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

Achievement of complete remission 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 tailor treatment concepts individually to disease biology. We used nine machine learning (ML) models to predict complete remission and 2-year overall survival in a large multicenter cohort of 1,383 AML patients who received intensive induction therapy. Clinical, laboratory, cytogenetic and molecular genetic data were incorporated and our results were validated on an external multicenter cohort. Our ML models autonomously selected predictive features 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 CEBPA, mutations of CEBPA-bZIP, NPM1, FLT3-ITD, ASXL1, RUNX1, SF3B1, IKZF1, TP53, and U2AF1, t(8;21), inv(16)/t(16;16), del(5)/del(5q), del(17)/del(17p), normal or complex karyotypes, age and hemoglobin concentration at initial diagnosis were statistically significant markers predictive of complete remission, while 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, and TP53 mutations, age, white blood cell count, peripheral blast count, serum lactate dehydrogenase level and hemoglobin concentration at initial diagnosis as well as extramedullary manifestations were predictive for 2-year overall survival. For prediction of complete remission and 2-year overall survival areas under the receiver operating characteristic curves ranged between 0.77-0.86 and between 0.63-0.74, respectively in our test set, and between 0.71-0.80 and 0.65-0.75 in the external validation cohort. We demonstrated 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.

A clinically applicable gene expression-based score predicts resistance to induction treatment in acute myeloid leukemia.

Prediction of resistant disease at initial diagnosis of acute myeloid leukemia (AML) can be achieved with high accuracy using cytogenetic data and 29 gene expression markers (Predictive Score 29 Medical Research Council; PS29MRC). Our aim was to establish PS29MRC as a clinically usable assay by using the widely implemented NanoString platform and further validate the classifier in a more recently treated patient cohort. Analyses were performed on 351 patients with newly diagnosed AML intensively treated within the German AML Cooperative Group registry. As a continuous variable, PS29MRC performed best in predicting induction failure in comparison with previously published risk models. The classifier was strongly associated with overall survival. We were able to establish a previously defined cutoff that allows classifier dichotomization (PS29MRCdic). PS29MRCdic significantly identified induction failure with 59% sensitivity, 77% specificity, and 72% overall accuracy (odds ratio, 4.81; P = 4.15 × 10-10). PS29MRCdic was able to improve the European Leukemia Network 2017 (ELN-2017) risk classification within every category. The median overall survival with high PS29MRCdic was 1.8 years compared with 4.3 years for low-risk patients. In multivariate analysis including ELN-2017 and clinical and genetic markers, only age and PS29MRCdic were independent predictors of refractory disease. In patients aged ≥60 years, only PS29MRCdic remained as a significant variable. In summary, we confirmed PS29MRC as a valuable classifier to identify high-risk patients with AML. Risk classification can still be refined beyond ELN-2017, and predictive classifiers might facilitate clinical trials focusing on these high-risk patients with AML.

The NPM1 mutation type has no impact on survival in cytogenetically normal AML.

NPM1 mutations represent frequent genetic alterations in patients with acute myeloid leukemia (AML) associated with a favorable prognosis. Different types of NPM1 mutations have been described. The purpose of our study was to evaluate the relevance of different NPM1 mutation types with regard to clinical outcome. Our analyses were based on 349 NPM1-mutated AML patients treated in the AMLCG99 trial. Complete remission rates, overall survival and relapse-free survival were not significantly different between patients with NPM1 type A or rare type mutations. The NPM1 mutation type does not seem to play a role in risk stratification of cytogenetically normal AML.