Genotypic and phenotypic resistance under zidovudine and lamivudine therapy in human immunodeficiency virus-infected children.

Abstract

Sir: With advancing duration of antiretroviral treatment in HIVinfected patients, appearance of resistant virus strains may cause therapeutic failure [1, 5]. Drug susceptibility can be assessed using genotypic and phenotypic assays, although their comparability and clinical signi®cance remains controversial [3]. We investigated 12 perinatally infected patients (5 females and 7 males), aged 2.5±13 years (median: 8.5 at the end of the study). Antiretroviral therapy was started in stage B (Center for Disease Control classi®cation, 1994). Dosage of zidovudine (AZT) was 6± 8 mg/kg body weight bid and of lamivudine (3TC), 4 mg/kg body weight bid. Seven patients received AZT (Retrovir) as initial monotherapy for a period of 11±49 months (median: 40), ®ve in combination with AZT and 3TC (Epivir) as a second-line therapy for 6±17 months (median: 8). The remaining ®ve patients were given AZT initially in combination with a second antiretroviral drug for 6±36 months (median: 8); three patients received 3TC and two patients didanosine (Videx). Phenotypic susceptibility to AZT and 3TC was assessed using the MTT-MT4 assay (Antivirogram) as previously described [2]. Results are expressed as the fold resistance (fold-increase in IC50) relative to the wild type. Maximal measurable resistance to AZT was 1500 fold and 27 fold to 3TC. Genotypic resistance was measured by determination of drugrelated point mutations in the pol gene. HIV 1-cDNA was obtained directly from ampli®ed reverse transcriptase coding regions including mutations at positions 41, 67, 70, 184, 215 and 219. A methionine to valine substitution at amino acid residue 184 (M184V) represents a highly increased 3TC resistance. The mutations M41L, K70R and T215F/Y are correlated with major and D67N and K219Q/E with minor AZT resistance [4, 6]. Resistance analyses were done after the above-mentioned courses of therapy; genoand phenotypic assays were performed simultaneously. Median phenotypic AZT resistance was 78 fold in patients with monotherapy and 3 fold in patients with ®rst-line combination therapy, showing a signi®cant di€erence (Table 1). Duration of AZT therapy correlated well with phenotypic and genotypic resistance. A 1±10 fold resistance was found in patients on treatment for 24 months representing drug resistance. Mutation frequency correlated with duration of AZT therapy and with values of phenotypic resistance. Maximal genotypic 3TC resistance was reached in almost all patients after 6±12 months of therapy, irrespective of whether 3TC was given in a ®rstor second-line combination. M184V substitution was found in four of ®ve patients with a second-line therapy and in two of three patient with a ®rst-line therapy. Antiretroviral monotherapy and/or long treatment duration (>24 months) led to highly increased AZT resistance, whereas highly increased 3TC resistance was seen within 6±12 months after onset of therapy in patients with ®rstand second-line combination. Frequency of mutations correlated with values of phenotypic resistance in patients with AZT therapy and secondas well as ®rst-line 3TC combination. Inpatients with second-line 3TC combination, a decrease in drug susceptibility caused an increased viral load and decrease in CD4 cell count. Our patients with a ®rst-line 3TC combination showed no decrease in CD4 cell count; although drug resistance to 3TC was demonstrated. These results raise the question whether decreased susceptibility to 3TC as a ®rst-line combination is of clinical relevance.