Background: The heterogeneity within the connective tissue disease patient cohort is not adequately reflected by current diagnostic labels(1,2). An immunophenotypic classification may be more appropriate for targeting therapy and basket therapeutic trials. DEFINITION is a multi-disease cohort (n=300) containing multiple clinical and immune datasets.

Objectives: To perform preliminary clustering analysis using the flow cytometric dataset from DEFINITION.

Methods: 142 patients with comprehensive flow cytometry data available were selected from the DEFINITION cohort. Highly sensitive flow cytometry was used to identify 7 B cell, T cell, NK cell and monocyte subsets and tetherin/SIGLEC1 expression on each subset.

ANA and ENA titres were assessed via the Bioplex assay. Demographic and clinical data was obtained from the DEFINITION cohort. Multiple classification algorithms were applied to flow cytometry data using R. Including K-means, hierachical, model-based and partitioning around medioids (PAM). Output clusters were retrospectively compared to their legacy diagnoses and serology.

Results: In the first 142 patients, PAM identified 3 clusters with memberships of 45, 49 and 48 patients respectively ( Table 1 ).

Cluster 1 ( UCTD/CD3+ dominant ) was characterised by a higher proportion of UCTD patients (27%) and larger populations of CD3+ T cells and NK bright cells.

Cluster 2 ( SSc dominant/Tetherin low ) was characterised by a higher population of systemic sclerosis (SSc) patients (10%) and a lower proportion of SLE patients (20%). High levels of classical and non-classical monocytes were demonstrated alongside low plasmablast populations and low SIGLEC1/Tetherin expression.

Cluster 3 ( SLE/pSS high, Ro/Tetherin high ) included a higher proportion of SLE and primary Sjogren’s syndrome (pSS) patients (52% and 19%). This cluster showed larger plasmablast numbers and higher SIGLEC1/Tetherin expression. Serologically this cluster had higher levels of Ro60, Ro52, Sm-RNP and La positivity.

There were statistically significant differences in Ro60, Ro52, La, and Sm antibody positivity between the 3 clusters. Significant differences in CD19+CD27+ memory B cell tetherin expression and CD14+CD16- classical monocyte SIGLEC1 expression were observed (both p < 0.05). A significant difference in SLE distribution was noted between the clusters (p < 0.01).

Conclusion: This preliminary study identified of a flow cytometric dataset identified 3 immunophenotypically distinct subgroups, each comprised of multiple legacy diagnoses within spectrum of ANA+ disease. Future work will evaluate other datasets within the full cohort with a view to conduct basket trials in the baskets defined.

REFERENCES:

[1]Mucke J, Alarcon-Riquelme M, Andersen J, Aringer M, Bombardieri S, Brinks R, et al. What are the topics you care about making trials in lupus more effective? Results of an Open Space meeting of international lupus experts. Lupus Sci Med [Internet]. 2021 May 20 [cited 2021 Jul 12];8(1):506. Available from: /pmc/articles/PMC8141446/

[2]Barturen G, Beretta L, Cervera R, Van Vollenhoven R, Alarcón-Riquelme ME. Moving towards a molecular taxonomy of autoimmune rheumatic diseases [Internet]. Vol. 14, Nature Reviews Rheumatology. Nature Publishing Group; 2018 [cited 2021 Jun 28]. p. 75–93. Available from: https://pubmed.ncbi.nlm.nih.gov/29362467/

Acknowledgements: Funding: AstraZeneca

Disclosure of Interests: Jack Arnold: None declared, Lucy Marie Carter: None declared, Md Yuzaiful Md Yusof: None declared, Zoe Wigston: None declared, Sabih-Ul Hassan: None declared, Katherine Dutton: None declared, Shouvik Dass Speakers bureau: Honoraria from Roche, Consultant of: Roche, Abbvie, UCB & Chugai, Antonios Psarras: None declared, Edward Vital Speakers bureau: GSK, AstraZeneca, Consultant of: AstraZeneca, Lilly, Roche, GSK, Aurinia, Iltoo, Novartis, Grant/research support from: Research grants paid to the employer: AstraZeneca and Sandoz