Soufiane Ajana1, Audrey Cougnard-Grégoire1, Johanna M Colijn2, Bénédicte M J Merle1, Timo Verzijden2, Paulus T V M de Jong3, Albert Hofman4, Johannes R Vingerling5, Boris P Hejblum6, Jean-François Korobelnik7, Magda A Meester-Smoor2, Marius Ueffing8, Hélène Jacqmin-Gadda1, Caroline C W Klaver9, Cécile Delcourt10. 1. Bordeaux Population Health Research Center, Inserm, Université de Bordeaux, Bordeaux, France. 2. Department of Ophthalmology, Erasmus University Medical Center, Rotterdam, The Netherlands; Department of Epidemiology, Erasmus University Medical Center, Rotterdam, The Netherlands. 3. Department of Retinal Signal Processing, Netherlands Institute of Neurosciences, KNAW, and Department of Ophthalmology, Amsterdam University MC, Amsterdam, The Netherlands; Department of Ophthalmology, Leiden University MC, Leiden, The Netherlands. 4. Department of Epidemiology, Erasmus University Medical Center, Rotterdam, The Netherlands; Department of Epidemiology, Harvard T. H. Chan School of Public Health, Boston, Massachusetts. 5. Department of Ophthalmology, Erasmus University Medical Center, Rotterdam, The Netherlands. 6. Bordeaux Population Health Research Center, Inserm, Université de Bordeaux, Bordeaux, France; Inria, SISTM, Bordeaux Sud-Ouest, Bordeaux, France. 7. Bordeaux Population Health Research Center, Inserm, Université de Bordeaux, Bordeaux, France; Department of Ophthalmology, Bordeaux University Hospital, Bordeaux, France. 8. Department of Ophthalmology, Institute for Ophthalmic Research, University of Tübingen, Tübingen, Germany. 9. Department of Ophthalmology, Erasmus University Medical Center, Rotterdam, The Netherlands; Department of Epidemiology, Erasmus University Medical Center, Rotterdam, The Netherlands; Department of Ophthalmology, Radboud University Medical Center, Nijmegen, The Netherlands; Institute of Molecular and Clinical Ophthalmology Basel, Basel, Switzerland. 10. Bordeaux Population Health Research Center, Inserm, Université de Bordeaux, Bordeaux, France. Electronic address: cecile.delcourt@u-bordeaux.fr.
Abstract
PURPOSE: Current prediction models for advanced age-related macular degeneration (AMD) are based on a restrictive set of risk factors. The objective of this study was to develop a comprehensive prediction model applying a machine learning algorithm allowing selection of the most predictive risk factors automatically. DESIGN: Two population-based cohort studies. PARTICIPANTS: The Rotterdam Study I (RS-I; training set) included 3838 participants 55 years of age or older, with a median follow-up period of 10.8 years, and 108 incident cases of advanced AMD. The Antioxydants, Lipids Essentiels, Nutrition et Maladies Oculaires (ALIENOR) study (test set) included 362 participants 73 years of age or older, with a median follow-up period of 6.5 years, and 33 incident cases of advanced AMD. METHODS: The prediction model used the bootstrap least absolute shrinkage and selection operator (LASSO) method for survival analysis to select the best predictors of incident advanced AMD in the training set. Predictive performance of the model was assessed using the area under the receiver operating characteristic curve (AUC). MAIN OUTCOME MEASURES: Incident advanced AMD (atrophic, neovascular, or both), based on standardized interpretation of retinal photographs. RESULTS: The prediction model retained (1) age, (2) a combination of phenotypic predictors (based on the presence of intermediate drusen, hyperpigmentation in one or both eyes, and Age-Related Eye Disease Study simplified score), (3) a summary genetic risk score based on 49 single nucleotide polymorphisms, (4) smoking, (5) diet quality, (6) education, and (7) pulse pressure. The cross-validated AUC estimation in RS-I was 0.92 (95% confidence interval [CI], 0.88-0.97) at 5 years, 0.92 (95% CI, 0.90-0.95) at 10 years, and 0.91 (95% CI, 0.88-0.94) at 15 years. In ALIENOR, the AUC reached 0.92 at 5 years (95% CI, 0.87-0.98). In terms of calibration, the model tended to underestimate the cumulative incidence of advanced AMD for the high-risk groups, especially in ALIENOR. CONCLUSIONS: This prediction model reached high discrimination abilities, paving the way toward making precision medicine for AMD patients a reality in the near future.
PURPOSE: Current prediction models for advanced age-related macular degeneration (AMD) are based on a restrictive set of risk factors. The objective of this study was to develop a comprehensive prediction model applying a machine learning algorithm allowing selection of the most predictive risk factors automatically. DESIGN: Two population-based cohort studies. PARTICIPANTS: The Rotterdam Study I (RS-I; training set) included 3838 participants 55 years of age or older, with a median follow-up period of 10.8 years, and 108 incident cases of advanced AMD. The Antioxydants, Lipids Essentiels, Nutrition et Maladies Oculaires (ALIENOR) study (test set) included 362 participants 73 years of age or older, with a median follow-up period of 6.5 years, and 33 incident cases of advanced AMD. METHODS: The prediction model used the bootstrap least absolute shrinkage and selection operator (LASSO) method for survival analysis to select the best predictors of incident advanced AMD in the training set. Predictive performance of the model was assessed using the area under the receiver operating characteristic curve (AUC). MAIN OUTCOME MEASURES: Incident advanced AMD (atrophic, neovascular, or both), based on standardized interpretation of retinal photographs. RESULTS: The prediction model retained (1) age, (2) a combination of phenotypic predictors (based on the presence of intermediate drusen, hyperpigmentation in one or both eyes, and Age-Related Eye Disease Study simplified score), (3) a summary genetic risk score based on 49 single nucleotide polymorphisms, (4) smoking, (5) diet quality, (6) education, and (7) pulse pressure. The cross-validated AUC estimation in RS-I was 0.92 (95% confidence interval [CI], 0.88-0.97) at 5 years, 0.92 (95% CI, 0.90-0.95) at 10 years, and 0.91 (95% CI, 0.88-0.94) at 15 years. In ALIENOR, the AUC reached 0.92 at 5 years (95% CI, 0.87-0.98). In terms of calibration, the model tended to underestimate the cumulative incidence of advanced AMD for the high-risk groups, especially in ALIENOR. CONCLUSIONS: This prediction model reached high discrimination abilities, paving the way toward making precision medicine for AMDpatients a reality in the near future.
Authors: Alfredo García-Layana; Maribel López-Gálvez; José García-Arumí; Luis Arias; Alfredo Gea-Sánchez; Juan J Marín-Méndez; Onintza Sayar-Beristain; Germán Sedano-Gil; Tariq M Aslam; Angelo M Minnella; Isabel López Ibáñez; José M de Dios Hernández; Johanna M Seddon Journal: Transl Vis Sci Technol Date: 2022-06-01 Impact factor: 3.048
Authors: Caroline Brandl; Felix Günther; Martina E Zimmermann; Kathrin I Hartmann; Gregor Eberlein; Teresa Barth; Thomas W Winkler; Birgit Linkohr; Margit Heier; Annette Peters; Jeany Q Li; Robert P Finger; Horst Helbig; Bernhard H F Weber; Helmut Küchenhoff; Arthur Mueller; Klaus J Stark; Iris M Heid Journal: BMJ Open Ophthalmol Date: 2022-01-04