Artikel
The impact of MRI slice thickness on the detection of spinal syndesmophytes in axialspondyloarthritis
Suche in Medline nach
Autoren
-
Kalliopi Klavdianou
- Rheumazentrum Ruhrgebiet Herne, Herne; Asklepieion' General Hospital, Department of Rheumatology, Voula, Athens
-
Styliani Tsiami
- Rheumazentrum Ruhrgebiet Herne, Herne; Ruhr Universität Bochum, Bochum
-
Alexaner Dieter Mewes
- University Düsseldorf, Medical Faculty, Department of Diagnostic and Interventional Radiology, Düsseldorf
-
Philipp Sewerin
- Rheumazentrum Ruhrgebiet Herne, Herne; Ruhr Universität Bochum, Bochum
-
Xenofon Baraliakos
- Rheumazentrum Ruhrgebiet Herne, Herne; Ruhr Universität Bochum, Bochum
| Veröffentlicht: | 31. August 2022 |
|---|
Gliederung
Text
Introduction: Conventional radiographs (CR) are the gold standard for detecting syndesmophytes in axial Spondyloarthritis (axSpA), mainly because magnetic resonance imaging (MRI) is not able to detect thin bony structures due to its slicing technique. We assessed the ability and performance of detection of syndesmophytes using different slice thicknesses on MRI and compare them with CR.
Methods: MRI with T1W and STIR images with slice thicknesses of 1–6 mm and CR of the lower thoracic and lumbar spine performed on the same day were prospectively performed in 43 axSpA patients. Each vertebral corner (VC) (anterior superior/posterior, inferior anterior/posterior) from the thoracic (Th11) to the lumbar (L5) vertebral body were assessed for the presence/absence of syndesmophytes but also fat lesion (FL) on MRI by two experienced readers .
Results: A total of 1.204 VCs was assessed from all patients. The mean time needed for scoring the MRIs varied between 2:27 and 4:12min for the MRI thicknesses of 1-6mm, respectively. Syndesmophytes were detected in 19.3% VCs on CR and in 38.3%, 37.5%, 34.8%, 33.7%, 31.4%, 28.7% VCs on MRI slice thicknesses of 1–6 mm, respectively (p≤0.001 for all MRI vs. CR evaluations and within MRI slice thicknesses). The anterior superior VC of L1 was the most affected site among all VCs in all MRI slice thicknesses and CRs. In thoracic spine, the anterior superior corner of T12 was the most frequently affected site in both MRI and CR. Although MRI could detect more syndesmophytes than CR, MRI at any slice thickness could not detect 15.4%–23.2% of syndesmophytes detected in CR (Table 1 [Tab. 1]). FLs were detected in 38.3%, 37.5%, 34.8%, 33.7%, 31.4%, 28.7% of VC with MRI slice thicknesses of 1–6 mm, respectively. MRI slice thickness had no role in detecting FLs (p>0.05 within MRI slice thicknesses).
Conclusion: MRI at any slice thickness detected more syndesmophytes than CR, but the best agreement and least false-positive findings on MRI based on CR as gold standard was found in the thicker slice thicknesses. The presence of FLs did not influence syndesmophyte detection on MRI. These results may influence the performance of MRI in identification of SpAspecific spinal lesions in daily practice.
