DOP33 Predictors of ileo-rectal anastomosis, permanent ileostomy and recurrence of disease after total colectomy for Crohn’s Colitis: Results over 20 years
Rottoli, M.(1);Melina, M.(1);Tanzanu, M.(1);Romano, A.(1);Belvedere, A.(1);Parlanti, D.(1);Pezzuto, A.P.(1);Salice, M.(2);Rizzello, F.(2);Gionchetti, P.(2);Poggioli, G.(1)
(1)Alma Mater Studiorum University of Bologna. IRCCS Azienda Ospedaliero-Universitaria di Bologna., Surgery of the Alimentary Tract. Department of Medical and Surgical Sciences, Bologna, Italy;(2)Alma Mater Studiorum University of Bologna. IRCCS Azienda Ospedaliero-Universitaria di Bologna., IBD Unit. Department of Medical and Surgical Sciences, Bologna, Italy
Background
Patients with surgical Crohn’s colitis (CC) might either undergo a total colectomy (TC) and ileo-rectal anastomosis (IRA), or TC and end ileostomy. Among the latter group, a second-stage IRA is not always performed. The predictors of the different outcomes after TC (including the recurrence of disease at the level of the anastomosis) are yet to be identified.
Methods
Retrospective study including 354 patients undergoing TC for CC (2000-2019), with a minimum preoperative follow-up of 2 years in our centre. The mean postoperative follow-up was 67.5+/-49.8 months (62 patients lost). The primary end-points were to identify the predictors for the following outcomes:
a) IRA (87 cases, 24.6%) vs. end ileostomy (267 cases, 75.4%) at the primary colectomy.
b) second-stage IRA (80 cases, 39%) vs. no IRA (125 cases, 61%) at the last follow-up.
c) recurrence of the disease after IRA (167 patients).
Considering the large number of regressors and the risk of over-fitting, the least absolute shrinkage and selection operator (LASSO) method was used. A multivariate analysis was carried out using the preselected covariates. The analysis was conducted using logistic regression and cox regression for dichotomus and time-dependent outcomes. A p-value<0.05 was considered significant.
Results
a) Predictors against the IRA at primary TC: preoperative biologic exposure (OR=1.96, CI 1.15-3.33, p=0.014), Crohn’s rectal location (OR=4.17, CI 2.0-8.33, p<0.0001), perianal disease (OR=3.33, CI 1.75-6.25, p<0.0001), and low hemoglobin concentration (OR=1.26, CI 1.01-1.58, p=0.037).
b) Predictors of the risk of still having the ileostomy at the follow-up: age (OR: 1.02, CI 1.00-1.04, p=0.045), exposure to biologics before (OR 1.85, CI 1.11-3.03, p=0.017) and after (OR 1.79, CI 1.11-2.89, p=0.018) the TC; postoperative use of azathioprine was associated with a greater chance of a second-stage IRA (OR 2.66, CI 1.21-5.83, p=0.014).
c) Risk of anastomotic recurrence was 8.7% and 30.8% at 5 and 10 years. Significant predictors were: female gender (OR 2.38, CI 1.11-5.55, p=0.046), use of more than one biologic (OR 5.65, CI 2.18-9.89, p<0.0001) and worsening of symptoms needing for drug escalation (OR 5.51, CI 2.06-8.59, p=0.001) during the follow-up.
Conclusion
The location and severity of the disease at diagnosis predict the long-term behavior of the disease. The exposure to biologics, especially if multiple drugs are required, does not represent a risk for worse outcomes per se, but rather identifies a population at higher risk of permanent ileostomy due to more severe disease. These predictors might be implemented in the assessment of patients affected by CC, in order to identify the population at risk of permanent ileostomy.