Primary Central Nervous System Involvement at Initial Diagnosis Remains an Independent Risk Factor for Relapse in Acute Lymphoblastic Leukemia after Allogeneic Hematopoietic Cell Transplantation in CR1

Background: A recent study from the Acute Leukemia Working Party of EBMT demonstrated that outcomes of allogeneic hematopoietic cell transplantation (allo-HCT) for adults with acute lymphoblastic leukemia (ALL) have improved significantly over time and that total body irradiation (TBI) should be considered as the preferable type of myeloablative conditioning (MAC). This study, however, did not compare outcomes of allo-HCT in patients with CNS involvement (CNS-pos) vs. those without CNS disease (CNS-neg).

Study population: Here, we evaluate post allo-HCT outcomes of 547 patients (CNS-pos at initial presentation=96, CNS-neg=451) who underwent the procedure in first complete remission (CR1) between 2009 and 2019 at an EBMT participating transplant center. The distribution of ALL subtypes were as follows: CNS-pos (Ph-neg B ALL=28%, Ph-pos B ALL=27%, and T-cell ALL=45%) and for CNS-neg (Ph-neg B ALL=21%, Ph-pos B ALL=44%, and T-cell ALL=35%), p=0.01. The primary endpoint was leukemia-free survival (LFS).

Results: The median follow up was not statistically different between the CNS-pos (78.7 months) and the CNS-neg group (67.2 months), p=0.58. Patients in the CNS-pos group were younger (median age 31.3 vs. 39.7 years, p=0.004), received the procedure more recently (median year 2012 vs. 2010, p=0.003), were less likely to have a Karnofsky score of equal or higher than 90 (70.8% vs. 81.9%, p=0.017), or to have received peripheral blood stem cells (PBSC) (61.5% vs. 72.7%, p=0.028). The groups did not differ in regards to donor source (URD, 50% vs. 56.5%, p=0.24) or the intensity of the preparative regimen (MAC, 82.3% vs. 85.6%, p=0.41). In multivariate analysis, CNS-pos were associated with higher cumulative incidence of relapse (HR=1.58 (95%CI=1.06-2.35), P=0.025) and a trend for an inferior leukemia-free survival (LFS) (HR=1.38 (95%CI=0.99-1.92), p=0.057), but did not adversely impact overall survival (OS) (HR=1.28 (95%CI=0.89-1.85), p=0.18). A subgroup multivariate analysis limited to patients with CNS-pos showed that prescribing a TBI MAC regimen (vs. others) results in a lower cumulative incidence of relapse (HR=0.35 (95%CI=0.15-0.79), p=0.012) and better LFS (HR=0.43 (95%CI=0.22-0.83), p=0.01) and OS (HR=0.44 (95%CI=0.21-0.92), p=0.03). Use of PBSC (vs. BM) was also independently associated with better OS (HR=0.53 (95%CI=0.29-0.99), p=0.046).

Conclusion: Notwithstanding the inherent limitations of registry data, particularly ascertaining the absence of CNS involvement in the CNS-neg group, our results show CNS involvement as an independent risk factor for relapse following allo-HCT. Our data support, nonetheless, the choice of a TBI-based MAC regimen in this group of patients but stresses the need for close monitoring of relapse after allo-HCT.