New scoring system for predicting long-term survival of AML patients over 70 years and suitable for intensive chemotherapy

Introduction

The median age at AML diagnosis is 70 years, making it a disease of advanced age. Patients older than 70 years have a poorer prognosis due to the accumulation of comorbidities, which increase the risk of treatment toxicity, and due to less favorable biological characteristics of the disease, which in turn increase the risk of therapy failure. Standard therapeutic options for AML include intensive chemotherapy and hypomethylating agents or low-dose cytarabine in combination with the Bcl2 inhibitor venetoclax.

Objective

The primary objective of this work was to create and validate a European scoring system for predicting long-term survival of AML patients over 70 years of age (ESS70+) who were selected for intensive chemotherapy in routine practice. The score was to be determined based on parameters available at diagnosis.

Methodology

Data from AML patients aged ≥ 70 years treated in the 1st line in routine practice with intensive chemotherapy (IC: mostly combinations of daunorubicin and cytarabine or idarubicin and cytarabine ± lomustine) from the French DATAML database (comprising 23 centers), the German SAL database (46 centers), and the Spanish PETHEMA database (88 centers) were used to create the scoring system.

The analyzed data included age, gender, diagnosis date, AML status (de novo or secondary), performance status (PS) according to ECOG, white blood cell count, proportion of blasts in peripheral blood and bone marrow, lactate dehydrogenase level, cytogenetic risk, mutational status NPM1, FLT3-ITD, CEBPA, IDH1, IDH2, TP53 at diagnosis, response to treatment, occurrence of allogeneic hematopoietic stem cell transplant after the first complete remission (CR), and relapse or death date. Evaluated parameters included CR achievement, early death (ED: within 30 and 60 days), relapse, relapse-free survival (RFS), and overall survival (OS).

After creating the score in a training set of patients, calibration, performance evaluation, and external validation were conducted using a validation set. The validity of ESS70+ was also verified by comparing its predictive performance with previously published scoring systems for older patients treated with IC.

Evaluated patient population

A total of 1199 European patients with AML diagnosed between 2007−2018 and treated with IC were included. Their median age was 74 years, 56% were men, 75% had de novo AML, PS ECOG ≤ 1 and intermediate cytogenetic risk. Mutations in NPM1 were found in 35.6% and FLT3-ITD mutations in 19.8% of patients. CR or CR without complete hematologic recovery was achieved in 56.1% of treated patients, 30-day mortality was 13.0%, 60-day mortality 20.6%, and median OS was 10.9 months, which was significantly similar between patients older and younger than 70 years. The median follow-up duration was 50.8 months.

Results

The training set comprised 636 patients. OS was variably associated with age, PS ECOG, secondary AML, leukocytosis, cytogenetic risk, and NPM1 (without FLT3-ITD) mutations. These parameters were used to develop a score for predicting long-term OS.

Three risk groups were identified:

• Low risk with a 5-year survival probability ≥ 12% included 51% of patients, with a median OS of 18 months.
• Intermediate risk with a 5-year survival probability of 3–12% included 41% of patients, with a median OS of 9 months.
• High risk with a 5-year survival probability < 3% included 8% of patients, with a median OS of 3 months.

The risk level also significantly correlated with CR achievement, ED risk, and RFS.

The results were similar in the external validation cohort (n = 563). Additionally, a lower rate of false positives was observed compared to previously published scoring systems.

Conclusion

ESS70+ created based on data from a large population of AML patients over 70 years offers a score easily applicable in routine clinical practice using basic clinical and molecular parameters available at diagnosis. It enables the prediction of long-term survival for these patients, in whom intensive chemotherapy is being considered.

(zza)

Source: Bérard E., Röllig C., Bertoli S. et al. A scoring system for AML patients aged 70 years or older, eligible for intensive chemotherapy: a study based on a large European data set using the DATAML, SAL, and PETHEMA registries. Blood Cancer J 2022; 12 (7): 107, doi: 10.1038/s41408-022-00700-x. 

Abbreviated product information can be found here.

2011-CZ-2400006