HIV drug resistance – a manageable risk or the harbinger of a “massive second global wave”?

Editor’s notes: The Paris conference [1] also saw the launch of the WHO report [2] and action plan on HIV drug resistance [3]. Laurie Garret writing in Business Insider [4] invoked drug resistance as one of the elements that could lead to a devastating reversal in progress against HIV. Inevitably with increasing scale up of ART, drug resistance levels will begin to rise.  In the absence of robust laboratory systems to detect virological failure, the risk is that drug resistance will continue to rise and could threaten progress in some countries.  Detection and accurate diagnosis of drug resistance is routine in well-resourced health systems, but still needs to be developed in many of the countries most affected.  While switching to a standard second line regimen provides a public health approach, there will be increasing need for tailored treatment and this requires both strong laboratory systems for sequencing relevant genes, but also good bioinformatic approaches to predict the optimum treatment regimen.

Svard et al. used a split genotyping procedure to explore resistance in people failing HIV treatment in Tanzania.  This term is used to describe a procedure where the nucleic acid amplification, which requires less sophisticated laboratory systems, is performed locally.  The amplified product was then sent to Sweden, where the sequencing and bioinformatics was performed.  Tanzania does not yet have widespread access to viral load testing, so treatment failure is defined using the WHO clinical and immunological criteria.  In this study, it became clear that as many as two thirds of people thought to be failing treatment were probably not failing virologically, and so were at risk of being changed onto second (or third) line treatments unnecessarily.  So the priority should clearly be to establish viral load assays to use to detect treatment failure and to make appropriate switches in treatment.  For the minority where failure was truly associated with a raised viral load, resistance mutations showed that people with failure on first line treatment often had viruses that were also resistant to second generation NNRTIs.  Current standard second line treatment in Tanzania is with a PI based regimen.  Among the first line failure cases, the probability predicted by the bioinformatics for successful second line treatment was around 85% using the medicines registered in Tanzania, which would improve to 95% if all current medicines were available. For people with second-line failure, the corresponding proportions were 79% and 94% respectively. 

Inzaule et al. also reported on drug resistance in East Africa.  They too showed that one in four patients failing second line treatment across Kenya could not be treated effectively with the medicines currently registered in Kenya.

While drug resistance is certainly a threat to the success of ART programmes, it is important to remember that resistance can be minimised by paying close attention to supporting those on treatment to maximize adherence.  Viral load assays need to be cheaper, simpler and more widely available to prevent unnecessary switching.  The results of the recent population-based HIV impact assessments (PHIA) are somewhat reassuring.  In random population-based surveys in Zimbabwe, Zambia and Malawi, 87%-91% of people taking ART had a suppressed viral load.  And in these three countries, the large majority of people are still taking first line treatment and viral load assays are not routinely available.  “Alert but not alarmed” was the message from the WHO press release [5] accompanying their report. 

Drug resistance testing through remote genotyping and predicted treatment options in human immunodeficiency virus type 1 infected Tanzanian subjects failing first or second line antiretroviral therapy.

Svärd J, Mugusi S, Mloka D, Neogi U, Meini G, Mugusi F,Incardona F, Zazzi M, Sönnerborg A. PLoS One. 2017 Jun 5;12(6):e0178942. doi: 10.1371/journal.pone.0178942. eCollection 2017.

Introduction: Antiretroviral therapy (ART) has been successfully introduced in low-middle income countries. However an increasing rate of ART failure with resistant virus is reported. We therefore described the pattern of drug resistance mutations at antiretroviral treatment (ART) failure in a real-life Tanzanian setting using the remote genotyping procedure and thereafter predicted future treatment options using rule-based algorithm and the EuResist bioinformatics predictive engine. According to national guidelines, the default first-line regimen is tenofovir + lamivudine + efavirenz, but variations including nevirapine, stavudine or emtricitabine can be considered. If failure on first-line ART occurs, a combination of two nucleoside reverse transcriptase inhibitors (NRTIs) and boosted lopinavir or atazanavir is recommended.

Materials and methods: Plasma was obtained from subjects with first (n = 174) or second-line (n = 99) treatment failure, as defined by clinical or immunological criteria, as well as from a control group of ART naïve subjects (n = 17) in Dar es Salaam, Tanzania. Amplification of the pol region was performed locally and the amplified DNA fragment was sent to Sweden for sequencing (split genotyping procedure). The therapeutic options after failure were assessed by the genotypic sensitivity score and the EuResist predictive engine. Viral load was quantified in a subset of subjects with second-line failure (n = 52).

Results: The HIV-1 pol region was successfully amplified from 55/174 (32%) and 28/99 (28%) subjects with first- or second-line failure, respectively, and 14/17 (82%) ART-naïve individuals. HIV-1 pol sequence was obtained in 82 of these 97 cases (84.5%). Undetectable or very low (<2.6 log10 copies/10-3 L) viral load explained 19 out of 25 (76%) amplification failures in subjects at second-line ART failure. At first and second line failure, extensive accumulation of NRTI (88% and 73%, respectively) and NNRTI (93% and 73%, respectively) DRMs but a limited number of PI DRMs (11% at second line failure) was observed. First line failure subjects displayed a high degree of cross-resistance to second-generation NNRTIs etravirine (ETR; 51% intermediate and 9% resistant) and rilpivirine (RPV; 12% intermediate and 58% resistant), and to abacavir (ABC; 49% resistant) which is reserved for second line therapy in Tanzania. The predicted probability of success with the best salvage regimen at second-line failure decreased from 93.9% to 78.7% when restricting access to the NRTIs, NNRTIs and PIs currently available in Tanzania compared to when including all approved drugs.

Discussion: The split genotyping procedure is a potential tool to analyse drug resistance in Tanzania but the sensitivity should be evaluated further. The lack of viral load monitoring likely results in a high false positive rate of treatment failures, unnecessary therapy switches and massive accumulation of NRTI and NNRTI mutations. The introduction of regular virological monitoring should be prioritized and integrated with drug resistance studies in resource limited settings.

Abstract [6] Full-text [free] access  [7]

 

Emergence of untreatable, multidrug-resistant HIV-1 in patients failing second-line therapy in Kenya.

Inzaule SC, Hamers RL, Mukui I, Were K, Owiti P, Kwaro D, Rinke de Wit TF, Zeh C. AIDS. 2017 Jun 19;31(10):1495-1498. doi: 10.1097/QAD.0000000000001500.

We performed a countrywide assessment of HIV drug resistance among 123 patients with virological failure on second-line antiretroviral therapy (ART) in Kenya. The percentage of patients harbouring intermediate-to-high-level resistance was 27% for lopinavir-ritonavir, 24% for atazanavir-ritonavir and 7% for darunavir-ritonavir, and 25% had complete loss of activity to all available first and second-line drugs. Overall, one in four patients failing second-line ART have completely exhausted available antiretrovirals in Kenya, highlighting the need for increased access to third-line drugs.

Abstract access [8] 

Africa [11]
Kenya [12], United Republic of Tanzania [13]
  • [14]