Polymorphism in Protease and Reverse Transcriptase and Phenotypic Drug Resistance of HIV-1 Recombinant CRF02_AG Isolates From Patients With No Prior Use of Antiretroviral Drugs in Abidjan, Côte d'Ivoire

To the Editor: Access to antiretroviral drug therapy (ART) is increasing in Africa. 1,2 For instance, today about 3200 HIV-infected patients have initiated ART in Côte d'Ivoire as part of the UNAIDS Drug Access Initiative. Because most drugs were designed, tested, and validated against HIV-1 subtype B viruses, a wider implementation of ART in Africa requires evaluation of the role of drug resistance since limited data exist regarding resistance profiles of HIV-1 non-B strains. In Côte d'Ivoire and most of West Africa, about 75% of HIV-1 strains are CRF02_AG, 3 which is a complex mosaic with subtypes A or G in the pol region. Limited studies of naive persons infected with HIV-1 non-B subtypes have examined mutational resistance profiles in the protease region 4,5 or protease and reverse transcriptase regions of HIV-1. 6,7 However, only limited phenotypic testing, which is a direct measure of resistance in the presence of drugs, has been done. Here we describe baseline polymorphism and phenotypic antiretroviral drug-resistant mutational profiles of patients infected with HIV-1 CRF02_ AG recombinant forms in Abidjan, Côte d'Ivoire. We analyzed ART drug resistance among a consecutively enrolled subset of 20 ART-naive HIV-1-infected patients who participated in the UNAIDS Drug Initiative (DAI). For sequencing of the Pol genes, we have used the TrueGene HIV-1 genotyping assay (version 2.5) (Visible Genetics, Toronto, Ontario, Canada). 3 Amino acid sequences obtained were compared with a reference HIV-1 sequence HXBr2, using the ADRA program from Los Alamos, New Mexico. Mutations were classified as primary or secondary and associated or not associated with ARV drug resistance, according to the consensus statement on ARV drug resistance. 3 The Bioedit (http://www.mbio.nesu.edu) program was used to align the 20 HIV pol sequence with the HXBr2 HIV-1 reference sequence. For identification of secondary HIV-1 resistance mutations or baseline polymorphisms we used the ADRA program from http://www.hiv.lanl.gov site. Phenotypic resistance was performed by a recombinant virus assay technology (Antivirogram, Virco NV, Mechelen, Belgium) as described previously. 8 Of the 20 patients, 8 (40%) were men; all the patients were infected with the HIV-1 CRF02_AG strains. Median age was 38 years (interquartile range [IQR] 30–40], median CD4+ T-cell count was 84 cells/μL (IQR, 13–137), and median viral load was 5.0 log10 RNA copies/mL (IQR, 5.0–6.0). We successfully sequenced all 20 protease and 19 RT regions. Compared with the HIV-1 subtype B (HXBr2) amino acid sequence, the RT gene was less conserved than protease gene. No primary genotypic resistance mutation was observed in any of the 20 protease or 19 RT sequences. Baseline polymorphisms were restricted to the following positions in the protease: M36I (n = 20, 100%), K20I (n = 19, 95%), L63P/H/L/S (n = 6, 30%), L10I/V (n = 4, 21%). Secondary mutations for the RT were at positions: L214F (n = 16, 84%), I135V (n = 12, 60%), R211K (n = 6, 31.6%), L283I (n = 2, 10.5%), V189I (n = 1, 5.3%), and W88S (n = 1, 5.3%) for the RT. Of the 20 patients' viruses, 12 (60%) had 2 protease mutations, and 8 (45%) had >2 mutations. Of the 19 RT sequences, 3 (15.8%) had a single mutation, 9 (47.4%) had 2 mutations, and 7 (36.8%) >2 mutations. We observed polymorphisms at positions not yet described to be genotypic resistance mutations in the protease and in the RT deduced amino acids. Of the 19 HIV-1 patients' viruses tested for phenotypic resistance, we observed low-level phenotypic resistance in 4 patients (21%) (Table 1). No correlation was observed between any secondary genotypic mutations and phenotypic resistance, as not all patients with the R211K, L214F, I135V, and L210M mutations had phenotypic resistance (Table 1). One patient had a 5.1-fold resistance to delarvidine (DLV) and had the I135V, the R211K, and L214F mutations. One patient had a 4.4-fold resistance to DLV but no sequence information was available for the RT gene. Lastly, one patient's virus had a 6-fold resistance to EFV and had the L214F and I135V mutations. Our results show that none of the 20 patients infected with the recombinant virus CRF02_AG had primary genotypic resistance mutations; however, low levels of phenotypic resistance were found in 4 patients' viruses. The high proportion of M36I (100%) and K20I (94%) in the protease region is comparable with what others have reported with other HIV-1 non-B subtypes. 5,6 We found a high percentage of the R211K (31.6%), L214F (84%), and I135V (60%) mutations in the 19 HIV-1 RT sequences analyzed. The R211K mutation has also been shown to be present in high percentage in non-B subtype. 6 The 135 and 283 RT positions have been strongly associated with reduced susceptibility to efavirenz, nevirapine, and delavirdine. 9 In the present study these mutations are observed in 13 samples but reduced susceptibility to drugs was observed in only 3 samples. Our findings that 4 ART-naive HIV-1 patients had low levels of phenotypic resistance are not clear. However, natural resistance to nevirapine (NVP) has been reported in HIV-1 ART-naive patients. 10 Taken together, our results suggest that persons infected with the CRF02_AG strains predominant in West Africa do not harbor naturally occurring genotypic mutations or phenotypic resistance and should respond appropriately to ART.TABLE 1: Phenotypic Sensitivity and Genotypic Mutations of 20 HIV-1 CRF02_AG Recombinant Viruses From Antiretroviral-Naive Patients in the UNAIDS-Drug Access Initiative in Abidjan, Cote d'IvoireChristiane Adjé-Touré Célestin E. Bilé Marie-Yolande Borget Kurt Hertog Chantal Maurice Monica L. Nolan John N. Nkengasong