Can endoscopic ultrasound predict early rectal cancers that can be resected endoscopically? A meta-analysis and systematic review

Abstract

Background Rectal cancers that are confined to the mucosa (T0) can be resected endoscopically. This can help the patient avoid transabdominal surgery. The published data on accuracy of endoscopic ultrasound (EUS) to predict T0 stage of rectal cancers has been varied. Aim To evaluate the accuracy of EUS in T0 staging of rectal cancers. Method (Study Selection Criteria) Only EUS studies confirmed by surgery were selected. T0 was defined as tumor confined to the mucosa. Data Collection and Extraction Articles were searched in Medline, PubMed, and CENTRAL. Statistical Method Pooling was conducted by both the fixed-effects model and random-effects model. Results An initial search identified 3,360 reference articles. Of these, 339 relevant articles were selected and reviewed. Eleven studies (N = 1,791) which met the inclusion criteria were included in this analysis. Pooled sensitivity of EUS in diagnosing T0 was 97.3% (95% CI: 93.7-99.1). EUS had a pooled specificity of 96.3% (95% CI: 95.3-97.2). The positive likelihood ratio of EUS was 21.9 (95% CI: 16.3-29.7) and negative likelihood ratio was 0.08 (95% CI: 0.04-0.15). All the pooled estimates, calculated by fixed and random effect models, were similar. The P-value for Chi-squared heterogeneity for all the pooled accuracy estimates was >0.10. Conclusions EUS has excellent sensitivity and specificity, this helps accurately diagnose T0 stage of rectal cancers. Over the past two decades, the sensitivity and specificity of EUS to diagnose T0 stage of rectal cancers has remained high. This can help physicians offer endoscopic treatment to these patients, therefore EUS should be strongly considered for staging of early rectal cancers. © Springer Science+Business Media, LLC 2009.