Can we predict regression of anal high-grade squamous intraepithelial lesions using molecular biomarkers?

ABSTRACT

High-grade squamous intraepithelial lesions (HSILs) of the anus are at risk of progression to anal squamous cell carcinoma (ASCC). However, most HSILs spontaneously regress, while treatment efficacy is debated and associated with side effects. No clear clinical predictors for anal HSIL regression have been validated to date. We aimed to identify molecular biomarkers that predict the regression of anal HSIL.In this retrospective translational study, 79 patients who were diagnosed with anal HSIL from a prospective clinical cohort in a tertiary unit of proctology were included. We performed RNA-Seq for 63 samples and whole-exome analysis for 52 samples. HSILs were divided into two groups according to their evolution: regressors or non-regressors. We used samples for training and others for validation, in accordance with the methods of the laboratory for RNA extraction. Cox proportional hazard models were fitted to identify the main prognostic factors associated with disease evolution.In the univariate analysis, the absence of tobacco consumption and the absence of HSIL treatment were associated with regression. The quality and quantity of the libraries were correct enough for analysis for 57.7% of the samples (30/52). A total of 162 genes were differentially expressed between the regressor and stable groups in both cohorts of training and validation. However, after adjustment for clinical factors and correction of the p-value for multiple testing, no molecular biomarker was strongly associated with disease evolution.No robust candidate genes predictive of HSIL regression were identified in this study.