Article
Use of magnetic resonance imaging of the pelvis to describe anatomical potentially specific inflammatory changes for polymyalgia rheumatica
Search Medline for
Authors
-
Martin Fruth
- Radiologie Herne, Herne
-
Marina Franklin
- Rheumazentrum Ruhrgebiet, Herne
-
Björn Bühring
- Rheumazentrum Ruhrgebiet, Herne
-
Xenofon Baraliakos
- Rheumazentrum Ruhrgebiet, Herne
-
Jürgen Braun
- Rheumazentrum Ruhrgebiet, Herne
| Published: | February 5, 2019 |
|---|
Outline
Text
Background: Pelvic girdle pain is a common clinical symptom of patients with polymyalgia rheumatica (PMR). It also occurs in patients with rheumatoid arthritis (RA). The origin of this characteristic pain is not really clear, even though some imaging findings have been reported. However, this has neither changed pathophysiologic thinking nor clinical practice related to diagnosis. Here we describe pelvic structures in PMR patients in detail which show signs of inflammation by magnetic resonance imaging (MRI) in order to find a disease specific pattern.
Methods: In a retrospective study we used MRIs of patients who had presented with clinical symptoms suggestive of PMR in our center between 2015 and 2017. Only patients with complete MRI examinations, including contrast enhanced scans in coronal and transversal planes were included to be carefully examined by an experienced musculoskeletal radiologist (MF). After having first described all findings in much detail, we developed a preliminary semiquantitative scoring system that assesses a total of 12 sites which appeared to be frequently involved.
Results: The MRI signal of interest for PMR is a post-gadolinium signal in T1 weighted pelvic images. The most frequently involved anatomic sites were: the hamstring tendon and the M.gluteus medius and minimus tendon near the greater trochanter in all cases, which were found to be bilaterally involved as was the proximal M.rectus femoris in all cases, and the insertion of the adductor muscles, especially the M.adductor longus at the inferomedial pubic symphysis in 90% of cases. Other sites were also, but less frequently, involved. We think that the involvement of 4 sites including either the M.rectus femoris or the M.adductor longus is rather specific for PMR. There was no major difference between patients with and without RA.
Conclusion: This study suggests that there may be a MRI pattern specific for PMR. The target structure of the characteristically inflamed anatomic site seems to be the paratenon which implies that the pattern observed in PMR differs from the enthesitis seen in patients with spondyloarthritis. Prospective randomised trials are needed to further test and prove the clinical usefulness of this approach.
