Two shared immune cell signatures stratify patients with Sjögren's syndrome and systemic lupus erythematosus with potential therapeutic implications

• Published in: Arthritis & rheumatology (Hoboken, N.J.)
• Main Authors: Martin-Gutierrez, Lucia, Peng, Junjie, Thompson, Nicolyn L, Robinson, George A, Naja, Meena, Peckham, Hannah, Wu, WingHan, J'bari, Hajar, Ahwireng, Nyarko, Waddington, Kirsty E, Bradford, Claire M, Varnier, Giulia, Gandhi, Akash, Radmore, Rebecca, Gupta, Vivek, Isenberg, David A, Jury, Elizabeth C, Ciurtin, Coziana
• Format: Journal Article
• Language: English
• Published: United States 01-03-2021
• EISSN: 2326-5205

Similarities in the clinical and laboratory features of patients with primary Sjögren's syndrome (pSS) and systemic lupus erythematosus (SLE) have led to attempts to treat pSS and SLE patients with similar biologic therapeutics. However, the results of many clinical trials are disappointing, and no biologic treatments are licensed in pSS, while few are available for SLE patients with refractory disease. Identifying shared immunological features between pSS and SLE could lead to better treatment selection using a stratification approach.
Immune-phenotyping of 29 immune-cell subsets in peripheral blood from patients with pSS (n=45), SLE (n=29) and secondary SS associated with SLE (SLE/SS) (n=14) with low disease activity or in clinical remission, and sex-matched healthy controls (n=31), was performed using flow cytometry. Data were analysed using supervised machine learning (balanced random forest, sparse partial least squares discriminant analysis), logistic regression and multiple t-tests. Patients were stratified by k-means clustering, and clinical trajectory analysis.
Patients with pSS and SLE had a similar immunological architecture despite having different clinical presentations and prognosis. K-means cluster analysis of the combined pSS, SLE and SLE/SS patient cohorts identified two endotypes characterized by distinct immune-cell profiles which spanned patient diagnoses. Logistic regression and machine learning models identified a signature of eight T-cell subsets that differentiated between the two endotypes with high accuracy (AUC=0.9979). Baseline and five-year clinical trajectory analysis identified differential damage scores and disease activity between the two endotypes.
An immune-cell toolkit could differentiate patients across diseases with high accuracy for targeted therapeutic approaches.