Insufficient Clarity of Statement 6 in the Consensus Opinions to Prevent Neurologic Complications after Epidural Steroid Injections

I read with interest the consensus opinions to prevent neurologic complications after epidural steroid injections in the May 2015 issue.¹ 

The working group deserves the thanks of practitioners of interventional pain medicine. The statements serve to direct care that has been complicated by statements from the Food and Drug Administration and recent evidence regarding the safety and efficacy of steroid injections.

The statement’s utility lies in the clarity of the recommendations that can be defined as current “standards of care.”

For this reason, statement number 6 should be reconsidered. It reads, “No cervical IL ESI [interlaminar epidural steroid injection] should be undertaken, at any segmental level, without reviewing, before the procedure, prior imaging studies that show there is adequate epidural space for the needle placement at the target level.”

Certainly, no interlaminar cervical epidural steroid injection should be undertaken at a level whose epidural space has been disrupted by surgery, infection, or metastasis.

However, neither clear definition of “adequate epidural space” is given nor is readily available from the literature or reference texts. The supporting references for statement number 6 include the following:

• - Hodges et al.²  present two case reports of spinal cord injury during a cervical epidural steroid injection placed at C5-C6, 1 cm left of midline while the patients were sedated with propofol.

• - Aldrete et al.³  review the imaging of 100 patients and determine the mean depth of the epidural space to be 0.4 cm at C7-T1.

• - Hogan⁴  reviews frozen cross-sections of 26 adult cadaveric cervical spines and finds “no posterior epidural space above C7-T1 level.” As well, the investigators find “the ligamenta flava…failed to fuse in about half of the cervical and thoracic levels examined.”

• - Goel and Pollan⁵  determined that 2 to 4 cc of contrast injected in 34 patients at C7/T1 reached to C3 bilaterally in 100% of patients. Interestingly, no mention of establishing an adequate epidural space is mentioned before the placement of the needles.

Most importantly, the most recent editions of texts by Benzon and Rathmell, two authors of the statement, do not include clear instruction for “determining that a cervical epidural space is adequate for needle placement at the target level.”

Benzon et al.⁶  include the advice: “It is almost always advisable to have magnetic resonance imaging (MRI) available before performing any cervical epidural injections. Once the safety of potential interlaminar approach is verified, a chlorhexidine alcohol preparation is performed and sterile drapes placed.” (No figures accompany these statements.)

Rathmell and Nelson⁷  state, “Caution should also be taken to avoid interlaminar epidural injection at any level where there is effacement of the epidural space (e.g., complete effacement of the epidural space and cerebrospinal fluid column surrounding the spinal cord within the thecal sac occurs in high-grade spinal stenosis, particularly that due to a large central or paramedian disc herniation).”

A review of several other available reference texts on MRI of the spine yields no clear direction for the visualization or measurement of the cervical epidural space.^8–10 

The International Spine Intervention Society Practice Guidelines (2013) mention this concern when presenting a case that resulted in a complication from direct cord injury by the needle. However, a specific process for evaluating imaging is not included, other than the statement, “preprocedural MRI was not taken into account.”¹¹ 

Perhaps, the authors of the statement mean that at least a 1-mm white epidural space is clear on MRI T1 axial and sagittal cuts at the level of needle entry. If so, must the MRI be less than a year old? Frequently, sagittal cuts do not include the exact laterality of the planned track of the needle. Frequently, practitioners have access to a written report by the radiologist, but not the images themselves. Frequently, the MRI has been read, and the presence of the epidural space not been commented upon. How should computed tomography images be evaluated for patients who cannot undergo MRI? Should new images be required if symptoms have changed in severity or distribution since previous imaging? Are practitioners of interventional pain required to interpret a millimeter-thick absence or presence of an epidural space on MRI or is this the appropriate responsibility of the radiologist?

Please clarify this recommendation, and offer a clear process for establishing adequate epidural space at the level of the injection, as this clarity would match the mandate that no such procedure should be attempted without certainty. Such clarification would include the type and timing of imaging and the exact measurements that, if not obtained, would render the injection contraindicated.

If such a clarification is not forthcoming, statement number 6 should be regarded as an important suggestion—not a mandate and not a standard of care.

The author declares no competing interests.