Multicenter Retrospective Evaluation of Treatment of High-Risk Acute Promyelocytic Leukemia (APL): Results from the Hematology Research Outcomes (HERO) Consortium

Background:

Regimens employing all-trans retinoid acid and arsenic (ATRA+ATO) are the backbone of high-risk APL treatment. The APML4 regimen consists of a three-phase treatment approach: induction with ATRA, ATO, and idarubicin; two cycles of consolidation with ATRA+ATO; and maintenance with ATRA, methotrexate, and mercaptopurine.¹ MD Anderson Cancer Center (MDACC) developed a regimen incorporating gemtuzumab ozogamicin (GO) in addition to ATRA+ATO.² Differences in tolerability and efficacy between these two approaches are unknown as no comparative data exists.

We aimed to describe real-world treatment outcomes for high-risk APL, comparing toxicity between APML4 and ATRA+ATO+GO.

Methods

We conducted a multicenter, retrospective cohort study including patients treated among six US centers participating in the pharmacist-led collaborative research consortium, HERO. Adults (>18 years) with newly diagnosed high-risk APL (t(15;17)(q24.1;q21.2)) according to the Sanz risk model³ between April 2014 and June 2022, who received at least one dose of an induction regimen under the institution’s high-risk protocol were compared based on induction regimen received. Patients with incomplete records were excluded. The primary endpoint was a composite of mortality, ICU admission, or confirmed infection during induction. Secondary endpoints included individual components of the primary endpoint, other induction-related toxicities, and complete response (CR) rate after induction. Analyses were performed as intent-to-treat which included all patients who received APML4 and ATRA+ATO+GO inductions regardless of consolidation strategies. Chi-square or Fisher's exact tests were utilized to evaluate dichotomous variables. Continuous variables were analyzed via Student's t-test or Mann-Whitney U test. Kaplan-Meier analysis with log-rank test was performed to estimate progression-free survival (PFS) and overall survival (OS).

Results

Fifty-seven patients with high-risk APL who started either APML4 (n=41) or ATRA+ATO+GO (n=16) as induction treatment met inclusion criteria. Baseline characteristics were balanced (Table 1) except for race and ethnicity, with a higher percentage of non-Hispanic white patients in the APML4 cohort (p=0.020) and a higher percentage of Hispanic/LatinX in the ATRA+ATO+GO cohort (p=0.011). Of evaluated molecular abnormalities, FLT3-ITD was most common (76%). Steroid prophylaxis was more common in the ATRA+ATO+GO cohort (p=0.015). Additional cytoreductive therapy was frequent in both arms: hydroxyurea used more often in the APML4 cohort (p=0.033) and idarubicin in the ATRA+ATO+GO cohort (p<0.001) (Table 1). Induction mortality was 7%, including 3 deaths in the APML4 cohort and 1 in the ATRA+ATO+GO (Table 2; p=1.000). Time to platelet recovery was longer with APML4 than ATRA+ATO+GO (median 27 vs 23 days, p=0.021) as was time to ANC recovery (median 38 vs. 29 days, p=0.020). No additional significant differences in individual safety outcomes were observed. Of note, there was no difference in differentiation syndrome that required dose changes or holds between the two cohorts. There were no detectable differences in treatment effectiveness, with 92.7% of patients in the APML4 arm and 93.8% of patients in the ATRA+ATO/GO arm achieving a CR at end of induction (Table 2). With a median follow-up of 35 months, neither group reached the median for overall survival and the estimated 36-month survival was similar between cohorts; 92.5% (95%CI 84.3-100) for APML4 and 94.1% (95%CI 83.1-100) for ATRA+ATO+GO.

Conclusion

ATRA+ATO-based regimens yield similar short- and long-term efficacy in a high-risk APL population, though hematologic recovery occurred earlier in the ATRA+ATO+GO cohort. No clinically significant differences in acute toxicities exist between APML4 and ATRA+ATO+GO induction strategies as demonstrated by this real-world multicenter analysis.