Machine Learning Incorporating Host Factors for Predicting Survival in Head and Neck Squamous Cell Carcinoma Patients

Simple Summary Among head and neck squamous cell carcinoma patients, the five-year survival rates have seen little improvement over the past decade. Prediction of a cancer patient's clinical outcome is challenging but important for patient counseling and treatment planning. In this work, we evaluated common machine learning models in predicting head and neck squamous cell carcinoma patients' overall survival based on clinical, demographic features and host factors. We identified the top-performing model and verified host factors can improve the model performance when proper methods are applied. The findings are of critical importance for improved risk stratification of head and neck squamous cell carcinoma patients and provide targeted supportive care for patients who are likely to have the worst outcome. Prognostication for cancer patients is integral for patient counseling and treatment planning, yet providing accurate prediction can be challenging using existing patient-specific clinical indicators and host factors. In this work, we evaluated common machine learning models in predicting head and neck squamous cell carcinoma (HNSCC) patients' overall survival based on demographic, clinical features and host factors. We found random survival forest had best performance among the models evaluated, which achieved a C-index of 0.729 and AUROC of 0.792 in predicting two-year overall survival. In addition, we verified that host factors are independently predictive of HNSCC overall survival, which improved the C-index by a margin of 0.026 and the AUROC by 0.034. Due to the strong correlation among host factors, we showed that proper dimension reduction is an important step before their incorporation into the machine learning models, which provides a host factor score reflecting the patients' nutrition and inflammation status. The score by itself showed excellent discriminating capacity with the high-risk group having a hazard ratio of 3.76 (1.93-7.32, p < 0.0001) over the low-risk group. The hazard ratios were further improved to 7.41 (3.66-14.98, p < 0.0001) by the random survival forest model after including demographic and clinical features.

Automated summary

Predicting the overall survival of head and neck squamous cell carcinoma patients remains challenging, but host factors can enhance the performance of prediction models. Proper dimension reduction of host factors before incorporating them into machine learning models is crucial for accurate risk assessment and targeted care for high-risk patients.