Drug resistance in HIV patients with virological failure or slow virological response to antiretroviral therapy in Ethiopia

The present study was part of a randomized controlled trial evaluating two nutritional supplements in adult patients initiating ART in Ethiopia, which was registered at http://​www.​controlled-trials.​com (ISRCTN32453477). There were no differences in virological suppression among intervention groups (Olsen MF et al., submitted). The Ethiopian ART program was initiated in 2005 free of charge in hospitals and followed by the expansion to health centers in the periphery in 2007. Patients initiating ART at Jimma University Hospital, Jimma Health Center and Agaro Health Center were enrolled between July 2010 and August 2012. The first two centers are located in Jimma (a city of 150,000 inhabitants), with an urban site managing a total of approximately 1,617 patients on ART in Jimma University Hospital and 317 in Jimma Health center. The third centre is located in Agaro (a small town of 28,000 inhabitants) with approximately 240 patients on treatment.

HIV-positive patients attending the ART clinics, who were ART naïve and eligible for ART according to the Ethiopian national guideline [13], were invited to participate in the study and followed for 6 months. Inclusion criteria were being adult (≥18 years), living within approximately 50 km of Jimma, and consenting to participate. Pregnant or lactating women and patients with known previous use of ART were excluded. Participants gave informed consent and protocols were approved by Jimma University Ethics Review Board and National Research Ethics Committee of Ethiopia.

Decision to initiate treatment was made according to WHO criteria: stage IV irrespective of CD4 count, stage III if CD4 < 350, or CD4 < 200 cells/μl at any stage. The choice of ART combination was guided by national treatment guideline [13] and generally consisted of one of three first line regimens [Tenofovir (TDF) + Lamivudine (3TC) + Nevirapine (NVP)/Efavirenz (EFV); or Zidovudine (AZT) + 3TC + NVP/EFV; or Stavudine (d4T) + 3TC + NVP/EFV]. As part of the ART program, patients collected drugs every month for free. CD4 counts were monitored at ART initiation and every 6 months. Viral load monitoring was performed as part of this study.

Whole blood was obtained for CD4 count and plasma samples were separated and stored at −80°C until analyzed for HIV viral load and genotype. CD4 count and viral loads were measured at baseline and after 3 and 6 months on therapy. Genotypic HIVDR was done at baseline and after 6 months among selected participants based on viral load levels.

HIV-1 viral load (VL) was quantified using a commercial PCR assay (RealTime HIV-1, Abbott Laboratories, Illinois, USA) and automated extraction system (m2000 Real Time System, Abbott Laboratories, Illinois, USA) was used. Extracted samples were amplified and detected on the m2000rt platform (Abbott Laboratories, Illinois, USA). Virological failure was defined as a confirmed VL >1000 copies/mL at 6 months and viral suppression is the success of attaining VL <1000 copies/mL at 6 months. Virological slow responders were defined as VL ≥5000 copies/mL at 3 months but with viral suppression at 6 months. CD4 cells were enumerated using the Facscount® automated cell counter (Becton-Dickinson, Franklin Lakes, New Jersey, USA). Immunological failure was defined as a decline in CD4 count from baseline after 6 months of ART [14].

Self-reported adherence to ART was documented monthly. In addition, efavirenz or nevirapine plasma concentrations were measured at 1 and 2 months using liquid chromatography-tandem mass spectrometry at Division of Clinical Pharmacology, university of Cape Town, South Africa. Limit of quantification was 0.01 μg/ml for both drugs.

Plasma samples were shipped to Copenhagen, Denmark on dry ice for drug resistance mutation analysis at Statens Serum Institut (SSI, Copenhagen DK). Amplification of the pol gene was performed in two steps using SuperScript™ III One-Step RT-PCR System with Platinum® Taq High Fidelity (Invitrogen) in accordance with manufacturer´s instructions. First cDNA synthesis and first PCR was performed using the primers JA203 [15] and IN3 [16] at the following conditions: 52°C 30 min, 94°C, 2 min followed by 40 cycles of: 94°C, 15 sec; 60°C, 45 sec; 66°C, 3 min and a final extension at: 66°C for 10 min. The second nested-PCR was performed using the Expand High Fidelity PCR System (Roche) in accordance with manufacturer´s instructions and with the primers JA204 [15] and ABI1_R [17] at the following conditions: 94°C, 2 min followed by 35 cycles of: 94°C, 30 sec; 60°C, 30 sec; 72°C, 60 sec, and a final extension at: 72°C for 7 min, yielding a PCR fragment of 1449 bp. PCR products were prepared for sequencing using Illustra^TM ExoProStar 1 step (GE Healthcare) and sequencing reactions were carried out using the BigDye Terminator version 1.1 (Applied Biosystems) and the sequencing primers:

INH-A:CTTCAGAGCAGACCAGAGCCAA,

INH-B:GTTAAACAATGGCCATTGACAGA,

INH-D:CAGRCYGGAGCCAACAGC,INH-C:GATGGAAAGGATCACCRGCAATA, INH-F:TTGGGCCATCCATTCCTGG,

INH-G:CCATCCCTGTGGAAGCACAT

INH-H:TTCTGTATTTCTGCTAYTAAGTCTTTTG.

Sequencing was performed on an ABI 377 DNA Sequencer (Applied Biosystems) and sequence assembly and analysis was performed in BioNumerics v. 6.6 (Applied Maths, Sint-Martens-Latem, Belgium). For six of the 29 analyzed samples (Id #_month; 25_6, 146_6, 301_6, 314_6, 360_6 and 25_0), sequences were obtained using the ViroSeq HIV-1 genotyping System v. 2 (Abbott Diagnostics, Foster City, CA) in accordance with manufacturer´s instructions.

For each sequence, the resistance profile was calculated using the current HIVdb algorithm [18] at Stanford’s HIV genotypic resistance profile (http://​sierra2.​stanford.​edu/​sierra/​servlet/​JSierra?​action=​sequenceInput) and as implemented in BioNumerics v. 6.6.

Statistical analysis was done using STATA/IC version 11.2 (StataCorp LP, College Station, USA). Differences in means, medians and proportions between men and women were tested using Pearson Chi-square test and the Fisher’s exact test. P-values <0.05 were considered significant.

Sequences obtained in this study were submitted to GenBank under the following accession numbers: KJ561122-KJ561136 (baseline samples) and KJ561137-KJ561150 (6 months samples).

A total of 312 adult patients on first-line ART were enrolled into the study. Of these, 281 and 273 had reached 3 and 6 months of follow up, respectively, while the rest dropped out for various reasons including withdrawal of consent, lost to follow up or death. A total of 275 participants had VL results available at 3 months and 265 at 6 months. At ART initiation, the 265 participants retained for 6 months of follow-up did not differ significantly from the rest of participants (n =47), in terms of baseline characteristics such as age, BMI, VL and WHO stage. The mean (±SD) age was 33.0 (±8.8) years and 67% were women (Table 1). A combination of tenofovir (TDF), lamuvidine (3TC) and efavirenz (EFV) or nevirapine (NVP) was the most commonly (82.6%) prescribed first-line regimen. The other first-line regimens contained zidovudine (AZT; 14.3%) or stavudine (d4T; 3%) instead of TDF.

Virological failure was observed in 14 (5.3%) of the participants. Three of the participants with good viral suppression at 6 months had a slow response. These participants (Id # 146, 301 and 314) had 184878, 6849 and 125562 copies/mL at 3 months, respectively (Table 2). It was found that 158/275 (57.5%) and 233/265 (87.9%) of the participants achieved VL <40 copies/mL at 3 and 6 months respectively. Patients experiencing virological failure had significantly lower CD4 count at 6 months mean (±SD) =169 (±85) compared to those with successful viral suppression 313 (±131); P = 0.002). However, only four of the 14 participants with virological failure (28.6%) experienced immunological failure at 6 month (Table 2) and, 22 of 251 (8.7%) participants with virological suppression experienced immunological failure.

HIVDR genotyping was performed in pre-ART plasma specimens from 12 participants with virological failure. All patients were infected with HIV subtype C. Resistances to NNRTIs were identified in six of 12 (50%) and none were resistant to NRTI. Mutations detected in three of the patients are associated with high level resistance. Five participants with HIVDR mutations at ART initiation received ART regimens that were only partially active (Additional file 1: Table S1). During the 6 months of follow-up, none of these patients were switched to a fully active first-line or second-line regimen.

A total of seven patients had HIVDR mutations detectable in the baseline samples; one mutation (V90IV) confers no resistance (Id # 339). The K103N/X was the most common mutation observed. However, one participant (Id # 25) had a mutation (A98G) that confers resistance to nevirapine. This patient was treated with a supposedly active efavirenz based regimen and the mutation had disappeared at 6 months (Additional file 1: Table S1).

HIVDR genotyping was performed in plasma from 11 participants with virological failure at 6 months. Nine of them (81.8%) harboured mutations that cause high level resistance to NNRTIs. Six participants had the mutations at baseline, and three (Id # 9, 339 and 346) acquired additional mutations during follow-up causing resistance to NRTI. Three of the participants with HIVDR had no resistance at baseline but acquired resistance mutations during the course of treatment and two of them became resistant to both NNRTIs and NRTIs. HIVDR genotyping could not be performed for 3 participants (Id # 164, 354 and 373) due to failed sequencing attempt or sample inadequacy (Additional file 1: Table S1).

Of the three slow responders, one (Id # 301) was resistant to all three drugs prescribed, while we found no evidence of HIVDR in the other two participants at 6 months (Id # 146 and 314) (Additional file 1: Table S1). The former had pre-existing mutation and acquired additional mutations developing high level resistance to all the three antiretroviral drugs during the course of treatment (Additional file 1: Table S1).

All cases of HIVDR involved at least one NNRTI mutation, with the K103N (3/9) being the most frequently observed. Combined NNRTI and NRTI mutations were seen in four of nine (44.4%) participants and involved M184V/I in all cases and K65R and D67G occurred in one participant in addition to the M184I mutation. A K65R mutation was also detected in a slow responding patient. While five of 11 participants with virological failure received ART regimens that were not or only partially active, the other four became resistant during the therapy in spite of a fully active ART regimens initiated at baseline (Additional file 1: Table S1).

A total of four participants, two from the virologicaly failing group and another two from the slow responders had no HIVDR mutations at 6 months. Self reported adherence data indicated that all participants with genotype data had good treatment adherence except two of the slow responders (Id # 301 and 314). Measurements of NNRTIs plasma concentrations indicated two out of nine participants with resistance mutations at 6 months (Id # 213 and 269) and one participant with no resistance mutation (Id # 357) had a drug concentration below the limit of quantification at 1 and 2 months (Table 2).