Predicting pathological lymph node status in clinical stage I/II tongue cancer.

BACKGROUND

It is difficult to accurately predict occult cervical lymph node metastasis, a major factor that significantly influences treatment outcomes in patients with clinical stage I/II tongue cancer. This study aimed to develop a predictive model for occult cervical lymph node metastasis in patients with stage I/II tongue cancer based on clinical and histopathological findings.

METHODS

This multicenter cross-sectional study analyzed the archived pathological specimens and clinical records of patients diagnosed with stage I/II tongue cancer between January 2012 and March 2020 at seven institutions in Japan. All patients underwent transoral partial glossectomy with elective neck dissection, and the histological depth of invasion (DOI) was confirmed to be 3-10 mm. The clinicopathological features of the primary tumors were compared with those of occult cervical lymph node metastases. Predictive factors for occult cervical lymph node metastasis were identified using the least absolute shrinkage and selection operator (LASSO) regression model, and the model performance was evaluated using receiver operating characteristic (ROC) analysis.

RESULTS

In total, 106 patients were included in this study. The LASSO regression analysis identified three significant predictors of occult cervical lymph node metastasis: vascular invasion, poorly differentiated clusters (PDC), and DOI. The predictive model incorporating these factors achieved an area under the curve (AUC) of 0.713.

CONCLUSION

Vascular invasion, PDC, and DOI are the key histopathological predictors of occult cervical lymph node metastasis in early-stage tongue cancer. External validation in larger cohorts is warranted to validate the utility of the predictive model developed in this study.