Preference criteria for regorafenib in treating refractory metastatic colorectal cancer are the small tumor burden, slow growth and poor/scanty spread

Abstract

Given the unclear preference criteria for regorafenib in treating refractory metastatic colorectal cancer (mCRC), this study aimed to construct an algorithm in selecting right patients for regorafenib. This was a multicenter retrospective cohort study. Patients with pathology confirmed mCRC and administered with regorafenib for > 3 weeks were enrolled. Patients with good response were defined to have progression-free survival (PFS) of ≥ 4 months. The Kaplan-Meier plot was used to analyze survival. A Cox proportional hazards model was used to analyze univariate and multivariate prognostic factors and was visualized using forest plot. A clustering heatmap was used to classify patients according to responses. The decision tree and nomogram were used to construct the approaching algorithm. A total of 613 patients was analyzed. The median PFS and overall survival (OS) were 2.7 and 10.6 months, respectively. The partial response and stable disease rate are 2.4% and 36.4%. The interval between metastasis (M1) and regorafenib, metastatic status (number, liver, and brain), and CEA level were independent prognostics factors of PFS that classifies patients into three groups: good, bad and modest-1/modest-2 group with PFS > = 4 months rates of 51%, 20%, 39% and 30%, respectively. Results were used to develop the decision tree and nomogram for approaching patients indicated with regorafenib. The preference criteria for regorafenib in treating patients with refractory mCRC are small tumor burden (CEA), slow growth (interval between metastasis and regorafenib) and poor/scanty spread (metastatic status: number and sites of metastasis): The 3S rules.TRIAL registration ClinicalTrials.gov Identifier: NCT03829852; Date of first registration (February 11, 2019).

Conflict of interest statement

Hao-Wei Teng, Hung-Chih Hsu, Kuo-Cheng Huang received a research grant in this PI initiated clinical investigation trial from Bayer Taiwan Co., Ltd (Bayer was not involved in the study design, data interpretation, and manuscript preparation and review.) Wei-Shone Chen, Jeng-Kai Jiang, Shung-Haur Yang, Huann-Sheng Wang, Shih-Ching Chang, Yuan-Tzu Lan, Chun-Chi Lin, Hung-Hsin Lin, Sheng-Chieh Huang, Hou-Hsuan Cheng, Tsai-Sheng Yang, Chien-Chih Chen, Yee Chao has no actual or potential competing interests.

© 2021. The Author(s).

Figures

Figure 1

Kaplan–Meier curves for (a) progression-free survival (PFS); (b) overall survival (OS); (c) PFS by metastatic sites; (d) PFS by liver metastasis; (e) PFS by interval between M1 and regorafenib; (f) PFS by Carcino Embryonic Antigen (CEA) level.

Figure 2

The forest plots summarize the (a) univariate and (b) multivariate analyses of prognostic factors in predicting progression-free survival (PFS). APM, abdominal/peritoneal metastasis.

Figure 3

The forest plots present the (a) univariate and (b) multivariate analyses of prognostic factors in predicting overall survival (OS). Abbreviation: APM, abdominal/peritoneal metastasis.

Figure 4

The clustering heatmap (k-means = 4) was used to classify patients into good, bad, modest 1, and modest 2 groups according to significantly independent prognostic factors by PFS >  = 4 months or not. (Top) Pie charts represent the distribution of patients with good and bad responses in four groups. (Middle) Clustering heatmap visualizes the different groups’ distribution of independent prognostic factors. (Bottom) Independent prognostic factors were classified into three patterns of cancer behaviors: growth rate, metastasis ability, and tumor burden. It was used to score the grouped patients. Abbreviations: ad, adenocarcinoma; HR, hazard ratio; CI, confidence interval.

Figure 5

Kaplan–Meier curves for (a) progression-free survival (PFS); (b) overall survival (OS) by patient with PFS >  = 4 months or not. (c) The decision tree in predicating the priority right of regorafenib for the treatment of refractory metastatic colorectal cancer according to liver metastasis, CEA, and interval between M1 and regorafenib. (d) The prognostic nomogram for OS of patients after regorafenib.