Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm

Amalie D Haue; Peter C Holm; Karina Banasik; Kenny Emil Aunstrup; Christian Holm Johansen; Agnete T Lundgaard; Victorine P Muse; Timo Röder; David Westergaard; Piotr J Chmura; Alex H Christensen; Peter E Weeke; Erik Sørensen; Ole B V Pedersen; Sisse R Ostrowski; Kasper K Iversen; Lars V Køber; Henrik Ullum; Henning Bundgaard; Søren Brunak

doi:10.1038/s43856-025-01077-1

Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm

Amalie D Haue
, Peter C Holm
, Karina Banasik
, Kenny Emil Aunstrup
, Christian Holm Johansen
, Agnete T Lundgaard
, Victorine P Muse
, Timo Röder
, David Westergaard
, Piotr J Chmura
, Alex H Christensen
, Peter E Weeke
, Erik Sørensen
, Ole B V Pedersen
, Sisse R Ostrowski
, Kasper K Iversen
, Lars V Køber
, Henrik Ullum
, Henning Bundgaard^*
, Søren Brunak^*

^*Corresponding author for this work

Clinical Immunology

Research output: Contribution to journal › Article › Research › peer-review

Abstract

BACKGROUND: Ischemic heart disease (IHD) is heterogeneous with respect to onset, burden of symptoms, and disease progression. We hypothesized that unsupervised clustering analysis could facilitate identification of distinct and clinically relevant multimorbidity clusters.

METHODS: We included IHD patients who underwent coronary angiography (CAG) or coronary computed tomography angiography (CCTA) between 2004 and 2016 and used the earliest procedure as the index date. Patient health records were obtained from the Danish National Patient Registry, the Danish National Prescription Registry, and two in-hospital laboratory database systems. Genetic data were obtained from the Copenhagen Hospital Biobank. Using registered pre-index diagnosis codes (n = 3046), patients were clustered by application of the Markov Cluster algorithm. Multimorbidity clusters were then characterized using Cox regressions (new ischemic events, non-IHD mortality, and all-cause mortality) and enrichment analysis to explore both risks and phenotypical characteristics.

RESULTS: In a cohort of 72,249 patients with IHD (mean age 63.9 years, 63.1% males), 31 distinct clusters (C1-31, 67,136 patients) are identified. Comparing each cluster to the 30 others, seven clusters (9,590 patients) have significantly higher or lower risk of new ischemic events (five and two clusters, respectively). A total of 18 clusters (35,982 patients) have higher or lower risk of death from non-IHD causes (12 and six clusters, respectively), and 23 clusters have a statistically significant higher or lower risk for all-cause mortality. Cardiovascular or inflammatory diseases are commonly enriched in clusters (13). Distributions for 24 laboratory test results differ significantly across clusters. Polygenic risk scores are increased in a total of 15 clusters (48.4%).

CONCLUSIONS: Based on prior disease profiles, unsupervised clustering robustly stratify patients with IHD in subgroups with similar clinical features and outcomes.

Original language	English
Article number	372
Number of pages	11
Journal	Communications medicine
Volume	5
Issue number	1
DOIs	https://doi.org/10.1038/s43856-025-01077-1
Publication status	Published - 26 Aug 2025

Funding

This work was financially supported by Novo Nordisk Foundation (Grants NNF17OC0027594 and NNF14CC0001) and the Innovation Fund Denmark via the NordForsk project PM Heart (5184-00102B). The authors would like to thank (1) research programmer, Troels Siggaard, Novo Nordisk. Foundation Center for Research, University of Copenhagen, Denmark, for continuous and reliable infrastructure support, and (2) Head of Cardiovascular Research, Hilma Holm, deCODE genetics, Icelan,d for insightful comments.

Funders	Funder number
Novo Nordisk Foundation	NNF17OC0027594, NNF14CC0001
Danish Council for Strategic Research	5184-00102B

Keywords

Association
Multimorbidity
Outcomes
Phenotypes
Patterns
System
Care
Tool

Access to Document

10.1038/s43856-025-01077-1Licence: CC BY

Cite this

Haue, A. D., Holm, P. C., Banasik, K., Aunstrup, K. E., Johansen, C. H., Lundgaard, A. T., Muse, V. P., Röder, T., Westergaard, D., Chmura, P. J., Christensen, A. H., Weeke, P. E., Sørensen, E., Pedersen, O. B. V., Ostrowski, S. R., Iversen, K. K., Køber, L. V., Ullum, H., Bundgaard, H., & Brunak, S. (2025). Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm. Communications medicine, 5(1), Article 372. https://doi.org/10.1038/s43856-025-01077-1

@article{9e8d13df683740a89f300f6add0f8c16,

title = "Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm",

abstract = "BACKGROUND: Ischemic heart disease (IHD) is heterogeneous with respect to onset, burden of symptoms, and disease progression. We hypothesized that unsupervised clustering analysis could facilitate identification of distinct and clinically relevant multimorbidity clusters.METHODS: We included IHD patients who underwent coronary angiography (CAG) or coronary computed tomography angiography (CCTA) between 2004 and 2016 and used the earliest procedure as the index date. Patient health records were obtained from the Danish National Patient Registry, the Danish National Prescription Registry, and two in-hospital laboratory database systems. Genetic data were obtained from the Copenhagen Hospital Biobank. Using registered pre-index diagnosis codes (n = 3046), patients were clustered by application of the Markov Cluster algorithm. Multimorbidity clusters were then characterized using Cox regressions (new ischemic events, non-IHD mortality, and all-cause mortality) and enrichment analysis to explore both risks and phenotypical characteristics.RESULTS: In a cohort of 72,249 patients with IHD (mean age 63.9 years, 63.1\% males), 31 distinct clusters (C1-31, 67,136 patients) are identified. Comparing each cluster to the 30 others, seven clusters (9,590 patients) have significantly higher or lower risk of new ischemic events (five and two clusters, respectively). A total of 18 clusters (35,982 patients) have higher or lower risk of death from non-IHD causes (12 and six clusters, respectively), and 23 clusters have a statistically significant higher or lower risk for all-cause mortality. Cardiovascular or inflammatory diseases are commonly enriched in clusters (13). Distributions for 24 laboratory test results differ significantly across clusters. Polygenic risk scores are increased in a total of 15 clusters (48.4\%).CONCLUSIONS: Based on prior disease profiles, unsupervised clustering robustly stratify patients with IHD in subgroups with similar clinical features and outcomes.",

keywords = "Association, Multimorbidity, Outcomes, Phenotypes, Patterns, System, Care, Tool",

author = "Haue, \{Amalie D\} and Holm, \{Peter C\} and Karina Banasik and Aunstrup, \{Kenny Emil\} and Johansen, \{Christian Holm\} and Lundgaard, \{Agnete T\} and Muse, \{Victorine P\} and Timo R{\"o}der and David Westergaard and Chmura, \{Piotr J\} and Christensen, \{Alex H\} and Weeke, \{Peter E\} and Erik S{\o}rensen and Pedersen, \{Ole B V\} and Ostrowski, \{Sisse R\} and Iversen, \{Kasper K\} and K{\o}ber, \{Lars V\} and Henrik Ullum and Henning Bundgaard and S{\o}ren Brunak",

note = "{\textcopyright} 2025. The Author(s).",

year = "2025",

month = aug,

day = "26",

doi = "10.1038/s43856-025-01077-1",

language = "English",

volume = "5",

journal = "Communications medicine",

issn = "2730-664X",

publisher = "Springer Nature",

number = "1",

}

Haue, AD, Holm, PC, Banasik, K, Aunstrup, KE, Johansen, CH, Lundgaard, AT, Muse, VP, Röder, T, Westergaard, D, Chmura, PJ, Christensen, AH, Weeke, PE, Sørensen, E, Pedersen, OBV, Ostrowski, SR, Iversen, KK, Køber, LV, Ullum, H, Bundgaard, H & Brunak, S 2025, 'Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm', Communications medicine, vol. 5, no. 1, 372. https://doi.org/10.1038/s43856-025-01077-1

TY - JOUR

T1 - Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm

AU - Haue, Amalie D

AU - Holm, Peter C

AU - Banasik, Karina

AU - Aunstrup, Kenny Emil

AU - Johansen, Christian Holm

AU - Lundgaard, Agnete T

AU - Muse, Victorine P

AU - Röder, Timo

AU - Westergaard, David

AU - Chmura, Piotr J

AU - Christensen, Alex H

AU - Weeke, Peter E

AU - Sørensen, Erik

AU - Pedersen, Ole B V

AU - Ostrowski, Sisse R

AU - Iversen, Kasper K

AU - Køber, Lars V

AU - Ullum, Henrik

AU - Bundgaard, Henning

AU - Brunak, Søren

PY - 2025/8/26

Y1 - 2025/8/26

N2 - BACKGROUND: Ischemic heart disease (IHD) is heterogeneous with respect to onset, burden of symptoms, and disease progression. We hypothesized that unsupervised clustering analysis could facilitate identification of distinct and clinically relevant multimorbidity clusters.METHODS: We included IHD patients who underwent coronary angiography (CAG) or coronary computed tomography angiography (CCTA) between 2004 and 2016 and used the earliest procedure as the index date. Patient health records were obtained from the Danish National Patient Registry, the Danish National Prescription Registry, and two in-hospital laboratory database systems. Genetic data were obtained from the Copenhagen Hospital Biobank. Using registered pre-index diagnosis codes (n = 3046), patients were clustered by application of the Markov Cluster algorithm. Multimorbidity clusters were then characterized using Cox regressions (new ischemic events, non-IHD mortality, and all-cause mortality) and enrichment analysis to explore both risks and phenotypical characteristics.RESULTS: In a cohort of 72,249 patients with IHD (mean age 63.9 years, 63.1% males), 31 distinct clusters (C1-31, 67,136 patients) are identified. Comparing each cluster to the 30 others, seven clusters (9,590 patients) have significantly higher or lower risk of new ischemic events (five and two clusters, respectively). A total of 18 clusters (35,982 patients) have higher or lower risk of death from non-IHD causes (12 and six clusters, respectively), and 23 clusters have a statistically significant higher or lower risk for all-cause mortality. Cardiovascular or inflammatory diseases are commonly enriched in clusters (13). Distributions for 24 laboratory test results differ significantly across clusters. Polygenic risk scores are increased in a total of 15 clusters (48.4%).CONCLUSIONS: Based on prior disease profiles, unsupervised clustering robustly stratify patients with IHD in subgroups with similar clinical features and outcomes.

AB - BACKGROUND: Ischemic heart disease (IHD) is heterogeneous with respect to onset, burden of symptoms, and disease progression. We hypothesized that unsupervised clustering analysis could facilitate identification of distinct and clinically relevant multimorbidity clusters.METHODS: We included IHD patients who underwent coronary angiography (CAG) or coronary computed tomography angiography (CCTA) between 2004 and 2016 and used the earliest procedure as the index date. Patient health records were obtained from the Danish National Patient Registry, the Danish National Prescription Registry, and two in-hospital laboratory database systems. Genetic data were obtained from the Copenhagen Hospital Biobank. Using registered pre-index diagnosis codes (n = 3046), patients were clustered by application of the Markov Cluster algorithm. Multimorbidity clusters were then characterized using Cox regressions (new ischemic events, non-IHD mortality, and all-cause mortality) and enrichment analysis to explore both risks and phenotypical characteristics.RESULTS: In a cohort of 72,249 patients with IHD (mean age 63.9 years, 63.1% males), 31 distinct clusters (C1-31, 67,136 patients) are identified. Comparing each cluster to the 30 others, seven clusters (9,590 patients) have significantly higher or lower risk of new ischemic events (five and two clusters, respectively). A total of 18 clusters (35,982 patients) have higher or lower risk of death from non-IHD causes (12 and six clusters, respectively), and 23 clusters have a statistically significant higher or lower risk for all-cause mortality. Cardiovascular or inflammatory diseases are commonly enriched in clusters (13). Distributions for 24 laboratory test results differ significantly across clusters. Polygenic risk scores are increased in a total of 15 clusters (48.4%).CONCLUSIONS: Based on prior disease profiles, unsupervised clustering robustly stratify patients with IHD in subgroups with similar clinical features and outcomes.

KW - Association

KW - Multimorbidity

KW - Outcomes

KW - Phenotypes

KW - Patterns

KW - System

KW - Care

KW - Tool

U2 - 10.1038/s43856-025-01077-1

DO - 10.1038/s43856-025-01077-1

M3 - Article

C2 - 40858818

SN - 2730-664X

VL - 5

JO - Communications medicine

JF - Communications medicine

IS - 1

M1 - 372

ER -

Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm

Abstract

Funding

Keywords

Access to Document

Fingerprint

Cite this