CT Delivers the Strongest Diagnostic Read in Suspected Axial Spondyloarthritis
Choosing the right imaging test for axial spondyloarthritis is not a cosmetic decision. It determines whether inflammatory or structural sacroiliac disease is recognized early, whether mimics are filtered out, and whether a patient is moved toward targeted treatment or left in diagnostic limbo. In this comparative study, radiography, CT, and MRI of the sacroiliac joints were tested head-to-head in 163 patients, including 89 with axSpA and 74 with degenerative conditions. Three blinded experts reviewed the scans in five separate reading rounds: XR alone, CT alone, MR alone, XR plus MR, and CT plus MR. The clinical diagnosis served as the reference standard.
The hierarchy was unmistakable. X-ray underperformed both cross-sectional techniques, with lower sensitivity and lower specificity than MRI and CT. CT emerged as the most reliable discriminator between axSpA and non-inflammatory disease, while MRI remained stronger than radiography but did not match CT for structural accuracy.
Radiography Missed Too Many Cases and Misclassified Too Many Others
Plain radiography has long held a place in the workup of suspected sacroiliitis because it is inexpensive, familiar, and widely available. Yet sacroiliac joint change evolves slowly on X-ray, and early disease can sit below the threshold of visible structural damage. In this cohort, that limitation translated into a sensitivity of 66.3% for XR, compared with 82.0% for MRI and 76.4% for CT. In practical terms, a normal radiograph left too much room for missed disease.
Specificity told a similar story. XR reached 67.6%, while MRI reached 86.5% and CT reached 97.3%. The false-positive burden on radiography matters because degenerative or mechanically altered joints can resemble inflammatory disease when images are read without the added detail offered by cross-sectional methods. In a disorder where classification and diagnosis can shape long-term treatment trajectories, a weak specificity profile carries real clinical consequences.
Cross-sectional imaging changed that picture. CT was the sharpest tool in the comparison, separating axSpA from degenerative pathology with the fewest errors. MRI improved both sensitivity and specificity over plain films, but it still left some diagnostic uncertainty when structural lesions were the key question.
Structural Lesions Carried More Diagnostic Weight Than Plain Film Change
The superiority of CT in this study points to the importance of structural sacroiliac lesions in the differential diagnosis of axSpA. Bone erosion, sclerosis, ankylotic change, and other chronic structural abnormalities are often easier to define on CT than on radiography or MRI. MRI remains valuable for detecting active inflammatory lesions, particularly bone marrow edema, but the data here suggest that inflammation alone does not always resolve the diagnostic question when degenerative disease is in the differential.
That distinction is clinically consequential. Axial spondyloarthritis is not diagnosed by a single imaging sign in isolation; the diagnosis depends on a synthesis of symptoms, laboratory findings, and imaging context. When the imaging question is whether the sacroiliac joints show disease consistent with axSpA or a non-inflammatory mimic, CT provides a clearer structural map. The study’s results support a more deliberate use of CT when the goal is to discriminate between chronic inflammatory damage and degenerative change.
MRI remained a strong modality, but not the final answer in every patient. Its sensitivity of 82.0% and specificity of 86.5% make it substantially better than XR, and its role in detecting active inflammation remains important. Still, when imaging interpretation hinges on structural definition, CT had the edge.
Combining Imaging Improved Some Readings, but MRI Did Not Gain Much From X-ray
The combined-reading arms sharpened the study’s message. XR plus MRI produced sensitivity and specificity similar to MRI alone, with 77.5% sensitivity and 87.8% specificity. In other words, adding plain radiography to MRI did not materially improve diagnostic performance. The image pair did not create a meaningful gain over MRI on its own.
CT plus MRI performed better, reaching 75.3% sensitivity and 97.3% specificity. That combination preserved CT’s high specificity and improved interpretive confidence. For clinicians, this points to a simple pattern: if MRI is already available, adding XR contributes little. If CT is added to MRI, the structural information becomes more decisive, especially in cases where degenerative change clouds the picture.
One sentence captures the core finding: CT was the most dependable imaging test in this study.
Reader Agreement Was Highest With CT and Lowest With X-ray
Diagnostic accuracy is only part of the story. A test that can be interpreted differently by different readers loses value at the bedside and in routine practice. Inter-rater reliability was highest for CT, with a kappa of 0.875. MRI followed at 0.665, while XR lagged at 0.517. The XR plus MRI combination reached 0.662, close to MRI alone, and CT plus MRI improved to 0.732.
The pattern is clinically familiar. Two-dimensional radiographs demand inference from limited anatomical information, and sacroiliac joints are especially difficult to assess on plain film because overlap, projection, and subtlety of change all interfere with reading consistency. CT provides a more anatomically faithful view, which likely explains why experts agreed more often when reviewing it. Higher agreement matters because diagnostic pathways become more reproducible when the scan itself offers less ambiguity.
This is not a minor technical issue. If two expert readers do not arrive at the same conclusion from the same XR, the value of the test for real-world diagnostic triage drops quickly. CT reduced that uncertainty.
What This Means for Diagnostic Pathways in Suspected axSpA
For patients with suspected axSpA, the findings support a move away from routine reliance on radiography when a more accurate cross-sectional test is available. MRI remains valuable, especially when active inflammation is the question and radiation exposure is a concern. But when the clinical challenge is separating axSpA from degenerative disease at the sacroiliac joints, CT offered the strongest overall performance in this study.
The results also strengthen the idea that structural lesions deserve more weight in the imaging assessment of axSpA than plain radiographs can reliably provide. Radiography may still appear early in some diagnostic algorithms because of accessibility, but this comparison shows its limits plainly. It missed more disease, misclassified more non-disease, and produced the weakest reader agreement.
Cross-sectional imaging changed the equation. CT was the most accurate and the most reproducible, while MRI remained clearly superior to X-ray and clinically useful on its own. The main unanswered issue is not whether XR is weaker; it clearly is. The more relevant question for practice is how often CT should be used as the preferred structural test when sacroiliac disease remains uncertain after clinical assessment and MRI is inconclusive or unavailable.