Transanal Ileal Pouch-Anal Anastomosis for Ulcerative Colitis - A Retrospective Single-Center Study Regarding Anastomotic Leakage and Episodes of Pouchitis in the Long-Term

Introduction: Colectomy with transanal ileal pouch-anal anastomosis (taIPAA) is a surgical technique that can be used to treat benign colorectal disease. Ulcerative colitis (UC) is the most frequent inammatory bowel disease (IBD) and although pharmacological therapy has improved, colectomy rates reach up to 15%. Objective of this study was to determine anastomotic leakage (AL) rates and treatment after taIPAA as well as short-and long-term pouch function. Patients and Methods: Data from a prospectively collected database of all patients undergoing taIPAA in our center between March 2015 and August 2019 was analyzed retrospectively. Patients with indications other than UC or with adjuvant chemotherapy following colectomy for colorectal carcinoma were excluded. Results: Of 23 met with a of 36 at the of Overall prevalence of AL was 10% with one early (11 days after and one late AL after In both patients pouches could be preserved with a multimodal approach based on endosponge therapy. Data on short-term pouch function could be obtained in 11 (55.8%) patients and was satisfactory in all cases. In the long-term we observed a pouchitis rate of 57.9% and a pouch failure rate of 5.3%. further studies


Introduction
Ulcerative colitis (UC) is the most common in ammatory bowel disease (IBD) and exceeds the incidence and prevalence of Crohn's disease in most countries around the world (1). Over the past few decades, pharmacologic therapy has improved, limiting the need for surgery to cases of refractory or steroiddependent disease, colorectal cancer or surgical emergencies such as toxic megacolon, perforation or lifethreatening haemorrhage (2). In the course of the disease, medical therapy refractory UC, requiring surgical treatment, leads to reconstructive surgery rates by restoring continuity of up to 15% (3). In this regard, the use of minimally invasive ileal pouch anal anastomosis (IPAA) is the favored technique (4,5).
Transanal total mesorectal excision (TaTME) has rst been established for sphincter-preserving rectal resection due to rectal carcinoma, providing better access and overview especially in obese patients or male individuals with naturally narrow pelvis (6, 7). Subsequently, transanal minimally invasive surgery (TAMIS) has also been implemented in operations for transanal IPAA (taIPAA) and, as already proven in various studies, also provides satisfactory surgical results and comparable complication rates to the pure transabdominal approach (8-10). Anastomotic leakage (AL) is a common complication in colorectal surgery with signi cant postoperative morbidity (11). In malignant as well as benign indications for surgery, AL results in chronic in ammation, stulae or stenosis, reduced quality of life and eventually anastomotic failure with permanent stoma (11)(12)(13). Additionally, in colorectal carcinoma, AL is associated with reduced disease-free and overall survival (14,15). Besides AL, pouchitis is the most frequent complication of IPAA signi cantly impairing pouch function and quality of life (16,17). AL in association to pouch complications might be of relevant impact for pouch failure after IPAA (9, 18).
The multimodal treatment procedures of AL management include -depending on the extent of ALantibiotic therapy, radiological or endoscopic interventions as well as reoperation with the intend to preserve continuity, as recently shown by different groups for AL for rectal cancer and in ammatory bowel disease (19)(20)(21).
Therefore, the aim of our retrospective study was to analyze patients after taIPAA for UC, regarding incidence and course of AL as well as episodes of pouchitis is and associated hospital readmissions, especially in the long-term.

Patient characteristics and operative details
All consecutive patients were analyzed retrospectively that had been subjected to a taIPAA in our department. All patients eligible for laparoscopic resection were offered the transanal approach. Those who chose taIPAA were prospectively recruited and gave written informed consent for the international LOREC® registry from December 2014 to January 2020. The study was performed in accordance with the Declaration of Helsinki and its amendments and an authorization has been granted by the Charité Instiutionals (Reg. No. 711/16). For clinical data such as sex, date of birth, preoperative data (therapy, BMI, extent of in ammation), operative data (operation time) and postoperative course as well as short and long-term follow-up data (complications after Clavien/Dindo; re-incurrences), patient data were collected anonymously.
From February 2015 to August 2019, taIPAA, performed in the context of a two or three staged restorative proctocolectomy (RPC), was applied in 23 patients for benign colorectal disease at our Department of Surgery, Campus Virchow Klinikum and Campus Charité Mitte at Charité -Universitätsmedizin Berlin. For this analysis, we only considered taIPAA in patients with ulcerative colitis (UC) and hence excluded three patients. Of those, one suffered from familial adenomatosis polyposis (FAP), one from Crohn´s disease and a third patient diagnosed with UC was excluded due to additional colorectal carcinoma with adjuvant chemotherapy after colectomy and therefore diverging postoperative care (Table 1 and Fig. 1).
Relevant data (clinical course, laboratory chemical parameters, hospital stay, readmissions) for our patient cohort were analyzed retrospectively. Postoperative complications were de ned according to Clavien/Dindo with a special focus on complications > II ( Table 2).
Occurrence of early and late anastomotic leakage within the rst 30 (eAL) and 31 to 90 (lAL) days or later was the primary endpoint. Secondary endpoints included pouch complications (pouchitis, stula, stenosis). These endpoints were assessed and evaluated at regular follow-up. One patient did not receive ileostomy closure at our Department of surgery, hence data on short-and long-term follow up could not be obtained.
Transanal Ileal Pouch-Anal Anastomosis (taIPAA) The whole operation procedure for UC was performed either as a two or three staged procedure with an interval of 3-6 months between the operations. While the rst operation comprised subtotal colectomy and terminal ileostomy, the second operation included ileal completion proctectomy and pouch-anal anastomosis with diverting loop-ileostomy. The last operation contained ileostomy closure. Before de nitive ileostomy closure endoscopy of the pouch function was performed to exclude severe in ammation (pouchitis) or AL. Ileostomy closure only took place in absence of in ammation and inconspicuous anastomosis region.
In one patient colectomy as well as pouch formation were performed at once due to an excellent remission of in ammation. One patient in our cohort did not undergo loop ileostomy due to severe adhesions. Basically, diverting loop ileostomy is standard at our institution, but may be considered optional in patients at very low risk for anastomotic leakage and/ or increased risk for complications. All operations for pouch formation were conducted in a hybrid technique including minimally invasive transabdominal approach and combined transanal approach.

Data collection
Data was collected in a prospective database for patients receiving taIPAA and analyzed retrospectively for perioperative complications and postoperative clinical course. For all patients the subsequent data were analyzed: demographics (age, gender), BMI, American society of Anaesthesiologists score (ASA), comorbidities, preoperative course of disease, operative details (Table 1), postoperative morbidity and mortality, complications with special focus on AL (time of diagnosis, treatment approach, treatment duration) ( Table 2 and 3) short-term ( rst 60 days after ileostomy closure) ( Table 4) and long-term (1 to 5 years after ileostomy closure) pouch function (Table 5), pouchitis (clinical symptoms, con rmed via endoscopy), stula, stenosis and pouch failure (table 6).

Statistical analysis
Statistical analysis was performed using Microsoft Excel. In descriptive statistics continuous variables were reported as mean or median values with range. Categorical variables were quanti ed using frequencies and percentage. Follow-up period started at the day of last operation.

Patient characteristics and operative details
Gender distribution presented more male patients (male 65%; female 35%) with median age of 36 years at the time of taIPAA. The median time from rst diagnosis of UC to subtotal colectomy was 36 months (1-408 months). 18 patients (90%) received a 3-step procedure as described earlier. One patient was operated in a 2-step procedure with colectomy and pouch-formation in the same operation. Another patient did not receive diverting loop-ileostomy due to multiple adhesions in the second step. Detailed information on patients' demographics and clinical features as well as on the operation are shown in Table 1.  Complications according to Clavien/Dindo > II took part over all three operation procedures, in total four patients (20%). We observed one patient with severe complications for colectomy (6.4%), two (10%) for pouch formation and three (15.8%) for ileostomy closure (Table 2).  The patient with lAL rst received two cycles of endosponge therapy followed by a transanal suture completing the multimodal therapy procedure. Recurrence of AL 51 days after transanal suture was again treated with four times of endosponge therapy and re-operation for performing a diverting loop-ileostomy after the rst cycle (Table 3). De nitive ileostomy closure could be achieved 380 days after last treatment for AL.  In the long-term follow-up, comprising 1 to 5 years, in total 11 (57.9%) patients out of 19 patients developed pouchitis, at a mean of 475 days (91-1231 days) after ileostomy closure (Table 5). Six (54.4%) patients were diagnosed within the rst year after ileostomy closure.
Additionally three (27.3%) patients developed stula, all in direct coincidence to ileal-pouch-anal anastomosis. Out of those stulas two were blind stulas and one was a pouch-vaginal stula with intermittent vaginal defecation.  Extraintestinal manifestation

Discussion
Transanal surgery has mainly been studied in TaTME as therapy for rectal carcinoma and provides comparable oncological outcomes, yet better pathologic resection status, shorter operation time, lower conversion rate and lower overall postoperative complication rate in comparison to laparoscopic TME (12,(22)(23)(24)(25)(26). Although the transanal approach for oncological surgery is well researched since rst performed for TME in 2010 (7) and offers improved surgical access to the otherwise often challenging, narrow pelvis (6), it is still not as widely established for benign indications such as UC. Currently, taIPAA is reported to provide comparable functional results and postoperative morbidity and mortality to the transabdominal approach with shorter operation time and conversion rate, but still study populations are relatively small and randomized controlled trials have not been conducted so far (8, 9,27).
In our study we observed a complication rate (Clavien/Dindo > II) of 10% for pouch formation, which is comparable to results reported recently for transabdominal IPAA (17) and even slightly lower than in two studies concerning taIPAA (8, 9). Yet, these studies also showed higher complication rates for transabdominal IPAA resulting in no signi cant difference between both approaches. AL occurred in 10% of all patients in our study. These results are consistent with ndings for IPAA before implementation of the transanal approach (18, 28), and similar AL rates have also been shown for taIPAA (8,9). Chandrasinghe et al. even reported a trend for lower AL rates in taIPAA compared to transabdominal IPAA, which did not reach statistical signi cance (8) and could not be veri ed in our study. Course of treatment for AL after taIPAA has not yet been reported in detail.
In our study, only one eAL and one lAL was observed. Treatment was aimed at preserving the anastomosis founding on endosponge therapy. While eAL could be treated only with endosponge therapy, lAL required a multimodal approach consisting of endosponge therapy, transanal suture and reopening of diverting loop-ileostomy. Our therapeutic approach relies on our experiences after TaTME for rectal carcinoma, where we developed a differentiated therapy strategy for AL (20). Endosponge therapy for AL after IPAA has not yet been widely researched, but the few studies existing show non-inferiority in comparison to conventional treatment and high rates of pouch preservation (21,29).
In our study, we did not observe any pouch failure due to AL with an overall pouch failure rate of 5.3%, which is also consistent with earlier ndings (8, 17). The only detected patient with pouch failure had an underlying severe pouchitis which was therapy refractory to pharmalogical treatment and therefore recommended for pouch removal and terminal ileostomy.
Pouchitis is one of the most frequent complications after IPAA (28, 30) and associated with signi cant morbidity and reduced quality of life (31,32). Pouchitis rates after IPAA reach up to 60% with increasing incidence during postoperative follow-up (16, 30,33) and may result in chronic pouch dysfunction and in severe cases pouch failure (34). The overall pouchitis rate was 11 (57.9%), with half of the patients developing symptoms within one year after last operation. Nevertheless, pharmacological treatment was successful in 72.7% resulting in a good functional outcome in long-term follow-up. There are still only few studies on treatment for acute and chronic pouchitis leaving therapeutic strategies with highly variable outcomes and lacking su cient scienti c evidence (35,36 Our study is limited by its retrospective design without option to rule out potential bias and restricted number of patients included. However, the monocentric design of our study ensured that the same experienced colorectal surgeons' team has performed all procedures, resulting in excellent comparability of postoperative outcomes.
In summary of our study, taIPAA for UC was a safe procedure at a low overall postoperative complication rate and satisfactory short-term and long-term outcome. AL rates were comparable to transabdominal IPAA with high rates of anastomosis preservation. Since taIPAA has been introduced only a few years ago, studies for long-term outcomes and evaluation of larger patient cohorts are necessary to fully understand possible advantages or downsides of this surgical approach. Authors contribution: AL collected the data and wrote the paper. FA and MB performed the surgery and supervised and corrected the manuscript. LD collected the data. AF and AA helped with data acquisition. FT supervised conservative patient treatment and corrected the manuscript. RS, SC, MK and JP supervised the study and corrected the manuscript. SGK designed and performed the research and wrote and corrected the paper. All authors have made substantial contributions to the study, including conception and design, the acquisition, analysis and interpretation of data, drafting or critical revision of the paper. All authors read and approved the nal manuscript.