Objective: To evaluate the development of phenotypic and genotypic resistance to zidovudine, didanosine and nevirapine as a function of the virologic response to therapy in a group of drug-naive individuals receiving various combinations of these agents.
Design: All patients were enrolled in a double-blind controlled randomized trial (the INCAS study) and were selected for detailed resistance studies based on specimen availability and virologic response.
Methods: Within the three study groups (zidovudine/nevirapine, zidovudine/didanosine or zidovudine/nevirapine/didanosine), 16, 19 and 24 patients, respectively, had evaluable baseline isolates and remained in the study > 24 weeks. Phenotypic resistance to all three drugs was evaluated using the VIRCO recombinant virus assay. Genotypic sequencing was done on selected specimens from patients receiving zidovudine/nevirapine/didanosine.
Results: After 24 weeks, all available isolates taken from patients receiving nevirapine were resistant to this agent, while 18/21 (86%) patients receiving triple therapy carried such isolates at 30--60 weeks. At 24 weeks, zidovudine resistance developed in 4/40 isolates but was more frequent after 30--60 weeks, especially in patients on two drugs. The degree of zidovudine resistance (rise in concentration required for 50% inhibition) appeared lower in the triple therapy group compared with zidovudine/didanosine (P = 0.0004). All nevirapine-resistant isolates that were sequenced carried at least one mutation associated with resistance, most often K103N and/or Y181C.
Conclusion: The use of highly active drug therapies may be associated with a beneficial effect on the development of antiretroviral drug resistance. The characteristics of virologic suppression that must be maintained to avoid resistance are currently being studied in hypothesis-driven clinical trials.