Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients
Background The growing awareness of cardiovascular toxicity from cancer therapies has led to the emerging field of cardio‐oncology, which centers on preventing, detecting, and treating patients with cardiac dysfunction before, during, or after cancer treatment. Early detection and prevention of canc...
| Main Authors: | , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-12-01
|
| Series: | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| Subjects: | |
| Online Access: | https://www.ahajournals.org/doi/10.1161/JAHA.120.019628 |
| _version_ | 1819178710122627072 |
|---|---|
| author | Yadi Zhou Yuan Hou Muzna Hussain Sherry‐Ann Brown Thomas Budd W. H. Wilson Tang Jame Abraham Bo Xu Chirag Shah Rohit Moudgil Zoran Popovic Leslie Cho Mohamed Kanj Chris Watson Brian Griffin Mina K. Chung Samir Kapadia Lars Svensson Patrick Collier Feixiong Cheng |
| author_facet | Yadi Zhou Yuan Hou Muzna Hussain Sherry‐Ann Brown Thomas Budd W. H. Wilson Tang Jame Abraham Bo Xu Chirag Shah Rohit Moudgil Zoran Popovic Leslie Cho Mohamed Kanj Chris Watson Brian Griffin Mina K. Chung Samir Kapadia Lars Svensson Patrick Collier Feixiong Cheng |
| author_sort | Yadi Zhou |
| collection | DOAJ |
| description | Background The growing awareness of cardiovascular toxicity from cancer therapies has led to the emerging field of cardio‐oncology, which centers on preventing, detecting, and treating patients with cardiac dysfunction before, during, or after cancer treatment. Early detection and prevention of cancer therapy–related cardiac dysfunction (CTRCD) play important roles in precision cardio‐oncology. Methods and Results This retrospective study included 4309 cancer patients between 1997 and 2018 whose laboratory tests and cardiovascular echocardiographic variables were collected from the Cleveland Clinic institutional electronic medical record database (Epic Systems). Among these patients, 1560 (36%) were diagnosed with at least 1 type of CTRCD, and 838 (19%) developed CTRCD after cancer therapy (de novo). We posited that machine learning algorithms can be implemented to predict CTRCDs in cancer patients according to clinically relevant variables. Classification models were trained and evaluated for 6 types of cardiovascular outcomes, including coronary artery disease (area under the receiver operating characteristic curve [AUROC], 0.821; 95% CI, 0.815–0.826), atrial fibrillation (AUROC, 0.787; 95% CI, 0.782–0.792), heart failure (AUROC, 0.882; 95% CI, 0.878–0.887), stroke (AUROC, 0.660; 95% CI, 0.650–0.670), myocardial infarction (AUROC, 0.807; 95% CI, 0.799–0.816), and de novo CTRCD (AUROC, 0.802; 95% CI, 0.797–0.807). Model generalizability was further confirmed using time‐split data. Model inspection revealed several clinically relevant variables significantly associated with CTRCDs, including age, hypertension, glucose levels, left ventricular ejection fraction, creatinine, and aspartate aminotransferase levels. Conclusions This study suggests that machine learning approaches offer powerful tools for cardiac risk stratification in oncology patients by utilizing large‐scale, longitudinal patient data from healthcare systems. |
| first_indexed | 2024-12-22T21:46:52Z |
| format | Article |
| id | doaj.art-52a1fd8ae6f64644adb36a6b2fa60502 |
| institution | Directory Open Access Journal |
| issn | 2047-9980 |
| language | English |
| last_indexed | 2024-12-22T21:46:52Z |
| publishDate | 2020-12-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| spelling | doaj.art-52a1fd8ae6f64644adb36a6b2fa605022022-12-21T18:11:28ZengWileyJournal of the American Heart Association: Cardiovascular and Cerebrovascular Disease2047-99802020-12-0192310.1161/JAHA.120.019628Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology PatientsYadi Zhou0Yuan Hou1Muzna Hussain2Sherry‐Ann Brown3Thomas Budd4W. H. Wilson Tang5Jame Abraham6Bo Xu7Chirag Shah8Rohit Moudgil9Zoran Popovic10Leslie Cho11Mohamed Kanj12Chris Watson13Brian Griffin14Mina K. Chung15Samir Kapadia16Lars Svensson17Patrick Collier18Feixiong Cheng19Genomic Medicine Institute Lerner Research InstituteCleveland Clinic Cleveland OHGenomic Medicine Institute Lerner Research InstituteCleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHCardio‐Oncology Program Division of Cardiovascular Medicine Medical College of Wisconsin Milwaukee WIDepartment of Hematology/Medical Oncology Taussig Cancer InstituteCleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHDepartment of Hematology/Medical Oncology Taussig Cancer InstituteCleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHDepartment of Radiation Oncology Taussig Cancer InstituteCleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHSchool of Medicine Dentistry and Biomedical Sciences Wellcome‐Wolfson Institute of Experimental MedicineQueen’s University Belfast United KingdomRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHDepartment of Cardiovascular Surgery Cleveland Clinic Cleveland OHRobert and Suzanne Tomsich Department of Cardiovascular Medicine Sydell and Arnold Miller Family Heart and Vascular Institute Cleveland Clinic Cleveland OHGenomic Medicine Institute Lerner Research InstituteCleveland Clinic Cleveland OHBackground The growing awareness of cardiovascular toxicity from cancer therapies has led to the emerging field of cardio‐oncology, which centers on preventing, detecting, and treating patients with cardiac dysfunction before, during, or after cancer treatment. Early detection and prevention of cancer therapy–related cardiac dysfunction (CTRCD) play important roles in precision cardio‐oncology. Methods and Results This retrospective study included 4309 cancer patients between 1997 and 2018 whose laboratory tests and cardiovascular echocardiographic variables were collected from the Cleveland Clinic institutional electronic medical record database (Epic Systems). Among these patients, 1560 (36%) were diagnosed with at least 1 type of CTRCD, and 838 (19%) developed CTRCD after cancer therapy (de novo). We posited that machine learning algorithms can be implemented to predict CTRCDs in cancer patients according to clinically relevant variables. Classification models were trained and evaluated for 6 types of cardiovascular outcomes, including coronary artery disease (area under the receiver operating characteristic curve [AUROC], 0.821; 95% CI, 0.815–0.826), atrial fibrillation (AUROC, 0.787; 95% CI, 0.782–0.792), heart failure (AUROC, 0.882; 95% CI, 0.878–0.887), stroke (AUROC, 0.660; 95% CI, 0.650–0.670), myocardial infarction (AUROC, 0.807; 95% CI, 0.799–0.816), and de novo CTRCD (AUROC, 0.802; 95% CI, 0.797–0.807). Model generalizability was further confirmed using time‐split data. Model inspection revealed several clinically relevant variables significantly associated with CTRCDs, including age, hypertension, glucose levels, left ventricular ejection fraction, creatinine, and aspartate aminotransferase levels. Conclusions This study suggests that machine learning approaches offer powerful tools for cardiac risk stratification in oncology patients by utilizing large‐scale, longitudinal patient data from healthcare systems.https://www.ahajournals.org/doi/10.1161/JAHA.120.019628anthracycline therapycancer therapy–related cardiac dysfunctioncardio‐oncologycardiotoxicityechocardiographymachine learning |
| spellingShingle | Yadi Zhou Yuan Hou Muzna Hussain Sherry‐Ann Brown Thomas Budd W. H. Wilson Tang Jame Abraham Bo Xu Chirag Shah Rohit Moudgil Zoran Popovic Leslie Cho Mohamed Kanj Chris Watson Brian Griffin Mina K. Chung Samir Kapadia Lars Svensson Patrick Collier Feixiong Cheng Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease anthracycline therapy cancer therapy–related cardiac dysfunction cardio‐oncology cardiotoxicity echocardiography machine learning |
| title | Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients |
| title_full | Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients |
| title_fullStr | Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients |
| title_full_unstemmed | Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients |
| title_short | Machine Learning–Based Risk Assessment for Cancer Therapy–Related Cardiac Dysfunction in 4300 Longitudinal Oncology Patients |
| title_sort | machine learning based risk assessment for cancer therapy related cardiac dysfunction in 4300 longitudinal oncology patients |
| topic | anthracycline therapy cancer therapy–related cardiac dysfunction cardio‐oncology cardiotoxicity echocardiography machine learning |
| url | https://www.ahajournals.org/doi/10.1161/JAHA.120.019628 |
| work_keys_str_mv | AT yadizhou machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT yuanhou machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT muznahussain machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT sherryannbrown machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT thomasbudd machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT whwilsontang machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT jameabraham machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT boxu machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT chiragshah machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT rohitmoudgil machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT zoranpopovic machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT lesliecho machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT mohamedkanj machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT chriswatson machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT briangriffin machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT minakchung machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT samirkapadia machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT larssvensson machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT patrickcollier machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients AT feixiongcheng machinelearningbasedriskassessmentforcancertherapyrelatedcardiacdysfunctionin4300longitudinaloncologypatients |