We are pleased to share our latest publication in Scientific Reports, entitled A lasso-based model combining miRNA and clinical variables predicts future risk of breast and ovarian cancer. The paper was published online on March 24, 2026. Co-authors include Konrad Stawiski and Wojciech Fendler from the Department of Biostatistics and Translational Medicine, Medical University of Lodz.
The study presents a lasso-based model that integrates circulating serum miRNA profiles with clinical variables to estimate “BRCAness”, a phenotype resembling BRCA1/2 mutation carrier status, and to assess long-term cancer risk. The model was trained in a cohort of 1,831 women and then evaluated in an independent cohort of 1,044 participants from the PLCO trial.
The results are highly encouraging: the model achieved an AUC of 0.98 for BRCA prediction, and in the external cohort it showed a strong correlation with 5-year relative ovarian cancer risk (R = 0.93) and an AUC of 0.75 for direct prediction of future ovarian cancer. The study highlights how combining miRNA biomarkers with advanced biostatistical modelling can help refine cancer risk assessment and strengthen translational oncology research.