1042 THE JOURNAL OF BONE AND JOINT SURGERY

R. Gehri, MD, Orthopaedic ResidentM. Zanetti, MD, Consultant RadiologistN. Boos, MD, Associate Professor, Consultant Spine SurgeonOrthopaedic University Hospital Zürich, Balgrist, Forchstrasse 340, CH-8008 Zürich, Switzerland.

Correspondence should be sent to Dr N. Boos.

©2000 British Editorial Society of Bone and Joint Surgery0301-620X/00/710169 $2.00

Subacute subdural haematoma complicatinglumbar microdiscectomyR. Gehri, M. Zanetti, N. BoosFrom the Orthopaedic University Hospital, Zürich, Switzerland

There have been no previous reports of a spinalsubdural haematoma occurring as a complication

of spinal surgery. We highlight the pitfalls in thediagnosis and management of a subacute subduralhaematoma resulting from a dural tear whichoccurred as a surgical complication ofmicrodiscectomy.

J Bone Joint Surg [Br] 2000;82-B:1042-5.Received 4 May 1999; Accepted 30 June 1999

A spinal subdural haematoma is a very rare cause of aspinal compression syndrome and only a few cases havebeen reported.1,2 Its development has been related to differ-ent factors such as vascular malformation, tumours, bleed-ing disorders, anticoagulant therapy, trauma andinfection.1,3 It has occasionally been seen after diagnosticlumbar puncture or spinal anaesthesia.4-6

We describe the pitfalls in the diagnosis and managementof a subacute subdural haematoma resulting from a duraltear which occurred as a complication of micro-discectomy.

Case report

A 77-year-old healthy woman presented with severe radicu-lar pain in the left leg with an accompanying sensory andmotor deficit (MRC grade 4). MRI of the lumbar spineshowed two sequestrated disc herniations at L5/S1 withcranial and caudal displacement, respectively (Fig. 1). Afterfailure of an adequate trial of non-operative treatment, amicrodiscectomy was carried out. The two free fragmentswere extracted and the nerve roots were decompressed. During the operation the dura below the lamina of L5 wasincidentally opened over a length of 2 mm with protrusionof the intact arachnoid and sutured with one stitch (Prolene

5-0). The area was covered with a large piece of gelfoam.No leak of CSF was observed. The patient immediately losther pain and the sensory and motor deficits improved. Sixdays later, severe radicular pain reappeared in the left legaccompanied by progressive numbness and weakness in thefoot. A further MR scan demonstrated a large collection offluid at the site of the operation. This was thought to be dueto a subacute epidural haematoma in the presence of blood-soaked gelfoam covering the thecal sac (Fig. 2). At asecond operation, clotted epidural blood and blood-soakedgelfoam were removed. The thecal sac was of normalcolour and pulsating. There was no evidence of a CSFfistula, recurrent herniation of the disc or infection. Asubdural haematoma was not considered at this time. Thepatient recovered uneventfully with improvement of her legpain and motor deficit. One week after discharge, however,she again experienced severe left-sided leg pain and anincomplete, but progressive, cauda equina syndrome withdecreased sensation below the L5 level on the left, bladderdysfunction, and moderate motor weakness in the distribu-tion of L5 and S1. Further MRI showed a large lumbarsubdural haematoma at the site of the previous operations

Fig. 1

Sagittal T2-weighted MRI (TR 5000 msec, TE 130msec) before surgery showing cranial and caudal dis-placement of the extruded disc (arrows).

which was compressing the cauda equina (Fig. 3). A reviewof the initial postoperative MR scan (Fig. 2) indicated thatthe subdural haematoma was already present then. A fur-ther operation was undertaken. There was no evidence ofextradural compression of the nerve roots, but the thecalsac was tensely expanded. Partial laminotomy of L5 and S1was carried out for better exposure. The dura was carefullyincised for 5 cm under microscopic magnification and10 ml of xanthochromic fluid and some dark organisedblood clots were evacuated. A subdural space had devel-oped and the arachnoid was visible in the depth. Thearachnoid was opened and the cauda equina explored. Noother abnormality was seen apart from the previous dural

suture, where the haematoma appeared to have started. Thedura was sutured with Prolene 5-0 and covered with fibringlue. The postoperative course was uneventful. One yearlater, the patient had almost completely recovered, but stillhad some decreased sensation in the L5 and S1 dermatomesand occasional mild back pain.

Discussion

This is the first report of a lumbar subacute subduralhaematoma subsequent to the repair of a dural tear duringmicrodiscectomy. A subdural haematoma must be con-sidered when severe leg and back pain with neurological

1043SUBACUTE SUBDURAL HAEMATOMA COMPLICATING LUMBAR MICRODISCECTOMY

VOL. 82-B, NO. 7, SEPTEMBER 2000

Fig. 2a Fig. 2b

Fig. 2c Fig. 2d

Early postoperative MRI showing the subdural haematoma. Figure 2a – The sagittal T2-weighted image (TR 5000 msec, TE 130msec) demonstrates a subdural haematoma (arrowheads) with the posteroinferior part of the haematoma having intermediate signalintensity and the anterosuperior part an equivalent signal indicating higher fluid. Figure 2b – The corresponding sagittal T1-weightedimage (TR/TE: 700/12) shows the haematoma as a slightly hyperintense oval structure (arrowheads). Figure 2c – An axial non-enhanced image (TR 800/12) cranial to the site of operation reveals the subdural haematoma (arrowheads) with a semicircular shape.Figure 2d – An axial T1-weighted contrast-enhanced image (TR/TE: 800/12) shows the haematoma on the right side (arrowheads) andpostoperative changes on the left side with a small air bubble (arrow). The dorsal dura (small arrows) lies outside the haematoma andindicates the subdural location.

deterioration occur after the operation. To date, only onecase of a similar postoperative course has been reported,but this was attributed to inadvertent trauma during pre-operative myelography.6

A subdural haematoma was not considered at the initialrevision procedure. The haematoma demonstrated on MRIposterior to the cauda equina was misinterpreted as beingepidural in the presence of a large blood-soaked piece ofgelfoam. At operation the dura appeared normal and therewas no evidence to suggest a subdural haematoma. Aretrospective analysis of the first repeat MR scan, however,indicated unequivocal signs of its presence.

The characteristic MRI findings of a subdural haemato-ma are areas of abnormal signal intensity within the duralsac. The signal intensity depends on the time since thedevelopment of the haematoma. Within the first few days,there is a higher signal intensity on T1-weighted imagescompared with the CSF.7 Later, the change in signal inten-sity is more variable.7 In our case, two weeks after thebleeding the subdural haematoma appeared more like fluid,with low signal intensity on T1-weighted images and highintensity on T2-weighted images. By contrast, Post et al8

found a striking low signal intensity on T2-weightedimages in their follow-up. It is possible that the decrease inintensity on these images was caused by the process offormalin fixation in their postmortem study.

The axial scan is usually helpful in differentiating sub-dural and epidural haematoma. The subdural location isclear when the haematoma extends within the border of thedura mater. Since spinal epidural haematomas spread wide-ly in the epidural space, the shape of the dura mater is oftenirregular.9 With a subdural haematoma the dura appears

smooth. The lack of direct continuity with the adjacentbony structures points to a subdural location.8 The lowsignal intensity of the dura may be obscured by the haema-toma itself.

The shape of the haematoma can sometimes be diag-nostic. A subdural haematoma has a semicircular appear-ance on axial scans (Fig. 2c) whereas an epiduralhaematoma is more convex. The differentiation between asubdural and epidural haematoma will remain difficult andat times only surgical exploration will prove the exactlocation of the haematoma.

Debate continues as to the presence of a subdural space.Haines10 stated that a subdural haematoma is most fre-quently found within the layer formed by ‘dural bordercells’ and that there is no evidence of a subdural space inthe region of the dura-arachnoid junction. When a spacedoes appear at this site, it is the result of pathological ortraumatic processes which have resulted in tissue damage.In our case, the onset of the subdural haematoma wasrelated to a tiny dural tear. The repair of the tear under thelamina L5 was technically difficult because of the limitedexposure without laminectomy. The dura was suturedsuperficially in order to avoid entrapment of the roots of thecauda equina. We assume that an injury to a dural capillaryoccurred at the time of the repair of the thecal sac. Sub-sequent bleeding may have developed and caused the for-mation of a haematoma between the dura and the intactarachnoid. A subdural haematoma has not been describedpreviously as resulting from unintended durotomy.11 Thiscomplication could have been avoided if the arachnoid hadbeen opened and both the dura and the arachnoid correctlysutured after an adequate exposure.

1044 R. GEHRI, M. ZANETTI, N. BOOS

THE JOURNAL OF BONE AND JOINT SURGERY

Fig. 3a Fig. 3b

The second postoperative MR scan shows persistence of the subdural haematoma. Sagittal a) T1- and b) T2-weighted images demonstrate the subdural haematoma at the same location with additional cranial extension(arrowheads). The signal characteristics indicated more fluid (dark on the T1-weighted image, bright on the T2-weighted image) than those of the first postoperative scan (Fig. 2).

No benefits in any form have been received or will be received from acommercial party related directly or indirectly to the subject of thisarticle.

References1. Russell NA, Benoit BG. Spinal subdural hematoma: a review. Surg

Neurol 1983;20:133-7.2. Shimada Y, Sato K, Abe E, Miyakoshi N, Tsutsumi Y. Spinal

subdural hematoma. Skeletal Radiol 1996;25:477-80.3. Roscoe MW, Barrington TW. Acute spinal subdural hematoma: a

case report and review of literature. Spine 1984;9:672-5.4. Edelson RN, Chernik NL, Posner JB. Spinal subdural hematomas

complicating lumbar puncture. Arch Neurol 1974;31:134-7.5. Jonsson LO, Einarsson P, Olsson GL. Subdural haematoma and

spinal anaesthesia: a case report and an incidence study. Anaesthesia1983;38:144-6.

6. Reinsel TE, Goldberg E, Granato DB, Wilkinson S, Penn R. Spinalsubdural hematoma: a rare cause of recurrent postoperative radiculo-pathy. J Spinal Disord 1993;6:62-7.

7. Kulkarni AV, Willinsky RA, Gray T, Cusimano MD. Serial mag-netic resonance imaging findings for a spontaneously resolving spinalsubdural hematoma: case report. Neurosurgery 1998;42:398-400.

8. Post MJ, Becerra JL, Madsen PW, et al. Acute spinal subduralhematoma: MR and CT findings with pathologic correlates. AJNR AmJ Neuroradiol 1994;15:1895-905.

9. Boukobza M, Guichard JP, Boissonet M, et al. Spinal epiduralhaematoma: report of 11 cases and review of the literature. Neuro-radiology 1994;36:456-9.

10. Haines DE. On the question of a subdural space. Anat Rec1991;230:3-21.

11. Jones AA, Stambough JL, Balderston RA, Rothman RH, BoothRE Jr. Long-term results of lumbar spine surgery complicated byunintended incidental durotomy. Spine 1989;14:443-6.

1045SUBACUTE SUBDURAL HAEMATOMA COMPLICATING LUMBAR MICRODISCECTOMY

VOL. 82-B, NO. 7, SEPTEMBER 2000