Long-term survival after intensive chemotherapy or hypomethylating agents in AML patients aged 70 years and older: a large patient data set study from European registries

Abstract

The outcome of acute myeloid leukemia patients aged 70 years or older is poor. Defining the best treatment option remains controversial especially when choosing between intensive chemotherapy and hypomethylating agents. We set up a multicentric European database collecting data of 3 700 newly diagnosed acute myeloid leukemia patients ≥70 years. The primary objective was to compare overall survival in patients selected for intensive chemotherapy (n = 1199) or hypomethylating agents (n = 1073). With a median follow-up of 49.5 months, the median overall survival was 10.9 (95% CI: 9.7-11.6) and 9.2 months (95% CI: 8.3-10.2) with chemotherapy and hypomethylating agents, respectively. Complete remission or complete remission with incomplete hematologic recovery was 56.1% and 19.7% with chemotherapy and hypomethylating agents, respectively (P < 0.0001). Treatment effect on overall survival was time-dependent. The Royston and Parmar model showed that patients treated with hypomethylating agents had a significantly lower risk of death before 1.5 months of follow-up; no significant difference between 1.5 and 4.0 months, whereas patients treated with intensive chemotherapy had a significantly better overall survival from four months after start of therapy. This study shows that intensive chemotherapy remains a valuable option associated with a better long-term survival in older AML patients.

Fig. 1. Study flowchart.

APL acute promyelocytic leukemia. LDA low-dose cytarabine.

Fig. 2. Survival according to intensive chemotherapy or HMA treatment.

A Kaplan–Meier curve of overall survival according to treatment (median OS: 10.9 months, 95% CI: 9.7–11.6 with IC and 9.2 months, 95% CI: 8.3–10.2 with HMAs). B Kaplan–Meier curve of relapse free survival according to treatment (median RFS: 11.5 months, 95% CI: 10.5–12.7 with IC and 11.0 months, 95% CI: 9.7–12.9 with HMAs). C Royston and Parmar adjusted* hazard ratio for overall survival in HMA vs. IC for each month from diagnosis. Before 1.5 months of follow-up, patients treated with HMAs had a significantly lower risk of death compared to IC patients. Between 1.5 months and 4.0 months of follow-up, there was no significant difference in survival between HMAs and IC patients. From 4.0 months of follow-up, patients treated with HMAs had a significantly higher risk of death compared to IC patients. Interaction between treatment (HMAs vs. IC) and age (< vs ≥75 y), performance status (≤ vs >1), cytogenetic risk (favorable vs. intermediate vs. adverse) or NPM1 mutation (yes vs. no) was not significant, showing that the effect of HMAs vs. IC was not significantly different according to age, performance status, cytogenetic risk and NPM1 mutation. Thus, there is no indication to stratify the analysis on age, performance status, cytogenetic risk and NPM1 mutation (Figure C was the same according to age (< vs. ≥75 y), performance status (≤ vs. >1), cytogenetic risk (favorable vs. intermediate vs. adverse) or NPM1 mutation (yes vs. no)). *Adjusted for age ≥75 y, performance status > 1, white blood cell count at diagnosis >30 giga per liter, cytogenetic risk, secondary vs de novo AML and NPM1 mutation. D Royston and Parmar adjusted* hazard ratio for relapse-free survival in HMAs vs. IC for each month from CR/CRi. Before 3 months of follow-up, patients treated with HMAs had a significantly lower risk of relapse or death compared to IC patients. Between 3 months and 8.5 months from CR/CRi, there was no significant difference between HMAs and IC patients. Beyond 8.5 months from CR/CRi, patients treated with HMA had a significantly higher risk of relapse or death compared to IC patients. Interaction between treatment (HMAs vs. IC) and age (< vs. ≥75 y), performance status (≤ vs. >1), cytogenetic risk (favorable/intermediate vs. adverse) or NPM1 mutation (yes vs. no) was not significant, showing that the effect of HMAs vs. IC was not significantly different according to age, performance status, cytogenetic risk and NPM1 mutation. Thus, there is no indication to stratify the analysis on age, performance status, cytogenetic risk and NPM1 mutation (Figure D was the same according to age (< vs. ≥75 y), performance status (≤ vs. > 1), cytogenetic risk (favorable/intermediate vs. adverse) or NPM1 mutation (yes vs. no)). *Adjusted for performance status >1, white blood cell count at diagnosis >30 giga per liter, cytogenetic risk, secondary vs. de novo AML, NPM1 and FLT3-ITD mutations.

Fig. 3. Survival according to intensive chemotherapy or HMA treatment in the pairwise population matched by the propensity score.

A Kaplan–Meier curve of overall survival according to treatment in 532 IC patients matched with 532 HMA patients (median OS: 10.5 months, 95% CI: 8.8–12.2, with IC and 9.6 months, 95% CI: 8.5–11.0, with HMAs). B Kaplan–Meier curve of relapse free survival according to treatment (median RFS: 11.9 months, 95% CI: 10.3–14.5, with IC and 10.0 months, 95% CI: 8.4–12.9, with HMAs). C Royston and Parmar hazard ratio for overall survival in HMAs vs. IC for each month from diagnosis. Before one month of follow-up, patients treated with HMAs had a significantly lower risk of death compared to IC patients. Between 1 month and 3.0 months of follow-up, there was no significant difference in survival between HMA and IC patients. From 3.0 months of follow-up, patients treated with HMAs had a significantly higher risk of death compared to IC patients. D Royston and Parmar hazard ratio for relapse-free survival in HMAs vs. IC for each month from CR/CRi. Before 7 months of follow-up from CR/CRi, there was no significant difference in relapse or death between HMA and IC patients. From seven months of follow-up after CR/CRi, patients treated with HMAs had a significantly higher risk of relapse or death compared to IC patients.