Historical resistance profile helps to predict salvage failure

Abstract

This study compared the predictive value for treatment failure of extended resistance detected in the current genotype resistance test (GRT) versus those from GRT history in patients with multiple combination antiretroviral therapy (cART) failures. Patients who underwent three GRT between 1999 and 2007 were included. Extended resistance at genotypic sensitivity score (GSS) using the Rega 7.1 interpretation system compared with a non-standard definition (defined as class-wide resistance [CWR] on the basis of International AIDS Society-USA mutations) was assessed both for current and historical GRTs (a combination of mutations was detected in all three tests). The predictive role of extended resistance for treatment failure was evaluated with an adjusted Cox proportional hazard model. Overall, 177 patients were included. The historical GRT increased the number of patients with extended resistance to all three major drug classes by 25% in comparison with the current GRT. Using the GSS method, the absence of detection of any active drug in any drug class was predictive of failure with both the current and historical GRTs. Similarly, the number of active drugs in the cART regimen after the third resistance test, used as continuous variable, was also predictive of failure. Using both GSS approaches, current genotype had a higher effect than historical genotype on risk of treatment failure. Using the non-standard definition (CWR), historical resistance predicted failure better than current resistance. Our results provide an epidemiological demonstration that analysis of a combined latest and historical GRT, which also considers archived mutations, might better identify of the more virologically impaired patients in order to assess the best salvage treatment.