Prediction of complete remission and survival in acute myeloid leukemia using supervised machine learning

Abstract

Achievement of complete remission (CR) signifies a crucial milestone in the therapy of acute myeloid leukemia (AML) while refractory disease is associated with dismal outcomes. Hence, accurately identifying patients at risk is essential to individually tailor treatment concepts to disease biology. We used nine machine learning (ML) models to predict CR and 2-year overall survival (OS) in a large multi-center cohort of 1383 AML patients who received intensive induction therapy using clinical, laboratory, cytogenetic and molecular genetic data and validated our results on an external multicenter cohort. Our ML models autonomously selected predictive features both including established markers of favorable or adverse risk as well as identifying markers of so-far controversial relevance. De novo AML, extramedullary AML, double-mutated (dm) CEBPA, mutations of CEBPA-bZIP, NPM1, FLT3-ITD, ASXL1, RUNX1, SF3B1, IKZF1, TP53, U2AF1, t(8;21), inv(16)/t(16;16), del(5)/del(5q), del(17)/del(17p), normal or complex karyotypes, age and hemoglobin at initial diagnosis were statistically significant markers predictive of CR, while t(8;21), del(5)/del(5q), inv(16)/t(16;16), del(17)/del(17p), dmCEBPA, CEBPA-bZIP, NPM1, FLT3-ITD, DNMT3A, SF3B1, U2AF1, TP53, age, white blood cell count, peripheral blast count, serum LDH and Hb at initial diagnosis as well as extramedullary manifestations were predictive for 2-year OS. For prediction of CR and 2-year OS, AUROCs ranged between 0.77 – 0.86 and 0.63 and 0.74, respectively in our test set and 0.71 – 0.80 and 0.65 – 0.75 in the external validation cohort. We demonstrate the feasibility of ML for risk stratification in AML as a model disease for hematologic neoplasms using a scalable and reusable ML framework. Our study illustrates the clinical applicability of ML as a decision support system in hematology.

Prognostic impact of the AML ELN2022 risk classification in patients undergoing allogeneic stem cell transplantation

“…The clinical relevance of this model has been validated [32][33][34][35] and shown to be superior to the ELN2017 risk prediction in patients consolidated with chemotherapy [33], but not with allogeneic HSCT [35]. In addition to this approach, the increased use of machine learning and artificial intelligence approaches in outcome prediction will likely further impact AML risk assessment in the future [36].…”

Acute myeloid leukemia risk models: Where do we stand?

“…Thus, combining large clinical and genomic data sets in the form of knowledge banks can provide an accurate prediction of relapse, remission, and overall survival (OS) and guide clinicians to precisely tailor the therapy for the individual patient 2 . Recently, ML approaches were applied to 1383 AML patients who received intensive induction therapy demonstrating their high accuracy in predicting CR and OS both in internal testing and external validation 17 . Even in this circumstance, clearly emerged the value of combined clinical, laboratory, and genetic factors in improving prediction.…”

“…2 Recently, ML approaches were applied to 1383 AML patients who received intensive induction therapy demonstrating their high accuracy in predicting CR and OS both in internal testing and external validation. 17 Even in this circumstance, clearly emerged the value of combined clinical, laboratory, and genetic factors in improving prediction. In particular, the ML models identified the following factors as predictive for CR: de novo AML, extramedullary AML, double-mutated CEBPA, mutations of CEBPA-bZIP, NPM1, FLT3-ITD, ASXL1, RUNX1, SF3B1, IKZF1, TP53, U2AF1, t(8;21), inv (16)/t(16;16), del (5)/del(5q), del (17)/del(17p), normal or complex karyotypes, age and hemoglobin.…”

“…In particular, the ML models identified the following factors as predictive for CR: de novo AML, extramedullary AML, double-mutated CEBPA, mutations of CEBPA-bZIP, NPM1, FLT3-ITD, ASXL1, RUNX1, SF3B1, IKZF1, TP53, U2AF1, t(8;21), inv (16)/t(16;16), del (5)/del(5q), del (17)/del(17p), normal or complex karyotypes, age and hemoglobin. 17 The same ML models autonomously selected the following factors as predictive for 2-year OS: t(8;21), del (5)/del(5q), inv(16)/t (16;16), del(17)/del(17p), double-mutated CEBPA, CEBPA-bZIP, NPM1, FLT3-ITD, DNMT3A, SF3B1, U2AF1, TP53, age, white blood cell count, peripheral blast count, serum LDH, hemoglobin and extramedullary involvement. 17 Several studies have shown the superiority of artificial intelligence as compared to the ELN 2017 score in predicting survival or personalized outcomes.…”

“…17 The same ML models autonomously selected the following factors as predictive for 2-year OS: t(8;21), del (5)/del(5q), inv(16)/t (16;16), del(17)/del(17p), double-mutated CEBPA, CEBPA-bZIP, NPM1, FLT3-ITD, DNMT3A, SF3B1, U2AF1, TP53, age, white blood cell count, peripheral blast count, serum LDH, hemoglobin and extramedullary involvement. 17 Several studies have shown the superiority of artificial intelligence as compared to the ELN 2017 score in predicting survival or personalized outcomes. 15 Using ML approaches to include posttherapy variables such as MRD is expected to further improve prediction accuracy.…”