Send to:

Choose Destination
See comment in PubMed Commons below
Pain Physician. 2012 Nov-Dec;15(6):515-23.

Digital subtraction angiography does not reliably prevent paraplegia associated with lumbar transforaminal epidural steroid injection.

Author information

  • 1Rehabilitation Institute of Chicago, Department of Physical Medicine and Rehabilitation, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA.


Digital subtraction angiography (DSA) has been touted as a radiologic adjunct to interventional neuraxial procedures where it is imperative to identify vascular compromise during the injection. Transforaminal epidural steroid injections (TFESI) are commonly performed interventions for treating acute and chronic radicular spine pain. We present a case of instantaneous and irreversible paraplegia following lumbar TFESI wherein a local anesthetic test dose, as well as DSA, were used as adjuncts to fluoroscopy. An 80-year-old man with severe lumbar spinal stenosis and chronic L5 radiculopathic pain was evaluated at a university pain management center seeking symptomatic pain relief. Two prior lumbar interlaminar epidural steroid injections (LESI) provided only transient pain relief, and a decision was made to perform right-sided L5-S1 TFESI. A 5-inch, 22-gauge Quincke-type spinal needle with a curved tip was used. Foraminal placement of the needle tip was confirmed with anteroposterior, oblique, and lateral views on fluoroscopy. Aspiration did not reveal any blood or cerebrospinal fluid. Digital subtraction angiography was performed twice to confirm the absence of intravascular contrast medium spread. Subsequently, a 0.5 mL of 1% lidocaine test dose was performed without any changes in neurological status. Two minutes later, a mixture of one mL of 1% lidocaine with 80 mg triamcinolone acetonide was injected. Immediately following the completion of the injection, the patient reported extreme bilateral lower extremity pain. He became diaphoretic, followed by marked weakness in his bilateral lower extremities and numbness up to his lower abdomen. The patient was transferred to the emergency department for evaluation. Magnetic resonance imaging (MRI) of the lumbar and thoracic spine was completed 5 hours postinjection. It showed a small high T2 signal focus in the thoracic spinal cord at the T7-T8 level. The patient was admitted to the critical care unit for neurological observation and treatment with intravenous methylprednisolone. Follow-up MRI revealed a hyper-intense T2 and short-tau inversion recovery signal in the central portion of the spinal cord beginning at the level of the T6 superior endplate and extending caudally to the T9-T10 level with accompanying development of mild spinal cord expansion. The patient was diagnosed with paraplegia from acute spinal cord infarction. At discharge to an acute inpatient rehabilitation program, the patient had persistent bilateral lower extremity paralysis, and incontinence of bowel and bladder functions. In the present patient, DSA performed twice and an anesthetic test dose did not prevent a catastrophic spinal cord infarction and resulting paraplegia. DSA use is clearly not foolproof and may not be sufficient to identify potentially life-or-limb threatening consequences of lumbar TFESI. We believe that this report should open further discussion regarding adding the possibility of these catastrophic events in the informed consent process for lumbar TFESIs, as it has for cervical TFESI. Utilizing blunt needles or larger bevel needles in place of sharp, cutting needles may minimize the chances of this event occurring. Considering eliminating use of particulate steroids for TFESI should be evaluated, although the use of nonparticulate agents remains controversial due to the perception that their respective duration of action is less than that of particulate steroids.

[PubMed - indexed for MEDLINE]
Free full text
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Pain Physician
    Loading ...
    Write to the Help Desk