Late Breaker--Amprenavir (141W94) + AZT/3TC: 16 week data in treatment-naive

This study was presented at ICAAC in the late breaker session. The data from this study is the first phase III study data available on amprenavir (APV). This randomized, double-blinded, placebo controlled study is being submitted by Glaxo Wellcome to the FDA as part of their New Drug Application (NDA) to receive accelerated approval. The application is expected to be submitted by November and if all goes well accelerated approval should be forthcoming in early '99. This study compares APV+AZT/3TC to AZT/3TC.

An expanded access program for amprenavir was announced two weeks ago by Glaxo Wellcome. An article is posted to the web site describing the program and giving the 800 telephone # (click here to link to the article). There is interesting and potentially important preliminary PK data which you should read that is also posted in an article to the web site (click here to link to the article). In preliminary PK studies: efavirenz reduces APV blood levels; IDV reduces APV blood levels, APV increases IDV blood levels; APV raises NFV Cmin appreciably; APV raises AZT blood levels 20-30%.

An APV double protease study was started and early data was reported at Retrovirus '98. The May issue of our newsletter, NATAP Reports, reported the preliminary 16 week data from this study--APV+IDV 4/6 <20 copies/ml; APV+NFV 3/6 <20 copies/ml; APV+SQV 2/5 <20 copies/ml; APV+AZT/3TC 2/3 <20 copies/ml. Obviously, a small amount of data. 48 week data is expected to be reported this Fall and NATAP will report it to you.

Dr Jeff Goodgame, of the Central Florida Research Initiative, reported the data on this study at ICAAC. He said --

• APV has a long half-life of 10 hours which supports bid dosing (I have seen the PK profile and it looks good. Drug levels stay elevated for a prolonged period lending itself to adherence flexibility)
• no food or hydration requirements
• he said there was a unique resistance profile, the main resistance mutation being I50V. However, cross-resistance to other PIs have yet to be clinically characterized. The key to cross-resistance is secondary mutations as well as primary mutations. APV may have secondary mutations in common with other PIs. ACTG 398 and 400 are about to start. They will explore using APV as salvage therapy for nelfinavir failures or failure of other protease inhibitors

At ICAAC it was reported by Drusano and others that abacavir has a synergy with APV, in vitro, that abacavir does not have with other PIs (abstract I-21, session 49-I).

Baseline Characteristics-- median baseline viral load was 4.61 log (40,738 copies/ml--range 4,047-1,230,269) in the APV+AZT/3TC group; in AZT/3TC control arm, the baseline median viral load was 4.74 log (54,954 copies/ml--range 1,148-2,041,738). Baseline CD4 was 435 in APV group and 409 in AZT/3TC alone group.

As with studies for all new drugs, there were 3 statistical analysis conducted:

Intent-To-Treat (ITT)--there were two types of ITT analysis. The first was a strict ITT in which all subjects who received drug in study are included. This includes people who discontinued for any reason. They are considered failures. They also conducted an ITT as-observed analysis which they described as including all individuals for whom they have data and excludes individuals for whom they do not have data. So, this analysis is a little less inclusive and the percent undetectable should be higher.

As-Treated-- this analysis includes people on drug only. Goodgame said in this analysis protocol violaters were excluded, specifically those who had been off drug for more than 28 days consecutively. This analysis includes people who failed but are still receiving drug in the study.

The safety analysis includes everyone who has received at least one dose of medication.

Viral Load Changes

The percent of patients with viral load <400 copies/ml at week 16--

• 88% using the As Treated analysis (n=164)
• 82% using the ITT: Observed Data analysis (n=186)
• 59% using the ITT analysis which includes all patients even if there is no observed data where all discontinuations were considered failures whether they actually failed or just decided not to continue the study (n=221)

The percent of patients with viral load <50 copies/ml at week 16--

• 66% had <50 copies/ml using the ITT analysis in which all discontinuations were considered failures but individuals were included if they dropped off study and still remained undetectable. Goodgame called this a Last Observation Carried Forward analysis.
• 59% had <50 copies/ml using a more stringent ITT Observed Data analysis (n=186). This analysis was more stringent because it did not include those who remained undetectable although they withdrew from the study.

Percent of Subjects with Viral Load <400 copies (As Treated analysis) by Baseline Viral Load

	APV+AZT/3TC	AZT/3TC
10,000-30,000 copies/ml	96% (25/26)	52% (13/25)
30,000-100,000 copies/ml	88% (30/34)	10% (4/42)
>100,000 copies/ml	71% (10/14)	0% (0/23)

Drug Related Adverse Events (Incidence in parenthesis)--

Nausea (70% in APV grp, 42% in AZT/3TC grp, 10% disct in APV grp due to nausea)

Flatulence (30% in APV grp, 40% in AZT/3TC grp)

fatigue (about 30% in each grp)

Headache (21% in APV, 27% in AZT/3TC)

Vomiting (21% vs 11%)

Diarrhea (20% vs 15%)

Oral Parasthesia (25% vs 2%)

Rash (20% vs 5%)

AZT related side effects may be increased because APV increases AZT blood levels.

Goodgame reported that the vast majority of adverse events required no change in study meds. Study meds were most often interrupted or stopped permanently because of grade 1 or 2 nausea or vomiting. Rash was rarely therapy limiting and most subjects rechallenged successfully. In terms of lab findings, he reported there were no remarkable changes or differences seen between groups and very few grade 3 or 4 lab toxicities were reported.