Background:  Darunavir (DRV) and brecanavir (BCV) are 2 protease inhibitors (PI) that are chemically related to amprenavir (APV), but have improved in vitro activity against PI-resistant clinical isolates. The development of the V82L or T protease (PRO) mutation represents the most common pathway leading to resistance to tipranavir (TPV) in treatment-experienced patients. This analysis investigated the change in phenotypic susceptibility to DRV and BCV following development of the V82L/T mutation after TPV treatment and the correlation between susceptibility to TPV, DRV, and BCV.

Methods:  Matched baseline and virological failure samples for isolates (n = 20) developing a V82L/T PRO mutation during the RESIST 1 and 2 clinical trials were evaluated for susceptibility to TPV, DRV, and BCV using the PHENOSCRIPT® Assay (Eurofins Viralliance). Fold change in susceptibility of viruses to drugs was determined relative to HIV-1_HXB-2.

Results:  At study entry, the median fold change values for TPV, DRV, and BCV were 2, 6, and 8, respectively. Following virologic failure of TPV, development of the V82L (n = 6) or T (n = 14) mutation occurred in combination with other PRO changes:  gain­s—I13V n = 5, I15V n = 3, L33F/I n = 5, K55R n = 3, I84V n = 8, L89M n = 3; losses—L33F n = 2, M46I/L n = 2, I50V n = 5. Susceptibility to TPV decreased (2- to 55-fold change); whereas overall susceptibility to DRV (6- to 6-fold change), BCV (8- to 7-fold change) remained unchanged. The median Δfold change (the fold change of failure sample relative to matched baseline sample’s fold change) of DRV and BCV susceptibility for individual isolates was 1.55 and 1.05, respectively. Virological failure of TPV was accompanied by loss of the APV-associated I50V mutation if present at baseline (n = 5). Acquisition of the V82L/T mutation accompanied by loss of the baseline I50V mutation, was associated with increased susceptibility to APV (20- to 9-fold change), DRV (14- to 8-fold change), and BCV (21- to 6-fold change). No correlation between susceptibility to TPV and DRV or BCV was observed (r² = 0.036 and 0.01) whereas a strong correlation between DRV and BCV (r² = 0.81), DRV and APV (r² = 0.85) and BCV and APV (r² = 0.67) was observed.

Conclusions:  Development of the V82L/T mutations following TPV therapy has limited effect on the susceptibility to DRV and BCV. The resistance profiles of APV, DRV, and BCV are closely related, whereas there is no correlation between the resistance profiles of TPV with those of DRV or BCV. These data support the option of using DRV or BCV if patients fail TPV-based therapy and develop a V82T/L PRO mutation.