Age-Based Probability of Ten-Year Overall Survival (OS) of Acute Myeloid Leukemia (AML): A National Cancer Database (NCDB) Analysis

Background: Clinical trial data suggest excellent short-term outcomes for younger patients and a subset of fit older patients. Translating findings of clinical trials into the real-world practice has challenges, and longer-term data are often not available. Hence, we utilized NCDB to determine age-based probability of ten-year OS of unselected cohorts of adults with AML.

Methods: NCDB captures about 70% of all new diagnosis of cancer. We utilized NCDB to analyze ten-year OS of 15,646 patients aged ≥18 years, who were diagnosed with AML during the years 2004-2007. Univariate and multiple logistic regression models were used to determine factors associated with ten-year OS. Kaplan Meier curves were generated for OS analysis.

Results: Ten-year OS was 12.7% for patients aged 18-59 years treated with chemotherapy without hematopoietic cell transplant (HCT) (Table 1, Fig. 1). Older age, male, Charlson-Deyo comorbidity score >0, insurance other than private, subtypes other than core-binding factor AML were associated with lower ten-year OS. Ten-year OS was 24.3% for patients aged 18-59 years treated with chemotherapy and HCT (Fig 2). Older age, male, and insurance other than private were associated with lower ten-year OS. Ten-year OS was 16.1% and 2.2% for patients older than 60 years and treated with and without HCT, respectively; multivariate analysis was not performed for older patients because of low sample size (HCT group) or low survival (non-HCT group). Detailed analysis will be presented.

Conclusions: Long-term OS of adults with AML is low with less than a quarter of patients being alive at ten years. Ten-year OS is particularly poor for older patients who are treated with chemotherapy alone. Whereas recent advances and approval of eight new drugs will likely improve short-term OS at 2-3 years, innovative strategies are necessary to improve long-term OS and cure. Further study to identify the cause of death will be insightful.