British Journal of Radiology 75 (2002),474-477 © 2002 The British Institute of Radiology
Contrast enhancement of cerebrospinal fluid on delayed MRI in a patient with an epidural abscess and renal failure
P Zatman, MB BSc, MRCS
and
M D Hourihan, MB BCh, FRCR
Department of Diagnostic Radiology, University Hospital of Wales, Heath Park, Cardiff CF14 4XW, UK
Correspondence: Dr M D Hourihan
 |
Abstract
|
|---|
Cerebrospinal fluid (CSF) is normally of low signal intensity on T1 weighted MRI. The case of a patient with renal failure who developed an epidural abscess, and in whom the CSF appeared of high signal intensity on delayed T1 weighted MRI is presented. Possible mechanisms for this are discussed.
|
Introduction
|
|---|
Intravenous contrast medium does not normally cross the blood–brain barrier (BBB) or enter the cerebrospinal fluid (CSF). It is almost exclusively excreted by the kidneys in patients with normal renal function. When renal function is impaired, the plasma half-life is increased. When a pathological process such as infection breaks down the BBB, there may be leakage of contrast into the CSF, however, this is rarely seen. We present the case of a patient who has renal failure and developed an epidural abscess due to Staphylococcus aureus infection, in whom the CSF appeared of high signal intensity on delayed T1 weighted imaging.
|
Case report
|
|---|
A 69-year-old female was admitted to hospital with a history of pain in her lower back and perineum for 8 days. She had undergone panproctocolectomy and ileostomy for Crohn's disease 35 years previously, and both a cadaveric renal transplant and a right total hip replacement 4 years before presentation. On admission the patient had a temperature of 37.7°C, a pulse of 110 per minute and a blood pressure of 90/47. Examination of the perineum revealed a large cavity from a ruptured abscess. Neurological examination was normal. The admission white blood count was elevated at 19.1 g l-1 (normal range 4–10.5 g l-1). Both urea and creatinine were elevated to 19.6 mmol l-1 (normal range 2.5–7.5 mmol l-1) and 273 µmol l-1 (normal range 70–120 µmol l-1), respectively, with a creatinine clearance of 15 ml min-1. Blood culture from admission grew Staphylococcus aureus and the patient was treated with iv cefuroxime and metronidazole.
The day after admission, the patient complained of increasing lower back pain and pain in the right hip radiating down the right thigh. Investigations, including plain radiography and a bone scan, suggested possible loosening of the acetabular component of the total hip replacement. She was treated with bed rest but complained of such increasing lower back pain, which radiated down the right thigh, that the acute pain team felt the use of morphine delivered by a patient controlled analgesia syringe driver was necessary.
10 days after admission, repeat neurological examination revealed loss of the right knee jerk reflex.
MRI of the lumbar spine was initially performed on a GE Signa 1.5 Tesla scanner (General Electric Medical Systems, Milwaukee, WI) using a phased array spine coil. T1 weighted (TIW) spin echo (500 ms repetition time, 14 ms echo time, 32 cm x 32 cm field of view (FOV), 512 x 256 matrix, number of excitations 2, 4 mm slice thickness) images were acquired pre- and post-iv gadopentetate dimeglumine (Gd-DTPA) (Magnevist; Schering, Bedin, Germany) at a dose of 0.1 mmol kg-1. An epidural abscess was demonstrated with focal collections anterior to the theca at L5/S1, posterior to the theca at L3/4 and anterior to the theca at T12 and L1 (Figure 1
). There was also enhancement of the dura up to the level of T10. Owing to increasing pain, the patient could not cooperate for further imaging at this time. As it was not clear if the upper level of the abscess had been demonstrated, a second MR scan of the thoracic spine was performed 6 h later, prior to the patient going to theatre. The delayed enhancement of the spinal CSF was identified as high signal intensity when the second T1W MR scan (Figure 2) was performed using the same parameters as the first, apart from a FOV of 48 cm x 48 cm. No other focal collection was identified and the upper level of the dural enhancement was confirmed to be at T10.
View larger version (155K):
[in this window]
[in a new window]
|
Figure 1. T1 weighted MR images pre-contrast (A,B,C) demonstrating normal signal from the cerebrospinal fluid, and post-contrast (D,E,F) demonstrating the epidural abscess (arrows) compressing the thecal sac.
|
|
View larger version (123K):
[in this window]
[in a new window]
|
Figure 2. T1 weighted MR images of the thoracic spinal canal demonstrating high signal from the cerebrospinal fluid on the delayed post-contrast scan.
|
|
Following the second MR scan the patient went to theatre and underwent a laminectomy from L3 to L5, and evacuation of pus from the epidural abscess. There was no subsequent deterioration in her renal function following either the MR scan or the surgical procedure. Post-operatively her lower back pain settled and she made a slow recovery, gradually improving her mobility.
|
Discussion
|
|---|
Spinal epidural abscess is a rare infection that can evolve to cause severe permanent neurologic impairment, or death if the diagnosis is delayed. Immunocompromised patients at high risk include those with diabetes, iv drug abuse, chronic renal failure, alcoholism or liver disease. The variable symptoms of epidural abscess may include fever, lower back pain, local spinal tenderness and neurological deficit. Previous studies have shown that infection at sites other than the spinal canal are common, and that Staphylococcus aureus is the usual infection organism [1].
In this case there was a delay of 10 days between presentation and demonstration of the abscess by MRI. MRI is the optimal imaging technique for the reliable diagnosis of epidural abscess and should be performed early if there is clinical suspicion of epidural abscess [1, 2].
Gadolinium agents are extracellular contrast media used in MRI and administered intravenously. The kidneys excrete iv Gd-DTPA almost exclusively and it has a normal plasma half-life of 9 h. Glomerular filtration is the main pathway of elimination. In renal failure the clearance of Gd-DTPA from the plasma is delayed with a prolongation of plasma half-life proportional to the degree of renal impairment. In severe renal failure (creatinine clearance <20 ml min-1), as in this case, the half-life may be up to 30 h. Gd-DTPA can be removed by extracorporeal dialysis, although no further impairment in renal function or other adverse effects attributable to administration of gadolinium have so far been observed [3].
Contrast enhancement of CSF is rarely seen after iv Gd-DTPA and it has not been reported in the normal population. Contrast enhancement of CSF on MRI following iv Gd-DTPA has been observed in animal experiments [4, 5]. There have been reports of Gd-DTPA enhancement of CSF in neoplastic meningeal infiltration [6, 7], meningeal fibrosis, cryptococcal infection [8] and neurosyphilis, each of which may represent a breakdown of the blood–brain barrier (BBB).
We have found no report of contrast enhancement of CSF on MRI following iv Gd-DTPA in patients with epidural abscess due to Staphylococcus aureus infection. Hagino et al [9], using ureter ligated rats as an experimental renal failure model, showed that iomeprol was maintained in high concentrations in the plasma secondary to renal impairment and this led to an increase in the CSF distribution of contrast. It was suggested that this might involve a non-specific process, such as passive diffusion [9]. CSF enhancement can be differentiated from meningeal contrast enhancement by observing the patterns of enhancement in each compartment [10]. In this case the meninges gave a higher signal than the CSF on initial images following injection, whereas on delayed images the signal from the CSF was increased whilst that from the meninges was unchanged.
The mechanism of CSF enhancement has not been established. Enhancement may be owing to passive diffusion into the CSF of contrast that has been retained in the plasma because of renal failure. However, it has also been suggested that mechanisms of BBB breakdown may also be involved in infective and malignant processes [10]. In metastatic disease of the meninges, new fenestrated capillaries develop without a blood–CSF barrier. These allow plasma proteins to leak into the CSF [5]. Severe infection may disrupt the endothelial tight junctions of the capillaries thereby allowing both plasma proteins and contrast medium to leak. Increase in the signal of CSF is delayed in onset and then progressively increases. In spirochaetal meningitis, this was seen 30 min after iv Gd-DTPA [10].
In the case reported, high signal intensity in the CSF was seen 6 h following iv Gd-DTPA. We have assumed that this was owing to contrast medium entering the CSF and causing T1 shortening. There are other causes for T1 shortening including altered blood and high protein content. However, the CSF signal intensity was normal 6 h previously and there was no interim event to suggest haemorrhage or other pathologies between the scans. We remain uncertain as to the reason for contrast medium entering the CSF but the possibilities considered were passive diffusion from the plasma in a patient with renal failure, which delayed the clearance of the Gd-DTPA, a breakdown of the BBB owing to the abscess, or a combination of renal failure and infection.
|
Acknowledgments
|
|---|
We wish to thank Dr K Baboolal (Consultant Physician) and Mr J Vafidis (Consultant Neurosurgeon) for allowing us to report their patient, Mr P A Braithwaite for his editorial help and Mrs K Teagle for her secretarial support.
Received for publication July 23, 2001.
Revision received November 30, 2001.
Accepted for publication December 6, 2001.
|
References
|
|---|
-
Teman AJ. Spinal epidural abscess. Early detection with gadolinium magnetic resonance imaging. Arch Neurol 1992;49:743–6.[Abstract/Free Full Text]
-
Obrador GT, Levenson DJ. Spinal epidural abscess in haemodialysis patients: report of three cases and review of the literature. Am J Kidney Dis 1996;27:75–83.[Medline]
-
Arsenault TM, King BF, Marsh JW, et al. Systemic gadolinium toxicity in patients with renal insufficiency and renal failure: retrospective analysis of an initial experience. Mayo Clin Proc 1996;71:1150–4.[Abstract]
-
Knutson R, Porier V, Gersovich E, Brock JM, Buonocore M. The effect of intravenous gadolinum on the magnetic resonance appearance of cerebrospinal fluid. Invest Radiol 1991;26:671–3.[Medline]
-
Frank JA, Girton M, Dwyer AJ, Wright DC, Cohen PJ, Doppman JL. Meningeal carcinomatosis in the VX2 rabbit tumour model: detection with Gd-DTPA enhanced MR imaging. Radiology 1988;167:825–9.[Abstract/Free Full Text]
-
Berger B, Ortiz O, Gold A, Hilal SK. Total cerebrospinal fluid enhancement following intravenous Gd-DTPA administration in a case of meningiomatosis. AJNR 1992;13:15–8.[Medline]
-
Pui MHW, Langston JW, Arai Y. Gd-DTPA enhancement of CSF in meningeal carcinomatosis. J Comput Assist Tomogr 1993;17:940–4.[Medline]
-
Sakamoto S, Kitagaki H, Ishii K, Yamaji S, Ikegiri Y, Mori E. Gadolinium enhancement of the cerebrospinal fluid in a patient with meningeal fibrosis and cryptococcal infection. Neuroradiology 1997;39:504–5.[Medline]
-
Hagino M, Yamamoto K, Ohashi K, et al. Effect of acute renal failure on the distribution of iomeprol, a nonionic contrast agent, to cerebrospinal fluid in rats. Biol Pharm Bull 1997;20:939–42.[Medline]
-
Good CD, Jäger HR. Contrast enhancement of the cerebrospinal fluid on MRI in two cases of spirochaetal meningitis. Neuroradiology 2000;242:448–50.
Top
Abstract
Introduction
