Tailoring TBI for vulnerable patients: 8 gy TBI shows comparable outcomes to 12 gy TBI in adult ALL allo-HSCT

Abstract Introduction: Total body irradiation (TBI) is a standard component of conditioning regimens for allogeneic hematopoietic stem cell transplantation (allo-HSCT) in acute lymphoblastic leukemia (ALL). While 12 Gy TBI is the myeloablative standard, data are limited regarding the efficacy of a reduced-intensity 8 Gy approach. A recent report by Spyridonidis et al. (Hemasphere. 2023) suggested non-inferiority of FLU/TBI 8 Gy to FLU/TBI 12 Gy in ALL patients in first complete remission (CR), which prompted our investigation into the utility of 8 Gy TBI compared to 12 Gy TBI. Methods: We retrospectively analyzed adult ALL patients who underwent a first allo-HSCT at Toranomon Hospital between 2000 and 2025. Major endpoints included overall survival (OS), leukemia-free survival (LFS), relapse incidence (CIR), non-relapse mortality (NRM), and acute and chronic graft-versus-host disease (aGVHD and cGVHD). Results: A total of 219 adult ALL patients underwent a first allo-HSCT at Toranomon Hospital between 2000 and 2025 and were included in this analysis. The median age was 46 years (range, 17-69 years), and 124 patients were male. Diagnoses included Ph+ B-ALL (n=100), Ph-negative B-ALL (n=86), and T-ALL (n=33). Of these, 83 patients received 8 Gy TBI, while 136 patients received 12 Gy TBI. The median follow-up time for the entire cohort was 827 days (range, 9-7649).Several baseline characteristics significantly differed between the two TBI dose groups. The 8 Gy TBI group was significantly older (median 58 vs. 37.5 years in 12 Gy group; p < 0.001). They also had a lower proportion of later-period transplants (2011-2025: 66.3% vs. 80.1% for 12 Gy group, p = 0.033) and a higher percentage of PS ≥2 (10.8% vs. 2.9%, p = 0.035). Conditioning regimens differed significantly (p < 0.001); 8 Gy patients predominantly received fludarabine/melphalan-based regimens (69.9%), while 12 Gy patients primarily received cyclophosphamide-based regimens (68.4%). Other factors, including donor source (p = 0.09), diagnosis (p = 0.33), disease status at allo-HCT (p = 0.43), and HCT-CI (p = 0.30), showed no significant differences. Regarding primary endpoints, the 2-year OS rates were comparable (66.5% [95% CI: 56.6-78.0%] for 8 Gy TBI vs. 66.3% [58.3-75.5%] for 12 Gy TBI; p = 0.34). Similarly, 2-year LFS rates showed no significant difference (64.3% [54.4-76.0%] vs. 57.0% [48.9-66.6%]; p = 0.64). The 2-year CIR was 18.4% in the 8 Gy TBI group and 31.5% in the 12 Gy TBI group, a non-significant difference (p = 0.22). The 2-year NRM was significantly higher in the 8 Gy TBI group (17.3% vs. 10.5%; p = 0.045). No significant differences were observed in the 100-day cumulative incidence of Grade II-IV aGVHD (58.2% vs. 64.0%; p = 0.53), Grade III-IV aGVHD (26.3% vs. 24.7%; p = 0.73), or 1-year cGVHD (31.9% vs. 30.6%; p = 0.77). Multivariate Cox regression analysis identified independent prognostic factors for outcomes. For OS, non-CR disease status at allo-HSCT (HR = 3.03 [95% CI: 1.89-4.86]; p < 0.001) and PS ≥2 (HR = 2.51 [1.19-5.29]; p = 0.015) were identified as significant adverse prognostic factors. For LFS, non-CR disease status (HR = 2.94 [1.84-4.68]; p < 0.001) and PS ≥2 (HR = 4.46 [2.16-9.22]; p < 0.001) were also significant adverse prognostic factors. Additionally, a fludarabine/melphalan/TBI regimen (HR = 0.48 [0.24-0.98]; p = 0.045) was identified as a favorable factor for LFS. Importantly, the TBI dose (8 Gy vs. 12 Gy) was not found to be a significant prognostic factor for either OS (HR = 1.01 [0.47-2.16]; p = 0.99) or LFS (HR = 0.83 [0.40-1.74]; p = 0.63).In subgroup analysis by disease status at allo-HSCT, for patients in CR, the 2-year OS rates were 80.9% (8 Gy) vs. 74.8% (12 Gy) (p = 0.82), and 2-year LFS rates were 79.3% vs. 67.4% (p = 0.95). In non-CR, 2-year OS rates were 29.2% vs. 33.8% (p = 0.33), while 2-year LFS rates were 24.5% vs. 17.7% (p = 0.74). When limiting the analysis to patients who received the fludarabine/melphalan/TBI regimen, 2-year OS was 69.9% (8 Gy) vs. 82.2% (12 Gy) (p = 0.09), and 2-year LFS rates were 66.2% vs. 79.7% (p = 0.040). Conclusion: Despite the 8 Gy TBI group comprising older and less fit patients with different baseline characteristics, our study demonstrated comparable long-term outcomes, including overall and leukemia-free survival, to the 12 Gy TBI group. These findings suggest that 8 Gy TBI may be a valuable conditioning option, particularly for patients who may not tolerate the standard 12 Gy dose.