Treatment of head and neck cancer (HNC) is associated with significant acute toxicity, and accurately identifying patients who will not tolerate aggressive treatment remains a challenge. Comorbidity indices are known prognostic factors for HNC survival outcomes. However, no index has been found to be superior, and their performance in predicting early outcomes in HNC patients is not known. This retrospective analysis aims to determine the role of comorbidity indices in predicting 90-day mortality after radical radiotherapy.
Study population included all non-thyroid, non-metastatic HNC patients who received curative intent radiotherapy, with or without chemotherapy, from 2016-2019 in a single tertiary oncology centre. 260 patients were randomly selected for model building and comparison. Multiple logistic regression was used to analyse the performance of six comorbidity scores – the Charlson Comorbidity Index (CCI), age-adjusted CCI (ACCI), head and neck CCI (HNCCI), Simplified Comorbidity Score (SCS), Adult Comorbidity Evaluation-27 (ACE-27) and the Washington University Head and Neck Comorbidity Index (WUNHCI)– in predicting risk of 90-day mortality after radical treatment. ROC analysis was done to identify the best performing index, on which further analysis was carried out to determine the best cut-off.
46 out of a total of 958 eligible patients (4.8%) died within 90 days after radical treatment in our cohort. Four of the six comorbidity scores were independent predictors of early mortality [CCI: odds ratio (OR) 1.298, p=0.004; ACE-27: OR 2.577, p<0.001; ACCI: OR 1.324, p<0.001; HNCCI: OR 2.056, p=0.001]. From ROC analysis of the remaining four comorbidity indices, ACCI was the best performing model (AUC 0.728). Calculation of the Youden Index yielded a threshold of ACCI score > 3 as the most discriminatory cut-off in predicting 90-day mortality. ACCI >3 was associated with more than 20% mortality in our cohort.
Age-adjusted CCI > 3 is associated with a higher risk of early mortality in HNC patients after radical radiotherapy. This may be taken as an easily accessible reference for discussion in the clinical setting.
The authors.
Has not received any funding.
All authors have declared no conflicts of interest.