Bevirimat: good results at 400 mg liquid dose, but formulation problems persist

Phase II trials of the HIV maturation inhibitor bevirimat continue to show good short-term safety and, at adequate doses in patients without key pre-existing HIV mutations, good antiviral efficacy. Results were presented Sunday by Dr. Jacob Lalezari at the 48^th Interscience Conference on Antimicrobial Agents and Chemotherapy in Washington.

Bevirimat is an investigational maturation inhibitor (NRTI) being developed by Panacos Pharmaceuticals. It targets a late stage in the HIV life cycle, as new viruses are being prepared for release from infected cells. The drug interferes with the maturation of these viruses. A previous 10-day Phase II study found potent anti-HIV effect and encouraging tolerability in small groups of treatment-experienced patients.

Development has been slowed by formulation and bioavailability issues. Bevirimat response appears to be highly dose-dependent; a 400 mg tablet resulted in unexpectedly low plasma levels, and earlier studies based on this formulation showed disappointing results.

The previous phase II trial, evaluating 200, 250 and 300 mg once-daily doses of oral solution, had good results at the highest dose, but responses were still uneven. Analyses also identified key polymorphisms (naturally occurring genetic variations) in the HIV Gag gene (Q369, V370 and T371) which resulted in poorer responses.

In this study, bevirimat Study 203, investigators opted to "push the dose" further in the hopes of better responses. The study consisted of two distinct phases: the original trial was a phase 2 double-blind, randomised dose escalation study, and the second phase focused on resistance.

In the first phase, 59 treatment-experienced patients with >1 primary resistance mutation and viral loads of at least 2000 copies/mL received two weeks of bevirimat (N=46) or placebo (N=13) on top of a failing background regimen. Patients initially received either a 400mg bevirimat tablet dose, a 250, 300, 350 or 400mg once-daily liquid dose, or placebo.

The relatively long half-life of bevirimat (50 – 60 hours) results in a slow increase in plasma concentrations over the first week, taking seven to ten days to reach steady-state levels. As previously, in this trial the 400 mg tablet yielded low plasma trough concentrations (approx. 20 µg/ml after seven days). Higher concentrations (approx. 35 µg/ml after seven days) were seen with the 250 mg liquid dose. The 300, 350 and 450 mg liquid doses yielded essentially similar trough concentrations (approx. 50 µg/ml after seven days), with no additional benefit at doses above 300 mg.

Modest viral load reductions were seen the original study group (in which bevirimat functionally served as monotherapy). After two weeks, the mean changes in viral load were -0.1 log copies/mL for the 13 patients given placebo, -0.36 log for the 12 on 400 mg tablets, -0.63 log for the 9 on 250 mg liquid, -1.02 log for the 8 on 300 mg liquid, -0.62 log for the 9 on 350 mg liquid, and -0.44 log for the 6 on 400 mg liquid.

Analysis found that response was not only dose-dependent but highly dependent on the key Gag resistance polymorphisms. Viral load declined (at week two) by only 0.21 log in those with any of the polymorphisms at 369, 370 or 371, but by 1.08 log in those without.

This retrospective analysis led to the second phase of the study, in which patients (treatment-experienced and –naïve) were prospectively screened for these key Gag polymorphisms. All received a 300 mg liquid dose. After two weeks, the 8 treatment-experienced patients with polymorphisms had a 0.65 log viral load decline; the 10 treatment-experienced patients without polymorphisms had a 1.10 log viral load decline, and the 9 treatment-naïve patients had a 1.04 log decline.

A screening of external data from 1034 patients showed that, overall, roughly 62% of patients show key Gag polymorphisms. These polymorphisms do not appear to be more common in treatment-experienced patients, although Lalezari commented that they may be more common in clade C virus. They tend to occur separately rather than together, with Q369 being the most detrimental.

As far as side-effects, 69% of bevirimat-treated patients and 65% placebo treated patients had >1 adverse event (AE). For BVM and placebo treated patients respectively, the most common AEs were diarrhea (20%; 38%), nausea (15%; 17%) and headache (23%; 9%); all were Grade 1. Abnormal dreams were reported by 7% of the bevirimat patients.

In conclusion, bevirimat appears to have a significant, dose-dependent anti-HIV effect. Viral load declines over 1 log are achievable at adequate trough concentrations in patients without pre-existing polymorphisms in the HIV Gag gene. Short-term safety appears comparable to placebo. Further trials will be based on a tablet formulation if it yields adequate drug levels.