Conference Coverage

New ACR-EULAR diagnostic criteria proposed for ANCA-associated vasculitides

 

Key clinical point: Provisional classification criteria have been developed to identify patients with granulomatosis with polyangiitis.

Major finding: The nine-item classification criteria model has an area under the curve of 0.98, with high sensitivity (90.7%) and specificity (93.5%).

Data source: More than 1,400 vasculitis cases submitted to the Diagnostic and Classification of the Systemic Vasculitides (DCVAS) study.

Disclosures: Dr. Robson reported having no conflicts of interest.


 

AT RHEUMATOLOGY 2017

– New criteria for classifying granulomatosis with polyangiitis (GPA) have been proposed by the American College of Rheumatology (ACR) and the European League Against Rheumatism (EULAR) .

GPA, a type of antineutrophil cytoplasmic antibody (ANCA)–associated systemic vasculitis, was formerly known as Wegener’s granulomatosis. When a small or medium vessel vasculitis has been diagnosed, the new criteria – which are provisional at present – are invoked to confirm the GPA diagnosis. Patients are assessed for five clinical and four laboratory variables.

Dr. Joanna Robson Sara Freeman/Frontline Medical News

Dr. Joanna Robson

“We know it is very important to have homogeneous groups [of patients] to put into clinical trials,” Joanna Robson, MBBS BSc (Hons), PhD, MRCP, explained at the British Society for Rheumatology annual conference. “The old criteria, developed in the 1990s, ... are not consistent with current disease definitions (for example, microscopic polyangiitis), and they do not include ANCA antibody testing which is now routine,” Dr. Robson of the University of the West of England in Bristol added.

The new criteria were developed as part of the Diagnostic and Classification of the Systemic Vasculitides (DCVAS) study. An international project set up to update the classification criteria for all types of systemic vasculitis, DCVAS involves more than 6,000 patients from 133 sites in 32 countries.

To develop the criteria for GPA, a team of 49 vasculitis experts was asked to review over 1, 400 clinical vignettes of newly diagnosed vasculitis based on cases submitted to the DCVAS. The expert panel was not told the suspected diagnosis. When the panel did not reach consensus on a case, it was further reviewed by seven members of the DCVAS study steering committee.

Using this approach, 85% of 578 cases submitted as GPA were confirmed. Compared with other vasculitides, GPA occurred in younger patients (53 vs. 58 years), was more likely to be proteinase 3–ANCA (81% vs. 3.4%) and c-ANCA (72% vs. 5.5%) positive, and was less likely to be p-ANCA (10.3% vs. 47.3%) and myeloperoxidase-antibody (8% vs. 59.3%) positive.

The next stage was to identify data-driven items that might be used to distinguish GPA from other vasculitides and obtain a clinical consensus on those that were the most important for a diagnosis. Of 1000 possible items that included clinical, laboratory, imaging, and biopsy findings, 91 were retained after regression analysis and 22 appeared independently predictive for GPA.

The top five data-driven items had both c-ANCA and PR3-ANCA antibodies, bloody nasal discharge, nasal ulcers, crusting, or sinonasal congestion or blockage; a high (greater than 1 x 109/L) eosinophil count; and the presence of nasal polyps. There were also some clinical and data-driven items that were considered and used in a final nine-item model, Dr. Robson explained.

A point-based risk score was subsequently developed, with a total score of 5 or more suggesting GPA.

The presence of c-ANCA and PR3-ANCA antibodies, an almost certain indicator of GPA with an odds ratio of 134.8 (95% confidence interval, 62.4–291.1; P less than .001), was given the highest score of 5.

Having bloody or nasal discharge or other nasal symptoms or seeing a granuloma on biopsy were both given a score of 3. A score of 2 was awarded if there were nodules, a mass or cavity on chest imaging, or if there was cartilaginous involvement. A score of 1 was given if there was a loss or reduction in hearing or if the patient had red or painful eyes. Two items – the presence of a high eosinophil count and nasal polyps – were given negative scores (-3 and -4, respectively).

Dr. Robson reported that the nine-item model had an area under the curve of 0.98 and high sensitivity (90.7%) and specificity (93.5%).

Dr. Robson reported having no conflicts of interest.

Next Article: