KEY POINTS
- This retrospective two-center UK study evaluated associations between CT-derived skeletal muscle characteristics and overall survival in patients receiving curative-intent treatment for urothelial bladder cancer between 2014 and 2024. The final population included 353 patients treated with radiotherapy-based bladder preservation and 117 treated with radical cystectomy. Median follow-up was 5.19 years in the radiotherapy cohort and 4.86 years in the surgical cohort, with 182 and 50 deaths, respectively.
- Muscle characteristics were measured using an AI-based CT segmentation pipeline at the L5 vertebral level, selected because this region was more consistently included in radiotherapy planning CT than the conventional L3 reference level. Muscle quantity was expressed as skeletal muscle index (SMI; cm²/m²), while muscle quality was assessed using the intramuscular adipose tissue ratio (IMATR), reflecting the proportion of fat within the skeletal muscle compartment. A prespecified directed acyclic graph (DAG) informed covariate adjustment for age, sex, comorbidity, nutritional status, systemic inflammation, and socioeconomic status.
- The radiotherapy and surgical populations differed substantially before outcome analysis. Radiotherapy patients were older (median 74 vs 72 years; p=0.026), had lower SMI (33.9 vs 42.1 cm²/m²; p<0.001), and higher IMATR (24.6% vs 12.7%; p<0.001). They also had less favorable performance status, with 85% versus 98% having ECOG 0–1. These differences reflect distinct treatment-selection pathways and imaging conditions, making direct comparisons between the groups inappropriate for evaluating the relative effectiveness of radiotherapy and cystectomy.
- Higher muscle quantity was associated with improved survival after bladder-preserving radiotherapy. In adjusted Cox models involving 336 evaluable RT patients and 166 deaths, each cohort- and sex-specific standard deviation increase in SMI was associated with lower mortality (HR 0.79, 95% CI 0.66–0.94; p=0.0068). Expressed in more clinically interpretable units, each 5 cm²/m² increase in SMI corresponded to HR 0.85 (95% CI 0.75–0.95). In unadjusted analyses, patients with SMI above the sex-specific median had approximately 7.0 years median OS versus 4.4 years below the median (p=0.011).
- Greater intramuscular fat infiltration was independently associated with worse survival in the radiotherapy cohort. Each cohort- and sex-specific standard deviation increase in IMATR corresponded to HR 1.24 (95% CI 1.04–1.48; p=0.014), while each 5-percentage-point increase corresponded to HR 1.13 (95% CI 1.02–1.24). Median OS was approximately 4.3 years among patients with IMATR above the median versus 7.0 years below the median (p=0.016). However, model discrimination remained modest, with concordance statistics of only 0.604 for SMI and 0.599 for IMATR, limiting their suitability for individual survival prediction.
- The cystectomy cohort showed different but statistically inconclusive associations. Higher SMI was associated with a nonsignificant trend toward higher mortality (HR 1.35, 95% CI 0.99–1.85; p=0.057), whereas higher IMATR showed a nonsignificant trend toward lower mortality (HR 0.86, 95% CI 0.62–1.21; p=0.39). These estimates were unstable because the primary surgical models involved 15 parameters for only 50 deaths. Reducing model complexity and applying bootstrap correction lowered apparent discrimination to 0.606 for SMI and 0.593 for IMATR, reinforcing that the surgical findings are exploratory.
- Important limitations include retrospective selection, incomplete nutritional and frailty measurements, lack of cancer-specific mortality data, and differences in CT acquisition between institutions. The L5-based intramuscular fat measurement had limited independent validation, and muscle measurements could have been influenced by neoadjuvant systemic treatment before imaging. Although sensitivity analyses broadly supported the radiotherapy findings, the study did not test treatment-effect modification and cannot determine whether patients with particular muscle characteristics benefit more from bladder-preserving RT than from cystectomy.
CLINICAL TAKEAWAY
CT-derived skeletal muscle quantity and fat infiltration may provide complementary prognostic information in patients undergoing bladder-preserving radiotherapy, potentially reflecting nutritional status, frailty, systemic inflammation, and treatment resilience. However, these observational associations are not sufficiently accurate for standalone risk prediction and do not support using body composition to choose between radiotherapy and radical cystectomy without prospective, externally validated comparative studies.