Abstract
BACKGROUND: This study evaluates the diagnostic accuracy of ultrasound-guided fine needle aspiration (US-FNA) and core needle biopsy (US-CNB) for detecting axillary lymph nodes in women with breast cancer. METHODS: Eligible studies and pertinent literature resources were identified in Cochrane, PubMed, Embase, CNKI, VIP, and Wanfang databases using subject-specific keywords. Study outcomes were tested for heterogeneity, and meta-analyses were performed to estimate sensitivity, specificity, and diagnostic odds ratios (DORs). The summary receiver operating characteristic (SROC) curve analysis was also performed. RESULTS: A total of 22 studies involving 3,548 patients were included to evaluate the diagnostic accuracy of US-FNA and 11 studies involving 758 patients were included to evaluate the diagnostic accuracy of US-CNB in identifying axillary lymph nodes in women with breast cancer. The accuracy of US-FNA in identifying suspicious axillary lymph nodes was as follows: overall sensitivity, 79% (95% CI: 73%-84%); global specificity, 96% (95% CI: 92%-98%); overall positive likelihood ratio, 18.55 (95% CI: 10.53-32.69); overall negative likelihood ratio, 0.22 (95% CI: 0.17-0.28); DOR, 71.68 (95% CI: 37.19-138.12); and the area under the SROC curve, 0.94 (95% CI: 0.92-0.96). The accuracy of US-CNB in identifying suspicious axillary lymph nodes was as follows: overall sensitivity, 85% (95% CI: 81%-89%); global specificity, 93% (95% CI: 87%-96%); overall positive likelihood ratio, 11.88 (95% CI: 6.56-21.50); overall negative likelihood ratio, 0.16 (95% CI: 0.12-0.21); overall DOR, 66.83 (95% CI: 33.28-134.21), and the area under SROC curve 0.96 (95% CI: 0.94-0.97). CONCLUSIONS: The results indicate that both US-FNA and US-CNB have high accuracy for suspicious axillary lymph nodes.