Editor,—We read the paper by Hammel et al (Gut 1997;40:356–61) with interest. The authors studied antibodies to the p53 tumour suppressor gene product in serum of patients with colorectal cancer and reported a p53 positive patient (CC49) who developed colorectal cancer after a history of >10 years of ulcerous pancolitis. Hammel et alstated that p53 tumour suppressor gene mutations are frequent abnormalities in colorectal cancer. p53 mutations can lead to the accumulation of the p53 gene product in the cell and thus may trigger an antigen driven humoral response against p53. The authors showed that, as in other tumours with p53 gene mutations, antibodies to p53 are found frequently in patients with colorectal cancer (26% of patients) and that the serum level of these antibodies might parallel tumour progression.

As p53 gene mutations have been detected in colonic biopsy samples of patients with chronic inflammatory bowel disease (IBD),1we tested whether antibodies to p53 can be detected in serum of these patients. Antibodies were detected using an ELISA (Dianova, Hamburg, Germany) and positivity was confirmed by immunoblotting. We examined 76 patients with Crohn’s disease, 61 with ulcerative colitis and 206 first degree relatives (105 first degree relatives of patients with Crohn’s disease and 101 first degree relatives of patients with ulcerative colitis). Serum p53 antibodies were found in three of 61 patients with ulcerative colitis—all three patients had a 5–23 year history of pancolitis—and two of 76 patients with Crohn’s disease (one patient with severe pancolitis and one with colitis of the left hemicolon for over five years) (fig 1). There was no histological evidence of dysplasia or malignancy in serial colonic biopsy samples taken at the time of investigation. Antibodies were not detected in any of the first degree relatives.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Detection of antibodies to p53 in serum of patients with IBD.

Patients with severe ulcerative colitis or Crohn’s disease are at increased relative risk of developing colon cancer.2 In a meta-analysis of prospective studies performed between 1980 and 1993 approximately 15% of patients with IBD had developed colon cancer or dysplasia,3 but no information exists about the frequency of p53 mutations or antibodies in these patients.

To date, antibodies to p53 have been detected in patients with cancer and those at high risk of developing malignant disease.4-6 Recently it was reported that antibodies to p53 are present in patients at risk of developing lung cancer (heavy smokers) prior to the clinical diagnosis of lung cancer and the authors concluded that these antibodies could be used as an early marker for lung cancer.7 In the same way, we postulate that antibodies to p53 could identify those patients with IBD at risk of developing bowel cancer. Further prospective studies on larger groups of patients with IBD are urgently needed to test the predictive value of p53 antibody levels.