The predictive value of cervical lymph node metastasis through sentinel lymph node biopsy in patients with oral cancer: A meta-analysis

Objective: The purpose of this meta-analysis was to evaluate the predictive value of cervical lymph node metastasis through sentinel lymph node biopsy (SLNB) in patients with oral cancer. Materials and Methods: Two authors independently searched the databases of PubMed, Cochrane central register of controlled trials, EMBASE, and CNKI to find the potential suitable diagnostic study related to SLNB. The diagnostic sensitivity, specificity, positive likely hood ratio (+LR), negative likely hood ratio (uLR), diagnostic odds ratio (DOR) were pooled by Meta-DiSc1.4 software. The publication bias was assessed by funnel plot and line regression test. Results: After electronic searching the related databases, we finally included twenty diagnostic studies. The pooled sensitivity, specificity, +LR, and DOR were 0.91 (95% confidence interval [CI]: 0.88u0.94), 1.00 (95% CI: 0.99u1.00), 35.52 (95% CI: 19.19u65.75), and 323.50 (95% CI: 148.27u705.83), respectively, with fixed-effect model. Moreover, the pooled uLR was 0.13 (95% CI: 0.07u0.23) by random-effect model. The area under the summary receiver operating characteristic curve of SLNB for cervical lymph node metastasis was 0.99. Conclusion: SLNB had very high sensitivity and specificity for productizing cervical lymph node metastasis in oral cancer patients.