Introduction

Spinal anesthesia is a well-established ordinary practice for patients undergoing orthopedic surgery. Spinal epidural hematoma (SEH) represents a rare, but potentially devastating complication of spinal anesthesia with an incidence, according to a recent 10-year closed claims analyses, of 1:775 000.1 After obtaining patient’s written consent, we report a case of SEH, far from the spinal needle puncture site, on the third postoperative day.

Case presentation

A 64-year-old woman, with BMI of 22.3 kg m−2, was scheduled for left total knee replacement. Her medical history was unremarkable. Her laboratory investigations, including renal function, platelet count and coagulation tests (activated partial thromboplastin time and international normalized ratio), were within normal range. She had spinal anesthesia, which is commonly preferred in our department over epidural mainly due to its rapid onset of surgical anesthesia, in the sitting position through a median approach with a 25-gauge Whitacre needle (Vygon, Ecouen, France) in the L2/L3 space. The technique was atraumatic and successful on first attempt. No paresthesia was elicited, clear cerebrospinal fluid was obtained and 3 ml bupivacaine 0.5% was injected intrathecally. Within 12 min, complete motor block of the lower extremities was achieved and the sensory block, evaluated using the pinprick test, was identified at the T8 level bilaterally. The surgical procedure was uneventful and the block was fully reversed 3 h later. She was mobilized on the same day. Prophylactic anticoagulation (subcutaneous bemiparin 3500 IU per day) was started the next day.

On the third postoperative day at 1200 hours (4 h after bemiparin administration), she complained of acute onset pain in both shoulders and upper extremities. Within 20 min, the pain increased and shortly after, paraparesis emerged. Blood pressure was 180/90 mm Hg and her electrocardiogram was unremarkable. Clinical examination revealed sensory loss below T10, flaccid paralysis of the lower extremities (muscle strength: grade 0) and muscle weakness of the upper extremities (muscle strength: grade 3 for forearm flexion, grade 2 for elbow extension, grade 0 for finger flexion–extension) with normal sensation. New clotting tests were normal. A magnetic resonance imaging (MRI) was performed at 1300 hours, which revealed a vascular malformation from C2 to T3, partially ruptured, compressing the spinal cord, mostly at C6/C7 and C5/C6 levels, where small disc herniations were additionally present (Figure 1). Since the clinical presentation was gradually improving after the MRI, conservative treatment was decided, using methylprednisolone 16 mg three times daily, which was continued for 10 days after neurological status was restored. At 2000 hours, sensation was normal in upper and lower extremities and at 0200 hours motor activity of the lower extremities was restored. Following diagnosis, bemiparin was discontinued and pneumatic calf compression devices were used instead. She was mobilized 2 days later and was discharged 9 days later without any symptoms. Follow-up was scheduled in order to reevaluate the patient and perform a selective spinal angiography that would be needed to confirm the diagnosis, allow classifying of the venous malformation and discuss further endovascular or surgical therapeutic possibilities but the patient did not attend the appointment.

Figure 1
figure 1

Early MRI. (a) T1-weighted sagittal image showing an intermediate signaling mass, corresponding to an epidural hematoma (arrows). (b) T2-weighted image showing a hyperintensive signaling mass depicting an abnormal clustering of vessels corresponding to a vascular malformation (arrows) and the hematoma. (c, d) T1 and T2-weighted axial images of the upper cervical spine showing the epidural hematoma (arrows) compressing posteriorly the spinal cord toward the vertebral bodies.

Discussion

A SEH may be either spontaneous or secondary to trauma, coagulation disorders or regional anesthetic techniques. Risk factors include advanced age, female gender, traumatic and multiple attempts at puncture for neuraxial anesthesia, vascular malformations, anticoagulation therapy and arterial hypertension.2 Regarding neuraxial techniques, the level of the hematoma is usually the level of puncture or up to the highest level of catheter insertion.3 Hemorrhagic complications, far from needle puncture site, attributed to anesthesia, have been reported mainly as cerebral subdural hematomas, where the possible mechanism is the caudal shift of intracranial structures due to low CSF pressure causing traction on dural veins with subsequent bleeding.3, 4

The etiological relation to bemiparin cannot be excluded, as onset of symptoms corresponds to the timing of its maximal plasma anti-Xa activity (2–6 h). The interval between initiation of low molecular weight heparin and onset of neurological dysfunction due to hematoma ranges between 15 h and 3 days, which is in agreement with our finding. On the other hand, the possibility of vessel damage during spinal anesthesia has been reported, but is difficult to document. In our case, multifactorial etiology is more likely, including vascular malformation and thromboprophylaxis, especially since the hematoma occurred on the third day after spinal anesthesia and on the second day after starting thromboprophylaxis, while mechanically induced vessel damage cannot be ruled out. The latter may be attributed to a caudal traction on the arachnoid matter and epidural venous plexus of cervical and thoracic regions secondary to loss of CSF after the lumbar puncture, causing stretching and tearing of the epidural veins. If an epidural anesthesia had been performed, this mechanical stress would have been avoided, unless an accidental dural puncture had occurred, but as we are not sure of which factor had a major role in the occurrence of the hematoma, it is difficult to exactly assess the degree of involvement of the anesthetic technique on the outcome.

Treatment of SEH includes either surgical or conservative measures. Conservative treatment with repeated neurologic examination was decided in our case, based on the gradual resolution of symptoms, after the MRI, which might be explained by the leakage of the fluid hematoma through the intervertebral foramina, relieving the pressure on the spinal cord. The dosage and duration of methylprednisolone treatment was as per the judgment and experience of the neurosurgeon and not as per any guidelines. Several cases of complete resolution have been described with conservative management and there is a growing body of evidence supporting it in selected cases. The majority of patients treated conservatively had already shown improvement during initial evaluation.5

Conclusions

The possibility of SEH should always be considered whenever neurological symptoms occur in the postoperative period, especially if neuraxial blockade has been performed. Detailed neurological examination should be carried out in any case of suspicion. MRI is the preferred diagnostic method and early neurosurgical consultation is imperative in order to decide on management. Conservative treatment may be considered in selected cases.

Data archiving

There were no data to deposit.