Long-Term Risk of GI Cancers Not Elevated in Celiac Patients



Overall, celiac disease patients are not at increased risk for gastrointestinal cancers beyond their first year after diagnosis, according to a review by Dr. Peter Elfström and colleagues in the journal Clinical Gastroenterology and Hepatology.

However, there is a significantly increased risk for GI cancer among celiac patients during that first year after the biopsy reveals the celiac diagnosis, according to this large, population-based study.

Dr. Elfström of Karolinska University Hospital in Stockholm and his colleagues reviewed data from 28,882 Swedish patients with celiac disease that was confirmed by biopsy showing villous atrophy in the small intestine (Clin. Gastroenterol. Hepatol. 2011 [doi:10.1016/j.cgh.2011.06.029]).

They also reviewed data on 12,860 patients whose biopsies did not show atrophy but did show inflammation of the small intestine, and another 3,705 patients with normal mucosa but serology indicative of celiac disease, including a positive IgA or IgG antigliadin, endomysial, or tissue transglutaminase test. The authors defined this third group as having "latent" celiac disease.

The study included children and adults. The percentage of children (aged 0-19 years) was 41% in the biopsy-confirmed celiac disease (villous atrophy) group, 10% in the inflammation group, and 25% in the latent celiac group.

Excluded from this study were all patients with a history of GI cancer before biopsy. Patients in the study were followed until the first incidence of GI cancer (according to the Swedish Cancer Registry), emigration, death, or the end of the study period on Dec. 31, 2007 – whichever came first.

Overall, during the entire follow-up period, the authors found 372 incident cancers in the villous atrophy–positive celiac disease patients, 347 cases in the inflammation-only cohort, and 38 cancers in the latent celiac group.

During the first year alone, there was a total of 125 incident cancers in the villous atrophy cohort, for a hazard ratio of 5.95 (95% confidence interval, 4.64-7.64).

In the inflammation-only group, the first year of follow-up after biopsy yielded 169 incident cancers (HR, 9.13; 95% CI, 7.19-11.6).

Similarly, latent celiac patients had a hazard ratio of 8.10 (95% CI, 4.69-14.0) for the first year of follow-up, with 23 incident cancers.

After the first year, however, the findings abruptly changed. "Excluding the first year of follow-up, patients undergoing small intestinal biopsy were at no increased risk of any GI cancer," wrote the authors, with the villous atrophy patients having a hazard ratio of 1.07 (95% CI, 0.93-1.23).

Similarly, the inflammation-only group’s HR fell to 1.16 (95% CI, 0.98-1.37), and the latent celiac disease group’s HR dropped to 0.96 (95% CI, 0.56-1.66).

This corresponded to an absolute risk for any GI cancer of 101 per 100,000 person-years in the group with villous atrophy, wrote the authors, or an excess risk over the nonceliac population of 2 per 100,000.

The absolute risk among the inflammation-only cohort was 192 per 100,000, whereas the latent disease group had an absolute risk of only 70 per 100,000.

The authors postulated that the lack of elevated risk after the first year could be due to patients starting and following a gluten-free diet. However, "this is unlikely, since a similar pattern was seen also in inflammation and latent" groups, they wrote," adding that "in Sweden, patients with inflammation and latent [celiac disease] have traditionally not received a gluten-free diet."

Additionally, "we cannot rule out that part of the early risk increase is due to confounding by indication, [that is,] that symptoms of a GI cancer may cause investigation leading to a diagnosis of [celiac disease]," they postulated, with a later diagnosis of cancer.

This study was supported by grants from several universities, foundations, and societies, including the Swedish Society of Medicine and the Swedish Celiac Society. The authors disclosed no individual conflicts relating to this study.

* This article was updated on Dec. 7, 2011.

Next Article: