Pain Medicine 2016; 17: 1628–1633
doi: 10.1093/pm/pnv075
Reliability and Accuracy of MRI Laminar Angle
Measurements to Determine Intra-Procedural
Contralateral Oblique View Angle for Cervical
or Thoracic Interlaminar Epidural
Steroid Injections
David S. Levi, MD* Scott Horn, DO,* and
Alexandra Collado, BS*
*APM Spine and Sports Physicians, Virginia Beach,
Virginia, USA
Correspondence to: David S. Levi, APM Spine and
Sports Physicians, Virginia Beach, VA 23502, USA.
Tel: 757-422-2966; Fax: 757-422-4563; E-mail:
[email protected].
Funding sources: No funding or other forms of support were provided for this study.
Disclosure and conflicts of interest: There are no
conflicts of interest to report.
Abstract
Background and Objective. Contralateral oblique (CLO)
angle view has been a useful addition to standard
views in fluoroscopically guided interlaminar epidural
injections. Determination of the appropriate CLO angle
is paramount in the usefulness of this technique.
Using MRI laminar angle measurements as a pre-procedural guide for the intra-procedural fluoroscopic CLO
angle has been proposed. The purpose of this study
was to help determine if using axial MRI laminar measurements prior to a cervical or thoracic epidural steroid
injection would be useful in predicting the appropriate
fluoroscopic CLO angle.
Study Design. A retrospective review was performed for patients who underwent cervical or
thoracic interlaminar injections. In the performance
of interlaminar injections, the authors had routinely
determined the true fluoroscopic contra-lateral oblique angle after epidural access was confirmed, for
use during any potential future injections. The fluoroscopic CLO angle measurements were obtained
from a chart review and compared blindly to each
patient’s MRI axial laminar angle measurements.
Results. 34 injections were included. Inter-rater
reliability comparing the two authors’ MRI angle
measurements was considered fair, ICC 5 0.395.
Accuracy was only 57% comparing MRI laminar
angle measurements to within five degrees of the
true fluoroscopic CLO angle as determined during
the injection procedure. Accuracy by ICC showed
only fair agreement, 0.47 and 0.22, for the two
authors.
Conclusions. The findings of this study indicate fair
inter-rater reliability in manual measurements of
laminar angle on axial MRI images. MRI laminar
angle measurements do not appear to be highly
accurate in determining the appropriate fluoroscopic CLO angle.
Key Words. Epidural (Injection Space); Fluoroscopy;
Contralateral Oblique; Spinal Cord; Block; Cervical;
Lamina
Introduction
Epidural steroid injections are often used in the treatment
of cervicalgia and cervical radicular pain [1]. To gain
access to the cervical epidural space an interlaminar
approach is often utilized [1]. In contrast to the historical
“blind” approach, fluoroscopy is now typically employed
for the purpose of improved safety and accuracy.
However, iatrogenic spinal cord injuries have occurred
even with the use of fluoroscopic imaging [2,3].
Determining needle depth appropriately during the performance of an interlaminar approach is paramount in
C 2016 American Academy of Pain Medicine. All rights reserved. For permissions, please e-mail: [email protected]
V
1628
Thoracic Interlaminar Epidural Steroid Injections
avoiding the catastrophic complication of spinal cord
injury through needle puncture or through direct injection into the cord [1,2]. Although obtaining loss-ofresistance (LOR) through the ligamentum flavum
remains the final step of needle placement into the epidural space, determining the needle depth on imaging
has two important roles. First, the LOR portion of the
procedure must be initiated near but not through the ligamentum flavum. Although some practitioners favor
contacting the lamina to help establish an initial depth
[1], others prefer a more direct approach. Regardless,
the depth of the needle tip prior to beginning the LOR
portion of the procedure should be accurately determined with the assistance of fluoroscopic imaging.
Secondly, as the needle tip is advanced during the LOR
portion, remaining cognizant of the needle depth anticipated to achieve epidural access serves as a safety
measure against false negative LOR. This is especially
important with a midline approach in light of the high
frequency of non-fusion of the ligamentum flavum in the
midline at the cervical and proximal thoracic levels [4].
Traditionally, the lateral view has been utilized to determine needle depth. When a midline approach is used the
lateral fluoroscopic image should demonstrate accurate
needle tip position at or just beyond the ventral margins
of the lamina [1], often referred to as the spinolaminar
line on lateral view. As there is usually a paucity of dorsal
epidural adipose at the mid and proximal cervical levels,
and consistent high cervical flow can be achieved with
access from the cervicothoracic junction [5,6], it is recommended that the epidural approach to the cervical
region should be obtained at C7/T1 or T1/T2 [1].
Unfortunately, visualization of the C7/T1 and T1/T2 levels
on lateral view is often difficult depending on patient positioning, body habitus, and anatomy [7]. Some practitioners prefer to use a non-midline lateral interlaminar
approach dependent upon the patient symptoms. With a
non-midline needle position, the lateral view can no longer provide an accurate needle tip depth. The needle will
appear to be more anterior than actual position secondary to the vertebral geometry [1,7,8].
Although anterior-posterior and lateral imaging has been
standard practice, more recent attention has been
directed toward the use of a contralateral oblique (CLO)
view for determining needle depth during interlaminar epidural steroid injections [1,7–10]. Obtaining a view in
which the x-ray beam is parallel to the obliquity of the
lamina in which the needle is passing provides a very
accurate image of the needle depth (Figure 1). To obtain
this view, the image intensifier is moved to a contralateral
oblique position (contralateral from the lamina the needle
tip is passing beyond). For example, during a right sided
C7/T1 interlaminar injection, the image intensifier is
rotated to a left oblique position to “look down” the right
C7 and T1 laminae. This concept is very well illustrated in
Landers et al. as well as the work of Furman et al. [7,8].
Ideally, LOR should occur at or just beyond the ventral
edge of the lamina, referred to as the ventral interlaminar
line (VILL), when using this CLO approach.
Some controversy does exist, however, in determining
the appropriate contralateral oblique angle. Different
authors have proposed various angles ranging from 30
to 60 degrees [7–10]. Guidelines from the International
Spine Intervention Society propose a starting angle of
45 degrees [1]. It is then recommended to adjust the
angle accordingly to obtain an appropriate laminar view
with a characteristically wide foramen and longitudinally
co-aligned superior articular process and lamina [1]. Gill
et al. [11] expressed that the previously described
method to determine the appropriate contralateral oblique angle is somewhat ambiguous and recommended
angular measurements of the lamina on axial MRI.
Furman et al. [12] replied that “measurement [of angle
of the lamina on the MRI] is not precise, practical, nor
fluoroscopically reproducible.” If an appropriate contralateral oblique angle can be accurately predicted by MRI
measurements as described by Gill et al. [11], it would
certainly increase practitioners’ confidence that an accurate needle depth is being visualized.
The purpose of this study is to evaluate the usefulness
of pre-procedural MRI laminar angle measurements in
determining the appropriate fluoroscopic contralateral
oblique angle to be used during cervical or thoracic
interlaminar epidural steroid injections.
Methods
The study was approved by an independent IRB,
Sterling ID#4516-001.
At the time of this study, the authors (DL and SH, both
fellowship trained interventional physiatrists), performed
cervical interlaminar epidural approach with AP and lateral views and had added the contralateral oblique
angle view with some apprehension. Seemingly appropriate contralateral oblique angles had, often times,
failed to be close to the correct angle. As confidence in
the contralateral oblique view was not strong, the
authors typically continued to use a lateral view to determine depth with the understanding that the further lateral from midline the needle tip, the more ventral the
position would be perceived on the lateral image. A
CLO view was also often employed using the
International Spine Intervention Society technique for
angle determination [1]. After the epidural space was
accessed by LOR, the true contralateral oblique angle
was determined by needle tip position and recorded.
The true angle was defined as the fluoroscopic angle in
which the needle tip lies within the epidural space and
at or just beyond the VILL. The true CLO angle was
determined after contrast confirmation of epidural
access. If the needle tip lies superficial or deep to the
VILL, the fluoroscope was rotated until an oblique angle
was such that the needle tip appeared to lie at or just
beyond the contralateral oblique ventral laminar line
(Figures 1 and 2). The fluoroscopic angle measurement
was adjusted appropriately if the patient’s true A–P
angle was not zero. This true CLO angle was then
recorded in the procedure report. The purpose of
1629
Levi et al.
A chart review was performed, retrospectively,
through the authors’ practice electronic medical
record (EMR). Patients undergoing cervical or thoracic
epidural steroid injections were included from
December 1, 2012 through October 15, 2013.
Patients were excluded if a true contralateral oblique
angle was not recorded.
For each patient found to have a documented CLO
angle recorded in the procedure note, the patient’s MRI
was then blindly reviewed to determine the laminar
angle measurement. The measurements were performed independently by two of the authors (DL and
SH). MRI measurements were not performed prior the
injection procedure, but only completed as part of this
review. One patient had a CT only, which was used in
place of the MRI.
Figure 1
Contralateral oblique view of C7/T1 interlaminar injection (pre-contrast).
Technique for MRI laminar measurements: The MRI was
blindly reviewed by two of the authors independently (DL
and SH) to determine the laminar angle measurement.
The authors agreed to use the axial MRI T2 sequence at
the level superior to the needle insertion (i.e., for a C7/T1
injection the axial image best demonstrating the C7 lamina was used). As some of the MRIs reviewed had no
software to include angular measurements, all MRI angles
were determined manually to maintain consistency.
(Figure 3) A line was drawn bisecting the spinous process and central portion of the vertebral body or disc
(segment A). A second line (segment B) was drawn
through the mid portion of the long axis of the left lamina (for a left sided procedure) and extended to cross
beyond “segment A.” The angle between segments A
and B (the contralateral oblique angle) was determined.
Note that opposite angles formed by two intersecting
lines are equal per the vertical angles theorem.
Main outcome measures:
MRI angle measurement inter-rater reliability by Intraclass correlation coefficient (ICC).
Accuracy by percentage of the MRI measurement compared with the true CLO angle determined during the
fluoroscopic procedure within five degrees. Five degrees
error was determined by the authors’ experience in that
beyond this amount (plus or minus), the needle tip
appeared clearly too anterior or posterior to be considered the true CLO angle.
Results
Figure 2
Contralateral oblique view of C7/T1 interlaminar injection (post-contrast).
routinely determining the true CLO angle measurement
was primarily to increase the safety and efficiency of a
possible subsequent injection. This also helped to confirm appropriate epidural flow pattern including the
cephalad and caudal extent of the contrast spread.
1630
Thirty-four injections were included. Twenty-five were
paramedian, and nine were midline. Thirty-one were
performed at C7/T1, two were at T1/2, and one was at
T9/10.
The average true fluoroscopic angle was 53 degrees
(range 44–61 degrees). The average measured angle on
MRI was 48.5 degrees (range 28–59 degrees).
Thoracic Interlaminar Epidural Steroid Injections
Figure 3 Axial MRI of C7
level. Contralateral oblique
angle determination for left
C7/T1 interlaminar ESI.
Segment A (vertical line)
bisects the vertebral body
and
spinous
process.
Segment B (oblique line) is
drawn along the long axis
of the C7 lamina. The corresponding
contralateral
oblique angle is measured.
70
60
Angle (Degrees)
50
40
30
20
10
Figure 4
Observer 1 compared with observer 2 MRI
laminar angle measurement.
Observer 1 MRI Angle
Observer 2 MRI Angle
0
0
5
Inter-rater reliability comparing the two authors MRI
(one CT) angle measurements was considered fair, ICC
¼ 0.395.
Accuracy of the MRI (one CT) angle measurement to
predict, within five degrees, the “true” contralateral oblique angle as determined during the injection procedure
was overall, 57% (CI 73.6–40.4).
Individual accuracy rates within five degrees for each of
the two authors were 65% and 50%.
Accuracy within five degrees measured by ICC also
showed only fair agreement, 0.47, and poor agreement,
0.22, for each of the two authors (Figures 4–6).
10
15
20
25
30
35
Paent Number
Discussion
The purpose of this study was to help determine if using
axial MRI laminar measurements prior to a cervical or
thoracic epidural steroid injection would be useful in predicting the appropriate fluoroscopic contralateral oblique
angle. This retrospective review provides evidence against
using a pre-procedural laminar measurement. The accuracy of the MRI measured contralateral oblique angle
allowing five degrees of error was poor. Stated in a different manner, had the MRI measurements been performed
prior to the procedure and utilized to determine the fluoroscopic CLO angle (plus or minus up to five degrees), the
needle tip would have been at or just beyond the ventral
portion of the lamina only 57% (CI 73.6–40.4) of the time.
1631
Levi et al.
70
60
Angle (Degrees)
50
40
30
Actual Fluoroscopic
Angle
Observer 1 MRI
Angle
20
10
0
0
5
10
15
20
25
30
35
Paent Number
Figure 5
Actual fluoroscopic
contralateral oblique angle determined during the injection procedure compared with observer 1
MRI laminar angle measurement.
70
60
Angle (Degrees)
50
40
30
20
10
Actual fluoroscopic Angle
Observer 2 MRI Angle
0
0
5
10
15
20
Paent Number
Our findings are in concordance with an outstanding
prospective study by Gill et al. [13] investigating the
appropriate angle for the contralateral oblique view. The
authors used 30, 40, 45, and 50 degree angles, an isointense angle with dorsal and ventral laminar lines of
equal intensity as well as the MRI laminar measured
angle prior to the injection. The MRI measurement technique was similar to that of our study except Gill et al.
used computer software rather than manual angular
measurements. They described foraminal zones as
anterior, middle, and ventral thirds of the foramen and
determined needle zone location at each of the above
angles once epidural access was achieved. Similar to
the findings in the current study, Gill et al. [13] found
that loss of resistance at the ventral interlaminar line
occurred in only 14/24 patients (58%) when using the
MRI measured angle. However, when using the MRI
measured angle, or 50 degrees, the needle tip position
was typically within the dorsal third of the foramen and
always within the dorsal half of the foramen when loss
of resistance occurred. There was no statistical difference between the measured MRI laminar angle and 50
1632
25
30
35
Figure 6
Actual fluoroscopic
contralateral oblique angle determined during the injection procedure compared with observer 2
MRI laminar angle measurement.
degrees in their study [13]. Although falling short of providing an exact measurement at which LOR is likely to
occur at the VILL, the work of Gill et al. does provide
rationale for using 50 degrees CLO angle with the
understanding that the needle tip may appear to traverse up to half of the foramen [13].
As previously mentioned, our intention in recording the
true fluoroscopic CLO angle was for use in possible
subsequent injections. The average angle was 53
degrees with a wide range of 44–61 degrees. As far as
we are aware, no other study has documented the
angle at which epidural access was achieved on CLO
fluoroscopic view. Although no formal study has yet to
be undertaken, anecdotally, using the recorded true fluoroscopic CLO from an initial injection appears to be
very accurate for subsequent injections. This is an area
for future research.
There are several limitations of our study. The measurements of the MRI laminar angle were performed manually as several of the MRIs did not have angle
Thoracic Interlaminar Epidural Steroid Injections
measurement on the viewing software. However, our
average MRI laminar angular measurement was similar
to foramen and laminar angle measurements of other
studies [13,14]. In addition, the laminar line is rarely perfectly straight, introducing error whether the angle is
measured manually or electronically. The authors also
did include a single CT in place of the MRI but this was
not expected to have altered the results. Although fluoroscopic angle measurements were felt to be accurate,
the angle demarcation lines on the c-arm are only at
five degree increments and this may have introduced
some measurement error as well.
5 Lee SE, Joe HB, Park JH, et al. Distribution range
of cervical interlaminar epidural injections: A comparative study with 2.5 mL, 5 mL, and 10 mL of contrast. Pain Physician 2013;16:155–64.
Conclusion
8 Landers MH, Dreyfuss P, Bodgduk N. On the
geometry of fluoroscopic views of cervical epidural
injections. Pain Med 2012;13(1):58–65.
The findings of this study indicate only fair inter-rater
reliability in manual measurements of laminar angle on
axial MRI images. MRI laminar angle measurements do
not appear to be highly accurate in determining the
appropriate fluoroscopic contralateral oblique angle.
This study does not support the use of pre-procedural
laminar angle measurements as a predictive tool to
determine the appropriate fluoroscopic CLO angle for
cervical or thoracic interlaminar injections.
References
1 International Spine Intervention Society. Cervical interlaminar access. In: Bogduk N, ed. Practice Guidelines
for Spinal Diagnostic and Treatment Procedures, 2nd
edition. San Francisco: International Spine Intervention
Society; 2012: 219–56.
2 Hodges SD, Castleberg RL, Miller T, Ward R,
Thornburg C. Cervical epidural steroid injection with
intrinsic spinal cord damage: Two case reports.
Spine 1998;23:2137–42. discussion 2141–2.
3 Khan S, Pioro EP. Cervical epidural injection complicated by syrinx formation: A case report. Spine
2010;35:E614–6.
4 Lirk P, Kolbitsch C, Putz G, et al. Cervical and high
thoracic ligamentum flavum frequently fails to fuse in
the midline. Anesthesiology 2003;99:1387–90.
6 Goel A, Pollan JJ. Contrast flow characteristics in
the cervical epidural space: An analysis of cervical
epidurograms. Spine 2006;31:1576–9.
7 Furman M, Jasper NR, Lin H. Fluoroscopic contralateral oblique view in Interlaminar interventions: A
technical note. Pain Med 2012;13:1389–96.
9 Whitworth M. Puttlitz line: A rapid and reproducible
fluoroscopic
needle
endpoint
for
cervical
Interlaminar epidural steroid injections. Pain Med
2008;9:136–7.
10 Vaisman J. Alternative view for the interlaminar
cervical epidural steroid injections. Pain Med 2010;
11:1743.
11 Gill J, Aner M, Simopoulos T. Intricacies of the contralateral oblique view for interlaminar epidural
access. Pain Med 2013;14:1265–6.
12 Furman MB, Jasper NR, Lin HT. In response
to “intricacies of the contralateral oblique
view for interlaminar epidural access”. Pain Med
2013;14:1267–8.
13 Gill JS, Aner M, Jyotsna N, Keel JC, Simopoulos
TT. Contralateral oblique view is superior to lateral
view for interlaminar cervical and cervicothoracic
epidural access. Pain Med 2015;16(1):68–80.
14 Simpson AK, Sabino J, Whang P, Emerson JW,
Grauer JN. The assessment of cervical foramina
with oblique radiographs: The effect of film angle
on foraminal area. J Spinal Disord Tech 2009;22
(1):21–5.
1633