Practice Patterns and Real-Life Outcomes for Patients with Acute Promyelocytic Leukemia

Background: Despite a potential for cure, a large proportion of patients with acute promyelocytic leukemia (APL) succumb early to hemorrhagic and thrombotic complications. Prompt initiation of all-trans retinoic acid (ATRA) has been associated with improved outcomes. However, little is known about the patterns of care and clinical outcomes of patients with APL in the United States (US) outside of the controlled trial setting.

Methods: We identified APL patients (pts) included in the Vizient Clinical Database/Resource Manager (CDB/RM™), which includes patient demographics, hospital characteristics, charge-level medication usage, procedures and mortality data from over 100 academic health centers and affiliated hospitals. Initial inpatient encounters of adult APL pts (≥18 years) during 2014-2015 (most recent years available) were identified by ICD-10 code (C92.40; C92.42) or ICD-9 codes for acute myeloid leukemia (AML; 250.00, 205.02) plus receipt of ATRA for ≥3 days as monotherapy, or in combination with arsenic trioxide (ATO) or an anthracycline. The 3-day minimum ATRA requirement is intended to exclude AML patients who may have therapy initiated while awaiting diagnostic confirmation. Pts were excluded if information on discharge disposition or medication administration were missing or if patients had an ICD-9/10 code consistent with relapsed AML/APL. Primary outcome was a composite of in-hospital death or discharge to hospice. Pearson’s Chi-square test was used to compare categorical variables in bivariate analysis. Multivariable logistic regression was used to examine associations between patient, hospital, and treatment characteristics with the primary outcome. We lacked information on laboratory test results needed to assess the impact of treatment on clinical outcomes.

Results: We identified 486 pts treated at 76 hospitals. Patient demographic, treatment and hospital characteristics are shown in Table 1. Median length of stay was 30 days (interquartile range [IQR]: 22-36 days). Pts received a median of 26 days of ATRA (IQR: 12-34 days). The majority of pts was treated with ATRA + ATO (240 pts; 49.4%) or ATRA + anthracycline (146 pts; 30.0%), which are the standard of care for lower- and higher-risk APL, respectively. Forty-two pts (9.3%) received ATRA monotherapy and 58 (11.9%) were treated with ATRA + ATO + anthracycline. Forty-eight pts (9.9%) required endotracheal intubation.

Overall, 54 (11.1%) pts experienced our composite adverse outcome of in-hospital death or transfer to hospice (45 pts [9.3%] died in the hospital; 9 pts [1.9%] were discharged to hospice). Among 45 pts who died in the hospital, median time to death was 17 days (IQR: 5-24 days) with 31.1% of deaths occurring during the first 7 days of hospitalization. Half of the pts (n=18; 50%) who died in the ATRA monotherapy group died within 5 days of admission. In bivariate analyses, age ≥66 years (22.9% vs. 7.9%; p<0.001), endotracheal intubation (64.6% vs. 5.3%; p<0.001), and ATRA monotherapy (42.9% vs. 8.1%; p<0.001) were associated with an increased risk of poor outcome (i.e. inpatient death or discharge to hospice). Similar results were obtained from the multivariable logistic regression model (Table 2).

Discussion: A substantial proportion (11%) of adults treated for APL in this large inpatient dataset died or were discharged to hospice. This likely underestimates the full mortality rate of APL, because our APL case definition excluded pts who received <3 days of ATRA therapy, which may have been due to early mortality. Age≥66 years, receipt of ATRA monotherapy or ATRA + anthracycline was associated with higher odds of adverse outcomes. The higher mortality with ATRA + anthracycline compared to ATRA + ATO is likely due to an inherently higher disease risk, but the database lacked laboratory results (e.g. white blood cell count) needed to confirm that hypothesis. Early complications related to coagulopathy or differentiation syndrome that precluded co-administration of ATO or anthracycline could be a potential explanation for the high early mortality rate in pts receiving ATRA monotherapy.