Defining risk of central nervous system relapse in acute myeloid leukemia following allogeneic stem cell transplantation.

Abstract

e19009 Background: Relapse following allogeneic stem cell transplant (AlloSCT) for acute myeloid leukemia (AML) remains the commonest cause of death post-transplant. Central nervous system (CNS) relapse is a devastating complication. With an estimated incidence of 2-8%, defining those patients at high risk and the management following CNS relapse remains controversial. We sought to identify demographic, clinical, molecular, and laboratory risk factors that would predispose AML patients to CNS relapse. Methods: This was a single institution retrospective study at Loyola University Medical Center. Inclusion criteria were adult patients with AML undergoing first AlloSCT from January 2012 to June 2021. Diagnosis of CNS relapse following AlloSCT was defined as leukemic blasts detected in spinal fluid analysis in accordance with established criteria. Chi-square tests were used to test the association between the experiencing of relapse and the categorical variables recorded. Wilcoxon Rank Sum test were used to assess the relationship between relapse experience and each of the demographic variables. Due to the low count of patients with CNS relapse and in order to avoid a poor estimation of confidence limits of any effect size, we limited our analyses to descriptive statistics and univariable analyses. Results: 198 patients met inclusion criteria with 8 cases (4.0% total) identified with CNS relapse following AlloSCT. Our analysis initially characterized 44 variables including age, race, HCT-CI risk, and day 100 chimerism. Univariate analysis is shown in the table below. Of note, it revealed that 62.50% of patients with CNS relapse had a chimerism value ≤98% at day 100 post-transplant as compared to 22.58% of patients without CNS relapse (p-value 0.02). Additionally, 87.50% of the CNS relapse group versus 51.38% of the group without relapse had favorable to intermediate AML risk scores (p-value 0.07). Conclusions: We observed a 4.0% rate of CNS relapse, which is consistent with what has been repored in the literature. Our results suggest that day 100 chimerism and disease risk index may be predictive of CNS relapse. Since there is a 10.6% incidence (5/47) of eventual CNS relapse amongst those with ≤98% chimerism at day 100, these patients could be considered for a lumbar puncture and bone marrow biopsy to prevent full relapse. Further research in this topic would be valuable to provide additional AML guidance in managing this uncommon complication, such as which patients may benefit from prophylactic strategies. [Table: see text]