Predictors of Survival in Core Binding Factor Acute Myeloid Leukemia: An Analysis of the National Cancer Database

Background

Core binding factor acute myeloid leukemia (CBF AML) is characterized by t(8;21) or inv(16). Though characterized as favorable risk due to high remission rates after cytarabine-based chemotherapy, prior population-based studies have found significantly lower survival rates of CBF AML compared with those reported in clinical trials. Using the National Cancer Database (NCDB), we evaluate the impact of socioeconomic, demographic, and clinical factors on long term survival rates in CBF AML among patients diagnosed in the United States between 2005 and 2020.

Methods

Demographic, socioeconomic, facility-level, and clinical covariates were compared between inv(16) and t(8;21) using the χ² or Fisher’s exact test for categorical variables and t-test or Wilcoxon rank test for continuous variables. Univariable and multivariable Cox proportional hazards regression were used to determine factors associated with survival. All factors univariably associated with survival differences were adjusted for in multivariable analysis. Survival analyses by subtype of CBF AML, age, year of diagnosis, treatment comparison, and time to treatment were performed using the Kaplan-Meier method with corresponding log-rank test.

Results

A total of 2354 CBF AML patients were identified for analysis, 1227 with inv(16) and 1127 with t(8;21). The median age at diagnosis was 55 years. Median follow up from diagnosis to last contact was 26.6 months and median overall survival (OS) was 15.7 months for the entire cohort, 19.8 months for inv(16), and 11.2 months for t(8;21).

In univariable analysis, there was no significant difference in mortality risk based on sex (p=0.0599), race (p=0.0541), metropolitan vs. urban vs. rural location of treatment (p=0.4770), geographic location (p=0.1833), or distance to the treatment facility (p=0.0550). OS did not vary significantly by year of diagnosis when comparing years 2005-2011 with 2015-2017 (HR 0.99 [0.83-1.18], p=0.4264) or 2018-2020 (HR 1.04 [0.85-1.27], p=0.6910). There was a trend towards worsened survival of t(8;21) compared with inv(16) though this was not significant (HR 1.11 [0.99 – 1.24], p=0.0678).

In multivariable analysis, the likelihood of mortality increased significantly with increasing age (HR 1.29 [1.00-1.65] for 50-59 years, 1.84 [1.43 – 2.36] for ages 60-69 years, 2.68 [2.01-3.56] for ages 70-79 years, and 2.77 [2.01 – 3.81] for ages ≥80, p<0.0001, reference 40-49 years). Mortality risk for patients insured by Medicaid (HR 1.28 [0.99 – 1.66], p=0.0644) may be worse compared with private insurance, though this was not significant. Compared with an income of <$46277, an income of >$74063 did not confer a survival benefit (p=0.2591). Treatment at an academic facility was associated with improved survival (HR 0.83 [0.72 – 0.95] p=0.0059) compared with treatment at a non-academic facility. A Charlson comorbidity index of ≥1 (HR 1.17 [1.02 – 1.35], p=0.0275) was associated with worsened survival. Increased time to treatment was not associated with survival (p=0.8679). Compared to no treatment, both multiagent (HR 0.19 [0.15 – 0.23], p<0.0001) and single agent (HR 0.39 [0.31 – 0.49], p<0.0001) chemotherapy were associated with survival benefit. Compared with those who only received multiagent chemotherapy, those who also underwent bone marrow transplant (BMT) had no significant difference in survival (HR 0.94 [0.69 – 1.26], p=0.6683). However, when examining Kaplan Meier estimators, t(8;21) patients who underwent BMT had better OS compared with t(8;21) patients who only received multiagent chemotherapy (HR 0.61 [0.43-0.87], p=0.0058).

Conclusion

In this large NCDB study of CBF AML, age remained a significant predictor of mortality risk. The study suggests that alternative therapeutic strategies should be considered for unfit patients, as single agent chemotherapy provided a significant survival benefit over no treatment. Though there was no significant survival difference between inv(16) and t(8;21), patients with t(8;21) had notably worsened median survival and demonstrated different survival outcomes based on the consolidation strategy. Overall, these findings describe the unique predictors of survival in CBF AML and emphasize the importance of identifying novel therapies and standardization of current therapy approaches to improve outcomes, especially as survival has not significantly improved in recent years.