P291. Clinical course of perianal fistulas in Crohn's disease: a retrospective study
L. Alessandroni1, C. Papi2, L.G. Papparella3, M.C. Addarii1, A. Kohn1, 1San Camillo-Forlanini Hospital, Rome, Italy, 2San Filippo-Neri Hospital, Rome, Italy, 3Campus Bio-Medico, Rome, Italy
Perianal fistulas are frequent complications of Crohn's disease (CD) that can result in significant morbidity and costly medical therapy. Little is known about risk factors for negative outcomes (NOs). Aim of the present study is to retrospectively determine the NOs, defined as defunctioning stoma or proctectomy, for these patients.
We selected all CD patients with perianal fistulas that were diagnosed and classified with examination under anesthesia at the Surgical Department-San Camillo Forlanini Hospital from 1980 to 2012. Recto-vaginal and recto-urethral fistulas were excluded. The follow-up was calculated from diagnosis to the onset of a NO or to the end of observation. Patients were divided in 3 cohorts according to the time of diagnosis (A: 1980-'89, B: 1990-'99, C: 2000-'12).
229 (47% females, median age 34 [range 9–74]) patients with perianal fistulas were analyzed; 19 with recto-vaginal or recto-urethral fistulas were excluded. Out of 210 patients: 56 (27%) had ileal disease, 55 (26%) colonic, 98 (47%) ileocolonic involvement. Complex fistulas were diagnosed in 181 patients (160 transphincteric, 10 suprasphincteric, 11 extrasphincteric), 30 had simple fistulas (5 superficial, 25 intersphincteric), 103 rectal involvement. The follow up was 72 months, within this period 99% of patients underwent 1 surgical procedure, 58% underwent 2 or more procedures. The cumulative probability of disease free from NOs was 0.7, not influenced by the type of fistula, by the age at diagnosis or by the sex, and was not significantly different for the three cohorts. Conversely, the risk of NO was significantly related to localization of disease (colonic vs ileal disease, p = 0.001) and to rectal involvement (p = 0.002). Two or more surgical procedures were related to an increased risk of NO compared to a single one (p < 0.001). The risk was not significantly reduced by immunosuppressive or biological therapy. At multivariate analysis, the risk of NO was independently predicted by the number of surgical procedures (P = 0.009) and colic disease (P = 0.04).
Our results suggest that in patients with perianal CD, the risk of NO is high, not influenced by the type of fistula but significantly related with disease localization, rectal involvement and need of more than one surgical procedure. Despite the introduction of biological and immunosuppressive treatments, the probability of negative outcomes did not change.