OUP user menu

Temporary Neurologic Deficit After Cervical Transforaminal Injection of Local Anesthetic

Michael Karasek MD, Nikolai Bogduk MD, PhD, DSc
DOI: http://dx.doi.org/10.1111/j.1526-4637.2004.04028.x 202-205 First published online: 1 June 2004


Objective To describe the effects of spinal cord block after injection of local anesthetic into a cervical radicular artery.

Design Case report.

Setting Neurology practice specializing in spinal pain.

Interventions A patient underwent a C6–7 transforaminal injection. Contrast medium indicated correct and safe placement of the needle.

Results After the injection of local anesthetic, the patient developed quadriplegia. The injection was terminated. The neurologic impairment resolved after 20 minutes observation.

Conclusion Despite correct placement of the needle for a cervical transforaminal injection, injectate may nevertheless enter a cervical radicular artery. Whereas local anesthetic, so injected, appears to have only a temporary effect on spinal cord function, particulate steroids may act as an embolus and cause permanent impairment.

  • Cervical
  • Radicular Pain
  • Treatment
  • Transforaminal Injection
  • Steroids


Cervical radicular pain has lacked a proven conservative treatment. Controlled studies have shown that exercises and traction have no attributable effect beyond that of natural history [1–3]. Surgery is the only conventional option for treating intractable cervical radicular pain [4,5].

Transforaminal injections of corticosteroids were developed as a means of treating cervical radicular pain without resorting to surgery. They have been promoted on the basis of observational studies, which provide modest but encouraging results [6–8]. However, no controlled studies have proven their efficacy; but this has not prevented their seemingly widespread adoption.

Meanwhile, the use of transforaminal injections of corticosteroids has been punctuated by reports of serious neurologic complications. One case report described death from spinal cord injury following cervical transforaminal injections [9]. Other publications have alluded to several other cases that are sub judice, for which reason they have not been described in the medical literature [10,11].

Unknown is the mechanism of injury in these cases. Although one report describes perforation of a vertebral artery [12], in other cases, the needle appears to have been correctly placed in the intervertebral foramen, where it should not encounter the vertebral artery. In such cases, it has been proposed that inadvertent injection into a radicular artery is the mechanism of injury [10,11].

In two cases, filling of a cervical radicular artery was recognized and documented following the injection of contrast medium during cervical transforaminal injection [10,11]. This observation provided grounds for the conjecture that neurologic injury could occur if particulate steroid preparations were injected into a radicular artery and acted as an embolus.

One report [10] recommended, in order to avoid this complication, that local anesthetic be injected as a test injection before any steroid preparation. However, it remarked that the effects of intra-arterial injection of local anesthetic into a radicular artery were unknown. The present report describes these effects.

Case Report

A 55-year-old woman had a good response to the cervical transforaminal injection of corticosteroids on a previous occasion. She was referred by her rehabilitation physician for another injection in order to reinstate relief of her cervical radicular pain.

An injection was initiated at the C6–7 level on the right. A spinal needle was accurately introduced along an oblique view into the target intervertebral foramen (Figure 1). In an anteroposterior (A–P) view, it was placed no further than half way across the width of the articular pillar (Figure 2). A test dose of contrast medium showed appropriate filling of the intervertebral foramen with no vascular injection (Figure 2).

Figure 1

A right anterior oblique view of the cervical spine, showing a needle in place, in the C6–7 intervertebral foramen, low in front of the superior articular process of C7.

Figure 2

An AP view of the cervical spine, showing (i) a needle in the right C6–7 intervertebral foramen, with its tip (arrow) no further than the sagittal bisector of the articular pillar; and (ii) the appearance of a test-dose of contrast medium surrounding the C7 spinal nerve.

A bolus of local anesthetic (0.8 mL of 2% lidocaine) was injected prior to the intended injection of corticosteroid. This injection flushed some of the previously injected contrast medium from the intervertebral foramen. No vascular filling was consistent with arterial injection, but close inspection of the angiogram did reveal filling of what appeared to be epidural and vertebral veins on the contralateral side (Figure 3). At no time during the procedure did the contrast medium indicate that intrathecal injection had occurred. At all times, the main deposit of the contrast medium remained around the target spinal nerve and outside its dural sleeve (Figure 3).

Figure 3

An AP view of the cervical spine after the injection of contrast medium into the C6–7 intervertebral foramen, and following injection of local anesthetic. Filling is evident of small vertebral veins (1). Faintly evident (2) is a vessel within the vertebral canal opposite the C5–6 disc space.

Approximately 60 seconds after the local anesthetic had been injected, the patient reported feeling unwell. The procedure was terminated, and the patient was assessed. Over the next 2–3 minutes, the patient reported weakness affecting all four limbs. Examination by the neurologist who performed the procedure revealed paralysis of both upper and lower limbs, loss of pin-prick sensation, but preservation of fine touch and proprioception. Respiration was not affected.

The patient was transferred to a recovery suite and closely monitored. After 20 minutes, all symptoms resolved. The patient had no persisting symptoms or signs and suffered no lasting impairment.


This case reinforces the observation that, despite correct technique being used, inadvertent intra-arterial injection can occur during cervical transforaminal injections. In this case, intra-arterial injection was not evident angiographically. However, the nature of the symptoms that the patient developed was clearly consistent with anesthesia of the spinal cord. The neurologic signs were restricted to tracts of the anterior and anterolateral quadrants of the spinal cord. This correlates with the distribution of the anterior spinal artery and indicates that a cervical radicular artery must have been infused, despite its angiographic invisibility. The fact that the angiogram revealed contralateral filling of veins is a suspicious, but indecisive, finding. Venous uptake, rather than arterial injection, could account for this. Nevertheless, it is also consistent with venous washout after a rapid intra-arterial injection.

This case vividly demonstrates the effects of anesthesia of the spinal cord. Motor signs dominated, consistent with long tracts having being anesthetized. Fortunately, and mercifully, the effects of the local anesthetic were temporary and brief.

The signs exhibited by the patient in this case are consonant with those developed by patients who have suffered permanent injury following cervical transforaminal injection of corticosteroids. In that regard, the present case provides further circumstantial evidence of the proposed mechanism of injury following transforaminal injection of corticosteroids. The signs developed by the patient were widespread, affecting all four limbs, and developed rapidly. This is not consistent with intraneuronal uptake of local anesthetic [13]. Nor can the signs be attributed to intrathecal injection. All radiographs showed no evidence of intrathecal spread. The only feasible mechanism for the neurologic impairment is injection into a cervical radicular artery that reinforced the anterior spinal artery.

The case underscores the merit and wisdom of injecting local anesthetic separate from, and before, any corticosteroid. It would appear that local anesthetic has a temporary pharmacological effect if injected to the spinal cord. The same cannot be presumed for particulate steroid preparations, which might act physically, as an embolus.

As recommended by Baker et al. [10], a preliminary injection of local anesthetic allows the operator to assess if the injectate has reached the spinal cord. Operators who take this step should ensure that they allow sufficient time to assess the patient for any response, and assess them care-fully. Numbness may not be the cardinal feature (because of sparing of the posterior columns). Paralysis is the cardinal feature of anterolateral spinal cord anesthesia; but this may not be evident if the patient does not attempt movements. Meanwhile, respiration may not be affected if low cervical spinal levels only are affected.

On the other hand, practitioners might take heed of this case. Although misadventure was avoided, it might have occurred had the operator presumptively injected corticosteroids. Cervical transforaminal injection of steroids is a hazardous procedure despite correct technique being used. If this procedure is to continue to be used, consummate care and vigilance is required.

An alternative technique has been described [14], in which a catheter is introduced to the target intervertebral foramen from below using an interlaminar insertion. Allegedly, this technique is safer than the transforaminal approach. However, it has not been used extensively, and its safety has not yet been demonstrated in a large number of patients. Nevertheless, some operators might prefer to adopt this technique in an effort to avoid possible injection into a radicular artery.


View Abstract