Re-occurrence of HIV-1 drug mutations after treatment re-initiation following interruption in patients with multiple treatment failure

Antiretroviral treatment interruption in 20 extensively pre-treated HIV-1 patients with treatment failure led to genotype viral reversion of at least one class of drug-mutation resistance in half of the patients. The only predictive factor of reversion was found to be the duration of interruption. The outgrowth of residual wild-type virus seems not to be a true genetic reversion because drug mutations are detected rapidly at salvage therapy re-initiation. Viral resistance to drugs is a major cause of treatment failure in HIV-1-infected patients. Previous studies have shown that periods of time off the relevant drug are required in order to achieve a change from a mutant to a wild-type virus population, probably in relation to an impaired fitness of resistant viruses [1,2]. In this context, treatment interruptions had been proposed before treatment re-initiation in chronically infected patients with multiple failure to antiretroviral therapy. Theoretically, removing drug pressure will allow viral reversion from a highly resistant virus to a fully susceptible wild-type virus. Recent studies [3–6] reported a decline in detectable HIV-1 drug resistance mutations after stopping therapy, and even a complete shift to wild-type phenotype associated with a greater short-term virological response to a salvage regimen. The major limitation of antiretroviral treatment interruptions is the subsequent decrease in CD4 lymphocytes in these chronically infected patients [7,8]. The objectives of this prospective study were to assess the potential benefit of treatment interruption in terms of virus resistance profile, the efficacy of subsequent mega-highly active antiretroviral therapy (HAART) and the safety of stopping therapy in treatment-experienced patients with multiple treatment failures. In particular, the study also investigated the genetic evolution of resistance mutations in the pol HIV-1 gene after treatment re-introduction. Twenty patients with a long history of treatment failure were prospectively studied during a period of treatment interruption followed by re-initiation of therapy with a combination of five to nine antiretroviral drugs. Patients had a plasma HIV-1-RNA level of more than 10 000 copies/ml under antiretroviral therapy for a median of 36 months, and the baseline median value was approximately 160 000 copies/ml. Of the 20 patients, 16 had an absolute value of CD4 lymphocytes of less than 200 cells/mm3 before the treatment interruption and belonged to group C of the Centers for Disease Control and Prevention classification. The median time on therapy was 73 months (range 29–118) and all patients had received the three classes of antiretroviral drugs. The genotypic characteristics of the virus at baseline were a median of 10 major resistance mutations in reverse transcriptase and protease HIV-1 genes, and 16 patients harboured at least one major resistance mutation conferring resistance to each class of drug. The median duration of treatment interruption was 8 weeks, ranging from 4 to 24 weeks. As previously described in controlled patients [9,10], a plasma HIV-1-RNA rebound was observed but with a moderate median increase of 0.4 log10 (Fig. 1). This was probably because in those patients failing on HAART, the plasma HIV-RNA level on therapy was close to the pre-therapy level. Concomitantly, the median CD4 cell decrease of 15 lymphocytes was observed in this study, but was slightly lower than has been reported (Fig. 1) [8,11]. Two AIDS-related clinical events were observed in two patients with CD4 cell counts of less than 200 cells/mm3: a Candida oesophagitis and a progression of Kaposi's sarcoma.Fig. 1.: (a) Median changes from baseline in plasma HIV-1-RNA concentrations (log10 copies/ml). (b) Median changes from baseline in CD4 lymphocytes (cells/mm3). Day 0 (D0) STOP was the sample at the beginning of treatment interruption and D0 ARV was the sample at the beginning of treatment reintroduction. The symbol ▪ indicates patients with genotype reversion, ▴ indicates patients without genotype reversion. An intent-to-treat, model-based approach was used. The (n) indicates numbers of patients in the models-based analysis.Before treatment interruption, the viruses were multi-resistant to almost all currently available antiretroviral drugs. Our study revealed the disappearance of drug-associated mutations in the reverse transcriptase and protease genes for at least one class of drug after treatment interruption in 11 out of 20 patients. We observed that genotype viral shifts occurred more consistently with regard to protease inhibitor (eight out of 11 patients) and non-nucleoside reverse transcriptase inhibitor-associated mutations (nine out of 11 patients) than with mutations conferring nucleoside-analogue resistance (four out of 11 patients). Boucher et al. [1] showed in 1993 that a period of one year without zidovudine might be required to achieve a change from a mutant or mixed virus population to a wild-type virus population. In contrast to previous studies [5,6], which showed in half the patients a predominantly wild-type virus after treatment interruption, we observed wild-type virus in only four patients. This could be caused by the severity of the clinical status of patients at baseline, with 80% of them being Centers for Disease Control and Prevention group C, with a median duration of previous therapy of 6 years. The only factor, among RNA-1 HIV, CD4 cells and mutations pre- and post-interruption, which significantly predicted resistance reversion was the duration of treatment interruption in patients with genotype reversion (P = 0.05). A salvage treatment, including a median number of seven drugs (range five to nine drugs) was reintroduced in all patients after treatment interruption. A median decrease of approximately 2 log10 copies/ml from baseline was observed at month 2 until month 6 (Fig. 1). The median CD4 lymphocyte count returned to baseline at month 2 of treatment reintroduction (Fig. 1). In contrast with previous studies [5,6], there was no significant difference in the response to therapy re-initiation between patients harbouring ‘revertant’ and patients harbouring resistant viruses. However, chronically infected patients harbour a diverse population of viruses, and a mega-HAART may suppress the replication of an important number but probably not all of the variants. The genotype results were also provided 3 months after salvage treatment re-initiation. Among the 11 patients in whom a genotype viral reversion occurred, the genotypes remained wild type in only three patients. In the eight other patients, baseline resistance mutations were re-selected simultaneously or successively. Overall, these results suggest that the ‘genotype viral reversion’ is probably just a replacement of the highly mutated quasi-species by the less mutated quasi-species, the most fit viruses without treatment. A low proportion of detectable wild-type strains is probably caused by the long period of antiretroviral treatment before therapy interruption and to the turnover of wild-type virus-infected cells. Once antiretroviral drugs are restarted, the resistant quasi-species are likely to recur in these patients with multidrug-resistant viruses. These data could not determine how long the resistant virus persists at low levels in plasma or as an archived virus in long-lived cells. Recently, despite the loss of detectable resistance in plasma, resistant virus was cultured from peripheral blood mononuclear cells in half the patients during treatment interruption [7]. In these advanced patients, frequent monitoring and opportunistic infection prophylaxis would be necessary when treatment is discontinued. Further prospective randomized studies will be required to determine whether treatment interruption can be used as a treatment strategy among patients with few remaining treatment options. Constance Delaugerrea Marc-Antoine Valantinb Mireille Mourouxa Manuela Bonmarchandc Ghislaine Carcelaind Claudine Duvivierb Roland Tubianab Anne Simonc François Bricaireb Henri Aguta Brigitte Autrand Christine Katlamab Vincent Calveza