Mixed Phenotype Acute Leukemias: Real-World Outcomes By Who Classifications 2010-2021

Introduction

Mixed phenotype acute leukemias (MPAL) are an exceptionally rare and aggressive form of leukemia characterized by the absence of markers specific to a single hematopoietic lineage. In 2008 the WHO 4^th edition recognized 5 distinct subclasses of MPAL, defined by recurrent genetic abnormalities (MPAL with BCR::ABL and MPAL with KMT2A rearrangements [rKMT2A]), or immunophenotype (MPAL B/myeloid, MPAL T/Myeloid, and acute undifferentiated leukemia [AUL]). Outcomes and optimal management of MPAL, particularly for distinct subtypes, remain poorly understood. Treatment of MPAL typically involves ALL-like regimens, with the addition of targeted agents based on the immunophenotype and genotype of the disease. In this study, we present the largest cohort of MPAL identified to date, examining real-world outcomes and identifying factors associated with prognosis.

Methods

Adult patients (pts) ≥18 years (yrs) with MPAL were identified in the National Cancer Database (NCDB) by WHO 4^th edition ICD-O-3 histology code, for all available yrs (diagnosed 2010-2021). Kaplan-Meier and Cox regression survival analyses compared overall survival (OS) by MPAL subtype, by age group, and by year of diagnosis. The effect of race, insurance status, and income of residential zip code on OS was evaluated by cox regression.

Results

Of 4,756 pts with MPAL, 373 (8.2%) had MPAL with BCR::ABL1, 42 (0.9%) had MPAL with rKMT2A, 497 (10.9%) had MPAL with B/myeloid, 460 had MPAL with T/myeloid phenotype. The majority, 3,195 (70%) were registered as AUL. The median age of diagnosis was 68 (IQR 53-80), with only 14.4% presenting from 18-39 yrs, 28.5% from 40-64 yrs, 19.1% from 65-74 yrs, and 38% diagnosed at ≥75 yrs. The majority were White (83.6%), non-Hispanic (90.3%), and male (55.6%). With a median follow up of 3.4 months, median OS was only 4.2 months (95% CI 3.6-4.9) for all MPAL. With a 30-day mortality rate of 33%, 1-, 3-, and 5-yr OS were 37%, 24%, and 20%, respectively. MPAL with BCR::ABL1 had the longest OS, with median OS of 53.6 months (95% CI 28.7-78.5) and 1-,3-,5-yr OS of 76%, 57%, and 49%, respectively. AUL had the shortest OS, with median OS of 1.4 months (95% CI 1.2-1.6), and 1-, 3-, 5-yr OS of 28%, 17%, and 14%, respectively. The remaining subtypes had similar survival, with MPAL with rKMT2A median OS 15.4 months (95% CI 0.0-35.7), MPAL B/myeloid median OS 12.1 months (95% 9.8-14.4), and MPAL T/myeloid median OS 17.7 months (95% CI 1.2-1.6). Survival was strongly associated with age of diagnosis, with pts aged 18-39 yrs surviving significantly longer (1-, 5- yr OS 78%, 55%) than 40-64 yrs (1-, 5-yr OS 58%, 33%; hazard ratio [HR] 1.95 [95% CI 1.69-2.25]), and 65-74 (1-, 5-yr OS 31%, 12%; HR 3.83 [95% CI 3.31-4.44]), and ≥75 yrs (1-, 5-yr OS 10%, 2%; HR 7.95 [95% CI 6.92-9.13]). Later yr of diagnosis was associated with improved OS (HR 0.97 [95% CI 0.96-0.98]). On cox regression model including yr of diagnosis, subtype, age, and comorbidity score, being uninsured (HR 1.68 [95% CI 1.35-2.08]), Medicaid insured (HR 1.21 [95% CI 1.02-1.43], reference private insurance), or residing in low-income zip-code (HR 1.24 [95% CI 1.10-1.39], reference highest quartile income zip code) was associated with worse OS. Minority race was not associated with survival disparities, and Hispanic ethnicity was protective (HR 0.76 [95% CI 0.65-0.89]). On subanalysis, treatment at academic facilities was strongly associated with improved OS, HR 0.54 (95% CI 0.50-0.58).

Discussion

In the largest study to date of MPAL, we report real-world outcomes in the United States, identifying the prognosis for distinct subtypes. While survival remains dismal for MPAL, except for pts with targetable BCR::ABL1, there is a trend towards improved OS in recent years, which may be due to improved adoption of ALL-like regimens and the availability of targeted therapies for select pts. However, survival remains remarkably poor in elderly patients, emphasizing the urgent need for research into optimal therapies for pts who cannot tolerate intensive chemotherapy and allogeneic transplant. Efforts to expand access to specialized medical care, which is strongly associated with improved OS, are crucial to improve outcomes for all pts, especially pts from traditionally underserved populations. Our data supports the crucial need for improved understanding of the biology of MPAL, and innovative treatment options and clinical trials to improve survival for MPAL.