Secondary Logo

Journal Logo

Paraplegia After Thoracic Epidural Steroid Injection

Loomba, Vivek MD; Kaveeshvar, Hirsh DO; Dwivedi, Samvid DO

doi: 10.1213/XAA.0000000000000402
Case Reports: Case Report

Epidural steroid injections are a common procedure performed by pain physicians. The American Society of Regional Anesthesia along with several other groups recently provided guidelines for performing epidural injections in the setting of anticoagulants. We present a case of a patient who developed an epidural hematoma and subsequent paraplegia despite strict adherence to these guidelines. Although new guidelines serve to direct practice, risks of devastating neurologic complications remain as evidenced by our case.

From the *Department of Anesthesiology, Henry Ford Hospital, Detroit, Michigan; and Department of Neurology, Henry Ford Hospital, Detroit, Michigan.

Accepted for publication June 20, 2016.

Funding: None.

The authors declare no conflicts of interest.

Address correspondence to Vivek Loomba, MD, Department of Anesthesiology, Henry Ford Hospital, 2799 West Grand Blvd, Detroit, MI 48202. Address e-mail to

Spine pain is a well-established condition routinely encountered by pain physicians. With known risks of opioid administration, the use of interventional pain procedures for chronic neuraxial pain has increased substantially.1 Epidural steroid injections are commonly utilized in the outpatient setting for the management of patients with chronic pain. A difficult scenario arises when performing interventional procedures on patients who are anticoagulated or on antiplatelet medications. One of the most catastrophic complications when performing neuraxial blocks in patients receiving anticoagulation is spinal/epidural hematoma, resulting in the potential for irreversible neurologic sequelae. The American Society of Regional Anesthesia (ASRA) along with several other expert societies recently published new guidelines for performing neuraxial injections in the setting of anticoagulation medications.2 We report a case that shows that severe complications may occur even when these guidelines are strictly followed.

Back to Top | Article Outline


A 63-year-old woman was referred to our pain clinic with a 4-year history of thoracic radiculopathy secondary to a T10 compression fracture as a result of a fall on concrete. Her medical history included bipolar disorder, renal stones, chronic kidney disease, hypertension, coronary artery disease, previous myocardial infarction, paroxysmal atrial fibrillation, and transient ischemic attacks. Her medications included amiodarone, furosemide, gabapentin, metoprolol, prednisone, aspirin (81 mg/d), and warfarin (5 mg/d). One year after the fall, she underwent vertebroplasty for deteriorating kyphosis but continued to experience worsening pain.

She was diagnosed with a thoracic radiculopathy based on clinical examination and diagnostic imaging and was subsequently scheduled for a T9–T10 epidural steroid injection (ESI). She had experienced good short-term pain relief after the ESI. Typically, her warfarin would be discontinued for 1 week, and she would be bridged with enoxaparin up until 1 day before the procedure. She would typically continue taking aspirin 81 mg/d for secondary stroke prevention. She would resume the warfarin that evening.

She had a trial for a spinal cord dorsal column stimulator under fluoroscopic guidance with an entry point of T12–L1. The tip of lead was located at the top of the T5 vertebrae in the left lateral gutter with the area of capture being the left thoracic area of pain. The neurostimulator trial was successful, achieving >50% improvement in pain in the left thoracic region without any unpleasant side effects. The temporary stimulator was removed after a trial period of 5 days. A permanent implantation was performed 20 days after the trial with 2 octad neurostimulator leads. The entry point was T12–L1, and the leads were guided under fluoroscopic guidance until the tip of the leads reached the top of T5 level in the right and left gutter. Because of persistent burning sensation and less than optimal stimulation, she had multiple revisions of spinal cord stimulator (SCS) after initial implantation. Three months after her initial revision, we were unable to capture the axial component of pain; thus, the leads were removed and placed peripherally in the thoracic paraspinal area spanning the T10–T12 region. The patient had good pain relief in her back; however, she continued to experience radicular pain.

Therefore, a decision was made to continue with thoracic ESI every 2 to 3 months. The aforementioned combined approach improved her overall pain score and function. In total, she had 11 ESIs performed to date without complication.

In preparation for her scheduled T9–T10 ESI as had been done for her previous procedures, warfarin was discontinued for 7 days, and a bridging therapy with 1 mg/kg enoxaparin twice daily was initiated. As per the ASRA guidelines, enoxaparin was discontinued for 24 hours before the scheduled injection. She continued her 81 mg daily dose of aspirin. The day before the procedure, she had an international normalized ratio (INR) of 1.10 and prothrombin time (PT) of 13.9 seconds. The most recent platelet count, measured 54 days before injection, was 267 × 103/µL. On the day of the procedure, her INR was 1.00.

Informed written consent was obtained before the thoracic ESI. After securing intravenous access, standard monitors were applied including oxygen saturation, noninvasive blood pressure, and electrocardiogram. She received intravenous sedation with 1 mg midazolam and 50 μg fentanyl for the procedure. Care was taken to avoid peripheral stimulator leads in the T5 level throughout the procedure (Figures 1 and 2). Using anteroposterior fluoroscopy, lateral fluoroscopy, and loss-of-resistance technique, the epidural space was entered with a 3-1/2-inch 20-gauge Tuohy needle. Care was taken to avoid peripheral stimulator leads in the thoracic area throughout the procedure.

Figure 1

Figure 1

Figure 2

Figure 2

After negative aspiration for cerebrospinal fluid, blood, air, and with no paresthesia, the epidural space was injected with 1 mL Iohexol 240 contrast, which was consistent with previous procedure records, and an epidurogram was excellent on visualization showing dye spread from T9 to T11. Again, after negative aspiration for cerebrospinal fluid, blood, air, and with no paresthesia, the epidural space was injected with a block solution containing 80 mg methylprednisolone acetate mixed with 9 mL preservative-free saline. Epidurogram washout was observed on fluoroscopy.

The entire procedure lasted 10 minutes with vital signs monitored by a pain management sedation nurse under staff supervision. The patient tolerated the procedure well with no immediate complications. She was discharged home after meeting discharge criteria from the recovery area with instructions to restart warfarin in the evening.

Postprocedure day 5, the patient reported to the scheduled anticoagulation clinic where her INR was 1.44, PT 17.5 seconds, and platelet count 221 × 103/µL. At this point, she remained on low-molecular-weight heparin (LMWH) while bridging her warfarin. She did not express any complaints of back pain, which was different than her chronic pain. The next day the patient presented to emergency department of an outside hospital with increasingly severe backache associated with numbness and weakness in bilateral lower extremities with urinary and fecal incontinence for the past 3 to 5 hours. After initial examination, she was transferred to a tertiary-level hospital. The time from presentation to transfer to our emergency department was approximately 5 hours. On arrival to our emergency department, clinical examination demonstrated complete paraplegia with a T4 sensory level and loss of anal sphincter tone. Coagulation profile results were INR 1.44, PT 17 seconds, and activated partial thromboplastin time 40 seconds. Platelet count was low at 145 × 103/µL. Because of the spinal cord stimulator, she was not a candidate for magnetic resonance imaging. A computed tomography myelogram with contrast demonstrated a large posterolateral epidural hematoma extending from T2 to T9 causing severe compression of the spinal cord at T2–T10, worse at the level of T7–T8 (Figure 3).

Figure 3

Figure 3

She received 10 mg intravenous vitamin K and 2 units of fresh frozen plasma and was emergently taken to the operating room where a laminectomy for evacuation of the epidural hematoma was performed. Total time elapsed from her initial outside emergency department presentation to the operating room was approximately 16 hours. After the incision, significant oozing of blood was noticed in the epidural space, and epidural fat was noticed to be infiltrated with blood clots. An evacuation of the epidural hematoma and T3–T8 thoracic laminectomy was performed. The surgeon believed that the hematoma appeared to be a result of a slow bleeding process, likely accumulated over several days.

She was extubated successfully the next day, and her neurologic examination was significant for 0/5 strength in her lower extremities bilaterally with a sensory level of T4, consistent with an American Spinal Injury Association T9A score. Coagulation profile on postoperative day 1 revealed INR 1.49, PT 17.4 seconds, activated partial thromboplastin time 39 seconds, and platelets 134 × 103/µL. Anticoagulation was not resumed at this time.

The patient was subsequently discharged to an inpatient rehabilitation facility 12 days after admission with continued paraplegia, T4 sensory level, and urinary and bowel incontinence. Six weeks after decompression, she had not regained any strength in her lower extremities. Minimal sensation returned to her right leg, and she continued to report severe back pain and spasms in the legs. Treatment with tramadol, lidocaine patches, and 5 mg baclofen three times daily provided some pain relief.

At 1-year follow-up, she had minimal improvement in her neurologic examination with persistent sequelae from urinary and bowel incontinence and continues to experience severe pain and spasticity.

Back to Top | Article Outline


Despite following ASRA guidelines for interventional spine and pain procedures in patients on antiplatelet and anticoagulant medications, our patient experienced an unexpected and devastating neurological injury postprocedure.2 As evidenced by this case report, although laboratory tests may be within an acceptable range for a neuraxial procedure, a possibility of an adverse outcome remains, especially with a patient on multiple blood-thinning medications.

ESI is less invasive than spinal surgery with proven functional improvement.3 Although thoracic pain occurs less often than low back or neck pain, it remains a chronic, disabling condition.3 Treatment options range from conservative treatment with physical therapy, to drug therapy with neuromodulators and opioids, to more invasive spinal surgery interventions. The risks of long-term opiate treatment include serious health issues and death.4 With increasing pressure to avoid opioids in the chronic noncancer pain patient, nonopioid modalities for pain relief have expanded, including the repeated use of interventional techniques, especially for spinal pain.

One of the most feared and devastating, albeit rare, complications of ESI is epidural hematoma formation.5 The incidence of neurologic dysfunction resulting from hemorrhagic complications associated with ESI is unknown. In 1993, the calculated incidence of epidural hematoma after epidural anesthesia was estimated to be <1 in 150,000 and <1 in 220,000 spinal anesthetics.1,6 More recent reports show that this risk is much higher, especially with the advent of LMWHs and antiplatelet agents.6 These studies suggest that epidural hematoma after epidural blockade likely ranges from 1:2700 to 1:19,505.6 Certain factors such as osteoporotic deformities altering the epidural anatomy, use of dual antiplatelet/anticoagulant therapies, unrecognized use of nonprescribed aspirin containing over-the-counter medications, or an accumulation of anticoagulant caused by undetectable or known compromised renal function7 may increase the risk of hematoma formation.6

As the use of interventional pain procedures has become more common, new and more refined anticoagulant/antiplatelet agents became available, although evidence remains limited on the safety of their use with neuraxial blocks. Concomitant administration of medications affecting hemostasis, including antiplatelet drugs and vitamin K antagonists, represents additional risks of hemorrhagic complications, including spinal hematoma.8 Simultaneous use of multiple anticoagulation medications is common in medical management, especially with bridging from one drug to another in preparation for surgery and procedures. In the case of a single anticoagulant or antiplatelet administration, the medication can be withheld for an established time before the scheduled procedure with relative certainty that hemostasis will be restored. Clinical laboratory measurements may be able to predict the anticoagulation status of a single medication with some certainty; however, concurrent use of other medications affecting the coagulation mechanism may further complicate matters such as a coagulopathy unidentified by measurement of PT or INR.9 For example, antiplatelet or oral anticoagulants administered in combination with LMWH may increase the risk of spinal hematoma.9

For patients receiving LMWH before neuraxial anesthesia, recommendations specify that needle placement should occur at least 10 to 12 hours after the last dose of prophylactic LMWH. In patients receiving therapeutic dose LMWH, neuraxial techniques should be delayed at least 24 hours after the last dose.2 Furthermore, the presence of blood during needle and catheter placement warrants delaying initiation of LMWH therapy for 24 hours after the procedure because traumatic needle or catheter placement may increase the risk of spinal hematoma.10

Platelet adequacy must also be addressed before the neuraxial block. Platelet function is thought to be more important than platelet count alone.7 No specific test exists to guide antiplatelet therapy, including bleeding time. Identifying patients with risk factors predisposing them to bleeding is crucial. Risk factors include a history of easy bruising/excessive bleeding, female sex, and increased age.10

Previous studies have supported the use of neuraxial blocks in patients taking aspirin; however, several cases of spinal/epidural hematoma during neuraxial block in combination with aspirin or other cyclo-oxygenase-1 inhibitors have been reported. In most of these cases, the concomitant use of a nonsteroidal anti-inflammatory drug with heparin or other antiplatelet agents had been a major implicating risk factor for spinal/epidural hematoma.7

To justify continued intervention in anticoagulated patients, benefits should be carefully weighed against possible risks. Objective measures of improvement should be sought such as pain scores, disability indices, and decreased analgesic use before further intervention. Patients with poor compliance and taking anticoagulation medications than otherwise recommended by their physician should forego neuraxial procedures, because this scenario could put them at unnecessary risk for bleeding.

Patients should be thoroughly educated on “red flag” symptoms such as increased sensory or motor block followed by bowel/bladder dysfunction11 so that prompt medical intervention can be obtained. Although Bateman et al11 found that a minority of patients with spinal hematoma regained partial or good neurologic recovery, spinal cord ischemia was reversible in those patients receiving laminectomy within 8 hours of the onset of neurologic dysfunction.

Despite the low reported incidence of spinal/epidural hematoma, the devastating resultant neurologic damage demands effective preventive strategies for patients receiving neuraxial blockade in the presence of anticoagulation.7 The ASRA anticoagulation guidelines serve as a tool for practitioners to safely place epidurals/spinals in patients receiving anticoagulation.2 Notably, these guidelines are based on case reports, small studies, and theoretical knowledge of the pharmacokinetics and dynamics of each anticoagulant rather than large randomized prospective studies.7

It is unclear what exactly led to our patient’s devastating outcome despite following ASRA guidelines. In our opinion, spinal cord stimulator placement had no role to play because leads had been removed from neuraxis over a year prior. Our clinic typically performs a 24- to 48-hour postprocedure follow-up to monitor for “red flag” symptoms. In our patient’s case, there is no documentation of this occurring; however, in her anticoagulation clinic notes 5 days postprocedure, she did not report any bleeding, bruising, melena, or changes in medication. There is, however, a possibility that the patient was not being completely truthful about holding anticoagulation or resuming anticoagulation too early, which cannot be necessarily determined with blood work. Furthermore, the patient’s aspirin was not held because of secondary stroke prevention in a patient with cerebrovascular risk factors as well as a previous transient ischemic attack. This may have been a causative factor for postprocedural hemorrhage from aspirin’s antiplatelet effects.

In conclusion, practitioners today face many challenges with respect to performing neuraxial techniques in the presence of anticoagulation or antiplatelet therapy. Not only has the introduction of many new anticoagulation/antiplatelet medications complicated management, but many patients receive more than one of these medications simultaneously. The current guidelines serve to direct practice; yet, risks of devastating neurologic complications remain as evidenced by our case.

Back to Top | Article Outline


1. Manchikanti L, Falco FJ, Singh V, et al. Utilization of interventional techniques in managing chronic pain in the Medicare population: analysis of growth patterns from 2000 to 2011. Pain Physician. 2012;15:E969–E982.
2. Narouze S, Benzon HT, Provenzano DA, et al. Interventional spine and pain procedures in patients on antiplatelet and anticoagulant medications: guidelines from the American Society of Regional Anesthesia and Pain Medicine, the European Society of Regional Anaesthesia and Pain Therapy, the American Academy of Pain Medicine, the International Neuromodulation Society, the North American Neuromodulation Society, and the World Institute of Pain. Reg Anesth Pain Med. 2015;40:182–212.
3. Enneking FK, Benzon H. Oral anticoagulants and regional anesthesia: a perspective. Reg Anesth Pain Med. 1998;23(6 suppl 2):140–145.
4. Manchikanti L, Cash KA, McManus CD, Pampati V, Benyamin RM. Thoracic interlaminar epidural injections in managing chronic thoracic pain: a randomized, double-blind, controlled trial with a 2-year follow-up. Pain Physician. 2014;17:E327–E338.
5. Bloodworth DM, Perez-Toro MR, Nouri KH. Neurological deficits after epidural steroid injection: time course, differential diagnoses, management, and prognosis suggested by review of case reports. Pain Med. 2008;9(suppl 1):S417–S57.
6. Tryba M. [Epidural regional anesthesia and low molecular heparin: Pro] [in German]. Anasthesiol Intensivmed Notfallmed Schmerzther. 1993;28:179–181.
7. Horlocker TT, Heit JA. Low molecular weight heparin: biochemistry, pharmacology, perioperative prophylaxis regimens, and guidelines for regional anesthetic management. Anesth Analg. 1997;85:874–885.
8. Ladha A, Alam A, Idestrup C, Sawyer J, Choi S. Spinal haematoma after removal of a thoracic epidural catheter in a patient with coagulopathy resulting from unexpected vitamin K deficiency. Anaesthesia. 2013;68:856–860.
9. Kreppel D, Antoniadis G, Seeling W. Spinal hematoma: a literature survey with meta-analysis of 613 patients. Neurosurg Rev. 2003;26:1–49.
10. Green L, Machin SJ. Managing anticoagulated patients during neuraxial anaesthesia. Br J Haematol. 2010;149:195–208.
11. Bateman BT, Mhyre JM, Ehrenfeld J, et al. The risk and outcomes of epidural hematomas after perioperative and obstetric epidural catheterization: a report from the Multicenter Perioperative Outcomes Group Research Consortium. Anesth Analg. 2013;116:1380–1385.
© 2016 International Anesthesia Research Society