Reply to Dr. Selinger et al.'s letter.

We appreciate the input from Selinger et al. They questioned the utility of surveillance for patients with colonic Crohn's disease (CD), finding a cumulative incidence of only 2% of colorectal cancer (CRC) at 30 years. In Australia they found 5 cases of CRC among 327 patients, and we in Finland 4 among 208 patients with colonic CD. The follow up periods were 1977–1992 and 1986–2007, respectively. These two cohorts1,2 probably differ in many aspects, and direct comparisons cannot be made. Nevertheless, we had a regular scheme for surveillance, whereas Selinger et al. identified the patients from different sources. In our study, the relative risk of CRC did not differ between patients with ulcerative colitis and CD, being approximately two-fold. We agree that the risk of CRC is at most only marginally increased in CD. This was also supported in a recent Danish study3, where the relative CRC was even decreased in CD. Apart from cancer surveillance, there are other challenges in the management of CD. This was demonstrated in the study of Salinger et al., as more than 50% of the patients with CD required surgical resection at 15 years. Again, we agree with Salinger et al. that the surveillance should be carried out by taking individual risk of CRC into account. A routine endoscopic surveillance is of limited benefit, and should be instead targeted on high-risk patients, comprising those with long-lasting disease, active endoscopic or histological inflammation, post-inflammatory polyps, family history of CRC, or concomitant primary sclerosing cholangitis. In the remainder, we advocate tailor-made surveillance; whether it