A retrospective survival outcomes analysis of acute erythroid leukemia patients treated with upfront 7+3 induction chemotherapy versus hypomethylating agents with venetoclax utilizing real-world data

Abstract Background: Acute erythroid leukemia (AEL) is considered a rare and aggressive subtype of acute myeloid leukemia (AML). There is some controversy surrounding the exact nature of AEL as a distinct clinicopathologic entity. Whilst the ICC classification system includes AEL under the broader category of AML with TP53 mutations, the WHO 2022 classification uses the term to describe cases of pure erythroid leukemia which requires ≥80% erythroid predominance in the bone marrow with ≥30% proerythroblasts. Owing to its relative rarity, accounting for only 3-5% of all AML cases, available information regarding optimal treatment and prognosis is limited. Although survival outcomes for TP53 mutant AML are still extremely poor, cumulative evidence has suggested a better efficacy for use of a hypomethylating agent with venetoclax in TP53 mutant AML. We therefore conducted a retrospective cohort study to determine survival outcomes of AEL compared to AML and further studied survival outcomes in AEL patients treated with traditional 7+3 induction regimen compared to use of a hypomethylating agent with venetoclax. Methods: We utilized databases from the TriNetXtmglobal federated health research network which included 107 different healthcare organizations. By utilizing ICD-10 and ICD oncology codes, patients with acute myeloid leukemia and acute erythroid leukemia were identified as separate cohorts. Separate cohorts were identified for acute erythroid leukemia cases based on time to initiation of either idarubicin or daunorubicin with cytarabine, or azacitidine or decitabine with venetoclax within 4 weeks of the index diagnosis event. Propensity score matching (PSM) was conducted, matching cohorts based on age, gender, race and prior exposure to antineoplastic agents. Primary outcome measure of interest was all-cause mortality, and Kaplan-Meier survival analysis was conducted excluding patients with primary outcome prior to index diagnosis event. Relapsed acute erythroid leukemia was a secondary outcome measure, and Pearson's chi-squared test was employed to analyze measures of association. Results: A total of 93,667 AML patients and 1,581 AEL patients were identified. 1,581 patients in each cohort were 1:1 PSM matched. Baseline characteristics including age, gender, race, and exposure to antineoplastic agents were well balanced in both cohorts. Kaplan-Meier analysis on estimates of overall survival probability at 1-year and 5-years was 72.4% and 54.5% in the AML cohort vs 69.7% and 54.6% in the AEL cohort respectively (log-rank test p=0.093 for 1-year analysis and p=0.73 for 5-year analysis). A total of 126 patients with AEL who received induction 7+3 regimen and 92 patients with AEL who received HMA + venetoclax were identified. 48 patients in each cohort were 1:1 PSM matched. Baseline characteristics including age, gender, race, and exposure to antineoplastic agents were well balanced in both cohorts. Kaplan-Meier analysis on estimates of overall survival at 1-year predicted a median overall survival of 343 days in the AEL who received 7+3 cohort versus 140 days in the AEL who received HMA + venetoclax cohort, and a survival probability of 49.5% versus 22.8% respectively at 1-year (log-rank p-value 0.012) and 28.6% versus 0% at 5-years (log-rank p-value=0.0008). An equal number of patients relapsed in both treatment cohorts (n=10), and no risk difference was subsequently noted. Conclusion: AEL represents a rare and aggressive subtype of AML with poorly defined outcomes. These patients may have better survival outcomes with upfront 7+3 chemotherapy induction regimens compared to upfront use of hypomethylating agents with venetoclax, and argues in favor of considering this rare subtype as a distinct clinicopathologic entity. Further prospective studies are warranted to explore differences in treatment outcomes.