Figure 3 from Cell-Free DNA–Derived Immune Cell Ratios Uncover Cancer-Associated Systemic Changes

Le résumé fourni par la source

Longitudinal dynamics of cfDNA-derived immune cell contributions during therapy. A, Schematic of the treatment timeline for patients with colorectal cancer (CRC), including surgery and ACT. B, Longitudinal changes in cfDNA-derived rank of classic monocytes in patients with CRC. Number of patients per group is indicated. P values were calculated using two-sided Wilcoxon rank-sum tests comparing baseline (pre-op) to each time point. ns, not significant. C, Same as B, for plasma cell ranks in CRC. D, Schematic of the treatment timeline for patients with MIBC, including NAC, pre-RC, and follow-up time points. E, Longitudinal classic monocyte ranks in patients with MIBC, as in B. F, Same as E, for plasma cell ranks in MIBC. G, Receiver operating characteristic (ROC) curves for LASSO-regularized logistic regression models predicting relapse in CRC (red) and MIBC (yellow) based on cfDNA-derived immune features. Models were trained using the rank change between the baseline sample and the first chemotherapy time point (pre-op vs. ACT day 90 for CRC; pre-treat vs. NAC day 20 for MIBC) for the 10 cell types with highest IQR in rank change. Models were evaluated using LOOCV. Area under the curve (AUC) and 95% CIs are indicated. H, Kaplan–Meier curves of RFS for patients with CRC stratified into high (brown) and low (yellow) relapse probability groups based on the median predicted probability from the CRC model in G. Time is shown in months since operation. I, Same as H, for MIBC.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.