Abstract

Background . Lumbar transforaminal epidural steroid injections (TFESIs) are often used in the treatment of radicular pain. In light of safety concerns, many practitioners have proposed adopting the retrodiscal (infraneural) approach with the needle tip positioned into Kambin’s triangle. With this technique, the needle may inadvertently be directed too far ventrally and enter the intervertebral disc. In addition, the risk of subarachnoid or subdural extra-arachnoid injection may be higher with this technique as well.

Objective . To determine the incidence of inadvertent intradiscal, intrathecal, and vascular injections during the performance of retrodiscal TFESI.

Study Design . Retrospective review

Methods . Retrospective review of all retrodiscal approach TFESIs performed from July 2012 to August 2014 by two of the authors (DL and SH).

Results . A total of 257 retrodiscal transforaminal injections were performed. There were no neurologic complications. There were no cases of discitis. Inadvertent intradiscal injections occurred in 12/257 injections, 4.7% (95% CI 2.1–7.3%). Intrathecal injections occurred in 8/257 injections, 3.1% (95% CI 0.99– 5.23%). Three were subarachnoid (SA), four were subdural extra-arachnoid (SDXA), and one was both SA and SDXA. Vascular injections occurred in 17/257, 6.6% (95% CI 3.6–9.6%).

Conclusion . This retrospective review demonstrates that a relatively high rate of inadvertent intradiscal injections occurs in the performance of the retrodiscal approach for TFESI. This has significant implications in terms of the potential risk of disc injury induced by the needle puncture. The high incidence of intrathecal injections may also be of great concern depending upon the injectate delivered.

Introduction

Lumbar transforaminal epidural steroid injections (TFESIs) are often used in the treatment of radicular pain [ 1 ]. Although the subpedicular approach is most often utilized for lumbar transforaminal injections, the supraneural, retroneural, or infraneural approaches may also be used [ 2 ]. Many practitioners favor one approach over another based on training, personal experience, and safety concerns.

Rare, but catastrophic, complications have been seen with lumbar transforaminal epidural steroid injections, one of which is spinal cord infarction [ 3 , 4 ]. The most likely etiology is injection of particulate matter into a radiculomedullary artery, which communicates with, and occludes, the anterior spinal artery [ 3 , 4 ]. Prior studies have demonstrated the most frequent vertebral levels and intraforaminal locations of radiculomedullary arteries, in particular the artery of Adamkiewicz (AKA) [ 5 , 6 ]. The AKA is found typically T8-L1 but may be located anywhere within the thoracic or lumbar region and even have been localized as inferior as S2 [ 5 , 6 ]. The most common foraminal location of the AKA is the upper third of the foramen and least common is the lower fifth [ 5 ]. The relationship of AKA and other radiculomedullary arteries to the dorsal root ganglion and ventral nerve root has also been studied. The arteries were located most frequently anterior and superior to the neural structure [ 6 ]. This location is the needle position for the most commonly employed lumbar transforaminal epidural approach, the subpedicular position, also referred to as the classic “safe triangle” technique [ 2 ]. The term “safe” is in reference to the location of the neural but not vasculature structures.

In light of safety concerns, many practitioners have proposed adopting the retrodiscal (infraneural) approach with the needle tip positioned into Kambin’s triangle [ 7–9 ]. Kambin’s triangle is defined as the right triangle bordered inferiorly by the superior end plate, medially by the traversing nerve root and thecal sac and laterally (obliquely) by the hypotenuse formed by the exiting nerve root [ 10 ]. Although this technique may place the needle in a position with less likelihood of radiculomedullary intra-arterial cannulation [ 7 ], there may be other risks associated with this needle tip placement. Using the retrodiscal approach, the needle tip may unintentionally be placed too far ventrally and enter the intervertebral disc [ 11 ]. Chang et al. performed a retrospective review of 50 cases of retrodiscal approach TFESIs and found 3 (6%) cases of disc entry [ 8 ]. In addition, although not previously investigated, the risk of subarachnoid or subdural injection may be higher with this technique, because theoretically, the axillary pouch of the nerve-root sleeve may encroach into Kambin’s triangle.

The objective of this study is to help determine the incidence of inadvertent intradiscal, intrathecal, or vascular injections during the performance of retrodiscal TFESI.

Methods

This study was approved by an independent institutional review board, Sterling IRB. No external funding or other means of support was received for this study. The investigation was conducted as a retrospective review of all retrodiscal approach of TFESIs consecutively performed from July 2012 to August 2014 by two of the authors (DL and SH). The written procedure reports were reviewed for documentation of the presence of intradiscal, intrathecal, or vascular flow. When required, the images saved from the procedure were reviewed for clarification.

Performance of retrodiscal TFESI procedure: The procedure was performed by two of the investigators (DL and SH). Both are board certified in physical medicine and rehabilitation, fellowship trained in spine injections, and serve as instructors for spinal injection procedures at the national level. Both investigators perform the “safe triangle” TFESI approach more frequently and are experienced discographers.

The decision to perform the retrodiscal approach rather than the “safe triangle” technique for a particular TFESI was based upon several factors. The most common reason was the unavailability of non-particulate steroid. During the time period of the study, the authors’ preferred steroid for TFESI, preservative-free dexamethasone (non-particulate), was intermittently commercially unavailable. In light of the possible safety advantage of the retrodiscal approach in avoiding a radicular artery [ 5–9 ], the authors typically chose this technique if particulate steroid was being used. The retrodiscal approach was also chosen at times, independent of the type of steroid being injected. In many instances, Kambin’s triangle position was felt to be more appropriate owing to the patient’s specific anatomy. Such instances include a foraminal disc herniation compressing the nerve root superiorly against the pedicle or a large zygapophyseal joint in combination with an inferiorly extending osteophyte from the medial portion of the transverse process. Kambin’s triangle approach was also occasionally chosen when an initial injection with a subpedicular technique provided an inadequate clinical response.

Prior to the procedure, the MRI was reviewed on every patient. The size and location of any disc herniation was taken into account for needle positioning. In addition, the physicians screened for the presence of any spinal anomaly such as nerve root sleeve cyst or conjoint nerve roots which might interfere with the performance of the retrodiscal approach. A 22- or 25-gauge Quincke (Becton-Dickinson, Franklin Lakes, NJ) spinal needle with stylet with a slight bend placed on the tip was utilized. The area was cleansed with chlorhexidine and a fenestrated drape was placed. Using an initial oblique approach similar to discography technique, the needle was advanced to contact the lateral portion of the superior articular process (SAP) of the inferior vertebrae. Anterior-posterior (AP) and lateral views were obtained and the needle was directed to pass just beyond the SAP. The position was slightly adjusted on lateral first then on AP imaging ( Figure 1 ). On the lateral image the needle was advanced into the anterior third of the foramen, no further ventral than the posterior vertebral line (representing the posterior aspect of the intervertebral disc). On AP view, the needle was placed in a position no further medial than the inter-pedicular line. Needle advancement beyond the SAP was very slow with attention to change in resistance should the outer annulus be unintentionally encountered. The stylet was removed and via tubing, contrast agent, 2-4 ml, was then injected. The flow pattern was then determined and, following the procedure, documented in the procedure report including any aberrant pattern such as intradiscal, vascular, subdural extra-arachnoid, or subarachnoid ( Figures 2–5 ). If the pattern was unclear, further contrast was injected. The needle was repositioned as required. After an appropriate epidurogram was obtained, a test dose of local anesthetic was administered. After a 1-minute waiting period, a steroid preparation with or without further anesthetic was administered.

Figure 1

A, B, C, D: AP and lateral retrodiscal TFESI before and after contrast injection with epidural flow pattern.

Figure 1

A, B, C, D: AP and lateral retrodiscal TFESI before and after contrast injection with epidural flow pattern.

Figure 2

A, B, C, D: AP and lateral retrodiscal TFESI before and after contrast injection with intradiscal flow pattern.

Figure 2

A, B, C, D: AP and lateral retrodiscal TFESI before and after contrast injection with intradiscal flow pattern.

Figure 3

AP retrodiscal TFESI with vascular flow pattern.

Figure 3

AP retrodiscal TFESI with vascular flow pattern.

Figure 4

A, B: AP and lateral retrodiscal TFESI with subarachnoid flow pattern.

Figure 4

A, B: AP and lateral retrodiscal TFESI with subarachnoid flow pattern.

Figure 5

A, B, C: AP and lateral retrodiscal TFESI with subdural extra-arachnoid flow pattern. Note the cephalad extent of the contrast from the lumbar injection into the subdural extra-arachnoid space with flow into the thoracic region.

Figure 5

A, B, C: AP and lateral retrodiscal TFESI with subdural extra-arachnoid flow pattern. Note the cephalad extent of the contrast from the lumbar injection into the subdural extra-arachnoid space with flow into the thoracic region.

If disc entry was encountered, 20 mg of gentamycin in an additional 2 ml of contrast agent was injected into the disc for discitis prophylaxis.

Results

A total of 257 retrodiscal transforaminal injections were performed during the 2-year study period by two of the investigators (DL and SH). 214 were performed by DL and 43 were performed by SH. 39 of the injections were bilateral. 14 procedures were two level. Two injections were performed at T12/L1, 7 at L1/2, 16 at L2/3, 40 at L3/4, 101 at L4/5, and 91 at L5/S1. 132 injections were left sided and 125 were right sided.

There were no neurologic complications. There were no cases of discitis.

Unintentional intradiscal injections occurred in 12/257 injections, 4.7% (95% CI 2.1–7.3%).

Intrathecal (including subarachnoid and/or subdural extra-arachnoid) injections occurred in 8/257 injections, 3.1% (95% CI 0.99–5.23%). Three were subarachnoid (SA), four were subdural extra-arachnoid (SDXA), and one was both SA and SDXA.

Vascular injections occurred in 17/257, 6.6% (95% CI 3.6–9.6%).

Discussion

This retrospective review demonstrates a relatively high rate of intradiscal injection of 4.7% using a retrodiscal TFESI approach. This is in contrast to the relatively low rate of intradiscal injection with the subpedicular, “safe triangle” technique of 0.17% in a review of over 8,000 cases [ 12 ] and 2.3% in a review of 251 injections [ 13 ]. Our findings are in accordance with the smaller review by Chang et al. of 6% disc entry based upon a review of 50 retrodiscal TFESIs [ 8 ].

Needle entry into the disc may be of concern for several reasons. The complication of infectious discitis is a rare occurrence following provocation discography [ 14 , 15 ]. To our knowledge, there have been no reports of discitis following an inadvertent intradiscal injection. Like the study of Plastaras et al., we utilized prophylactic antibiotic for our cases of intradiscal injection [ 12 ]. The second, and possibly more concerning issue regarding needle disc entry, may be the increased risk for future disc degeneration or herniation induced by needle puncture as suggested by the work of Carragee et al. [ 16 ]. This safety concern may offset the theoretical benefit of the retrodiscal approach, the decreased risk of encountering a radiculomedullary artery.

Proponents of the retrodiscal TFESI approach submit that the technique has a safety advantage in that the needle is placed in a foraminal position that is distant to the typical location of a radiculomedullary artery [ 9 , 6 ]. The catastrophic neurologic events that have occurred during TFESIs are felt to be the result of injection of particulate steroid into a radiculomedullary artery using the “safe triangle” approach [ 3 , 4 ]. More recently, many physicians have adopted the use of non-particulate steroid for TFESIs due to these safety issues. Although it was generally believed that particulate steroids were superior in efficacy, recent trials have demonstrated essentially equivalent results from non-particulate and particulate preparations [ 17 , 18 ]. Although there have been no reported cases of paralysis using non-particulate steroid at the time of this manuscript, the number of injections performed with non-particulate steroid are likely too few, at this point, to confirm, with any great confidence, the elimination of the risk of a radiculomedullary artery injection. Therefore, currently, the retrodiscal approach likely confers some safety advantage over the classic “safe triangle” approach with respect to the extremely rare but catastrophic risk of neurologic injury from a radiculomedullary artery injection, even with the use of non-particulate steroid. The use of non-particulate steroid with the “safe triangle” approach, however, is increasing. Hopefully, this will demonstrate an absence of catastrophic events from the non-particulate preparation, and, as such, the safety advantage of the retrodiscal technique may no longer exist.

Surprising was the relatively high, 3%, incidence of intrathecal injections in this review of retrodiscal injections. Although rare, with the “safe triangle” approach, case reports of subdural and sub arachnoid injections have been published [ 19 ]. A recent review was performed finding an overall very low incidence of dural punctures, 0.04% for a total of nearly 15,000 transforaminal injections (lumbar, thoracic, and cervical levels) [ 20 ]. It is possible, however, that the incidence of intrathecal injections is equally high in the “safe triangle” approach but goes unrecognized. In particular, subdural extra-arachnoid, flow may be difficult to identify especially if combined with a partial epidural pattern.

Although not specifically related to epidural steroid injections, the high dural puncture rate also has implications for discography and potential future intradiscal treatments. The technique and needle trajectory of the retrodiscal approach is identical to the approach for disc access (except the needle tip is intended to remain outside the disc in the retrodiscal TFESI). In light of the findings of this study, dural puncture likely occurs with discography/disc injections with equivalent frequency to the retrodiscal TFESI approach of 3.1% (95% CI 0.99–5.23%). This may be of consequence in the delivery of future intradiscal therapeutics, particularly if a substance to be injected is contraindicated in the subarachnoid space. Historically, the severe neurologic complications seen with chymopapain disc injection were likely due to dural puncture during disc access with the lateral, extrapedicular approach [ 21 ]. Interestingly, Dabezies and Murphy published a case series in which they removed the stylet just prior to entering the disc, as a safety measure for chymopapain intradiscal injections and described three cases of CSF return with the needle tip within Kambin’s triangle [ 21 ].

As previously discussed, one of the most concerning risks with epidural injection is radiculomedullary arterial injection leading to spinal cord infarct. The perceived benefit of a retroneural approach is a lower rate of encountering a radiculomedullary artery. Our study found intravascular uptake during a retroneural approach for TFESI at 7%. This is similar to the findings by Furman and colleagues demonstrating vascular uptake during lumbar subpedicular TFESI at approximately 8% [ 22 ]. It should be noted, however, that no distinction in our study was made between arterial and venous flow based on the difficultly of reliably determining the difference. It is presumed that based upon prior studies [ 5 , 6 ], the vascular flow encountered in this review was not a radiculomedullary artery but the current study was not designed to confirm this.

Beyond safety issues in choosing a transforaminal approach, many physicians believe one approach is superior to another for a variety of reasons. At this point in time, it remains unclear if the retrodiscal or traditional subpedicular technique is superior in efficacy. Two small trials have investigated this issue without demonstrating any significant clinical difference between the two approaches [ 11 , 23 ].

This study does have limitations. The retrospective nature does not allow for the control of bias that a prospective trial may provide. The decision to perform a retrodiscal approach rather than subpedicular might be considered a potential bias. Although the authors did not collect data on the rationale for each TFESI approach, most commonly, the retrodiscal technique was performed due to the non-availability of non-particulate steroid. This is very unlikely to confound the findings of this study. However, the investigators also used the retrodiscal approach when it was felt to be more appropriate due to a patient’s specific anatomy or occasionally, during a subsequent injection following a poor clinical response to a prior subpedicular approach. The physician’s decision to use the retrodiscal approach in these two latter situations could potentially be viewed as confounding.

It is unclear if our findings are applicable to practitioners with less experience in the practice of discography as the retrodiscal technique is similar to that used for disc access. This review also did not address if a different type of needle, e.g., blunt tip, could decrease the incidence of unintentional disc or thecal puncture. Future research may be needed to address this issue.

Conclusion

This retrospective review demonstrates that a relatively high rate of unintentional intradiscal injection occurs in the performance of the retrodiscal approach for TFESI. This is in contrast to low rate of intradiscal injection with the traditional subpedicular approach [ 13 ]. This has significant implications in terms of the potential risk of disc injury induced by the needle puncture [ 16 ]. The incidence of intrathecal injections was also relatively high. If unrecognized, this may also be of great concern depending upon the injectate delivered.

Acknowledgments

The authors would like to thank Tabathia Thacker, Christina Stone, Diana Carnahan, and Kim Gerrity for their generous assistance with this project. The authors also wish express our gratitude to the editorial staff for their guidance in improving this manuscript.

Conflicts of interest: Ms Corcoran, after completion of the study, discontinued employment at APM Spine and Sports, and is currently employed by Medtronic.

References

1
MacVicar
J
King
W
Landers
MH
Bogduk
N.
The effectiveness of lumbar transforaminal injection of steroids: A comprehensive review with systematic analysis of the published data
.
Pain Med
 
2013
;
14
:
14
28
.
2
International Spine Intervention Society.
Lumbar transforaminal access . In:
Bogduk
N
, ed.
Practice Guidelines for Spinal Diagnostic and Treatment Procedures
  ,
2nd edition
.
San Francisco, CA
:
International Spine Intervention Society
;
2012
:
459
536
.
3
Kennedy
DJ
Dreyfuss
P
Aprill
CN
Bogduk
N.
Paraplegia following image-guided transforaminal lumbar spine epidural steroid injection: Two case reports
.
Pain Med
 
2009
;
10
:
1389
94
.
4
Glaser
SE
Falco
F.
Paraplegia following a thoracolumbar transforaminal epidural steroid injection
.
Pain Physician
 
2005
;
8
:
309
14
.
5
Murthy
NS
Maus
TP
Behrns
CL.
Intraforaminal location of the great anterior radiculomedullary artery (artery of Adamkiewicz): A retrospective review
.
Pain Med
 
2010
;
11
(
12
):
1756
64
.
6
Kroszczynski
AC
Kohan
K
Kurowski
M
Olson
TR
Downie
SA.
Intraforaminal location of thoracolumbar anterior medullary arteries
.
Pain Med
 
2013
;
14
(
6
):
808
12
.
7
Glaser
SE
Shah
RV.
Root cause analysis of paraplegia following transforaminal epidural steroid injections: The ‘unsafe' triangle
.
Pain Physician
 
2010
;
13
(
3
):
237
44
.
8
Chang
J.
Lumbar transforaminal epidural steroid injections through a reverse safe triangle approach-technical consideration
.
Pain Med
 
2011
;
12
(
9
):
1439
40
. (International Spine Intervention Society 2011 Abstract).
9
Jasper
JF.
Lumbar Retrodiscal Transforaminal Injection
.
Pain Physician
 
2007
;
10
:
501
10
.
10
Kambin
P.
Arthroscopic microdiskectomy
.
Mt Sinai J Med
 
1991
;
58
:
159
64
.
11
Kim
C
Moon
CJ
Choi
HE
Park
Y.
Retrodiscal approach of lumbar epidural block
.
Ann Rehab Med
 
2011
;
35
(
3
):
418
26
.
12
Plastaras
CT
Casey
E
Goodman
BS
, et al.   .
Inadvertent intradiscal contrast flow during lumbar transforaminal epidural steroid injections: A case series examining the prevalence of intradiscal injection as well as potential associated factors and adverse events
.
Pain Med
 
2010
;
11
(
12
):
1765
73
.
13
Hong
JH
Kim
SY
Huh
B
Shin
HH.
Analysis of inadvertent intradiscal and intravascular injection during lumbar transforaminal epidural steroid injections: A prospective study
.
Reg Anesth Pain Med
 
2013
;
38
(
6
):
520
5
.
14
Fraser
RD
Osti
OL
Vernon-Roberts
B.
Discitis after discography
.
J Bone Joint Surg Br
 
1987
;
69
:
26
35
.
15
Osti
OL
Fraser
RD
Vernon-Roberts
B.
Discitis after discography. The role of prophylactic antibiotics
.
J Bone Joint Surg Br
 
1990
;
72
:
271
4
.
16
Carragee
EJ
Don
AS
Hurwitz
EL
, et al.   .
Does discography cause accelerated progression of degeneration changes in the lumbar disc: A ten-year matched cohort study
.
Spine
 
2009
;
34
:
2338
45
.
17
Kennedy
DJ
Plastaras
C
Casey
E
, et al.   .
Comparative effectiveness of lumbar transforaminal epidural steroid injections with particulate versus nonparticulate corticosteroids for lumbar radicular pain due to intervertebral disc herniation: A prospective, randomized, double-blind trial
.
Pain Med
 
2014
;
15
(
4
):
548
55
.
18
El-Yahchouchi
C
Geske
JR
Carter
RE
, et al.   .
The noninferiority of the nonparticulate steroid dexamethasone vs the particulate steroids betamethasone and triamcinolone in lumbar transforaminal epidural steroid injections
.
Pain Med
 
2013
;
14
(
11
):
1650
7
.
19
Goodman
BS
Bayazitoglu
M
Mallempati
S
Noble
BR
Geffen
JF.
Dural puncture and subdural injection: A complication of lumbar transforaminal epidural injections
.
Pain Physician
 
2007
;
10
(
5
):
697
705
.
20
El-Yahchouchi
CA
Plastaras
CT
Maus
TP
, et al.   .
Complication rates of transforaminal and interlaminar epidural steroid injections: A multi-institutional study
.
Pain Med
 
2014
;
15
(
8
):(abstract only).
21
Dabezies
EJ
Murphy
CP.
Dural puncture using the lateral approach for chemonucleolysis
.
Spine
 
1985
;
10
(
1
):
93
22
Furman
MB
O'Brien
EM
Zgleszewski
TM.
Incidence of intravascular penetration in transforaminal lumbosacral epidural steroid injections
.
Spine
 
2000
;
25
(
20
):
2628
32
.
23
Park
CH
Lee
SH
Park
HS.
Lumbar retrodiscal versus post-ganglionic transforaminal epidural steroid injection for the treatment of lumbar intervertebral disc herniations
.
Pain Physician
 
2011
;
14
(
4
):
353
60
.