High prevalence of resistance to zidovudine among people in sub-Saharan Africa experiencing virological failure with tenofovir-containing ART

A significant proportion of people with virological failure after treatment with tenofovir-containing antiretroviral therapy (ART) have resistance to thymidine analogues (stavudine and zidovudine), and in many cases other classes of antiretrovirals, investigators report in Lancet Infectious Diseases. The retrospective study involved over 700 people in sub-Saharan Africa who entered care between 2005 and 2011.

“We found TAMs [thymidine analogue mutations] that are specifically selected by zidovudine and stavudine in roughly 16% of patients with failure of tenofovir-based first-line antiretroviral regimens,” comment the authors. “TAMs were associated with greater drug resistance to all components of WHO recommended, tenofovir-containing first-line treatment. The prevalence of resistance to tenofovir reached 80% in individuals with TAMs, a result that is concerning.”

The authors believe the findings have progamme implications, showing the importance of viral load monitoring and point-of-care resistance testing for people with suspected virological failure and those switching from tenofovir to zidovudine, as recommended after failure of first-line treatment.

An estimated 15-35% of people experience virological failure within 12 months of starting first-line ART containing thymidine analogues plus lamivudine and nevirapine or efavirenz. Most cases of resistance to lamuvidine and non-nucleoside reverse transcriptase inhibitors (NNRTIs) occur in settings where there is limited routine monitoring of viral load. Antiretroviral resistance could be responsible for up to 425,000 HIV-related deaths and over 300,000 new infections over the next five years, the World Health Organization estimated in its Global Action Plan for HIV Drug Resistance, published in 2016.

Since 2012, the World Health Organization (WHO) has recommended that first-line ART should include tenofovir, a nucleotide reverse transcriptase inhibitor, in place of a thymidine analogue. Of the 17 million people accessing ART in 2016, an estimated 3.5 million have previous experience of therapy with thymidine analogues. In most resource-limited settings, viral load monitoring is rarely performed before therapy is switched from a thymidine analogue to tenofovir. Single drug substitutions involve a significant risk of resistance and virological failure of therapy.

Investigators wanted to determine the prevalence and correlated factors of TAMs among people experiencing virological failure after tenofovir-containing first-line ART. Virological failure was defined as a viral load above 1000 copies/ml.

They identified 712 people in sub-Saharan Africa with no known previous exposure to stavudine or zidovudine who were then enrolled in 20 separate studies. The median age at baseline was 35 years and 58% were female. At baseline, median CD4 cell count and viral load were 92 cells/mm³ and approximately 175,000 copies/ml, respectively.

TAMs were detected in 16% of individuals.

People with TAMs were more likely to have resistance to tenofovir compared to those without TAMs (81% vs 59%, p < 0.001). Individuals with TAMs were also more likely than individuals without TAMs to have resistance to emtricitabine/lamivudine (87% vs 63%, p = 0.002) and NNRTIs (93% vs 77%).

There was a significant correlation between TAMs and tenofovir resistance (p < 0.001). People with TAMs were approximately 30% more likely to have tenofovir resistance compared to those without TAMs (OR = 1.29; 95% CI 1.13-1.47; p < 0.001). Unexpectedly, up to 80% of patients in some cohorts had both TAMs and the L65A mutation conferring resistance to. tenofovir; these two resistance patterns had been considered mutually exclusive.

The authors found that viral load monitoring was rarely performed before substitution of a thymidine analogue with tenofovir.

“This study has important policy implications for the limitation of drug resistance,” comment the authors. They recommend point-of-care viral load and resistance testing for people dropping thymidine analogues for tenofovir.

The investigators estimate that if this pattern of resistance development holds true for the entire population of people starting tenofovir-containing treatment in sub-Saharan Africa, 2-6% of people would have resistance to thymidine analogues after one year and 2-5% would have resistance to thymidine analogues, FTC and 3TC, to tenofovir and to efavirenz or nevirapine after one year. A further 8%-18% would have resistance to tenofovir, to FTC and to efavirenz and nevirapine after one year. This level of drug resistance will pose significant challenges for prescribing second-line treatment without drug resistance testing, unless a second-line regimen combining agents from two new drug classes (a protease inhibitor and an integrase inhibitor) is universally available.

The author of an accompanying editorial suggests that the use of the integrase inhibitor dolutegravir – which has a high barrier to resistance – in place of an NNRTI in first-line therapy could limit the risk of resistance. “This high barrier against the development of resistance in initial therapy could be life-saving in sub-Saharan Africa if properly used,” he concludes.