A machine learning (ML)-based approach may help clinicians to determine the optimal treatment regimen among newly diagnosed patients with acute myeloid leukemia (AML), according to research published in JCO Clinical Cancer Informatics.
The standard of care regimen for patients diagnosed with AML has been induction therapy with an anthracycline plus cytosine arabinoside (7+3), sometimes followed by consolidation therapy with high-dose cytosine arabinoside or allogeneic stem cell transplant (alloSCT). The age of many patients, as well as the frequent presence of comorbidities, has rendered many ineligible for this aggressive therapy.
In the past several years, many patients ineligible for aggressive therapy have received venetoclax plus hypomethylating agents (ven/HMAs). This group of therapies has proved optimal particularly among patients older than 75 years or in whom there are many comorbidities.
Continue Reading
Choosing between 7+3 and alloSCT or ven/HMAs has, however, proved to be a difficult task in many patients. For this study, researchers evaluated whether an ML approach might help in choosing between 7+3 and alloSCT or ven/HMAs among newly diagnosed patients with AML.
Overall, data from 221 patients were included, 120 vs 101 of whom received 7+3 vs ven/HMAs, respectively. In the 7+3 vs ven/HMAs groups, the median age was 53 vs 73 years, respectively, 56% vs 48% of patients were female sex, and 98% vs 70% of patients had primary AML.
The authors used the ML strategy to determine predictors of short- and long-term outcomes among patients in both cohorts. The resulting algorithms showed promise in determining outcomes, length of hospital stays, and transfusion requirements among patients treated with either regimen.
“These initial tools set the stage for data collection and analysis in larger cohorts to further refine the toolset as well as to validate the ML approaches for newly diagnosed patients with AML,” the authors wrote.
Disclosure: Some study authors declared affiliations with biotech, pharmaceutical, or device companies. Please see the original reference for a full list of authors’ disclosures.
Reference
Islam N, Reuben JS, Dale J, et al. Machine learning-based exploratory clinical decision support for newly diagnosed patients with acute myeloid leukemia treated with 7 + 3 type chemotherapy or venetoclax/azacitidine. JCO Clin Cancer Inform. 2022;6:e2200030. doi:10.1200/CCI.22.00030
