Disappointing amprenavir results: only 30% undetectable at 48 weeks

Glaxo Wellcome is reported to be reconsidering its marketing strategy for amprenavir, the company's protease inhibitor, following the announcement of disappointing results from a randomised, open-label, Phase III study, PROAB3006. 48 week data were recently communicated to trial investigators.

PROAB3006 compared amprenavir with indinavir in a group of 504 nucleoside analogue experienced, PI naïve patients. At entry, median viral load was around 8,000 copies, with the median CD4 count around the 400 mark. The study protocol encouraged participants to switch one of their nucleoside analogues at the same time as starting their PI, and 73% did so. d4T/3TC and d4T/ddI were the most common nucleoside backbones in participants’ regimens.

At 48 weeks, by intent-to-treat analysis, 30% of amprenavir recipients had viral load below 400 copies, compared to 46% in the indinavir arm, a statistically significant difference. These relatively poor results appear to have been influenced by the fairly high number of people who discontinued their allocated study medication – whether because of adverse events, clinical progression, withdrawal of consent, or loss to follow-up. By week 48, discontinuations were 44% in the amprenavir arm and 36% in the indinavir arm, and the proportions stopping due to adverse events were 18% and 15% respectively. GI symptoms and rash were the most common toxicity-related reasons for discontinuation in the amprenavir arm, whilst indinavir recipients were more likely to stop because of urological symptoms, which tended to occur later.

Amprenavir, which is currently unlicensed within the European Union, has been associated with a relatively high drop-out rate in another Glaxo Wellcome study, PROAB3001, which was presented at the recent 39th ICAAC. Though dosed twice daily, the drug’s pill burden of eight large capsules is one of the highest of currently available antiretrovirals.

Whilst these factors count against amprenavir’s use in first-line therapy, its resistance profile may allow its use as a second-line PI, though the efficacy of such a strategy has not been proven as yet. Similarly, amprenavir has so far been less strongly associated with the lipodystrophy syndrome than other PIs. Whether this may simply be a factor of its later development is unclear at present. Abnormal fat redistribution syndrome was reported more frequently in those in the indinavir arm of the PROAB3006 study than the amprenavir arm, (28 events versus eight, a significant difference). However, though toxicity data were gathered prospectively, the study was not designed to measure the relative risk of lipodystrophy between the two regimens. It is possible that differences may be the result of a bias in investigator reporting.