Participation in an exercise program after primary glioblastoma diagnosis is associated with attenuated tumor volume, performance status decline, and development of sarcopenic adiposity at recurrence.

OBJECTIVE: Here, we investigate whether participation in a structured exercise program after initial diagnosis is linked to modified muscle and fat measurements, performance status, tumor characteristics, and, ultimately, overall survival (OS) in recurrent glioblastoma patients. METHODS: Included in this study were 201 consecutive glioblastoma patients (isocitrate dehydrogenase 1 (IDH1) wildtype) diagnosed between 2010 and 2023 with available primary and recurrence imaging. Of these, 25 (12.4%) participated in a structured exercise program after diagnosis. Patient characteristics at primary diagnosis and at recurrence were collected and muscle, inter-/intramuscular adipose tissue (IMAT), and subcutaneous adipose tissue (SAT) compartments of the neck were quantified on initial and recurrence radiation simulation CT scans. Time-to-event analyses were then run for OS from recurrence. RESULTS: Participants in the exercise program had lower tumor volumes (p = 0.025) and better Eastern Cooperative Oncology Group (ECOG) scores (p = 0.016) at recurrence than non-participants despite no differences at initial diagnosis (p > 0.8 for both). Exercise program participants also demonstrated reduced development of sarcopenic adiposity (i.e., a decline in muscle and an increase in IMAT and SAT measurements) in univariable and multivariable analyses. Exercise program participation, tumor volume, ECOG score, and sarcopenic adiposity measures showed univariable associations with post-recurrence OS (p < 0.05). Tumor volume at recurrence and increase in sarcopenic adiposity during treatment retained significance in multivariable models (p < 0.01) while exercise program participation itself did not (p = 0.11). CONCLUSION: This exploratory retrospective observational cohort study comparing prospectively enrolled exercise program participants with non-participants finds a close interplay between exercise program participation, muscle and fat dynamics, performance status, and tumor volume, linking structured exercise to prognostically favorable changes in patient and tumor characteristics. While results encourage exercise in glioblastoma, prospective randomized trials are needed to substantiate findings.