TEMPORAL BONE CARCINOMA: A NOVEL PROGNOSTIC SCORE BASED ON CLINICAL AND HISTOLOGICAL FEATURES

Background This study aimed to develop a novel temporal bone carcinoma (TBSCC) prognosis scoring system and compare it with the revised Pittsburgh staging system. Methods Forty-four consecutive TBSCC patients were assessed to identify predictors of recurrence. Each predictor’s hazard ratio for recurrence was used to develop our novel scoring system. Results Based on variables with p<0.10 in Cox’s regression model, our score included: revised Pittsburgh stage; non-anterior spread of T4 carcinoma; dural involvement; and histological grade. A higher recurrence rate (p=0.000), and shorter disease-free survival (p=0.000) were associated with scores of ≥5. The AUC of our score was larger than that of the revised Pittsburgh stage for both recurrence and disease-specific mortality (p=0.0178 and p=0.0193, respectively). Conclusions Our TBSCC scoring system is based on variables which are obtainable preoperatively from clinical and radiological data, and biopsies. Its prognostic value should be validated for published TBSCC series, then in prospective settings.