Mutations and migration : molecular and epidemiological aspects of HIV-1 in Sweden

Abstract

The Swedish HIV-1 epidemic is characterized by a wide variation in subtypes due to migration and travel. The aims of this thesis were to investigate molecular methods for estimating time to diagnosis and epidemiological tools for estimating the proportion of undiagnosed individuals with HIV-1 in Sweden, as well as to update the knowledge on transmitted and pre-treatment drug resistance (TDR and PDR) among persons living with HIV-1 in Sweden.

In Paper I we found that viral diversity estimated by the proportion of mixed base calls in pol gene sequences is an indicator of whether a HIV-1 infection is recent (≤1 year) or chronic. A cutoff of 0.47% mixed base calls identified recent infections with a sensitivity of 89% and a specificity of 75%. In Paper II we studied TDR in newly diagnosed individuals (n=1,713, 71% coverage) in 2010-2016 in Sweden. The prevalence of TDR was 7.1% (95% CI 5.8-8.3%) and resistance to non-nucleoside reverse transcriptase inhibitors (NNRTI) increased significantly from 1.5% in 2010 to 6.2% in 2016 and was related to infection/origin in sub-Saharan Africa. A transmission cluster in men having sex with men with the M41L mutation was traced back to the 1990´s. In Paper III we evaluated two mathematical models for estimation of the proportion of undiagnosed individuals with HIV-1 in Sweden, SSOPHIE and the ECDC Modelling tool. The model estimates for 2013 were 26% (n=2,100; 90% plausibility range 900-5,000) and 21% (n=2,013; 95% CI 1,831-2,189), respectively. The wide confidence bounds and problems to account for migration limited the value of the results. We concluded that improvements of models and better data on migration are needed. In Paper IV we investigated PDR in individuals diagnosed in Sweden 2017-2019 (n=224) using an in-house high-throughput sequencing (HTS) assay with a detection limit of 1% for minority viral variants. HTS was successful in 87% of samples and failure was associated with low viral loads. Drug resistance mutations (DRMs) in protease/reverse transcriptase were detected in 49% of patients and in 18% of patients the DRMs were at levels ≥20% of the viral population. DRMs above 20% were mainly towards NNRTIs and correlated with previous ART exposure (p=0.001) and origin in Asia (p=0.007). Integrase strand transfer inhibitor (INSTI) DRM were rare and only found in levels <20% of the total viral population.

In conclusion, we have confirmed that viral diversity is a useful biomarker of time to diagnosis in HIV-infection and have identified important complications with estimating the proportion of undiagnosed HIV-1 infections in settings with high levels of migration. We have shown that the global increase in NNRTI-resistance is mirrored among HIV-positive migrants in Sweden and we have contributed to the knowledge about the utility of HTS in drug resistance testing of HIV-1.