Comprehensive analysis of proton pump inhibitors and risk of digestive tract cancers

• Published in: European journal of cancer (1990) Vol. 156; pp. 190 - 201
• Main Authors: Zeng, Ruijie, Cheng, Yunjiu, Luo, Dongling, Wang, Jinghua, Yang, Jun, Jiang, Lei, Zhuo, Zewei, Guo, Kehang, Wu, Huihuan, Leung, Felix W., Sha, Weihong, Chen, Hao
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.10.2021; Elsevier Science Ltd
• Subjects: Cancer; Carcinogens; Causality; Causation; Colorectal cancer; Colorectal carcinoma; Confidence intervals; Criteria; Digestive tract cancer; Dose; Duration; Esophageal cancer; Esophagus; Evaluation; Gastric cancer; Gastrointestinal tract; Liver cancer; Meta-analysis; Pancreatic cancer; Polynomials; Proton pump inhibitor; Proton pump inhibitors; Protons; Risk; Risk assessment; Duration; Digestive tract cancer; Dose; Proton pump inhibitor
• DOI: 10.1016/j.ejca.2021.07.030
• ISSN: 0959-8049; 1879-0852

For the past two decades, dispute on whether proton pump inhibitor (PPI) leads to digestive tract cancer remains, and emerging studies in recent years still demonstrate inconsistent results, which continues to perpetuate concerns over the safety of PPI use. We performed a systematic review and meta-analysis, with comprehensive evaluation by Bradford Hill criteria of causation, to assess the effect of PPI use on digestive tract cancers. Medline, Embase and Web of Science databases were searched for observational studies published up to 15th January 2021. Pooled relative risks (RRs) were estimated via random effects models. Cumulative defined daily dose- and duration-risk relationships using restricted cubic spline and fractional polynomial models were investigated. Bradford Hill criteria were applied to evaluate causation. PROSPERO Registration: CRD42020211103. Thirty-two publications containing 4,355,254 participants were included. PPI use is associated with an increased risk of overall digestive tract cancers (RR = 1.63, 95% confidence interval (CI) 1.33 to 2.00). PPI use is correlated with increased risks of gastric cancer (RR = 1.78, 95% CI 1.38 to 2.31), pancreatic cancer (RR = 1.72, 95% CI 1.05 to 2.82) and liver cancer (RR = 1.62, 95% CI 1.04 to 2.52), but not of esophageal cancer (RR = 2.06, 95% CI 0.65 to 6.57) and colorectal cancer (RR = 1.24, 95% CI 0.93 to 1.66). The association between PPI and digestive tract cancers is stronger in people with minimal exposure. When cumulative defined daily dose or duration increases, the risks decline and become non-significant. Evaluation by Bradford Hill criteria indicates weak evidence of causation. A causal relationship between PPI use and digestive tract cancers is not supported by the evidence in the current review. Concerns over carcinogenic side-effects of PPI might be unfounded. •Thirty-two studies containing 4.3 million participants are included.•Correlation is evaluated by dose-response analysis and Bradford Hill Criteria.•The risks are non-significant in the highest dose or duration of PPI use.•The causality between PPI use and digestive tract cancers might not be supported.•Dose-response analysis and Bradford Hill Criteria are recommended for interpretation.