Intestinal cancer in patients with Crohn's disease: A systematic review and meta-analysis

Motoi Uchino,Hiroki Ikeuchi,Keisuke Hata,Tomohiro Minagawa,Yuki Horio,Ryuichi Kuwahara,Shiro Nakamura,Kenji Watanabe,Masayuki Saruta,Toshimitsu Fujii,Taku Kobayashi,Ken Sugimoto,Fumihito Hirai,Motohiro Esaki,Sakiko Hiraoka,Katsuyoshi Matsuoka,Shinichiro Shinzaki,Minoru Matsuura,Nagamu Inoue,Hiroshi Nakase,Mamoru Watanabe

DOI: https://doi.org/10.1111/jgh.15229

Abstract:Background and aim: Although surveillance colonoscopy is recommended by several guidelines for Crohn's disease (CD), the evidence is insufficient to support the validity of this recommendation. Moreover, the efficacy of surveillance colonoscopy for anorectal cancer remains unclear. Therefore, we performed a systematic review of cancer in patients with CD before considering the proper surveillance methods. Methods: We conducted a systematic review and meta-analysis examining the incidence of intestinal cancer and a literature review to clarify the characteristic features of cancer in CD. We performed the systematic literature review of studies published up to May 2019. Results: Overall, 7344 patients were included in eight studies. The standardized incidence ratios (95% confidence intervals) of colorectal cancer (CRC) and small bowel cancer (SBC) were 2.08 (1.43-3.02) and 22.01 (9.10-53.25), respectively. The prevalence of CRC and SBC was 57/7344 (0.77%) and 17/7344 (0.23%), respectively, during a median follow-up of 12.55 years. Additionally, 54 studies reporting 208 anorectal cancer cases were identified. In patients with anorectal cancer, the prognosis for survival was 2.1 ± 2.3 years, and advanced cancer greater than stage T3 occurred in 46/74 patients (62.1%). Many more reports of anorectal cancer were published in Asia than in Western countries. Conclusion: Although we were unable to state a recommendation for surveillance for SBC, we should perform cancer surveillance for CRC in patients with CD. However, the characteristics of cancer may differ according to geography or race. We must establish proper and effective surveillance methods that are independently suitable to detect these differences.