January 1997
Dr. Christine Katlama from the Hopital Pitie Salpetriere in Paris presented a late breaker (Abstract LB4) on two pilot studies of D4T plus 3TC (Altis 1 and Altis 2). Altis 1 involved antiviral naive patients with a median CD4 count of 258 and median viral load at baseline of 76,502 copies (about 4.8 log 10). The maximal viral load response was about 2.0 log and at 24 weeks, there was a sustained reduction of 1.66 log accompanied by a CD4 rise of about 100 cells. In Altis 2, the patients were heavily pretreated with AZT, ddI , and ddC, but had to be naive to d4T and 3TC. These patients had slightly more advanced disease with baseline median CD4 count of 172 and median viral load at baseline of 91,000 copies (about 4.9 log 10). The responses to D4T plus 3TC were more modest in this group with a maximal viral load decrease of 1.4 and CD4 increase of about 50 cells. By week 24, the viral load decrease was about 0.5 log below baseline. Among the naive patients, 21% had undetectable viral loads at week 24 (using the standard Amplicor assay). Toxicity was limited and only 1 patient discontinued D4T due to neuropathy. These results were quite comparable to results from the 4 pivotal trials for AZT plus 3TC, presented as late breakers at this meeting 2 years ago.
A head to head comparison of AZT plus 3TC with D4T plus 3TC focusing on the antiviral effect and drug levels in the cerebrospinal fluid was reported in another late breaker presented by Dr. Foudraine from the University of Amsterdam (Abstract LB5). They randomized 31 drug naive patients to begin either D4T plus 3TC or AZT plus 3TC and performed lumbar punctures at baseline and at week 12. The change in plasma viral load in both groups was identical. Fifteen patients had lumbar punctures; all had detectable virus in their CSF at baseline. By week 12, all had undetectable viral RNA and negative cultures. Drug levels in the CSF were 41, 56, and 67ng/mL for AZT, D4T, and 3TC respectively. Absolute drug levels in the CSF relative to the concentration needed for antiviral activity may be more telling than blood to CSF ratios, which are commonly reported. When one drug has rapidly changing blood levels over the dosing interval (as in the case of AZT) the reported ratio may be very dependent on the timing of specimens. The Amsterdam group speculated that both D4T plus 3TC and AZT plus 3TC would be expected to protect against HIV-related dementia. The effects of protease inhibitors on CSF viremia are less clear, and one consideration in the design of triple combination regimens may be the inclusion of drugs with good activity in this compartment
Several posters also confirmed the activity of D4T plus 3TC. A pilot open label trial of 41 patients performed in at the BC Center for Excellence in Vancouver yielded similar results (Abstract #557). Importantly, viral load reductions were greatest among those who had not received either drug and were least among patients who had been treated with 3TC. Cal Cohen from Community Research Initiative of New England presented a retrospective study of 330 patients (abstract # 556). The median initial viral load response was 1.1 log, and as in the Vancouver study, patients who had already received AZT and 3TC had a significantly reduced response.
Each of these studies has at least one major limitation, including small sample size, lack of randomization, lack of concurrent controls, or retrospective data collection, but taken together they paint a consistent picture.