| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Head and Neck Imaging |
1 From the Departments of Radiology (K.S., C.W.A.P., M.S., J.H., M.Z.) and Orthopedic Surgery (N.B.), Orthopedic University Hospital, Zurich, Switzerland. Received March 26, 2003; revision requested June 18; final revision received February 3, 2004; accepted February 17. Address correspondence to K.S., Department of Radiology, Cantonal Hospital Lucerne, Spitalstrasse, 6000 Lucerne 16, Switzerland (e-mail: klaustro@bluewin.ch).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Sixty consecutive patients with cervical radicular pain were examined with MR imaging and then treated with CT-guided cervical nerve root blocks (CNRBs). Various MR imaging findings were assessed and compared in terms of associated pain relief after CNRB. Pain relief was graded (0%–100%) by using a visual analogue scale (VAS). The relationship between MR imaging findings and level of pain relief was tested by using Mann-Whitney U and Kruskal-Wallis tests.
RESULTS: The mean percentage of pain reduction at VAS grading was 46%. There was a significant relationship between pain relief level and both location of disk herniation (mean pain reductions of 41% at median or mediolateral locations and 64% at foraminal locations, P = .034) and location of nerve root compromise (mean pain reductions of 19% at intraspinal, 45% at foraminal entrance, and 58% at foraminal locations; P = .019). There was an inverse relationship between pain relief level and absence or presence of spinal canal stenosis (mean pain reductions of 29% when stenosis present and 53% when stenosis absent, P = .013). Other parameters were not significantly related to pain relief.
CONCLUSION: MR imaging of the cervical spine assists in the appropriate selection of patients suitable for CNRB. Patients with foraminal disk herniation, foraminal nerve root compromise, and no spinal canal stenosis appear to have the greatest pain relief after this procedure.
© RSNA, 2004
Index terms: Nerves, roots • Spinal canal, stenosis, 31.142 • Spine, abnormalities, 31.142, 31.148, 31.783 • Spine, CT, 31.12112 • Spine, MR, 31.121411, 31.121416
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Magnetic resonance (MR) imaging of the cervical spine combined with standard radiography may represent the reference-standard examination for the work-up of patients with cervical radiculopathy (5,6). The diagnostic performance of MR imaging of cervical spine degeneration has been documented in several studies (7–9). To our knowledge, the role of MR imaging in the selection of patients who will benefit from nerve root blocks has not been formally assessed. Thus, the purpose of this study was to examine whether MR imaging findings of the cervical spine can predict pain relief after selective CT-guided nerve root block and thus assist in the appropriate selection of patients who are suitable for this procedure.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Exclusion criteria were as follows: (a) relevant motor deficit diagnosed on the basis of a Medical Research Council of Great Britain grade of M3 or lower (no patients), (b) clinical and/or MR imaging findings of myelopathy (no patients), (c) previously performed surgery of the cervical spine (five patients), and (d) injection in more than one neural foramen during a single session (two patients). These exclusions resulted in a final study population of 60 patients: 26 women with a mean age of 51 years (age range, 33–69 years) and 34 men with a mean age of 48 years (age range, 30–71 years).
Informed consent was obtained for all CNRBs. On the basis of a general permit issued by the responsible state agency, our institutional review board allows the retrospective analysis of patient data relating to standard diagnostic or therapeutic procedures. Thus, informed consent was not required for our retrospective review of data.
MR Imaging
MR imaging was performed for the evaluation of cervical radicular pain. Thirty-five patients were examined with a 1.5-T (Symphony; Siemens Medical Solutions, Erlangen, Germany) or 1.0-T (Impact Expert; Siemens Medical Solutions) MR imaging unit at our institution. The cervical spine was placed in a dedicated receive-only spine coil, with the patient in the supine position. The imaging protocol with the 1.5-T unit, which was used to examine 19 patients, included the acquisition of sagittal T1-weighted spin-echo (350/12 [repetition time msec/echo time msec], 4-mm section thickness), sagittal T2-weighted turbo spin-echo (4500/121, 4-mm section thickness), and transverse T2*-weighted multiecho data image combination (1250/26, 30° flip angle, 3-mm section thickness) MR images. With the 1.0-T unit, which was used to examine 16 patients, sagittal T1-weighted spin-echo (500/12, 4-mm section thickness), sagittal T2-weighted turbo spin-echo (4000/96, 4-mm section thickness), and transverse two-dimensional T2*-weighted fast low-angle shot (666/22, 20° flip angle, 3-mm section thickness) MR images were acquired.
Twenty-five MR imaging examinations were performed at other institutions but with sequences similar to those used at our institution: 1.0- or 1.5-T units were used, and the following examinations were performed: sagittal T1-weighted (350–720/11–20) and T2-weighted (3500–4800/90–130) MR imaging and transverse T2-weighted (3500–4800/90–300) or T2*-weighted (580–710/20–40, 18°–22° flip angle) MR imaging. In 21 of the total of 60 patients, T2- or T1-weighted oblique sagittal MR imaging perpendicular to the relevant neural foramen was performed in addition to the standard MR imaging examinations.
Analysis of MR Images
The MR images were retrospectively evaluated in consensus by two radiologists (C.W.A.P., M.S.), each of whom had at least 10 years of experience in musculoskeletal MR imaging, including MR imaging of the spine. In 39 patients, film hard-copy MR images were reviewed. In the remaining 21 patients, MR images were reviewed on a picture archiving and communication system, or PACS, workstation (ID.Station Report; Image Devices, Idstein, Germany). The reviewers were informed of the clinical diagnosis of radiculopathy, including the clinically diagnosed level and side of the abnormality. However, they were blinded with regard to the level of pain relief that resulted after the CNRBs.
The following findings were assessed: The presence or absence of any disk herniations, as well as the location of these abnormalities (ie, median or mediolateral, or foraminal) was reported. Nerve root compromise, regardless of its origin, was described according to location (intraspinal, foraminal entrance, or foraminal), extent (ie, radiculopathy has no contact with nerve root, radiculopathy has contact with nerve root, nerve root deviation, or nerve root compression), and origin (ie, diskogenic, mixed, or osseous). Stenosis of the spinal canal was diagnosed when the sagittal diameter of the cervical canal was 10 mm or less. Any spinal cord deformity (eg, appearance of the anterior contour of the spinal cord on sagittal images and loss of symmetric form on transverse images, both caused by disk material or osteophytes) was also noted.
CNRB and Assessment of Pain Reduction
All CNRBs were performed as outpatient procedures by seven radiologists (including K.S., C.W.A.P., M.S., J.H., and M.Z.) experienced in performing spinal interventions. Each radiologist performed between five and 11 interventions. All radiologists had fellowship training in musculoskeletal radiology, including imaging of musculoskeletal interventions. Five of the seven radiologists had an additional 10 years (J.H.), 10 years (M.Z.), 5 years (C.W.A.P.), 2 years (M.S.), and 1 year (K.S.) of experience in performing musculoskeletal interventions. All of the radiologists adhered to a standardized protocol to guarantee the consistency of the CNRB procedures.
The injections were performed with CT fluoroscopic (Somatom Plus 4; Siemens Medical Solutions) guidance. The patients lay supine, with their heads turned to the side opposite of the side of the injection in the foramen. The head was fixed with tape to prevent motion during the procedure. After skin disinfection and administration of subcutaneous local anesthetics, a 23-gauge needle was introduced with fluoroscopic guidance by using a lateral or slightly anterolateral approach dorsal to the large cervical vessels. The needle was aimed at the posterior border of the neural foramen, dorsal to the vertebral artery.
Initially, 0.3 mL of iopamidol (Iopamiro 300, 300 mg of iodine per milliliter; Bracco Diagnostics, Princeton, NJ) was injected to verify the correct position of the needle tip. The intraforaminal distribution of the contrast material was documented with a single CT fluoroscopic scan. A maximum of 40 mg (1.0 mL) of crystalloid corticosteroid suspension—specifically, 40 mg of triamcinolone (Kenacort A 40; Bristol-Myers Squibb, New York, NY)—and 1 mL of 0.2% ropivacaine (Naropin 0.2%; AstraZeneca, Westborough, Mass) were slowly injected. Fifteen minutes after the injection, pain relief was assessed by using a visual analogue scale (VAS) (10), on which 0% represented no pain reduction and 100% represented complete pain relief.
Statistical Analyses
The presence or absence of an MR imaging finding was compared with the level of pain reduction after CNRB by using the Mann-Whitney U test. If gradings that included more than two possible values had to be compared, the Kruskal-Wallis test was used instead. P < .05 was considered to indicate significance. We used a computer software package (SPSS, release 11.0.0; SPSS, Chicago, Ill) to perform statistical analyses.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Although the method of selective nerve root block is the subject of great controversy (17), there is evidence that lumbar nerve root blocks and CNRBs are effective treatment methods (1,18). However, most published investigations have focused on lumbar nerve root blocks (19–21); relatively few have focused on CNRBs. Berger et al (14) performed CT-guided foraminal injections with a lateral approach in 18 patients with cervical radiculopathy. They reported effective (
50%) long-term pain relief in 11 of their patients (61%). Bush and Hillier (1) injected corticosteroids in 68 patients with cervical radiculopathy, all of whom were potential surgical candidates. They used three fluoroscopically guided periradicular and epidural injection techniques and performed serial (average of 2.5 injections; range, 1–6 injections) injections. All patients recovered satisfactorily, without the need for surgical intervention (1).
In contrast to our study, in which we concentrated on the short-term effects of local anesthetics, the Bush and Hillier study (1) was an investigation of the long-term pain relief related to the use of steroids. They reported a mean pain score of only 0.6 (score range of 0–5, with 10 representing the worst pain) at VAS grading after a mean follow-up period of 7 months. The mean short-term pain relief of 46% at VAS grading that was achieved in our study population is comparable to the mean short-term pain relief of 47% that has been reported after fluoroscopically guided lumbar nerve root blocks (20).
The potential for severe complications with cervical spine interventions is higher than that with lumbar spine interventions. Damage to the cervical spinal cord following epidural steroid injections has been described (22,23). Williams et al (2) reported a cervical epidural hematoma that developed after repeated epidural steroid injections and required surgical decompression. Epidural abscess is another possible complication (24). Foraminal injections may be less prone to induce severe complications than epidural injections. However, Brouwers et al (3) reported the occurrence of a spinal cord infarction immediately following a fluoroscopically guided C6 nerve root block. They considered the infarction to have been caused by an anterior spinal artery syndrome. At our institution, to reduce the risk of injury to the vertebral artery and the cervical spinal cord, CNRBs are performed with CT fluoroscopic guidance. Despite such precautions, side effects are a potential problem, and, thus, the appropriate selection of patients for CNRB is important.
Because MR imaging is commonly performed before patients are referred for CNRB, we intended to relate various MR imaging findings to the patient outcomes after the injections. The outcome after CNRB was best when any disk herniation or any nerve root compromise (regardless of origin) was foraminal. This more favorable outcome could be anticipated because the injection site was closest to the site of the abnormality. The inverse relationship between spinal canal stenosis and outcome after CNRB may have an analogous explanation—namely, that the main nerve root compromise was remote from the injection site. Several other MR imaging findings, including presence versus absence of disk herniation, extent of nerve root compromise, and presence versus absence of spinal cord deformity, were not significantly associated with CNRB outcome. Owing to the retrospective nature of this investigation, oblique MR images were not available for all patients. These views may add important information regarding foraminal nerve root compromise that is not available on standard sagittal and transverse images (25).
Our study results suggest that when "hard" bone changes are the dominant responsible pathologic entity, the pain relief after a CNRB is markedly better than that in patients with a mainly "soft" diskogenic abnormality (mean pain reduction of 57% vs 44%); however, the difference was not significantly different. This trend is supported by the results of a study of CT-guided lumbar foraminal steroid injections: In the presence of osteophytes, the injections were more successful (95% pain reduction) than were those in patients with disk herniations (45% pain reduction) (21). We are aware that the VAS grade determined 15 minutes after the intervention may not always predict the final therapeutic effect. On the other hand, this is the only measurement that was routinely obtained in the described study population and therefore that was routinely available for review. This is one of the drawbacks of a retrospective investigation. Obtaining reliable midterm data in a retrospective manner would have been quite difficult considering that patient data were included in the study up to several years after the nerve block injections had been performed.
The retrospective study design contributed to another weakness of the study: Oblique views were not available for all patients, although they add important information concerning foraminal nerve root compromise (25). In addition, the diagnosis of stenosis in the cervical spine is difficult to make. Only moderate interobserver agreement in the assessment of cervical spine stenosis has been documented (26).
In summary, MR imaging of the cervical spine is helpful in the appropriate selection of patients who are suitable for a CNRB. Patients with foraminal disk herniation, foraminal nerve root compromise, and no spinal canal stenosis appear to have the greatest pain relief after this procedure.
|
FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, K.S., J.H., M.Z.; study concepts, K.S., C.W.A.P., M.Z.; study design, K.S., C.W.A.P., M.S., M.Z.; literature research, K.S., N.B.; clinical studies, K.S., N.B.; data acquisition and analysis/interpretation, K.S., C.W.A.P., M.S.; statistical analysis, C.W.A.P., J.H., M.Z.; manuscript preparation, K.S., M.Z., J.H.; manuscript definition of intellectual content, J.H., N.B., M.Z.; manuscript editing and revision/review, all authors; manuscript final version approval, K.S., N.B., J.H., M.Z.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K.P. Schellhas, S.R. Pollei, B.A. Johnson, M.J. Golden, J.A. Eklund, and R.S. Pobiel Selective Cervical Nerve Root Blockade: Experience with a Safe and Reliable Technique Using an Anterolateral Approach for Needle Placement AJNR Am. J. Neuroradiol., November 1, 2007; 28(10): 1909 - 1914. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
