MAGNETIC-RESONANCE-IMAGING IN A SPINAL ABSCESS MODEL - PRELIMINARY-REPORT

Citation
Vm. Runge et al., MAGNETIC-RESONANCE-IMAGING IN A SPINAL ABSCESS MODEL - PRELIMINARY-REPORT, Investigative radiology, 33(4), 1998, pp. 246-255
Citations number
20
Categorie Soggetti
Radiology,Nuclear Medicine & Medical Imaging
Journal title
ISSN journal
00209996
Volume
33
Issue
4
Year of publication
1998
Pages
246 - 255
Database
ISI
SICI code
0020-9996(1998)33:4<246:MIASAM>2.0.ZU;2-L
Abstract
RATIONALE AND OBJECTIVES. Magnetic resonance (MR) scan technique and l esion detectability were evaluated using a newly developed spinal absc ess model in the New Zealand White rabbit. METHODS. TO create the lesi on, an epidural needle was inserted under fluoroscopic guidance in the lumbar region and advanced to penetrate the ligamentum flavum. Next, polyethylene tubing was fed through the needle into the epidural space . A mixed suspension of Staphylococcus aureus (Cowan I) and blue polys tyrene microspheres then was injected. Lesions were evaluated by MR im aging in four animals at multiple time points (3, 6, and 9 days), Imag ing was performed at 1.5 tesla using a surface coil. Precontrast T2- a nd T1-weighted scans were first obtained. The T1-weighted scans were a cquired both with and without fat saturation, and were repeated after intravenous contrast administration, The contrast agent used was gadot eridol (gadolinium HP-DO3A or ProHance(R)) at a dose of 0.3 mmol/kg. R ESULTS. On prospective film review, postcontrast scans proved superior for lesion detection. A spinal abscess could be identified postcontra st in all cases, irrespective of the use of fat saturation. The next b est imaging technique for lesion detection was the T2-weighted scan, w ith 5 of 8 lesions noted thereon. Visualization of lesion margins prov ed to be a primary factor in prospective lesion identification, Region of interest image analysis demonstrated the postcontrast scans to be superior to all precontrast scan techniques for conspicuity of the int erface between the abscess and the compressed spinal cord, with these results statistically significant. The lesions were characterized hist ologically by infiltrates of heterophils into the meninges and outer s pinal cord with accompanying mild hemorrhage, fibrin exudation, and ba cterial colonies. The lesions in three animals were confirmed to be in the epidural space, with the lesion in one animal in the subdural spa ce. CONCLUSIONS. The current animal model was developed to study spine infection and, specifically, imaging characteristics and lesion detec tability on MR, With the increased use of epidural catheters for pain management and the large number of acquired immunodeficiency syndrome cases, epidural infection is becoming an increasingly important clinic al problem. Imaging technique, in particular the use of intravenous co ntrast, is critical for lesion detection and evaluation.