Background:  Etravirine (TMC125) is a next-generation NNRTI with demonstrated activity against NNRTI-resistant HIV-1. Patients in the DUET trials whose viruses contained ≥3 TMC125 resistance associated mutations (RAM:  V90I, A98G, L100I, K101E/P, V106I, V179D/F, Y181C/I/V, G190A/S) at baseline had a response comparable to that of the overall placebo group. The likelihood of response to TMC125 was estimated by determining the prevalence of each TMC125 RAM and the frequency of various combinations in a large number of samples received for routine clinical resistance testing.

Methods:  From January 1999 to June 2007, 226,491 samples were submitted to Virco for routine clinical resistance testing. Samples were defined as NNRTI-resistant if they carried at least 1 International AIDS Society (IAS) –USA-defined NNRTI mutation (September 2006) or if the predicted fold-change in EC₅₀ for any currently approved NNRTI was greater than the respective vircoTYPE (vT) biological cut-off (BCO).

Results:  Respectively, 89,113 and 95,019 samples met the definition of having NNRTI resistance according to the IAS-USA mutation list or vT BCO. Of isolates with NNRTI resistance, 40% defined by the IAS-USA list, contained no TMC125 RAM, while 36.7%, 16%, and 7.3% harbored 1, 2, and ≥3 TMC125 RAMS, respectively. In the same subset (n = 89,113), the TMC125 RAM occurring most frequently were Y181C (27.6%), G190A (21.1%), K101E (8.9%), L100I (7.6%), V90I (6.9%), A98G (6.4%), G190S (4.3%), and V106I (3.5%). The other 5 TMC125 RAM had a prevalence of <3%. Among all NNRTI-resistant samples with 2 TMC125 RAM (n = 14,258) the most frequent combinations were:  Y181C+G190A (27.1%), K101E+G190A (12.5%), and V90I+Y181C (7.0%). While the proportion of NNRTI-resistant samples with TMC125 RAM declined over time (69.9% in first semester 1999 to 54% in first semester 2007) and the prevalence of some TMC125 RAM changed (Y181C declined from 44.9 to 20.7%, V106I increased from 2.1 to 4.8%), the ratio between isolates with 1, 2, or ≥3 TMC125 RAM remained relatively constant (~5:2:1). Similar results were observed in samples with NNRTI resistance defined by the vT BCO (n = 95,019).

Conclusions:  In 89,113 clinical isolates with known NNRTI resistance submitted for routine clinical resistance testing, only 7.3% harbored ≥3 TMC125 RAM. These data suggest that the co-existence of ≥3 TMC125 RAM, which may confer a diminished virologic response to TMC125, is infrequent even in patients with evidence of resistance to the currently approved NNRTI.