Background:  Quantitative phenotypic resistance information interpreted via clinical cut-offs can facilitate optimization of combination antiretroviral therapy. Virtual phenotype-LM (Linear Model) predictions of TMC125 (etravirine, ETR) drug susceptibility and clinical cut-offs are described.

Methods: Based on clinical isolates with both drug susceptibility phenotypes (Antivirogram) and viral genotypes, linear regression models were developed to predict TMC125 fold-change in IC₅₀ from the viral genotype (Virtual phenotype-LM, vPT). Using treatment response data from the 2 phase III DUET trials and 4 phase IIb trials, a separate linear regression model was developed to predict 8-week change in viral load for regimens including TMC125 as a function of baseline viral load, TMC125 fold change, activity of the background regimen, and enfuvirtide use. We defined 2 clinical cut-offs, corresponding to predicted fold-change values associated with 20% or 80% loss of the response of subjects infected with HIV-1 wild type strains.

Results: vPT predictions of TMC125 fold change weigh the contributions of 284 mutations and mutation pairs in reverse transcriptase (RT), and provide an accurate prediction of the measured fold change (R = 0.87, n = 20447). vPT predicted TMC125 resistance ranged from 0.9 fold change for wild type isolates to 200 fold change. We observed 20% (clinical cut-off 1) and 80% (clinical cut-off 2) loss of TMC125 response in treated patients at 1.6 and 27.6 TMC125 fold change. In the analysis dataset, patients receiving TMC125 with baseline TMC125 fold change £clinical cut-off 1 (n = 355), between clinical cut-off 1 and clinical cut-off 2 (n = 413), and > clinical cut-off 2 (n = 85) had median viral load reductions of –2.6, –2.3, and –1.3 log₁₀ HIV RNA and 38%, 24%, and 15% of patients achieved <50 copies/mL at 8 weeks, respectively. By week 24, the percentage of patients achieving <50 copies/mL on their complete treatment regimen increased to 55%, 37%, and 26%, respectively (drop-outs considered as non-responders).

Conclusions:  We have integrated complex interactions among multiple mutations in RT to provide accurate quantitative predictions of TMC125 susceptibility, which can be interpreted utilizing drug-specific clinical cut-offs. These analyses will help guide clinicians in making decisions on the use of this compound and backbone therapy by taking into account its spectrum of activity.