Novel Scoring Formula to Predict Survival in Patients of Primary Tongue Cancer Belonging to Tobacco Chewing Population

Indian J Surg Oncol. 2023 Dec;14(4):928-934. doi: 10.1007/s13193-023-01799-5. Epub 2023 Aug 7.

Abstract

Worldwide and in India head and neck malignancies are a major contributor to cancer mortality and morbidity. Tongue cancer predominates oral cavity cancers worldwide but in India it comes next to buccal mucosa. OPD patients after completing treatment tend to ask about the prognosis of their disease where they want an objective answer to "How long will I live?" His scoring system is intended to answer this question and guide patients for adjuvant therapy. This study enrolled all patients between 20 and 85 years old with a history of tobacco chewing at least for the last 1 year before diagnosis. Patients should have primary tongue cancer amenable to surgical resection. For survival calculation, date of diagnosis was taken as reference time. Using Kaplan-Meier survival analysis, clinicopathological factors significantly associated with survival were ascertained. Then using logit regression, a scoring system predicting patient survival in years based on clinicopathological risk factors was formulated and internal validation was done. A total 241 were enrolled and there were 69 cancer-related deaths. T stage, N stage, LVSI, and DOI were found to be significantly associated with cancer-related survival in tongue cancer patients. Another factor affecting survival was defaulting adjuvant radiation therapy. Using these variables, a survival predicting score was developed. On internal validation and regression, the score was found 80% accurate with error limits ± 6 months. It is a concise comprehensive score applicable on Indian population with history of tobacco chewing. It will not only help clinicians to tell patients about their survival expectancy but also help to counsel them for adjuvant therapy. However, external validation and if required recalibration incorporating other factors need to be done for this score.

Keywords: Logit regression; Prognostic scoring; Survival score; Tobacco chewing; Tongue cancer.