TL;DR
The DTG RESIST study is a major international effort to track how HIV-1 develops resistance to dolutegravir (DTG), the current "gold standard" for global antiretroviral therapy. By analyzing viral genomes from 1,500 patients across 17 countries, the study aims to uncover how different HIV-1 subtypes and clinical contexts influence the emergence of drug resistance.
What Problem Does This Paper Address?
Dolutegravir is an Integrase Strand Transfer Inhibitor (InSTI) with a high "genetic barrier" to resistance, meaning the virus usually needs multiple specific mutations to bypass it. Because of its effectiveness and safety, the WHO endorsed it in 2019 as the primary choice for HIV treatment worldwide.
However, most of our data on HIV drug resistance comes from subtype B viruses, which are prevalent in high-income countries. The global epidemic is dominated by non-B subtypes (like Subtype C in Southern Africa or CRF01_AE in Southeast Asia). Early evidence suggests that the mutational pathways to resistance might differ by subtype. As DTG use scales up globally—often in resource-limited settings with less frequent viral load monitoring—there is a critical need to understand if and how resistance will emerge in these diverse viral backgrounds to safeguard long-term treatment success.
Methods
DTG RESIST is a multicentre longitudinal study embedded within the IeDEA (International epidemiology Databases to Evaluate AIDS) network. The study employs a multi-pronged technical approach:
- Genotyping: Using Next-Generation Sequencing (NGS) on the Illumina MiSeq platform to generate whole HIV-1 genomes. This allows researchers to look for resistance mutations not just in the pol gene (the standard target), but across the entire genome.
- Phenotypic Testing: Performing in vitro assays (PhenoSense Integrase) on approximately 100 selected samples to measure the actual inhibitory concentration (IC50) of DTG against specific mutant strains.
- Computational Modeling:
- Genome-Wide Association Studies (GWAS): To identify novel mutations outside the integrase gene that correlate with treatment failure.
- Conjunctive Bayesian Networks: To map the "evolutionary pathways" of mutations, determining the order in which they accumulate and how they depend on one another.
- Statistical Analysis: Using random-intercept logistic regression and parametric survival models to identify clinical risk factors (e.g., age, sex, previous drug exposure, and viral load monitoring frequency).
Datasets
The study leverages the IeDEA global cohort collaboration, which has collected data since 2006.
- Primary Cohort: The study aims to recruit 1,500 people living with HIV (PWH) who experience virologic failure while on a DTG-based regimen.
- Geographic Scope: 17 countries across sub-Saharan Africa, Southeast Asia, and Central/South America.
- Control Data: Viral sequences will be compared against phylogenetically matched controls from the Los Alamos HIV sequence database.
Key Insights and Ideas
- Subtype-Specific Pathways: The researchers hypothesize that the "genetic landscape" of different HIV subtypes makes certain resistance mutations more or less likely. For example, the Gln148 mutation appears less common in subtype C than in others.
- Beyond the Target Gene: While most resistance research focuses on the integrase enzyme (the drug's target), this study uses GWAS to explore if mutations in other parts of the virus—like the envelope or gag proteins—contribute to resistance or help the virus compensate for the "fitness cost" of primary mutations.
- The "Public Health Approach" Paradox: Standardized, simplified treatment regimens are great for scaling up care but may facilitate resistance if monitoring (like viral load testing) is infrequent. The study specifically looks at how these programmatic factors influence resistance.
Key Results
As this is a study protocol, the paper focuses on the design and power calculations rather than final clinical findings. However, it notes:
- Recruitment Progress: As of the paper's writing, over 550 participants have already been enrolled.
- Statistical Power: The study is powered to detect a 6% difference in resistance prevalence between major subtypes (e.g., Subtype B vs. C) with 81% power, assuming a sample size of 250 per group.
- Expected Outcome: The study expects to identify between 84 and 157 patients with confirmed InSTI drug resistance mutations, providing a robust dataset for evolutionary modeling.
Limitations and Future Work
- Baseline Data Gap: A major limitation is the lack of genotypic data from participants before they started DTG. This makes it harder to distinguish between transmitted resistance and acquired resistance. The authors plan to use imputation models to address this.
- Adherence Data: Consistent, harmonized data on patient adherence to medication is lacking across the various international sites.
- Subtype Uncertainty: The exact distribution of subtypes across sites is not perfectly known in advance, which could impact the statistical power for rarer subtypes.
- Future Directions: The results from the GWAS and Bayesian modeling will likely be "hypothesis-generating," requiring further validation in future clinical trials or laboratory studies.
Summary updated