Hirschsprung disease (HD) is a congenital anomaly of the enteric nervous system causing intractable constipation which requires surgical correction. Following the first successful operation for HD [1], several surgical techniques have been developed. Most patients with HD have favorable outcomes, although some experience long-term postoperative complications, including constipation, fecal incontinence, recurrent Hirschsprung-associated enterocolitis (HAEC), and failure to thrive (FTT) [234]. Obstructive symptoms after a pull-through (PT) occur around 8%–30%, and the causes can be classified as anatomical, pathologic, and functional [4]. Although conservative medical treatment or minor surgical intervention can improve these symptoms, some patients may require redo PT [5]. Indications for redo PT surgery can include pathologic and/or anatomic factors. Anatomical problems can include anastomotic stricture, stenosis, obstructing Duhamel pouch, remnant septum, and obstructing Soave cuff, whereas pathologic problems include retained aganglionosis and transition zone pathology [25]. Most of the anatomical problems are related to the initial PT procedure, which can be confirmed by rectal examination or contrast enema. In the absence of mechanical obstruction on rectal examination and contrast enema, a rectal biopsy should be performed to determine pathologic reasons [4].

Several studies on redo PT surgery in HD patients have been previously published. However, the outcomes vary by report and long-term outcomes were scarce. Therefore, we aimed to evaluate the clinical outcomes of redo PT surgery in HD patients at a single center.

The medical records of patients who underwent redo PT for HD at a single tertiary center, between June 2003 and October 2019 were retrospectively reviewed. Data regarding clinical characteristics, pathologic results, initial surgical techniques, redo PT techniques, symptoms and indications for redo surgery, complications after redo PT, and follow-up clinical outcomes were collected. The study was approved by the Institutional Review Board of Asan Medical Center (No. 2020-0188). It was performed in accordance with the Declaration of Helsinki and written informed consent was waived due to its retrospective nature.

All redo PT procedures were performed by a single surgeon. Patients were included if they had postoperative obstructive symptoms after initial PT, including refractory constipation and episodes of enterocolitis or FTT. For work-up, contrast enema (Figs. 1, 2) was done in every patient, except for those confirmed to have pathologic problems by biopsy.

The surgical technique for redo PT was determined by the surgeon according to the patient’s state. Posterior sagittal approach was performed in 1 patient in 2003 and transanal endorectal PT (TERPT), which has a similar basic procedure to the Soave procedure, was performed thereafter [6]. TERPT underwent in lithotomy position, and laparoscopy was performed for the mobilization of the colon if necessary.

Although surgery is usually successful in patients with HD, some patients experience postoperative morbidities, including constipation, fecal incontinence, and enterocolitis [234]. Improvements are often provided by medical treatment or minor surgical interventions, but some patients may require redo PT [5]. The present study evaluated long-term clinical outcomes in patients who underwent redo PT for various indications, with all procedures performed by a single surgeon.

Of the 13 cases requiring redo PT during the study period, 8 (61.5%) did so for pathologic issues, as confirmed by examination of the removed tissue specimens. Pathologic problem is usually caused by a pathologist’s incorrect assessment of presence or absence of ganglion cells on frozen section of specimen, or by a surgeon’s technical failure not going far enough above the pathological transition zone [4]. Three patients underwent redo PT based on biopsy results without contrast enema. Case 4 in Table 3 had regular ganglion cells in the frozen biopsy specimen during initial PT. However, the patient had no stool passage after the surgery, and the final pathology report indicates that, with a few ganglion cells present, this patient underwent redo PT 18 days after the initial PT. Case 8 in Table 3, who suffered from refractory constipation, was found to have a decrease in ganglion cell density only upon reassessment of the pathology slides 26 months after initial PT; this patient underwent redo PT without re-biopsy during the work-up period. Two sequential rectal biopsies of case 2 in Table 3 with recurrent enterocolitis and FTT after initial PT confirmed that the bowel was normally ganglionated. Contrast enema of this patient showed functional obstruction, for which this patient underwent redo PT, and examination of the final specimen showed transition zone pathology. These cases indicate the importance of the evaluation of adequate and sufficient specimens by experienced and reliable pathologists [8].

Two patients (15.4%) experienced early complications within 30 days after redo PT. One patient experienced postoperative ileus, which improved after conservative management, whereas the other experienced a ureter injury during redo PT due to severe adhesion, with this patient later requiring an additional ureteroureterostomy. None of the patients experienced an anastomotic leak, abscess, or perforation. This incidence was lower than in previous studies [910], which regarded preoperative stoma formation as advantageous. All 7 patients who underwent diverting ostomy formation prior to redo PT were referred from other institutions after initial PT. Since obstructive symptoms such as refractory constipation or HAEC were severe in patients who were referred from other institutions, diverting ostomy was performed to prevent obstructive symptoms during preoperative evaluation period. Five of the 6 patients who did not undergo stoma formation prior to redo PT had undergone initial PT at our institution, with initial PT in the sixth performed by a pediatric surgeon trained in our center. The absence of detailed medical records made it difficult to perform redo PT shortly after initial PT. Diverting ostomy formation prior to redo PT may be a safe option, enabling preoperative work-up with sufficient time to prevent the recurrence of obstructive symptoms. In addition, early complications after redo PT can be reduced by delayed stoma closure [11]. However, 42.9% of the patients with diverting ostomy in this study and 28% of pediatric patients with colostomy in a previous study had stoma-related complications [12], suggesting the need for careful patient selection.

Three patients (23.1%) experienced incontinence after redo surgery, with 1 showing symptom improvement during follow-up, an incidence lower than in previous studies [210]. It is important to preserve the anal canal during PT procedures [413], as the low incidence of incontinence in this study appears to be due to careful preservation of the dentate line throughout the procedure.

Two patients (15.4%) diagnosed with TCA reported recurrent diarrhea after redo PT. One of these patients (case 10 in Table 3), who underwent redo PT for transition zone pathology, was diagnosed by a pediatric gastroenterologist with chronic active ileitis and possibility of Crohn disease (Fig. 3). This patient is currently on medication for Crohn disease and has been showing improvement in symptoms since. A meta-analysis reported that, of HD patients who developed inflammatory bowel disease (IBD) after PT operation, 41% had TCA and 45% showed aganglionic involvement of the long segment [14]. The possibility of IBD should be considered in patients with extensive colonic aganglionosis if uncontrolled diarrhea persists after successful PT operation. The other TCA patient (case 11 in Table 3) with suspected sphincter muscle impairment eventually required a permanent ileostomy due to repeated diarrhea after redo PT. This patient was also diagnosed with chronic ileitis by the same gastroenterologist and received medical treatment, but symptoms did not improve. Findings in this patient once again indicate the importance of preserving the anal canal during initial PT [413].

The strengths of this study were its relatively long-term follow-up period after redo PT and that all operations were performed by a single expert surgeon, which makes all surgeries comparable by similar surgical skill. However, the present study had several limitations. Firstly, its retrospective design can lead to selection bias and 5 patients referred from other centers might not include sufficient data. Secondly, it has wide variability and heterogeneity of the patient population including initial length of aganglionosis, type of primary PT, highly variable patient age, highly variable interval between initial and redo surgery, and the presenting symptoms leading to redo PT. Thirdly, included number of patients is too small to reach any conclusions, though we attempted to show long-term follow-up results of redo PT.

In conclusion, redo PT procedures could be effective approach for improving obstructive symptoms in HD patients with anatomic or pathologic reasons following primary PT. Careful selection of patients and discreet indications for redo PT are crucial.