Reading the Sacroiliac Joint Early: What MRI Reveals in axSpA Save

Axial spondyloarthritis (axSpA) is notorious for being diagnosed late. Imaging is one of the best tools for catching it sooner, and the 2025 ASAS-SPARTAN revised classification criteria reflect that: the most heavily weighted variable is “MRI indicative of axSpA in the sacroiliac joints (SIJ).” Three new abstracts from the SPACE and SPARTAN CLASSIC cohorts show what that finding means in practice.

Damage starts early, and MRI sees it first (Abstr 0890)

In the SPACE cohort, 366 patients with early axSpA were compared with 363 patients with non-axSpA chronic back pain over two years. On conventional radiographs, progression in axSpA was minimal, though more patients met the modified New York criteria at two years (8%). MRI told a different story: erosions, fat lesions, ankylosis, and total structural lesions all increased in axSpA, while patients with non-axSpA back pain showed no change on either modality. Structural progression, though limited, begins early in axSpA and helps separate it from other causes of back pain.

Who is most at risk? (Abstr 0894)

A second SPACE analysis (249 patients with radiographs, 251 with MRI; mean age 30, 54% male) asked which baseline features predict progression. Patients with MRI positivity by ASAS criteria progressed fastest, and HLA-B27 positivity added to the risk. For total structural lesions, patients positive for averaged 1.46 units of progression per year, while patients negative for both averaged essentially none (0.01). Progression of erosions was also greater in patients who were mNY-positive, ASAS-MRI-positive, or HLA-B27-positive. Even in early disease, combining imaging and clinical markers can identify those most likely to accumulate damage.

Which lesions matter for the diagnosis? (Abstr 0891)

The third abstract looked at the SPARTAN CLASSIC inception cohort of 501 North American patients under 45 with undiagnosed back pain, of whom 148 (29.5%) were diagnosed with axSpA. The question was which structural lesions best support an “MRI indicative of axSpA” call. Only erosion alone met the prespecified targets of at least 70% sensitivity and 90% specificity. Fat lesions, backfill, and ankylosis were highly specific but less sensitive overall. Combinations of structural and inflammatory findings did better, with erosion plus bone marrow edema giving the strongest signal for MRI indicative of axSpA. Lesions in isolation were uncommon: most positive MRIs showed both structural and inflammatory changes.

Together, these studies point in one direction. Imaging is key for early diagnosis of axSpA and recognizing structural lesions that increase risk for progression along with clinical manifestations including HLA-B27 presence should aid clinicians for earlier and more tailored treatment.