14-05-2025
Revised AxSpA Classification Criteria Ready for Prime Time
TORONTO — After almost 10 years, the much-anticipated revision of the current classification criteria for the diagnosis of axial spondyloarthritis (axSpA) is ready for prime time.
An abstract of the modified criteria is being submitted for presentation at American College of Rheumatology 2025 Annual Meeting in October, said Walter P. Maksymowych, MBChB, professor of rheumatology at the University of Alberta, Edmonton, Alberta, Canada.
Walter P. Maksymowych, MBChB
The revised criteria 'have a specificity of almost 90% and will ensure homogeneous trial populations,' he said during a presentation at Spondyloarthritis Research and Treatment Network (SPARTAN) 2025 Annual Meeting.
The modifications were developed using a worldwide dataset from the Classification of Axial Spondyloarthritis Inception Cohort (CLASSIC) trial. Maksymowych is the principal investigator.
The criteria are not meant to be used for the diagnosis of axSpA in clinical practice, he told Medscape Medical News . Admittedly, the revisions will provide greater diagnostic certainty for clinicians wanting to enroll patients in trials. However, the criteria were modified to help investigators identify classic cases of axSpA suitable for clinical trials.
Help Is On the Way to 'Identify an MRI That's Indicative of AxSpA'
'These criteria may limit generalizability, but regulatory bodies such as the FDA [US Food and Drug Administration] and the EMA [European Medicines Agency] will see this as of benefit due to reduced misclassification,' Maksymowych said.
'We want to help radiologists in particular, but also rheumatologists, identify an MRI that's indicative of axSpA. We also thought it was important to identify the clinical factors associated with a diagnosis of axSpA when the MRI imaging is negative.'
Concerns about the current classification criteria, developed in 2009 by the Assessment of SpondyloArthritis International Society (ASAS), surfaced shortly after implementation. Although the criteria made it possible to distinguish axial from peripheral disease, and to diagnose axSpA earlier using an MRI, there were reports of inconsistencies and lower clinical trial response rates.
The prospective CLASSIC trial was an ASAS-SPARTAN initiative launched to test the performance of the criteria using prespecified sensitivity and specificity targets of ≥ 75% and ≥ 90%, respectively.
The trial recruited 1015 patients aged 45 years or younger from 61 centers in 27 countries. All had been consecutively referred to a rheumatologist for lower back pain of at least 3 months' duration and suspected spondyloarthritis.
The findings showed that the sensitivity of the 2009 criteria varied from 73.8% to 82.4% and the specificity ranged from 77.1% to 84.3%, depending on the source of imaging data and the stage of global diagnostic evaluation.
Revising the 2009 criteria is a very important follow-up project, Maksymowych said.
In 2016, the work of modifying the criteria to a specificity threshold of 90% or more was taken on by the ASAS-SPARTAN criteria classification steering committee. The approach was data-driven, with some minor expert consensus modifications along the way.
A multivariate analysis of data from the CLASSIC trial showed that inflammation in the sacroiliac joints (SIJs) on an MRI — even before visible evidence of the structural change — was the most important independent variable associated with the diagnosis of axSpA. The second most important independent variable was inflammatory back pain, followed by the presence of the HLA-B27 allele.
Lianne Gensler, MD
Revisions to the current criteria are long overdue, said Lianne Gensler, MD, professor of medicine at the University of California San Francisco (UCSF) and a member of the steering committee. Some of the issues surrounding the 2009 criteria were the result of clinicians using them as diagnostic criteria, she said in an interview.
In clinical practice, the focus is on diagnostic sensitivity not specificity, explained Gensler, who is director of the UCSF Ankylosing Spondylitis Clinic. 'The goal is to capture all patients with axSpA, including the atypical cases, because you need to treat them. But with classification criteria, the goal is to identify typical cases that will allow researchers to draw an inference. The patient population is much more restricted.'
The modified criteria have a table format, which brings them into alignment with classification criteria for all rheumatic diseases, Maksymowych said. There is also a scoring system with a cutoff of 11 points.
'Classification Criteria Scoring System Really Reflects Clinical Decision-Making'
Liron Caplan, MD, PhD
'Mathematically, it's a simple concept, but the classification criteria scoring system really reflects clinical decision-making,' said Liron Caplan, MD, PhD, associate professor of rheumatology at the University of Colorado, Aurora, Colorado. 'The data from CLASSIC show that confidence takes a big leap when rheumatologists have the imaging data available, specifically the MRI,' he told Medscape Medical News .
The clinical features selected for a diagnosis of axSpA all have an objective manifestation, such as uveitis — 'that painful, red, inflamed eye' — and inflammatory bowel disease, said Caplan, who leads the steering committee and is a past chair of SPARTAN.
'Interpretation of MRI Imaging of Certain Negative States Is Key'
The revised criteria rely very much on the imaging, 'even though a lot of people like to think that, 'Yes, well, we can do this clinically,'' Caplan pointed out. 'Interpretation of the MRI imaging of certain negative states is key.'
Rheumatologists should be talking to radiologists about how to get the MRI done, Maksymowych emphasized in his presentation. 'Bone marrow edema seen in two consecutive slices is no longer sufficient for a diagnosis of axSpA.'
In the CLASSIC trial, MRI was performed at all SPARTAN and most ASAS sites using the first standardized MRI image acquisition protocol (IAP) to diagnose inflammation of the SIJs, he said. 'When you're looking at an MRI, you're not just looking at inflammatory lesions, you're looking at all the sequences simultaneously and ascertaining other structural and inflammatory lesions at the same time.'
The IAP, to which 91% of delegates voted 'yes' at the 2022 annual meeting of ASAS, includes at least four sequences of the SIJs in two planes orthogonal to the sacrum. Inflammation-, fat-, and erosion-sensitive sequences are considered critical for the optimal visualization of inflammation, structural damage, and the bone-cartilage interface.
The standardized IAP for SIJ MRI diagnosis can be applied to any MRI scanner and delivers 'superb' imaging — 'better than you'd see in any clinical trial today,' Maksymowych said. It should be performed within 6 weeks of the patient visit, he added, 'whilst the clinical data is still fresh in the mind of the rheumatologist.'
So what's next?
'We've got this phenomenal dataset. Our imaging dataset is unique,' Maksymowych said.
'We need to provide clear guidance around interpretation of an MRI and transfer knowledge to rheumatologists and radiologists about what constitutes inflammatory back pain.'
