Are PPIs Associated With an Increased Risk for Colorectal Cancer?

David A. Johnson, MD, FACG, FACP


October 29, 2007

Chronic Proton Pump Inhibitor Therapy and the Risk of Colorectal Cancer

Yang YX, Hennessy S, Propert K, Hwang WT, Sedarat A, Lewis JD
Gastroenterology. 2007;133:748-754


Proton Pump Inhibitor Use and Risk of Colorectal Cancer: A Population-Based, Case-Control Study

Robertson DJ, Larsson H, Friis S, Pedersen L; Baron JA; Sorensen HT
Gastroenterology. 2007;133:755-760


Editor's Note: Dr. Johnson's summary and commentary collaboratively address these 2 reports on chronic proton-pump inhibitor therapy and the risk for colorectal cancer.

Proton-pump inhibitors (PPIs) are widely prescribed by physicians who provide care for patients with gastric-acid-related diseases. This includes those patients with peptic ulcer disease, nonsteroidal anti-inflammatory drug-related ulcer, and gastroesophageal reflux-related diseases. Although PPIs are considered to be extremely safe[1] and have been approved for long-term use, concerns have been raised about the long-term effect of prolonged acid inhibition and elevation of serum gastrin level. Gastrin has trophic effects on tissue throughout the gastrointestinal tract. In particular, there has been significant controversy regarding the potential effect of increased gastrin levels and colorectal cancer risk. High gastrin levels have been associated with increased growth and proliferation of colon cancer cells in vitro.[2] In contrast, antagonism of gastrin has been shown to inhibit the growth of colorectal cancer.[3] Additionally, a nested case-control study suggested that elevated gastrin levels were associated with a 4-fold increased risk for development of colorectal cancer.[4] Given that PPIs increase serum gastrin levels, the authors of these 2 reports examined whether there is a potential link between PPIs and increased colorectal cancer risk.

The first study by Yang and colleagues was a nested case-control analysis of the General Practice Research Database, a computerized medical record system of approximately 700 general medical practices in the United Kingdom. The analysis was restricted to exclude patients with less than 5 years of follow-up, those younger than 50 years of age at database enrollment, and those with a history of colorectal cancer before the start of the standard database follow-up. Accordingly, a cohort of 890,368 remained from the total database of more than 8 million patients in the registry. "Cases" were defined as any patient with an initial diagnosis of colorectal cancer who had accumulated at least 5 years of follow-up. For each case, at least 10 controls were selected from the database and additionally carefully matched for practice site, time period, and duration of follow-up. History of PPI use was documented from the same database. From among the 4432 incident cases of colorectal cancer, the study authors found no evidence of increased risk for colorectal cancer associated with long-term use of PPIs over 5 years (odds ratio [OR], 1.1; 95% confidence interval [CI], 0.7-1.9). There was an association found between recent (less than 1 year) PPI use and the risk for colorectal cancer (OR, 2.6; 95% CI, 2.3-2.9); P < .001). However, this risk was seen exclusively with short-term use and there was no risk associated with longer-term use, thus suggesting alternative factors affecting the diagnosis and not a true drug effect. The authors also found no significant overall association between a history of pernicious anemia and the risk of colorectal cancer developing (OR, 0.9; 95% CI, 0..6-1.3).

In the second report, Robertson and colleagues conducted a population-based case-control study in a county of 500,000 inhabitants in Denmark. The population in this country has access to free hospital care. As with the Yang study, the main outcome interest was incident cases of colorectal cancer and the possible association with PPI exposure. To minimize the possible other reasons for exposure to PPIs, patients with PPI use of less than 1 year were excluded from the analysis. A total of 5589 cases of colorectal cancer were compared with 55,890 appropriately matched controls. There was no evidence of associated colorectal cancer risk when comparing "ever" users of PPIs to "never/rare" users (OR, 1.11; 95% CI, 0.97-1.27). Further support for this observation was evident by the lack of association of increased colorectal cancer risk among long-term (more than 7 years) users of PPIs (OR, 1.09; 95% CI, 0.58-2.06).

These studies are a welcome addition to the literature on the topic for 2 reasons. First, they are extremely well designed, drawing from excellent databases and employing a population-based design, thus avoiding/minimizing the potential for sampling bias. Second, at least the study by Yang avoids a major flaw of study analysis seen with the recent highly publicized reports attempting to associate PPI use with other complications, including community-acquired pneumonias,[5,6] Clostridium difficile-associated disease,[7] and hip fractures.[8] These studies all attempted to implicate a detrimental PPI effect on acid inhibition as a risk factor for related infectious or metabolic bone consequences. The hypothesis of all of these reports was that acid inhibition was detrimental and facilitated the attributed infectious and metabolic consequences. If, however, the effect of acid inhibition was true and causally detrimental, then profound or total acid inhibition (as seen with pernicious anemia or surgical gastric changes) should have a more evident effect. However, none of those studies used a nested case-control analysis to evaluate this possibility. The study by Yang did indeed look at patients with pernicious anemia to evaluate the effects of more profound acid inhibition, and accordingly, there was a much higher serum gastrin elevation than was seen with PPI use, and no associated increased risk for colorectal cancer was evident.

Limitations of both of these studies include lack of evaluation for potential confounders that might influence colorectal cancer risks, including colonoscopy history, smoking status, and body mass index. Nonetheless, both studies are reassuring from a public health standpoint. Although there is a high prevalence of PPI use and colorectal cancer in the general population, there is now reassuring information and strong evidence to suggest that there is no causal link between long-term use of PPIs and the risk of this malignancy developing.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.