Resistance testing in children changing human immunodeficiency virus type 1 protease inhibitor

Objective. To assess genotypic and phenotypic resistance testing in HIV-1-infected children failing a first protease inhibitor (PI) therapy. Methods. In a multicenter observational study 21 children, ages 3 to 16 years, were given two reverse transcriptase inhibitors and one PI (mainly ritonavir, n = 18). They were subsequently treated with single or dual PI-based therapy (predominantly nelfinavir, n = 10, or ritonavir-saquinavir, n = 7). Resistance testing was performed at the time of therapy switch via direct sequencing and a recombinant virus susceptibility assay. Results. A total of 21 genotypic and 15 phenotypic resistance profiles were obtained. Most viruses displayed several reverse transcriptase mutations; however, 7 isolates maintained a wild-type protease. Ritonavir targeted the well-known pathway containing 82, 54, 46 and other secondary (nonactive site) mutations including T74A. No in vitro cross-resistance, i.e. greater than or equal to8-fold resistance to saquinavir or amprenavir, was encountered. Secondary mutations enhanced the prediction of ritonavir resistance (i.e. L10I) and in vitro nelfinavir cross-resistance (i.e. K20R/I conferred by primary (active site) resistance mutations. Either the 82, 54, 46 mutational genotype or the phenotype showing greater than or equal to8-fold nelfinavir cross-resistance predicted a poorer virologic response to nelfinavir salvage therapy. Conclusion. In a small cohort of heavily pretreated pediatric patients, resistance testing appears to predict the response to nelfinavir as salvage for a ritonavir-based therapy. This is further supported by the correlation between ritonavir-selected mutations and in vitro nelfinavir cross-resistance. Prospective studies should assess clinical outcome in children undergoing regimen changes based on resistance testing.