Machine learning-driven immunophenotypic stratification of mixed connective tissue disease corroborating the clinical heterogeneity

Shinji Izuka; Toshihiko Komai; Takahiro Itamiya; Mineto Ota; Yasuo Nagafuchi; Hirofumi Shoda; Kosuke Matsuki; Kazuhiko Yamamoto; Tomohisa Okamura; Keishi Fujio

doi:10.1093/rheumatology/keae158

Machine learning-driven immunophenotypic stratification of mixed connective tissue disease corroborating the clinical heterogeneity

Rheumatology (Oxford). 2024 Mar 13:keae158. doi: 10.1093/rheumatology/keae158. Online ahead of print.

Authors

Shinji Izuka¹, Toshihiko Komai¹, Takahiro Itamiya^{1

2}, Mineto Ota^{1

2}, Yasuo Nagafuchi^{1

2}, Hirofumi Shoda¹, Kosuke Matsuki³, Kazuhiko Yamamoto⁴, Tomohisa Okamura^{1

2}, Keishi Fujio¹

Affiliations

¹ Department of Allergy and Rheumatology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan.
² Department of Functional Genomics and Immunological Diseases, Graduate School of Medicine, The University of Tokyo, Japan.
³ Research Division, Chugai Pharmaceutical Co., Ltd, Yokohama, Kanagawa, Japan.
⁴ Laboratory for Autoimmune Diseases, RIKEN Center for Integrative Medical Sciences, Yokohama, Kanagawa, Japan.

PMID: 38479808
DOI: 10.1093/rheumatology/keae158

Abstract

Objectives: To stratify patients with mixed connective tissue disease (MCTD) based on their immunophenotype.

Methods: We analyzed the immunophenotype and transcriptome of 24 immune cell subsets from patients with MCTD, systemic lupus erythematosus (SLE), idiopathic inflammatory myopathy (IIM), and systemic sclerosis (SSc) from our functional genome database, ImmuNexUT (https://www.immunexut.org/). MCTD patients were stratified by employing machine learning models including Random Forest, trained by immunophenotyping data from SLE, IIM, and SSc patients. Transcriptomes were analyzed with gene set variation analysis (GSVA) and clinical features of MCTD subgroups were compared.

Results: This study included 215 patients, including 22 patients with MCTD. Machine learning models, constructed to classify SLE, IIM, and SSc patients based on immunophenotyping, were applied to MCTD patients, resulting in 16 classified as SLE-immunophenotype and 6 as non-SLE-immunophenotype. Among MCTD patients, patients with the SLE-immunophenotype had higher proportions of Th1 cells [2.85% (interquartile range (IQR) 1.54-3.91) vs 1.33% (IQR 0.99-1.74) p= 0.027] and plasmablasts [6.35% (IQR 4.17-17.49) vs 2.00% (IQR 1.20-2.80) p= 0.010]. Notably, the number of SLE-related symptoms was higher in patients with the SLE-immunophenotype [2.0 (IQR 1.0-2.0) vs 1.0 (IQR1.0-1.0) p= 0.038]. Moreover, GSVA scores of interferon-α and -γ responses were significantly higher in patients with the SLE-immunophenotype in central memory CD8+ T cells, while hedgehog signalling was higher in non-SLE-immunophenotype patients in 5 cell subsets.

Conclusion: This study describes the stratification of MCTD patients based on immunophenotyping, suggesting the presence of distinct immunological processes behind the clinical subtypes of MCTD.

Keywords: idiopathic inflammatory myopathies; immunophenotyping; machine learning; mixed connective tissue disease; systemic lupus erythematosus; systemic sclerosis.