Published online before print November 15, 2012, doi: 10.3174/ajnr.A3393
AJNR 2012 33: E138

P.P. Ng^a, M.J. Wilder^a and P.A. Jenkins^a
aDepartment of Radiology
University of Utah Health Sciences Center
Salt Lake City, Utah

We read with great interest the article by Kranz et al in the August 2012 issue ofAJNR entitled, “CT Fluoroscopy-Guided Cervical Interlaminar Steroid Injections: Safety, Technique, and Radiation Dose Parameters.”¹ The authors concluded that CT fluoroscopy (CTF)– guided epidural steroid injections (ESIs) can be performed with a low rate of procedural complications and short procedural and CTF times, making these a practical alternative to using conventional fluoroscopy (CF).

While we agree that CTF provides superior visualization of anatomic structures in relation to the needle tip compared with CF, we do not believe that such precise visualization is necessary or increases the safety of the procedure. Our group of 3 neurointerventionalists has performed hundreds of interlaminar cervical ESIs by using CF for years without a single clinical complication.

We use single-plane CF (Inova 4100IQ; GE Healthcare, Milwaukee, Wisconsin), starting with a near-anteroposterior tube position until the needle is advanced to touch the spinal lamina, at which time the tube is rotated steeply in the contralateral oblique position to permit live CF visualization of needle advancement to the spinolaminar line. A loss-of-resistance technique is then used to advance the needle tip into the epidural space, which is confirmed with injection of 1 mL of contrast. Inadvertent subarachnoid injection is rare but is easily identified by rapid dissipation of contrast from the needle tip and pooling in the dependent subarachnoid space.

A review of our last 16 cases revealed a mean CF time of 0.9 minutes, procedural time of 10 minutes, and mean entrance skin air kerma (ESAK) of 16.7 mGy. If one assumes a mean shaded surface display of 64 cm, skin exposed area of 49 cm², beam quality of 4.3 mmAl, mean kilovolt (peak) of 70, and effective dose per milligray of ESAK to be 0.015–0.03 mSv/mGy (depending on tube angulation), the effective dose was 0.3 mSv, and the skin dose, 23 mGy. By comparison, by using the authors’ parameters of mean CTF time = 24 seconds with a median tube current of 70 mA and assuming a 1-cm CTF scan range, the effective dose would be approximately 2.1 mSv, depending on scan position. Moreover, if one assumes a mean neck circumference of 38 cm, the approximate skin dose of a CTF procedure, excluding planning scans, is close to 0.5 Gy. These doses are already 7 and 20 times the CF effective and skin doses to the patient respectively, without accounting for CT planning scans. Higher doses to the operator with CTF can also be expected and should be monitored.²

The authors stated that many operators using CF inject at the C7-T1 interspace or below where the spinal canal is wider, to increase safety, and that this location limits the amount of medication reaching stenotic levels more cranially. We disagree that this is a significant limitation due to the small volume and relatively short length of the cervical epidural space. We typically observe contrast migration for several spinal levels with just 1 mL of contrast injection. A nuclear medicine study revealed impressive cranially directed spread of technetium Tc99m-labeled red blood cells from the site of the epidural injection.³ Regardless, we also perform CF-guided cervical ESIs from C2–3 to C7-T1, unless MR imaging shows such severe spinal stenosis that the spinal cord directly abuts the ligamentum flavum; then, we target an adjacent level.

The authors describe waiting for 3–4 seconds after contrast injection before rescanning with CTF to detect washout in the event of vascular injection. We have observed cases of mixed injection into the epidural space and an epidural vein that would be missed by using this approach.

While we appreciate the authors’ work and the proposition to use CTF as an alternative to CF for cervical interlaminar ESIs, we believe that they overestimate the safety of CTF, particularly with regard to radiation exposure. CF provides adequate visualization of pertinent anatomy and allows superior confirmation of needle-tip position during contrast injection. CF procedures can be completed in less time, with less expense and a lower radiation dose without compromising procedural safety.

References

1. Kranz PG, Raduazo P, Gray L, et al. CT fluoroscopy-guided cervical interlaminar steroid injections: safety, technique, and radiation dose parameters. AJNR Am J Neuroradiol 2012; 33: 1221–24 » Abstract/FREE Full Text
2. Keat N. Real-time CT and CT fluoroscopy. Br J Radiol 2001; 74: 1088–90 » FREE Full Text
3. Szeinfeld M, Ihmeidan IH, Moser MM, et al. Epidural blood patch: evaluation of the volume and spread of blood injected into the epidural space. Anesthesiology 1986; 64: 820–22 » Medline

Reply

Published online before print November 15, 2012, doi: 10.3174/ajnr.A3430
AJNR 2012 33: E139

P.G. Kranz^a, P. Raduazo^a, L. Gray^a, R.K. Kilani^a and J.K. Hoang^a
aDepartment of Radiology
Duke University Medical Center
Durham, North Carolina

We very much appreciate the comments of Drs. Ng et al in response to our article “CT Fluoroscopy-Guided Cervical Interlaminar Steroid Injections: Safety, Technique, and Radiation Dose Parameters.”

Dr. Ng and colleagues question the necessity of CT guidance given the large number of successful injections they have performed under conventional fluoroscopic (CF) guidance, and suggest that injection above the C7-T1 level can be safe under CF guidance. We do not doubt that such procedures can be performed safely in experienced hands. Many authors would disagree with this opinion, however.^1⇓⇓–4Such authors cite a number of anatomic and technical reasons for their objections, as summarized in our manuscript, which may be overcome by using CT fluoroscopy (CTF). Nevertheless, the purpose of our article is not to advocate for 1 type of image guidance over another, because comparative safety and effectiveness data have not been established. Any position taken on this issue would be based on opinion (ie, level 5 evidence), at best. Nor is it our intent to supplant CF guidance for those interventionalists who successfully use it to perform injections, but rather to provide an alternative method for those with reservations about the safety of the procedure.

We fully agree with Dr. Ng and colleagues that radiation dose must be closely scrutinized with any procedure. We take issue with the dose estimates they propose, however. Dose comparison between CF and CTF is not straightforward, due to narrow beam collimation and scan lengths of <1 cm used with CTF. In CT, conversion factors have been used to convert dose-length product to effective dose (ED), but these are not reliable when scan lengths of less than 2 cm are used, and can result in the incorrect estimation of ED by up to a factor of 30.⁵ For this reason, estimates based on phantom studies may be more reliable. A direct comparison of lumbar CF and CTF guided injections at our institution using an anthropomorphic phantom demonstrated that the ED from the CTF portion of a lumbar injection was half that of the ED for the same procedure performed with CF.⁶ Ultimately, the small differential risk of radiation-related cancer induction between procedures must be balanced against any potential safety benefit related to the procedure, which as yet remains undetermined. Such overall safety comparisons are therefore premature.

Next, the authors argue that injection at C7-T1 is equivalent to injections at higher cervical levels, though the authors admit that despite this argument they do perform injections as high as C2–3 when anatomy permits. One may speculate as to the validity of this assertion, which is based on indirect evidence, and whether the delivered concentration of injected steroid medication is really equivalent when it reaches a level by remote epidural spread as opposed to direct injection. In the end, though, we are not aware of any study that directly addresses the fundamental question of clinical outcomes using these 2 different approaches.

As for the potential for mixed injection into an epidural vein and the epidural space, we have observed no adverse events to date involving such an injection in several hundred cervical interlaminar epidural steroid injections and several thousand lumbar interlaminar epidural steroid injections.

Drs. Ng et al nicely highlight several topics related to cervical steroid injections where there is insufficient evidence to draw firm conclusions. We believe that these questions will be best addressed with further investigation rather than opinion. The fact that these issues are unresolved, however, does not detract from the primary purpose of our manuscript, which was to report our initial experience with a technique that may have potential advantages—and potential drawbacks—compared with CF, as a basis for future experience.

References

1. Watanabe AT, Nishimura E, Garris J. Image-guided epidural steroid injections. Tech Vasc Interv Radiol 2002; 5: 186–93 » CrossRef » Medline
2. Derby R, Lee SH, Kim BJ, et al. Complications following cervical epidural steroid injections by expert interventionalists in 2003. Pain Physician 2004; 7:445–49 » Medline
3. Abbasi A, Malhotra G, Malanga G, et al. Complications of interlaminar cervical epidural steroid injections: a review of the literature. Spine (Phila Pa 1976) 2007; 32: 2144–51 » CrossRef
4. Reiter TM, Baker RM, DePalma MJ, et al. Epidural steroid instillation for cervical radiculopathy. In: DePalma MJ , ed. iSpine: Evidence-Based Interventional Spine Care. New York: Demos Medical Publishing 2011: 405–18 » Search Google Scholar
5. Huda W, Ogden KM, Khorasani MR. Converting dose-length product to effective dose at CT. Radiology 2008; 248: 995–1003 » Abstract/FREE Full Text
6. Hoang JK, Yoshizumi TT, Toncheva G, et al. Radiation dose exposure for lumbar spine epidural steroid injections: a comparison of conventional fluoroscopy data and CT fluoroscopy techniques. AJR Am J Roentgenol 2011;4: 778–82 » Search Google Scholar