Pre-Treatment Risk Assessment for Elderly Patients with Acute Myeloid Leukemia

Abstract

The treatment of AML in older adults is limited by the high mortality related with induction chemotherapy; however, those who tolerate an intensive treatment will have better outcomes; therefore, selecting this group of patients through the use of functionality scales is a fundamental part of the initial therapeutic approach. Risk assessment scales have been designed and validated by other authors; in our country they have not been routinely used until now. Objective: To describe 8-week treatment related and 1- year mortality in AML patients, older than 60 years, after selecting treatment based on functionality risk scores (FRS), at two hospitals in Bogotá. Design: An observational study was performed, analyzing early mortality in two cohorts; a retrospective, including patients treated from 2010-2015 and a prospective one, from 2015 to 2018, in which the treatment was selected according FRS (SPPB, CCI and MD Anderson Predictive Score). Setting: Patients were treated in two university hospitals in Bogotá, Colombia. Patients: AML patients older than 60 years; acute promyelocytic leukemia patients were excluded. Interventions: FRS were assessed at diagnosis, high risk patients received supportive care, intermediate risk received 5-Azacitidine or low dose ARA-C, low risk patient wereconsidered eligible for standard induction chemotherapy (7+3). Main Outcomes Measures: We evaluated 8-week mortality as predicted by a combination of 3 different scales and compared it with a control retrospective cohort. Results: Sixty patients were included, median age 72 years (range: 62 - 84), 78% had intermediate cytogenetic risk and 20% high risk. 35% had a history of another hematological neoplasm. Only 38.3% received high intensity chemotherapy. Survival at 8 weeks was 70% without differences between treatment groups. One-year mortality was high, 73.9% of patients treated with 7x3 died, 80% in the low intensity group and 85.7% in the best support treatment. The ICC scale was predictive of 1-year mortality, but not the MD Anderson scale. Conclusions: In this high-risk group, 7+3 was well tolerated when patients were selected using FRS. The CCI scale was predictive of one-year mortality and could be used to optimize the selection of elderly patients with AML.