Analysis of HIV cross-resistance to protease inhibitors using a rapid single-cycle recombinant virus assay for patients failing on combination therapies

Objective: To assess the patterns of HIV phenotypic cross-resistance to pro tease inhibitors (PI) in patients experiencing viral load rebound on combin ation therapy including a PI. Methods: Phenotypic analysis of sensitivity to indinavir, nelfinavir, saqui navir, ritonavir and amprenavir was carried out using a single-cycle recomb inant virus assay. Viral protease was sequenced by automated dideoxynucleot ide chain termination. Results: Of the 108 patients studied, 68 had received indinavir, 50 ritonav ir, 25 saquinavir and eight nelfinavir. The majority (77%) had received onl y one PI. The incidence of cross-resistance between indinavir, nelfinavir, ritonavir and saquinavir was high (60-90%). Cross-resistance to amprenavir was less frequent (37-40%). However there was some correlation between leve ls of sensitivity to amprenavir and indinavir (r(2) = 0.34; P < 0.01). Conv ersely, the correlation between levels of sensitivity to indinavir and saqu inavir was poor (r(2) = 0.25), particularly for patients who had not receiv ed saquinavir. The degree of cross-resistance correlated with the lever of resistance and with the total number of mutations in the protease gene (P < 0.05, chi square test) but could not be significantly correlated to any on e particular mutation or combination of mutations. Mutation 184V was signif icantly associated with cross-resistance to amprenavir, with no mutations a t codon 50 observed, while mutations associated with cross-resistance to sa quinavir differed according to the treatment received. Conclusions: These results suggest that, although the total number of prote ase mutations correlates with the degree of cross-resistance, the specific mechanisms accounting for primary resistance and for cross-resistance may b e different. (C) 1999 Lippincott Williams & Wilkins.