Comparison of computed tomography imaging analyses for evaluation after chemotherapy in patients with colorectal cancer: a retrospective pooled analysis of six phase II clinical trials.

BACKGROUND: There are several methods for analyzing computed tomography (CT) images to evaluate chemotherapy efficacy in clinical studies. However, the optimal analysis method for each drug is still under debate. We conducted a pooled analysis using data from six phase II studies to evaluate four analysis methods in colorectal cancers (CRCs): morphological responses (MRs), early tumor shrinkage (ETS), depth of response (DpR), and response evaluation criteria in solid tumors (RECIST) ver.1.1. METHODS: We included 249 patients in this analysis. Pretreatments and findings of subsequent CT imaging were analyzed based on the MR, ETS, DpR, and RECIST ver.1.1. Differences in overall survival (OS) between the responders and non-responders according to each method were evaluated using survival analysis. RESULTS: The responders had significantly better hazard ratios (HRs) for OS, in terms of DpR (≥ median), ETS, objective response rate (ORR) [complete response (CR) + partial response (PR)], and disease control rate [CR + PR + stable disease (SD)]. Patients with right-sided colon cancers showed better HRs for DpR, but not for ETS and ORR. Contrastingly, patients with left-sided CRCs had better HRs for ETS, DpR, and ORR. MR was not associated with outcomes in this study, even in cases where bevacizumab was used. In patients with liver metastasis, ETS, DpR, and ORR showed better HRs, but not in those with lung metastasis. CONCLUSION: Early tumor shrinkage and DpR might be predictive markers only in left-sided CRCs with liver metastasis. Each imaging analysis has a different value based on the primary and metastatic sites.