- Email: [email protected]
MRI findings in spinal subdural and epidural hematomas
European Journal of Radiology 64 (2007) 119–125
Review
MRI findings in spinal subdural and epidural hematomas Petra Braun a,∗ , Khuram Kazmi b , Pablo Nogu´es-Mel´endez c , Fernando Mas-Estell´es c , Fernando Aparici-Robles c a b
Department of Radiology, Hospital La Plana, Ctra. De Vila-real a Borriana km. 0,5, 12540 Vila-real (Castell´o), Spain Department of Radiology, Penn State Milton S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033, USA c Department of Radiology, La Fe Hospital, Avenida Campanar, 21, 46009 Valencia, Spain Received 20 July 2006; received in revised form 6 February 2007; accepted 13 February 2007
Abstract Background: Spinal hematomas are rare entities that can be the cause of an acute spinal cord compression syndrome. Therefore, an early diagnosis is of great importance. Patients and Methods: From 2001 to 2005 seven patients with intense back pain and/or acute progressive neurological deficit were studied via 1.5 T MRI (in axial and sagittal T1- and T2-weighted sequences). Follow-up MRI was obtained in six patients. Results: Four patients showed the MRI features of a hyperacute spinal hematoma (two spinal subdural hematoma [SSH] and two spinal epidural hematoma [SEH]), isointense to the spinal cord on T1- and hyperintense on T2-weighted sequences. One patient had an early subacute SEH manifest as heterogeneous signal intensity with areas of high signal intensity on T1- and T2-weighted images. Another patient had a late subacute SSH with high signal intensity on T1- and T2-weighted sequences. The final patient had a SEH in the late chronic phase being hypointense on T1and T2-weighted sequences. Discussion: MRI is valuable in diagnosing the presence, location and extent of spinal hematomas. Hyperacute spinal hematoma and the differentiation between SSH and SEH are particular diagnostic challenges. In addition, MRI is an important tool in the follow-up in patients with conservative treatment. © 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Spinal subdural hematoma; Spinal epidural hematoma; Hyperacute spinal hematoma; Subacute spinal hematoma; Chronic spinal hematoma; MRI
Contents 1. 2. 3. 4. 5.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patients and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Signs of acute or progressive, painful paraparesis and/or paraparesis with sphincter dysfunction are well known indications for emergency MRI. Inflammatory [1,2], infectious [3,4]
∗
Corresponding author at: C/Garb´ı 11, 46116 Masias (Moncada), Spain. Tel.: +34 679288874. E-mail address: [email protected] (P. Braun). 0720-048X/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2007.02.014
119 120 120 121 124 124
and metastatic lesions, as well as intraspinal hematomas are diagnostic considerations [5]. Intraspinal hematomas can be post-traumatic [6–8], post-procedural [9–11], due to anticoagulant therapy [12–14] or even spontaneous [13,15–19]. Magnetic resonance imaging (MRI) is considered to be the technique of choice for diagnosis [20,21]. Even though MRI is able to evaluate the location, extent, chemical state and compressive effects of hemorrhage, the diagnosis of hyperacute intraspinal hemorrhage may be difficult due to its imaging characteristics.
120
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
Table 1 Overview of the patient demographics, clinical symptoms, hematoma etiology, location and extent Case #
Age/sex
Initial symptoms
Etiology
Location
Extent (vertebrae)
1 2 3 4 5 6 7
87/F 76/F 72/F 81/F 22/M 72/F 43/M
Paraplegia and intense back pain Thoracic pain with irradiation into arms Cervical pain, tetraparesis Thoracic pain, left hemiparesis Lumbago Intrascapular pain, flaccid tetraparesis Chronic cauda equina compression
Anticoagulation Spontaneous Spontaneous Anticoagulation Spontaneous Spontaneous Epidural catheter
Thoracic Cervicothoracic Cervical to lumbar Cervicothoracic Thoracolumbar Cervicothoracic Thoracolumbar
4 13 23 7 4 8 2
The purpose of this study is to review the literature and describe our experience in seven patients with spinal subdural and epidural hematoma.
- late subacute phase—MRI performed from 1 to 2 weeks; - early chronic phase—MRI performed from weeks to months; - late chronic phase—MRI performed from months to years.
2. Patients and methods
3. Results
A retrospective study of clinical and MRI findings in patients with SSH or SEH was carried out from 2001 to 2005. Seven patients were identified who had spinal hematomas on MRI. Six of them were imaged between 10 h and 3 weeks after onset of symptoms. In one patient, the onset could not be clearly identified. Imaging was performed on 1.5 T scanner (Signa CVi GE Medical Systems, Milwaukee, WI; Siemens Symphony, Erlangen, Germany) and consisted of sagittal and axial T1-weighted (450–600/10–20 [TR range/TE range]), sagittal and axial T2weighted (2800–4200/70–110 [TR range/TE range]), axial gradient-echo (500–600/15–30 [TR range/TE range]; flip angle 20–30◦ ) and short inversion-time inversion recovery (STIR) images (3500/35/140 [TR/TE/TI]). In addition, T1-weighted images with fat suppression were obtained. Contrast enhanced T1-weighted images were acquired in three patients (cases 1, 4 and 5). A follow up MRI was performed between 2 weeks and 2 months after the first MRI in all patients. Spinal hematomas were classified as subdural if epidural fatty tissue without displacement of the dura mater was seen. If dural displacement towards the spinal cord was observed, the hematoma was classified as epidural. Additionally, spinal hematomas were classified depending on their symptom onset in relation to the MRI study as follows:
Three patients had a SSH. All patients were female, between 72 and 87 years. Initial symptoms, etiology, location and extent are summarized in Table 1. MRI in the hyperacute phase demonstrated an isointense lesion on T1-weighted and hyperintense lesion on T2-weighted images (Fig. 1). MRI in the late subacute phase showed a hyperintense lesion on T1- and T2-weighted images (Fig. 2) (Table 2). Spinal angiography was carried out in two patients and ruled out vascular malformations. Surgery was performed only in one case with evacuation of the SSH resulting in partial recovery of neurologic function. The other two patients were treated conservatively and showed a complete recovery. In all cases, follow up MRI several weeks later demonstrated a complete resorption of the SSH (Table 2). Four patients had a SEH. Three out of four were male, between 22 and 81 years. Initial symptoms, etiology, location and extent are summarized in Table 1. MRI in the hyperacute phase demonstrated an isointense lesion on T1-weighted and hyperintense lesion on T2-weighted images (Fig. 3). MRI in the early subacute phase showed a heterogeneous lesion with areas of high signal intensity on T1- and T2-weighted images (Fig. 4). MRI in the chronic phase demonstrated a hypointense lesion on T1- and T2-weighted (Fig. 5) (Table 2). All patients were treated conservatively, one showed a partial recovery and two a complete recovery with normal MRI about 1 month later (Table 2).
- hyperacute phase—MRI performed <12 h after onset; - acute phase—MRI performed between 12 h and 3 days; - early subacute phase—MRI performed from 3 to 7 days;
Table 2 Description of the spinal hematoma: sub- or epidural location, MRI features, time from symptom onset until MRI, treatment, evolution and MRI follow-up Case #
Space
Symptoms-MRI
T1 signal
T2 signal
Surgery
Evolution
Follow-up MRI
1 2 3 4 5 6 7
Subdural Subdural Subdural Epidural Epidural Epidural Epidural
<12 h <12 h 2 weeks <12 h <12 h 7 days Months
Isointense Isointense Hyperintense Isointense Isointense Heterogenous Hypointense
Hyperintense Hyperintense Hyperintense Hyperintense Hyperintense Heterogenous Hypointense
Yes No No No No No No
Partial recovery Complete recovery Complete recovery Partial recovery Complete recovery Complete recovery Partial recovery
Normal Normal Normal Normal Normal Normal Unchanged
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
121
have escaped medical attention [22]. In a large metaanalysis of 613 patients with spinal hematoma, the mean age at presentation was 63 years [23] with a male predominance (ratio, male:female, 1.7:1). SSH and SEH normally present with signs of spinal cord or cauda equina compression, often with an acute onset preceded by back pain or radicular pain [24–26]. Isolated pain is less common [12]. It is difficult to differentiate between SSH and SEH only by clinical history and examination, but SSH seems to have a more rapid evolution [27]. Prognosis for neurologic recovery depends on the patient’s preoperative neurologic status and duration of neurologic dysfunction [28]. Computed tomography, which is usually readily available in the acute setting, is helpful in differentiating the often hyperdense, frequently crescent-shaped hyperacute hematoma from the adjacent fat of low density and osseous structures of high density [29]. However, the subacute hematoma becomes isodense and clearly difficult to identify. Like CT imaging, MRI provides information on the localization and shape of the hematoma in the axial plane. However, the craniocaudal extent in the sagittal plane is better evaluated by MRI. Additionally, MRI more readily identifies other common causes of spinal cord dysfunction. Therefore, MRI is considered the preferred imaging modality to evaluate spinal hematomas. Spinal angiography is especially useful in the detection of vascular malformation and should be performed in cases of diagnostic suspicion on MRI images. The evolution of the MRI findings for spinal hematomas is in keeping with that described for brain hematomas [30,31] (Table 3).
Fig. 1. MRI of an 87-year-old patient presenting a hyperacute spinal subdural hematoma from T6 to T9 due to anticoagulant therapy. (a) Sagittal T2-weighted image (TR 3320, TE 117.2) of the thoracic spine demonstrates a hyperintense collection in the posterior spinal subdural space (long arrow). It is anterior to the posterior epidural fat (short arrow), displacing and compressing the spinal cord anteriorly. (b) Axial T1-weighted image (TR 440, TE 9.2) after intravenous administration of contrast material demonstrates nodular enhancement centrally in the posterior spinal subdural hematoma (arrow).
4. Discussion The recent increase of publications of SSH and SEH may be explained by the introduction of MRI in daily medical practice. As a result, more patients with a mild clinical course caused by SSH and SEH are diagnosed. Previously those patients would
1. The hyperacute hemorrhage is iso- to hypointense on T1weighted images and hyperintense on T2-weighted images due to the fresh oxygenated arterial blood containing intracellular oxyhemoglobin. A peripheral rim of hypointensity might be seen due to the degraded blood products. 2. The acute hematoma is slightly hypo-/isointense on T1weighted imaged (as the paramagnetic deoxyhemoglobin within clotted intact hypoxic red blood cells does not cause T1-shortening) and hypointense on T2-weighted images (as progressive concentration of red blood cells, blood clot retraction and fibrin production shortens T2). 3. The early subacute hematoma is very hyperintense on T1weighted images (as the oxidation of deoxyhemoglobin to methemoglobin results in marked shortening of T1) and hypointense on T2-weighted images (due to intracellular methemoglobin causing T2-shortening). Table 3 MR appearance of intraspinal hematoma Phase
Age
Hemoglobin
T1-T2
Hyperacute Acute Early subacute
<12 1–3 days 3–7 days
Iso-hyper Hypo-hypo Hyper-hypo
Late subacute
1–2 weeks
Chronic
>2 weeks
Oxyhemoglobin Deoxyhemoglobin Intracellular methemoglobin Extracellular methemoglobin Hemosiderin
Hyper-hyper Hypo-hypo
122
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
Fig. 2. MRI of a 72-year-old patient presenting a spontaneous late subacute spinal subdural hematoma from clivus to L4. (a) Sagittal T2-weighted (A) and T1weighted fat suppressed images (B and C) (TR 2640, TE 109.1 (A); TR 440, TE 8.1 (B); TR 320, TE 9.0 (C)) of the cervical (A and B) and lumbar spine (C) demonstrate a hyperintense collection in the anterior spinal subdural space of the cervical spine and posterior spinal subdural space of the lumbar spine (arrows) that cannot be distinguished from the CSF on the T2-weighted sequence. (b) Axial T1-weighted fat suppressed image (TR 540, TE 9.2) of the thoracolumbar spine shows a hyperintense collection in the posterior spinal subdural space (arrow) without cord compression. The asterisk points out the posterior epidural fat.
4. The late subacute hematoma is hyperintense on T1- and T2weighted images due to extracellular methemoglobin after lysis of red blood cells. 5. The chronic hematoma is mainly hypointense on T1- and T2weighted images because of the paramagnetic hemosiderin and ferritin within macrophages. While most articles about SSH described findings in the acute, subacute and chronic phases [29], the experience with hyperacute SSH and SEH is sparse. Pedraza et al. [11] described the MRI findings of two patients in the hyperacute phase. The SSHs were isointense on T1-weighted images and showed high signal intensity on T2-weighted images. Additionally, Pedraza et al. [11] described the presence of a low-signal rim to the lesion on T2-weighted images. In the hyperacute phase it is difficult to differentiate between the lesion and the adjacent structures due to their isointensity on T1-weighted images. In all four of our patients (Figs. 1 and 3), the MRI findings in the hyperacute phase were concordant with the findings described by Pedraza et al. [11]. In one patient, we also found a low-signal rim around the lesion on gradient-echo imaging (Patient 2). Felber et al. [32] proposes performing gradient-echo images to help characterize the hematoma in the hyperacute phase. In our opinion, the best sequence protocol to evaluate hyperacute spinal hematoma are T1-weighted images in sagittal and axial planes, T2-weighted images in sagittal planes and T2-weighted gradient-echo images in axial planes. Another diagnostic difficulty exists in imaging the T1 hyperintense subacute SSH. In one of our patients studied 2 weeks after symptom onset, we performed a T1-weighted fat suppressed sequence that clearly differentiated the hyperintense subacute subdural hematoma from epidural fat (Fig. 2). Thus, in patients studied in the subacute phase, we recommend the performance of T1-weighted images in sagittal planes, T1-weighted fat
suppressed images in sagittal and axial planes and T2-weighted images in sagittal and/or axial planes. In the early chronic phase (weeks–months), more diffuse and lightly inhomogeneous signal intensities are seen within the hematoma on T1- and T2-weighted images [7,10]. In the late chronic phase (months–years), MRI demonstrates low-signal intensity on T1- and T2-weighted images, as is seen in one of our patients (Fig. 5). After contrast material administration peripheral enhancement was observed in two patients with SEH (Fig. 3). Thus far, only a few cases with contrast enhancement have been described [33–36]. Fukui et al. [35] described similar findings in four of five patients. This peripheral enhancement is thought to be due to hyperemia of the dura resulting in thickening of adjacent meninges [34]. Nodular enhancement was seen in one patient with SSH after contrast material administration (Fig. 1). Some authors [35,37] postulate that the central pattern of enhancement in spinal hematomas may result from extravasation of contrast material by leaking vessels. It is important to be aware of this pattern of contrast enhancement as a similar appearance can be seen with tumor, aneurysm and vascular malformation [35]. The necessity of a surgical intervention versus conservative management has been discussed in the literature. Conservative management is preferred in cases with mild neurological deficits [15,38], progressive improvement in the early period [7,8,10,39–42] or if coagulopathy is present [29]. In patients in whom conservative management is chosen, evolution of spinal hematomas can be monitored using serial MRI [10,43]. In our study, one patient underwent emergency surgery for a SSH in the hyperacute phase. The other six patients with SSH and SEH were treated conservatively. Follow-up MRI demonstrated complete resorption of the hematoma in five patients, all of whom had clinical improvement. One patient had a chronic hematoma which did not change on follow-up MRI. In our opinion, MRI
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
Fig. 3. MRI of a 22-year-old patient presenting with a spontaneous hyperacute spinal epidural hematoma from T11 to L2. (a) Sagittal T2-weighted image (TR 3320, TE 117.2) demonstrates a hyperintense, slightly inhomogenous collection in the anterior spinal epidural space (asterisk) with posterior displacement of the dura (arrow) and slight compression of the spinal cord. (b) Axial T1-weighted image (TR 560, TE 9.5) after intravenous administration of contrast material demonstrates an isointense (to the spinal cord) collection in the anterior spinal epidural space (arrow) with peripheral enhancement.
123
Fig. 4. MRI of a 72-year-old patient presenting with a spontaneous early subacute spinal epidural hematoma from C3 to T2. (a) Sagittal T2-weighted fat suppressed image (TR 4000, TE 43.4) shows a heterogeneous collection in the posterior spinal epidural space (long arrow) with anterior displacement of the dura and compression of the spinal cord. (b) Axial T1-weighted image (TR 519, TE 9.2) after intravenous administration of contrast material demonstrates a mainly hypointense collection in the posterior spinal epidural space with peripheral enhancement (arrow), displacing the spinal cord anterolaterally.
124
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
imaging characteristics. T2-weighted gradient-echo images may be helpful to demonstrate a low-signal rim. In our study, T1weighted fat suppressed images proved helpful in the subacute phase to differentiate epidural fat from hematoma. We recommend performance of sagittal planes to evaluate the full extent of the hematoma and axial planes to differentiate a SSH from a SEH by means of the epidural fat and the visualization of the dura. In patients in whom conservative management is chosen, MRI monitoring may be performed. References
Fig. 5. MRI of a 43-year-old patient presenting with a late chronic spinal epidural hematoma from T12 to L1 probably due to an epidural catheter. (a) Sagittal T1weighted image (TR 560 TE 9.5) shows a hypointense collection in the posterior epidural space (arrow) with anterior displacement of the dura. (b) Axial T2weighted image (TR 3300, TE 110.6) demonstrates a hypointense collection in the posterior spinal epidural space (arrow) displacing the cauda equina slightly anterior.
is adequate to monitor subdural hemorrhage in patients when conservative treatment is decided. 5. Conclusion MRI is regarded as the imaging modality of choice to evaluate SSH and SHE. However, there are important considerations to take into account. Spontaneous SSH is a rare lesion that is difficult to diagnose in the hyperacute phase due to the lesion’s
[1] K¨uker W, Mull M, Mayfrank L, Topper R, Thron A. Epidural spinal infection. Variability of clinical and magnetic resonance imaging findings. Spine 1997;22:544–50. [2] K¨uker W, Schaade L, Ritter K, Naciemiento W. MRI follow-up of herpes simplex virus (type 1) radiculomyelitis. Neurology 1999;52:1102–3. [3] Boukobza M, Guichard JP, Boissonet M, et al. Spinal epidural haematoma: report of 11 cases and review of the literature. Neuroradiology 1994;36:456–9. [4] Lloret-Garc´ıa J, Almenar-Garc´ıa V, Correa-Lac´arcel J, Garc´ıa de Quir´os J, Ruiz-Garc´ıa F. Hematomas extradurales espinales espont´aneos: presentaci´on de dos casos. Rev Neurol 1997;25:1936–8. [5] Sklar EML, Donovan Post JM, Falcone S. MRI of acute spinal epidural hematomas. J Comput Assist Tomogr 1999;23:238–43. [6] Hung KS, Lui CC, Wang CH, Wang CJ, Howng SL. Traumatic spinal subdural hematoma with spontaneous resolution. Spine 2002;27(24):E534–8. [7] Tillich M, Kammerhuber F, Reittner P, Leber KA, Szolar DH. Chronic spinal subdural haematoma associated with intracranial subdural haematoma: CT and MRI. Neuroradiology 1999;41:137–9. [8] Juvonen T, Tervonen O, Ukkola V, Klintrup HE. Wide-spread posttraumatic spinal subdural hematoma-imaging findings with spontaneous resolution: case report. J Trauma 1994;36:262–4. [9] Gehri R, Zanetti M, Boos N. Subacute subdural haematoma complicating lumbar microdiscectomy. J Bone Joint Surg Br 2000;82:1042–5. [10] Kulkarni AV, Willinsky RA, Gray T, Cusimano M. Serial magnetic resonance imaging findings for a spontaneously resolving spinal subdural haematoma: case report. Neurosurgery 1998;42:398–401. [11] Pedraza Guti´errez S, Coll Masfarr´e S, Casta˜no Duque CH, Suesc´un M, Rovira Ca˜nellas A. Hyperacute spinal subdural haematoma as a complication of lumbar spinal anaesthesia: MRI. Neuroradiology 1999;41:910–4. [12] Boukobza M, Haddar D, Boissonet M, Merland JJ. Spinal subdural haematoma: a study of three cases. Clin Radiol 2001;56:475–80. [13] Domenicucci M, Ramieri A, Ciappetta P, Delfini R. Nontraumatic acute subdural haematoma. Report of five cases and review of the literature. J Neurosurg (Spine 1) 1999;91:65–73. [14] Wysowski DK, Talarico L, Bacsanyi J, Botstein P. Spinal and epidural hematoma and low-molecular weight heparin. N Engl J Med 1998;338:1774–5. [15] Yamada K, Nakahara T, Yamamato K, Muranaka T, Ushio Y. Nontraumatic spinal subdural haematoma occurring in a postpartum period. Acta Neurochir (Wien) 2003;145:151–5. [16] K¨uker W, Thiex R, Friese S, et al. Spinal subdural and epidural haematomas: Diagnostic and therapeutic aspects in acute and subacute cases. Acta Neurochir (Wien) 2000;142:777–85. [17] Kirsch EC, Khangure MS, Holthouse D, McAuliffe W. Acute spontaneous spinal subdural haematoma: MRI features. Neuroradiology 2000;42:586–90. [18] Cohen JE, Ginsberg HJ, Emery D, Schwartz ML. Fatal spontaneous spinal epidural hematoma following thrombolysis for myocardial infarction. Surg Neurol 1998;49:520–3. [19] Calhoun JM, Boop F. Spontaneous spinal subdural hematoma: case report and review of the literature. Neurosurgery 1991;29:133–4. [20] Saito S, Katsube H, Kobayashi Y. Spinal epidural hematoma with spontaneous recovery demonstrated by magnetic resonance imaging. Spine 1994;19:483–6.
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125 [21] Post MJ, Becerra JL, Madsen PW, et al. Acute spinal subdural hematoma. MR and CT findings with pathological correslates. Am J Neuroradiol 1994;15:1895–905. [22] Groen RJ. Non-operative treatment of spontaneous spinal epidural hematomas: a review of the literature and a comparison with operative cases. Acta Neurochir (Wien) 2004;146(2):103–10. [23] Kreppel D, Antoniadis G, Seeling W. Spinal hematoma: a literature survey with meta-analysis of 613 patients. Neurosurg Rev 2003;26:1–49. [24] Chen CJ, Fang W, Chen CM, Wan YL. Spontaneous spinal epidural haematoma with repeated remission and relapse. Neuroradiology 1997;39:737–40. [25] Horcajadas A, Katati M, Arraey MA, et al. Hematoma epidural espinal espont´aneo: presentaci´on de 2 casos y revisi´on de la bibliograf´ıa. Neurolog´ıa 1998;13:401–4. [26] Rovira A, Griv´e E, Pedraza S, et al. Hematomas espinales epidurales agudos no traum´aticos: diagn´ostico por resonancia magn´etica. Radiolog´ıa 2000;42(4):213–20. [27] Johnston RA. The management of acute spinal cord compression. J Neurol Neurosurg Psychiatry 1993;56:1046–54. [28] Rhode V, K¨uker W, Reinges MHT, Gilsbach JM. Microsurgical treatment of spontaneous and non-spontaneous spinal epidural haematomas: neurological outcome in relation to aetiology. Acta Neurochir (Wien) 2000;142:787–93. [29] Post MJ, Becerra JL, Madsen PW, et al. Acute spinal subdural hematoma: MR and CT findings with pathologic correlates. AJNR 1994;15(10):1895–905. [30] Gomori JM, Grossman RI, Goldberg HI, Zimmerman RA, Bilaniuk LT. Intracranial hematomas: imaging by high-field MR. Radiology 1985;157:87–93. [31] D¨ahnert W. Radiology review manual. Baltimore. 3rd ed; 1996. [32] Felber S, Langmaier J, Judmaier W, et al. Magnetresonanztomographie bei epiduralen und subduralen spinalen H¨amatomen. Radiologe 1994;34:656–61.
125
[33] Caldemeyer K, Mocharla R, Moran C, Smith R. Gadolinium enhancement in the center of a spinal epidural hematoma in a hemophiliac. J Comput Assist Tomogr 1993;17:321–3. [34] Crisi G, Sorgato P, Colombo A, Scarpa M, Falasca A, Angiari P. Gadolinium-DTPA-enhanced MR imaging in the diagnosis of spinal epidural haematoma. Report of a case. Neuroradiology 1990;32:64– 6. [35] Chen C, Ro L. Central gadolinium enhancement of an acute spontaneous spinal epidural hematoma. Neuroradiology 1996;38:114–6. [36] Fukui MB, Swarnkar AS, Williams RL. Acute spontaneous spinal epidural hematomas. AJNR 1999;20(7):1365–72. [37] Chang F-C, Lirng J-F, Chen S-S, et al. Contrast enhancement patterns of acute spinal epidural hematomas: a report of two cases. AJNR 2003;24(3):366–9. [38] Torres A, Acebes JJ, Cabiol J, et al. Revisi´on de 22 casos de hematomas epidurales espinales. Factores pron´osticos y manejo terap´eutico. Neurocirug´ıa 2004;15:353–9. [39] Shimada N, Sugawara T, Itoh Y, et al. Four cases of spontaneous spinal epidural hematoma. No Shinkei Geka 2005;33(2):163–8. [40] Duffill J, Sparrow OC, Millar J, Barker CS. Can spontaneous spinal epidural haematoma be managed safely without operation? A report of four cases. J Neurol Neurosurg Psychiatry 2000;69:816–9. [41] Schwerdtfeger K, Caspar W, Alloussi S, Strowitzki M, Loew F. Acute spinal intradural extramedullary hematoma: a nonsurgical approach for spinal cord decompression. Neurosurgery 1990;27(2):312–4. [42] Hentschel SJ, Woolfenden AR, Fairholm DJ. Resolution of spontaneous spinal epidural hematoma without surgery: report of two cases. Spine 2001;26(22):E525–7. [43] Lecouvet FE, Annet L, Duprez TP, Cosnard G, Scordidis V, Malghem J. Uncommon magnetic resonance imaging observation of lumbar subdural hematoma with cranial origin. J Comput Assist Tomogr 2003;27(4): 530–3.
Review
MRI findings in spinal subdural and epidural hematomas Petra Braun a,∗ , Khuram Kazmi b , Pablo Nogu´es-Mel´endez c , Fernando Mas-Estell´es c , Fernando Aparici-Robles c a b
Department of Radiology, Hospital La Plana, Ctra. De Vila-real a Borriana km. 0,5, 12540 Vila-real (Castell´o), Spain Department of Radiology, Penn State Milton S. Hershey Medical Center, 500 University Drive, Hershey, PA 17033, USA c Department of Radiology, La Fe Hospital, Avenida Campanar, 21, 46009 Valencia, Spain Received 20 July 2006; received in revised form 6 February 2007; accepted 13 February 2007
Abstract Background: Spinal hematomas are rare entities that can be the cause of an acute spinal cord compression syndrome. Therefore, an early diagnosis is of great importance. Patients and Methods: From 2001 to 2005 seven patients with intense back pain and/or acute progressive neurological deficit were studied via 1.5 T MRI (in axial and sagittal T1- and T2-weighted sequences). Follow-up MRI was obtained in six patients. Results: Four patients showed the MRI features of a hyperacute spinal hematoma (two spinal subdural hematoma [SSH] and two spinal epidural hematoma [SEH]), isointense to the spinal cord on T1- and hyperintense on T2-weighted sequences. One patient had an early subacute SEH manifest as heterogeneous signal intensity with areas of high signal intensity on T1- and T2-weighted images. Another patient had a late subacute SSH with high signal intensity on T1- and T2-weighted sequences. The final patient had a SEH in the late chronic phase being hypointense on T1and T2-weighted sequences. Discussion: MRI is valuable in diagnosing the presence, location and extent of spinal hematomas. Hyperacute spinal hematoma and the differentiation between SSH and SEH are particular diagnostic challenges. In addition, MRI is an important tool in the follow-up in patients with conservative treatment. © 2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Spinal subdural hematoma; Spinal epidural hematoma; Hyperacute spinal hematoma; Subacute spinal hematoma; Chronic spinal hematoma; MRI
Contents 1. 2. 3. 4. 5.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Patients and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Signs of acute or progressive, painful paraparesis and/or paraparesis with sphincter dysfunction are well known indications for emergency MRI. Inflammatory [1,2], infectious [3,4]
∗
Corresponding author at: C/Garb´ı 11, 46116 Masias (Moncada), Spain. Tel.: +34 679288874. E-mail address: [email protected] (P. Braun). 0720-048X/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejrad.2007.02.014
119 120 120 121 124 124
and metastatic lesions, as well as intraspinal hematomas are diagnostic considerations [5]. Intraspinal hematomas can be post-traumatic [6–8], post-procedural [9–11], due to anticoagulant therapy [12–14] or even spontaneous [13,15–19]. Magnetic resonance imaging (MRI) is considered to be the technique of choice for diagnosis [20,21]. Even though MRI is able to evaluate the location, extent, chemical state and compressive effects of hemorrhage, the diagnosis of hyperacute intraspinal hemorrhage may be difficult due to its imaging characteristics.
120
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
Table 1 Overview of the patient demographics, clinical symptoms, hematoma etiology, location and extent Case #
Age/sex
Initial symptoms
Etiology
Location
Extent (vertebrae)
1 2 3 4 5 6 7
87/F 76/F 72/F 81/F 22/M 72/F 43/M
Paraplegia and intense back pain Thoracic pain with irradiation into arms Cervical pain, tetraparesis Thoracic pain, left hemiparesis Lumbago Intrascapular pain, flaccid tetraparesis Chronic cauda equina compression
Anticoagulation Spontaneous Spontaneous Anticoagulation Spontaneous Spontaneous Epidural catheter
Thoracic Cervicothoracic Cervical to lumbar Cervicothoracic Thoracolumbar Cervicothoracic Thoracolumbar
4 13 23 7 4 8 2
The purpose of this study is to review the literature and describe our experience in seven patients with spinal subdural and epidural hematoma.
- late subacute phase—MRI performed from 1 to 2 weeks; - early chronic phase—MRI performed from weeks to months; - late chronic phase—MRI performed from months to years.
2. Patients and methods
3. Results
A retrospective study of clinical and MRI findings in patients with SSH or SEH was carried out from 2001 to 2005. Seven patients were identified who had spinal hematomas on MRI. Six of them were imaged between 10 h and 3 weeks after onset of symptoms. In one patient, the onset could not be clearly identified. Imaging was performed on 1.5 T scanner (Signa CVi GE Medical Systems, Milwaukee, WI; Siemens Symphony, Erlangen, Germany) and consisted of sagittal and axial T1-weighted (450–600/10–20 [TR range/TE range]), sagittal and axial T2weighted (2800–4200/70–110 [TR range/TE range]), axial gradient-echo (500–600/15–30 [TR range/TE range]; flip angle 20–30◦ ) and short inversion-time inversion recovery (STIR) images (3500/35/140 [TR/TE/TI]). In addition, T1-weighted images with fat suppression were obtained. Contrast enhanced T1-weighted images were acquired in three patients (cases 1, 4 and 5). A follow up MRI was performed between 2 weeks and 2 months after the first MRI in all patients. Spinal hematomas were classified as subdural if epidural fatty tissue without displacement of the dura mater was seen. If dural displacement towards the spinal cord was observed, the hematoma was classified as epidural. Additionally, spinal hematomas were classified depending on their symptom onset in relation to the MRI study as follows:
Three patients had a SSH. All patients were female, between 72 and 87 years. Initial symptoms, etiology, location and extent are summarized in Table 1. MRI in the hyperacute phase demonstrated an isointense lesion on T1-weighted and hyperintense lesion on T2-weighted images (Fig. 1). MRI in the late subacute phase showed a hyperintense lesion on T1- and T2-weighted images (Fig. 2) (Table 2). Spinal angiography was carried out in two patients and ruled out vascular malformations. Surgery was performed only in one case with evacuation of the SSH resulting in partial recovery of neurologic function. The other two patients were treated conservatively and showed a complete recovery. In all cases, follow up MRI several weeks later demonstrated a complete resorption of the SSH (Table 2). Four patients had a SEH. Three out of four were male, between 22 and 81 years. Initial symptoms, etiology, location and extent are summarized in Table 1. MRI in the hyperacute phase demonstrated an isointense lesion on T1-weighted and hyperintense lesion on T2-weighted images (Fig. 3). MRI in the early subacute phase showed a heterogeneous lesion with areas of high signal intensity on T1- and T2-weighted images (Fig. 4). MRI in the chronic phase demonstrated a hypointense lesion on T1- and T2-weighted (Fig. 5) (Table 2). All patients were treated conservatively, one showed a partial recovery and two a complete recovery with normal MRI about 1 month later (Table 2).
- hyperacute phase—MRI performed <12 h after onset; - acute phase—MRI performed between 12 h and 3 days; - early subacute phase—MRI performed from 3 to 7 days;
Table 2 Description of the spinal hematoma: sub- or epidural location, MRI features, time from symptom onset until MRI, treatment, evolution and MRI follow-up Case #
Space
Symptoms-MRI
T1 signal
T2 signal
Surgery
Evolution
Follow-up MRI
1 2 3 4 5 6 7
Subdural Subdural Subdural Epidural Epidural Epidural Epidural
<12 h <12 h 2 weeks <12 h <12 h 7 days Months
Isointense Isointense Hyperintense Isointense Isointense Heterogenous Hypointense
Hyperintense Hyperintense Hyperintense Hyperintense Hyperintense Heterogenous Hypointense
Yes No No No No No No
Partial recovery Complete recovery Complete recovery Partial recovery Complete recovery Complete recovery Partial recovery
Normal Normal Normal Normal Normal Normal Unchanged
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
121
have escaped medical attention [22]. In a large metaanalysis of 613 patients with spinal hematoma, the mean age at presentation was 63 years [23] with a male predominance (ratio, male:female, 1.7:1). SSH and SEH normally present with signs of spinal cord or cauda equina compression, often with an acute onset preceded by back pain or radicular pain [24–26]. Isolated pain is less common [12]. It is difficult to differentiate between SSH and SEH only by clinical history and examination, but SSH seems to have a more rapid evolution [27]. Prognosis for neurologic recovery depends on the patient’s preoperative neurologic status and duration of neurologic dysfunction [28]. Computed tomography, which is usually readily available in the acute setting, is helpful in differentiating the often hyperdense, frequently crescent-shaped hyperacute hematoma from the adjacent fat of low density and osseous structures of high density [29]. However, the subacute hematoma becomes isodense and clearly difficult to identify. Like CT imaging, MRI provides information on the localization and shape of the hematoma in the axial plane. However, the craniocaudal extent in the sagittal plane is better evaluated by MRI. Additionally, MRI more readily identifies other common causes of spinal cord dysfunction. Therefore, MRI is considered the preferred imaging modality to evaluate spinal hematomas. Spinal angiography is especially useful in the detection of vascular malformation and should be performed in cases of diagnostic suspicion on MRI images. The evolution of the MRI findings for spinal hematomas is in keeping with that described for brain hematomas [30,31] (Table 3).
Fig. 1. MRI of an 87-year-old patient presenting a hyperacute spinal subdural hematoma from T6 to T9 due to anticoagulant therapy. (a) Sagittal T2-weighted image (TR 3320, TE 117.2) of the thoracic spine demonstrates a hyperintense collection in the posterior spinal subdural space (long arrow). It is anterior to the posterior epidural fat (short arrow), displacing and compressing the spinal cord anteriorly. (b) Axial T1-weighted image (TR 440, TE 9.2) after intravenous administration of contrast material demonstrates nodular enhancement centrally in the posterior spinal subdural hematoma (arrow).
4. Discussion The recent increase of publications of SSH and SEH may be explained by the introduction of MRI in daily medical practice. As a result, more patients with a mild clinical course caused by SSH and SEH are diagnosed. Previously those patients would
1. The hyperacute hemorrhage is iso- to hypointense on T1weighted images and hyperintense on T2-weighted images due to the fresh oxygenated arterial blood containing intracellular oxyhemoglobin. A peripheral rim of hypointensity might be seen due to the degraded blood products. 2. The acute hematoma is slightly hypo-/isointense on T1weighted imaged (as the paramagnetic deoxyhemoglobin within clotted intact hypoxic red blood cells does not cause T1-shortening) and hypointense on T2-weighted images (as progressive concentration of red blood cells, blood clot retraction and fibrin production shortens T2). 3. The early subacute hematoma is very hyperintense on T1weighted images (as the oxidation of deoxyhemoglobin to methemoglobin results in marked shortening of T1) and hypointense on T2-weighted images (due to intracellular methemoglobin causing T2-shortening). Table 3 MR appearance of intraspinal hematoma Phase
Age
Hemoglobin
T1-T2
Hyperacute Acute Early subacute
<12 1–3 days 3–7 days
Iso-hyper Hypo-hypo Hyper-hypo
Late subacute
1–2 weeks
Chronic
>2 weeks
Oxyhemoglobin Deoxyhemoglobin Intracellular methemoglobin Extracellular methemoglobin Hemosiderin
Hyper-hyper Hypo-hypo
122
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
Fig. 2. MRI of a 72-year-old patient presenting a spontaneous late subacute spinal subdural hematoma from clivus to L4. (a) Sagittal T2-weighted (A) and T1weighted fat suppressed images (B and C) (TR 2640, TE 109.1 (A); TR 440, TE 8.1 (B); TR 320, TE 9.0 (C)) of the cervical (A and B) and lumbar spine (C) demonstrate a hyperintense collection in the anterior spinal subdural space of the cervical spine and posterior spinal subdural space of the lumbar spine (arrows) that cannot be distinguished from the CSF on the T2-weighted sequence. (b) Axial T1-weighted fat suppressed image (TR 540, TE 9.2) of the thoracolumbar spine shows a hyperintense collection in the posterior spinal subdural space (arrow) without cord compression. The asterisk points out the posterior epidural fat.
4. The late subacute hematoma is hyperintense on T1- and T2weighted images due to extracellular methemoglobin after lysis of red blood cells. 5. The chronic hematoma is mainly hypointense on T1- and T2weighted images because of the paramagnetic hemosiderin and ferritin within macrophages. While most articles about SSH described findings in the acute, subacute and chronic phases [29], the experience with hyperacute SSH and SEH is sparse. Pedraza et al. [11] described the MRI findings of two patients in the hyperacute phase. The SSHs were isointense on T1-weighted images and showed high signal intensity on T2-weighted images. Additionally, Pedraza et al. [11] described the presence of a low-signal rim to the lesion on T2-weighted images. In the hyperacute phase it is difficult to differentiate between the lesion and the adjacent structures due to their isointensity on T1-weighted images. In all four of our patients (Figs. 1 and 3), the MRI findings in the hyperacute phase were concordant with the findings described by Pedraza et al. [11]. In one patient, we also found a low-signal rim around the lesion on gradient-echo imaging (Patient 2). Felber et al. [32] proposes performing gradient-echo images to help characterize the hematoma in the hyperacute phase. In our opinion, the best sequence protocol to evaluate hyperacute spinal hematoma are T1-weighted images in sagittal and axial planes, T2-weighted images in sagittal planes and T2-weighted gradient-echo images in axial planes. Another diagnostic difficulty exists in imaging the T1 hyperintense subacute SSH. In one of our patients studied 2 weeks after symptom onset, we performed a T1-weighted fat suppressed sequence that clearly differentiated the hyperintense subacute subdural hematoma from epidural fat (Fig. 2). Thus, in patients studied in the subacute phase, we recommend the performance of T1-weighted images in sagittal planes, T1-weighted fat
suppressed images in sagittal and axial planes and T2-weighted images in sagittal and/or axial planes. In the early chronic phase (weeks–months), more diffuse and lightly inhomogeneous signal intensities are seen within the hematoma on T1- and T2-weighted images [7,10]. In the late chronic phase (months–years), MRI demonstrates low-signal intensity on T1- and T2-weighted images, as is seen in one of our patients (Fig. 5). After contrast material administration peripheral enhancement was observed in two patients with SEH (Fig. 3). Thus far, only a few cases with contrast enhancement have been described [33–36]. Fukui et al. [35] described similar findings in four of five patients. This peripheral enhancement is thought to be due to hyperemia of the dura resulting in thickening of adjacent meninges [34]. Nodular enhancement was seen in one patient with SSH after contrast material administration (Fig. 1). Some authors [35,37] postulate that the central pattern of enhancement in spinal hematomas may result from extravasation of contrast material by leaking vessels. It is important to be aware of this pattern of contrast enhancement as a similar appearance can be seen with tumor, aneurysm and vascular malformation [35]. The necessity of a surgical intervention versus conservative management has been discussed in the literature. Conservative management is preferred in cases with mild neurological deficits [15,38], progressive improvement in the early period [7,8,10,39–42] or if coagulopathy is present [29]. In patients in whom conservative management is chosen, evolution of spinal hematomas can be monitored using serial MRI [10,43]. In our study, one patient underwent emergency surgery for a SSH in the hyperacute phase. The other six patients with SSH and SEH were treated conservatively. Follow-up MRI demonstrated complete resorption of the hematoma in five patients, all of whom had clinical improvement. One patient had a chronic hematoma which did not change on follow-up MRI. In our opinion, MRI
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
Fig. 3. MRI of a 22-year-old patient presenting with a spontaneous hyperacute spinal epidural hematoma from T11 to L2. (a) Sagittal T2-weighted image (TR 3320, TE 117.2) demonstrates a hyperintense, slightly inhomogenous collection in the anterior spinal epidural space (asterisk) with posterior displacement of the dura (arrow) and slight compression of the spinal cord. (b) Axial T1-weighted image (TR 560, TE 9.5) after intravenous administration of contrast material demonstrates an isointense (to the spinal cord) collection in the anterior spinal epidural space (arrow) with peripheral enhancement.
123
Fig. 4. MRI of a 72-year-old patient presenting with a spontaneous early subacute spinal epidural hematoma from C3 to T2. (a) Sagittal T2-weighted fat suppressed image (TR 4000, TE 43.4) shows a heterogeneous collection in the posterior spinal epidural space (long arrow) with anterior displacement of the dura and compression of the spinal cord. (b) Axial T1-weighted image (TR 519, TE 9.2) after intravenous administration of contrast material demonstrates a mainly hypointense collection in the posterior spinal epidural space with peripheral enhancement (arrow), displacing the spinal cord anterolaterally.
124
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125
imaging characteristics. T2-weighted gradient-echo images may be helpful to demonstrate a low-signal rim. In our study, T1weighted fat suppressed images proved helpful in the subacute phase to differentiate epidural fat from hematoma. We recommend performance of sagittal planes to evaluate the full extent of the hematoma and axial planes to differentiate a SSH from a SEH by means of the epidural fat and the visualization of the dura. In patients in whom conservative management is chosen, MRI monitoring may be performed. References
Fig. 5. MRI of a 43-year-old patient presenting with a late chronic spinal epidural hematoma from T12 to L1 probably due to an epidural catheter. (a) Sagittal T1weighted image (TR 560 TE 9.5) shows a hypointense collection in the posterior epidural space (arrow) with anterior displacement of the dura. (b) Axial T2weighted image (TR 3300, TE 110.6) demonstrates a hypointense collection in the posterior spinal epidural space (arrow) displacing the cauda equina slightly anterior.
is adequate to monitor subdural hemorrhage in patients when conservative treatment is decided. 5. Conclusion MRI is regarded as the imaging modality of choice to evaluate SSH and SHE. However, there are important considerations to take into account. Spontaneous SSH is a rare lesion that is difficult to diagnose in the hyperacute phase due to the lesion’s
[1] K¨uker W, Mull M, Mayfrank L, Topper R, Thron A. Epidural spinal infection. Variability of clinical and magnetic resonance imaging findings. Spine 1997;22:544–50. [2] K¨uker W, Schaade L, Ritter K, Naciemiento W. MRI follow-up of herpes simplex virus (type 1) radiculomyelitis. Neurology 1999;52:1102–3. [3] Boukobza M, Guichard JP, Boissonet M, et al. Spinal epidural haematoma: report of 11 cases and review of the literature. Neuroradiology 1994;36:456–9. [4] Lloret-Garc´ıa J, Almenar-Garc´ıa V, Correa-Lac´arcel J, Garc´ıa de Quir´os J, Ruiz-Garc´ıa F. Hematomas extradurales espinales espont´aneos: presentaci´on de dos casos. Rev Neurol 1997;25:1936–8. [5] Sklar EML, Donovan Post JM, Falcone S. MRI of acute spinal epidural hematomas. J Comput Assist Tomogr 1999;23:238–43. [6] Hung KS, Lui CC, Wang CH, Wang CJ, Howng SL. Traumatic spinal subdural hematoma with spontaneous resolution. Spine 2002;27(24):E534–8. [7] Tillich M, Kammerhuber F, Reittner P, Leber KA, Szolar DH. Chronic spinal subdural haematoma associated with intracranial subdural haematoma: CT and MRI. Neuroradiology 1999;41:137–9. [8] Juvonen T, Tervonen O, Ukkola V, Klintrup HE. Wide-spread posttraumatic spinal subdural hematoma-imaging findings with spontaneous resolution: case report. J Trauma 1994;36:262–4. [9] Gehri R, Zanetti M, Boos N. Subacute subdural haematoma complicating lumbar microdiscectomy. J Bone Joint Surg Br 2000;82:1042–5. [10] Kulkarni AV, Willinsky RA, Gray T, Cusimano M. Serial magnetic resonance imaging findings for a spontaneously resolving spinal subdural haematoma: case report. Neurosurgery 1998;42:398–401. [11] Pedraza Guti´errez S, Coll Masfarr´e S, Casta˜no Duque CH, Suesc´un M, Rovira Ca˜nellas A. Hyperacute spinal subdural haematoma as a complication of lumbar spinal anaesthesia: MRI. Neuroradiology 1999;41:910–4. [12] Boukobza M, Haddar D, Boissonet M, Merland JJ. Spinal subdural haematoma: a study of three cases. Clin Radiol 2001;56:475–80. [13] Domenicucci M, Ramieri A, Ciappetta P, Delfini R. Nontraumatic acute subdural haematoma. Report of five cases and review of the literature. J Neurosurg (Spine 1) 1999;91:65–73. [14] Wysowski DK, Talarico L, Bacsanyi J, Botstein P. Spinal and epidural hematoma and low-molecular weight heparin. N Engl J Med 1998;338:1774–5. [15] Yamada K, Nakahara T, Yamamato K, Muranaka T, Ushio Y. Nontraumatic spinal subdural haematoma occurring in a postpartum period. Acta Neurochir (Wien) 2003;145:151–5. [16] K¨uker W, Thiex R, Friese S, et al. Spinal subdural and epidural haematomas: Diagnostic and therapeutic aspects in acute and subacute cases. Acta Neurochir (Wien) 2000;142:777–85. [17] Kirsch EC, Khangure MS, Holthouse D, McAuliffe W. Acute spontaneous spinal subdural haematoma: MRI features. Neuroradiology 2000;42:586–90. [18] Cohen JE, Ginsberg HJ, Emery D, Schwartz ML. Fatal spontaneous spinal epidural hematoma following thrombolysis for myocardial infarction. Surg Neurol 1998;49:520–3. [19] Calhoun JM, Boop F. Spontaneous spinal subdural hematoma: case report and review of the literature. Neurosurgery 1991;29:133–4. [20] Saito S, Katsube H, Kobayashi Y. Spinal epidural hematoma with spontaneous recovery demonstrated by magnetic resonance imaging. Spine 1994;19:483–6.
P. Braun et al. / European Journal of Radiology 64 (2007) 119–125 [21] Post MJ, Becerra JL, Madsen PW, et al. Acute spinal subdural hematoma. MR and CT findings with pathological correslates. Am J Neuroradiol 1994;15:1895–905. [22] Groen RJ. Non-operative treatment of spontaneous spinal epidural hematomas: a review of the literature and a comparison with operative cases. Acta Neurochir (Wien) 2004;146(2):103–10. [23] Kreppel D, Antoniadis G, Seeling W. Spinal hematoma: a literature survey with meta-analysis of 613 patients. Neurosurg Rev 2003;26:1–49. [24] Chen CJ, Fang W, Chen CM, Wan YL. Spontaneous spinal epidural haematoma with repeated remission and relapse. Neuroradiology 1997;39:737–40. [25] Horcajadas A, Katati M, Arraey MA, et al. Hematoma epidural espinal espont´aneo: presentaci´on de 2 casos y revisi´on de la bibliograf´ıa. Neurolog´ıa 1998;13:401–4. [26] Rovira A, Griv´e E, Pedraza S, et al. Hematomas espinales epidurales agudos no traum´aticos: diagn´ostico por resonancia magn´etica. Radiolog´ıa 2000;42(4):213–20. [27] Johnston RA. The management of acute spinal cord compression. J Neurol Neurosurg Psychiatry 1993;56:1046–54. [28] Rhode V, K¨uker W, Reinges MHT, Gilsbach JM. Microsurgical treatment of spontaneous and non-spontaneous spinal epidural haematomas: neurological outcome in relation to aetiology. Acta Neurochir (Wien) 2000;142:787–93. [29] Post MJ, Becerra JL, Madsen PW, et al. Acute spinal subdural hematoma: MR and CT findings with pathologic correlates. AJNR 1994;15(10):1895–905. [30] Gomori JM, Grossman RI, Goldberg HI, Zimmerman RA, Bilaniuk LT. Intracranial hematomas: imaging by high-field MR. Radiology 1985;157:87–93. [31] D¨ahnert W. Radiology review manual. Baltimore. 3rd ed; 1996. [32] Felber S, Langmaier J, Judmaier W, et al. Magnetresonanztomographie bei epiduralen und subduralen spinalen H¨amatomen. Radiologe 1994;34:656–61.
125
[33] Caldemeyer K, Mocharla R, Moran C, Smith R. Gadolinium enhancement in the center of a spinal epidural hematoma in a hemophiliac. J Comput Assist Tomogr 1993;17:321–3. [34] Crisi G, Sorgato P, Colombo A, Scarpa M, Falasca A, Angiari P. Gadolinium-DTPA-enhanced MR imaging in the diagnosis of spinal epidural haematoma. Report of a case. Neuroradiology 1990;32:64– 6. [35] Chen C, Ro L. Central gadolinium enhancement of an acute spontaneous spinal epidural hematoma. Neuroradiology 1996;38:114–6. [36] Fukui MB, Swarnkar AS, Williams RL. Acute spontaneous spinal epidural hematomas. AJNR 1999;20(7):1365–72. [37] Chang F-C, Lirng J-F, Chen S-S, et al. Contrast enhancement patterns of acute spinal epidural hematomas: a report of two cases. AJNR 2003;24(3):366–9. [38] Torres A, Acebes JJ, Cabiol J, et al. Revisi´on de 22 casos de hematomas epidurales espinales. Factores pron´osticos y manejo terap´eutico. Neurocirug´ıa 2004;15:353–9. [39] Shimada N, Sugawara T, Itoh Y, et al. Four cases of spontaneous spinal epidural hematoma. No Shinkei Geka 2005;33(2):163–8. [40] Duffill J, Sparrow OC, Millar J, Barker CS. Can spontaneous spinal epidural haematoma be managed safely without operation? A report of four cases. J Neurol Neurosurg Psychiatry 2000;69:816–9. [41] Schwerdtfeger K, Caspar W, Alloussi S, Strowitzki M, Loew F. Acute spinal intradural extramedullary hematoma: a nonsurgical approach for spinal cord decompression. Neurosurgery 1990;27(2):312–4. [42] Hentschel SJ, Woolfenden AR, Fairholm DJ. Resolution of spontaneous spinal epidural hematoma without surgery: report of two cases. Spine 2001;26(22):E525–7. [43] Lecouvet FE, Annet L, Duprez TP, Cosnard G, Scordidis V, Malghem J. Uncommon magnetic resonance imaging observation of lumbar subdural hematoma with cranial origin. J Comput Assist Tomogr 2003;27(4): 530–3.