Redo Pull-through in Hirschsprungs Disease - A Meta-Analysis of Outcomes Based on type of Pull-through

ABSTRACT

Purpose:

The meta-analysis investigates the effects of redo pull-through (PT) surgeries on bowel function outcomes in patients with rectosigmoid and long-segment Hirschsprung’s disease (HD) who have already undergone an initial PT. By comparing different redo PT techniques, this study explores whether specific approaches yield better postoperative bowel function.

Methodology:

A literature search was conducted for articles on redo PT for HD; this included PubMed and ScienceDirect databases. The search terms or MeSH words “Hirschsprung disease” OR “congenital megacolon” AND “re-do pull through” OR “re-do” OR “reoperation,” connected by two Boolean operators “OR” and “AND,” were used to search the databases for relevant articles.

Results:

Fifteen articles of interest comprising 374 patients receiving redo PT were selected for analysis. The results suggest that selecting a redo PT method tailored to the patient’s unique anatomical and surgical history may substantially improve functional outcomes, highlighting the potential for individualized strategies to enhance recovery and long-term quality of life for those requiring reoperation.

Conclusion:

The findings indicate that the Swenson redo PT approach yielded better bowel function outcomes compared to other techniques. While redo PT procedures were effective in reducing fecal incontinence, the incidence of enterocolitis remained consistent across the various approaches.

INTRODUCTION

The surgical management of Hirschsprung’s disease (HD) by a pull-through (PT) procedure is well-established with a variety[1] of PT procedures including trans-anal endorectal,[2] Soave,[3] Swenson,[4] and Duhamel being used[5] to achieve continuity of gut. Despite the encouraging short-term outcomes such as low morbidity, short hospitalization, absence of anastomotic leaks, or wound infections,[3,4,6] about 30%–50% of patients experience long-term complications including constipation, recurrent obstructive symptoms, fecal incontinence, and enterocolitis.[1,7] Both anatomical and functional etiologies remain the cause for persistent obstructive symptoms following initial PT.[7] In most cases, postoperative disturbance in bowel function is managed by nonsurgical methods such as the use of laxatives or enemas.[8] However, in patients with postoperative stricture, retraction of bowel, transition zone PT, fecal fistula, or persistent aganglionosis, a redo PT is performed.[9,10] Different redo PT procedures such as endorectal,[11] Soave,[12] Swenson,[13] or Duhamel[14] procedures by open,[15] laparoscopic,[13] or robotic[12] approaches have been described. Redo PTs have been shown to be safe and effective in resolving the obstructive symptoms following primary PT.[13,15,16]

In recent decades, the debate on the preferable initial PT has been addressed by systematic reviews comparing patient outcomes after primary PT.[17,18] However, to date, analysis of outcomes following different redo PT procedures for rectosigmoid HD is limited,[19] and the relative efficacy of the different redo PT approaches for both general and specific outcomes remains unclear. With this background, the research question of this systematic review is “which is the best redo operation for HD following a primary PT, with a focus on functional outcomes?.” The aim of this systematic review and meta-analysis is to assess the outcomes and complications in patients undergoing redo PT surgery for HD. The review is followed by a meta-analysis to assess whether the selection of redo PT procedure affects patient outcomes, particularly in the context of bowel function.

METHODOLOGY

Literature search strategy

The search for publications from 2004 to 2024 on redo PT in HD was performed systematically based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) chart [Supplementary Figure 1 ^{(112.9KB, tif)} ]. PubMed and ScienceDirect were searched for relevant articles and case reports of patients with HD who underwent redo PT. The following search terms or MeSH words connected by two Boolean operators “OR” and “AND” were used: “Hirschsprung disease” OR “congenital megacolon” AND “re-do pull through” OR “re-do” OR “re-operation.” Further, the retrieved articles were manually searched for any more additional eligible articles.

Study selection

The selected articles were assessed for inclusion based on certain criteria. The inclusion criteria were as follows: (i) patients who underwent redo PT for various indications after initial PT for rectosigmoid HD including long segment; (ii) patients with a median age of <15 years; and (iii) case reports – patients aged below 18 years and (iv) articles published in English. The exclusion criteria were as follows: (i) adults with HD; (ii) patients presented with perforations, peritonitis, and diagnostic difficulties; (iii) immediate postoperative complications of initial pull-through or colostomy related; (iv) studies not related to redo PT and total colonic aganglionosis (TCA); (v) studies related to redo PT but without any data on outcome of redo PT; and (vi) review articles, comments, and letters to editors.

Data extraction

Data extracted from the selected studies were the following – author, year of publication, country, type of study, total number of patients, number of patients receiving redo PT, median age of patients at the time of redo surgery, type of initial PT, pathological or anatomical indications for redo PT, type of redo PT, surgical approach (posterior sagittal, open, laparoscopic, or robotic), follow-up period, and complications after redo PT. The outcomes included short-term outcomes and long-term outcomes [Table 1].

Table 1.

Characteristics of included studies

Author	Year	Country		Type of Study (compartive study. RCT, observational)			Total number of patients (n, M: F)	Requiring redo	No of patients receiving redo PT (M: F)			Reopeartion/abnormal HPE			Age of patients at the time of redo surgery (median)			biopsy at the time of re-do PT
Haikal	2022	Indonesia		Case report			1	1	1 (Female)			1//1			6 y			full-thickness rectal biopsy
Shankar	2021	India		Prospective			11	1	1 (male)			1//11			2 months			circumferential doughnut biopsy of proximal anastomosing margin
Jiang	2018	China		Retrospective			836	72	31 (22:9)			72//31			3 y (12 mo - 5 y)			full-thickness rectal biopsy
Xia	2016	China		Retrospective			18	18	LSR - 10 (9:1); TAR -8 (7:1); Total: 18 (16:2):			18//18			34.4±20.8 months for LSR; 43±22.3 months for TAR			suction biopsy of rectal mucosa
Dingemans	2016	The Netherlands		Retrospective			16	16	16 (10:6)			16//8			4.6 years (2 months-21 years)			Intra-operative biopsy
Dickie	2014	USA		NA			36	36	17 (10:7) (Soave cuff as the source of obstruction)			36//0			3.5 y (2 mo - 8.5 y)			rectal biopsy
Khope	2013	India		Case report			1	1	1
Sheng	2012	China		Retrospective			24	24	24			24//5						rectal biopsy
Chatoorgoon	2011	USA		Retrospective			17	17	17			17//0			7 y (2-16 y)
Lawal	2011	USA		NA			93	93	16 (14:2)			93//25			7.5 y (2-17 y)			rectal biopsy
Sowande	2008	Nigeria		Retrospective (case-reports)			4	4	4 (3:1)			4//0			10 mo - 15 y			rectal biopsy
Gobran	2007	Egypt		NA			7	7	7 (4:3)			7//6			4.2 y (2.5 to 6 y)			rectal biopsy
Pena	2007	USA		Retrospective			51	51	45			51//8						rectal biopsy
Schweizer	2007	Germany		NA			17	17	17			17//16			4.6 y (2-9 y)
Antao	2005	UK		Retrospective			53	17	17 (15:2)			17//0			42.8±10.8 mo
Author	Type of primary PT (no. of patients)				Clinical problems after primary PT or indications for Redo									Additional treatment		Type of redo PT
					Pathologic						Anatomical
Haikal	TEPT										Anastomotic dehiscence					Modified Swenson-like method using a temporary stump
Shankar	TERP				Enterocolitis (1); soiling (1)						X-ray examination - no spur or stricture			Laxatives/Bowel management program/Metrogyl and botox injection		Duhamel
Jiang	TERPT (18); Laprascopic Duhamel (5); laprascopic Soave (8)															Duhamel (14); Soave (17)
Xia	Transanal soave (18)				intractable constipation (18); soiling (6)									conservative treatment prior to reoperation: diet control, laxative, anal dilation, enema and colonic irrigation for 3–6 months		Laparoscopic (10): Soave (8); Swenson (2); transabdominal (10): Soave (7); Swenson (1)
Dingemans	TERPT (Soave-like) (8); TERPT (Swenson-like) (2); Rehbein (3); Duhamel (3)				obstructive symptoms (16), recurrent enterocolitis (1), fecal incontinence (4)											TERPT (Swenson-like) (15); TERPT (Soave-like) (1)
Dickie	Soave (36)				enterocolitis (10), constipation (6); failure to thrive (1)											transanal dissection (due to Soave cuff) (17)
Khope	Soave										Rectal stricture; recurrent stricture					Duhamel (combined with Malone procedure)
Sheng	Duhamel (9); Soave (12); Swenson (1); Rehbein (2)				Constipation (15); frequent fecal soiling (8); volutary bowel movement with ocassional soiling (1); enterocolitis (2)						anastomotic stricture (5); residual anganglionic segment (5); gate syndrome after Duhamel (5); fistula from rectum (6) and vagina (1); complex fistulae (2)					Posterios sagittal approach with laparotomy (7); Soave (7); Duhamel (1); Rehbein (3); re-using stapling device (5); rectovagnial fistual via laparotomy (1)
Chatoorgoon	Duhamel (17)				constipation/impaction (17); soiling (9); malnutrition and weight loss (1); enterocolitis (4)											posterior sagittal approach (8); Transanal Swenson-type resection (9)
Lawal	transabdominal Soave (7); Transanal Soave-like (7); transanal Swenson like (1); Duhamel (1)				enterocolitis (n=9), constipation (n=7), failure to thrive (n=5), impaction (n=5)											transanal pull through (15); posterior sagittal approach (1)
Sowande	Swensons (4)										rectovaginal fistula (1) rectourethral fistula (1), high trans-sphincteric fistula-in-ano (1) and complete anastomotic disruption (1)					posterior sagittal trans-sphincteric approach (4)
Gobran	TEPT (5); transabdominal Soave (2)				persistent obstructive symptoms (5); recurrent enterocolitis (1).						tight anastomotic stricture (1)					transanal endorectal pull-through (TEPT) (7)
Pena	Soave procedures (17) Duhamel (14), transanal (6), Swenson (5), unknown type (3), myotomy and/or myectomy (3), Soave with a right colonic patch (2); Swenson with a J pouch (1)										Stricture, acquired atresia, and dehiscence (21); Megarectal pouch (post-Duhamel) (12); Aganglionosis on rectal biopsy (8); Cutaneous fistula (8); Urethral fistula (2); Pouchitis (2); Vaginal fistula (1)					Posterior sagitta approach (40); Transanally (5)
Schweizer	Soave (3), Rehbein (13), Duhamel (1)										Incomplete resection of TZ (16), anastomotic stricture (9), fistulas (2)					Duhamel pull through (17)
Antao	Duhamel (53)				Constipation (17); Encropesis (14); enterocolitis (15); soiling (17)											Modified Duhamel (17)
Author	Kind of surgery									Follow-up in months (median; range)			Outcome after redo surgery
	Open/lapratomy (n)		Laprascopic (n)			Robot (n)/Other (n)							Short outcome				Long outcome
Haikal	1									NA							Normal digestive function; no incontinence, an excellent anal outcome
Shankar	NA		NA			NA				3 y and 8 mo			Symptom-free at 6 months				continent toilet trained at 3 yr and 8 months follow-up
Jiang			31							40 mo (10 mo - 8 y)			enterocolitis (8), death (3), thrived with weight gain (5)				<1 y to >5 y followup: Bowel function: excellent/good (77.4%); fair (16.1%); poor (6.5%)
Xia			10			transabdominal (10)				13 to 75 mo			(< 1 y follow-up) Perianal dermatitis (LSR-3, TAR-4); wound disruption (TAR-1); enterocolitis (LSR-4, TAR-4); soiling (LSR-5, TAR-5)				(< 1 y follow-up) adhesive bowel obstruction (LSR-0, TAR-2); enterocolitis (LSR-3; TAR-3); soiling (LSR-2, TAR-2)
Dingemans	7 (additional laparatomy)					8 (protective/temporary stoma)				3 y (9 mo-7 y).			one month followup (short-term): Anastomotic dehiscence (3), wound abscess (2), rectovaginal fistula (1), enterocutaneous fistula (1).				final followup (long term): enterocolitis (1); reversal stoma (8); additional surgery (9); soiling and fecal incontinence (6)
Dickie	5					12 (transanally)				7 mo (1-17 mo)			Enterocolitis resolved; other obstructive symptoms were also resolved
Khope													no soiling
Sheng	Laparotomy (8)									2.5 y (7 mo to 6 y)			21 patients (>3 years): normal or near normal bowel habits with a stool frequency of 1–5 times per day (19); voluntary bowel movements but occasional soiling (once or twice per week) and without significant incontinence (2); 2 patients (<3 years): 2-4 bowel movements per day; no symptoms of constipation, incontinence, soiling or enterocolitis.
Chatoorgoon	17 (8 -PSA; 9 -transanal)									2 y (2 mo to 9 y)			at 2 months follow-up: voluntary bowel movements (12); requirement of small dose of laxatives (8); patients diverted (2); enema requirement (1); enterocolitis (1) but managed by conservative treatment; gain in weight
Lawal	Yes (10)		Yes (10)			5 (transanally)				16 mo (3 mo-10 y)			Voluntary bowel movement (11); bowel anastomosed to the anocutaneous junction and presence of stoma (1); soiling (1); lost to follow-up (1); not tested for fecal incontinence because patient were below 3 y (2); no patient had recurrent enterocolitis; good weight gain
Sowande	2									6 mo - 8 y			6 mo follow-up (1): Bowel movement -8 times/day (1); 3 y follow-up (1): fully continent at day with a Kelly score of 4				6 y follow-up (1): fully continent of flatus and feces with a Kelly score of 6; 8 y follow-up (1): fully continent of feces and flatus with a Kelly score of 6
Gobran	NA		NA			NA				12 mo (8-16 mo)			Relieve from obstructive symptoms (7); no soiling (7); first month followup: 2-10 bowel movements per day (7);				3rd month followup: 1-3 motions pers day (7)
Pena	NA		NA			NA				6 mo -25 y			Voluntary bowel movements (14); Stomaa (11); On rectal irrigations (6); Voluntary bowel movements, occasional soiling (6); Fecal incontinence (on bowel management) (6); Lost to follow-up (2)
Schweizer										9 y (1-23 y)			normal stool pattern (15), constipation with ocassional use of laxatives (2); soiling (1)
Antao										73.9±31.2 mo			No constipation, encopresis and enterocolitis; normal sensation and urge to defecate (17); moderate soiling (2); stool frequency (1-5 per day)				quality of life improved - full freedom of movements and engaged in physical and social activities and better self-esteem (15)

Quality assessment

The quality of the included studies was assessed using the Newcastle–Ottawa Scale (NOS).[20]

Statistical analysis

Data gathered were analyzed using RevMan software (v5.4). The pooled odds ratios and their corresponding 95% confidence intervals (CIs) were calculated for individual and pooled statistics. Heterogeneity was quantified using I² statistic, and the I² value of 25%, 50%, and 75% represented low, moderate, and substantial heterogeneity, respectively.[21] Statistical significance was determined by a probability value of <0.05.

RESULTS

Literature screening

The systematic literature search identified 125 articles [Supplementary Figure 1 ^{(112.9KB, tif)} ]. After scanning, 7 duplicates were removed, and the remaining 118 articles were screened to assess the eligibility for inclusion. After reviewing the title and abstract, 85 articles were removed based on review – nonredo PT and articles without data on histopathology. Further, after reviewing the remaining 33 articles, 18 articles were excluded based on articles with patients with TCA, no outcome data on redo PT, lack of information of type of redo PT, and articles without full text. Finally, 15 studies were accepted for the qualitative analysis and 13 articles were considered for meta-analysis. PRISMA[14] flowchart outlining the article selection procedure is shown in Supplementary Figure 1 ^{(112.9KB, tif)} .

Study characteristics

The characteristics including the quality assessment of included studies are presented in Table 1. Among the selected 15 articles, 3 were case reports,[14,15,22] 7 were retrospective studies,[13,16,23,24,25,26,27] 1 was prospective,[28] and the remaining 4[29,30,31,32] had no information on the type of study. These studies were conducted at hospitals from different countries such as Indonesia,[15] India,[14,28] China,[13,25,27] The Netherlands,[16] Germany,[32] Egypt,[30] Nigeria,[22] the UK,[23] and the USA.[24,26,29,31] The selected studies were conducted between 2005 and 2022 and included a total of 1185 patients. Three hundred and seventy-four underwent redo PT. The type of primary PT was reported for 285 patients. The initial PT was transanal endorectal pull-through (TERPT)/Soave (145), Duhamel (109), Rehbein (18), Swenson (13), transanal (6), myotomy (3), and unknown in the remaining. The information on the patient’s gender was not reported in five articles.[14,24,26,27,32] Out of 374 patients who underwent redo PT, the gender was known for 128 patients (55.2%), with a male-to-female ratio of 2.8:1. Patients age at the time of redo PT, with a median age range from 2 months to 17 years. The main clinical symptoms necessitating a redo PT after the initial PT were as follows: constipation/impaction (85), enterocolitis (43), soiling (42), obstructive symptoms (21), encopresis (14), fecal incontinence (4), failure to thrive (2), and malnutrition/weight loss (1). Data were available for 120 patients regarding their abnormal histopathology (102), absent RAIR (9), and mega rectal pouch (9). Among the anatomical issues, patients reported to have anastomotic dehiscence, rectal stricture, recurrent stricture, fistula from rectum, residual aganglionic bowel, complex fistulae, megarectal pouch (post-Duhamel), and urethral fistula. Patients also received additional treatment such as use of laxative and diet control, anal dilation, enema, and colonic irrigation before reoperation[13,28] and myomectomy. The various details of redo surgeries and their approaches were available for 289 patients, which included Soave (67), Duhamel (51), Swenson (39), modified Duhamel (17), TERPT (43), posterior sagittal route with or without laparotomy (64), laparoscopy (51), reapplication of staples (5), Duhamel with Malone (1), Soave cuff excision (17), and modified Swenson (1). Ten patients responded to conservative treatment alone. Details of the remaining patients were not available.

Follow-up was reported in 13 studies with a follow-up time ranging from 1 month to 25 years [Table 2]. The majority of short-term outcomes of the redo PT suggested resolution of enterocolitis and obstructive symptoms, occasional or absence of soiling, and normal or near-normal stool frequency. A very few patients reported occasional use of laxative or the requirement for enema. The long-term outcomes included normal digestive function, no incontinence, improved quality of life, and self-esteem. However, some instances of bowel obstruction, soiling, enterocolitis, and additional surgery were also reported. Out of 374 patients who underwent redo PT procedure, follow-up data were available for 300 patients. One hundred and forty-nine patients in short-term follow-up (<1 year) and 107 in long-term follow-up (>1 year) were having normal bowel patterns. Enterocolitis was observed in 17 (<1 year) and 7 (>1 year) patients, supportive treatment in the form of enemas or laxatives was needed in 21 (<1 year) and 15 patients (>1 year), diversion was needed in 21 patients initially, and additional surgeries (stoma closure, fistula closure, and wound-related surgeries) in 28 patients in long term. In total, 287 of 374 redo patients achieved normal bowel function in long-term follow-up of more than 1 year. Three deaths were reported in immediate postoperative period.

Table 2.

Summary of analysis

Summary of study
Number of publications analyzed	13
Number of patients with initial PT studied	1185
Number of patients requiring Redo PT	374
Abnormal histopathology at redo surgery, absent RAIR, mega rectal pouch Remaining are stricture, obstruction, fistulas, enterocolitis, dehiscence, etc.	120
Duration of follow-up after redo	1 month–25 years
Follow-up	Short term <1 year	Long term >1 year
Normal bowel pattern	149	107
Enterocolitis	17	7
Supportive care ( enemas, laxatives)	21
Soiling	9	15
Death	3
Stoma	21
Stoma reversal, additional surgeries		28
Lost or unknown	3	No data
Primary/initial PT details available for		285/374
TEPT, Soave		145
Duhamel		109
Swenson		13
Rehbein		18
Myotomy, unknown		89
Transanal		6
Redo PT details available for		289/374
Soave		67
Duhamel		51
Swenson		39
Reapplication of staples		5
Posterior sagittal route±laparotomy		64
Laparoscopy		10
TERPT		43
Soave cuff excision		17
Conservative (enemas, dilatations, laxatives etc.)		10

Data not available or a combination of the above methods were done in the remaining. PT: Pull-through, TEPT/TERPT: Transanal endorectal PT

Meta-analysis

Comparison between different types of redo pull-through for complications

Duhamel versus Swenson

The data for the comparison of total complications of redo PT between Duhamel and Swenson procedures were retrieved from eight articles.[14,15,16,23,25,27,28,32] There was no significant difference in the total complications or redo PT between Duhamel and Swenson procedures (MD = 3.37, 95% CI: 0.35–31.95, P = 0.29); however, the data showed statistically moderate heterogeneity (I² = 65%, P = 0.09) [Figure 1].

Figure 1.

Forest plot comparing complications of redo Duhamel with Swenson. CI: Confidence interval

Duhamel versus Soave

The data for the comparison of total complications of redo PT between Duhamel and Soave procedures were retrieved from three articles.[16,25,27] There was a significant difference in the total complications in redo PT between Duhamel and Soave procedures (MD = 9.23, 95% CI: 1.16–73.66, P = 0.04) with no statistical heterogeneity (I² = 0%, P = 0.78). As shown, the total complications were higher in Duhamel than that in the Soave procedure [Figure 2].

Figure 2.

Forest plot comparing total complications of redo pull-through of Duhamel with Soave. CI: Confidence interval

Soave versus Swenson

The data for the comparison of total complications of redo PT between Soave and Swenson procedures were retrieved from six articles.[15,16,25,27,30,31] There was a significant difference in the total complications or redo PT between Soave and Swenson procedures (MD = 0.08, 95% CI: 0.01–0.55, P = 0.01) with statistical moderate heterogeneity (I² = 66%, P = 0.09). As shown, the total complications were higher in Soave than that in the Swenson procedure [Figure 3].

Figure 3.

Forest plot comparing total complications of redo pull-through of Swenson with Soave. CI: Confidence interval

Comparison of total complications between primary and redo pull-through

Overall, the comparison of total complications between primary and redo PT revealed no significant difference in the events (MD = 1.57, 95% CI: 0.87–2.85, P = 0.14) with statistically moderate heterogeneity (I² = 65%, P = 0.01) [Figure 4].

Figure 4.

Forest plot comparing total complications of overall primary with redo pull-through. CI: Confidence interval

Comparison of specific outcomes between different redo pull-through procedures

Continence or voluntary bowel movement

The data for the comparison of continence between Duhamel and Swenson were obtained from two studies.[16,27] The analysis revealed a significant difference in the continence between Duhamel and Swenson redo PT (MD = 10.82, 95% CI: 0.67–173.7, P = 0.09) with no statistical heterogeneity (I² = 11%, P = 0.29) [Figure 5]. Similarly, three studies[16,25,27] were included for the comparison of continence between Duhamel and Soave. The analysis revealed no significant difference in the continence between Duhamel and Soave redo PT (MD = 4.63, 95% CI: 0.71–30.31, P = 0.11) with no statistical heterogeneity (I² = 0%, P = 0.95) [Figure 6]. In addition, the comparison[16,27] showed no significant difference in the continence between Swenson and Soave redo PT (MD = 0.21, 95% CI: 0.02–2.24, P = 0.20) with no statistical heterogeneity (I² = 26%, P = 0.25) [Figure 7].

Figure 5.

Forest plot comparing continence between Duhamel and Swenson redo pull-through. CI: Confidence interval

Figure 6.

Forest plot comparing continence between Duhamel and Soave redo pull-through. CI: Confidence interval

Figure 7.

Forest plot comparing continence between Swenson and Soave redo pull-through. CI: Confidence interval

Enterocolitis

The data for the comparison of enterocolitis between Duhamel and Swenson were obtained from two studies.[16,27] The analysis revealed no significant difference in enterocolitis between Duhamel and Swenson redo PT (MD = 3.73, 95% CI: 0.22–63.40, P = 0.36) with no statistical heterogeneity (I² = 0%, P = 0.62) [Figure 8].

Figure 8.

Forest plot comparing enterocolitis between Duhamel and Swenson redo pull-through. CI: Confidence interval

Similarly, the comparison of enterocolitis between Duhamel and Soave[16,25,27] revealed no significant difference in enterocolitis between Duhamel and Soave redo PT (MD = 0.28, 95% CI: 0.04–1.91, P = 0.19) with no statistical heterogeneity (I² = 0%, P = 0.88) [Figure 9].

Figure 9.

Forest plot comparing enterocolitis between Duhamel and Soave redo pull-through. CI: Confidence interval

DISCUSSION

HD presents complex management challenges that extend beyond the initial surgical intervention. Although the primary surgery might seem straightforward in principle – resect or bypass the aganglionic segment – its success is not always guaranteed. When persistent symptoms or complications such as constipation, incontinence, or enterocolitis occur postoperatively, a redo PT procedure becomes necessary. Several techniques are employed for these secondary interventions, including Duhamel, Soave, Swenson, and TERPT each either by open method or laparoscopically with distinct benefits and limitations.

In our analysis, the most frequently chosen operation for initial PT is TERPT and Soave procedures, followed by the Duhamel approach. For redo PT, the sequence often involves Soave, Swenson, and Duhamel in that order. The need for redo PT can frequently stem from issues like an incomplete resection of the aganglionic bowel in Soave or Swenson procedures or an abnormally large pouch in Duhamel cases. To address these, the modified Duhamel technique has evolved, with a small-pouch low-rectal anastomosis approach now recommended for optimal outcomes.

Selecting the most suitable redo PT technique typically becomes evident only after surgery commences. The exact approach often becomes clear only once the redo surgery has begun, allowing the surgeon to assess the scarring, tissue planes, and overall tissue condition as in frozen pelvis. When the initial operation was a TERPT, Swenson, or Soave, it is sometimes possible to attempt another TERPT. However, this is contingent on the absence of factors such as severe scarring, poor surgical planes, infection, or fibrosis, which can complicate redo TERPT and make it less viable. If these challenges arise, the surgeon may resort to a Duhamel procedure, which can bypass the problematic anatomy.[33] In cases where the initial procedure was a Duhamel and revision is required, options such as Swenson or TERPT may be more effective. Redo Duhamel is challenging in these situations due to the presence of two thickened colonic walls, along with fibrosis and layers of interspersed muscle from previous surgeries. Be it a reapplication of staples or a complex redo operation in the form of Swenson or TERPT.

Meticulous inspection of the colon is essential before performing any type of PT. This inspection includes checking for kinks, torsion, vascularity, and the presence of ganglionated segments that could impair bowel function, ensuring adequate vascularity to support healing. The colon must be reevaluated at each stage–particularly before pulling it through the pelvic inlet–to ensure that no issues have developed that could compromise the success of the procedure.

In the present study, abnormal histopathological findings in patients highlight the necessity of full-thickness biopsies to confirm the normal zone and confirm complete resection. To ensure a comprehensive diagnosis, this study advocates for a laparoscopic or open full-thickness biopsy before the primary PT operation as advised by De la Torre et al.[34]

By prioritizing these two factors – technical precision during surgery and accurate tissue identification – surgeons can significantly decrease or avoid postoperative complications and the need for reoperations, leading to better patient outcomes and smoother recovery paths.

A primary goal in redo PT is to identify the technique that offers optimal outcomes with minimal complications. While few systematic reviews have examined the nuanced differences among redo PT methods, recent studies provide insight into the general effectiveness of these techniques. Meta-analysis data reveal that redo PT surgeries yield satisfactory functional outcomes, with complication rates comparable to those seen in primary PT surgeries. Swenson redo PT, which involves minimal residual rectal tissue, demonstrated a lower incidence of complications relative to Duhamel and Soave methods, attributed partly to its straightforward dissection and minimal residual rectal tissue. It has shown favorable results with a reduced risk of stool stasis and associated complications. The laparoscopic-modified Swenson technique further minimizes invasiveness and postoperative complications, making it a recommended choice.[35]

Enterocolitis remains one of the most serious complications following both primary and redo PT surgeries. The meta-analysis reveals similar incidence rates of enterocolitis across various redo PT techniques, including Duhamel, Soave, and Swenson. One notable limitation with the Duhamel method is the residual rectal pouch, composed partially of aganglionic bowel, which can cause stool stasis and recurrent enterocolitis. Studies, such as those by Chatoorgoon et al.,[24] recommend converting the Duhamel pouch to a Swenson-like TERPT to reduce constipation and associated enterocolitis. The data indicate that among other complications, enterocolitis remains a major post-operative complication after redo PT by any approach. The early postoperative episodes of enterocolitis may be controlled or prevented by regular colonic washes with normal saline starting 1^st week to 3^rd and 4^th weeks.

Fecal incontinence remains a critical outcome to monitor after redo PT. The current review suggests that the likelihood of achieving continence or voluntary bowel movements varies across techniques, with Swenson redo PT, which spares the sphincter complex, showing better continence rates than Duhamel. However, there was no statistically significant difference in continence outcomes between Duhamel and Soave or between Swenson and Soave PT.

In some cases, impaired fecal continence preexists after initial PT operation, raising questions about the redo PT’s influence on this outcome. Some studies suggest that postoperative medical management following primary PT can improve continence and minimize redo surgery complications. It is thus essential to initiate a trial of conservative treatment and perform histopathological assessment before considering redo surgery.[28] Despite some patients benefiting from improved continence and reduced obstructive symptoms, short-term outcomes of redo PT can still include soiling and fecal incontinence, warranting careful consideration before proceeding with surgery.[16] On the contrary, poor outcomes in terms of continence after redo PT when compared to the primary PT were reported in another study.[36]

The choice between open and laparoscopic redo PT also impacts patient outcomes and complication rates. Laparoscopy provides advantages in terms of visualization, allowing for adhesiolysis, segmental biopsies, and tension-free mobilization of the proximal colon. For cases involving prior laparoscopic or transanal surgery, the minimal abdominal cavity disruption results in fewer adhesions, facilitating redo procedures with reduced complications.[25]

The present study also highlights the posterior sagittal approach as particularly beneficial for managing low rectal strictures, complex rectovaginal or rectocutaneous fistulae, or retained aganglionic bowel. The fistulae pose significant challenges during redo. They can arise following Duhamel, Soave, or Swenson procedures, leading to severe, often unexpected complications. One-fourth of the present analysis of redo operations utilized the posterior sagittal route with or without laparotomy. This technique is advantageous as it has direct access, allowing for precise dissection and avoiding previously scarred pelvic regions, especially in patients who have undergone multiple pelvic surgeries or present with a frozen pelvis due to various causes. For additional support, a protective colostomy is often recommended to facilitate controlled healing and reduce patient discomfort.[26,37]

Finally, in a small subset of patients with HD, a permanent colostomy or a Malone antegrade continence enema procedure may be necessary to support bowel management and enhance quality of life (Khope et al.).[14] This is particularly relevant for patients with underlying neurological deficits or associated syndromes, where standard PT procedures may not fully restore bowel function. In these cases, a colostomy or Malone procedure provides a tailored approach to ongoing bowel care, promoting comfort and independence in daily life. While this study contributes valuable insights into redo PT techniques for HD, it has certain limitations. First, the variability in surgical outcomes can be attributed to the fact that all redo PT surgeries were performed by different surgeons across multiple institutions. This variation may introduce discrepancies in surgical expertise and the availability of instruments. Given the heterogeneity of surgical techniques used, more comprehensive studies are required to accurately analyze the possible associations between long-term outcomes and the different types of redo PT procedures. In the present study, it was not possible to pinpoint a specific redo operation as universally suitable for each type of failed initial procedure. Moreover, the retrospective design of the study, along with the broad age range of the patient cohort, could introduce bias into the meta-analysis. To mitigate this potential bias, future research should consider performing separate analyses in pediatric populations by making age- and sex-matched comparisons and larger, standardized datasets.

CONCLUSION

Redo PT for HD offers substantial benefits in cases where primary PT is inadequate and must be offered as soon as indicated. With Swenson and modified laparoscopic Swenson redo techniques showing favorable outcomes, this technique seems to be promising in providing the least postoperative side effects. Surgeons now have variable approaches (open, PS route, and laparoscopy) to reduce complications and enhance patient quality of life. Redo TERPT is safe and effective to resolve obstructive symptoms and relieve persistent symptoms of HD.

Resecting of Duhamel pouch or change of technique (to Swenson, Soave) is also found to work well in appropriate clinical settings.

Majority of redo patients do well in terms of stool control and general well-being, though, few with enterocolitis and fecal incontinence continue to pose challenges across all redo PT techniques, but many in the redo category seem to get better in the long term. A multidisciplinary approach, integrating histopathological, anatomical, and procedural insights, is essential for tailoring redo PT. All efforts must be made to perform the initial PT with utmost perfection.

Further research involving comprehensive datasets and broader inclusion criteria will be instrumental in refining redo PT criteria and improving long-term outcomes in HD patients. By continuing to investigate the complex interplay of technical and pathological factors in redo PT, clinicians can optimize interventions and reduce the burden of postsurgical complications in HD.

Summary

This meta-analysis reviewed 13 publications involving 1185 patients initially treated with pull-through (PT) surgery [Table 2]. It was found that 374 required redo PT and among them, 120 had abnormal histopathology or dysfunction, and the remaining had obstruction, fistulas, enterocolitis, etc., Follow-up duration ranged from 1 month to 25 years, with data available for 300 patients. Of these, details are available for 287 who achieved normal voluntary bowel movements in the long term [Table 2].

Conflicts of interest

There are no conflicts of interest.

Supplementary Figure 1

Preferred Reporting Items for Systematic Reviews and Meta-Analyses chart. PT: Pull-through, TCA: Total colonic aganglionosis

Funding Statement

Nil.