Meta-Analysis of the Prognostic and Predictive Role of the CpG Island Methylator Phenotype in Colorectal Cancer

Abstract

Background. Various studies have produced contradictory results on the prognostic role of the CpG island methylator phenotype (CIMP) among colorectal cancer (CRC) patients. Although a meta-analysis published in 2014 reported a worse prognosis of CIMP among CIMP-high (CIMP-H) CRC patients, the sample sizes of the major included studies were small. In this study, we included the most recent studies with large sample sizes and performed an updated meta-analysis on the relationship between CIMP and CRC prognosis. Methods. A search of MEDLINE, Web of Science, and Cochrane for studies related to CIMP and CRC published until July 2021 was conducted based on the PICO (participant, intervention, control, outcome) framework. Data extraction and literature analyses were performed according to PRISMA standards. Results. In the present update, 36 eligible studies (20 recently published) reported survival data in 15315 CRC patients, 18.3% of whom were characterized as CIMP-H. Pooled analysis suggested that CIMP-H was associated with poorer overall survival (OS) ( $\text{hazard}\text{ratio}\left[\text{HR}\right]=1.37$ , 95% CI: 1.26–1.48) and disease-free survival/progression-free survival/recurrence-free survival (DFS/PFS/RFS) ( $\text{HR}=1.51$ , 95% CI: 1.19–1.91) among CRC patients. Subgroup analysis based on tumor stage and DNA mismatch repair (MMR) status showed that only patients with stages III-IV and proficient MMR (pMMR) tumors showed a significant association between CIMP-H and shorter OS, with HRs of 1.52 and 1.37, respectively. Three studies were pooled to explore the predictive value of CIMP on CRC patient DFS after receiving postoperative chemotherapy, and no significant correlation was found. Conclusion. CIMP-H is associated with a significantly poor prognosis in CRC patients, especially those with stage III-IV and pMMR tumors. However, the predictive value of CIMP needs to be confirmed by more prospective randomized studies.