External validation of OroGrams as a predictive model for overall and progression-free survival in Scottish patients with oropharyngeal squamous cell carcinoma: a retrospective cohort study.

The aims of this study were to externally validate the OroGrams (oropharyngeal cancer survival calculator) nomogram in a Scottish cohort with high endemic smoking rates, and to compare survival rates with the original validation cohort. A retrospective cohort study in Greater Glasgow and Clyde ENT and OMFS departments was performed to investigate survival outcomes of patients with oropharyngeal squamous cell carcinoma (OPSCC) from January 2012 - December 2017. The main outcome measures were progression-free (PFS) and overall survival (OS), and Kaplan Meier curves. The predictive accuracy of OroGrams was investigated for survival probabilities at one, three, and five years. Smoking and HPV-negative cancer rates were significantly higher in Scottish patients than in the UK consortium cohort. A greater proportion of Scottish patients had advanced UICC8 stages. PFS in patients with HPV-negative cancer appeared better in the Scottish cohort than in the original cohort. Calculated mean standardised Brier scores for PFS and OS were below 0.2 at all three follow-up points, suggesting good overall prognostic accuracy, but there was wide variation between predictive accuracy in individual patients. OroGrams shows prognostic accuracy at one-year follow up in Scotland. The accuracy decreases with longer follow-up periods.