HIV-GRADE: a publicly available, rules-based drug resistance interpretation algorithm integrating bioinformatic...

Area of Science:

• Medical Science
• Virology
• Pharmacology

Background:

• Genotypic drug resistance testing is crucial for guiding HIV treatment and managing drug resistance.
• Existing interpretation systems often do not account for the combined effects of multiple antiretroviral drugs.
• HIV-GRADE (Genotypischer Resistenz Algorithmus Deutschland) was developed to standardize drug resistance interpretation in Germany, including drug combinations.

Purpose of the Study:

• To develop and evaluate HIV-GRADE, a novel rules-based system for interpreting HIV drug resistance.
• To establish standardized interpretation rules for single drugs and drug combinations.
• To provide a platform for comparing different drug resistance interpretation algorithms.

Main Methods:

• HIV-GRADE rules were derived from literature, clinical data, and the bioinformatic system geno2pheno.
• An online tool was created based on the Stanford HIVdb interpretation tool.
• Clinical validation involved analyzing treatment episodes using multiple linear regression.

Main Results:

• HIV-GRADE demonstrated performance comparable to other leading systems (R(2) = 0.40).
• The system allows for simultaneous comparison of various drug resistance interpretation algorithms.
• Coefficients of determination (R(2)) showed similar performance across HIV-GRADE, Stanford HIVdb, REGA, and ANRS.

Conclusions:

• HIV-GRADE's performance is on par with existing rule-based interpretation systems for HIV drug resistance.
• The system is publicly available, offering a new approach to interpreting resistance to single drugs and combinations.
• Further improvements are anticipated, enhancing its utility in clinical practice.