Long-Term Outcomes of Total Body Irradiation Plus Cyclophosphamide Versus Busulfan Plus Cyclophosphamide As Conditioning Regimen for Acute Lymphoblastic Leukemia: A Comparative Study

Introduction: Total body irradiation-Cyclophosphamide (TBI-Cy) and Busulfan-Cyclophosphamide (Bu-Cy) are the most commonly used conditioning regimens in allogeneic hematopoietic stem cell transplantation (allo-HCT) for acute lymphoblastic leukemia (ALL). TBI-based conditioning regimens may have a better antitumor effect at sanctuary cites compared to chemotherapy regimens. However, some controversy still remains concerning the role of TBI in adult ALL. Therefore, we aimed to retrospectively investigate long-term treatment outcomes of ALL patients post allo-HCT and identify prognostic factors and the impact of the conditioning.

Methods: We enrolled consecutive patients with ALL who underwent alloHCT from 1990 to 2016 in our center, following a TBI- or a Busulfan (Bu)-based conditioning regimen. TBI at 14.4Gy was divided at 6 fractions over 3 days and Bu (mainly oral) at 4mg/kg/day. Both regimens included 120mg/kg Cy. Graft source was mainly mobilized peripheral blood stem cells.

Results: We studied 151 ALL (114 B-ALL, 35 T-ALL and 2 biphenotypic) patients aged 28±12 years, transplanted in CR1 (60), other CR (33) or relapsed/refractory disease (58) from sibling (87), HLA matched (42) or mis-matched (17) unrelated and alternative donors (5). TBI-based conditioning was administered in 84, whereas non-TBI in 67. No differences were observed in baseline characteristics, except for disease phase, disease risk index (DRI) and graft source. The majority of non-TBI patients had relapsed/refractory disease at transplant (50.7% versus 28.6% in TBI-conditioning, p=0.043) and subsequently significantly higher DRI (p=0.025). Furthermore, bone marrow was the graft source in significantly higher number of non-TBI patients (28.4% versus 8.3% in the TBI group, p=0.001).

With a follow-up of 19.0 (0.5-170.5) in TBI and 14.5 (1.2-319.1) months in the non-TBI group, there was no difference in the incidence of acute (Grade 2-4) or chronic GVHD, thrombotic microangiopathy, bacterial or fungal infections. Only viral infections were significantly increased in the non-TBI group (59% versus 41% in the TBI group, p=0.03). Therefore, there was no significant difference in the cumulative incidence (CI) of treatment-related mortality (5-year CI of 24.1% in non-TBI versus 27.7% in the TBI group, p=0.678) or relapse mortality (5-year CI of 29.2% in non-TBI versus 37.1% in the TBI group, p=0.283).

Overall survival (OS) was similar in the TBI and non-TBI group (5-year 46.7% compared to 35.8%, p=0.201). In the multivariate analysis, unfavorable pre-transplant predictors of OS were age (p=0.024), higher disease risk index (DRI) (p=0.007) and female to male donor (p=0.006). Disease-free survival (DFS) was similar in the TBI and non-TBI group (5-year 47.1% compared to 35.4%, p=0.134), with age, higher DRI and female to male donor being unfavorable predictors. When the analysis was limited to patients transplanted in CR1 (40 in the TBI versus 20 in the non-TBI group), there were also no significant differences.

Interestingly, age predicted OS and DFS only in the TBI group. TBI patients younger than 40 years had significantly high OS (55.1%, p=0.023) and DFS (55.5%, p=0.02) versus the older ones. Older TBI patients had higher TRM (43.3% versus 20.1%, p=0.135) and relapse mortality (45.8 versus 24.7, p=0.238) that did not reach levels of statistical significance. However, the same was not true for the non-TBI group (p=0.233 for DFS, p=0.326 for OS).

Conclusion: In our retrospective data of the last 3 decades, the type of conditioning regimen did not influence the outcome of transplantation. Based on our results, high dose TBI is a feasible conditioning in terms of a prolonged DFS and OS for younger patients and seems an efficient tool to cure high-risk ALL in CR1. It should be kept in mind that intravenous BU is also used with stable pharmacokinetics. Since our data were not powered to appraise superiority of one regimen in terms of efficacy and toxicity, the choice between TBI or Bu conditioning regimens is still under consideration. Treatment is complex for adult ALL patients and remains a field of debate and a challenge for prospective studies in the future.