Thank you for your interest in CCO content. As a guest, please complete the following information fields. These data help ensure our continued delivery of impactful education.
Become a member (or login)? Member benefits include accreditation certificates, downloadable slides, and decision support tools.
Atul Deodhar, MD, MRCP
In this commentary, I discuss 4 key studies presented at EULAR 2021 that address practical issues that practicing rheumatologists face daily concerning axSpA.
C-OPTIMISE: Certolizumab Pegol Therapy Withdrawal in Patients With axSpA Associated With Increased Disease Activity in Patients Without Flares
The C-OPTIMISE study by Gensler and colleagues is important because it deals with the issue of drug‑free remission, which is the closest we come to a cure in patients with axSpA. Studies in RA have shown that the earlier a patient is treated with a biologic, the greater the chance of drug withdrawal or dose reduction. As rheumatologists, we see patients inadvertently reducing the dose of their biologic by forgetting to take it as often as prescribed. It was initially thought that such dose reduction could lead to disease progression, but x‑rays of patients with RA showed that the disease was not progressing. In general, I am fine if patients with RA in remission tells me that they are taking etanercept once every 2 weeks or adalimumab once per month.
In the C-OPTIMISE study, patients with nonradiographic axSpA and patients with ankylosing spondylitis with a disease duration of less than 5 years received 48 weeks of open‑label certolizumab pegol. The enrollment of patients with early-stage disease is an advance, as our experience with RA has taught us that the earlier you treat, the better.
After the patients went into remission, they were randomized in blinded fashion to either continue certolizumab at 400 mg every 2 weeks, at a reduced (off-label) dosage of 200 mg once every 4 weeks, or placebo. Patients were followed from 48-96 weeks.
The investigators found that, at the end of 1 year, 86% of patients receiving the full dose of 200 mg every 2 weeks were flare free, 80% of those receiving the reduced dose were flare free, and 23% of the patients receiving placebo remained flare free.
Gensler and colleagues also determined that the patients who were flare free and receiving placebo had subclinical inflammation with slightly elevated C‑reactive protein (CRP) compared with those receiving certolizumab. Since patients with axSpA also have inflammation in the gut, investigators also measured fecal calprotectin. Fecal calprotectin levels were slightly higher in the placebo arm vs in the full‑dose or half-dose certolizumab arms. The BASDAI and ASDAS were also higher but did not reach flare levels in the placebo arm compared with the treatment arms.
This study raises several questions: Is drug‑free remission an achievable goal in axSpA? Is subclinical inflammation necessarily bad? Is subclinical inflammation going to cause other problems later on?
The COAST‑Y study evaluated blinded withdrawal of ixekizumab in patients with nonradiographic axSpA who had reached remission. It found that 83% of the patients who received some level of ixekizumab remained flare-free after 40 weeks vs 55% in the placebo group.
So that’s where we are heading—flare-free dose reduction or drug withdrawal—and that is what many of our patients are already doing on their own.
What about underlying disease progression and x-ray progression in patients who dose reduce or discontinue therapy? Gensler and colleagues did not look at x‑rays. I would say that x‑ray progression in axSpA is very slow and that at least a 2‑year study would be required to see any x‑ray progression in the spine (not in the sacroiliac joint).
Of course, patients come to us because they cannot do the physical activities they want to do (eg, bending down to get into a crawl space). They don’t come to us and say, “I have developed a small syndesmophyte.” If we are able to control their CRP and their BASDAI score—which is a completely patient‑driven disease activity index—and they’re able to function, any very minute radiographic progression may not be that important.
Should we, as rheumatologists, consider withdrawing or reducing the dosage of a drug in patients with axSpA? Based on the results of the C-OPTIMISE study, I would advise that if patients have low disease activity or inactive disease, with ASDAS <1.3 (the primary entry point for people into the study), and if the patients are in remission after 1 year of treatment, consider halving the dose. At the end of 1 year, there is an 80% chance these patients will still be in remission. If patients do flare and if you treat them with full-dose certolizumab again, there is a good chance that you can return them to flare-free status.
MRI Criteria for Diagnosis of axSpA Consistent With Rheumatologic Diagnosis
The diagnosis of axSpA is very difficult; we have classification criteria but no diagnostic criteria. In addition, there are many reasons for back pain, and HLA-B27 positivity occurs in 6.1% of the general population in the United States.
Patients with RA have swollen joints that are relatively easy to palpate, but in patients with axSpA, the sacroiliac joint is affected, which is difficult to examine by touch. Therefore, MRI is the only way active inflammation of the sacroiliac joint can be assessed.
In this study, Maksymowych and colleagues used 8 expert central readers to read 62 scans from the original ASAS classification criteria cohort, and 5 of the 8 had to agree whether a candidate lesion was compatible with a diagnosis of axSpA. Keep in mind that they had no other clinical information about the patients—just the MRI scans.
The old “positive MRI” definition was based on a “consensus” approach of looking for evidence of active inflammation on the anterior vertebrae, which appear as a bright signal on Short Tau Inversion Recovery (STIR) images in ≥3 sites. This has been considered highly suggestive of axSpA, but it’s not data-driven evidence. Furthermore, this signal is also found in people with nonspecific back pain and healthy controls without backache. For the current study, the central readers’ decisions regarding the presence of spinal MRI findings consistent with axSpA were compared with the current “gold standard,” which is the rheumatologist’s expert opinion as to whether a patient has axSpA.
The investigators found that MRI evidence of bone marrow edema in 4 vertebral corners—or bone marrow edema in 3 vertebral corners in the setting of additional inflammatory lesions or the presence of corner fat—was sufficient for diagnosis of axSpA. This was consistent with a rheumatologist diagnosis of axSpA at 4.4 years of follow-up
This data‑driven approach will be very important for the diagnosis of axSpA. Despite the limitation of small sample size, this study is important, as we do see patients who are referred with a high index of suspicion for axSpA, which is exactly the population addressed in this study.
Structural Lesion Scoring by MRI of Sacroiliac Joints in axSpA Captured Primarily in Central 5 Slices
MRI of the sacroiliac joint must also be done to determine whether a person has axSpA. But how many slices are optimal to identify inflammatory and structural lesions? In another study by Maksymowych and colleagues, they used MRI T1W images with the Digital Imaging and Communications in Medicine series performed in the ASAS classification cohort to compare the “all-slice” approach vs the 5-slice method recommended by the Spondyloarthritis Research Consortium of Canada (SPARCC). Interreader variability, detection of lesions, and the frequency of cases with positive MRI for structural lesions were examined. The structural lesions included erosions, fat deposition (in noneroded bone marrow), sclerosis, “backfill” (fat lesion that appears in the excavated area caused by erosion), and ankylosis. Structural lesion frequency was assessed descriptively by a majority agreement (≥4/7 readers had to agree that the structural lesions suggested axSpA) and by consensus of any 2 central readers.
The investigators found that >75% of all structural lesions (except for ankylosis) were captured by analyzing the 5 central slices advocated by SPARCC.
In summary, this and the other study by Maksymowych and colleagues are moving us away from consensus approaches in axSpA toward data‑driven approaches. We need more of such studies.
I should add that practicing rheumatologists need to understand MRI much better than we currently do. Unfortunately, MRI interpretation is not taught in rheumatology fellowships, so my message to rheumatologists is to learn more about MRI by posing questions to your radiologist colleagues!
Safety and Immunogenicity of BNT162b2 COVID-19 Vaccine Among Adults With AIIRD vs the General Population in Israel
We cannot discuss what happened at EULAR 2021 and not speak about COVID-19. I like this study by Furer and colleagues because it is the largest study to date evaluating the safety, efficacy, disease activity, and immunogenicity of the BNT162b2 mRNA COVID-19 vaccine in patients with AIIRD, including psoriatic arthritis (PsA) and axSpA. The control population comprised people who don’t have AIIRD and who are not receiving immunosuppressive therapy.
In my practice, patients ask such questions as:
This study addresses those questions. The primary endpoint was immunogenicity of the vaccine as determined by serum SARS-CoV-2 anti-S1/S2 IgG seropositivity (cutoff of >15 BAU/mL). Secondary endpoints included the effect of immunosuppressive treatment on the vaccine’s immunogenicity, vaccination efficacy defined as prevention of COVID‑19 disease, safety, and the effect of vaccination on clinical disease activity.
The investigators found that patients with AIIRD showed a substantial but significantly lower postvaccination seropositivity rates than the control population (86% vs 100%, respectively). Lower titers of anti-S1/S2 IgG were also attained by patients with AIIRD compared with healthy controls. In RA, 82% of people were seropositive, whereas people with PsA or axSpA were >90% seropositive. Lower seropositivity rates were associated with being older than 65 years of age, a diagnosis of RA, idiopathic inflammatory myositis, antineutrophil cytoplasmic antibody-associated vasculitis, or other vasculitides, and treatment with rituximab, glucocorticoids, mycophenolate mofetil, or abatacept (with or without methotrexate).
Patients with RA are more likely to receive glucocorticoids, abatacept, and methotrexate, whereas patients with axSpA or PsA don’t receive methotrexate and are less likely to receive glucocorticoids. Anti–IL‑17 therapy did not suppress seropositivity, which is important for patients with axSpA and PsA.
Regarding efficacy, no symptomatic COVID-19 cases occurred in these patients, and only 1 case of mild COVID-19 with full recovery occurred in a control subject. Was this because Israel was so successful in immunizing their population and, therefore, not many cases of COVID-19 occurred there? Does immunogenicity (or antibody titers) correlate with efficacy, or is it a product of population level immunity?
These study results are reassuring and allow me to answer important questions from my patients. The information provided by this study can also be used to reduce vaccine hesitancy.