Analyzing HIV Mutations and Treatment Response
Author Information
Author(s): Wittkop Linda, Commenges Daniel, Pellegrin Isabelle, Breilh Dominique, Neau Didier, Lacoste Denis, Pellegrin Jean-Luc, Chêne Geneviève, Dabis François, Thiébaut Rodolphe
Primary Institution: INSERM, Unit 897, Bordeaux, France
Hypothesis
Can principal component analysis (PCA) and partial least square (PLS) regression provide better insights into the impact of HIV mutations on treatment response compared to traditional genotypic scoring methods?
Conclusion
PCA and PLS showed good performance in predicting virological response, but their predictive ability was not clinically superior to that of the genotypic score.
Supporting Evidence
- Virological failure was observed in 46 (53%) patients at week 12.
- The cross-validated AUCs for PCA, PLS, and genotypic score were 0.880, 0.868, and 0.863, respectively.
- PCA and PLS regression helped to describe mutation clusters.
Takeaway
This study looked at how different methods can help understand how HIV mutations affect treatment success. It found that while new methods are useful, they don't work better than the traditional scoring system.
Methodology
The study used data from the ANRS-CO3 Aquitaine Cohort, analyzing 87 patients' genotypes and virological responses using PCA, PLS regression, and traditional genotypic scoring.
Potential Biases
Potential biases may arise from the limited sample size and the specific patient population analyzed.
Limitations
The study's findings may not be generalizable beyond the specific cohort and treatment regimens studied.
Participant Demographics
Patients were HIV-1 infected individuals with a complete baseline genotype and plasma HIV-1 RNA available.
Statistical Information
P-Value
< 10-4
Confidence Interval
[2.4; 7.0]
Statistical Significance
p < 10-4
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website