New ways of grouping AAV may better predict symptoms, risks
Study finds clinical signs, antibody type, and data-based clusters add insight
Using both traditional and data-driven ways of grouping patients may give a clearer, more accurate picture of how different types of ANCA-associated vasculitis (AAV) behave over time and how best to treat each person, according to a study from Japan.
Traditional ways of classifying AAV, based on symptoms or on ANCA antibodies, “provide complementary, not competing, prognostic insights in Japanese patients with AAV,” the researchers wrote. “Data-driven clustering [grouping] revealed additional clinical [variability] not captured by traditional systems, underscoring the need for integrated, multi-dimensional stratification approaches to improve personalised risk assessment and treatment strategies.”
The study, “Phenotype, serotype, and data-driven clustering reveal complementary dimensions of heterogeneity in ANCA-associated vasculitis: a multicentre Japanese cohort (J-CANVAS),” was published in Rheumatology International.
Understanding the different ways AAV can be classified
AAV is a group of autoimmune diseases that cause inflammation in small blood vessels. It is mainly driven by self-antibodies, called ANCAs, that target the proteins proteinase 3 (PR3) or myeloperoxidase (MPO).
Traditionally, AAV types such as granulomatosis with polyangiitis (GPA) and microscopic polyangiitis (MPA) were defined by clinical findings. More recently, doctors often classify AAV based on whether ANCAs against PR3 or MPO are present in the blood.
ANCAs targeting PR3 are most often found in people with GPA, while MPO-ANCAs are more commonly associated with MPA.
However, there is still debate about the best way to classify AAV because neither clinical features nor ANCA type alone fully captures how the disease appears or how it may change over time.
“Given the limitations of current AAV classification systems and the ongoing debate surrounding them, comparative evaluations of their prognostic value are warranted,” the researchers wrote, adding that simple two-way groupings are not enough, and a more integrated approach is needed.
How the study compared clinical, antibody-based, and data-driven groupings
To address this need, the researchers compared three ways of grouping patients: one based on symptoms and organ involvement (phenotype), one based on ANCA type (serotype), and one using computer-generated clusters built from large sets of clinical and laboratory data.
The study included 729 adults who were newly diagnosed with GPA or MPA who tested positive for ANCAs against PR3 or MPO. All were enrolled in a large Japanese registry called J-CANVAS.
A total of 557 people had MPA, while 172 had GPA. Among all participants, 650 were MPO-ANCA–positive and 79 were PR3-ANCA–positive.
“Phenotype and serotype classifications were strongly associated, although some discordance was observed; for example, 16.8% of MPO-ANCA-positive patients had GPA, and 20.3% of PR3-ANCA-positive patients had MPA,” the researchers wrote.
Classifying patients by phenotype helped predict the risk of death: people with MPA were more than twice as likely to die as those with GPA. When phenotype and serotype were combined, the group with MPA and anti-MPO ANCAs had the highest risk of death, while those with GPA and PR3 ANCAs faced the greatest risk of severe relapses.
These findings contribute to the ongoing debate on optimal AAV classification and emphasise the need for approaches that more accurately reflect the clinical diversity of patients.
Some groups did not fit the usual patterns. For example, GPA patients who were MPO-ANCA-positive showed unexpected clinical features, suggesting that these “mixed” or discordant groups may need closer attention in clinical practice.
Further analyses found no significant difference in remission rates at 24 weeks (six months) — whether patients were treated with rituximab (approved for MPA and GPA) or cyclophosphamide (commonly used off-label). This pattern held true overall and within every classification system.
“In this extensive Japanese [group of patients], we discovered that the traditional classification based on phenotype (MPA vs. GPA) and ANCA serotype (MPO vs. PR3) each offer unique prognostic insights, indicating that they are complementary rather than competing,” the researchers wrote. “However, neither approach fully captured the [variability] among patients.”
What data-driven clusters revealed about AAV symptom patterns
The data-driven approach identified four patient clusters, each defined by different patterns in symptoms and lab findings.
Cluster 1, the largest group, included mostly MPA patients who were MPO-ANCA positive, typically older people with lung and kidney involvement. Cluster 2 consisted mainly of younger GPA patients, positive for PR3- or MPO-ANCAs, and was characterized by lung nodules/cavities and ear, nose, and throat (ENT) involvement.
Cluster 3, which included some GPA patients with PR3-ANCAs, showed more severe disease, higher C-reactive protein, and involvement of the ENT, kidneys, and cardiovascular system.
Cluster 4 included people from multiple traditional subgroups. They were generally younger, had milder disease and low inflammation, but showed more general symptoms, including issues affecting the musculoskeletal system and the peripheral nervous system (PNS).
“A data-driven clustering analysis identified four distinct subgroups, which demonstrated limited concordance with conventional classifications, underscoring the limitations of existing classification frameworks,” the researchers wrote. “These findings contribute to the ongoing debate on optimal AAV classification and emphasise the need for approaches that more accurately reflect the clinical diversity of patients. Enhanced stratification may require the integration of multiple data dimensions, including phenotype, serotype, data-driven clusters, and biomarkers.”
About the Author
