Dr Fardina Malik - Rheumatology Advisor

Dr Fardina Malik

Expert Perspectives


Expert Perspectives on Non-Radiographic Axial Spondyloarthritis

Fardina Malik, MD

Practice Community
New York, New York
Hospital and Institutional Affiliations
Instructor, Division of Rheumatology, Department of Medicine
NYU School of Medicine



What are the limitations of the Assessment of SpondyloArthritis international Society (ASAS) classification criteria for describing nonradiographic axial spondyloarthritis (nr-axSpA)?


First of all, I wanted to bring to your attention that the current axial spondyloarthritis (AxSpA) criteria are for axial spondyloarthritis, which includes both ankylosing spondylitis (AS) with radiographic changes and nonradiographic axial SPA (nr-AxSpA). However, these criteria are classification criteria and not diagnostic criteria. Clinicians may want to look at the components to help them reach a diagnosis more quickly, but diagnosis ultimately depends on clinical expertise.

These criteria are primarily designed for research purposes, such as clinical trials or epidemiologic studies, and not for diagnosing a patient who is sitting in front of you, although the criteria can be helpful. I think that we should always underscore this point as sometimes clinicians do not differentiate between classification and diagnostic criteria. That is point number 1.

Having said that, I think the criteria do have good sensitivity and specificity if we combine both the imaging arm and the clinical arm; however, the sensitivity of an individual arm goes down significantly if taken on its own. In addition, I think the clinical arm criteria, which start with HLA- B27 positivity, lacks specificity. I feel that especially in a population in which you may not have HLA-B27 dominance — for example, nonwhites or Hispanics, especially in the United States — we actually don’t know how well those work, especially in a clinical setting. On the other hand, one possible limitation of the imaging arm criteria is that they rely only on specific magnetic resonance imaging (MRI), which has to do with active inflammation, or bone marrow edema. That can also affect sensitivity in a patient who has evidence of prior disease. Maybe they do not have active inflammation at the time of the MRI, but they have erosions, fat changes, and/or mild ankylosis, but these are not included in the imaging arm of the criteria. However, the imaging arm also includes radiographs, and although radiographic examination is not as sensitive as MRI, it can pick up erosions and AS. However, the radiographic changes by themselves do not overcome the limitation of not showing the chronic changes seen on MRI. The clinical arm requires a patient to be HLA-B27-positive. In fact, if a patient is HLA-B27-positive and meets 2 other SpA criteria, they meet the classification criteria for AxSpA, but I definitely would not recommend these for diagnosis.


Imaging remains an important tool for the diagnosis of axSpA, but the role of imaging for monitoring disease activity is less well‐defined. Can you comment on the utility of spine and/or pelvis MRI for patients in whom the degree of active inflammation is unknown?


There is no established guideline for clinicians regarding how, where, and when to perform spine and/or pelvis MRI. As you said, if a clinician is unsure about what is going on, on a case-by-case basis I think there is utility for MRIs, especially to rule out other possibilities. For instance, if the diagnosis was made based on radiographic examination and the patient’s symptoms, but the patient is not responding to first or subsequent tumor necrosis factor (TNF) treatment, MRI could be useful.

Sometimes, we also get lumbar spine imaging as having AS or AxSpA, which does not preclude a patient from having a degenerative disease such as a disk problem that was not initially identified. Once again, it would be great to have an established guideline. As you know, in clinical trials the researchers usually obtain pre- and post-treatment MRI data as one of their outcomes. They also use serial radiographic data to document radiographic progression. Although there are no clear guidelines, imaging has great potential. Clinicians can utilize MRI in these patients on a case-by-case basis.


The Ankylosing Spondylitis Disease Activity Score (ASDAS) has been used to assess disease activity in nr-axSpA. Can you discuss the limitations of ASDAS when applied to nr-axSpA?


RAPID-3 is not disease-specific. It can be used to assess axSpA activity, but it was primarily developed and validated in rheumatoid arthritis. Although the ASDAS and BASDAI are more specific for axSpA, researchers and experts believe that they are slightly more applicable to AS than nr-axSpA. In the most recent certolizumab trial, the researchers used ASDAS and BASDAI as measures of disease activity. I think the burden of information is higher in AS compared with nr-axSpA based on several past studies. I do not think there is a single best measure for pure axSpA, primarily because both the BASDAI and ASDAS also include at least 2 components of peripheral joint disease.

In general, we need a much more streamlined disease activity index that is particular to axSpA regardless of whether it is AS or nr-axSPA because peripheral disease activity sometimes interferes with your ability to capture the disease activity. I am glad you did not mention C- reactive protein (CRP), which creates a problem because CRP is usually higher in AS. There is no ceiling effect. The values go from high to low, which is perfect for a clinician or a researcher as you have much more data to work with. In axSpA, mean CRPs are not as high as you would see in AS. It is hard to use mean CRP levels in clinical day-to-day practice. ASAS, ASAS 20, or ASAS 40 to capture the phenomenon of the disease are slightly better because they have multiple domains.


Skeletal disease in axSpA can be accompanied by extra-articular disease including uveitis, psoriasis, and inflammatory bowel disease. Can you discuss changes in treatment regimens when these specific comorbidities arise?


Once again, we do not have much of a guideline in patients with inflammatory bowel disease (IBD). There are several layers of problems in patients with IBD because they are quite complicated. If you are seeing a patient with IBD who also has axSpA, treatment options can be limited as the axSpA treatment guideline recommends nonsteroidal anti-inflammatory drugs (NSAIDs), tumor necrosis factor (TNF) inhibitors, and interleukin-17 (IL-17). You cannot really use NSAIDs due to the concern that they worsen colitis.

On the other hand, IL-17 blockers actually make Crohn disease (CD) worse when included in treatment regimens when these specific comorbidities arise. We do not know if this is true for ulcerative colitis (UC). So you only have the option of TNF inhibitors. Moreover, CD disease activity does not correlate with axSpA disease activity (meaning IBD can be inactive but they may have active spine inflammation). If such a patient is already receiving treatment with a TNF inhibitor and has active inflammation in their spine despite this, then you may have to switch to a different TNF inhibitor. However, you should not use etanercept because this does not work for IBD.

Uveitis is also important, and adalimumab is the only medication approved for uveitis, both for panuveitis and posterior uveitis when topical steroids do not work. However, if a patient with AS or nr-axSPA with uveitis is having recurrent flares I would probably give them a TNF inhibitor that has known efficacy in uveitis. We avoid etanercept in this scenario again.

We have a lot more flexibility with psoriasis because there are numerous therapies approved for psoriasis and psoriatic arthritis, although we do not know if all of them work for axSpA. For example, ustekinumab failed in both radiographic and nr-axSpA trials, but it is approved for psoriasis and psoriatic arthritis (PsA). There is some indication that patients who have PsA and concomitant axSpA tend to do well with ustekinumab. If I had a patient who developed axSpA while on ustekinumab, of course I would change this medication. I could even try therapies that are not approved for axSpA such as tofacitinib. There are ongoing tofacitinib trials, and it is approved for psoriasis and PsA. I think PsA gives you a little more room to try therapies even though technically these medications are not yet individually approved for axSpA.


What are the current knowledge gaps in axSpa, including education gaps in diagnosis and treatment?


Delay in diagnosis is an age-old problem. We are trying our best to try to minimize the time between disease onset and diagnosis. Unfortunately, we are still seeing delay in diagnosis despite significant advances in the field. We need an integrated approach and to raise awareness without suggesting that everyone with back pain should be assessed for AS. We have to do it in a systematic way that does not overburden the health system, but at the same time creates a referral pipeline. Most of the time, patients go to chiropractors, orthopedic surgeons, or spine surgeons and perhaps, we could encourage that radiographic examination and MRIs of the sacroiliac joints be performed earlier.

I think there is also a gap in the literature, especially with regard to co-existing disease, such as IBD. Also, how do we monitor patients when we keep changing therapy? What do we do when they do not respond to 2 or 3 biologics? We need more information and knowledge on patients who are HLA-B27 negative or of different ethnic backgrounds. Because the United States is a melting pot of different races and ethnicities, I think it would be great to start looking at various patient populations and their disease progression.