A validated predictive model for risk of nodal metastasis in node negative oral squamous cell carcinoma of the buccal mucosa and tongue. (31st March 2022)

• Record Type: Journal Article
• Authors: Kallarakkal, Thomas George; Siriwardena, Bogahawatte S. M. S.; Samaranayaka, Ariyapala; De Silva, Rohana; Tilakaratne, Wanninayake M.
• Is Part Of: Journal of oral pathology & medicine. Volume 51:Number 5(2022)
• Page Start: 436
• Page End: 443
• Publication Date: 2022-03-31
• Subjects: depth of invasion -- nodal metastasis -- oral squamous cell carcinoma -- pattern of invasion -- predictive model -- tumor site -- tumor size; Dentistry -- Periodicals; Teeth -- Diseases -- Periodicals; 617
• DOI: 10.1111/jop.13294
• Languages: English
• ISSNs: 0904-2512

Abstract

Abstract: Background: Lymph node metastasis in oral squamous cell carcinoma (OSCC) is influenced by clinical and histopathological variables. The aim of this study was to develop a simple model to predict nodal metastasis of OSCC in clinically negative necks (cN0). Methods: Data from patients who underwent surgery for treatment of OSCC of the tongue or buccal mucosa with neck dissection were used for model development and validation. Results: Nodal metastasis was significantly associated with gender, age, tumor size, site, pattern of invasion and depth of invasion on univariate analysis. All the five variables except age were retained at the variable selection step of the model development and were used in the final model because it was not significant at 0.10 significance level after adjusting for other variables. Regression coefficients of the model were used to estimate risks of nodal metastases for each combination of clinicopathological characteristics. A 10‐fold cross‐validation was used to assess the model. The average of the resultant 10 AUCs (along with its 95% confidence interval estimated using bootstrap) was used as the overall validated measure of the model. A risk chart was produced using probability of nodal metastasis predicted by the model for each combination of five characteristics. The model's ability to identify patients with nodal metastases as assessed by the area under the ROC curve (AUC) was 0.752. Conclusion: The model based on established clinicopathological variables has been internally validated on a large cohort of patients and offers practicability for use in OSCCs of the tongue and buccal mucosa.