Rectal suction biopsy for the diagnosis of Hirschsprung's disease: a systematic review of diagnostic accuracy and complications

Rectal suction biopsy (RSB) combined with acetylcholinesterase (AChE) staining is currently the gold standard for the diagnosis of Hirschsprung's disease (HD). However, some pathologists are still reluctant to diagnose HD in RSBs as it requires thick submucosal tissue to confirm the absence of ganglion cells. Furthermore, insufficient specimens and adverse events have been reported. The objective of this study was to determine the diagnostic accuracy and incidence of complications of RSBs in patients suspected of HD based on a systematic review of the published literature. A literature-based search for relevant publications was conducted using multiple online databases. Staining for AChE/hematoxylin and eosin should have been used to evaluate RSBs. The number of true-positive, false-positive, true-negative and false-negative results was recorded and data on RSB-related complications was extracted. Pooled incidence rates and odds ratios (ORs) with 95 % confidence intervals (CI) were calculated using standardized statistical methodology. Fifty-eight studies met defined inclusion criteria, reporting a total of 14,053 RSBs. The median cohort size consisted of 72 patients (range 1-766) with a median age of 14.4 months (range 1 day-66 years) at time of biopsy. In 89.93 % (CI 89.11-90.70 %), RSB provided adequate tissue for the diagnosis of HD. Insufficient specimens were obtained in 10.07 % (CI 9.30-10.89 %) and in 8.46 % (CI 7.55-9.46 %) RSB was repeated. The incidence of HD was 19.13 % (CI 18.25-20.03 %). Mean sensitivity of RSB was 96.84 % (CI 95.57-97.47 %) and mean specificity was 99.42 % (CI 99.17-99.57 %). The overall complication rate was 0.65 % (CI 0.46-0.91 %) with persistent rectal bleeding requiring blood transfusion in 0.53 % (CI 0.36-0.77 %), bowel perforation in 0.06 % (CI 0.02-0.18 %) and pelvic sepsis in 0.06 % (CI 0.02-0.18 %). Complications were significantly more frequent in newborns and infants compared to older children (OR 9.00 [CI 4.75-17.07], p < 0.0001). RSB combined with AChE staining is a simple, safe and accurate method for the diagnosis of HD. The risk for RSB-related complications is higher in newborns and infants.