Radiation Oncology

Dyadic yoga during thoracic RT improves 6‑MWT distance and physical QoL

In a randomized trial of 222 thoracic cancer patients and caregivers, dyadic yoga increased 6‑minute walk distance versus education/support (LSM 469 m v 441 m, P=.03). Patients assigned to yoga reported higher physical QoL across end‑of‑treatment and follow‑up (LSM 44.7 v 41.6, P=.03). Exploratory benefits included better sleep and coping, and caregiver involvement may have supported adherence. This intervention applies to patients receiving ≥5 weeks of thoracic RT and their family caregivers; implementation requires trained instructors and schedule integration. 1

Framework for estimands and methods to handle missing PROs after progression

This methodological paper recommends specifying explicit estimands for PRO endpoints when disease progression causes missing data. It contrasts hypothetical and treatment‑policy strategies and demonstrates methods like implicit multiple imputation, joint models, and control‑based imputation. Authors provide a tutorial showing supplementary analyses can increase confidence in PRO interpretations despite intercurrent events. Apply these approaches in trial planning and analysis; treat results from any single method as assumption‑dependent. 2

Radiomics models predict ART response variably; strongest data for brain mets

A scoping review of 29 studies (3946 patients) found radiomics AUCs for ablative RT response ranged 0.69–0.95, with deep learning often highest (AUC 0.85–1.00) mainly for brain metastases. Methodological quality was moderate and external validation was limited, especially for extracranial sites. Authors conclude models are promising but not ready for routine clinical decision‑making without prospective multicentre validation and standardization. Consider radiomics findings as hypothesis‑generating and require robust local validation before altering ablative RT candidacy. 3