Sarcopenia and body composition predict survival and reveal immunometabolic dysfunction via single-cell profiling in anti-BCMA CAR T-treated myeloma

Abstract Introduction: Chimeric Antigen Receptor (CAR) T-cell therapy has revolutionized the treatment of relapsed/refractory multiple myeloma (RRMM), achieving unprecedented responses even in heavily pretreated patients. However, outcomes remain highly variable, shaped by a multifactorial interplay of patient characteristics and disease biology. While these determinants are increasingly well-defined, the role of body composition - particularly adipose tissue and skeletal muscle distribution - has not been systematically evaluated in RRMM. Methods: We retrospectively evaluated 108 RRMM patients treated with Idecabtagene vicleucel (ide-cel) or Ciltacabtagene autoleucel (cilta-cel) between 2021 and 2025. BMI was calculated before CAR T-cell infusion and pre-treatment CT scans were used to quantify total (TAT), subcutaneous (SAT), and visceral adipose tissue (VAT) as well as skeletal muscle (SM) area. For survival analyses we used cohort-derived optimal cut-off values for adipose tissue (TAT: 227.24 cm², SAT: 118.85 cm², VAT: 108.39 cm²) to stratify groups, while established cut-offs were used to define sarcopenia (SM <52.4 cm²/m² for men and <38.5 cm²/m² for women). Immune profiling of peripheral blood was performed at leukapheresis (LP) and on days 0, 7, 14, 30, and 100 after CAR T-cell infusion, including assessments of CAR T-cell expansion. Single-cell RNA and T-cell receptor sequencing (scRNA/TCR-seq) was conducted on 178 longitudinal samples obtained at LP, as well as on days 30 and 100 post-infusion from 75 patients with available CT data, to further explore immunological effects of body composition. Results: The median age was 64.3 years, with 67% receiving cilta-cel and 33% ide-cel. Median PFS was 20.6 months, while median OS was not reached (12-month PFS and OS: 70.1% and 81.7%, respectively). Sarcopenia was present in 30% of patients, while low TAT, SAT, and VAT was observed in 25%, 28%, and 33%, respectively. Body composition parameters were independent of R-ISS, Endothelial activation and stress index (EASIX), CAR T-cell product, number of prior therapy lines and CRS or ICANS. However, low SAT was linked to elevated levels of soluble BCMA (sBCMA) before infusion (161 vs 59.7ng/mL, p = 0.03). Normal (n=47) or low (n=3) BMI was associated with shorter overall survival (OS) compared to overweight (n=44) or obese (n=14) patients (HR=3.38, 95% CI: 1.05–11.01, p=0.03). Similarly, low TAT predicted inferior OS (HR=4.02, 95% CI:1.35–11.98, p=0.007), primarily driven by low SAT (HR=4.45, 95% CI: 1.45–13.63, p=0.004), while VAT had less impact on OS (HR=2.54, 95% CI:0.85-7.59, p=0.08). Sarcopenia predicted adverse OS (HR=3.35, 95% CI: 1.13–10.0, p=0.02), whereas progression-free survival (PFS) was not affected by adipose tissue parameters or sarcopenia. In a univariate cox regression analysis, low SAT as well as sarcopenia were significantly associated with reduced OS (p<0.05 respectively) and multivariate models identified low SAT as strongest independent predictor on OS (p=0.02) with no effects on PFS. At LP, both low SAT (CD3+: 174 vs 478*106/L, p=0.005; CD8+: 84 vs 275*106/L, p=0.002 and CD8+ Effector memory (CD8+EM): 38.6 vs 102*106/L, p=0.01) and sarcopenia (CD3+: 37 vs 96.3*106/L, p=0.03) correlated with reduced bystander T-cell counts, which was confirmed at single-cell resolution revealing diminished CD4+ granulysin-expressing cytotoxic (CD4+CTL_GNLY), CD8+EM, CD8+EM re-expressing CD45RA 1 (EMRA.1) and CD8+EMRA.2 T-cell populations, alongside reduced T-cell clonality (p < 0.05 respectively). Post infusion, T cells from low SAT and sarcopenic patients similarly exhibited impaired cytotoxic and interferon signaling, together with NF-κB activation and elevated mitochondrial OXPHOS pathway, rendering exhaustion and functional immunological impairment in CD4+ and CD8+ T-cell subsets. Analysis of the effector (CAR T-cell counts) to target (sBCMA) ratio revealed a significantly decreased ratio in low SAT patients at day 30 after infusion (1.0 vs 7.8, p=0.04). Conclusion: Low SAT and sarcopenia emerge as prognostic factors in RRMM CAR T-cell therapy affecting survival, response and immunometabolic profiles. Their quantification through routine imaging enables cost-effective integration into clinical workflows providing an opportunity for early risk stratification. Targeted interventions such as exercise, nutritional support and metabolic optimization may enhance immune fitness and therapeutic outcomes.