INTESTINAL NEURONAL DYSPLASIA: A PROSPECTIVE MULTI-CENTER STUDY ON INTEROBSERVER VARIATION AND CLINICAL OUTCOME

Abstract

Intestinal neuronal dysplasia (IND) of the colonic submucous plexus is considered to be a congenital disorder causing symptoms resembling those of Hirschsprung's disease (HD). The diagnostic criteria using enzyme-histochemistry on rectal biopsies were published by a group of German speaking pathologists (Pathologe 1991;12:171-4), but have been questioned by others due to absence of morphological data in healthy children. Patients and methods: Biopsies from 108 children (62 male, median age 2.1 yrs, range 4 days - 15.3 yrs.) with colonic motility problems were sent from 7 University hospitals (4 departments of paediatric surgery, 3 of paediatric gastroenterology) and examined by 3 pathologists for histology (acetylcholinesterase, LDH, SDH). In a total of 377 biopsies the 3 pathologists independently coded 21 features and the final diagnosis without knowledge of clinical symptoms. Interobserver variation was analysed by kappa statistics. Clinical data were recorded at the time of biopsy and 6 and 12 months later. Results:Observer-variability: There was full agreement between the 3 pathologists with respect to final diagnosis in all 20 children with HD (Kappa=1), but in only 14% of the remaining patients. In 15 children, at least one pathologist judged the case as,,normal”, while another diagnosed,,classic IND”. Kappa values were very low for more,,subjective” items (i.e. signs of immaturity, Kappa = 0.00 - 0.020), but unsatisfactory even for,,objective” items like,,presence of ganglion with > 8 ganglion cells” (kappa = 0.52 - 0.58). Clinical outcome: The 88 children without HD were clinically divided into 2 groups: children with constipation only (n=57) and additional different complications i.e. ileus, malformations etc. (n=31). In the 57 constipated patients, there were no significant differences between the 30 children looked after by pediatric surgeons compared to the 27 from pediatricians regarding items indicating severity of constipation (duration of symptoms, stool frequency, presence of encopresis, abdominal and rectal fecal mass). However, 13 sphinctermyectomies and 1 colonic resection were performed in those children that were looked after by surgeons. Cure rate (normal bowel pattern, off laxatives) after 1 year was higher in children classified as,,IND” or,,not normal” (40% & 60%) compared to histologically,,normal” or,,no IND” (23% & 27%). Operated children had a lower cure rate (17%) than conservatively treated patients (30%). Conclusions: In contrast to HD, there is an unacceptably high interobserver variation regarding the different morphological signs and final diagnosis of IND with the criteria and conditions used here. Abnormal findings have no predictive value for long-term prognosis in constipated children and seem to provoke a surgical pro-cedure. With the current knowledge, rectal biopsies for enzymehistochemistry should only be performed in constipated children to diagnose or exclude HD.