Geospatial and temporal analysis of HIV-1 viral loads and drug resistance in Kwazulu-Natal, South Africa.

Abstract

Human immunodeficiency virus type 1 (HIV-1) remains a global public health problem, with certain hyperendemic geographic regions contributing significantly to the pandemic, such as the province of KwaZulu- Natal (KZN) in South Africa. Introduction of dolutegravir (DTG) into South African first-line antiretroviral therapy (ART) regimens in December 2019, was a huge stride towards achieving the 95-95-95 UNAIDS targets. However, progressive improvements in standard-of-care must be accompanied by equally dynamic treatment monitoring tools that proactively survey HIV-1 viral load (VL) outcomes and emergent HIV-1 drug resistance (HIVDR). Linking VL and HIVDR data to Geographic Information System (GIS) coordinates, could provide visual insights into areas requiring directed interventions, as HIV-1 treatment programmes mature. In this study, a surveillance database platform was developed to monitor routinely processed VL and HIVDR data from KZN public health facilities and linked to GIS coordinates for geospatial analysis. Over 7 million VLs were spatiotemporally mapped prior to- and during the transition to DTG-based first-line ART in KZN. Despite an overall decrease in VLs observed between 2018 and 2022 (median log10 copies/ millilitre (mL); 4.09 vs. 3.56), approximately one in every eight people living with HIV (PLWH) had detectable VLs ≥400 copies/mL, with 80% of those having VLs ≥1000 copies/mL. This underscored the importance of closely monitoring VLs in PLWH, including those with low-level viraemia (i.e., VLs 400–999 copies/mL), a potential marker for treatment failure and emergence of INSTI mutations on DTG-based ART. This study further investigated patterns of HIVDR from across KZN public health facilities and showed that ~35% (9/26) of adults on DTG with INSTI genotypes had DTG-associated resistance mutations. This represents one of the highest reported prevalences of DTG resistance to date, warranting heightened surveillance of DTGassociated mutations, to prevent accumulation of HIVDR mutations, especially given that four out of five genotypes already exhibited dual- or triple-drug class resistance. Moreover, analyses of GRTs from the paediatric population revealed that HIVDR was detected in ~90% of children aged <10 years (p<0.05) and that approximately 3 out of every 4 children aged 5-9 years had HIVDR to >1 ARV drug class. Through geospatial analysis of ~4,000 HIVDR genotypes, this study also identified geographic differences in the burden of HIVDR in KZN province, providing proof of concept that big data geospatial mapping can be leveraged for data-driven public health decision making and health systems strengthening. Overall, the findings presented in this thesis send a broader message on the need to direct efforts to known communities with high levels of viraemia on ART and/or HIVDR, enhancing VL re-suppression efforts, improving HIV-1 prevention activities, and ultimately implementing strategies to curb the emergence of HIVDR. This is particularly crucial in resource-limited HIV-hyperendemic settings, where epidemic intelligence can guide the strategic allocation of limited resources for maximum impact in controlling the HIV-1 epidemic.