Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML
Abstract Risk stratification in acute myeloid leukemia (AML) is driven by genetics, yet patient age substantially influences therapeutic decisions. To evaluate how age alters the prognostic impact of genetic mutations, we pooled data from 3062 pediatric and adult AML patients from multiple cohorts....
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-05-01
|
| Series: | HemaSphere |
| Online Access: | https://doi.org/10.1002/hem3.70132 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839630632483815424 |
|---|---|
| author | Jan‐Niklas Eckardt Waldemar Hahn Rhonda E. Ries Szymon D. Chrost Susann Winter Sebastian Stasik Christoph Röllig Uwe Platzbecker Carsten Müller‐Tidow Hubert Serve Claudia D. Baldus Christoph Schliemann Kerstin Schäfer‐Eckart Maher Hanoun Martin Kaufmann Andreas Burchert Johannes Schetelig Martin Bornhäuser Markus Wolfien Soheil Meshinchi Christian Thiede Jan Moritz Middeke |
| author_facet | Jan‐Niklas Eckardt Waldemar Hahn Rhonda E. Ries Szymon D. Chrost Susann Winter Sebastian Stasik Christoph Röllig Uwe Platzbecker Carsten Müller‐Tidow Hubert Serve Claudia D. Baldus Christoph Schliemann Kerstin Schäfer‐Eckart Maher Hanoun Martin Kaufmann Andreas Burchert Johannes Schetelig Martin Bornhäuser Markus Wolfien Soheil Meshinchi Christian Thiede Jan Moritz Middeke |
| author_sort | Jan‐Niklas Eckardt |
| collection | DOAJ |
| description | Abstract Risk stratification in acute myeloid leukemia (AML) is driven by genetics, yet patient age substantially influences therapeutic decisions. To evaluate how age alters the prognostic impact of genetic mutations, we pooled data from 3062 pediatric and adult AML patients from multiple cohorts. Signaling pathway mutations dominated in younger patients, while mutations in epigenetic regulators, spliceosome genes, and TP53 alterations became more frequent with increasing age. Machine learning models were trained to identify prognostic variables and predict complete remission and 2‐year overall survival, achieving area‐under‐the‐curve scores of 0.801 and 0.791, respectively. Using Shapley (SHAP) values, we quantified the contribution of each variable to model decisions and traced their impact across six age groups: infants, children, adolescents/young adults, adults, seniors, and elderly. The highest contributions to model decisions among genetic variables were found for alterations of NPM1, CEBPA, inv(16), and t(8;21) conferring favorable risk and alterations of TP53, RUNX1, ASXL1, del(5q), ‐7, and ‐17 conferring adverse risk, while FLT3‐ITD had an ambiguous role conferring favorable treatment responses yet poor overall survival. Age significantly modified the prognostic value of genetic alterations, with no single alteration consistently predicting outcomes across all age groups. Specific alterations associated with aging such as TP53, ASXL1, or del(5q) posed a disproportionately higher risk in younger patients. These results challenge uniform risk stratification models and highlight the need for context‐sensitive AML treatment strategies. |
| format | Article |
| id | doaj-art-2a055feee1fa40b5b6c1eb2b0d0f75a4 |
| institution | Matheson Library |
| issn | 2572-9241 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | Wiley |
| record_format | Article |
| series | HemaSphere |
| spelling | doaj-art-2a055feee1fa40b5b6c1eb2b0d0f75a42025-07-13T08:46:18ZengWileyHemaSphere2572-92412025-05-0195n/an/a10.1002/hem3.70132Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AMLJan‐Niklas Eckardt0Waldemar Hahn1Rhonda E. Ries2Szymon D. Chrost3Susann Winter4Sebastian Stasik5Christoph Röllig6Uwe Platzbecker7Carsten Müller‐Tidow8Hubert Serve9Claudia D. Baldus10Christoph Schliemann11Kerstin Schäfer‐Eckart12Maher Hanoun13Martin Kaufmann14Andreas Burchert15Johannes Schetelig16Martin Bornhäuser17Markus Wolfien18Soheil Meshinchi19Christian Thiede20Jan Moritz Middeke21Department of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyCenter for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI) Dresden/Leipzig Dresden GermanyTranslational Science and Therapeutics Division Fred Hutchinson Cancer Research Center Seattle Washington USADepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyDepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyDepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyDepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyDepartment of Hematology, Cellular Therapy, Hemostaseology and Infectious Disease University of Leipzig Medical Center Leipzig GermanyDepartment of Medicine V University Hospital Heidelberg Heidelberg GermanyDepartment of Medicine 2, Hematology and Oncology Goethe University Frankfurt Frankfurt GermanyDepartment of Hematology and Oncology University Hospital Schleswig Holstein Kiel GermanyDepartment of Medicine A University Hospital Münster Münster GermanyDepartment of Internal Medicine V Paracelsus Medizinische Privatuniversität and University Hospital Nürnberg Nürnberg GermanyDepartment of Hematology University Hospital Essen Essen GermanyDepartment of Hematology, Oncology and Palliative Care Robert Bosch Hospital Stuttgart GermanyDepartment of Hematology, Oncology and Immunology Philipps‐University Marburg Marburg GermanyDepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyDepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyCenter for Scalable Data Analytics and Artificial Intelligence (ScaDS.AI) Dresden/Leipzig Dresden GermanyTranslational Science and Therapeutics Division Fred Hutchinson Cancer Research Center Seattle Washington USADepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyDepartment of Internal Medicine I University Hospital Carl Gustav Carus, TUD Dresden University of Technology Dresden GermanyAbstract Risk stratification in acute myeloid leukemia (AML) is driven by genetics, yet patient age substantially influences therapeutic decisions. To evaluate how age alters the prognostic impact of genetic mutations, we pooled data from 3062 pediatric and adult AML patients from multiple cohorts. Signaling pathway mutations dominated in younger patients, while mutations in epigenetic regulators, spliceosome genes, and TP53 alterations became more frequent with increasing age. Machine learning models were trained to identify prognostic variables and predict complete remission and 2‐year overall survival, achieving area‐under‐the‐curve scores of 0.801 and 0.791, respectively. Using Shapley (SHAP) values, we quantified the contribution of each variable to model decisions and traced their impact across six age groups: infants, children, adolescents/young adults, adults, seniors, and elderly. The highest contributions to model decisions among genetic variables were found for alterations of NPM1, CEBPA, inv(16), and t(8;21) conferring favorable risk and alterations of TP53, RUNX1, ASXL1, del(5q), ‐7, and ‐17 conferring adverse risk, while FLT3‐ITD had an ambiguous role conferring favorable treatment responses yet poor overall survival. Age significantly modified the prognostic value of genetic alterations, with no single alteration consistently predicting outcomes across all age groups. Specific alterations associated with aging such as TP53, ASXL1, or del(5q) posed a disproportionately higher risk in younger patients. These results challenge uniform risk stratification models and highlight the need for context‐sensitive AML treatment strategies.https://doi.org/10.1002/hem3.70132 |
| spellingShingle | Jan‐Niklas Eckardt Waldemar Hahn Rhonda E. Ries Szymon D. Chrost Susann Winter Sebastian Stasik Christoph Röllig Uwe Platzbecker Carsten Müller‐Tidow Hubert Serve Claudia D. Baldus Christoph Schliemann Kerstin Schäfer‐Eckart Maher Hanoun Martin Kaufmann Andreas Burchert Johannes Schetelig Martin Bornhäuser Markus Wolfien Soheil Meshinchi Christian Thiede Jan Moritz Middeke Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML HemaSphere |
| title | Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML |
| title_full | Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML |
| title_fullStr | Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML |
| title_full_unstemmed | Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML |
| title_short | Age‐stratified machine learning identifies divergent prognostic significance of molecular alterations in AML |
| title_sort | age stratified machine learning identifies divergent prognostic significance of molecular alterations in aml |
| url | https://doi.org/10.1002/hem3.70132 |
| work_keys_str_mv | AT janniklaseckardt agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT waldemarhahn agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT rhondaeries agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT szymondchrost agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT susannwinter agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT sebastianstasik agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT christophrollig agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT uweplatzbecker agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT carstenmullertidow agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT hubertserve agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT claudiadbaldus agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT christophschliemann agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT kerstinschafereckart agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT maherhanoun agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT martinkaufmann agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT andreasburchert agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT johannesschetelig agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT martinbornhauser agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT markuswolfien agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT soheilmeshinchi agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT christianthiede agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml AT janmoritzmiddeke agestratifiedmachinelearningidentifiesdivergentprognosticsignificanceofmolecularalterationsinaml |