Ana Pina1,2,3, Saga Helgadottir4, Rosellina Margherita Mancina5, Chiara Pavanello6, Carlo Pirazzi7, Tiziana Montalcini8, Roberto Henriques9, Laura Calabresi6, Olov Wiklund5, M Paula Macedo1,2,3, Luca Valenti10, Giovanni Volpe4, Stefano Romeo5,7,8. 1. CEDOC - Centro de Estudos de Doenças Crónicas, NOVA Medical School/Faculdade de Ciências Médicas, Universidade Nova de Lisboa, Portugal. 2. Portuguese Diabetes Association, Education and Research Center (APDP-ERC), Portugal. 3. Department of Medical Sciences, University of Aveiro, Portugal. 4. Department of Physics, University of Gothenburg, Sweden. 5. Department of Molecular and Clinical Medicine, Institute of Medicine, Sahlgrenska Academy, Wallenberg Laboratory, University of Gothenburg, Sweden. 6. Centro E. Grossi Paoletti, Dipartimento di Scienze Farmacologiche e Biomolecolari, Università degli Studi di Milano, Italy. 7. Department of Cardiology, Sahlgrenska University Hospital, Sweden. 8. Clinical Nutrition Unit, Department of Medical and Surgical Sciences, University Magna Graecia, Italy. 9. NOVA Information Management School, Campus de Campolide, Portugal. 10. Translational Medicine, Department of Transfusion Medicine and Hematology, Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico and Department of Pathophysiology and Transplantation, Università degli Studi di Milano, Italy.
Abstract
AIMS: Familial hypercholesterolemia (FH) is the most common genetic disorder of lipid metabolism. The gold standard for FH diagnosis is genetic testing, available, however, only in selected university hospitals. Clinical scores - for example, the Dutch Lipid Score - are often employed as alternative, more accessible, albeit less accurate FH diagnostic tools. The aim of this study is to obtain a more reliable approach to FH diagnosis by a "virtual" genetic test using machine-learning approaches. METHODS AND RESULTS: We used three machine-learning algorithms (a classification tree (CT), a gradient boosting machine (GBM), a neural network (NN)) to predict the presence of FH-causative genetic mutations in two independent FH cohorts: the FH Gothenburg cohort (split into training data (N = 174) and internal test (N = 74)) and the FH-CEGP Milan cohort (external test, N = 364). By evaluating their area under the receiver operating characteristic (AUROC) curves, we found that the three machine-learning algorithms performed better (AUROC 0.79 (CT), 0.83 (GBM), and 0.83 (NN) on the Gothenburg cohort, and 0.70 (CT), 0.78 (GBM), and 0.76 (NN) on the Milan cohort) than the clinical Dutch Lipid Score (AUROC 0.68 and 0.64 on the Gothenburg and Milan cohorts, respectively) in predicting carriers of FH-causative mutations. CONCLUSION: In the diagnosis of FH-causative genetic mutations, all three machine-learning approaches we have tested outperform the Dutch Lipid Score, which is the clinical standard. We expect these machine-learning algorithms to provide the tools to implement a virtual genetic test of FH. These tools might prove particularly important for lipid clinics without access to genetic testing.
AIMS: Familial hypercholesterolemia (FH) is the most common genetic disorder of lipid metabolism. The gold standard for FH diagnosis is genetic testing, available, however, only in selected university hospitals. Clinical scores - for example, the Dutch Lipid Score - are often employed as alternative, more accessible, albeit less accurate FH diagnostic tools. The aim of this study is to obtain a more reliable approach to FH diagnosis by a "virtual" genetic test using machine-learning approaches. METHODS AND RESULTS: We used three machine-learning algorithms (a classification tree (CT), a gradient boosting machine (GBM), a neural network (NN)) to predict the presence of FH-causative genetic mutations in two independent FH cohorts: the FH Gothenburg cohort (split into training data (N = 174) and internal test (N = 74)) and the FH-CEGP Milan cohort (external test, N = 364). By evaluating their area under the receiver operating characteristic (AUROC) curves, we found that the three machine-learning algorithms performed better (AUROC 0.79 (CT), 0.83 (GBM), and 0.83 (NN) on the Gothenburg cohort, and 0.70 (CT), 0.78 (GBM), and 0.76 (NN) on the Milan cohort) than the clinical Dutch Lipid Score (AUROC 0.68 and 0.64 on the Gothenburg and Milan cohorts, respectively) in predicting carriers of FH-causative mutations. CONCLUSION: In the diagnosis of FH-causative genetic mutations, all three machine-learning approaches we have tested outperform the Dutch Lipid Score, which is the clinical standard. We expect these machine-learning algorithms to provide the tools to implement a virtual genetic test of FH. These tools might prove particularly important for lipid clinics without access to genetic testing.
Entities:
Keywords:
Familial hypercholesterolemia; cardiovascular disease; dyslipidemia; machine learning; prediction model
Authors: Laney K Jones; Nicole Walters; Andrew Brangan; Catherine D Ahmed; Michael Gatusky; Gemme Campbell-Salome; Ilene G Ladd; Amanda Sheldon; Samuel S Gidding; Mary P McGowan; Alanna K Rahm; Amy C Sturm Journal: J Pers Med Date: 2021-06-21