Search

Background: Increasingly large and complex biomedical data sets challenge conventional hypothesis-driven analytical approaches, however, data-driven unsupervised learning can detect inherent patterns in such data sets., Methods: While unsupervised analysis in the medical literature commonly only utilizes a single clustering algorithm for a given data set, we developed a large-scale model with 605 different combinations of target dimensionalities as well as transformation and clustering algorithms and subsequent meta-clustering of individual results. With this model, we investigated a large cohort of 1383 patients from 59 centers in Germany with newly diagnosed acute myeloid leukemia for whom 212 clinical, laboratory, cytogenetic and molecular genetic parameters were available., Results: Unsupervised learning identifies four distinct patient clusters, and statistical analysis shows significant differences in rate of complete remissions, event-free, relapse-free and overall survival between the four clusters. In comparison to the standard-of-care hypothesis-driven European Leukemia Net (ELN2017) risk stratification model, we find all three ELN2017 risk categories being represented in all four clusters in varying proportions indicating unappreciated complexity of AML biology in current established risk stratification models. Further, by using assigned clusters as labels we subsequently train a supervised model to validate cluster assignments on a large external multicenter cohort of 664 intensively treated AML patients., Conclusions: Dynamic data-driven models are likely more suitable for risk stratification in the context of increasingly complex medical data than rigid hypothesis-driven models to allow for a more personalized treatment allocation and gain novel insights into disease biology., (© 2023. The Author(s).)

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.

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.

Background: The aim of this study was to analyze the long-term survival of AML patients with CEBPA mutations., Patients and Methods: We investigated 88 AML patients with a median age of 61 years and (1) cytogenetically normal AML (CN-AML), (2) monoallelic (moCEBPA) or biallelic (biCEBPA) CEBPA mutation, and (3) intensive induction treatment. 60/88 patients have been described previously with a shorter follow-up., Results: Median follow-up time was 9.8 years (95% CI: 9.4-10.1 years) compared to 3.2 and 5.2 years in our former analyses. Patients with biCEBPA mutations survived significantly longer compared to those with moCEBPA (median overall survival (OS) 9.6 years vs. 1.7 years, p = 0.008). Patients ≤ 60 years and biCEBPA mutations showed a favorable prognosis with a 10-year OS rate of 81%., Conclusions: In this long-term observation we confirmed the favorable prognostic outcome of patients with biCEBPA mutations compared to moCEBPA-mutated CN-AML. The high probability of OS (81%) in younger patients is helpful to guide intensity of postremission therapy.

Purpose: Cytogenetically normal (CN) acute myeloid leukemia (AML) is the largest and most heterogeneous cytogenetic AML subgroup. For the practicing clinician, it is difficult to summarize the prognostic information of the growing number of clinical and molecular markers. Our purpose was to develop a widely applicable prognostic model by combining well-established pretreatment patient and disease characteristics., Patients and Methods: Two prognostic indices for CN-AML (PINA), one regarding overall survival (OS; PINAOS) and the other regarding relapse-free survival (RFS; PINARFS), were derived from data of 572 patients with CN-AML treated within the AML Cooperative Group 99 study (www.aml-score.org)., Results: On the basis of age (median, 60 years; range, 17 to 85 years), performance status, WBC count, and mutation status of NPM1, CEBPA, and FLT3-internal tandem duplication, patients were classified into the following three risk groups according to PINAOS and PINARFS: 29% of all patients and 32% of 381 responding patients had low-risk disease (5-year OS, 74%; 5-year RFS, 55%); 56% of all patients and 39% of responding patients had intermediate-risk disease (5-year OS, 28%; 5-year RFS, 27%), and 15% of all patients and 29% of responding patients had high-risk disease (5-year OS, 3%; 5-year RFS, 5%), respectively. PINAOS and PINARFS stratified outcome within European LeukemiaNet genetic groups. Both indices were confirmed on independent data from Cancer and Leukemia Group B/Alliance trials., Conclusion: We have developed and validated, to our knowledge, the first prognostic indices specifically designed for adult patients of all ages with CN-AML that combine well-established molecular and clinical variables and that are easily applicable in routine clinical care. The integration of both clinical and molecular markers could provide a basis for individualized patient care through risk-adapted therapy of CN-AML., (© 2014 by American Society of Clinical Oncology.)

Chemotherapy (CT) induced critical neutropenia remains a major dose limiting problem in acute leukemias. In order to reduce the phase of risk we gave recombinant human GM-CSF to 30 patients at high risk of early death with acute myeloid leukemia (AML). 19 patients with untreated AML and 1 patient with AML late relapse were 65+ years of age and were treated for CT by the TAD9 regimen. 10 patients at all ages had AML early or second relapse and received S-HAM CT. Starting on day 4 after CT GM-CSF 250 micrograms/m2/d was given by continuous i.v. infusion until neutrophils recovered. GM-CSF reduced the median recovery time of neutrophils by 4 days in the TAD9 and 9 days in the S-HAM CT group when compared to controls. After the CT induced aplasia 3 patients with AML showed a regrowth of their blasts which after the stop of GM-CSF was reversible in 1 patient and unaffectedly continued in 2 patients. 57% of patients attained a complete remission, and the median age of the responders was 65 (34-84) years. Remission duration was not found to be reduced. Thus, GM-CSF reduces CT toxicity with a low risk of promoting the disease and may allow more effective antileukemic treatment.

Chemotherapy (CT) induced critical cytopenia remains as the major dose limiting problem in the treatment of acute leukemias. This is especially true in patients at high risk of early death due to high intensity chemotherapy for relapse or to higher age. In an attempt to reduce the phase of risk we administered human recombinant GM-CSF to 23 patients at a median age of 65 (range 17-84) years including 8 acute myelogenous (AML) and 5 lymphoblastic (ALL) leukemias after early or second relapses and 10 patients of 65 years and older in primary induction treatment for AML. 3 patients with AML had prior bone marrow transplantation. 4 AML's were secondary to tumor chemotherapy (2) or myelodysplastic syndrome (2). The study was part of chemotherapeutic phase II and III studies using TAD9 for primary induction chemotherapy and S-HAM for relapses. Starting on day 4 after the end of intensive CT GM-CSF 250 micrograms/m2/day was given by continuous i.v. infusion and after recovery of neutrophils was deescalated in two 4 day steps and discontinued. 12/23 patients achieved a complete remission (CR) and median age of responders is 61 (range 17-84) years. In the group of the non-transplanted patients the median recovery time of neutrophils is reduced by one week when compared to controls receiving the same chemotherapy (p = 0.002). We observed a leukemic regrowth in 3 patients of 61, 70, and 78 years, 2 of whom having secondary AML. After discontinuation of GM-CSF the regrowth of blasts was reversible in 1 patient and continued unaffectedly in its kinetics in 2 patients.(ABSTRACT TRUNCATED AT 250 WORDS)