People with celiac disease do not have an increased risk of developing gastrointestinal (GI) cancers, according to a large population-based study from Peter Elfström et al. in the January issue of Clinical Gastroenterology and Hepatology.
Celiac disease has been associated with GI cancers in small studies, but there have been no estimates from large populations or based on histopathology analyses.
Elfström et al. collected data from 28 pathology departments in Sweden on biopsy reports from 28,882 individuals with celiac disease, 12,860 with intestinal inflammation and 3705 individuals with latent celiac disease (based on serology test results, but with normal mucosa). They linked the data with those from the Swedish Cancer Registry to identify individuals that eventually developed cancers of the oropharynx, esophagus, stomach, small intestine, large intestine, rectum and anus, liver, or pancreas.
They found that over the study period, patients with these disorders did not have a significantly higher risk of developing GI cancers than the general population, matched for age and sex—in fact, patients with latent celiac disease had a lower absolute cancer risk than the population.
“We saw an association between having a biopsy and GI cancer in the first year, but then after that we saw no increased risk”, said the study’s senior author Jonas F. Ludvigsson in a video abstract:
Video abstract from Jonas Ludvigsson
Ludvigsson believes that the observed increase in cancer within the first year after biopsy analysis resulted from confounding factors—a patient with GI cancer would have symptoms that lead to biopsy analysis and identification of celiac disease or inflammation, and then the cancer.
Small intestinal cancer made up 6%–7% of all GI cancers that developed in individuals with celiac disease, with hazard ratio of 2.2 after the first year of follow up. However, Elfström et al. propose that the increased incidence of small intestinal cancer in these patients, compared with the population, resulted from a combination of surveillance bias or possibly the effects of persistent inflammation.
This was one of the largest studies of cancer risk in children with celiac disease. Of almost 12,000 children with celiac disease, only 4 developed cancer after the first biopsy sample was collected. Although this corresponded to more than 5-fold increase in cancer risk compared with the general pediatric population, the authors urged caution in interpretation of these data, because the increased risk estimate was based on different subtypes of cancer, indicating surveillance bias rather than a specific biologic mechanism.
Overall, Elfström et al. report lower risk estimates for GI cancer than previous studies. They attribute this difference to their use of biopsy data to identify patients with celiac diseases—all patients in Sweden diagnosed with celiac disease undergo biopsy analysis, which reduced selection bias in this particular study. Also, most patients included in this study were diagnosed with celiac disease in the late 1990s or after 2000. The increased use of serologic analyses to diagnose celiac disease during this time period meant that more patients with few or no symptoms of the disease were included in the analysis.
In an editorial that accompanies this article, Rupa Mukherhee et al. state that “up to this point, the prevailing belief has been that patients with severe, inflammatory celiac disease are likely to have a greater potential for malignancy, including small intestinal adenocarcinoma, than patients with milder forms of inflammation or latent celiac disease.” They state that, as a consequence, clinicians have lacked guidance about how to predict or reduce cancer risks, such as through dietary advice or celiac disease–specific cancer screening strategies.
Ludvigsson pointed out that the study was unable to include factors such as body mass index and smoking—important risk factors for cancer—because these data were not recorded in the registries analyzed. Mukherjee says that longer-term studies are needed, along with those to assess risk of other solid organ malignancies, in a similar prospective and population-based manner.
More Information on Celiac Disease:
Read the article online.
Elfström P, Granath F, Ye W, et al. Low risk of gastrointestinal cancer among patients with celiac disease, inflammation, or latent celiac disease. Clin Gastroenterol and Hepatol 2012;10:30–36.
Read the accompanying editorial.
Mukherjee R, Kelly CP, Leffler DA. Gastrointestinal cancer in celiac disease: “the first days are the hardest days, don’t you worry anymore?” Clin Gastroenterol and Hepatol 2012;10:4–6.
Read a related Gastroenterology article.
Mårild K, Stephansson O, Montgomery S, et al. Pregnancy outcome and risk of celiac disease in offspring: a nationwide case-control study. Gastroenterology 2012; 142:39-45.e3.
Kristine Novak, PhD, Science Editor
agajournals.wordpress.com 
Dear Kristine,
Thanks for commenting our our paper about GI cancer in celiac disease.
I welcome more studies in this field, but hope that they focus on mechanisms rather than exactly replicating our paper. It is unlikely that future studies (even those with long follow-up) will find substantially different hazard ratios for GI cancer than in our study for a number of reasons.
1. Although lack of data on smoking and body mass index is a weakness in our study, adjustment for these variables will probably not change the risk estimates for GI cancer. The reason is that the risk pattern (very high risk for GI cancer independently of the mucosal appearance in the first year and no risk increase at all beyond the first year) cannot reflect dramatic changes in either smoking or body mass index. Instead this pattern probably reflects another factor, and we suggest it is confounding by indication.
2. We presented data on follow-up divided in three groups 0-1 years; 1-5 years, and more than 5 years. However the last category includes patients followed for 5 years, but also those being followed for 25 years. The HR in this category is therefore a combination of follow-up between e.g. 5-10 years and 20-25 years etc. Although we did not present data on duration of follow-up in our paper in CGH, the follow-up in this study exceeded 25 years as can be seen in a similar paper on breastcancer (range of follow-up was 0-26 years) [Ludvigsson JF, West J, Ekbom A, et al: Reduced risk of breast, endometrial, and ovarian cancer in women with celiac disease. Int J Cancer ePub 2011.] Since the HR beyond 5 years of follow-up was fully neutral in our celiac-GI-paper, the only way celiac disease could be associated with positive HRs more than 25 years after CD diagnosis, would be if the HR first decreased below 1 between say 5-10 years after celiac disease diagnosis, only to increase substantially after more than 25 years of follow-up.
On the behalf of the authors,
Jonas F Ludvigsson
Fulbright Research Scholar, Mayo Clinic, Rochester, US
Chairman, Swedish Society of Epidemiology
Pingback: New Insights into Celiac Disease and GI Cancer Risks « G-Free Foodie