 |
|
|
| |
Are Week-12 Results Adequate for Phase 2 Integrase Inhibitor Trials?
|
| |
| |
ICAAC 2015, September 17-21, 2015, San Diego
Mark Mascolini
Comparison of week 12 and week 24 results of two phase 3 trials showed little virologic change after week 12 in the integrase inhibitor arms but substantial changes in the comparison arms, which tested efavirenz or a boosted protease inhibitor (PI) [1]. Gilead Sciences investigators who conducted the analysis suggested that week 12 data "may be a reliable surrogate for week 24 outcomes in patients receiving [integrase inhibitor] agents."
Phase 2 trials of new antiretrovirals usually use 24-week data for primary efficacy and safety endpoints. Because integrase inhibitors suppress viremia faster than previous antiretroviral classes, Gilead investigators asked whether week 12 data would provide adequate endpoints in phase 2 trials, which aim to establish an active and safe dose for phase 3 study.
To address this question, the Gilead team analyzed week 12 and 24 results of two phase 3 trials of Stribild, the coformulation of the integrase inhibitor elvitegravir plus cobicistat, tenofovir (TDF), and emtricitabine (FTC). The comparison arms in these double-blind, active-controlled trials assessed Atripla (coformulated efavirenz, TDF, and FTC) and atazanavir/ritonavir plus TDF/FTC. Participants entered both studies naive to antiretrovirals, with any CD4 count, and with an estimated glomerular filtration rate at or above 70 mL/min.
While 701 people got randomized to Stribild in the two trials, 352 received Atripla and 355 received atazanavir/ritonavir. Age averaged 38 or 39 in each of the four study arms, about 90% of participants were men, and 22% to 39% across the four arms were nonwhite. About one third of all study participants had a pretreatment viral load above 100,000 copies.
After 12 weeks of treatment, 89% in the Stribild (integrase inhibitor) arm had a viral load below 50 copies, compared with 71% in the Atripla (nonnucleoside) arm and 61% in the atazanavir/ritonavir (PI) arm. At week 48 respective proportions with a sub-50-copy viral load were 92%, 93%, and 92%. That meant that people randomized to the integrase inhibitor averaged only a 2.8% viral load change from week 12 to week 24, compared with a 22.0% change in the Atripla arm and a 31.0% change in the atazanavir/ritonavir arm. The virologic response change between week 12 and 24 was statistically significant for the Atripla and atazanavir/ritonavir groups (P < 0.0001 for both) but not for the Stribild group (P = 0.35).
Among people who entered the trials with a viral load above 100,000 copies, sub-50-copy response rates were 79% and 85% at weeks 12 and 24 in the Stribild arm, 49% and 89% in the Atripla arm, and 38% and 86% in the atazanavir/ritonavir arm. Changes in virologic response from week 12 to week 24 averaged 6.3% with Stribild, 39.9% with Atripla, and 48.2% with atazanavir/ritonavir.
The overall virologic response data for Stribild meant that week 12 data had 92% sensitivity in predicting week 24 data and a positive predictive value of 95%. But a viral load above 50 copies at week 12 did not predict a detectable viral load at week 24.
Changes in adverse event rates from week 12 to week 24 proved low and consistent for the three regimens studied. For each regimen, rates of treatment-related adverse events rose only 1% between week 12 and 24: from 43% to 44% for Stribild, from 65% to 66% for Atripla, and from 54% to 55% for atazanavir/ritonavir. For all regimens, the top 10 most frequent treatment-related adverse events stayed stable from week 12 through week 96.
The Gilead team proposed that "week 12 may be a useful endpoint in phase 2 clinical trials of novel HIV-1 agents that produce rapid decline in HIV-1 RNA."
Reference
1. Martin H, Garner W, Porter J, Szwarcberg J. Week 12 as an early predictor of week 24 outcomes in trials of an integrase strand transfer inhibitor (INSTI) for HIV. ICAAC 2015, September 17-21, 2015, San Diego. Abstract H-1210.
|
| |
|
|
|
|
|
