Premium
MRI Findings of Arachnoiditis, Revisited. Is Classification Possible?
Author(s) -
El Homsi Maria,
Gharzeddine Karem,
Cuevas Jordan,
ArevaloPerez Julio,
Rebeiz Karim,
Khoury Nabil J.,
Moukaddam Hicham
Publication year - 2021
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.27583
Subject(s) - medicine , arachnoiditis , kappa , radiology , magnetic resonance imaging , neuroradiologist , population , cohen's kappa , nuclear medicine , philosophy , linguistics , environmental health , machine learning , computer science
Background Prior imaging studies characterizing lumbar arachnoiditis have been based on small sample numbers and have reported inconsistent results. Purpose To review the different imaging patterns of lumbosacral arachnoiditis, their significance, and clinical implications. Study type Retrospective. Population A total of 96 patients (43 women; average age 61.3 years) with imaging findings of arachnoiditis (postsurgical: N = 49; degenerative: N = 29; vertebral fracture: N = 6; epidural and subdural hemorrhage: N = 3, infectious: N = 1; other: N = 8) from January 2009 to April 2018. Field strength/Sequence Sagittal and axial T2‐weighted Turbo Spin Echo at 1.5 T and 3 T. Assessment Chart review was performed to assess the cause of arachnoiditis, and imaging was reviewed by two musculoskeletal and three neurology radiologists, blinded to the clinical data and to each other's imaging interpretation. Previous classification included a three‐group system based on the appearance of the nerve roots on T2‐weighted images. A fourth group was added in our review as “nonspecified” and was proposed for indeterminate imaging findings that did not fall into the classical groups. The presence/absence of synechiae/fibrous bands that distort the nerve roots and of spinal canal stenosis was also assessed. Statistical tests The kappa score was used to assess agreement between readers for both classification type and presence/absence of synechiae. Results Postsurgical (51%) and degenerative changes (30%) were the most common etiologies. About 7%–55% of arachnoiditis were classified as group 4. There was very poor classification agreement between readers (kappa score 0.051). There was also poor interreader agreement for determining the presence of synechiae (kappa 0.18) with, however, strong interreader agreement for the presence of synechia obtained between the most experienced readers (kappa 0.89). Data Conclusion This study demonstrated the lack of consensus and clarity in the classification system of lumbar arachnoiditis. The presence of synechia has high interreader agreement only among most experienced readers and promises to be a useful tool in assessing arachnoiditis. Evidence Level 3 Technical Efficacy Stage 2