Clinical, cytogenetic and molecular predictors of early mortality in acute promyelocytic leukemia (APL): Real-world analysis

Abstract

Abstract Introduction The advent of all-trans retinoic acid (ATRA) and arsenic trioxide revolutionized APL treatment, yet early mortality (EM) rates among patients (pts) treated in real-world practice remain as high as 10-15%. Factors driving EM are not well understood, and the prognostic impact of genetic abnormalities beyond t(15;17) remains largely unknown Methods In this retrospective analysis, we identified all pts with APL diagnosed between 2013 and 2024 at Yale New Haven Hospital. EM was defined as death within 30 days of diagnosis. Continuous variables were compared using the Mann-Whitney U Test and categorical variables by using Fisher's exact test. Survival probabilities were estimated using the Kaplan–Meier method, and the impact of prognostic factors on survival was evaluated through multivariate analysis with Cox proportional hazards model Results We identified 64 APL pts (53% male) with a median age of 52 years (Range [R]: 19-89; 31% ≥60 years). Eleven (17%) had high-risk disease per Sanz criteria and 32 (50%) had coagulopathy at presentation. Aside from t(15;17), karyotypic analysis revealed additional cytogenetic abnormalities (ACA) in 15 (24%) of 64 pts. The most common ACA was isochromosome of the long arm of the derivative chromosome 17 (ider17) detected in 4 (6%) pts. Trisomy 8 and trisomy 21 were detected in 2 (3%) pts each, whereas complex t(3;15;17) was detected in 1pt. FLT3 mutations were detected in 23 (49%) of 47 evaluable pts. Among these, 12 had FLT3-ITD, 7 had FLT3-D835V, and 4 had both. Targeted exome sequencing using a myeloid malignancy panel was conducted in 32 pts, identifying mutations in 10 (31%). Most common mutations involved RAS in 3 pts, followed by CALR and WT1 in 2 pts each. Mutations were also detected in the ARID1A, ETV6, EZH2, PTPN11 and SF3B1 genes in addition to a FLT3 N676K mutation Fifty-six pts completed induction therapy. Among them, 53 (95%) received ATRA + arsenic-based regimens (16[29%] received additional GO, 3[5%] received additional idarubicin) and 3 (5%) received ATRA with idarubicin. Post-induction bone marrow biopsies were performed in 53 pts, all of whom achieved complete remission. Molecular complete remission (mCR) was achieved in 52 of 53 evaluable pts (98%) after a median of 2 treatment cycles (R 1–5), with 20 pts achieving mCR after induction EM occurred in 8 (12.5%) pts. Of these, 2 died before starting treatment, 5 within 48 hours of ATRA initiation, and 1 on day 21 of induction. The primary cause of EM was intracranial hemorrhage (4 pts). Additional causes included tumor lysis syndrome with multiorgan failure (2 pts), respiratory failure possibly related to differentiation syndrome (1pt), and ischemic stroke (1pt). At a median follow-up of 3.3 years, the 3-, 5- and 10-year overall survival (OS) for the cohort were 86%, 86% and 67%, respectively. The 3-, 5- and 10-year OS rates among pts who completed induction were 98%, 98% and 79%, respectively Pts who experienced EM were older (median age 71 vs. 50 years, P=0.03), had lower platelet counts (median 8.5 vs. 27.5 ×10⁹/L, P=0.003), higher percentages of circulating blasts (median 64% vs. 3%, P=0.003), and higher WBC counts (median 6.0 vs. 1.5 ×10⁹/L, P=0.06) at presentation, compared to those who completed induction. Sex, baseline hemoglobin, fibrinogen, and D-dimer levels were similar between the two groups. FLT3 mutations were significantly more common among pts with EM compared to those who completed induction (any FLT3 mutation: 86% vs. 42.5%, P=0.04; FLT3-ITD: 71% vs. 28%, P=0.03). Rates of ACA (12.5% vs. 6%, P=0.7) and other genetic mutations (50% vs. 30%, P=0.5) did not differ significantly between the two groups In a multivariate analysis including age, platelet count, WBC, peripheral blast percentage, D-dimer, and FLT3 mutational status, age emerged as the only independent predictor of inferior OS (HR=1.11; P=0.04). While FLT3 mutations were associated with HR of 30.8 for poorer OS, the limited number of events resulted in a wide confidence interval, rendering this finding statistically non-significant (P=0.07) Conclusion EM remains a major challenge of APL treatment, especially among older pts; however, pts who survive through induction therapy experience excellent long-term outcomes. Our findings highlight a possible prognostic role of genetic abnormalities beyond t(15;17). Larger studies may further clarify the significance of these mutations and their contribution to APL risk startification