Extramedullary Relapses after Allogeneic Stem Cell Transplant (allo-SCT) in Adult Patients with Acute Lymphoblastic Leukemia (ALL)

Background: Extramedullary (EM) relapses were sometimes observed in acute leukemia patients both after chemotherapy and allo-SCT. Our recent study described that the rate of EM relapses after allo-SCT was significantly higher when comparing with that after chemotherapy in acute myeloid leukemia (AML) patients. Since more potent graft-versus-leukemia (GVL) effect in EM lesion than bone marrow (BM) is proposed as potential biological basis of this phenomenon, it is expected that EM relapses after allo-SCT more frequently occurred than after chemotherapy also in ALL patients. However, this hypothesis has not been examined, and risk factors of EM relapses after allo-SCT have not been elucidated. So, we conducted this retrospective study to address this unsolved issue. Patients and methods: The study population included in this study was 215 adult patients who were diagnosed as ALL between 1990 and 2017 and received intensive chemotherapy. In the first part of this study, to compare the rates of EM relapses between after chemotherapy and allo-SCT, the initial relapses of the 88 patients were analyzed. In the second part, to investigate risk factors for EM relapses after allo-SCT, 110 patients who underwent allo-SCT against ALL were analyzed. EM relapses included both one only in EM lesions and in concurrent EM and BM lesions. Fisher's exact test was used to compare binary variables. Cumulative incidences (CIs) of EM relapse were compared using the stratified Gray test, considering relapse without EM lesions and death without the event as a competing risk. The logistic regression model and the Fine-Gray proportional hazard model were used for multivariate analysis of risk factors of EM relapses among the initial relapses and after allo-SCT, respectively. Values of p < 0.05 were considered significant. Results: Of the 88 relapsed patients included in the first part of this study, the median age at diagnosis was 47 years (range, 15-79 years), and the median duration of the first complete remission (CR1) was 7.1 months (range, 0.7-105.7 months). Philadelphia chromosome (Ph) and EM lesions at diagnosis were observed in 21 and 21 patients, respectively. Allo-SCT in CR1 was undergone in 12 patients. EM relapses occurred in 21 patients, and the sites of EM relapses were central nervous system (CNS) in 13, mediastinum in two, and bone in two. The median durations of CR1 were not significantly different between relapses with and without EM lesions (16.8 vs. 6.7 months, respectively; p = 0.295). In univariate analysis for risk factors of EM relapses, there was no significant difference in EM relapse rates between relapses after allo-SCT and chemotherapy (8.3% vs. 26.3%, respectively; p = 0.279), and in multivariate analysis, only EM lesion at diagnosis was identified as independent risk factor (odds ratio 4.21; p = 0.008). Of the 110 allo-SCT recipients included in the second part, the median age at diagnosis was 43 years (range, 16-66 years). Ph and EM lesions at diagnosis were observed in 43 and 21 patients, respectively. Disease status at the time of transplant was CR1 in 67, advanced CR in 17, and non-CR in 26. Stem cell sources were related, unrelated, and cord blood in 30, 50, and 25 patients, respectively, and almost all patients were conditioned with total body irradiation-containing myeloablative regimens. EM relapse after allo-SCT occurred in nine patients, and the 2-year CI of EM relapses was 6.5%. The sites of EM relapses after allo-SCT were CNS in three, lymph node in two, and skin in two. In univariate analysis for EM relapses after allo-SCT, the significantly higher CI of EM relapses after allo-SCT was observed in patients with EM lesion at diagnosis when comparing with those without EM lesion (28.6% vs. 1.1%, respectively; p = 0.279). Multivariate analysis extracted only EM lesion at diagnosis as an independent risk factor for EM relapses after allo-SCT (hazard ratio 24.09; p = 0.004). Conclusion: As a higher frequency of EM relapse after allo-SCT in ALL patients was not confirmed in this study, the hypothesis, more potent GVL effect in EM lesion than BM, was not able to apply to these patients. To determine whether this hypothesis is correct or not, further investigation in patients with other hematologic malignancy such as chronic myeloid leukemia is warranted. The vigilance is required regarding EM relapses in adult ALL patients with EM lesion at diagnosis both after chemotherapy and allo-SCT.