Clinical value of event-free survival in acute myeloid leukemia

The value of event-free survival (EFS) as an end point in acute myeloid leukemia (AML) trials has been questioned. We hypothesized that rather than a surrogate for overall survival (OS), improvement in EFS may decrease the use of health care. In this retrospective study, we identified 400 patients with AML who were treated on first-line therapy trials and had OS between 2 and 36 months. We captured health care use from diagnosis until death or until the patient was censored at stem cell transplantation (SCT). We used correlation and regression analysis to determine the relation between health care use and EFS. Among patients with newly diagnosed AML, 35% had adverse-risk AML, 48% received intensive chemotherapy, and 28% received hypomethylating agents. The median EFS censored at SCT was 9.7 months. Longer EFS led to a significant decline in health care use regardless of OS. This held true for all observations, including overall health care use (r = -0.45), sum of clinic visits, emergency room visits, hospitalizations, consultations (r = -0.44), sum of invasive procedures, laboratory and imaging studies (r = -0.51), and blood product transfusions (r = -0.19). These correlations were stronger for patients who achieved a complete remission and held true across age, treatment, and disease risk subgroups. In patients with newly diagnosed AML, improvement in EFS correlates with a decrease in all health care use irrespective of OS duration.