Epidural Hematoma after Epidural Steroid Injection in a Patient Withholding Enoxaparin per Guidelines

We present a case of a patient with renal insufficiency in whom an epidural hematoma developed after an epidural steroid injection while enoxaparin was withheld per guidelines.

An 85-yr-old woman referred to the Anesthesia Pain Clinic for epidural steroid injection (ESI). She reported worsening left-sided lumbar radicular pain with left foot numbness for approximately 1 yr. Conservative therapy had provided minimal relief.

Neurologic examination results were normal, except for a positive straight leg raise test result on the left. No magnetic resonance imaging study of the lumbar spine was available. However, two-view radiography demonstrated a wedge compression deformity of the L1 vertebra, mild to moderate degenerative joint disease, and mild scoliosis.

The patient was taking warfarin for chronic atrial fibrillation and a St. Jude aortic valve. Warfarin was withheld 6 days before the ESI, and the patient received 1 mg/kg subcutaneous enoxaparin every 12 h for 4 days before her appointment. On the day before her appointment, she received only her morning dose. Therefore, at the time of injection, it was more than 24 h since her last dose. Her international normalized ratio on the day of injection was 1.2.

Epidural steroid injection was performed atraumatically with use of an 18-gauge Tuohy needle and the loss-of-resistance technique into the L4–L5 interspace. The vertebral level of the injection was not confirmed radiologically. It is possible that the injection was performed at the L3–L4 interspace. No paresthesias or blood were encounter during needle placement. Methylprednisolone, 80 mg, was injected into the epidural space. The patient experienced no exacerbations of her sciatica and was discharged from the clinic with no acute distress.

She received strict written instructions to resume her warfarin treatment on the evening of the injection and to return to the anticoagulation clinic the next day for measurement of her prothrombin time. At that time, she was instructed to continue to take 1 mg/kg subcutaneous enoxaparin every 12 h and 2.5 mg oral warfarin once daily.

Again, the enoxaparin regimen was reinstituted 24 h after the procedure. The patient had received a total of three doses when, 48 h after the procedure, she reported “100-times-worse” back pain in the same location as before ESI. History and physical examination revealed no other neurologic findings and no other signs or symptoms of infection. Her blood pressure in the emergency department was 145/43 mmHg. The patient had a history of well-controlled hypertension, but no blood pressure readings were available from the day of her ESI. Straight leg raise test results were positive at 30° on the left and negative on the right, with no change from previous to ESI. An initial magnetic resonance imaging study (fig. 1) showed severe central spinal canal stenosis centered at L3–L4 from a posterior epidural hematoma. The craniocaudal extent of the hematoma was from L2–L3 through mid L4. There was mild to moderate central canal narrowing from the hematoma above L3–L4. The patient was admitted to the hospital, and neurosurgery was consulted. Her anticoagulation regimen was discontinued. Intravenous opioids, steroids, and baclofen were instituted, and serial neurologic examinations were ordered to monitor for development of neurologic signs or symptoms.

On the second day of hospitalization, the patient experienced increasing numbness and weakness in the lower extremities and urinary retention. Subsequent magnetic resonance imaging (fig. 1) revealed increasing size of the epidural hematoma, extending from the inferior L1 level to L4–L5. Decompression laminectomies were performed at L2–L4, and dexamethasone was continued for 3 days. On postoperative day 1, the patient’s urinary retention resolved, and her pain improved dramatically. After 1 week of physical rehabilitation, she was ambulating well without assistance and reporting only mild intermittent low back pain and numbness in her toes on the left.

This patient developed an epidural hematoma after ESI, despite strict adherence to current American Society of Regional Anesthesia guidelines for neuraxial anesthesia and anticoagulation regarding administration of low-molecular-weight heparin (LMWH). These guidelines state that patients should discontinue warfarin for 4–5 days before neuraxial procedures and check prothrombin time and international normalized ratio before the procedure. These guidelines also recommend delay of the procedure until 24 h after the last dose of LMWH. If needle placement is traumatic or if blood is encountered during needle placement, a 24-h delay is recommended before restarting LMWH.1 

It is well recognized that when neuraxial anesthesia (epidural/spinal anesthesia) is used, patients receiving LMWH are at risk of development of an epidural hematoma. The risk of this event is increased by the use of indwelling epidural catheters or by concomitant use of other drugs affecting homeostasis, such as nonsteroidal antiinflammatory drugs, platelet inhibitors, or other anticoagulants. The risk also seems to be increased by traumatic or repeated epidural or spinal puncture.2 

In our case, the needle placement was atraumatic, and no catheter was placed. Although the patient’s warfarin had been withheld for 6 days previously, on the day of the ESI, her international normalized ratio was still mildly increased at 1.2. Warfarin treatment was resumed that evening, and the patient’s international normalized ratio was 1.2 the next day in the anticoagulation clinic. Of course, the administration of warfarin, even in subtherapeutic concentrations, in combination with LMWH may have contributed to the development of the epidural hematoma.

No imaging studies were available to confirm that a hematoma was not present before the ESI. It is possible that a preexisting hematoma expanded after the ESI.

Although the patient was receiving the twice-daily treatment dose (1 mg/kg) of LMWH, the ESI was approximately 28 h after her last dose in accordance with current American Society of Regional Anesthesia consensus statement recommendations. The postprocedural dose of LMWH was administered 24 h afterward, again in compliance with current recommendations.1 

The therapeutic anticoagulant effect of enoxaparin, which correlates with plasma antifactor Xa activity, peaks within 3–5 h, is 50% at 12 h, and is 0% at 24 h. The patient had taken her third post-ESI enoxaparin dose before admission. Her plasma anti-Xa concentration was increased at 0.6 U/ml, which is still within therapeutic range, 12 h after her last dose.3The potential risk factors that may prolong the half-life of LMWH and predispose to bleeding complications are renal insufficiency and advanced age.4Initially, the studies done on the effects of renal insufficiency on the half-life of LMWH were contradictory.5,6However, it has been shown recently that renal insufficiency delays the elimination of enoxaparin, and patients with renal dysfunction are at increased risk for major bleeding complications.7–11 

There is currently no clear consensus to which degree of renal insufficiency requires dose adjustment.7,8,10,12In 2002, Becker et al.  7determined that patients with marked renal impairment (creatinine clearance < 40 ml/min) had higher trough and peak anti-Xa activity compared with those with normal renal function and were more likely to have major hemorrhagic events. A study conducted by the manufacturer concluded that the elimination half-life increased with a degree of renal impairment, and this relation was more evident after repeated dosing.8One manufacturer of enoxaparin recommends dosage adjustment in severe renal impairment (creatinine clearance ≤ 30 ml/min). Although dosage adjustment is not recommended in patients with moderate (creatinine clearance = 30–50 ml) and mild (50–80 ml/min) renal impairment, they caution that such patients should be observed carefully for signs and symptoms of bleeding.3Using the Cockcroft-Gault equation, our patient’s creatinine clearance at the time of instituting her regimen and at the time of admission to the hospital were 41 and 38 ml/min, respectively. We recognize that this equation allows only for an estimation of creatine clearance from the plasma creatinine in patients with stable plasma creatine and is not a direct measure of renal function. The American Society for Regional Anesthesia’s Consensus Statement on Neuraxial Anesthesia and Anti-Coagulation recognizes concern where sustained therapeutic levels of anticoagulation are present but does not recommend monitoring of the anti-Xa concentration or increasing the time from last dose of enoxaparin when performing neuraxial procedures in these patients.1In 2002, Bastani and Gonzales11presented a patient who had moderate degree of renal insufficiency with a very prolonged duration of anticoagulation after treatment with subcutaneous enoxaparin. This case emphasizes that even a low dose of enoxaparin administered over the course of days could produce a very prolonged anticoagulation effect such that even 36 h after its discontinuation, the patient may remain anticoagulated. In fact, the authors suggest that it is prudent that physicians caring for patients with a higher risk of the delayed elimination of LMWH (i.e. , patients with renal insufficiency and the elderly) monitor antifactor Xa concentration on a regular basis, particularly before any invasive procedures. In conclusion, current American Society for Regional Anesthesia recommendations for treatment of patients receiving LMWH may not apply to patients with renal insufficiency. To reduce the risk of bleeding complications, one should check for sustained therapeutic concentrations of enoxaparin by measuring factor Xa concentrations in the elderly and in patients with renal insufficiency.

The authors thank Joe Neal, M.D. (Internal Program Director), and Spencer Liu, M.D. (Professor, Department of Anesthesiology, Virginia Mason Medical Center, Seattle, Washington), for their assistance.

* Virginia Mason Medical Center, Seattle, Washington. burrisv@hotmail.com