This study used meta-analytical techniques to compare the incidence of recurrence and the indication for reoperation in patients with Crohn's disease (CD) who underwent their first operation, due to perforating disease versus patients who underwent their first operation due to nonperforating disease.
Comparative studies published between 1988 and 2005 of perforating versus nonperforating CD were included. Using a random effects model, end points evaluated were recurrence of CD given as reoperation, and the indication for reoperation, i.e., perforating or nonperforating. Heterogeneity (HG) was assessed and a sensitivity analysis was performed to account for bias in patient selection.
Thirteen studies (12 nonrandomized retrospective, 1 nonrandomized prospective) reported on 3,044 patients, of which 1,337 (43.9%) had perforating indications (P group) and 1,707 (56.1%) had nonperforating indications (NP group) for surgery. The recurrence was found to be significantly higher in the P group compared to the NP group (HR 1.50, P= 0.002), with significant HG among studies (P < 0.001). The recurrence remained significantly higher in the P group compared with the NP group during sensitivity analysis of high-quality studies (HR 1.47, P= 0.005) and more recent studies (HR 1.51, P= 0.05), but still demonstrating significant HG (P= 0.08 and P < 0.001, respectively). At reoperation, concordance was found in the disease type of those patients re-presenting with perforating disease (OR 5.93, P < 0.001, without significant HG among studies P= 0.15) and those with nonperforating disease (OR 5.73, P < 0.001, with significant HG among studies P < 0.001). Concordance in disease type remained when considering only high-quality studies (P: OR 7.48, P < 0.001; NP: OR 7.48, P < 0.001) and more recent studies (P: OR 5.95, P < 0.001; NP: OR 5.95, P < 0.001), both not associated with HG among studies (P= 0.47 and P= 0.60, respectively).
The indication for reoperation in CD tends to be the same as the primary operation, i.e., perforating disease tends to re-present as perforating disease, and nonperforating as nonperforating. Also, perforating CD appears to be associated with a higher recurrence rate compared with nonperforating CD. However, because of significant HG among studies, further studies should be undertaken to confirm this finding.
1Department of Surgical Oncology and Technology, Imperial College London, London, U.K.; and 2Inflammatory Bowel Disease Centre, Yokkaichi Social Insurance Hospital, Yokkaichi, Mie, Japan
Reprint requests and correspondence: Takayuki Yamamoto, M.D., Inflammatory Bowel Disease Centre, Yokkaichi Social Insurance Hospital, 10–8 Hazuyamacho, Yokkaichi, Mie 510–0016, Japan.
Received December 14, 2006; accepted July 8, 2007.
CONFLICT OF INTERESTGuarantor of the article: Paris P. Tekkis, M.D., F.R.C.S.
Specific author contributions: Constantinos Simillis, Takayuki Yamamoto: performed meta-analysis and wrote the paper. George E. Reese, Satoru Umegae, Koichi Matsumoto, Ara W. Darzi, Paris P. Tekkis: reviewed the paper and advised on the methods and results of the paper.
Financial support: None.
Potential competing interests: None.