Proton Pump Inhibitors and the Risk of Colorectal Cancer: a Meta-Analysis.
Shicai Chen,Xuejiao Song,Xiang Gao,Mingzhe Li,Zhihui Chen,Yulong He,Wenbin Zhan
DOI: https://doi.org/10.1097/mcg.0b013e3181ea181e
2011-01-01
Journal of Clinical Gastroenterology
Abstract:To the Editor: Proton pump inhibitors (PPIs) have been becoming the most widely used medications for the management of many acid-associated peptic disorders, such as peptic ulcer and gastroesophageal reflux disease. Although PPIs are considered to be safe for long-term use, one important concern referring to a possible link between PPI-induced hypergastrinemia and gastrointestinal cancers have been raised. PPIs may cause an obvious reduction in gastric acid secretion. This hypochlorhydria leads to a 2 to 6-fold increased secretion of serum gastrin from the G cells in gastric antrum. Gastrin is likely to have growth-promoting effects on the development of colorectal cancer (CRC). But the voices from clinical studies were controversial. The effect of PPIs on the risk of CRC remains unclear so far. To address this issue, we conducted a systematic literature search of following databases: MEDLINE, EMBASE, and Web of Science. Search terms included ‘‘proton pump inhibitor (s),’’ ‘‘omeprazole,’’ ‘‘lansoprazole,’’ combined with ‘‘cancer(s),’’ ‘‘neoplasm(s),’’ or ‘‘malignancy (ies).’’ The studies included in this meta-analysis could be either randomized controlled trails (RCTs) or observational studies (case-control or cohort) that evaluated the association between exposure to PPIs and the risk of CRC. Articles were excluded if there was insufficient published data for determining an estimate of odds ratio (OR) and a confidence interval (CI), or if the full text could not be found. Given that quality scoring in meta-analysis of observational studies is controversial, we did not intend to assess the methodologic quality of the primary studies. All data were independently abstracted by 2 reviewers with standardized data abstraction tool. Disagreements in data extraction were resolved by consensus, referring back to the original article. The statistical process was performed according to the guidelines proposed by the Meta-Analysis of Observational Studies in Epidemiology group. ORs with 95% CIs were calculated using Review Manager 4.2. Statistical heterogeneity was measured using the Q statistic (P<0.05 was considered representative of significant statistical heterogeneity). The potential for publication bias was assessed by using the Begg rank correlation method and the Egger weighted regression method (P<0.05 was considered representative of statistically significant publication bias). Without RCTs and cohort studies, 4 independent case-control studies met the predefined inclusion criteria. All these 4 studies, including about 100,000 participants, evaluated exposure to PPIs and risk of CRC and were controlled for potential confounding factors (age, sex, body mass index, smoking, H2-receptor antagonist use, nonsteroidal anti-inflammatory drugs/ aspirin and statin use, etc). PPI use was statistically significantly associated with an increase in the risk of CRC assuming a fixed-effects model (OR=1.37; 95%CI=1.28-1.47), but not statistically significant assuming a random-effects model (OR=1.19; 95% CI=0.90-1.57). When heterogeneity is present (P<0.01), the random-effects model is considered to be more appropriate than a fixed-effects model, resulting in wider intervals and a more conservative estimate of treatment effect. The result showed that PPI use was not significantly associated with an increased risk of CRC. The Begg’s as well as the Egger’s tests revealed no relation between the estimate of OR and study size, existence of an important selection or potential publication bias in our results is unlikely to occur. Nevertheless, there were some limitations to our meta-analysis. RCTs are considered to provide the best evidence on the effects of any intervention. But only 4 case-control studies were found in this meta-analysis and the strength of evidence might be not stronger enough. Further more, most of the studies included in this study had an intervention and follow-up time <5 years, which could be thought as a short period—compared with the latency time for at least 10 years between the initiation and the clinical detection of a cancer. The shorter the follow-up time is, the bigger effect of potential leading time bias is. In summary, the result of our study does not support the hypothesis that PPIs may increase the risk of CRC for a short-term use under usual doses for management of peptic ulcer disease. However, we cannot rule out an increase associated with higher doses of PPIs. Thus, it will be important to continue more high quality studies with longer follow-up time.