Research finds five distinct clusters of ANCA vasculitis

Reference: October 2024 | Issue 10 | Vol 10 | Page 33

Leveraging advanced artificial intelligence (AI) and big data techniques, researchers have made a significant breakthrough in our understanding of vasculitis.

A study published in the Lancet Rheumatology identified phenotypically distinct subgroups of ANCA-associated vasculitis, using a large real-world dataset.

An international team of researchers from Ireland and Sweden retrieved records of patients with ANCA-associated vasculitis from six European vasculitis registries in the Czech Republic, France, Germany, Poland, Ireland, and Sweden.

Model-based clustering of 17 mixed-type clinical variables was performed using a parsimonious mixture of two latent Gaussian variable models. Clinical validation of the optimal cluster solution was made through summary statistics of the clusters’ demography, phenotypic and serological characteristics, and outcome.

A total of 3,868 patients diagnosed with ANCA-associated vasculitis between November 1, 1966, and March 1, 2023, were included. Of these, 2,434 (62.9 per cent) had granulomatosis with polyangiitis (GPA) and 1,434 (37.1 per cent) had microscopic polyangiitis (MPA).

Five clusters, with distinct phenotype, biochemical presentation, and disease outcome, were identified.

Three of these were characterised by kidney involvement. One cluster was characterised by severe kidney involvement (14.3 per cent), with high C-reactive protein (CRP) and serum creatinine concentrations, and variable ANCA specificity (SK cluster). Another demonstrated myeloperoxidase (MPO)-ANCA-positive kidney involvement (20.2 per cent), with limited extrarenal disease (MPO-K cluster). The other cluster had proteinase 3 (PR3)-ANCA-positive kidney involvement (17.7 per cent), with widespread extrarenal disease (PR3-K cluster).

The two remaining clusters were characterised by relative absence of kidney involvement. One was a predominantly PR3-ANCA-positive cluster (31.1 per cent) with inflammatory multisystem disease (IMS cluster), and the other cluster (16.7 per cent) was mainly younger patients with predominantly ear, nose, and throat involvement and low CRP (YR cluster).

Compared with models fitted with clinical diagnosis or ANCA status, cluster-assigned models demonstrated improved predictive power with respect to both patient and kidney survival.

The study highlights the transformative potential of AI in medical research, particularly in addressing the complexities of rare diseases, where it has previously been impossible to generate sufficiently large cohorts to enable meaningful research. By enabling more precise identification of disease patterns, AI can revolutionise how clinicians approach diagnosis and treatment, offering hope for better outcomes not only for vasculitis patients but also for those suffering from other rare and challenging diseases.

This research provides a blueprint for using advanced technologies to tackle similar challenges in the broader field of rare diseases, potentially leading to breakthroughs that could benefit countless patients worldwide.

Co-author Prof Mark Little, Professor of Nephrology at Trinity College Dublin and Consultant Nephrologist at Tallaght and Beaumont Hospitals, said: “Our research shows that by leveraging advanced AI systems and broad datasets, we can uncover new patterns of this rare autoimmune disease, which have impacts on the probability of adverse outcomes. This allows us to focus potentially toxic therapies on those most likely to benefit.”

The research is part of the EU-funded FAIRVASC project, which connects vasculitis patient registries across Europe, enabling data sharing and advanced analysis to drive forward research and improve patient care.