Non-nucleoside phenotypic hypersusceptibility cut-point determination from ACTG 359

Purpose: Non-nucleoside reverse transcriptase inhibitor (NNRTI) hypersusceptibility (HS) improves virologic response to those agents, but phenotypic susceptibility cutpoints, and methods to determine the cut-points, have not been completely defined. Method: Phenotypic drug susceptibility (fold change in IC₅₀ [FC]) was determined for 96 randomly selected antiretroviral-experienced, NNRTI-naive patients who received a delavirdine (DLV)-containing regimen in ACTG 359. A weighted FC score was used to account for other regimen agents. Regression models were used to define baseline DLV HS cut-points using week 4 or week 16 responses. Results: At study entry, DLV HS was present in 36% (35/96) of patients. Models explored HS cut-points from 0.2-1.0 using the week 4 virologic response. Using either a binary or continuous endpoint, DLV HS cut-points between 0.3 and 0.4 were identified. The classification and regression tree (CART) analysis identified baseline DLV FC <0.44 as a predictor of week 4 response. Conclusions: In relating drug HS to virologic response, several different analytic methods identified a DLV HS FC cut-point of 0.3-0.4. In refining phenotypic cut-points, early virologic responses (not confounded by rebound) may be better metrics than later responses, especially for drugs with low genetic barriers, such as DLV.