Higher Total Body Irradiation Dose Intensity in Fludarabine/TBI-Based Reduced-Intensity Conditioning Regimen Is Associated with Inferior Survival in Non-Hodgkin Lymphoma Patients Undergoing Allogeneic Transplantation

Abstract

Disease relapse is the most common cause of therapy failure in patients with non-Hodgkin lymphoma (NHL) undergoing reduced-intensity conditioning (RIC) allogeneic hematopoietic cell transplantation (allo-HCT). It is not known whether or not increasing total body irradiation (TBI) dose from 2 to 4 Gy in a RIC platform can provide improved disease control without increasing nonrelapse mortality (NRM). Using the Center for International Blood & Marrow Transplant Research (CIBMTR) database, we evaluated the outcomes of patients with NHL receiving RIC allo-HCT with either fludarabine (Flu)/2-Gy TBI versus Flu/4-Gy TBI. In the CIBMTR registry, 413 adult patients with NHL underwent a first allo-HCT using either a matched related or unrelated donor between 2008 and 2017, using a RIC regimen with either Flu/2-Gy TBI (n = 349) or Flu/4-Gy TBI (n = 64). The primary endpoint was overall survival (OS). Secondary endpoints included acute (a) and chronic (c) graft-versus-host disease (GVHD), NRM, relapse/progression, and progression-free survival (PFS). At baseline, the Flu/2-Gy TBI cohort had significantly fewer patients with Karnofsky performance status 90 and significantly more patients had a higher HCT-comorbidity index. On multivariate analysis, the 2 conditioning cohorts were not significantly different in terms of risk of grade 3 to 4 aGVHD or cGVHD. Compared to Flu/2-Gy TBI, the Flu/4-Gy TBI conditioning was associated with a significantly higher risk of NRM (hazard ratio [HR], 1.79; 95% confidence interval [CI], 1.11 to 2.89; P = .02) and inferior OS (HR, 1.51; 95% CI, 1.03 to 2.23, P = .03). No significant differences were seen in the risk of relapse/progression (HR, 0.78; 95% CI, 0.47 to 1.29, P = .33) or PFS (HR, 1.09; 95% CI, 0.78 to 1.54, P = .61) between the 2 regimens. Comparing Flu/2-Gy TBI versus Flu/4-Gy TBI cohorts, the 5-year adjusted outcomes were NRM (28% versus 47%; P = .005), relapse/progression (35% versus 29%; P = .28), PFS (37% versus 24%; P = .03), and OS (51% versus 31%; P = .001), respectively. Relapse was the most common cause of death in both cohorts. In patients with NHL undergoing Flu/TB I-based conditioning, augmenting TBI dose from 2 to 4 Gy is associated with higher NRM and inferior OS, without any significant benefit in terms of disease control. The optimal dose is 2-Gy in the RIC Flu/TBI platform for lymphomas.

Reduced-intensity conditioning hematopoietic stem cell transplantation for chronic lymphocytic leukemia and Richter’s transformation

“…Interestingly, the CIBMTR Lymphoma Working Committee recently presented registry data illustrating that a higher dose of TBI (400 cGy vs 200 cGy) for patients with NHL undergoing TBI-containing RIC-HSCT led to significantly higher rates of NRM and overall mortality without increasing the risk of GVHD, concluding that 200 cGy was the optimal TBI dosing in those patients. 27 In our cohort, 36 (90%) of 40 patients who underwent TBI-containing RIC received 200 cGy dosing and only 4 (10%) received 400 cGy. Thus, the TBI dosing intensity is unlikely the reason for our finding.…”

Machine Learning for the Prediction of Survival Post-Allogeneic Hematopoietic Cell Transplantation: A Single-Center Experience

“…The significance of some of these variables is well established in the literature, such as the decreased risk associated with the use of HLA-matched related donors [32] and a higher FEV1 pretransplant [5], or the increased risk associated with older age at HCT [32]. Other variables determined to have a significant impact on outcome in our study have not been extensively examined, such as the increased risk associated with lower baseline HGB at HCT [33], or with an increased dose of TBI with conditioning [34]. Of note, data were available to calculate the laboratory value-based pre-HCT EASIX score [17] for 30% of the cohort; however, the inclusion of this variable did not impact the performance of the tested ML models.…”