Systematic review with meta-analysis: neoplasia detection rate and post-endoscopy Barrett’s neoplasia in Barrett’s oesophagus

Abstract

Background: Neoplasia detection rate, the proportion of Barrett's oesophagus patients with high-grade dysplasia or oesophageal adenocarcinoma detected at index surveillance endoscopy has been proposed as a quality metric. However, the correlation between neoplasia detection rate and a clinically relevant outcome like post-endoscopy Barrett's neoplasia remains unknown. Post-endoscopy Barrett's neoplasia refers to the rate of high-grade dysplasia or oesophageal adenocarcinoma on repeat endoscopy within one year of an index screening examination revealing non-dysplastic Barrett's oesophagus or low-grade dysplasia. Aim: To assess correlation between neoplasia detection rate and post-endoscopy Barrett's neoplasia. Methods: We performed a systematic search of multiple databases from date of inception to June 2021 to identify cohort studies reporting both neoplasia detection rate and post-endoscopy Barrett's neoplasia. Data from each study were pooled using a random effects model, and their correlation assessed using meta-regression. Heterogeneity was assessed and a priori planned subgroup analyses were conducted. Results: Ten studies with 27 894 patients with Barrett's oesophagus were included. The pooled neoplasia detection rate and post-endoscopy Barrett's neoplasia were 5.0% (95% CI: 3.4%-7.1%, I2 = 97%) and 19.6% (95% CI: 10.1%-34.7%, I2 = 96%), respectively. Meta-regression revealed a statistically significant inverse relationship between the two variables (coefficient -3.50, 95% CI: -4.63 to -2.37, P < 0.01). With every 1% increase of neoplasia detection rate, post-endoscopy Barrett's neoplasia decreased by 3.50%. Heterogeneity was high despite adjusting for study quality and performing several subgroup analyses. Conclusion: We observed a statistically significant inverse correlation between neoplasia detection rate and post-endoscopy Barrett's neoplasia. Additional studies are needed to further validate this correlation.