Abstract
OBJECTIVE: To access the comparative effectiveness and safety of different oral Chinese patent medicine (OCPM) versus oxaliplatin-based chemotherapy regimen (C) alone for colorectal cancer (CRC) through network meta-analysis (NMA). METHODS: Several electronic databases were searched for randomized controlled trials (RCTs) concentrated on the use of OCPM to treat CRC with C from the inception of the databases to January 10, 2021. We performed frequentist NMA and indirect comparison to compare study outcomes from the included RCTs. The risk of bias of each study was assessed using the Cochrane risk of bias tool. Confidence in evidence was assessed using Confidence in Network Meta-Analysis (CINeMA). RESULTS: A total of 31 RCTs with 1985 participants comparing 10 OCPM, namely, Antike (ATK), Shenyi (SY), Huachansu (HCS), Boerning (BEN), Xiaoaiping (XAP), Jinlong (JL), Compound matrine (CC), Pingxiao (PX), Xihuang pill (XHW), Kangaiping (KAP) were identified. The methodological quality of included RCTs was not very high. The results of the NMA showed that the comparisons were all indirect. Among diverse OCPM, ATK + C had the highest objective response rate (ORR) with a P-score of .63 with risk ratio (RR) of 1.37 (95% CI 1.12-1.66); with a RR of 1.96 (1.26-3.05), SY + C had the highest performance status with a P-score of .73; KAP + C had the lowest nausea and vomiting with a P-score of .91 and with a RR of 0.29 (0.10-0.79); and JL + C had lowest leukopenia with a P-score of .95 with a RR of 0.47 (0.31-0.72). The results of pairwise comparison suggested no difference in outcomes among 10 kinds of OCPM + C. The comparison-adjusted funnel plots suggested that there might not be small-study effects for outcomes. According to the CINeMa approach, the confidence rating of this NMA ranged from "very low" to "low" for various comparisons. CONCLUSION: Based on the NMA, ATK + C, SY + C, KAP + C and JL + C were associated with more preferable and options for CRC patients when referring to ORR, performance status, nausea and vomiting, and leukopenia, respectively. However, owing to the limitations of this research, the above conclusions require further verification by more high-quality RCTs. PROSPERO REGISTRATION: CRD42020160658.