Clinical Trials of High-intensity Focused Ultrasound for Cancer: Concerns Arising from Low Post-Hoc Power.

Abstract

Prospective controlled trials of high-intensity focused ultrasound (HIFU) for cancers were evaluated. Post-hoc power was <0.80 in 30/46 trials and in 22/38 trials with positive results, indicating low quality in most trials. Unscientific endpoints, small sample sizes, and high dropout rates led to low post-hoc power that caused inter-trial heterogeneity and overestimated the therapeutic effect. The objective response rate was not a substitute for survival time for estimating the sample size and assessing the efficacy. The present data can interpret a paradox: HIFU is considered to have slighter cytotoxicity to noncancer tissues and no radiation but is frequently combined with chemotherapy and/or radiotherapy in practice.