Elsevier

International Journal of Cardiology

Volume 327, 15 March 2021, Pages 117-124
International Journal of Cardiology

Predicting the development of adverse cardiac events in patients with hypertrophic cardiomyopathy using machine learning

https://doi.org/10.1016/j.ijcard.2020.11.003Get rights and content

Highlights

  • •

    We applied modern machine learning methods to hypertrophic cardiomyopathy.

  • •

    Machine learning improves risk prediction of adverse cardiac events.

  • •

    Such novel methods enhance identification of high-risk patients with hypertrophic cardiomyopathy.

Abstract

Background:

Only a subset of patients with hypertrophic cardiomyopathy (HCM) develop adverse cardiac events – e.g., end-stage heart failure, cardiovascular death. Current risk stratification methods are imperfect, limiting identification of high-risk patients with HCM. Our aim was to improve the prediction of adverse cardiac events in patients with HCM using machine learning methods.

Methods:

We applied modern machine learning methods to a prospective cohort of adults with HCM. The outcome was a composite of death due to heart failure, heart transplant, and sudden death. As the reference model, we constructed logistic regression model using known predictors. We determined 20 predictive characteristics based on random forest classification and a priori knowledge, and developed 4 machine learning models.

Results

Of 183 patients in the cohort, the mean age was 53 (SD = 17) years and 45% were female. During the median follow-up of 2.2 years (interquartile range, 0.6–3.8), 33 subjects (18%) developed an outcome event, the majority of which (85%) was heart transplant. The predictive accuracy of the reference model was 73% (sensitivity 76%, specificity 72%) while that of the machine learning model was 85% (e.g., sensitivity 88%, specificity 84% with elastic net regression). All 4 machine learning models significantly outperformed the reference model – e.g., area under the receiver-operating-characteristic curve 0.79 with the reference model vs. 0.93 with elastic net regression (p < 0.001).

Conclusions:

Compared with conventional risk stratification, the machine learning models demonstrated a superior ability to predict adverse cardiac events. These modern machine learning methods may enhance identification of high-risk HCM subpopulations.

Section snippets

Study sample

Patients who were seen at the Center for Advanced Cardiac Care at Columbia University Irving Medical Center (New York, USA) and ≥ 18 years of age with a clinical diagnosis of HCM were prospectively enrolled between 1998 and 2016. The diagnosis of HCM was established on the basis of echocardiographic evidence of LV hypertrophy with increased wall thickness (≥15 mm) out of proportion to the degree of loading conditions without chamber dilatation [2]. HCM phenocopies such as Fabry disease and

Results

During the study period of 1998–2016, a total of 183 patients with HCM were enrolled into the prospective cohort. Table 1 summarizes characteristics of study participants. Overall, mean age was 53 years and 45% were female. At baseline, patients who developed an adverse cardiac event had more severe heart failure symptoms, lower systolic blood pressure at rest and peak exercise, and lower exercise capacity, and were more likely to have a history of adverse cardiovascular events and evidence of

Discussion

In the present study, we applied 4 modern ML approaches (i.e. Lasso regression, elastic net regression, random forest, and gradient boosted decision tree) to the data from 183 well-characterized adults with HCM to predict the development of adverse cardiac events. Compared to the reference model based on logistic regression of 8 previously established clinical parameters, all 4 of the ML models demonstrated superior performance in predicting the primary outcome event. Over a range of clinical

Declaration of competing interest

No author has a relationship with industry to disclose.

Acknowledgements

None.

Funding

Dr. Shimada was supported in part by research grants from the National Institute of Health (Bethesda, MD; R01 HL157216), the American Heart Association (Dallas, TX) National Clinical and Population Research Awards and Career Development Award, Korea Institute of Oriental Medicine (Daejeon, Republic of Korea), and Columbia University Irving Medical Center Irving Institute for Clinical & Translational Research (New York, NY) Precision Medicine Pilot Award. Dr. Maurer was supported by the National

References (30)

  • S.J. Al’Aref et al.

    Determinants of in-hospital mortality after percutaneous coronary intervention: a machine learning approach

    J. Am. Heart Assoc.

    (2019)
  • J.A. Hernesniemi et al.

    Extensive phenotype data and machine learning in prediction of mortality in acute coronary syndrome – the MADDEC study

    Ann. Med.

    (2019)
  • Y. Raita et al.

    Emergency department triage prediction of clinical outcomes using machine learning models

    Crit. Care

    (2019)
  • S.J. Shah et al.

    Phenomapping for novel classification of heart failure with preserved ejection fraction

    Circulation

    (2015)
  • M. Kuhn et al.

    Applied Predictive Modeling

    (2013)

    • Cited by (0)

    1

    Drs. Kochav and Raita contributed equally to this manuscript.

    View full text
    © 2020 Published by Elsevier B.V.