

While on the surface these numbers appear relatively good, the authors do not mention the rather horrendous intent-to-treat numbers, which convert to 45% for VIRGO I and 30% for VIRGO II at the end of three years of treatment, nor do they mention that at least some of the patients lost during the three years of this trial were lost due to virologic failure.

So, as in many studies, the on-treatment analysis presented by the authors, which is frequently the most favorable analysis, must be viewed with some degree of skepticism, especially when there is such a marked divergence from the intent-to-treat analysis. The bottom line for this study is that VIRGO I and VIRGO II are relatively similar regarding their on treatment results at the end of 36 weeks; however, the small number of patients left and the huge drop-out rate makes it very difficult to state with any certainty that the regimens are in fact equivalent in effectiveness.