Impact of sociodemographic factors on early mortality in acute promyelocytic leukemia in the United States: A time-trend analysis.

ABSTRACT

BACKGROUND: The survival of patients with acute promyelocytic leukemia (APL) has dramatically improved with the use of all-trans retinoic acid (ATRA) and arsenic trioxide (ATO). However, because of the complexity of the initial management, early mortality (EM) remains a major contributor to treatment failure. It is less known whether advances in treatment, urgent access to specialized care, and broad availability of ATRA/ATO have reduced EM in the last 2 decades. Furthermore, the influence of sociodemographic factors on the risk of EM also remains unclear. METHODS: This study used the Surveillance, Epidemiology, and End Results program to characterize the impact of sociodemographic factors on the rates of EM and overall survival (OS) in patients with APL diagnosed between 1992 and 2015. RESULTS: In all, 2224 cases were identified (895 who were younger than 40 years and 1329 who were 40 years old or older); 47.9% had a county-level median household income of $59,630 or higher, 49.0% belonged to counties where more than 31% of adults held at least a bachelor's degree, and 86.0% resided in urban areas. The rate of EM declined from 31.5% in 1992-1995 to 15.9% in 2012-2015 for all patients. It improved for patients younger than 40 years (27.4% in 1992-1995 vs 5.4% in 2012-2015; P < .001) and for patients 40 years old or older but not to the same extent (35.2% in 1992-1995 vs 22.2% in 2012-2015; P = .02). Importantly, improvements in EM were not seen among patients residing in rural areas, with the rate remaining higher than 20% in 2012-2015. The 3-year OS rate was 49.2% for patients with APL diagnosed in 1992-1995 and 76.4% for patients diagnosed in 2012-2015. CONCLUSIONS: These findings confirm consistent improvements in EM and OS for patients with APL and point to the challenge of further extending these improvements in EM rates to older patients and to those living in rural areas.