Intestinal cancer risk in Crohn’s disease: reporting, eligibility, and inference considerations in a meta-analysis

Dear Editor, We read with interest the meta-analysis by Giusto et al. on intestinal cancer risk in Crohn’s disease (CD). However, several reporting and methodological issues merit clarification because they may affect interpretation of the pooled estimates.1 First, inconsistencies across the abstract, main text, figures, and graphical abstract should be corrected to ensure a single coherent analytic version. The abstract reports pooled standardized incidence ratios (SIRs) of 2.20 for colorectal cancer (CRC) and 17.18 for small-bowel cancer (SBC), whereas the results, figures 2 and 6, and the graphical abstract report 2.21 and 17.51. Similar discordance appears in rectal cancer (1.75 vs 1.95) and ileocolonic CRC (4.13 vs 4.53). Section 3.3 also refers to the “remaining 11 studies,” whereas figure 6 and Q(12) indicate 13 effect estimates. These discrepancies complicate identification of the estimates underpinning the authors’ conclusions. Second, some study-level data appear difficult to reconcile and may affect the empirical basis of the pooled estimates. For Jess et al., table 1 reports a CRC SIR of 0.85 (95% CI 0.67-1.07), whereas figure 2 plots the same study at 1.72 (95% CI 0.89-3.32). For Cheddani et al., table 1 reports a CRC SIR of 1.2 with a 95% CI of 1.16-8.23; however, the primary study reported 1.20 (95% CI 0.57-2.52) in CD, whereas 1.16-8.23 referred to malignant lymphoproliferative disorders.2 A corrected extraction table aligned with the figures would facilitate independent verification and clarify the basis of the pooled estimates.