A scoring system to predict early oral cavity squamous cell carcinoma without long-term survival

Purpose: Oral cavity squamous cell carcinomas (OCSCCs) are the largest subgroup of head and neck squamous cell carcinoma (HNSCC), up to 50% of all. More than half of them are early OCSCCs (pT1-2N0M0). Despite higher curative rates of these early OCSCCs, nearly one-fifth of early OCSCCs patients did not have long-term survival and rare biomarkers are available to predict these events. Method: A total of 376 early OCSCC diagnosed at Chung Shan Medical University Hospital were enrolled. All of them were divided into training (225 patients, from 2010 to 2013) and validation (151 patients, from 2014-2016) cohorts. The clinical-pathological variables were compared using the χ^2 test. Cox proportional-hazards analyses were performed for overall survival (OS). A scoring system was established to predict the death occurring within five years after diagnosis. Results: The 5-year locoregional-free survival (LFS) and OS were 70.6% and 81.6%, respectively. Each of the training cohort's independent factors for the OS, including age, surgical margin, and lymphocytemonocyte ratio, was scored one point to establish a scoring system. The sensitivity, specificity, and area under the curve for the training and validation cohorts were 83.3%, 41.7%, and 0.694, and 70.0%, 50.0%, and 0.645, respectively. Conclusions: A scoring system to predict long-term survival was established for early OCSCC.