Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study
Background Cardiac involvement can be an initial manifestation in sarcoidosis. However, little is known about the association between various clinical phenotypes of cardiac sarcoidosis (CS) and outcomes. We aimed to analyze the relation of different clinical manifestations with outcomes of CS and to...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2023-08-01
|
| Series: | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| Subjects: | |
| Online Access: | https://www.ahajournals.org/doi/10.1161/JAHA.123.029481 |
| _version_ | 1797738384167272448 |
|---|---|
| author | Emanuele Bobbio Per Eldhagen Christian L. Polte Clara Hjalmarsson Kristjan Karason Araz Rawshani Pernilla Darlington Susanna Kullberg Peder Sörensson Niklas Bergh Entela Bollano |
| author_facet | Emanuele Bobbio Per Eldhagen Christian L. Polte Clara Hjalmarsson Kristjan Karason Araz Rawshani Pernilla Darlington Susanna Kullberg Peder Sörensson Niklas Bergh Entela Bollano |
| author_sort | Emanuele Bobbio |
| collection | DOAJ |
| description | Background Cardiac involvement can be an initial manifestation in sarcoidosis. However, little is known about the association between various clinical phenotypes of cardiac sarcoidosis (CS) and outcomes. We aimed to analyze the relation of different clinical manifestations with outcomes of CS and to investigate the relative importance of clinical features influencing overall survival. Methods and Results A retrospective cohort of 141 patients with CS enrolled at 2 Swedish university hospitals was studied. Presentation, imaging studies, and outcomes of de novo CS and previously known extracardiac sarcoidosis were compared. Survival free of primary composite outcome (ventricular arrhythmias, heart transplantation, or death) was assessed. Machine learning algorithm was used to study the relative importance of clinical features in predicting outcome. Sixty‐two patients with de novo CS and 79 with previously known extracardiac sarcoidosis were included. De novo CS showed more advanced New York Heart Association class (P=0.02), higher circulating levels of NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) (P<0.001), and troponins (P<0.001), as well as a higher prevalence of right ventricular dysfunction (P<0.001). During a median (interquartile range) follow‐up of 61 (44–77) months, event‐free survival was shorter in patients with de novo CS (P<0.001). The top 5 features predicting worse event‐free survival in order of importance were as follows: impaired tricuspid annular plane systolic excursion, de novo CS, reduced right ventricular ejection fraction, absence of β‐blockers, and lower left ventricular ejection fraction. Conclusions Patients with de novo CS displayed more severe disease and worse outcomes compared with patients with previously known extracardiac sarcoidosis. Using machine learning, right ventricular dysfunction and de novo CS stand out as strong overall predictors of impaired survival. |
| first_indexed | 2024-03-12T13:42:59Z |
| format | Article |
| id | doaj.art-527424d7c5e04e8ca6937b2c1952aa28 |
| institution | Directory Open Access Journal |
| issn | 2047-9980 |
| language | English |
| last_indexed | 2024-03-12T13:42:59Z |
| publishDate | 2023-08-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| spelling | doaj.art-527424d7c5e04e8ca6937b2c1952aa282023-08-23T10:41:23ZengWileyJournal of the American Heart Association: Cardiovascular and Cerebrovascular Disease2047-99802023-08-01121510.1161/JAHA.123.029481Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter StudyEmanuele Bobbio0Per Eldhagen1Christian L. Polte2Clara Hjalmarsson3Kristjan Karason4Araz Rawshani5Pernilla Darlington6Susanna Kullberg7Peder Sörensson8Niklas Bergh9Entela Bollano10Department of Cardiology Sahlgrenska University Hospital Gothenburg SwedenDepartment of Medicine Solna Karolinska Institutet Stockholm SwedenInstitute of Medicine at Sahlgrenska Academy University of Gothenburg Gothenburg SwedenDepartment of Cardiology Sahlgrenska University Hospital Gothenburg SwedenInstitute of Medicine at Sahlgrenska Academy University of Gothenburg Gothenburg SwedenDepartment of Cardiology Sahlgrenska University Hospital Gothenburg SwedenDepartment of Internal Medicine Södersjukhuset Stockholm SwedenDepartment of Respiratory Medicine, Theme Inflammation and Ageing Karolinska University Hospital Stockholm SwedenDepartment of Respiratory Medicine, Theme Inflammation and Ageing Karolinska University Hospital Stockholm SwedenDepartment of Cardiology Sahlgrenska University Hospital Gothenburg SwedenDepartment of Cardiology Sahlgrenska University Hospital Gothenburg SwedenBackground Cardiac involvement can be an initial manifestation in sarcoidosis. However, little is known about the association between various clinical phenotypes of cardiac sarcoidosis (CS) and outcomes. We aimed to analyze the relation of different clinical manifestations with outcomes of CS and to investigate the relative importance of clinical features influencing overall survival. Methods and Results A retrospective cohort of 141 patients with CS enrolled at 2 Swedish university hospitals was studied. Presentation, imaging studies, and outcomes of de novo CS and previously known extracardiac sarcoidosis were compared. Survival free of primary composite outcome (ventricular arrhythmias, heart transplantation, or death) was assessed. Machine learning algorithm was used to study the relative importance of clinical features in predicting outcome. Sixty‐two patients with de novo CS and 79 with previously known extracardiac sarcoidosis were included. De novo CS showed more advanced New York Heart Association class (P=0.02), higher circulating levels of NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) (P<0.001), and troponins (P<0.001), as well as a higher prevalence of right ventricular dysfunction (P<0.001). During a median (interquartile range) follow‐up of 61 (44–77) months, event‐free survival was shorter in patients with de novo CS (P<0.001). The top 5 features predicting worse event‐free survival in order of importance were as follows: impaired tricuspid annular plane systolic excursion, de novo CS, reduced right ventricular ejection fraction, absence of β‐blockers, and lower left ventricular ejection fraction. Conclusions Patients with de novo CS displayed more severe disease and worse outcomes compared with patients with previously known extracardiac sarcoidosis. Using machine learning, right ventricular dysfunction and de novo CS stand out as strong overall predictors of impaired survival.https://www.ahajournals.org/doi/10.1161/JAHA.123.029481cardiac sarcoidosisendomyocardial biopsyheart failureinflammatory heart diseasemachine learningright ventricular function |
| spellingShingle | Emanuele Bobbio Per Eldhagen Christian L. Polte Clara Hjalmarsson Kristjan Karason Araz Rawshani Pernilla Darlington Susanna Kullberg Peder Sörensson Niklas Bergh Entela Bollano Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease cardiac sarcoidosis endomyocardial biopsy heart failure inflammatory heart disease machine learning right ventricular function |
| title | Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study |
| title_full | Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study |
| title_fullStr | Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study |
| title_full_unstemmed | Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study |
| title_short | Clinical Outcomes and Predictors of Long‐Term Survival in Patients With and Without Previously Known Extracardiac Sarcoidosis Using Machine Learning: A Swedish Multicenter Study |
| title_sort | clinical outcomes and predictors of long term survival in patients with and without previously known extracardiac sarcoidosis using machine learning a swedish multicenter study |
| topic | cardiac sarcoidosis endomyocardial biopsy heart failure inflammatory heart disease machine learning right ventricular function |
| url | https://www.ahajournals.org/doi/10.1161/JAHA.123.029481 |
| work_keys_str_mv | AT emanuelebobbio clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT pereldhagen clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT christianlpolte clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT clarahjalmarsson clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT kristjankarason clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT arazrawshani clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT pernilladarlington clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT susannakullberg clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT pedersorensson clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT niklasbergh clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy AT entelabollano clinicaloutcomesandpredictorsoflongtermsurvivalinpatientswithandwithoutpreviouslyknownextracardiacsarcoidosisusingmachinelearningaswedishmulticenterstudy |