Can Elective Neck Dissection for Oral Cancer Be Planned on the Basis of Depth of Invasion? Retrospective Analysis.

PURPOSE: The literature suggests that elective neck dissection (END) may be avoided in primaries with a depth of invasion (DOI) <3 mm. This study evaluated the accuracy of DOI in predicting nodal metastases (lymph node metastasis [LNM]) in oral squamous cell carcinoma (OSCC). MATERIALS AND METHODS: In this retrospective institutional chart review, pathologic N classification (pN) was correlated with DOI using Spearman's rank-order (pN1-3) and point-biserial (pN0 v pN+) correlation statistics. The receiver operating characteristic (ROC) curve was analyzed to determine the accuracy of DOI in predicting LNM. Chi-square statistics were used to measure the association between the DOI groups and pN status (pN0; pN+). Post hoc analysis of variance with Bonferroni statistics was used to compare the three DOI groups. RESULTS: DOI failed to demonstrate a strong positive correlation with the pN classification (rho = 0.31; P < .001). DOI was not strongly correlated with pN status (r(pb) = 0.27; P < .001). ROC curve analysis suggested a poor accuracy of DOI in predicting nodal metastases (AUC = 0.67 [95% CI, 0.6 to 0.73]). A statistically significant difference in LNM (pN+) was found between the DOI >10 mm and DOI </=10 mm primaries (P < .001 for DOI </=5 mm; P = .016 for DOI >5 mm and </=10 mm). However, the sensitivity and specificity achieved by ROC analysis for a DOI cutoff value of 10 mm in the present cohort were 69.3% and 57.9%, respectively. CONCLUSION: DOI is a poor indicator of LNM in OSCC. The decision to perform END should not be based on the tumor DOI.