Does calretinin immunohistochemistry reduce inconclusive diagnosis in rectal biopsies for Hirschsprung disease?

The Diagnostic Value of Immunohistochemistry Markers in Hirschsprung Disease; A Systematic Review and Meta-analysis

BACKGROUND

Immunohistochemistry (IHC) markers are employed to improve the diagnostic yield when testing for Hirschsprung disease (HSCR). Yet, a superior test has not been identified.

OBJECTIVES

We aimed to determine the diagnostic test accuracy (DTA) of IHC markers.

METHODS

We conducted database search for studies reporting IHC staining on rectal biopsy investigating for HSCR. We constructed 2 × 2 contingency tables, and calculated DTA estimates in pooled and paired testing using random-effect model meta-analysis.

RESULTS

Twenty eight IHC markers from 107 studies were used to investigate for HSCR in 10891 children. In pooled analysis; calretinin sensitivity and specificity were superior to acetylcholinesterase, S100, and peripherin [98 % (CI; 0.95-0.99) & 99 % (CI; 0.97-0.99)], [94 % (CI; 0.86-0.97) & 99 % (CI; 0.96-0.99)], [92 % (CI; 0.85-0.96) & 97 % (CI; 0.89-0.99)] and [91.7 % (CI; 0.54-0.98) & 94.8 % (CI; 0.59-0.99)], respectively. In paired analysis calretinin diagnostic odds ratio was superior to hematoxylin and eosin (H&E), acetylcholinesterase and S100: [3349 (PI; 551.3-22667.2) vs 345.3 (PI; 54.9-2394.2)], [300.9 (PI; 13.3-4146.9) vs 34.6 (PI; 2.2-363.9)] and [696.9 (PI; 91.2-3401.7) vs 196.9 (PI; 29.8-890.5)], respectively. In biopsies labelled inadequate for H&E testing, calretinin specificity to rule out HSCR reached 92 % (CI; 0.288-0.998).

CONCLUSIONS

IHC provides additional diagnostic value over H&E. Calretinin appears to be, currently, a superior IHC marker. The available literature is of variable quality, cautious interpretation of the findings should be considered.

LEVEL OF EVIDENCE

III.

Full-Thickness Rectal Biopsy in Children Suspected of Having Hirschsprung's Disease: The Inconclusive Biopsy

The diagnosis of Hirschsprung's disease relies on histologically proven aganglionosis and nerve trunk hypertrophy in rectal biopsies. Although the frequency of inconclusive biopsies is relatively low, it is a relevant clinical problem. The aim of the present study was to investigate whether a re-evaluation of archived full-thickness biopsies (FTBs) stained with hematoxylin and eosin (HE), together with immune histochemical (IHC) staining, would be diagnostic in biopsies otherwise deemed inconclusive at initial examination with HE only. A total of 34 inconclusive biopsies in 31 patients were identified. From each tissue block, three slices were cut and stained with HE, S100 and calretinin. A blinded pathologist examined the tissue samples. At re-evaluation, one patient was found positive for HD and 11 negative for HD with both HE and IHC staining, respectively. In all 12 cases, the result was confirmed by the final diagnosis at a 5-year follow-up. The rest of the cases were deemed inconclusive. A re-evaluation of the remaining tissue from the biobank might have saved one third of the children from a re-biopsy. The value of adding IHC to conventional HE staining is dubious.

Diagnostic accuracy of calretinin and acetylcholinesterase staining of rectal suction biopsies in Hirschsprung disease examined by unexperienced pathologists

Rectal suction biopsy (RSB) is a gold standard for diagnosing Hirschsprung disease (HD). Calretinin staining of RSB is increasingly used by experienced pathologists due to non-complex examination and comparable diagnostic accuracy with acetylcholinesterase (AChE). However, the diagnostic accuracy of calretinin examined by unexperienced pathologists remains to be elucidated. Therefore, we aim to compare diagnostic accuracy of calretinin with AChE on RSB for diagnosing HD when examined by unexperienced pathologists. We prospectively analyzed sections from RSB stained with AChE + HE and calretinin. Blinded examination was done by five unexperienced pathologists (pathology residents) and three experienced pathologists (senior pediatric gastro-enterology pathologists) assessing for the presence of HD. Cases for the study included ones proven to be HD on resection specimens and cases without HD. Diagnostic accuracy was determined calculating area under the curve (AUC), sensitivity, specificity, likelihood ratio, and posttest probability. Fleiss’ kappa analysis was performed to assess interobserver agreement between reviewers. Eleven of 18 included patients (61%) were diagnosed with HD. Comparing the diagnostic accuracy of unexperienced pathologists, calretinin versus AChE + HE showed sensitivity of 80.0% versus 74.5% and specificity of 100% versus 65.4%, AUC of 0.87 (0.78–0.96) versus 0.59 (0.45–0.72). Unexperienced pathologists showed substantial agreement with calretinin (kappa 0.72 [0.61–0.84]) and fair agreement with AChE + HE (kappa 0.34 [0.23–0.44]). We found calretinin having higher diagnostic accuracy in diagnosing HD compared to AChE + HE when examined by unexperienced pathologists. Therefore, we recommend to use calretinin as the standard technique for staining RSB in diagnosing HD.

Rectal Biopsy Technique for the Diagnosis of Hirschsprung Disease in Children: A Systematic Review and Meta-Analysis

ABSTRACT

The diagnosis of Hirschsprung disease (HD) depends on the histopathological analysis of rectal biopsies. This review aims to define the best rectal biopsy technique. A systematic literature review and proportional meta-analysis of the available case series studies of rectal biopsies were performed in this study. All case series with more than five rectal biopsies in children younger than 18 years of age suspected of HD that described at least one type of rectal biopsy were included. The studies that did not specify the rate of conclusive results and the rate of complications of the biopsy procedures were excluded. According to the literature review, there were four different techniques of rectal biopsy: open, suction, punch, and endoscopic. In the title and abstract screening process, we assessed 496 articles, 159 fulfilled the eligibility criteria, and 71 studies reported our outcomes of interest and were included in the meta-analysis. The pooled proportion of conclusive results was 94% in open biopsies (95% CI 0.89-0.98), 95% in punch (95% CI 0.90-0.98), and 88% in suction group (95% CI 0.85-0.92). The pooled proportion of complication rates was 2% in open biopsies (95% CI 0.00031-0.04), 0.039% in suction (95% CI 0.00023-0.0006), and 2% in punch biopsies (95% CI 0.00075-0.04). Suction, punch, and open techniques presented comparable rates of conclusive results. In the suction group, the association between different methods of histopathological analysis increased conclusive results rates; however, the punch biopsy was associated with significantly higher complication rates than the suction technique.

Barium Enema Revisited in the Workup for the Diagnosis of Hirschsprung's Disease

OBJECTIVE

To analyse the diagnostic capacity of barium enema (BE) in the diagnostic investigation for Hirschsprung's disease (HD) was analyzed for transition zone (TZ) identification and rectosigmoid index (RSI) ≤1.0 determination.

PATIENTS AND METHODS

BE images were analyzed retrospectively by 2 examiners and the results were compared with the histopathology of rectal biopsies.

RESULTS

TZ identification and RSI ≤1.0 were assessed separately and combined in 43 patients. Twelve (27.9%) patients had the diagnosis of HD based on rectal biopsies. TZ identification presented better diagnostic capacity for the 2 examiners than RSI ≤1.0. However, interexaminer agreement was higher for RSI ≤1.0 than for TZ identification. The combination of TZ identification and RSI ≤1.0 increased the sensitivity (83.3%-92.3%) and the negative predictive value (90.4%-92.3%).

CONCLUSION

Therefore, the high diagnostic sensitivity of TZ identification combined to RSI ≤1.0 reinforces the usefulness of these BE parameters in the screening for Hirschsprung's disease.

Optimal time for single-stage pull-through colectomy in infants with short-segment Hirschsprung disease

OBJECTIVE

Short-segment Hirschsprung disease (HSCR) is the predominant type of HSCR that affects approximately 75% of patients. Whether single-stage endorectal pull-through (ERPT) surgery is appropriate for neonatal patients with HSCR has not been definitively determined. This retrospective cohort study concerning infants with short-segment HSCR investigated the optimal age for single-stage ERPT surgery, regardless of the operative approach.

METHODS

The 198 patients were stratified by operative age ≤ 3 or > 3 months (groups A or B, respectively, n = 62 and 136, respectively). Diagnoses of short-segment HSCR were conducted by preoperative contrast enema and rectal suction biopsy with acetylcholinesterase immunohistochemical staining. The perioperative clinical course for all patients was reviewed and the accuracy rate of the preoperative diagnoses and postoperative short- and midterm outcomes were assessed.

RESULTS

The rates of diagnostic accuracy, according to the results of the preoperative contrast enema or rectal suction biopsy, were lower in group A (67.2 and 93.5%, respectively) than in group B (81.4 and 94.9%, respectively). In groups A and B, 49 (79.1%) and 108 (79.4%) infants, respectively, completed follow-up examinations. The short-term outcomes were postoperative HSCR-associated enterocolitis, adhesive bowel obstruction, anastomosis leakage, and anal stenosis during the first 12 months after surgery. The midterm outcomes were incontinence and constipation at ~24 months after surgery. Compared with group B, group A experienced more incidences of anastomotic leakage in the short-term and more soiling in the midterm. In groups A and B, the rates of constipation recurrence were nil and 1.9%, respectively.

CONCLUSION

Infants with HSCR ≤3 months old at the time of single-stage ERPT surgery showed lower rates of accurate and conclusive diagnostic results and poorer postoperative outcomes. Waiting to perform this surgery until infants are older might be more beneficial.

Ganglion Cells Are Frequently Present in Pediatric Mucosal Colorectal Biopsies: H&E Sections and Calretinin Immunohistochemistry in the Evaluation of Children With Chronic Constipation

Hirschsprung disease (HD) rarely presents as chronic constipation after the newborn period. At our institution, calretinin immunohistochemistry (CAL) is frequently requested by clinicians on rectal mucosal biopsies (RMBs) taken during colonoscopy in older children in whom suspicion for HD is low. We hypothesized that review of these biopsies would frequently reveal ganglion cells (GCs). We reviewed features of mucosal biopsies (November 2013 to September 2015) from children ≥1 year of age on which clinicians had requested CAL on at least one specimen. A total of 93 biopsies with paired CAL from 83 patients were suitable for study (ages 1-18 years, M:F 1.2). Submitted clinical indication was constipation in 62 patients (75%). GCs were found within or subjacent to muscularis mucosa in 63 biopsies (68%), 12 (19%) of which were designated from a specific anatomic site, eg, 2 or 3 cm. In 25 of 63 (40%) cases, GCs were identified on one of the first 3 sections (median 5th, range 1st-54th). Forty-six cases (73%) contained no or <0.5 mm of submucosa (SM, range 0-2 mm). All but one case (62/63, 98%) with identified GCs showed positive CAL staining; a single case showed equivocal staining. Among the 30 biopsies with no observed GCs, none (0%) had >1 mm of SM, and 21 (70%) had no SM. CAL was positive in 28 (93%) and equivocal/weak in 2 (7%); no additional work-up for HD was pursued. The data suggest that H&E sections of RMBs can exclude HD at a specified site in many cases and provide the basis for a future study examining the utility of CAL in RMBs without SM as a means for excluding HD.

Rectal suction biopsy with calretinin immunohistochemistry in patients suspected with residual aganglionosis after operation for Hirschsprung disease

INTRODUCTION

This study investigates the use of rectal suction biopsy (RSB) with calretinin immunohistochemical staining (CIS) in patients suspected of having abnormally innervated bowel after pull through operation for Hirschsprung disease (HD).

METHOD

This study was conducted in Children's Hospital 2, Ho Chi Minh City from January 2015 through June 2016. Patients suspected with abnormally innervated bowel after pull through operation for HD were submitted for a RSB with CIS. Following histological results, the patients underwent an appropriate treatment (redo pull through operation or non-operative medical management) and followed up at least 6 months to evaluate their outcomes.

RESULTS

Ten patients were enrolled in our study. The median age was 5.6 (range 1.4-20) years. The diagnosis of normally innervated bowel was made in five patients by showing positive reactivity of CIS on RSB, whereas five others were diagnosed with abnormally innervated bowel. Final diagnosis of the five latter patients was confirmed by analyses of the resected bowel after a redo pull through operation (including 4 cases with residual aganglionosis and one case with transition zone bowel).

CONCLUSIONS

RSB with CIS is a reliable and simple method for diagnosis of abnormally innervated bowel after pull through operation for HD.

LEVEL OF EVIDENCE

III TYPE OF STUDY: Study of diagnostic test (evaluate the effectiveness of a diagnostic test or outcome assessment).

Utility of Peripherin Versus MAP-2 and Calretinin in the Evaluation of Hirschsprung Disease

Hirschsprung disease (HD) is a congenital malformation resulting from the lack of migration of ganglion cells in the colon. The absence of ganglion cells in rectal suction biopsies aids in diagnosis, but evaluation of these small biopsies can be difficult. In this study, we compare the microtubule-associated protein-2 (MAP-2), calretinin, and peripherin immunohistochemical stains in 237 selected biopsies performed to rule out HD. By H&E stain, a total of 78 biopsies had ganglion cells, whereas 83 biopsies had no ganglion cells, and an additional 76 biopsies were equivocal for ganglion cells. Of the 78 biopsies with ganglion cells, MAP-2 was positive in 73 (94%), calretinin in 76 (97%), and peripherin in 78 (100%). Of the 83 biopsies with no ganglion cells, calretinin and peripherin highlighted nerve fibrils and ganglion cells, respectively, in 3 biopsies, whereas MAP-2 was positive in only 1 biopsy. Of the 76 biopsies equivocal for ganglion cells, 16 cases were positive by all 3 stains, an additional case by both calretinin and peripherin, and 2 cases by peripherin only. All of the newly positive biopsies were from patients without HD. This study demonstrates that peripherin is superior in helping to rule out HD in these small biopsies, highlighting ganglion cells in virtually all cases with ganglion cells, whereas MAP-2 and calretinin are less sensitive for identification of ganglion cells and nerve fibrils, respectively. In patients with HD, a panel using calretinin and peripherin with or without MAP-2 may be most helpful in identifying transition zones.

A critical appraisal of the morphological criteria for diagnosing intestinal neuronal dysplasia type B

Intestinal neuronal dysplasia type B is a controversial entity expressed by complex changes in the enteric nervous system. Diagnosis depends on rectal biopsy histopathology and diagnostic criteria, both qualitative and quantitative, have changed over time, hindering the diagnostic practice. We analyzed the morphological criteria for the histological diagnosis of intestinal neuronal dysplasia type B in a series of patients with intestinal neuronal dysplasia type B according to the 1990 Frankfurt Consensus criteria and verified the applicability of the numerical criteria proposed by Meier-Ruge et al in 2004 and 2006. Qualitative criteria adopted for the histological diagnosis of intestinal neuronal dysplasia type B included hyperplasia of the submucous plexus with hyperganglionosis and hypertrophy of the nerve trunks. Quantitative criteria considered more than 20% giant ganglia in the submucosa, with more than eight neurons each on 25 ganglia, and children aged over 1 year. Distal colon surgical specimens from 29 patients, aged 0-16 years, diagnosed with intestinal neuronal dysplasia type B were retrospectively analyzed using sections processed for conventional histology (H&E) and calretinin immunohistochemistry. Hyperplasia of the submucosal nerve plexi with hyperganglionosis and hypertrophy of the nerve trunks was observed in all cases. Ganglia with small, immature neurons were detected in the majority of cases. Quantitative analysis confirmed hyperganglionosis (mean number=10.7 neurons per ganglion) and hypertrophy of the nerve trunks (median=44.6 μm thickness). Neurons showed immunostaining for calretinin, but neuron counts in calretinin-stained sections were lower compared with H&E (P<0.01). No significant differences were verified between children aged under and over 1 year regarding hyperganglionosis (P=0.79), neuron counts (P=0.36), and immature ganglia (P=0.66). Only one patient met the numerical criteria proposed by Meier-Ruge et al in 2004 and 2006. In conclusion, the numerical criteria showed limited applicability when transposed to conventional histopathology. Children aged over 1 year presented very similar histological features of neuronal immaturity to younger children, questioning the need for an age criterion when diagnosing intestinal neuronal dysplasia type B.

Diagnostic value of rectal suction biopsies using calretinin immunohistochemical staining in Hirschsprung's disease

BACKGROUND

The study investigates the diagnostic value of calretinin immunohistochemical staining (CIS) on rectal suction biopsies (RSB) in Hirschsprung's disease (HD).

METHODS

A prospective study was conducted at Children's Hospital 2 in Ho Chi Minh City, Vietnam, from January through December 2015. Patients suspected of HD during this period underwent RSB and were followed in order to assess the accuracy of the diagnostic test with CIS compared with conventional histology (H&E).

RESULTS

A total of 188 children with RSB were investigated. Median age was 7.1 (range 0.2-159) months with 65.4% boys. HD was confirmed in 80 (42.6%) children. There were 1 false positive and no false-negative cases. The sensitivity and specificity were 100% (80/80) and 99.1% (107/108) for CIS and 100% and 85.2% for H&E, respectively. Cohen's kappa coefficient was 0.9891 with a diagnostic accuracy of 99.5% for CIS, compared with 0.8303 and 91.5% for H&E, respectively. There were no serious complications related to the RSB.

CONCLUSION

RSB with CIS is a useful diagnostic method for HD, with easy interpretation and no need for cryostat. CIS has a high diagnostic accuracy and should be considered as the primary method for the diagnosis of HD by RSB.

LEVEL OF EVIDENCE

Diagnostic Studies - Level I.

Effectiveness of calretinin and role of age in the diagnosis of Hirschsprung disease

OBJECTIVE

Gold standard in Hirschsprung disease (HD) diagnosis is rectal suction biopsy (RSB) with hematoxylin and eosin and acetylcholinesterase (ACE) stainings. Sensitivity is 83-93 % and specificity is between 95 and 99 %. The rate of non-conclusive results (NC) due to inadequate samples or variability in the interpretation is about 11-37.8 %. Interpretation is still difficult in neonates. Calretinin (C) is a calcium-binding protein, expressed in the nervous system. It has been introduced as a marker to improve the diagnosis accuracy in HD. We compare sensitivity, specificity between ACE and ACE + C and investigate whether the introduction of C could reduce the rate of inconclusive results. We also tried to define the most adequate age to obtain accurate results from RSB.

METHODS

Data from patients undergoing rectal suction biopsies from 2005 to 2014 was analyzed. Until 2010 we used ACE; since 2011 we used ACE + C. The ROC curve analysis of the NC results regarding the age, showed an optimal cut-off point at 1.5 months (m). It divides the sample into two groups which we compared.

RESULTS

We analyzed 91 patients. Results of the RSB: 40 HD (44 %), 34 no Hirschsprung (37.4 %) and 17 NC (18.7 %). Sensitivity = 97.5 %, specificity = 97.1 %, not including the NC (1 false positive, 1 false negative). Results depending on the staining: ACE (n = 58) (%) ACE + C (n = 33) (%) p Sensitivity 96 100 1 Specificity 94.7 100 1 NC 24.1 9.1 0.077 Results depending on the age: <1.5 m (n = 27) (%) >1.5 m (n = 64) (%) p Sensitivity 92.3 100 0.325 Specificity 100 96.8 1 NC 40.7 9.4 0.001 CONCLUSION: Calretinin decreases the rate of inconclusive results, but not significantly. The percentage of inconclusive results decreases in patients olders than 1.5 m. Further studies are necessary to determine if this technique is useful to improve RSB results in infants younger than 1.5 months.