Age and Population-Based Early Mortality Rate (EMR) in Acute Myeloid Leukemia (AML) Versus the EMR Reported By SWOG

Background: In their seminal paper SWOG described the expected EMR for patients by age enrolled on clinical trials in AML (Appelbaum et al Blood 2006; 107: 3481-5). We sought to contrast these to population based outcomes.

Methods: Using case listing session of SEER 18 (1973-2010) we identified patients (pt) 18 years of age and older diagnosed (dx) from 1990-2005 with AML (Site recode ICD-O-3/WHO 2008 Acute Myeloid Leukemia which includes ICD-0-3 9861/3, 9873/3, 9920/3, 9910/3, 9840/3, 9891/3, 9867/3, and 9895/3 and includes all FAB and WHO subtypes except M3 ). This was done in order to match the accrual periods of the SWOG studies. We then examined the EMR of pt with non-M3 AML and compared them to those reported by SWOG. EMR in SEER was defined as death within 0 to 1 month from dx; EMR in the SWOG studies was defined as death within 30 days of initiation of induction chemotherapy. Analyses were conducted with SEER*Stat 8.1.2, Microsoft Excel 2007 and GraphPad Prism 6. All p-values were 2-sided.

Results: 26,272 pt were identified within SEER and 955 pt were in the SWOG EMR cohort. 54% of patients were male in SEER and 55% were male in SWOG; 85% of patients were white in SEER and 89% were white in SWOG. The EMR was 38.6% (10,130) in the SEER cohort; EMR was 12.2% (116) in the SWOG cohort (chi-squared with Yates correction p<0.001). Age distribution and mortality by age are presented in the table below. In general the SEER patients were older and experienced a two-fold greater EMR than SWOG patients in each age subgroup.

Conclusion: EMR in pt with non-M3 AML is much higher across all studied age groups in population based data than reported in SWOG studies.