Supplementary Materials from Validation of miR-31-3p Expression to Predict Cetuximab Efficacy When Used as First-Line Treatment in <i>RAS</i> Wild-Type Metastatic Colorectal Cancer

Abstract

<p>Supplementary Table 1. Baseline characteristics in "Low" and "High" miR-31-3p expression levels subgroups by treatment arm (FOLFIRI plus bevacizumab [FOLFIRI+Beva] and FOLFIRI plus cetuximab [FOLFIRI+Cet]. Supplementary Table 2. Treatment effect on PFS, OS and iOR in right-sided tumor miR-pop subjects - Multivariate Hazard Ratio estimates with Cox regression (95% confidence intervals, p-values) adjusted on BRAF mutational status for all analyses. Supplementary Table 3. Treatment effect on PFS and OS in left-sided tumor subjects miR-population - Multivariate Hazard Ratio estimates with Cox regression (95% confidence intervals, p-values) adjusted on BRAF mutational status and number of metastatic sites for PFS and BRAF mutational status, ECOG score and number of metastatic sites for OS. Treatment effect on iOR: rates by treatment arms and odds ratios estimated through multivariate logistic regressions adjusted on ECOG score. Patients with missing response were excluded. Supplementary Figure 1. Distribution of normalized expression of miR-31-3p according to Î"Î"Ct method in FOLFIRI plus bevacizumab and in FOLFIRI plus cetuximab arms. Supplementary Figure 2. Treatment effect on OS in miR-population - Kaplan-Meier Survival curves and Univariate Hazard Ratio estimates with Cox regression (95% confidence intervals, p-values). Supplementary Figure 3. Treatment effect on PFS in miR-population - Kaplan-Meier Survival curves and Univariate Hazard Ratio estimates with Cox regression (95% confidence intervals, p-values). Supplementary Figure 4. Treatment effect on PFS in miR-population - hazard ratios estimated through univariate and multivariate Cox regressions (95% confidence intervals, p-values). Supplementary Figure 5. Treatment effect on OS in miR-population - hazard ratios estimated through univariate and multivariate Cox regressions (95% confidence intervals, p-values). Supplementary Figure 6. Treatment effect on PFS in Low miR-population - hazard ratios estimated through univariate and multivariate Cox regressions (95% confidence intervals, p-values). Supplementary Figure 7. Treatment effect on OS in Low miR-population - hazard ratios estimated through univariate and multivariate Cox regressions (95% confidence intervals, p-values). Supplementary Figure 8. Treatment effect on PFS in High miR-population - hazard ratios estimated through univariate and multivariate Cox regressions (95% confidence intervals, p-values). Supplementary Figure 9. Treatment effect on OS in High miR-population - hazard ratios estimated through univariate and multivariate Cox regressions (95% confidence intervals, p-values). Supplementary Figure 10. Correlation between miR-31-3p expression and tumor shrinkage in subjects treated with FOLFIRI plus cetuximab. Supplementary Figure 11. Biological hypotheses for miR-31-3p role: EGFR suppress the maturation process of pre-mir-31 via the phosphorylation of AGO2. Left: in tumor sensitive to anti-EGFR therapy, low miR-31-3p expression level is a consequence of the of anti-EGFR treatment. Right: in tumor resistant to anti-EGFR therapy, no suppression of the maturation process induces high miR-31-3p expression level.</p>