Altering the Traditional Approach to Restorative Proctocolectomy After Subtotal Colectomy in Pediatric Patients

Abstract

Purpose:

Restoration of intestinal continuity by ileal pouch-anal anastomosis (IPAA) following subtotal colectomy may not require a temporary, protective ileostomy. Diversion contributes to patient discomfort, cost, and additional operative risk at the time of subsequent reversal. We compared the outcomes of pediatric patients undergoing modified two-stage to three-stage IPAA after recovering from subtotal colectomy.

Methods:

We reviewed children (age <18) who underwent IPAA creation for ulcerative or indeterminate colitis from January 1, 2007 to December 31, 2017. Patient characteristics, operative details, 30-day complications, and postoperative length of stay (LOS) were abstracted. Total LOS for the three-stage group included both the IPAA and the ileostomy reversal operations. Univariate comparisons between patients undergoing modified two-stage and three-stage operations were performed.

Results:

A total of 43 patients underwent IPAA after subtotal colectomy; 32 (74%) underwent a three-stage approach, and 11 (26%) had a modified two-stage approach. Operative approach was laparoscopic in 33 (77%), planned open in 9 (21%), and converted to open in 1 (2%). Single-incision technique was used in 12 of 33 (36%) laparoscopic cases. Modified two-stage procedures had shorter total median LOS (7 days versus 9 days, P = .005). Incidence of postoperative leak, readmission, return to the operating room, and maximum 30-day Clavien–Dindo scores at the time of IPAA creation did not differ between modified two- and three-stage approaches (all P > .05).

Conclusion:

The modified two-stage approach to IPAA creation resulted in fewer hospital days compared to the three-stage approach. Considering the risks and patient burdens of diversion, further research is needed to assist in decisions regarding protective ileostomy after completion proctectomy with IPAA.

Introduction

Ileal pouch-anal anastomosis (IPAA) is the procedure of choice for medically refractory ulcerative colitis (UC) in children. Total proctocolectomy cures UC and this understanding has allowed for operative treatment of UC in elective and urgent/emergent settings. Operative approaches have trended toward minimally invasive techniques with excellent short- and long-term outcomes.^1–3 For patients who present with severe or fulminant disease, a three-stage approach is the traditional method—(1) total abdominal colectomy with end ileostomy, (2) completion proctectomy with IPAA and diversion, and (3) reversal of diverting ileostomy. A three-stage IPAA after subtotal colectomy has become increasingly more common due to the number of patients with severe disease, significant malnutrition, and concomitant use of biologic agents.^4,5 Three-stage procedures allow a patient's health status to recover and immunosuppression to cease before forming the IPAA, thus minimizing complications. Recently the need for three-stage procedures has been questioned in adults and children, noting limited data to support the use of three- over two-stage procedures and wide variation in diversion practices among surgeons.^6–8 In the pediatric literature, there is a paucity of data regarding the additional operative risk resulting from diversion. Because of the patient discomfort, cost, and additional operative risk at the time of ileostomy reversal, our group has increasingly performed modified two-stage (initial colectomy followed by IPAA without diverting ileostomy) procedures in children with severe medically refractory UC.

Modified two-stage IPAA takes advantage of the improvement normally seen following subtotal colectomy in patients with severe, medically refractory UC. Immunosuppression has often ceased and the approach shortens the number of operations, overall cost, and, potentially, the time spent diverted.^8,9 The decision to forgo diversion is weighed against the risk of leak, and there has been considerable controversy in adult studies and limited data in children regarding whether patients with proximal diversion have a lower leak rate.^10,11 A small number of modified two-stage IPAA procedures have been reported in the pediatric literature, demonstrating similar short-term outcomes, leak rates, and longer term function when compared with IPAA protected by diversion.^8,12–14 We reviewed our experience with this procedure and compared the outcomes of pediatric patients undergoing modified two-stage to three-stage IPAA after recovering from subtotal colectomy.

Methods

After approval from the Mayo Clinic Institutional Review Board, we reviewed pediatric patients (age <18 years at operation) with a diagnosis of ulcerative or indeterminate colitis who underwent subtotal colectomy at the Mayo Clinic Rochester campus between January 1, 2007 and December 31, 2017. Mayo Clinic is a large, academic tertiary referral hospital.

Patient cohort

Patients undergoing modified two- and three-stage procedures for UC were included. Patients who underwent a one-stage or traditional two-stage procedure (total proctocolectomy IPAA with diversion followed by ileostomy reversal) were excluded. Patients were also excluded if they did not undergo all stages at Mayo Clinic or within the study period. We excluded patients without a diagnosis of UC or indeterminate colitis. Follow-up was from the date of their first operation until the date of their last follow-up visit.

Variables

Patient demographics, including age, sex, and body mass index (BMI) at the time of the IPAA attempt, were obtained from the medical chart. Operative details, 30-day complications, and postoperative length of stay (LOS) were abstracted for each of the procedures following recovering from subtotal colectomy (IPAA creation and, in the case of three-stage procedures, ileostomy reversal). Total LOS for the three-stage group included both the IPAA and the ileostomy reversal operations.

The primary outcome of interest was pouch leak after IPAA creation. Pouch leaks were defined as either radiographic evidence of abscess or fluid collection within the pelvis in communication with the pouch or evidence of pouch dehiscence on examination under anesthesia.

Secondary outcomes included time spent diverted, postoperative LOS, readmission, and 30-day maximum Clavien–Dindo scores following the procedure. For all secondary outcomes, each outcome event was assumed to be independent, and patients undergoing three-stage operations were able to contribute up to 2 of each outcome (1 after each surgery).

Statistical analysis

Continuous variables with normal distribution were summarized with mean and standard deviation, while nonparametric continuous variables were summarized with medians and interquartile ranges (IQRs). Categorical variables were summarized with frequency counts and percentages. Patient characteristics and outcomes were compared between modified two- and three-stage operations at the time of IPAA creation using Fisher's exact test for categorical variables, Student's t-test for continuous normally distributed variables, and Wilcoxon rank sum for continuous nonparametric variables. Statistical significance was set at P < .05. All data analyses were performed using STATA 15.1 (College Station, Texas).

Results

During the study period, a total of 43 patients underwent IPAA after subtotal colectomy (and met inclusion criteria). Just over half of patients were female (n = 22, 51%) and median age at IPAA creation was 15 years (IQR: 14, 17). The majority of patients (n = 32, 74%) underwent a three-stage approach, and 11 (26%) underwent a modified two-stage approach. Most children (n = 38, 88%) had a confirmed diagnosis of UC with the remaining 5 (12%) patients classified as indeterminate colitis. Operative approach was laparoscopic in 33 (77%), planned open in 9 (21%), and converted to open in 1 (2%). Single-incision technique was used in 12 of 33 (36%) of laparoscopic cases.

At the time of IPAA creation, there were minimal differences in demographic characteristics, Table 1. No differences were seen in age (P = .17), sex (P < .99), BMI percentile for age and sex (P = .06), diagnosis (P = .09), primary sclerosing cholangitis (P < .99), or American Society of Anesthesiologists (ASA) class (P < .99). One (9%) patient in the modified two-stage group was on steroids at the time of IPAA creation, while no (0%) patients in the three-stage group were taking steroids (P = .26). No patients were taking 6-Mercaptopurine (6MP) or Azathioprine at the time of IPAA, and no patients in the modified two-stage group were exposed to biologic agents at the time of pouch creation. Two patients in the three-stage group (6%) had IPAA performed within 12 weeks of exposure to infliximab.

Table 1.

Preoperative Demographics and Medical Comorbidities at Time of Ileal Pouch-Anal Anastomosis Creation

Factor	Modified two-stage	Three-stage	P
n	11	32
Age in years, median (IQR)	15 (12, 16)	16 (14, 17)	.17
Sex
Female	6 (55%)	16 (50%)	.79
Male	5 (45%)	16 (50%)	.79
BMI percentile for age and sex, median, (IQR)	33.4 (28.9, 45.6)	67 (41.2, 82.0)	.06
Diagnosis
Ulcerative colitis	11 (100%)	27 (84%)	.10
Indeterminate colitis	0 (0%)	5 (16%)
Diabetes	0 (0%)	0 (0%)
PSC	0 (0%)	2 (6%)	<.99
ASA class
1	1 (9%)	4 (13%)	.76
2	10 (91%)	28 (88%)
Steroids at operation	1 (9%)	0 (0%)	.26
Immunomodulators at operation	0 (0%)	0 (0%)
Biologic exposure within 12 weeks	0 (0%)	2 (6%)	<.99
Time between subtotal colectomy and IPAA in days, median (IQR)	161 (91, 206)	109 (83.5, 221.5)	.44

ASA, American Society of Anesthesiologists; BMI, body mass index; IPAA, ileal pouch-anal anastomosis; IQR, interquartile range; PSC, primary sclerosing cholangitis.

Overall, there were 4 (9%) pouch leaks following IPAA creation. Two (18%) leaks occurred among patients undergoing modified two-stage procedures, while 2 (6%) leaks were in patients undergoing three-stage procedures (P = .27). Pouch leak in both modified two-stage patients required reoperation with peritoneal washout and diverting ileostomy. The two leaks in patients with three-stage procedures were managed with percutaneous drainage. Operative approaches, estimated blood loss, readmission, return to the operating room, and maximum 30-day Clavien–Dindo scores at the time of IPAA creation did not differ between modified two- and three-stage approaches, Table 2 (all P > .05).

Table 2.

Operative Details and Complications Following Ileal Pouch-Anal Anastomosis Creation

Factor	Modified two-stage	Three-stage	P
n	11	32
Operative approach
Laparoscopic	7 (64%)	26 (81%)	.41
Open	4 (36%)	5 (16%)
Converted to open	0 (0%)	1 (3%)
EBL in mL, median (IQR)	50 (25, 50)	100 (40, 112.5)	.05
Infectious complication
None	9 (82%)	28 (88%)	.44
Leak	2 (18%)	2 (6%)
Superficial soft tissue	0 (0%)	2 (6%)
Ileus	1 (9%)	2 (6%)	<.99
Urinary retention	1 (9%)	2 (6%)	<.99
OR return within 30 days of operation	2 (18%)	3 (10%)	.59
Readmission within 30 days of discharge	1 (11%)	5 (16%)	<.99
30-day maximum Clavien–Dindo score
No complications	5 (45%)	22 (69%)	.37
1	4 (36%)	6 (19%)
2	0 (0%)	1 (3%)
3b	2 (18%)	3 (9%)

EBL, estimated blood loss; IQR, interquartile range.

Three (9%) of 32 three-stage patients had complications within 30 days of ileostomy reversal. These included partial small bowel obstruction, renal colic/nephrolithiasis, and reoperation for small bowel obstruction with internal hernia. After accounting for complications and additional hospital days occurring at the time of ileostomy reversal for these patients, modified two-stage procedures had shorter total median LOS (7 days versus 9 days, P = .005). Time spent diverted, readmission, return to the operating room, and maximum 30-day Clavien–Dindo scores did not differ between modified two- and three-stage approaches after accounting for the ileostomy reversal outcomes in the three-stage group (all P > .05), Table 3.

Table 3.

Secondary Outcomes Accounting for Ileostomy Reversal

Factor	Modified two-stage	Three-stage	P
n	11	32
Time spent diverted in days, median (IQR)	161 (91, 206)	184 (149, 318)	.14
Total hospital length of stay in days, median (IQR)	7 (4, 8)	9 (7, 11)	.005
OR return within 30 days of operation	2 (18%)	4 (13%)	.64
Readmission within 30 days of discharge	1 (9%)	5 (16%)	<.99
30-day maximum Clavien–Dindo score
No complications	5 (45%)	19 (59%)	.71
1	4 (36%)	8 (25%)
2	0 (0%)	1 (3%)
3b	2 (18%)	4 (13%)

IQR, interquartile range.

Discussion

The need for protective ileostomy following IPAA creation has been called into question in the adult and pediatric literature, but the number of children studied is small. Modified two-stage and three-stage IPAA have seldom been compared in the pediatric literature, despite the fact that many children are not candidates for a traditional two-stage procedure due to their disease severity. Reviewing our recent experience with forgoing diversion using a modified two-stage procedure in children with UC, patients had decreased hospital LOS without a significant increase in pouch leak or 30-day complications compared to the three-stage approach. Given the small number of modified two-stage procedures reported in children, more research is needed to determine whether foregoing protective ileostomy after completion proctectomy with IPAA has similar risk of leak compared to the three-stage approach.

Complication, reoperation, and readmission rates are high in children undergoing IPAA. The complications in both study groups were similar to previous reports in the pediatric literature.^1,2,8 Leak is the most feared complication since it is associated with the risk of long-term pouch failure and need for permanent ileostomy.¹⁵ Dolgin et al. initially reported 12 modified two-stage procedures in children. Two of 12 (17%) patients had ileoanal separation requiring diversion.¹² Ryan and Doody reported an additional 12 patients with modified two-stage IPAA. While their overall analysis found no difference in morbidity between patients who did and did not have diversion, their study did not directly compare modified two-stage approach with patients who had an ileostomy.⁸ In the analysis by Bismar et al., there were no reported leaks in either the 14 modified two-stage patients or 27 traditional two-stage patients.¹³ Most adult studies have compared the modified two-stage patients against either the traditional two-stage or restorative proctocolectomy with diversion. Recently, 17 children undergoing modified two-stage procedures were found to have more complications, longer hospital LOS, and a 30% leak rate compared to a 5% leak rate in 20 children undergoing three-stage procedures, prompting caution in the use of such an approach.¹⁴ The results of these comparisons are discordant on whether modified two-stage procedures have an increased risk of leak.^10,11,16 More comparisons of the modified two- and three-stage approaches are necessary since the approaches are the only IPAA options after urgent colectomy in both adult and pediatric UC patients.

Diversion has been regarded as the safer approach in IPAA formation; however, it brings with it additional operative risk, stoma complications, and body image issues in children. Swenson et al. found that forgoing diversion in adult patients resulted in lower resource utilization without increased complications.⁹ There was a trend toward less time spent diverted in the modified two-stage group, but this result did not reach statistical significance. Children undergoing this approach in our study also had slightly more time (although not statistically significant) between subtotal colectomy and formation of IPAA compared to three-stage patients. As our group has moved toward performing more modified two-stage IPAA, the time between urgent colectomy to second stage has increased similar to other reports in the literature.¹⁴ The increased time has allowed the patient to improve nutritionally, wean from immunosuppressant or biologic medications, and recover health. The ideal time to wait for reconstruction following a colectomy with end ileostomy is not yet known.

Diversion is thought to decrease the contamination associated with a pouch leak; however, there is limited evidence that ileostomy prevents a leak.¹⁷ Appropriate management of anastomotic leaks appears to result in similar functional outcomes in patients treated with and without diversion.^11,14 Only two pediatric studies have looked at long-term outcomes of the modified two-stage approach, one comparing it to the traditional two-stage procedure and the other comparing it to three-stage approach. Each reported no differences in pouchitis or long-term pouch outcomes between the modified two- and traditional two-stage patients or three-stage patients.^13,14 Further research is necessary to determine long-term functional outcomes with the modified two-stage approach in children.

The results of this study should be interpreted with its limitations in mind. First, the small number of patients undergoing modified two-stage procedures leaves open the possibility of type II statistical error. Therefore, while we did not see a difference in leak rates or morbidity in general between the procedures, readers should interpret this result with caution. Second, authors collected the data retrospectively and the modified two-stage approach is relatively new for our group. Because of these limitations, we were unable to compare long-term morbidity, such as pouchitis and pouch failure, between approaches. Finally, there is a selection bias at play in those patients who are chosen for the modified two-stage approach in that there is a longer waiting time between subtotal colectomy and IPAA creation. This fact no doubt influences a surgeon's decision to forgo diversion.

Conclusion

The present study compared the short-term outcomes in modified two- and three-stage IPAA in children and found no difference in leak rate or overall morbidity between approaches. Modified two-stage procedures resulted in shorter hospital LOS. Future study of modified two-stage procedures should compare outcomes against both traditional two-stage and three-stage approaches and examine long-term risk of bowel obstruction and pouch failure.

Footnotes

Acknowledgment

Dr. M.D.T. salary is funded by the Mayo School of Graduate Medical Education Clinician Investigator program. No specific grant number is associated with the work.

Disclosure Statement

No competing financial interests exist.

References

Polites

, Zarroug

, Moir

, Potter

. Single-incision laparoscopic ileal pouch-anal anastomosis in children—How does it compare to a laparoscopic-assisted approach?. J Laparoendosc Adv Surg Tech A, 2015; 25:167–171.

McKenna

, Potter

, Bews

, Glasgow

, Mathis

, Habermann

. Ileal-pouch anal anastomosis in pediatric NSQIP: Does a laparoscopic approach reduce complications and length of stay?. J Pediatr Surg, 2019; 54:112–117.

Polites

, Potter

, Moir

, et al. Long-term outcomes of ileal pouch-anal anastomosis for pediatric chronic ulcerative colitis. J Pediatr Surg, 2015; 50:1625–1629.

Geltzeiler

, Lu

, Diggs

, et al. Initial surgical management of ulcerative colitis in the biologic era. Dis Colon Rectum, 2014; 57:1358–1363.

Kennedy

, Potter

, Moir

, Zarroug

, Faubion

, Tung

. Pediatric chronic ulcerative colitis: Does infliximab increase post-ileal pouch anal anastomosis complications?. J Pediatr Surg, 2012; 47:199–203.

Hicks

, Hodin

, Bordeianou

. Possible overuse of 3-stage procedures for active ulcerative colitis. JAMA Surg, 2013; 148:658–664.

Richardson

, deMontbrun

, Johnson

. Surgical management of ulcerative colitis: A comparison of Canadian and American colorectal surgeons. Can J Surg, 2011; 54:257–262.

Ryan

, Doody

. Restorative proctocolectomy with and without protective ileostomy in a pediatric population. J Pediatr Surg, 2011; 46:200–203.

Swenson

, Hollenbeak

, Poritz

, Koltun

. Modified two-stage ileal pouch-anal anastomosis: Equivalent outcomes with less resource utilization. Dis Colon Rectum, 2005; 48:256–261.

10.

Samples

, Evans

, Chaumont

, Strassle

, Sadiq

, Koruda

. Variant two-stage ileal pouch-anal anastomosis: An innovative and effective alternative to standard resection in ulcerative colitis. J Am Coll Surg, 2017; 224:557–563.

11.

Weston-Petrides

, Lovegrove

, Tilney

, et al. Comparison of outcomes after restorative proctocolectomy with or without defunctioning ileostomy. Arch Surg, 2008; 143:406–412.

12.

Dolgin

, Shlasko

, Gorfine

, Benkov

, Leleiko

. Restorative proctocolectomy in children with ulcerative colitis utilizing rectal mucosectomy with or without diverting ileostomy. J Pediatr Surg, 1999; 34:837–839; discussion 839–840.

13.

Bismar

, Knod

, Patel

, Schindel

. Outcomes following two-stage surgical approaches in the treatment of pediatric ulcerative colitis. J Pediatr Surg October 2018 [Epub ahead of print]; DOI: 10.1016/j.jpedsurg.2018.09.021.

14.

Chen

, Grant

, Lindholm

, et al. Is fecal diversion necessary during ileal pouch creation after initial subtotal colectomy in pediatric ulcerative colitis?. Pediatr Surg Int, 2019; 35:443–448.

15.

Galandiuk

, Scott

, Dozois

, et al. Ileal pouch-anal anastomosis. Reoperation for pouch-related complications. Ann Surg, 1990; 212:446–454.

16.

Zittan

, Wong-Chong

, Ma

, McLeod

, Silverberg

, Cohen

. Modified two-stage ileal pouch-anal anastomosis results in lower rate of anastomotic leak compared with traditional two-stage surgery for ulcerative colitis. J Crohns Colitis, 2016; 10:766–772.

17.

Garbus

, Potenti

, Wexner

. Current controversies in pouch surgery. South Med J, 2003; 96:32–36.