Statistics from Altmetric.com
We read with interest the article by Fidder et al1 reporting, in a single-centre cohort study, the frequency of severe adverse events (SAEs) in 734 infliximab- and 666 non-infliximab treated patients with inflammatory bowel disease (IBD) followed-up in the long term (median 58 and 144 months, respectively). While no differences between the two groups were detected in terms of SAEs, mortality, infection rate and malignancies, we do believe that the type of cancer observed deserves further consideration. A comparable cancer incidence was detected in the two groups, confirming previous studies.2 3 However, while in the non-infliximab-treated group 8 out of the 666 (1.2%) patients had a newly diagnosed colorectal cancer, no patients in the infliximab group had a new diagnosis of colorectal cancer (0%) during the study period. In the infliximab group, one metastatic adenocarcinoma of unknown origin was, however, detected in one patient with primary sclerosing cholangitis (PSC), defined as “most probably from intestinal origin”. The difference is most probably related to an occasional observation from a single-centre cohort study. However, we were impressed by this finding, as in our multicentre matched pair study including 808 patients with Crohn’s disease (CD) treated (n = 404) or not treated (n = 404) with infliximab, matched for clinical variables, a comparable frequency of neoplasia was observed from June 1999 to October 2004 (2.22% vs 1.73%, respectively; p = NS).2 However, when focusing on colorectal cancer, while among the 404 non-infliximab-treated patients with CD, three had a diagnosis of colorectal adenocarcinoma (caecum 2, rectum 1), none of the 404 infliximab-treated patients had a diagnosis of colorectal cancer.2
While providing no evidence of a cause–effect relationship between antitumour necrosis factor (TNFα) treatment and colon cancer, the data of these studies suggest that the frequency of colorectal cancer in infliximab-treated patients with IBD may be relatively lower than expected. Taking together all the reported data (clinical trials, retrospective analyses, case reports) published in 2007, the frequency of colorectal cancer was low (<0.5%) in infliximab-treated patients.3 One of the main biases in this analysis is the incidence of unknown pre-existing colorectal cancers before infliximab.
In contrast to IBD, a meta-analysis considering placebo-controlled trials in rheumatoid arthritis (RA) reported an increased risk of malignancies in patients receiving two anti-TNFα treatments (infliximab or adalimumab) for ⩾12 weeks (anti-TNFα vs placebo: 35/3493; 1% vs 3/1512, 0.19%, respectively). The pooled odds ratio for malignancy was 3.3 (95% CI 1.2 to 9.1) and the number needed to harm was 154 (95% CI 91 to 500) for one additional malignancy within 6–12 months. In RA, colorectal adenocarcinoma was detected in 3/3493 patients treated with an anti-TNFα antibodies (2 infliximab, 1 adalimumab), unspecified gastrointestinal adenocarcinoma in 2 (both adalimumab) and in none of the 1512 patients receiving placebo.4 These discrepant findings between patients with IBD and RA may be related to the different study design, sample size, anti-TNFα treatment and, mainly, to disease differences in terms of the possible role of anti-TNFα treatment in colon cancer prevention.
Supporting a relationship between anti-TNFα and colorectal cancer in IBD but not in RA, recent experimental evidence suggests that blocking TNFα in mice reduces colorectal carcinogenesis associated with chronic colitis5 and therefore chronic intestinal inflammation, as observed in IBD. In a mouse model of ulcerative colitis, administration of azaxymethane (AOM) followed by dextran sulfate sodium (DSS) caused severe colonic inflammation and multiple colonic cancers.5 Treatment of wild-type (WT) mice with AOM and DSS increased TNFα expression and the number of leucocytes expressing its major receptor (p55) infiltrating the colonic mucosa and submucosa. This was followed by multiple colonic tumours. Administration of etanercept (a TNFα antagonist) to WT mice after AOM and DSS markedly reduced the number and size of tumours and leucocyte infiltration. It was concluded that TNFα is a crucial mediator of the initiation and progression of chronic inflammation-mediated colon carcinogenesis.5 As blocking of TNFα reversed carcinoma progression, even after established carcinoma, the authors concluded that targeting TNFα may be useful for treating colonic cancers, particularly those arising from chronic inflammation, as in IBD.5 Although these observations represent only indirect preliminary suggestions for a possible relationship between TNFα and colitis-associated colon cancer, we believe that this hypothesis needs to be investigated further.
Funding “Fondazione Umberto Di Mario”, Largo Marchiafava 1, Rome.
Competing interests None.
Provenance and peer review Not commissioned; not externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.