The impact of institutional clinical trial recruitment versus hospital volume on survival outcomes of patients with head and neck cancer: An analysis of the PET-NECK trial outcomes, UKCRN portfolio, and Hospital Episode Statistics (HES) in England.

OBJECTIVES: High institutional clinical trial recruitment and high hospital volume are reported to be independent indicators of better patient outcomes following cancer treatment. However, their relationship in head and neck cancers (HNC) remains less clear.

METHODS: We aimed to assess the relationship between institutional clinical trial recruitment, hospital throughput of HNC cases, and survival of patients with advanced HNC treated with primary chemoradiotherapy at hospitals which recruited to the PET-NECK trial (2008-2012). The impact on outcome was assessed using Cox's proportional hazards regression analysis and multivariate analysis.

RESULTS: HNC RCT recruitment positively correlated with hospital throughput (r = 0.57, p < 0.0001). Low-recruiters (1-5 patients) had a 107% increased risk of death when compared to high-recruiters (>5 patients) (HR = 2.07, p = 0.05). There was no significant impact of hospital throughput on overall or disease-specific HNC survival. Multivariate analysis identified p16 status, N-stage, smoking, and RCT recruitment volume as the only significant predictors of survival. There was a significant difference in chemotherapy regimen between low and high-recruiters (p = 0.003) where a higher proportion of patients (50%, n = 13) in low-recruiting compared to high-recruiting hospitals (29%, n = 92) received neoadjuvant chemotherapy. A higher proportion of these patients died at low-recruiting hospitals (46% versus 23%).

DISCUSSION: A significant association exists between high recruitment and better OS for patients with HNC. However, no significance was found between hospital throughput and outcomes. The significance of individual centre differences in chemotherapy regimen needs further investigation. Future studies need a greater number of patient outcome events to support the trends found in this study.