Restorative versus non-restorative total proctocolectomy: Who doesn't pouch and how do they do?

David Liska: Supervision; writing – review and editing; writing – original draft. Arielle Kanters: Writing – original draft; writing – review and editing; supervision. Imran Khan: Data curation; formal analysis; writing – original draft; writing – review and editing. Scott R. Steele: Supervision; writing – review and editing; writing – original draft. Olga Lavryk: Supervision; writing – review and editing; writing – original draft. Jeremy Lipman: Supervision; writing – review and editing; writing – original draft. Nicholas G. Berger: Conceptualization; data curation; formal analysis; writing – original draft; writing – review and editing. Stefan D. Holubar: Conceptualization; formal analysis; supervision; writing – review and editing; writing – original draft; project administration.

Ethics approval was obtained from the Cleveland Clinic Institutional Review Board prior to performing the study with waiver of informed consent given its retrospective, minimal‐risk nature.

The authors have nothing to report.

We identified 821 patients with UC who underwent proctectomy and were included in our institutional NSQIP program: 689 (83.9%) underwent IPAA, and 132 (16.1%) underwent TPC‐EI. A subgroup analysis from 2019 to 2023, during which proctectomies were sampled at 100%, revealed that IPAA was performed in 207 (75.3%) patients and TPC‐EI in 68 (24.7%) patients. Of the 689 patients who underwent IPAA, 531 (77.1%) had 3‐stage, 141 (20.5%) had 2‐stage, and 17 (2.5%) had a modified 2‐stage restorative proctocolectomy. Of the 132 TPC‐EI patients, 69 (52.3%) underwent a 2‐stage and 63 (47.7%) a 1‐stage non‐restorative proctocolectomy. The baseline characteristics of the patients are shown in Table  1 . Patients in the TPC‐EI group were significantly older than those in the IPAA group (median age at surgery, 57 vs. 38 years; p  < 0.0001), whereas the sex distribution was similar between the two groups ( p  = 0.63). Patients with TPC‐EI had a higher median body mass index (BMI) of 27.6 vs. 25.5 kg/m 2 in the IPAA group ( p  < 0.0001). Comorbidities were more common in the TPC‐EI than the IPAA group: at least one comorbidity (68.8% vs. 44%, p  < 0.0001), and/or ≥ 2 comorbidities (20.9% vs. 7.4%, p  < 0.0001). Anaemia was similar between the groups, whereas malnutrition was more common in the TPC‐EI group ( p  < 0.0001). Although more patients in the TPC‐EI group received preoperative steroids (16.7% vs. 11.2%), the difference was not statistically significant ( p  = 0.08). However, the TPC‐EI group was more likely to have received biologics prior to proctectomy, while a minority of IPAA patients were on biologics at the time of pouch construction (18.2% vs. 6.8%, p  < 0.0001). Baseline characteristics of patients who underwent IPAA versus TPC‐EI ( N  = 821). Note : Figures represent frequency (proportion) or medians (interquartile range). None of the patients were on dialysis, had myocardial infarction within the past 6 months, or had rest pain/gangrene. Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease; IPAA, ileal pouch‐anal anastomosis; TPC‐EI, total proctocolectomy with end ileostomy. Anaemia = HCT < 35 or any transfusion within 72 h prior to surgery. Malnutrition = albumin  10% loss of body weight in 6 months prior to surgery. The reasons why patients with UC underwent TPC‐EI in lieu of IPAA are shown in Table  2 . The most common reason was patient preference (53%); of these, the most common (50%) rationale cited by patients as their motivation not to undergo IPAA was that they were content with their stoma and had no desire for restoration. Other common reasons included malignancy or dysplasia (10.6%) and incontinence or other pelvic floor disorders (9.1%). Other factors included obesity, advanced age, comorbidities, surgeon advice, deep infiltrating endometriosis and intraoperative abandonment of IPAA. Documented reasons UC patients underwent TPC‐EI in lieu of IPAA ( N  = 132). Note : Figures represent frequencies (proportions). Abbreviations: BMI, body mass index; IPAA, ileal pouch‐anal anastomosis. In terms of laboratory values (Table  3 ), albumin levels were lower in the TPC‐EI group than in the IPAA group ( p  = 0.001). However, the haematocrit and creatinine levels did not differ between the two groups. The operative variables are presented in Table  4 . In both groups, the most common indication for surgery was medically refractory disease (85.9% vs. 84.9% of IPAA and TPC‐EI patients, respectively; p  = 0.58), while neoplasia associated with UC was the indication in 15.9% vs. 14.1% of TPC‐EI and IPAA patients, respectively. The proportion of patients with ASA class > 2 was higher in the TPC‐EI group (76.5% vs. 44.8%, p  < 0.0001). The mean operating time was not significantly different between the two groups ( p  = 0.09). The proportion of patients with wound class 4 (contaminated, dirty or infected) was 5.2% vs. 1.6% ( p  = 0.04); any other concurrent procedure performed (22.9% vs. 10.6%, p  < 0.01); and intraoperative blood transfusion (13.7% vs. 4.6%, p  < 0.0001) were higher in the TPC‐EI group than in the IPAA group. Preoperative laboratory values of patients who underwent IPAA versus TPC‐EI ( N  = 821). Note : Figures represent frequencies (proportions). Abbreviations: AST, aspartate aminotransferase; BUN, blood urea nitrogen; WBC, white blood cells. Operative characteristics of patients who underwent IPAA versus TPC‐EI ( N  = 821). Note : Data are presented as frequencies (proportions) or medians (IQR). Abbreviations: ASA, American Society of Anesthesiology; CRC, colorectal cancer; RBC, red blood cells; TPC‐EI, total proctocolectomy with end ileostomy. The short‐term outcomes are presented in Table  5 . The overall postoperative complication rate was higher in the TPC‐EI group than in the IPAA group (37.1% vs. 19.2%, p  < 0.0001). There was no significant difference between the two groups in terms of complications classified as Clavien‐Dindo grade 3 or higher ( p  = 0.14). After adjusting for complications by excluding perineal wound complications (18.2%) and anastomotic leaks (4.5%), the adjusted overall rates in the TPC‐EI and IPAA groups were 22.7% and 17.9%, respectively ( p  = 0.23). Length of stay, return to the operating room, readmission and mortality rates were similar between the two groups. Short‐term (30‐day) surgical outcomes and morbidity in patients who underwent IPAA versus TPC‐EI ( N  = 821). Note : Data are presented as frequencies (proportions) or medians (IQR). Abbreviations: CPR, cardiopulmonary resuscitation; SBO, small bowel obstruction. Excluding anastomotic leaks and perineal wound dehiscence. Same patient counted only once. After PSM (Figure  1 ), 107 TPC‐EI patients were successfully matched with 188 IPAA patients. In the matched cohort, multivariable logistic regression analysis was performed to assess the overall complications, including pouch leaks and perineal wound complications (Table  6 ). In this model, TPC‐EI was a significant predictor of any complication (OR 2.84, 95% CI 1.37–6.08, p  = 0.006). In the adjusted complication model, which excluded pouch leaks and perineal wound healing complications (as these are unique to each operation), age (OR 1.04, 95% CI 1.002–1.082; p  = 0.04) and anaemia (OR 3.10, 95% CI 1.01–9.47, p  = 0.05) emerged as independent predictors of complications within 30 days. Love plot of covariate balance before and after 1:2 propensity score matching. Multivariable analysis of complications ( N  = 295). Abbreviations: ASA, American Society of Anesthesiology; BMI, body mass index; CI, confidence interval; NA, not applicable; NS, not significant; TPC‐EI, total proctocolectomy‐end ileostomy. Excluding pouch leaks and perineal wound complications.

Approximately 25% of patients with ulcerative colitis (UC) require surgery during their lifetime, most commonly due to medically refractory disease (80%) and less commonly for dysplasia or adenocarcinoma (20%) [ 1 , 2 , 3 , 4 ] When faced with surgery in the modern era, most patients with UC choose to undergo a restorative approach of total proctocolectomy (TPC) with ileal pouch‐anal anastomosis (IPAA) rather than a permanent end ileostomy (TPC‐EI). However, some patients undergo a non‐restorative approach with TPC‐EI and intersphincteric dissection (ISD) in lieu of a pouch. Both IPAA and TPC‐EI address the colorectal manifestations of UC and restore patient quality of life (QoL) for most patients, but there is a paucity of modern literature assessing the indications and patient profiles for patients undergoing either procedure, and the short‐term complication rates of the two procedures have not been widely described [ 5 , 6 , 7 , 8 , 9 ]. The American Society of Colon and Rectal Surgeons (ASCRS) Clinical Practice Guideline and the European UC Crohn's and Colitis Organization's (ECCO) UC Surgical Guidelines both recommend IPAA as the procedure of choice, while simultaneously recognizing that end ileostomy is a viable option for selected patients, as both procedures have comparable complications and patient‐reported outcome measures, highlighting the importance of shared decision‐making [ 3 , 10 ]. Modern literature comparing IPAA and TPC‐EI is scarce. In particular, perineal wound complications after proctectomy have received renewed attention with the recent TOpClass classification, yet contemporary short‐term morbidity data after TPC‐EI remain limited [ 11 , 12 ]. Therefore, we aimed to (a) report the reasons why patients with UC underwent TPC‐EI instead of IPAA, (b) describe the incidence of perineal wound complications and (c) compare the short‐term outcomes of TPC‐EI and IPAA at our center. Previous reports have demonstrated that perineal wound dehiscence after intersphincteric dissection for IBD (including both Crohn's disease and UC) is common, with only 70% healed at one year, and we aimed to quantify its contribution to overall short‐term morbidity in UC specifically [ 13 ]. We hypothesized that TPC‐EI was independently associated with an increased overall short‐term (30‐day) postoperative complication rate.

Overall, we observed that despite the widespread adoption and acceptance of IPAA for patients with UC, TPC‐EI is still performed in up to 25% of patients with UC, with patient preference being the most common reason for forgoing IPAA. In terms of complications, although the overall (unadjusted) complication rates were higher in the TPC‐EI group, this was driven by procedure‐specific perineal wound complications, which occurred in almost 20% of patients. Apart from leaks, which were uncommon at our center, and perineal wound healing problems, which were quite common, the procedures were equally safe in terms of short‐term complications. These perineal wound data are timely given the renewed focus on perineal morbidity following the TOpClass classification [ 11 , 12 ]. Both IPAA and TPC‐EI are associated with specific postoperative complications; AL is a concern after IPAA, whereas perineal wound complications occur after TPC. Hasab et al. compared these approaches in patients > 65 years of age and found that patients undergoing TPC with ileostomy were generally older and had higher rates of preoperative steroid use and significant weight loss (> 10% over 6 months). IPAA was not associated with an increased risk of postoperative morbidity, although this group had a higher readmission rate than the TPC‐EI group; however, NSQIP does not capture perineal wound complications, a key limitation of this study [ 17 ]. A study of patients undergoing 1‐stage TPC for UC found a non‐healing perineal wound rate of 18.5%, which is very similar to that observed in our study [ 18 ]. In terms of the motivation for patients to undergo TPC‐EI in lieu of IPAA at our center, patient preference, rather than advanced age or other factors, largely drives the decision. In fact, 53% of the patients in our study who were presumably candidates for IPAA underwent TPC‐EI; advanced age/comorbidities were reasons for not undergoing the IPAA in only 5.3% of the TPC‐EI patients. We previously reported 44 patients with UC who underwent TPC‐EI for UC; factors contributing to patient and surgeon choice of a non‐restorative TPC included increased age ( ≥ 65) in 41%, preference for an ileostomy due to lifestyle reasons (such as occupation) in 30%, comorbidities in 25%, incontinence in 23%, pelvic irradiation in 6.8%, obesity ( ≥ 30 kg/m 2 ) in 6.8% and ≥ 1 factor in 41% [ 2 ]. In terms of age, a cross‐sectional NSQIP study by Cohan et al. examined the impact of patient age on procedure choice and found patients > 50 years old were more likely to undergo TPC‐EI after adjusting for age, sex, smoking, BMI, frailty and ASA class (all p  < 0.001); authors reported a novel decision aid and stressed the importance of patient engagement and the importance of educational aids which describe complications and lifestyle implications of each operation [ 19 , 20 , 21 ]. In terms of risk factors for complications, as expected by the 20% perineal wound complication rate after EI compared with the 4.5% leak rate after IPAA, the PSM analysis demonstrated TPC‐EI was independently associated with unadjusted complications, whereas anaemia and increasing age were significant predictors of both adjusted and unadjusted complications. These findings underscore the importance of staging TPC‐EI, as is typically done for IPAA, as an opportunity for preoperative optimization, particularly in optimizing risk factors (e.g., anaemia and obesity), and highlight the protective benefits of staging in high‐risk patients [ 3 , 10 , 22 ]. Importantly, staging is a non‐restorative approach that also allows patients to experience life with an ileostomy. Anaemia was independently associated with complications after both TPC‐EI and IPAA. Nfonsam et al. used NSQIP data to show that anaemia was an independent predictor of overall complications, serious morbidity and increased LOS [ 23 , 24 ]. Malnutrition is also a risk factor for complications. Several studies have shown an association between weight loss, hypoalbuminemia and preoperative steroid use and adverse postoperative outcomes [ 22 , 25 , 26 , 27 , 28 ]. We previously reported a strong association between hypoalbuminemia and postoperative intra‐abdominal septic complications in patients undergoing ileocolic resection for Crohn's disease (OR 1.76 95% CI, 1.03–3.01; p  = 0.04) [ 27 ]. Preoperative anaemia and nutritional optimization significantly improve postoperative outcomes in UC patients [ 29 , 30 ]. However, biologics do not appear to be a risk factor for complications after surgery for IBD, as shown by the PUCCINI study [ 31 ]. We examined the role of postoperative steroid weaning after IPAA and found that both 2‐stage IPAA (compared with 3‐stage) and postoperative oral steroid tapering were independently associated with pelvic sepsis (OR 2.2, 95% CI 1.01–4.7, p  = 0.05 and OR 2.3; CI 1.06–5.1; p  = 0.04, respectively). Quality of life after IPAA has been widely reported. Studies have shown that QoL in ostomates is marginally lower, although not significantly lower than that in the general population, and the longer a person lives with an ostomy, QoL is seen to improve [ 32 ]. Interestingly, patients with UC who choose to undergo a TPC‐EI compared to those who undergo IPAA have similar QoL scores [ 33 ]. It has previously been shown that living with a stoma for 40 months is significantly associated with a higher QoL. Men were more likely to have higher QoL scores than women. While some studies have reported minimal effects on QoL, it is important to note that many patients with ostomies and pouches experience sexual dysfunction due to both functional and psychological factors. Importantly, while short‐term complication rates may appear similar between the two approaches, the nature and long‐term consequences of these complications differ substantially. Perineal wound complications after TPC‐EI, although common, are self‐limited and most ultimately heal, whereas pouch‐specific complications such as anastomotic leak, chronic pouchitis and pouch failure may necessitate reoperation or permanent diversion, with significant impact on long‐term QoL. This distinction is critical for preoperative counselling and shared decision‐making, as patients must weigh not only the likelihood of complications but also their potential severity and reversibility when choosing between IPAA and TPC‐EI. Our study had some limitations. Although most of our data were obtained using the validated ACS‐NSQIP methodology, some variables were abstracted by retrospective chart review; thus, our study was subject to selection bias. While propensity score matching was used to adjust for measurable baseline differences and diminish the influence of patients who were not IPAA candidates, it cannot account for unmeasured confounders influencing surgical decision‐making for those considering one or the other procedure, and residual bias may persist. Our 30‐day outcome window may not capture delayed anastomotic complications; Moojen et al. reported that 21% of pouch leaks were diagnosed beyond 21 days, particularly with a 3‐stage approach, which could underestimate the true IPAA complication rate [ 34 ]. Additionally, ileostomy closure rates following IPAA were not captured; a minority of patients (< 5%) may not undergo reversal, typically due to occult/delayed leaks, representing additional morbidity not reflected in 30‐day outcomes. Additionally, detailed characterization of perineal wound complications by type, grading and management was beyond the scope of the present study. A limitation of our methodology introduced by using ACS‐NSQIP was the difference in proctectomy sampling before (30%) and after (100%) 2019, which was chosen to utilize readily available data and given the high volume of surgery for UC at our center, limiting the burden of retrospective chart review. The retrospective nature of our study also introduced a form of recall bias, which somewhat limited our ability to discern the rationale for surgeons and patients selecting TPC‐EI over IPAA. Referral bias may have influenced our results, as our results were obtained from a specialized, high‐volume IBD center, which may limit the generalizability of our findings to other settings with varying levels of expertise and resources. Another limitation is the limited number of TPC‐EIs, which precluded multivariable analysis of perineal wound complications. A larger study of TPC‐EI is planned to elucidate independent risk factors for perineal wound complications in IBD, including the role of proctectomy staging. Further study of the predictors, classification and management of perineal wound dehiscence in both UC and CD is ongoing. Finally, it was beyond the scope of this study to include patient‐reported outcome measures, such as quality of life. While short‐term morbidity is an important component of preoperative counselling, we acknowledge it represents only one aspect of the shared decision‐making process, which must also incorporate long‐term functional outcomes and quality of life [ 8 ]. Despite these limitations, our data represent one of the largest examinations of what may reasonably be expected after both restorative and non‐restorative TPC in the modern era and provide clinically useful insights to guide surgical decision‐making and inform UC patients to allow for a well‐educated, shared decision regarding their options. In conclusion, after TPC for UC, the overall short‐term complication rates were similar between IPAA and TPC‐EI, although perineal wound complications occurred in 18.2% of TPC‐EI cases. Across both surgical groups, anaemia and older age emerged as predictors of adjusted postoperative complications, underscoring the critical role of thorough preoperative optimization in improving outcomes. Both IPAA and TPC‐EI are valid surgical options for UC. Notably, nearly one‐quarter of proctectomies at our high‐volume center were non‐restorative in recent years, and the choice was primarily driven by patient preference, underscoring the importance of individualized counselling, shared decision‐making and preoperative optimization to enhance short‐term surgical outcomes.

Originality: This paper is original and has not been published previously. An early version of this work was presented at the 2021 American Society of Colon and Rectal Surgeons Annual Scientific Meeting on April 25th, 2021, as an Oral Quickshot presentation. Use of AI: This work was written by the authors without the use of generative AI. Paperpal was used to edit the final draft for grammatical and punctuation errors.

We performed an institutional review board (IRB)‐approved retrospective case–control study between 2005 and 2023. Using our institutional American College of Surgeons National Surgical Quality Improvement Program (ACS‐NSQIP) proctectomy dataset and International Classification of Diseases version 9 and 10 (ICD‐9/10) codes, we identified all patients with UC (ICD‐9/−10: 556.x, K51.x, respectively) who underwent proctectomy at our center between 2005 and 2023. The ACS‐NSQIP is a multi‐institutional international registry of patients undergoing surgery in 720 participating hospitals, with > 1 million patients as of 2022. Details of data abstraction, validation methods and variable definitions are available on the ACS‐NSQIP website ( https://www.facs.org/quality‐programs/acs‐nsqip/program‐specifics/participant‐use ). The ACS NSQIP and the hospitals participating in the ACS NSQIP are the source of the data used herein; they have not been verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors. At Cleveland Clinic, 30% of colorectal cases were sampled between 2005 and 2018; after 2019, colorectal cases were captured at a 100% sample rate. The resultant dataset was used to identify all patients with UC who underwent proctectomy, and each patient was then stratified into the IPAA and TPC‐EI groups based on Current Procedural Terminology (CPT) codes (Table  S1 ) [ 14 , 15 ]. Patients were also stratified according to whether proctectomy was performed at the time of colectomy or staged and performed at a later date (i.e. completion proctectomy with or without IPAA). We reported data on baseline preoperative characteristics, preoperative laboratory values, operative variables and 30‐day surgical outcomes. Several composite variables relevant to inflammatory bowel disease (IBD) patients were constructed: anaemia was defined as a haemoglobin level < 12 g/dL or transfusion within 72 h of surgery; malnutrition was defined as an albumin level  10% weight loss within 6 months preoperatively; indications were neoplastic if a preoperative diagnosis of dysplasia or colorectal cancer was identified, and all others were designated as medically refractory UC as the indication for surgery. All other definitions are available from ACS‐NSQIP. In addition to our robust internal NSQIP data, we conducted a focused retrospective chart review to supplement the dataset by capturing information that was not specified in the NSQIP dataset. Corticosteroid and biologic use were defined as systemic administration within 30 and 90 days before surgery, respectively. Surgeon and patient preoperative rationale for the choice of a non‐restorative approach were collected for all patients who underwent TPC‐EI. We defined perineal wound complications as any wound dehiscence, infection or haematoma within 30 days of surgery. In the IPAA patients, anastomotic leak (AL) was defined as a defect in intestinal wall integrity at the ileoanal anastomotic site resulting in a communication between intra‐ and extraluminal compartments or a peri‐anastomotic pelvic abscess [ 16 ]. All complications were categorized according to the Clavien‐Dindo system [ 13 ]. Our primary outcome was the overall 30‐day complication rate after proctectomy for UC. Given that each procedure has unique complications (IPAA, anastomotic leak and TPC‐EI, perineal wound dehiscence), our secondary outcome was an adjusted complication rate, excluding these specific complications, to identify predictors of complications after either procedure. Other outcomes included total operating time in minutes, specific complications (including anastomotic leak and perineal wound dehiscence), length of hospital stay, return to the operating room, readmission and mortality, all within 30 days of proctectomy. Potential differences between the groups (IPAA and TPC‐EI), including preoperative characteristics, preoperative laboratory values, operative variables and outcomes, were assessed using univariate analysis as appropriate. To minimize baseline differences between patients undergoing TPC‐EI and those undergoing IPAA, we first performed propensity score matching (PSM) using a 1:2 nearest‐neighbour matching algorithm without replacement. Propensity scores (PS) were estimated using a logistic regression model incorporating clinically relevant covariates: age (continuous), body mass index (BMI; kg/m 2 , continuous), preoperative steroid use (yes/no), presence of any comorbidity (yes/no), American Society of Anaesthesiologists (ASA) class 3 or 4 (yes/no), haematocrit (continuous) and albumin levels (continuous). A calliper width of 0.1 standard deviations of the logit of the PS was applied to optimize the matching quality while minimizing residual confounding. Following PSM, the matched cohort was used for univariate and multivariable logistic regression analyses to identify independent predictors of complications. The first model examined clinically important variables independently associated with the overall (unadjusted) complication rate, with TPC‐EI versus IPAA as the primary predictor variable. Since patients with IPAA are not at risk of perineal wound complications and those with TPC‐EI are not at risk of IPAA leak, we performed a second adjusted model, excluding pouch leaks and perineal wound healing complications. Odds ratios (ORs) were reported with 95% confidence intervals (CIs). Statistical significance was set at p  < 0.05. All analyses were performed using R version 4.0.0.

Table S1. Current procedural terminology (CPT) codes for IPAA and TPC‐EI. [13, 14]. IPAA, ileal pouch‐anal anastomosis; TPC‐EI, total proctocolectomy with end ileostomy; DLI, diverting loop ileostomy. *There is no CPT code specific to lap completion proctectomy with IPAA; thus, surrogate codes (45397, lap proctectomy and colonic J‐pouch) were used and confirmed through chart review. **There is no CPT code specific to modified 2‐stage IPAA, but it was confirmed by chart review. ***The CPT code for lap completion proctectomy with abdominal perineal includes with colonic J‐pouch and colostomy, confirmed by chart review to be after a prior total abdominal colectomy with ileostomy (CPT code 44150 [open] and 44210 [laparoscopic]). None of the patients had 45499, 45999 (unlisted laparoscopic procedure, rectum) or 46999 (unlisted procedure, rectum) codes. Data S1. STROBE Statement—Checklist of items that should be included in reports of case–control studies.