Latest advances in diagnosis and management of axial spondyloarthritis
Advances in axSpA imaging
“axSpA represents a spectrum of diseases, it includes nonradiographic [nr] axial SpA at one end to ankylosing spondylitis (AS) at the other end. Some patients may not progress to AS at all,” said Dr Ho-Yin Chung of the University of Hong Kong. “Early and accurate diagnosis of nr-axSpA is important because biologic therapies are more effective in the early stage.”
In the GO-AHEAD study, the use of the tumour necrosis factor (TNF) inhibitor golimumab was associated with significant symptomatic improvements in patients with nr-axSpA, with more patients on golimumab vs placebo achieving 20 percent (71.1 percent vs 40.0 percent with placebo; p<0.0001) and 40 percent improvements (56.7 percent vs 23.0 percent with placebo; p>0.0001) according to the Assessment of SpondyloArthritis International Society (ASAS) criteria. [Arthritis Rheum 2015;67:2702-2712] These improvements were sustained in the open-label extension of GO-AHEAD, and the responses were greater in patients with objective inflammation at baseline. [Van der Heijde D, et al, ACR 2015, abstract 2854; Sieper J, et al, ACR 2016, abstract 712]
“Similarly, adalimumab and secukinumab were also shown to be effective in the treatment of nr-axSpA,” Chung noted. [Arthritis Res Ther 2018;20:61; Arthritis Rheum 2020, doi: 10.1002/art.41477]
“While the presence of bone marrow oedema in sacroiliac joint MRI is listed by ASAS as a key feature for axSpPA diagnosis, it may also appear in other conditions, such as postpartum women with back pain, athletes, and military recruits,” noted Chung. [Arthritis Rheumatol 2018;70:1042-1048]
“MRI spine is useful to diagnose axSpA,” he continued. “Among different types of MRI lesions, fatty corner lesion [FCL] is the commonest and can be used for axSpA diagnosis.”
In a study conducted by Chung and colleagues, incorporating ≥5 whole spine FCLs into ASAS classification criteria for axSpA improved the diagnostic sensitivity to 91.6 percent while the specificity remained at 91.9 percent. Applying ≥3 anterior thoracic FCLs into the criteria showed similar sensitivity of 92.0 percent and specificity of 93.5 percent. “Our findings suggest when ≥5 whole spine FCLs or ≥3 anterior thoracic FCLs are observed, MRI SI joints may not need to be performed,” Chung advised. [BMC Rheumatol 2019;3:17]
“Our team also showed that facet and costovertebral joint lesions were associated with Bath Ankylosing Spondylitis Metrology Index [BASMI] scores. However, rheumatologists frequently overlooked these lesions,” he noted. [Rheumatology 2020;59:2591-2602]
Assessment of inflammation intensity
“Clinical disease assessment tools such as the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), could be affected by medical conditions such as fibromyalgia and depression. Therefore, more objective disease assessment tools are needed for axSpA.” [Clin Exp Rheumatol 2017;35(Suppl 105):50-53]
“Spondyloarthritis Research Consortium of Canada [SPARCC] MRI index, a commonly used MRI scoring system, described the intensity of inflammation in a subjective way,” said Chung. [Arthritis Rheum 2005;53:703-709] The score is mainly used in axSpA research.
“The computer-generated apparent diffusion coefficient [ADC], an index of diffusivity, quantifies the intensity of spinal inflammation,” he pointed out. “In our studies, ADC was shown to correlate with BASDAI, Bath Ankylosing Spondylitis Functional Index (BASFI), patient global assessment, and Ankylosing Spondylitis Disease Activity Score [ASDAS].” [Radiology 2019;291:121-128; RMD Open 2019;5:e001008]
“The combination of STIR images and ADC would allow rheumatologists to assess both the extent and intensity of spinal inflammation in axSpA. In the future, with the development of artificial intelligence, STIR images and ADC could be combined to serve as an imaging biomarker,” concluded Chung.
Differentiating SpA from mimicking conditions
“Delays in axSpA diagnosis have been significantly reduced with improvements in classification criteria,” said Professor Mitsumasa Kishimoto of the Kyorin University School of Medicine, Tokyo, Japan. [Ann Rheum Dis 2009;68:777-783; Ann Rheum Dis 2015;74:e12]
“About 60 percent of healthy women have SpA-like MRI lesions up to 6 months after giving birth. This suggests that in postpartum women with back pain, MRI of sacroiliac joints should be performed only at 6 months after delivery,” he advised. [Ann Rheum Dis 2020;79:929-934]
“Notably, a number of other pathological conditions, such as insufficiency bone fracture, osteosarcoma and Hodgkin’s lymphoma, may mimic the inflammatory lesions seen in SpA,” Kishimoto noted. [Ann Rheum Dis 2009;68:1520-1527]
“Imaging modalities, including MRI, are important tools for early diagnosis of AS and differentiation of mimicking cases,” he continued. “Researchers have identified ASAS- defined erosion in ≥2 consecutive slices, ASAS-defined erosion in ≥3 sacroiliac joint quadrants, and ASAS-defined fat lesion with depth >1 cm in ≥1 sacroiliac joint quadrant, as three high priority candidates for defining an MRI structural lesion typical of axSpA.” [Ann Rheum Dis 2020;79:53-54]
“About 11 percent of SpA patients in Asia have nr-axSpA. Presence of peripheral arthritis and/or extra-articular manifestations could help with diagnosis in these patients, as about 80 percent of nr-axSpA patients have ≥1 peripheral symptom,” suggested Kishimoto. [J Rheumatol 2019;46:896-903; Joint Bone Spine 2011;78:598- 603]
“Compared with AS, patients with diffuse idiopathic skeletal hyperostosis are relatively old [>50 years old], have comorbid diabetes, and do not have facet or sacroiliac joint involvement,” he added.
AS imaging & treatment guidelines: Taiwan experience
“The prevalence of AS is 0.4 percent in the population in Taiwan, with a male-to- female ratio of 3:1,” said Professor James Wei of Chung Shan Medical University Hospital, Taichung, Taiwan. “The mean age at AS onset is 25.6 years, with a mean disease duration of 5.1 years at diagnosis. Extra-articular manifestations are observed in 40 percent of cases and include uveitis, psoriasis and inflammatory bowel disease.” [Semin Arthritis Rheum 2012;42:246-253; Curr Rheumatol Rev 2008;4:81-86; Clinical Rheumatology 2007;26:1685-1691]
Clinical practice recommendations for axSpA diagnosis and management were developed recently in Taiwan. [Int J Rheum Dis 2020;23:24-36; Int J Rheum Dis 2020;23:7-23]
Imaging and diagnosis
“Conventional radiography [CR] of the sacroiliac joint and lumbar spine [to T10] is recommended for the diagnosis of AS or other types of axSpA in patients with inflammatory lower back pain,” said Wei. “If CR is negative and nr-axSpA is suspected, MRI is strongly recommended for assessing whether active inflammation and structural changes are present.”
“If MRI is not available, CT scan can be performed to detect erosions or structural abnormalities of the sacroiliac joint,” he continued. “Ultrasound is recommended for evaluation of peripheral SpA, but not for axSpA. Scintigraphy without accompanying CT scan is not recommended for axSpA diagnosis.”
A diagnostic algorithm for axSpA was also proposed in the consensus recommendations, with consideration for the accessibility of local clinical imaging modalities. (Figure 1) [Int J Rheum Dis 2020;23:24-36]
In terms of axSpA management, the Taiwan consensus recommendations listed five overarching principles as well as 11 recommendations.
“Treatment should be individualized according to signs and symptoms of disease, patient characteristics and treatment goals,” noted Wei. “axSpA patients should be treated with the view of achieving clinical target. While minimal disease activity for axSpA has not been defined yet, achieving an ASDAS score of <2.1 and preferably <1.3 is recommended.”
“Treatment of axSpA should follow a stepwise approach. NSAIDs are the first-line treatment for symptomatic axSpA and should be used at optimal doses. Long-term treatment with systemic glucocorticoids should be avoided, although short-term systemic use and local injections may be beneficial,” he continued. “While conventional synthetic disease-modifying antirheumatic drugs are not recommended for axSpA, they can be effective against peripheral arthritis and extra-articular manifestations.”
“When conventional therapy fails and other causes have been ruled out, biologic therapy with TNF inhibitors or interleukin-17 inhibitors should be considered for axSpA. The choice between these two drug classes depends on comorbidities, such as uveitis, and consideration of the individual agents’ risk-benefit ratio,” Wei said. “Intra- or inter-class switching between biologics or small-molecule therapies may be considered in cases of inadequate response or treatment intolerance.” (Figure 2) [Int J Rheum Dis 2020;23:7-23]
Accurate axSpA diagnosis is challenging. However, rigorous patient selection and elimination of mimicking conditions is necessary in order to recommend the most appropriate treatment, especially when considering biologic agents. Relying on imaging alone leads to low sensitivity and has high false positivity; interpretation of MRI findings has to be carried out in the context of clinical manifestations, along with detailed history taking and a thorough physical exam.
Uveitis is one of the most common comorbidities associated with axSpa, occurring in as many as quarter of the patients. While currently, adalimumab is the only biologic specifically approved for uveitis treatment, phase IV real-world data show that treatment with another monoclonal antibody, golimumab, is associated with a significantly reduced uveitis occurrence rate and disease activity. [J Rheumatol 2019;46:153-159]