TY - JOUR
AU - MD, Mark T. Osterman,
AB - Piton G, Cosnes J, Monnet E, et al. Risk factors associated with small bowel adenocarcinoma in Crohn's disease: a case-control study. Am J Gastroenterol. 2008;103:1730–1736. Piton and a number of his colleagues from the GETAID (Groupe d'Etude Therapeutique des Affections Inflammatoires du Tube Digestif) performed a retrospective case–control study to evaluate potential risk factors for small bowel adenocarcinoma (SBA) in patients with Crohn's disease (CD). Cases, defined as CD patients with SBA, were selected from medical records collected between 1986 and 2006 from the various university hospitals affiliated with GETAID. Controls, defined as CD patients without SBA, were selected from the MICISTA Registry, a tertiary medical database of inflammatory bowel disease (IBD) patients treated at Rothschild Hospital (in Paris) from 1974 to 2002 and subsequently at the Saint-Antoine Hospital (also in Paris). Of the 3492 total CD patients in this database seen between 1986 and 2002, 3 controls were selected per case and matched on age (±5 years), sex, location of disease during the first 6 months after diagnosis, and duration of disease (±5 years). A number of both nontreatment and treatment variables were examined as possible risk factors for the development of SBA. Nontreatment variables included history of smoking, cholecystectomy, or appendectomy, family history of IBD, date of symptom onset and date of diagnosis of CD, and extraintestinal manifestations. In addition, the disease behavior was classified according to the Montreal classification (B1 = inflammatory, B2 = stenosing, B3 = penetrating) and examined at 3 different timepoints: 5 years after CD diagnosis, 3 years before SBA diagnosis, and at SBA diagnosis. Treatment variables analyzed in this study included duration of use of 5-aminosalicylates, corticosteroids, 6-mercaptopurine/azathioprine, methotrexate, or infliximab, date and site of previous small bowel resection as well as disease duration without resection, and history of stricturoplasty or intestinal bypass. The authors identified 29 cases and 87 controls, with roughly equal numbers of men and women. Although the median duration of disease prior to SBA diagnosis was 11 years, 8 patients developed SBA within 1 year of diagnosis of CD, and 3 of these patients did not even have symptoms of CD prior to diagnosis of SBA. Among the nontreatment variables on univariate analysis, only disease behavior at SBA diagnosis seemed to show promise as a potential risk factor, as none of the cases versus 39% of controls had an inflammatory phenotype, while 45% of cases versus 24% of controls had a stenosing phenotype. Thus, although fibrostenotic disease may be a risk factor for the development of SBA, it may not manifest itself until it is too late (i.e., diagnosis of SBA). Among the treatment variables on univariate analysis, only use of salicylates for >2 years and small bowel resection showed significant associations. 5-Aminosalicylate use for >2 years was observed in 29% of cases versus 57% of controls, with an odds ratio (OR) of 0.29 (95% confidence interval [CI], 0.10–0.82; P = 0.02). Small bowel resection was also less frequent in cases, occurring in 24% versus 73% in controls, with an OR of 0.07 (95% CI, 0.01–0.32; P = 0.0007). Not surprisingly, cases had a significantly longer disease duration without resection (15.6 years versus 8.2 years; P = 0.02). On multivariate analysis, these same 2 treatment variables remained significantly associated with SBA, with an OR of 0.16 (95% CI, 0.03–0.79; P = 0.02) for 5-aminosalicylate use >2 years and an OR of 0.04 (95% CI, 0.01–0.28; P = 0.001) for small bowel resection. Finally, since cases and controls were not selected from the same population, a restriction analysis was performed on 11 cases and 33 controls seen at Saint-Antoine Hospital and similar results were reported. Comment The first case of SBA in CD was published by Ginzburg et al1 in 1956, and to date fewer than 200 cases have been reported in the literature.2 Palascak-Juif et al3 quantified this low absolute risk of disease in a recent publication using the GETAID data and found that the cumulative risk of SBA in CD patients with small bowel disease was 0.2% (95% CI, 0–0.8%) at 10 years and 2.2% (95% CI, 0.7–6.4%) at 25 years after CD diagnosis. However, although the absolute risk of SBA in CD is very low, the relative risk seems to be quite high. A recent meta-analysis of 5 population-based studies by Jess et al4 reported a 27-fold increased risk of small bowel cancer in CD (pooled standardized incidence ratio, 27.1 [95% CI, 14.9–49.2; range, 3.4–66.7]). Another meta-analysis, by Canavan et al,5 which included 4 of these 5 population-based studies and 4 additional hospital-based studies, calculated a similar 31-fold increased risk of small bowel cancer in CD patients (pooled relative risk, 31.2 [95% CI, 15.9–60.9; range, 3.4–100]). Moreover, Palascak-Juif et al3 showed that although the absolute risk of SBA in CD patients was low at 10 and 25 years after CD diagnosis, this cancer accounted for 25% and 45% of the gastrointestinal cancers in CD patients at 10 and 25 years, respectively. A disease with a rarity of this magnitude poses immediate problems in identification of risk factors simply due to the paucity of events. Identification of independent risk factors is even more challenging because model overfitting becomes an issue when trying to control for other variables when there are so few outcomes. It comes as no surprise, then, that only 2 other case–control studies have attempted to identify risk factors for SBA in CD patients.6,7 The first, by Lashner,6 included 7 cases of SBA matched to 28 controls and found that proximal small bowel disease, use of 6-mercaptopurine, and hazardous occupational exposures were associated with an increased risk of SBA. The other case-control study, by Solem et al,7 involved 9 patients with SBA matched to 18 controls and did not identify any significant risk factors for SBA. One population-based study, by Jess et al,8 in which only 4 patients had small bowel cancer, suggested that disease confined to the ileum may highly predispose CD patients to small bowel cancer. Finally, Palascak-Juif et al3 characterized SBA in CD by comparing SBA in 20 CD patients to de novo SBA in 40 patients without CD from the general population. The authors found that SBA in CD occurred at a younger age (median 43 versus 68 years), always occurred in an area of chronic inflammation (19 cases in the ileum, 1 in the jejunum), was diagnosed preoperatively much less commonly (5% versus 55%), and was associated with signet ring cells (35% versus none). The study by Piton et al, therefore, is valuable, as it represents the largest study of risk factors for SBA in CD patients. The authors showed that use of 5-aminosalicylates for >2 years and prior small bowel resection were protective against the development of SBA in CD patients. Even though a recent meta-analysis of 9 observational studies demonstrated a protective effect of 5-aminosalicylates in the prevention of colorectal cancer in patients with ulcerative colitis (pooled OR 0.51 [95% CI, 0.37–0.69])9 and another suggested that regular 5-aminosalicylate use may protect against colorectal cancer in patients with Crohn's colitis (OR 0.30 [95% CI, 0.05–1.17]; P = 0.10),10 it remains to be determined whether 5-aminosalicylates are truly protective against the development of SBA, as the efficacy data of 5-aminosalicylates for the treatment of small bowel inflammation in CD have been underwhelming. Also, patients whose small bowel disease is controlled with 5-aminosalicylates are likely to have milder disease and thus a lower risk of SBA. With respect to small bowel resection, it seems biologically plausible that resection of inflamed tissue would protect against the subsequent development of cancer if SBA in CD develops as a consequence of chronic inflammation, as some studies suggest.3,7,8 Prospective randomized studies evaluating potential risk factors for SBA in CD will almost certainly never be done, as outcomes are so rare and tend to occur after a median of 11–21 years.4,5,7,8 The question then remains: should we screen for SBA in CD patients? The answer at the present time (and likely in the future as well) is “no.” First, while the relative risk of SBA in CD patients is high, the absolute risk is very low and thus many patients would have to be screened unnecessarily to pick up the few cancers that eventually develop. Second, we do not have effective screening tests for SBA, and with a disease this rare, even a highly specific screening test will have a high false-positive rate, thereby leading to unnecessary worry and work-up, or potentially even surgery. Third, we do not understand the mechanism by which SBA arises, both in CD and in general. One study suggested that SBA may develop as a consequence of dysplasia.11 However, small bowel surveillance for dysplasia is no small undertaking, especially in patients with fibrostenotic disease. Finally, the diagnosis of SBA in CD is rarely made preoperatively, as symptoms are relatively nonspecific, and an alarming proportion of patients present with SBA within 1 year of CD diagnosis (some of whom did not even have prior symptoms of CD). For all these reasons, it is very difficult and cost-ineffective to screen for SBA in patients with CD. What is most sorely needed is a way to detect these cancers earlier. References 1. Ginzburg L, Schneider KM, Dreizin DH, et al.   Carcinoma of the jejunum occurring in a case of regional enteritis. Surgery.  1956; 39: 347– 351. PubMed  2. Feldstein RC, Sood S, Katz S. Small bowel adenocarcinoma in Crohn's disease. Inflamm Bowel Dis.  2008; 14: 1154– 1157. CrossRef Search ADS PubMed  3. Palascak-Juif V, Bouvier AM, Cosnes J, et al.   Small bowel adenocarcinoma in patients with Crohn's disease compared with small bowel adenocarcinoma de novo. Inflamm Bowel Dis.  2005; 11: 828– 832. CrossRef Search ADS PubMed  4. Jess T, Gamborg M, Matzen P, et al.   Increased risk of intestinal cancer in Crohn's disease: a meta-analysis of population-based cohort studies. Am J Gastroenterol.  2005; 100: 2724– 2729. CrossRef Search ADS PubMed  5. Canavan C, Abrams KR, Mayberry J. Meta-analysis: colorectal and small bowel cancer risk in patients with Crohn's disease. Aliment Pharmacol Ther.  2006; 23: 1097– 1104. CrossRef Search ADS PubMed  6. Lashner BA. Risk factors for small bowel cancers in Crohn's disease. Dig Dis Sci.  1992; 37: 1179– 1184. CrossRef Search ADS PubMed  7. Solem CA, Harmsen WS, Zinsmeister AR, et al.   Small intestinal adenocarcinoma in Crohn's disease. A case-control study. Inflamm Bowel Dis.  2004; 10: 32– 35. CrossRef Search ADS PubMed  8. Jess T, Winther KV, Munkholm P, et al.   Intestinal and extra-intestinal cancer in Crohn's disease: follow-up of a population-based cohort in Copenhagen County, Denmark. Aliment Pharmacol Ther.  2004; 19: 287– 293. CrossRef Search ADS PubMed  9. Velayos F, Terdiman JP, Walsh JM. Effect of 5-aminosalicylate use on colorectal cancer and dysplasia risk: a systematic review and meta-analysis of observational studies. Am J Gastroenterol.  2005; 100: 1345– 1353. CrossRef Search ADS PubMed  10. Siegel CA, Sands BE. Risk factors for colorectal cancer in Crohn's colitis: a case-control study. Inflamm Bowel Dis.  2006; 12: 491– 496. CrossRef Search ADS PubMed  11. Sigel JE, Petras RE, Lashner BA, et al.   Intestinal adenocarcinoma in Crohn's disease: a report of 30 cases with a focus on coexisting dysplasia. Am J Surg Pathol.  1999; 23: 651– 655. CrossRef Search ADS PubMed  Copyright © 2009 Crohn's & Colitis Foundation of America, Inc.
TI - Big Risk, Small Risk: Small Bowel Cancer in Crohn's Disease
JO - Inflammatory Bowel Diseases
DO - 10.1002/ibd.20888
DA - 2009-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/big-risk-small-risk-small-bowel-cancer-in-crohn-s-disease-mdWpC0oc5R
SP - 1434
EP - 1435
VL - 15
IS - 9
DP - DeepDyve
ER -