Diagnostic accuracy of Node-RADS for lymph node metastasis in rectal cancer: a systematic review and meta-analysis
TL;DR
Node-RADS shows good diagnostic accuracy and clinical discriminatory ability for lymph node metastasis in RC and may serve as a standardized imaging-based adjunct for preoperative nodal assessment, however, substantial heterogeneity, post hoc threshold selection, and limited external validation indicate that standalone use requires further prospective validation.
Abstract
Objective To systematically evaluate the diagnostic accuracy and clinical value of the Node Reporting and Data System (Node-RADS) for lymph node metastasis in rectal cancer (RC). Methods The study followed PRISMA-DTA guidelines and was registered in PROSPERO (CRD420261372147). PubMed, Embase, Web of Science, Cochrane Library, CNKI, Wanfang, VIP, and the Chinese Biomedical Literature Database were searched from inception to April 2026. Two reviewers independently screened studies, extracted data, and assessed methodological quality using QUADAS-2. Sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and the summary receiver operating characteristic (SROC) area under the curve (AUC) were pooled using a bivariate random-effects model. Heterogeneity, publication bias, and clinical utility were assessed using subgroup and sensitivity analyses (including exclusion of the post-neoadjuvant chemoradiotherapy cohort), Deeks’ funnel plot, and Fagan’s nomogram based on the observed prevalence. Results Seven studies involving 994 patients with RC were included, comprising 374 with and 620 without lymph node metastasis. Pooled sensitivity, specificity, PLR, NLR, and DOR were 0.82 (95% confidence interval [CI], 0.74-0.87), 0.84 (95% CI, 0.73-0.91), 5.15 (95% CI, 3.06-8.68), 0.22 (95% CI, 0.16-0.30), and 23.66 (95% CI, 12.92-43.34), respectively; the SROC AUC was 0.88 (95% CI, 0.85-0.91). Excluding the clinically distinct post-neoadjuvant chemoradiotherapy study yielded sensitivity of 0.79 (95% CI, 0.73-0.84), specificity of 0.90 (95% CI, 0.84-0.94), and AUC of 0.88 (95% CI, 0.85-0.90), while specificity heterogeneity decreased from 89.07% to 69.79%. At the observed prevalence of 37.6%, a positive Node-RADS result increased the post-test probability to approximately 76%, whereas a negative result reduced it to approximately 12%. No significant funnel-plot asymmetry was detected, although publication bias could not be excluded because only seven studies were included. Conclusion Node-RADS shows good diagnostic accuracy and clinical discriminatory ability for lymph node metastasis in RC and may serve as a standardized imaging-based adjunct for preoperative nodal assessment. However, substantial heterogeneity, post hoc threshold selection, and limited external validation indicate that standalone use requires further prospective validation. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/View/CRD420261372147, identifier CRD420261372147.