LAUSR

INTRODUCTION This study aimed to analyze whether immunohistochemistry (IHC) is more sensitive than hematoxylin-eosin (H&E) staining for identifying perineural invasion (PNI) or lymphovascular invasion (LVI) in oral squamous cell carcinoma (OSCC). MATERIAL AND METHODS In this systematic review and meta-analysis (Prospective Register of Systematic Reviews - CRD 42021256515), data were obtained from six databases (PubMed, Scopus, LILACS, Web of Science, EBSCO, LIVIVO, Embase) and the grey literature. Cross-sectional observational studies of the diagnostic sensitivity of IHC for PNI and LVI were included. Studies were selected in two phases: first collection and reference retrieval. The Quality Assessment of Diagnostic Accuracy Studies-2 tool assessed study quality, while the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach assessed evidence quality. The meta-analysis (random effects model) was performed using MedCalc 18.2.1 software (MedCalc®) (p<0.05). RESULTS Four studies (560 patients with 295 biopsies) were analyzed. The combined sensitivity was 76% (95% confidence interval [CI], 44.30-97.19%) and specificity was 42% (95% CI, 23.40-62.02%). The positive predictive value (PPV) and negative predictive value (NPV) were 61% (95% CI, 49.78-71.53%) and 70% (95% CI, 37.63-94.43%). The overall accuracy was 58% (95% CI, 45.17-70.65%). The risk of bias was low, and GRADE analysis showed a very low certainty of evidence. CONCLUSIONS Our data suggest that IHC staining to highlight PNI/LVI may be useful in cases in which H&E analysis results in a negative decrease in the prevalence of false-negative cases and underestimated treatment.