We aimed to investigate the prognostic value of tumour-infiltrating lymphocytes’ (TILs) expression in pretreatment specimens from patients with head and neck squamous cell carcinoma (HNSCC) treated with definitive chemoradiotherapy (CRT). The prevalence of CD3+, CD8+, CD4+ and FOXP3+ TILs was assessed using immunohistochemistry in tumour tissue obtained from 101 patients before CRT and was correlated with clinicopathological characteristics as well as local failure-free- (LFFS), distant metastases free- (DMFS), progression-free (PFS) and overall survival (OS). Survival curves were measured using the Kaplan–Meier method, and differences in survival between the groups were estimated using the log-rank test. Prognostic effects of TIL subset density were determined using the Cox regression analysis. With a mean follow-up of 25 months (range, 2.3–63 months), OS at 2 years was 57.4% for the entire cohort. Patients with high immunohistochemical CD3 and CD8 expression had significantly increased OS (P=0.024 and P=0.028), PFS (P=0.044 and P=0.047) and DMFS (P=0.021 and P=0.026) but not LFFS (P=0.90 and P=0.104) in multivariate analysis that included predictive clinicopathologic factors, such as age, sex, T-stage, N-stage, tumour grading and localisation. Neither CD4 nor FOXP3 expression showed significance for the clinical outcome. The lower N-stage was associated with improved OS in the multivariate analysis (P=0.049). The positive correlation between a high number of infiltrating CD3+ and CD8+ cells and clinical outcome indicates that TILs may have a beneficial role in HNSCC patients and may serve as a biomarker to identify patients likely to benefit from definitive CRT.

Radiomics applies machine learning algorithms to quantitative imaging data to characterise the tumour phenotype and predict clinical outcome. For the development of radiomics risk models, a variety of different algorithms is available and it is not clear which one gives optimal results. Therefore, we assessed the performance of 11 machine learning algorithms combined with 12 feature selection methods by the concordance index (C-Index), to predict loco-regional tumour control (LRC) and overall survival for patients with head and neck squamous cell carcinoma. The considered algorithms are able to deal with continuous time-to-event survival data. Feature selection and model building were performed on a multicentre cohort (213 patients) and validated using an independent cohort (80 patients). We found several combinations of machine learning algorithms and feature selection methods which achieve similar results,e.g., MSR-RF: C-Index = 0.71 and BT-COX: C-Index = 0.70 in combination with Spearman feature selection. Using the best performing models, patients were stratified into groups of low and high risk of recurrence. Significant differences in LRC were obtained between both groups on the validation cohort. Based on the presented analysis, we identified a subset of algorithms which should be considered in future radiomics studies to develop stable and clinically relevant predictive models for time-to-event endpoints.

Because of the unprecedented disruption of health care services caused by the COVID-19 pandemic, the American Society of Radiation Oncology (ASTRO) and the European Society for Radiotherapy and Oncology (ESTRO) identified an urgent need to issue practice recommendations for radiation oncologists treating head and neck cancer (HNC) in a time of limited resources and heightened risk for patients and staff. A panel of international experts from ASTRO, ESTRO, and select Asia-Pacific countries completed a modified rapid Delphi process. Topics and questions were presented to the group, and subsequent questions were developed from iterative feedback. Each survey was open online for 24 hours, and successive rounds started within 24 hours of the previous round. The chosen cutoffs for strong agreement (≥80%) and agreement (≥66%) were extrapolated from the RAND methodology. Two pandemic scenarios, early (risk mitigation) and late (severely reduced radiation therapy resources), were evaluated. The panel developed treatment recommendations for 5 HNC cases. In total, 29 of 31 of those invited (94%) accepted, and after a replacement 30 of 30 completed all 3 surveys (100% response rate). There was agreement or strong agreement across a number of practice areas, including treatment prioritization, whether to delay initiation or interrupt radiation therapy for intercurrent SARS-CoV-2 infection, approaches to treatment (radiation dose-fractionation schedules and use of chemotherapy in each pandemic scenario), management of surgical cases in event of operating room closures, and recommended adjustments to outpatient clinic appointments and supportive care. This urgent practice recommendation was issued in the knowledge of the very difficult circumstances in which our patients find themselves at present, navigating strained health care systems functioning with limited resources and at heightened risk to their health during the COVID-19 pandemic. The aim of this consensus statement is to ensure high-quality HNC treatments continue, to save lives and for symptomatic benefit.