Hirschsprung\'s disease: a 20-year experience

Hirschsprung’s By Raymond

Reding,

Disease: A 20-Year Experience

Jean de Ville de Goyet, Serge Gosseye, Philippe Clapuyt, Paul Gibbs, and Jean Bernard Otte Brussels, Belgium

l During the period from 1972 to 1992, 59 children received surgical treatment at the University of Louvain Medical School for biopsy-proven Hirschsprung’s disease (HD). The extent of aganglionosis was as follows: short segment restricted to the rectosigmoid or descending colon (n = 44, 75%); long segment In = 9,15%); ultra-short segment (n = 3, 5%); unknown length because of death without autopsy (n = 3,5%). The median age at operation was 7 months for short-segment disease compared with 14 months for those with long-segment disease. Surgical procedures used for short-segment disease were Swenson with colostomy (n = 161, Swenson-Pellerin without colostomy (n = 271, Duhamel (n = I), and for long-segment disease were Martin (n = 3), Swenson-Delayers (n = 2), Swenson-Boley (n = 2) and ileostomy only (n = 2). Lynn’s sphincteromyotomy was performed in the three ultra-short cases. There were six deaths (10%) at a median age of 86 days (range, 28 to 1545 days), three had long-segment disease, and the others were not classified because of death before curative surgery. Enterocolitis (EC) was the most common cause of death (five cases) and was also the major source of morbidity after curative surgery (12 of 44, 27%) in short-segment patients, three of seven (43%) in long-segment patients. The functional success of the procedure was evaluated in 70% of the surviving patients (37 of 53; mean follow-up, 8.7 years; range, 1.2 to 21.51, using a novel semiquantitative scoring system, specifically designed for children who have HD. This system assesses normal stool evacuation, abdominal distention, soiling, and severe incontinence. The results were compared with those from a population of 39 healthy children and adolescents and demonstrated progressive improvement in function during childhood and adolescence (P = .04) for patients treated for short-segment disease. However, function was found to be consistently poorer in all age groups when compared with healthy controls (5 to 10 years, P < .Ol; 10 to 15 years, P < .05; >I5 years, P < .Ol). Copyright o 1997 by W.B. Saunders Company

INDEX WORDS: son’s procedure,

Hirschsprung’s functional

disease, evaluation.

enterocolitis,

Swen-

H

IRSCHSPRUNG’S DISEASE (HD) is a malformation of the distal bowel, characterized by the absence of ganglion cells, beginning at the internal anal sphincter and extending proximally for varying lengths.‘m3 The cause of this aganglionosis remains uncertain but has been attributed to a failure of migration or differentiation of neural crest cells, a result of microenvironmental changes, or to the destruction of enteric neurons secondary to ischemic or immunologic injury.4-6 The absence of the myenteric neural plexuses results in an imbalance of bowel innervation, with impaired local production of nitric oxide and excessive extrinsic cholinergic stimulation leading to permanent spasm of the aganglionic segment.’ The diagnosis of HD is based on radiological Jotrmal

ofPediatric

Surgery,

Vol32,

No 8 (August),

1997: pp 1221.1225

Etienne

Sokal, Jean Paul Buts,

studies, anorectal manometry findings, and histological examination of rectal wall biopsies.* Curative surgery consists of total or subtotal resection of the aganglionic segment followed by a pull-through of the normally innervated intestine.1,9-11We report the results including long-term functional assessment of a series of patients who had HD managed over a period of 20 years. In a subgroup of 27 patients, the results of Pellerin’s procedure, avoiding the use of a colostomy recommended in Swenson’s technique, were evaluated. MATERIALS

AND

METHODS

Between 1972 and 1992, 59 children were treated for biopsy-proven HD at the University of Louvain Medical School Hospital, Brussels. The records of these patients were reviewed retrospectively to assess presenting symptoms, associated abnormalities, investigative and surgical procedures, complications, and functional results of treatment. The extent of aganglionosis was determined during the curative surgical procedure, based on intraoperative frozen section biopsy findings, and was characterized as follows: ultra-short (n = 3); short segment, involving the rectosigmoid (n = 42) or descending colon (n = 2): long segment, involving the transverse colon (n = 1) or the total colon with a variable extension of the aganglionosis to the small bowel (n = 8). The precise length of aganglionosis could not be determined in three cases because of patient death without subsequent autopsy before curative surgery could be performed. Enterocolitis (EC) was defined as the occurrence of severe diarrhea associated with at least two of the following findings: positive stool culture, fever, intestinal pneumatosis at radiology, or septicaemia. Long-term follow-up was carried out by means of a detailed mail questionnaire as well as by routine postoperative examinations. A scoring system, specifically adapted for patients who have HD, was derived from the Templeton and Ditesheim score proposed for patients designed for patients with imperforate an~s,‘~ and was specifically operated on for HD. This score was used in children older than 4 years (mean age + SD at evaluation, 11.8 ? 5.3 years; range, 5.0 to 25.8 years) and assessed by one investigator (R.R.), taking into account four items related to the quality of anal control (normal stool evacuation, abdominal distension, soiling, severe stool incontinence with major accidents; Table 1). A control population of 39 healthy children and adolescents of comparable age (mean +- SD, 11.3 2 4.5 years: range, 5.2 to 19.1 years) was evaluated using the same system. Numerical variables with or without a normal distribution were expressed as mean (tSD) or median (range), respectively; differences between groups were tested by the Student’s t test or the Wilcoxon rank sum test. Categorical variables were compared with x2 or Fisher’s Exact

From the Departments of Paediatric Surgery, Pathology, Radiology, and Gastroenterology, St-Luc Urukersity Clinics, University of Louvain Medical School, Brussels, Belgium. Address reprint requests to Raymond Reding, MD, PhD, Department of Surgery {1401), St-Luc University Clinics, 10, Hippocrate Ave, B-1200 Brussels, Belgium. Copyright o 1997 by WB. Saunders Company 0022-3468/97/3208-0018$03.00/O

1221

1222

REDING

Table

1. Quantitative Assessment of Long-Term Functional After Surgery for Hirschsprung’s Disease 1. Normal

stool

evacuation

(toilet

1.0 0.5 0

distension

(A) Never (B) Occasionally

1.0 0.5

(C) Daily or continuous 3. Soiling (A) None, or rare (B) Less than three

0

Diagnosis of HD 2.0 1.0

per week

(C) Three per week, or more 4. Severe stool incontinence (major

0 accidents)

(A) Never (B) Occasionally/daily Total score (range) NOTE. assessable

Good,

spectively, failure to pass meconium within 24 and 48 hours of life was reported to have occurred in 41 (70%) and 31 (53%) cases, respectively. This information was missing in 12 cases (20%). The overall incidence of preoperative EC was 13 of 59 (22%), and was significantly more common in long-segment disease (7 of 44 [16%] short segment against five of nine [56%] in long segment, P < .Ol).

training)

(A) Once a day or more (B) Three per week or more (C) Less than three per week 2. Abdominal

Results

4-5 points;

in children

1.0 0 o-5 fair, 2-3.5

less than

ET AL

points;

5 years

old,

poor,

O-l.5

or with

points.

Not

ileostomy

or

colostomy.

tests. Spearman rank correlation coefficient was used as a nonparametric method to quantitate the degree of linear association between age and functional score.

RESULTS

Demography and Presenting Symptoms There was a significant preponderance of boys in the series, the male to female (M/F) ratio being 2.5 to 1 (42 boys, 71%; 17 girls, 29%). A family history of HD was noted in four cases (7%), three of them involving siblings, although interestingly only one child of the single pair of monozygous twins observed in this series was affected. Both the sex ratio and incidence of positive family history of HD varied depending on the severity of the disease. In short-segment disease the M/F ratio was 4.5 with a family incidence of 2% compared with 0.8 and 33% in the long-segment subgroup. Mean (*SD) birth weight was 3,380 (2486) g, with a gestational age of 39.6 (+ 1.3) weeks. Seventeen children (29%) had one or more associated abnormalities: seven children had Down’s syndrome, four had genitourinary abnormalities, three had cardiac abnormalities, one had Waardenburg’s syndrome, and the remaining two had other minor abnormalities. Clinical manifestation of HD is detailed in Table 2. For neonates (within 1 week of birth, n = 30), initial symptoms included acute intestinal obstruction in 14 cases, EC in seven cases (including bowel perforation in two), failure to pass meconium in five cases, and constipation in four. Constipation was more common in children older than 1 month, and there was a significant delay in correctly diagnosing the disease when constipation was the main symptom (median time, 244 days; range, 28 to 3,942 compared with 7 days; range, 0 to 1,400 for patients with other symptomatology; P < .OOOl). Retro-

Contrast enemas and anorectal manometry were performed to confirm the diagnosis in 50 and 34 cases, respectively. Preoperative surgical or suction biopsy of the rectum was available in all cases, the latter being performed in most cases after 1976. The sensitivity of contrast studies was 40 of 50 (80%), the median age of children with positive radiological diagnosis was 63 days (range, 2 to 3,832) versus 22 days (range, 2 to 545) for false negatives (NS). The sensitivity of anorectal manometry was 27 of 34 (79%), the median age of those who had positive diagnosis was 225 days (range, 7 to 3942) versus 76 days (range, 15 to 545) for false negatives (NS). Operative Procedures All 44 patients who had short-segment disease underwent a curative surgical procedure at a median age of 7 months (body weight, 7.5 kg). Swenson’s pull-through with protective transverse colostomy was carried out in 16 patients (Swenson’s three-stage procedure). Swenson’s operation was combined with the Pellerin’s technique in 27 cases without protective colostomy (Swenson’s two-stage procedure); in the Pellerin’s technique, the pulled-through ganglionic colon remains exteriorized for 10 to 15 days before being removed when a secondary cola-anal anastomosis is performed.r3 Duhamel’s pullthrough14 was performed in the remaining case. Among the nine patients who had long-segment disease, seven Table 2. Presenting Symptoms of Hirschsprung’s Age at Presentation and the lime Interval Symptoms

Symptoms

n (%I

Failure to pass meconium

5 (9)

2 (2-3)

Bowel NOTE. patients.

25 (42) 16 (27)

perforation Exact

age at onset

9 (15) 4 (7)

3 38 4 10

of symptoms

*Median (range). tsignificantly longer diagnostic cases with another symptomatology

to

and Diagnosis Age at Onset tdj*

Acute intestinal obstruction Constipation Enterocolitis

Disease, Related Between Onset of

(I-120) (2.1,461) (2-90) (2.2,461) was

Interval Between Symptom and Diagnosis (dj*

2 (O-13) 7 (O-178) 244 (28.3,942)t 13 (O-1,400) 8 (O-44) not available

delay when compared (P < .OOOl).

in three

with

the

HIRSCHSPRUNG’S

DISEASE

1223

underwent a curative surgery at a median age of 14 months (body weight, 5.6 kg). Surgery consisted of Martin’s operation l5 in three cases, Swenson’s operation combined with either the Boley’s technique16 or with the Deloyers’ procedure in two cases each; in the Deloyers’ technique, the right colon is mobilized from behind and turned upside down to reach the pelvis for the cola-anal anastomosis. l7 Histological examination of the operative specimens confirmed the diagnosis of HD without residual disease in the proximal colon or ileum. The three patients who had an ultra-short segment had Lynn’s sphincteromyotomy performed.18 In five patients, including two who had a long-segment disease, only an ileostomy or colostomy could be carried out because they died before further surgical treatment (Table 3). Perioperative Mortality and Morbidity

Assessment of Long-Term Function

Six children (10%) died at a median age of 86 days, three had long-segment disease and the other three were those who didn’t have an autopsy. Five of the six deaths (83%) occurred in the first half of the series between 1972 and 1982. There were no deaths in the short-segment patients, 0 of 44 against three of nine in long-segment cases (P < .Ol). Enterocolitis was the cause of death in five, pneumonia in the final case (Table 3). Death was not related to a prolonged delay between clinical presentation and bowel decompression (median delay, 1 day; range, 0 to 21). However, in two cases (Table 3, children AA and JB), death was caused by inappropriate surgical management because the ileostomy or colostomy was opened in an aganglionic segment in the absence of intraoperativi frozen section histology tests. In one child who had Down’s syndrome (DE), death was secondary to withdrawal of active therapy. The overall incidence of postoperative EC was 17 of 59 (29%) including two cases treated with a temporary ileostomy or colostomy, the mortality rate of EC being 5 of 17 (29%). There was a higher incidence in patients who had long-segment disease (5 of 9, 56%) than in thqse who had shortsegment disease (12 of 44, 27%), but this did not reach statistical significance. Enterocolitis recurred postoperatively in one of seven short-segment cases and three of five long-segment cases. Leakage of the ileo-anal anastomosis occurred in two children who had long-segment Table 3. Surgical Patient (Year of Death)

DR (1976) AA (1979) RG (1979) JB (1980) DE (1980) AS (1988)

disease and resulted in a pelvic abcess and a cola-vaginal fistula. No postoperative leakage occurred in the series of short-segment aganglionosis. Although postoperative anal dilatation was not included in routine management, anal dilatations under general anaesthetia were necessary both postoperatively or long term in 15 of 44 (34%) shortsegment cases. Eight children required more than one dilatation and Lynn’s operation had to be performed in six of them to improve bowel evacuation. Even after this extra procedure, one patient required further anal dilatation because of recurrent episodes of EC. Similarly, additional anal dilatations were performed in four of the seven long-segment cases (57%, NS when compared with the short-segment group). Three required Lynn’s procedure.

Associated Abnormalities

Functional evaluation was completed in 37 of 53 survivors (70%), after a mean follow-up period of 8.7 years (range, 1.2 to 21.5). Twenty-seven cases of operated short-segment disease in patients older than 4 years were compared with a healthy populatiqn of 39 children and adolescents (Fig 1). Median functional scores in children between 5 and 10 years old were 2.25 (range, 1 to 5) in the HD group (n = 12) versus 5 (range, 3.5 to 5) in the control group (n = 18; P < .Ol). The corresponding figures were 4 (range, 1.5 to 5; n = 7) versus 4.5 (range,4to5;n= 11;P5 years) operated on for short-segment Hirschsprung’s disease 1+1, according to age at evaluation, compared with scores obtained in 39 healthy subjects in the same age range (0). A significant positive linear correlation between the functional score and age at assessment was found in the HD group (P = .044), whereas no such correlation existed in the control group (NS).

recurrent postoperative EC after a Soave-Boley operation. DISCUSSION

This series confirms several previously reported demographic parameters of HD 9~19-21 : the male predominance (71%), particularly in short-segment disease (82%), the low rate of prematurity (mean gestational age, 39.6 weeks), the incidence of associated abnormalities (29%) especially Down’s syndrome (12%), and the possible occurrence of HD discordance in monozygotic twins. Analysis of the clinical presentations of HD confirms the heterogeneity of symptoms, with, however, a high diagnostic value of a history of failure to pass meconium within 24 hours of birth. Constipation was the most frequent misleading sign of HD, leading to a significantly longer diagnostic delay. The sensitivity of contrast enema and anorectal manometry was higher in older children, although this difference did not reach statistical significance. Surgery for HD must relieve the intestinal obstruction, provide early curative treatment with minimal morbidity, giving optimal long-term stool control and quality of life. In this series, most infants presenting as neonates or with acute complications (acute obstruction, EC, intestinal perforation) were initially treated with either an ileostomy or a colostomy. Subsequently, Swenson’s operation with or without the Pellerin’s modification was performed at 6 months or at a minimum weight of 5 to 6 kg. In patients experiencing chronic constipation, a two-stage strategy was used, combining Swenson’s operation with Pellerin’s procedure. In children treated for shortsegment disease, there was no postoperative mortality, and the surgical morbidity was very low, with an absence of anastomotic leaks or pelvic abcesses. These results compare favorably with other series in which cuff or perineal abscesses were encountered after endorectal

ET AL

pull-through operation or Swenson’s operation.22.23Potential developments include the performance of the operation in the neonatal period, as well as the use of the celioscopic technique. The surgical management of long-segment HD remains a matter of debate, with the main alternatives being Martin’s operation (three cases’in our series), which is a development of Duhamel’s technique,15 ileo-anal anastomosis leaving part of the aganglionic colon as a pedunculated onlay patch (Boley’s technique, two cases in our series),r6 and total colectomy and ileorectal anastomosis (modified Rehbein’s technique).” No procedure clearly stands out as being the most effective from either our experience or the literature, and morbidity remains high in most reported series of long-segment HD.24,25 However, in our most recent experience, the three-stage Martin’s technique has given very satisfactory short- and middle-term results and constitutes the procedure of choice in our institution. In most centers, EC still represents the most serious complication of HD both preoperatively and postoperatively resulting in high morbidity and mortality rates. The reported incidence of EC in the literature ranges from 20% to 58%.2 In our series, the overall incidence of preoperative and postoperative EC was 22% and 29%, with a higher but not statistically significant postoperative incidence in long-segment HD. Postoperative relapses of preoperative EC only occurred in four cases, three with long-segment disease. The overall mortality rate of EC in this series reached 29%. Therefore, our current policy in all children at risk is to perform prophylactic postoperative rectal irrigation, as well as to initiate prompt courses of vancomycin given by rectal enema or nasogastric tube in patients with symptoms of EC.26,27

Long-term results after surgery for HD usually report persisting constipation or anal incontinence, the latter complication ranging from 2.5% to 13.6% according to a recent review.28.2gHowever, the quality of bowel control cannot be assessed only by the presence or absence of incontinence, because gaseous distension of the abdomen represents another major indicator of poor bowel emptying. Few scoring systems exist to evaluate bowel function, but one of the more accurate systems, which allows semiquantitative assessment was specifically designed for patients with high imperforate anus.” To properly compare the components of bowel function after surgery for HD, we have adapted this score originally described by Templeton and Ditesheim (Table 1). Ideally, such a score should allow the evaluation of overall results in a single series, the follow-up of one particular patient, and the comparison of results between centers. However, several aspects of the scoring system used in this work

HIRSCHSPRUNG’S

1225

DISEASE

constitute a matter of debate. First, as confirmed in the control group, both A and B in parameter 1 (Table 1) represent normal stool frequency in healthy subjects, with 11 of 39 of the controls falling into category B. However, in our experience, it is essential to obtain at least one stool emission per day in HD patients. Second, the grading of abdominal distension can be difficult and highly subjective. However, it constitutes a relevant clinical finding of fecal impaction or bacterial overgrowth. Third, the long-term analysis results do not provide a longitudinal follow-up of the patients in this report; in addition, the assessment is dependent on the child’s or parents’ perception of bowel control. Despite these limits and drawbacks, the results of the long-term functional evaluation after surgery for short-segment disease suggest a significant improvement of bowel control during childhood and adolescence, which is in

accord with the conclusion of Swenson’s multinational retrospective study.29 This improvement may be related to the better motivation of adolescents as well as to a possible benefit from muscular development at puberty. Our long-term functional results also validate the use of the combined Swenson-Pellerin’s strategy, which avoids the need for a protective colostomy. Finally, it should be mentioned that the validity of the scoring system in patients operated on by alternative techniques (Soave, Duhamel) still needs to be evaluated. This study shows that Swenson’s pull-through operation performed in either two or three stages allows safe and efficient correction of short-segment HD. Postoperative EC constitutes the main complication in HD (29%), with a high mortality rate (29%). Long-term functional evaluation confirms a continuous improvement in bowel control during childhood and adolescence.

REFERENCES 1. Swenson 0, Bill AH: Resection of the rectum and rectosigmoid with preservation of the sphincter for benign spastic lesions producing megacolon. Surgery 24212-220, 1948 2. Puri P: Hirschspnmg’s disease: Clinical and experimental observations. World J Surg 17:374-384, 1993 3. Reding R: La maladie de Hirschsprung en 1996: Physiopathologie, diagnostic et thkrapeutique. Louvain Med 115:20-30, 1996 4. Edery P, Lyonnet S, Mulligan LM, et al: Mutations of the RET proto-oncogene in Hirschsprung’s disease. Nature 397:378-380, 1994 5. Parikh DH, Tam PKH, Van Velzen D, et al: The extracellular matrix components, tenascin and libronectin, in Hirschsprung’s disease: An immunohistochemical study. J Pediatr Surg 29:1302-1306, 1994 6. Hirobe S, Doody DP, Ryan DP, et al: Ectopic class II major histocompatibility antigens in Hirschsprung’s disease and neuronal intestinal dysplasia. J Pediatr Surg 27:357-363, 1992 7. Vanderwinden JM, De Laet MH, Schiffman SN, et al: Nitric oxide synthase distribution in the enteric nervous system of Hirschsprung’s disease. Gastroenterology 105:969-973, 1993 8. Bonham JR, Dale G, Scott DJ, et al: A 7-year study of the diagnostic value of rectal mucosal acetylcholinesterase measurement in Hirshsprtmg’s disease. J Pediatr Surg 22:150-152, 1987 9. Swenson 0, Sherman JO, Fisher JH, et al: The treatment and postoperative complications of congenital megacolon. Ann Surg 182: 266-213, 1975 10. Holschneider AM, Soylet Y: Rehbein anterior resection in the treatment of Hirschsprung’s congenital megacolon: Manual or stapler anastomosis-a comparative study. Z-Kinderchir 44:216-220, 1989 11. Bergmeijer JH, Tibboel D, Molenaar JC: Total colectomy and ileorectal anastomosis in the treatment of total colonic aganglionosis: A long-term follow-up study of six patients. .I Pediatr Surg 24:282-285, 1989 12. Templeton JM, Ditesheim JA: High imperforate anus: Quantitative results of long-term fecal continence. .I Pediatr Surg 20:645-652, 1985 13. Pellerin D: Le traitement chirurgical de la maladie de Hirschsprung par r§ion anastomose ext&iorisCe saris suture. J Int Co11 Surg 37:591-593, 1962 14. Duhamel B: A new operation for the treatment of Hirschsprung’s disease. Arch Dis Child 35:38, 1960

15. Martin LW: Total colon aganglionosis preservation and utilization of entire colon. J Pediatr Surg 17:635-637, 1982 16. Boley SJ: A new operative approach to total aganglionosis of the colon. Surg Gynecol Obstet 159:481-484, 1984 17. Prevot J: Hirschsprung’s disease: Deloyers’ technique. Ann Chir Infant 11:81-84, 1970 18. Lynn HB, Van Heerden JA: Rectal myectomy in Hirschsprung’s disease. Arch Surg 110:991-994, 1975 19. Jung PM: Hirschsprung’s disease: One surgeon’s experience in one institution. J Pediatr Surg 30:646-65 1, 1995 20. Ryan ET, Ecker JL, Christakis NA, et al: Hirschsprung’s disease: Associated abnormalities and demography. J Pediatr Surg 27:76-81, 1992 21. Hannon RJ, Boston VE: Discordant Hirschsprung’s disease in monozygotic twins: A clue to pathogenesis? J Pediatr Surg 23:10341035,1988 22. Joseph VT, Sim C-K: Problems and pitfalls in the management of Hirschsprung’s disease. J Pediatr Surg 23:398-402, 1988 23. Carcassonne M, Guys JM, Morisson-Lacombe G, et al: Management of Hirschsprung’s disease: Curative surgery before 3 months of age. J Pediatr Surg 24:1032-1034, 1989 24. Levy M, Reynolds M: Morbidity associated with total colon Hirschsprung’s disease. J Pediatr Surg 27:364-367, 1992 25. Endo M, Watanabe K, Fuchimoto Y. et al: Long-term results of surgical treatment in infants with total colonic aganglionosis. J Pediatr Surg29:1310-1314,1994 26. Marty TL, Seo T, Sullivan JJ, et al: Rectal irrigations for the prevention of postoperative enterocolitis in Hirschsprung’s disease. J Pediatr Surg 30:652-654, 1995 27. Bagwell CE, Langham MR, Mahaffey SM, et al: Pseudomembranous colitis following resection for Hirschsprung’s disease. J Pediatr Surg 27:1261-1264, 1992 28. Marty TL, Seo T, Matlak ME, et al: Gastrointestinal function after surgical correction of Hirschsprung’s disease: Long-term follow-up in 135 patients. J Pediatr Surg 30:655-658, 1995 29. Sherman JO, Snyder ME, Weitzman JJ, et al: A 40.year multinational retrospective study of 880 Swenson procedures. J Pediatr Surg 24:833-838, 1989