Statistical Methods

Multiple imputations were performed because exclusion of patients with at least one missing variable may cause bias.²⁵ Logistic regression was used for propensity score calculation from baseline patient characteristics including age, performance status, white blood cell count, percentage of blasts in peripheral blood and bone marrow, cytogenetics, percentage of CD20 and CD22 positive blasts, and the presence of central nervous system (CNS) disease. Propensity score analysis with 1:1 matching was performed with the nearest neighbor matching method using calipers of width equal to 0.2 of the standard deviation of the logit of the propensity score to balance baseline differences between cohorts.²⁶ Using prematched cohorts, univariate and multivariate analyses Cox regression analysis was performed to identify prognostic factors for overall survival (OS).²⁷

Time from therapy to stem cell transplant (SCT) was handled as a time dependent variable. Event-free survival (EFS) was calculated from the time of treatment initiation until the date of no response (after 2 cycles), relapse, or death. OS was calculated from the time of treatment initiation until death. Survival curves were plotted by the Kaplan-Meier method and compared with the log-rank test. Differences in subgroups by different covariates were evaluated with the χ2 test and Fisher exact test for nominal values and the Mann-Whitney U and Fisher exact tests for continuous variables. All the statistical data analyses were performed with SPSS version 24.0 software (SPSS, Chicago, IL) and R version 3.2.4. The cutoff date for follow-up was 7/2/2018.