Page 165 - 2020_02-Haematologica-web

Comparing front-line therapies in older AML patients
that is more frequently associated with chemotherapeutic resistance, unfavorable cytogenetics, increased frequency of somatic mutations, and preceded by myelodysplastic syndromes, making therapeutic decisions difficult.4,5 Many patients have their treatment chosen more on the basis of chronological age rather than the inherent disease biology (i.e. karyotype, molecular heterogeneity, antecedent hematologic disorders, and leukocyte count at diagnosis) and overall fitness of patients.6
Like many other malignancies, optimizing medical care of patients with AML is dependent on clinical trials. Unfortunately, older patients, particularly those who are ≥70 years of age, are under-represented in randomized controlled trials. The lack of clear clinical data in this sub- set of patients often leads to uncertainty regarding opti- mal treatment strategies.
Over the past decade, new treatment strategies have emerged targeting the biological challenges in AML; how- ever, there has been a lack of significant progress in opti- mizing strategies in the older AML population. Numerous studies have assessed risk stratification of this subgroup of patients with a goal toward building a com- prehensive approach; however, a model to help guide treatment has yet to be validated.2,7-10 The lack of a vali- dated decision model has led to individualized and vari- able care of older AML patients. In an attempt to create a comprehensive decision analysis model, we present the results of a very large, single institution retrospective study of 980 patients aged ≥70 years. The aim of our study was to compare survival outcomes of older AML patients treated with various induction regimens. Such a study offers the advantage of combining cytogenetics, comorbidities, and functional status information and importantly accounts for therapeutic decisions in older patients with AML.11-14
Methods
Data collection
We retrospectively analyzed patients ≥70 years of age who presented to Moffitt Cancer Center between 1995 and 2016 for evaluation of newly diagnosed and previously untreated AML. The study was approved by the University of South Florida institutional review board. Inclusion criteria for the study were age 70 years or older and diagnosis of AML that was untreated prior to patient presentation at our institution. Patients with antecedent hematologic malignancies were included regardless of treatment. Compiled data were supplemented by direct review of medical records as necessary. A dual data entry tech- nique was used to ensure data accuracy and quality. Baseline patient characteristics collected included vital status, age at diag- nosis, sex, race/ethnicity, comorbidities for calculation of Charlson comorbidity index (CCI), Eastern Cooperative Oncology Group (ECOG) performance status, and antecedent hematologic disease and its treatment. Collected disease-specific characteristics included baseline cytogenetics, type of AML (de novo or secondary AML), complete blood count with peripheral blood blast percentage at time of diagnosis, choice of therapy, responses to treatment including complete remission (CR), com- plete remission with incomplete count recovery (CRi), relapsed disease, partial remission, and whether allogeneic hematopoietic stem cell transplant was performed. We defined secondary AML as an AML arising from an antecedent hematologic disorder or therapy-related AML.
Treatment groups
Patients were categorized into four different treatment groups: high-intensity therapy [defined as cytarabine and daunorubicin/idarubicin (7+3) or “7+3” equivalent], low-intensity therapy (defined as low-dose cytarabine or similar but not includ- ing hypomethylating agents), hypomethylating agent (HMA) ther- apy, and supportive care (including hydroxyurea). “7+3” equiva- lent regimens included high-dose cytarabine-based regimens, specifically CLAG+/-M (cladribine, cytarabine, granulocyte colony stimulating factor (G-CSF), with or without Mitoxantrone), MEC (Mitoxantrone, etoposide, cytarabine), and HIDAC (high-dose cytarabine) regimens. A categorical distinction between low-intensity therapy and HMA therapy was made on the basis of recent randomized reports suggesting the modest superiority of HMA versus conventional care regimens (including low-dose cytarabine), in addition to practice pattern differences worldwide that utilized either HMA or low-dose cytarabine as standard front-line therapy for older adults with newly diagnosed AML. Patients enrolled in clinical trials were assigned to one of the four treatments groups depending on the intensity of treatment received as part of the clinical trial.15,16
Definition of clinical end points
Response to therapy was defined as those who achieved CR or CRi as per the 2003 International Working Group response criteria for AML.17 Overall survival was defined as time from date of diag- nosis of AML to date of death if known or censored at the time of last follow up. Relapse-free survival was calculated as time from achievement of CR or CRi to date of relapse as defined by International Working Group 2003 criteria.
Statistical analysis
Survival function was estimated by the Kaplan-Meier method and compared across groups using the log-rank test. Cox propor- tional hazards regression model was used to determine the asso- ciation between the variables and overall survival. Variables with P<0.25 in the univariate model were included in the initial multi- variate analysis. The backward elimination method was used to select the variables for the ultimate multivariate model. Variables with P>0.05 were excluded. Pairwise comparisons of survival between different treatment groups were performed using the stratified log-rank test and propensity score matching to adjust for potential treatment indication bias between groups. Within pair- wise comparison groups, the stratified Cox proportional hazards regression model was used to assess correlations between clinical variables and overall survival. Patients who had no information on response were considered as non-responders per the intention-to- treat approach. For treatment-related mortality (TRM) at day 30, patients who were censored before 30 days (n=5) were not eligible and were excluded from the analysis. Raw P-values were comput- ed by the χ2 test, and the Bonferroni method was used to adjust for multiplicity. A two-sided P<0.05 was considered significant. Statistical analysis was performed using SAS version 14.3 (Cary, NC,USA).
Results
In the total cohort of 980 patients, 360 (36.7%) received high-intensity therapy, 255 (26.0%) received HMA thera- py, 91 (9.3%) received low-intensity therapy, and 274 (28.0%) received supportive care; their baseline character- istics are represented in Table 1. Median age of patients when first diagnosed with AML was 75.6 years (range, 70- 95.7 years). Among patients with antecedent hematologic
haematologica | 2020; 105(2)
399