CERVICAL epidural steroid injections (CESI) are used commonly for the treatment of cervical radiculopathy. Significant subjective patient satisfaction has been reported, although controlled trials have not yet delineated the effectiveness of this procedure. 1–3
Complications of CESI are uncommon, the most frequent being dural puncture and transient paresthesia. 1,3
Two recent reports of more serious sequelae associated with CESI bring into question the risk–benefit ratio of attempting this technique in conservative management of cervical pain. 3,4
We report three cases of severe pain consistent with nerve injury immediately after CESI.
A 33-yr-old man with a history of worsening neck pain that radiated to the left second and third fingers was referred for consideration of CESI. The patient reported no weakness or sensory change in the arms or legs. Results of physical examination were normal, except for a slight decrease in sensation to light touch over the dorsal side of the left second finger and absence of the left triceps deep tendon reflex. Magnetic resonance imaging (MRI) showed a large central herniated nucleus pulposus at the C6–C7 interspace, lateralizing to the left, with displacement of the spinal cord toward the right (fig. 1
). Posterior displacement of the spinal cord and effacement of the cerebrospinal fluid signal that is normally anterior to the spinal cord was evident on the sagittal image (fig. 2
Cervical epidural steriod injection was performed with the patient sitting, with his neck flexed and his head resting on a cushion. A 17-gauge Weiss epidural needle was inserted through the C6–C7 interspace via a midline approach. The epidural space was identified using the hanging-drop technique, with saline administered at a depth of 5 cm from the skin. Immediately after entry of the needle into the epidural space, the patient experienced a shock-like sensation, extending to the right upper and lower extremities that resolved instantaneously. No blood or cerebrospinal fluid was aspirated from the needle. The transient paresthesia had completely resolved, and instillation of an additional 1 ml preservative-free saline produced no further symptoms. A mixture of 80 mg methylprednisolone (2 ml) and 0.5% lidocaine (2 ml) was administered through the epidural needle during gravity flow, without further discomfort. Then, the needle was withdrawn. Several minutes after the procedure, the patient reported a dull ache in the web space between the right first and second fingers. Examination showed pronounced allodynia in the same area. The pain and allodynia in the right hand persisted through the patient’s discharge from the clinic, 30 min later.
The day after CESI, the patient reported worsened pain and extreme sensitivity to touch in his right hand. Physical examination showed marked allodynia, warmth, and erythema in the web space between the right first and second fingers. Strength and deep tendon reflexes were normal in the right upper extremity. He was prescribed a tapering dose of oral steroids and oral oxycodone–acetaminophen, and his symptoms resolved over 3 days. Seven days after CESI, there was no further allodynia in the right hand. The original radicular symptoms in the left C7 distribution slowly resolved over 3 weeks. At the last follow-up, 6 months after CESI, the left C7 radicular symptoms continued to increase and decrease in intensity, but the patient experienced no further severe exacerbations.
A 49-yr-old man with a 3-month history of worsening neck pain that radiated to both arms was referred for CESI. Five years previously, the patient had undergone anterior cervical discectomy and fusion at the C5–C6 and C6–C7 levels.
The patient reported no weakness or sensory change in the arms or legs. At physical examination, marked limitation of cervical range of motion in flexion and extension was noted. Sensory, motor, and deep tendon reflex examination in all extremities yielded normal results. MRI showed intact fusion at the C5–C6 and C6–C7 levels. An osteophytic ridge extended posteriorly into the anterior epidural space but caused no displacement of the spinal cord or effacement of the cerebrospinal fluid signal anterior to the spinal cord. There was no stenosis of the central spinal canal or the lateral recesses at any of the cervical levels.
Cervical epidural steroid injection was performed as described in the first case. Immediately after entry of the needle into the epidural space, the patient experienced an aching sensation that extended from the left shoulder to the left fifth finger. No blood or cerebrospinal fluid was aspirated from the needle. The transient parasthesia resolved after several seconds. A mixture of 120 mg methylprednisolone (3 ml) and 1% lidocaine (2 ml) was administered through the epidural needle, and the needle was withdrawn without further discomfort. The patient had ongoing discomfort in the left arm at the time of discharge from the clinic, 30 min later.
The day after CESI, the patient returned to the clinic, reporting pain that extended from the left forearm into the third through fifth fingers of the left hand, with burning and sensitivity to light touch, and swelling of the dorsum of the hand. Physical examination showed marked allodynia, warmth, erythema, and swelling of the left hand, including the third through fifth fingers. There was decreased sensation to pin prick in the same area. He was prescribed a tapering dose of oral steroids and analgesics (hydrocodone–acetaminophen). He returned weekly during the next month, and his symptoms gradually resolved without further intervention. Sensitivity to cold that was not otherwise painful persisted in the left third through fifth fingers for more than 6 weeks after CESI. At the last follow-up, 3 months after CESI, the left-sided symptoms had resolved; however, the neck and right arm pain persisted.
A 54-yr-old woman with a 10-yr history of pain and tingling in the left arm and hand, which was diagnosed as cervical radiculopathy, was referred for CESI. Her symptoms had responded to previous CESI and recurred 5 months before her presentation to the clinic. Her history was significant for stroke 3 months previously, which had initially caused left-sided weakness and decreased sensation. Physical examination showed weakness in all hand muscles and decreased sensation to pin prick in the entire left hand to the mid forearm, consistent with residua from the stroke. MRI of the cervical spine showed moderate C4–C5 disk herniation, with mild compression of the spinal cord at that level, and moderate disk protrusion at C6–C7, with facet joint hypertrophy, causing foraminal narrowing at the left C6–C7 level. CESI was performed at the C6–C7 level without complication and produced relief for approximately 2 weeks. The patient returned 1 month later for a subsequent CESI. The epidural space was located using the hanging-drop technique at C6–C7. A solution of 80 mg methylprednisolone (2 ml) and 2 ml lidocaine, 0.5%, was administered through the epidural needle during gravity flow. The solution stopped flowing after half of the solution had been instilled. The needle was withdrawn and reinserted through the same interspace, this time using loss of resistance to air to locate the epidural space. During needle advancement, the patient reported a severe shooting pain to the left hand. The procedure was discontinued, and paresthesia resolved completely after several minutes. The patient returned 4 days later, reporting severe pain, swelling, and allodynia in the left hand. Examination results were not appreciably changed from baseline, except for hand swelling. She was administered oral dexamethasone and gabapentin. Symptoms gradually resolved over 3 months. She continued to experience neck pain that radiated into her left arm, and, 10 months later, she underwent anterior cervical discectomy at the C6–C7 level. The appearance of the C7 nerve root was unremarkable intraoperatively. Neck and left arm pain resolved after surgery.
There have been two previous reports of neurologic sequelae similar to those we described. 3,4
reported a patient with lancinating pain in the right arm after CESI as the epidural needle was withdrawn. Allodynia occurred 6 h later in the same extremity, associated with symptoms of increased sympathetic activity (cool, swollen arm). CESI was performed without sedation using a midline approach and the hanging-drop technique at the level of spinal cord impingement shown by computed tomography (C5–C6). No cerebrospinal fluid or blood was aspirated during the procedure, but, because a local anesthetic test dose was not used, dural puncture could not be excluded.
Hodges et al.3
reported two cases of CESI with use of fluoroscopic guidance in sedated patients. For both, injections were performed at the level of disk herniation shown by MRI (C5–C6) and resulted in subarachnoid placement of the needle, with return of cerebrospinal fluid. Although neither patient reported symptoms during the procedure, both awoke with new neurologic symptoms and subsequently had spinal cord injury shown by MRI and permanent neurologic sequelae. The authors concluded that use of fluoroscopic guidance does not prevent dural puncture and that excessive sedation may mask warning signs of neural impingement during CESI.
We report three patients in whom signs and symptoms of nerve injury developed after CESI using a standard technique without sedation. The cases of nerve injury occurred with different practitioners in different practice locations. No patient had evidence of dural puncture or intraneural injection. A test dose of local anesthetic was not used in any of these cases. However, a small dose of local anesthetic was mixed with the steroid in all cases, and none of the patients had evidence of spinal block after injection.
At least two mechanisms can explain the neuropathic pain experienced by these patients. Each patient reported painful paresthesia during CESI, suggesting that the needle tip was near an exiting nerve root. The symptoms were transient, and we administered methylpredisolone in each case. The methylprednisolone formulation used (Depo-Medrol; Pharmacia and Upjohn Companies, Kalamazoo, MI) is a particulate suspension of steroid in polyethylene glycol. This formulation has been implicated as a cause of neural toxicity. Perhaps deposition of the steroid near the exiting nerve root led to nerve root irritation.
Compromise of the spinal canal by a large herniated intervertebral disk or other anatomic abnormality that causes posterior displacement of the spinal cord toward the ligamentum flavum can result in spinal cord damage during epidural placement, without dural puncture. 2,3,5
Imaging studies in case 1 showed marked posterolateral displacement of the spinal cord caused by a herniated disk (figs. 1 and 2
); compression of the spinal cord was also shown by MRI in case 3. In case 2, there was no spinal cord effacement or displacement. However, this patient had undergone previous cervical fusion, and a small osteophytic ridge was present, extending from C5–C7 along the posterior part of the vertebral bodies. Loss of cervical mobility, combined with this osteophytic ridge, may have resulted in cervical cord displacement when the neck was flexed forward during CESI.
We hypothesize that neuropathic pain after CESI may result from either a direct nerve root irritation caused by the steroid solution or damage to the spinal cord or nerve roots without dural puncture by minor compression of neural elements. When a patient reports paresthesia during CESI, even if it is transient, the needle should be removed and placed at an adjacent level. This will avoid insertion of the steroid solution immediately adjacent to a nerve root and possibility of direct chemical neuritis. Clinicians should be fully aware of the nature of the disease, including any distortion of the spinal cord before performing CESI. Evaluation of imaging studies will allow rational decisions about where the injection should be positioned.
1. Abram SE, O’Connor TC: Complications associated with epidural steroid injections. Reg Anesth 1996; 21: 149–62
2. Manchiakanti L: Letter in cervical epidural steroid injection with intrinsic spinal cord by Hodges et al. Spine 1999; 24: 1170–1
3. Hodges SD, Castleberg RL, Miller T, Ward R, Thornburg C: Cervical epidural steroid injection with intrinsic spinal cord damage: Two case reports. Spine 1998; 23: 2137–42
4. Siegfried RN: Development of complex regional pain syndrome after cervical epidural steroid injection. A nesthesiology 1997; 86: 1394–6
5. Derby R: Point of view on cervical epidural steroid injection with intrinsic spinal cord damage by Hodges et al. Spine 1998; 23: 2141–2
© 2000 American Society of Anesthesiologists, Inc.