A machine learning approach to predict mortality due to immune-mediated thrombotic thrombocytopenic purpura

Mouhamed Yazan Abou-Ismail; Chong Zhang; Angela P Presson; Shruti Chaturvedi; Ana G Antun; Andrew M Farland; Ryan Woods; Ara Metjian; Yara A Park; Gustaaf de Ridder; Briana Gibson; Raj S Kasthuri; Darla K Liles; Frank Akwaa; Todd Clover; Lisa Baumann Kreuziger; Meera Sridharan; Ronald S Go; Keith R McCrae; Harsh Vardhan Upreti; Radhika Gangaraju; Nicole K Kocher; X Long Zheng; Jay S Raval; Camila Masias; Spero R Cataland; Andrew D Johnson; Elizabeth Davis; Michael D Evans; Marshall Mazepa; Ming Y Lim

doi:10.1016/j.rpth.2024.102388

A machine learning approach to predict mortality due to immune-mediated thrombotic thrombocytopenic purpura

Res Pract Thromb Haemost. 2024 Mar 19;8(3):102388. doi: 10.1016/j.rpth.2024.102388. eCollection 2024 Mar.

Authors

Mouhamed Yazan Abou-Ismail¹, Chong Zhang², Angela P Presson², Shruti Chaturvedi³, Ana G Antun⁴, Andrew M Farland⁵, Ryan Woods⁵, Ara Metjian⁶, Yara A Park⁷, Gustaaf de Ridder^{7

8}, Briana Gibson^{7

9}, Raj S Kasthuri¹⁰, Darla K Liles¹¹, Frank Akwaa¹², Todd Clover¹³, Lisa Baumann Kreuziger^{14

15}, Meera Sridharan¹⁶, Ronald S Go¹⁶, Keith R McCrae¹⁷, Harsh Vardhan Upreti^{3

18}, Radhika Gangaraju¹⁹, Nicole K Kocher¹⁹, X Long Zheng^{20

21}, Jay S Raval²², Camila Masias²³, Spero R Cataland²⁴, Andrew D Johnson²⁵, Elizabeth Davis²⁶, Michael D Evans²⁷, Marshall Mazepa²⁶, Ming Y Lim¹

Affiliations

¹ Division of Hematology and Hematologic Malignancies, Department of Internal Medicine, University of Utah, Salt Lake City, Utah, USA.
² Division of Epidemiology, Department of Internal Medicine, University of Utah, Salt Lake City, Utah, USA.
³ The Department of Medicine, Johns Hopkins University, Baltimore, Maryland, USA.
⁴ Department of Medicine, Emory University, Atlanta, Georgia, USA.
⁵ Department of Medicine, Wake Forest University, Winston-Salem, North Carolina, USA.
⁶ Department of Medicine, University of Colorado, Denver, Colorado, USA.
⁷ Department of Pathology and Laboratory Medicine, University of North Carolina, Chapel Hill, North Carolina, USA.
⁸ Geisinger Medical Laboratories, Danville, Pennsylvania, USA.
⁹ Department of Pathology and Laboratory Medicine, Emory University, Atlanta, Georgia, USA.
¹⁰ Department of Medicine, University of North Carolina, Chapel Hill, North Carolina, USA.
¹¹ Department of Medicine, East Carolina University, Greenville, North Carolina, USA.
¹² Department of Medicine, University of Rochester, Rochester, New York, USA.
¹³ St Charles Healthcare, Bend, Oregon, USA.
¹⁴ Versiti, Milwaukee, Wisconsin, USA.
¹⁵ Department of Medicine, Medical College of Wisconsin, Milwaukee, Wisconsin, USA.
¹⁶ Department of Medicine, Mayo Clinic, Rochester, Minnesota, USA.
¹⁷ Department of Medicine, Cleveland Clinic, Cleveland, Ohio, USA.
¹⁸ Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas, USA.
¹⁹ Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA.
²⁰ Department of Pathology and Laboratory Medicine, University of Kansas Medical Center, Kansas City, Kansas, USA.
²¹ Institute of Reproductive Medicine and Developmental Sciences, University of Kansas Medical Center, Kansas City, Kansas, USA.
²² Department of Pathology, University of New Mexico, Albuquerque, New Mexico, USA.
²³ Baptist Health South Florida, Miami, Florida, USA.
²⁴ Department of Medicine, The Ohio State University, Columbus, Ohio, USA.
²⁵ Department of Laboratory Medicine and Pathology, University of Minnesota, Minneapolis, Minnesota, USA.
²⁶ Department of Medicine, University of Minnesota, Minneapolis, Minnesota, USA.
²⁷ Clinical & Translational Science Institute, University of Minnesota, Minneapolis, Minnesota, USA.

Abstract

Background: Mortality due to immune-mediated thrombotic thrombocytopenic purpura (iTTP) remains significant. Predicting mortality risk may potentially help individualize treatment. The French Thrombotic Microangiopathy (TMA) Reference Score has not been externally validated in the United States. Recent advances in machine learning technology can help analyze large numbers of variables with complex interactions for the development of prediction models.

Objectives: To validate the French TMA Reference Score in the United States Thrombotic Microangiopathy (USTMA) iTTP database and subsequently develop a novel mortality prediction tool, the USTMA TTP Mortality Index.

Methods: We analyzed variables available at the time of initial presentation, including demographics, symptoms, and laboratory findings. We developed our model using gradient boosting machine, a machine learning ensemble method based on classification trees, implemented in the R package gbm.

Results: In our cohort (n = 419), the French score predicted mortality with an area under the receiver operating characteristic curve of 0.63 (95% CI: 0.50-0.77), sensitivity of 0.35, and specificity of 0.84. Our gradient boosting machine model selected 8 variables to predict acute mortality with a cross-validated area under the receiver operating characteristic curve of 0.77 (95% CI: 0.71-0.82). The 2 cutoffs corresponded to sensitivities of 0.64 and 0.50 and specificities of 0.76 and 0.87, respectively.

Conclusion: The USTMA Mortality Index was acceptable for predicting mortality due to acute iTTP in the USTMA registry, but not sensitive enough to rule out death. Identifying patients at high risk of iTTP-related mortality may help individualize care and ultimately improve iTTP survival outcomes. Further studies are needed to provide external validation. Our model is one of many recent examples where machine learning models may show promise in clinical prediction tools in healthcare.

Keywords: artificial intelligence; machine learning; mortality; statistical models; thrombotic thrombocytopenic purpura.

Abstract

Grants and funding