Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Original Article
Ultrasound guided versus fluroscopic guided pulsed radiofrequency
therapy of the stellate ganglion in neuropathic pain: A prospective
controlled comparative study
Mohamed H. Shaaban
b
, Raafat M. Reyad
a
, Hossam Z. Ghobrial
a
, Rania H. Hashem
b,
⇑
a
Department of Anesthesia and Pain Management, National Cancer Institute, Cairo University, Egypt
b
Department of Diagnostic and Interventional Radiology, Kasr El Aini, Faculty of Medicine, Cairo University, Egypt
article info
Article history:
Received 24 January 2017
Accepted 18 June 2017
Available online 28 March 2018
Keywords:
Stellate ganglion block (SGB)
Neuropathic pain
Pulsed RF
abstract
Objective: To assess the efficacy and safety of fluoroscopic–guided versus ultrasound (US)-guided tech-
niques for pulsed radiofrequency (RF) therapy of stellate ganglion for refractory neuropathic pain syn-
dromes.
Methods: 40 patients with severe chronic neuropathic pain syndromes, Visual Analogue Scale (VAS)
score > 7, with poor response to medical treatment were randomly integrated into 2 groups: Group
(F): (20 patients) in whom pulsed R.F. therapy is done under fluoroscopy, group (U): (20 patients) in
whom pulsed R.F. therapy is done under US guidance.
Results: The current study revealed that there is significant reduction of VAS, and of the medical treat-
ment consumption after the block as compared with pre block values, there is no statistically significant
difference between the guidance techniques of RF treatment in pain relief. However, the procedure time
was significantly lower in U group.
Conclusion: Pulsed R.F. blockade of the stellate ganglion in patients with refractory neuropathic pain syn-
dromes can be done safely and efficiently under the guidance of either ultrasound or fluoroscopy. Both
radiological techniques provide similar satisfactory guidance without significant complications.
Ó2018 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier. This
is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/
4.0/).
1. Introduction
The stellate (cervico-thoracic) ganglion is the result of fusion of
the inferior cervical sympathetic ganglion with the first thoracic
one and this fusion occurs in 80% of population. It is star shaped
and measures 2.5 cm long, 1 cm wide and 0.5 cm thick and lies
in front of C7 transverse process and the head of first rib [1]. Stel-
late ganglion blockade is utilized as diagnostic, prognostic or ther-
apeutic intervention for sympathetic-mediated (maintained) pain,
neuropathic pain syndromes and a lot of clinical implications [2].
Stellate ganglion blockade has been proven to be of value in
vascular insufficiency of the upper limb such as Raynaud’s disease,
vasospasm, embolic vascular disease, Paget’s disease, scleroderma,
palmar hyperhydrosis, and in many pain syndromes like phantom
limb pain, complex regional pain syndrome (CRPS), post-herpetic
neuralgia, diabetic neuropathy, vascular headache, atypical facial
pain and tic douloureux [3,4]. Other indications of left-stellate
block are prinzemetal angina, prolonged Q-T syndrome and mas-
sive pulmonary embolism (bilateral block) [4].
On the other hand, stellate ganglion blockade is not a risk-free
technique due to close proximity of a variety of vital structures.
The vertebral artery originates from subclavian artery and lies
anterior to the stellate ganglion at C7 level, then passing over the
ganglion to enter the vertebral foramen. It lies posterior to C6-
anterior tubercle. The ganglion is bounded medially by longus colli
muscles, laterally by scalene muscles, anteriorly by subclavian
artery, posteriorly by prevertebral fascia and transverse process,
inferiorly by the pleura. Other important nervous structures
related to the ganglion include the phrenic nerve (lateral), the
recurrent laryngeal nerve (antero-medial) and the C8-T1 anterior
divisions (posterior) [3,5].
Different modalities have been tried to block the stellate
ganglion including local anesthetics, steroids, neurolytic agents
https://doi.org/10.1016/j.ejrnm.2017.06.008
0378-603X/Ó2018 The Egyptian Society of Radiology and Nuclear Medicine. Production and hosting by Elsevier.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer review under responsibility of The Egyptian Society of Radiology and Nuclear
Medicine.
⇑
Corresponding author.
E-mail addresses: mohamedhamed24672@yahoo.com (M.H. Shaaban),
dr_raafat2006@hotmail.com (R.M. Reyad), hossam_zarif@yahoo.com (H.Z. Ghobrial),
rania.hachem@yahoo.com (R.H. Hashem).
The Egyptian Journal of Radiology and Nuclear Medicine 49 (2018) 71–75
Contents lists available at ScienceDirect
The Egyptian Journal of Radiology and Nuclear Medicine
journal homepage: www.sciencedirect.com/locate/ejrnm
(phenol in saline 3%) [6] and radiofrequency therapy (pulsed or
thermal) [7].
Radiofrequency neurolysis is an extension of continuous regio-
nal sympathetic blockade [8] or chemical neurolysis with long-
term efficacy and more safety together with less morbidity than
open surgical techniques [9]. Multiple imaging guidance are in
use to perform stellate block whether ultrasound (which provide
clear visualization of vascular and soft tissues structures) [10],
MRI, CT and plain fluoroscopy [2]. Fluoroscopic approaches to
block the stellate are either anterior (C6–C7), oblique or posterior
[1,3,6].
In this prospective controlled study, we tried to compare the
efficacy and safety of fluoroscopic–guided versus U/S-guided tech-
niques for pulsed radiofrequency therapy of stellate ganglion aim-
ing that the resulting sympathectomy may help to alleviate
refractory neuropathic pain syndromes.
2. Patients and methods
After approval of local ethical committee and obtaining
informed consent, 40 patients were randomly selected from the
pain clinic of National Cancer Institute (NCI) Cairo University
between August 2011 and February 2014. All patients had chronic
neuropathic pain syndromes in the upper limb with severe pain
(VAS score > 7) refractory to strong opioids (morphine sulfate
tablets) and adjuvant therapy (pre-gabalin Capsules) or experienc-
ing intolerable side effects. Patients with local and systemic sepsis,
coagulopathy, local anatomical distortion (post-operative or post-
radiotherapy) making the procedure difficult or hazardous are
excluded from the study. Also patients with history of contralateral
chest disease or pneumonectomy, glaucoma, recent M.I. or severe
bradyarrythmias or heart block and allergy to the used medica-
tions, were excluded from the study.
3. Patients were randomly integrated into 2 equal groups
Group (F): (20 patients) in whom pulsed R.F. therapy is done
under fluoroscopy.
Group (U): (20 patients) in whom pulsed R.F. therapy is done
under US guidance.
ASA-standard monitors (ECG, non-invasive blood pressure and
pulse oxymitery) were connected to all patients. I.V. line (G-20)
and O2 (3 L/min) through nasal pronge were used. Midazolam
0.02 mg/kg and fentanyl 1 Ug/kg (conscious sedation) were used.
The patient was asked to lie supine over radiolucent table with
the neck extended and a small pillow under shoulders. The field
was sterilized with 10% betadine (povidone-iodine) and draped.
The patient was foretold to communicate by moving the contralat-
eral hand and not to speak or swallow during needling.
4. Fluoroscopic-technique (anterior approach)
Visualization of C6-C7 level was attained through PA after good
alignment was obtained by caudocephalic orientation (C7 level is
identified by the nearby T1-transverse process ballooning). Then,
the C-arm was turned 5-10o ipsilateral to open the vertebro-
transverse junction at C7. At this point of entry, 1% lidocaine was
infliterated S.C. using 22 G needle. Then R.F. needle (curved, sharp,
22 G, 50 or 100 ml length with 10 mm active tip) was advanced
using tunnel technique until bony contact was made at the antero-
lateral side of C7 vertebra (Fig. 1A). After negative aspiration (For
blood, CSF or air), 3 ml of contrast medium (iohexol, omnipaque)
was injected. It should outline the retropharyngeal space, longitu-
dinal, huking the lateral vertebral margin, within the vertebral
shadow (on lateral view), not taking vascular, epidural, intrathecal
or muscular pattern (Fig. 1B). Then the suitable R.F. electrode was
inserted and connected to Bailys generator. Impedance should be
250–350 and no paresthesia is felt with sensory stimulation (50
Hz to 1.0–1.5 V) particularly in the upper limb and motor stimula-
tion should be negative (while the patient saying E-E) at 2 Hz and
3 V. 3 ml lidocaine 2% plus 1 ml diprofos (5mg betamethosone)
was injected. After 30–60 s, we used pulsed RF protocol with
time = 8 min, temperature = 42 °C and pulse width = 10 m s.
5. Ultrasound technique [10]
The patient was prepared as before. High-resolution ultrasound
imaging for identification of small nerves and the interface
between bone and soft tissues, with Doppler for the nearby vessels
(vertebral, superior and inferior thyroid vessels). Siemens Acuson
700 U/S machine with high frequency linear transducer was
used for superficial targets. Anterolateral margin of C6 body with
the transverse process was identified. The target point is identified
by the 4–12 MHz linear-array probe and check the R.F. needle ori-
entation (looking at thyroid anteriorly and esophagus posteriorly)
we used out of plane technique.
Then the needle was withdrawn and reinserted obliquely so
that the needle tip lie anterior to longus coli muscle (anterior to
C6 transverse process). After negative aspiration, 1 ml of normal
saline was injected which should spread adequately up and down
without vascular uptake (Fig. 2). Then pulsed R.F. was done as pre-
viously after sensory and motor stimulation then 3 ml of lidocaine
2% plus 1 ml diprofos was injected.
After stellate ganglion block was performed, to confirm stellate
ganglion block, touch temperature thermometer was used to com-
pare between both sides, then the site of procedure was draped
with sterile pad and ice pack is applied to reduce hematoma. The
patient is monitored for 2 h vitally and all patients of both groups
are screened 2 h after the procedure by plain radiography to
exclude pneumothorax and by neck ultrasound for hidden hema-
toma possibility .The patients were instructed before discharge to
call the physician urgently if severe chest pain, dyspnea, CVS col-
lapse, dysphonia, severe pain and motor deficit develop.
6. Patient evaluation
The patients were assessed for pain relief (VAS score), opioid
and pregabalin consumption prior to block and at 1 day 1, 4, 12
weeks afterwards. Both morphine and pregabalin were stopped
and the patient had free access to immediate release morphine
(Sevradol 10 mg) and reassessed after 2 and 7 days to estimate
the new escalating dose. Complications including Horner’s, nerve
palsies (recurrent laryngeal, phrenic and lower brachial plexus),
vascular (vertebral and carotid arteries) and pleural injuries, epidu-
ral or subarachnoid injection, esophageal puncture, hematoma and
osteomyelitis all were reported.
7. Statistical analysis
Descriptive tables and statistical analysis were made by soft-
ware SPSS (Statistical package for social science) version11.0 sta-
tistical program. Parametric data were represented as mean and
standard deviation; meanwhile, nonparametric data were repre-
sented as median and interquartile range. Within group compar-
ison for the difference of VAS score, morphine consumption, and
pregabalin consumption was done using paired ttest. Meanwhile,
comparison between the groups at specific time intervals was
made by Mann-Whitney U test. A significant difference was
accepted at P < .05.
72 M.H. Shaaban et al. / The Egyptian Journal of Radiology and Nuclear Medicine 49 (2018) 71–75
(A) RF needle at C7 level both at AP (end-on) and lateral view
Fig. 1A. RF needle at C7 level both at AP (end-on) and lateral view.
(B) Both A-P and lateral view after injection of contrast medium showing free up and down
para vertebral spread
Fig. 1B. Both A-P and lateral view after injection of contrast medium showing free up and down para vertebral spread.
B A
ICA
TV
LCM
Fig. 2. Sonar guidance showing needle pass (arrowed) A. spread after injectate (arrowed) B. ICA = internal carotid artery TV = transverse process LCM = Longus colli muscle.
M.H. Shaaban et al. / The Egyptian Journal of Radiology and Nuclear Medicine 49 (2018) 71–75 73
8. Results
The study included 40 cases with sympathetic mediated pain in
the upper limb. The demographic characteristics of the two study
groups are presented in Table 1. The two groups did not differ sta-
tistically in age, weight, distribution of sex; however the duration
of procedure is significantly longer in F group with P value .03
(see Table 2).
VAS values didn’t differ significantly among the two groups
allover the study period, but both groups showed significant reduc-
tion of pain scores after block at all assessment time points, com-
pared to pre-block baseline values (Table 3).
Regarding morphine consumption, there was a significant
reduction in mean daily consumption at 1 week, 1 month and 3
month compared to pre-block consumption in both groups, how-
ever, no significant difference in morphine consumption was
noticed between the two studied groups at all time of measure-
ments (Table 4).
Regarding pregabalin consumption, there was a significant
reduction in mean daily consumption at 1 week, 1 month and 3
month compared to pre-block consumption in both groups, how-
ever, no significant difference in pregabalin consumption was
noticed between the two studied groups at all time of measure-
ments (Table 5).
The recorded side effects were minimal, reversible and disap-
peared in few hours to few days period (Table 6). No serious com-
plications were recorded as pneumothorax, epidural or intrathecal
injection or permanent Horner syndrome. No significant difference
was detected between groups as regard side effects or complica-
tions. No technical difficulties or complications were faced during
procedure performance.
9. Discussion
The anatomical location of stellate ganglion and its close prox-
imity to vital neurovascular and other structures making the stel-
late blockade vulnerable to a lot of serious hazards, hence, an
imaging guidance is a routine nowadays. Traditionally interven-
tional pain procedures including stellate ganglion blockade are
done under fluoroscopy. However this standard guidance cannot
visualize the soft tissues and exposes both the patient and workers
to risk of irradiation. CT-guidance on the other hand, is speculated
to major blocks such as coeliac plexus block and percutaneous cer-
vical cordotomy as it exposes the patient and operators to risk of
radiation and it is not easily available [11]. MRI-guided work is
time consuming, cost-ineffective and impractical [10].
More recently, the use of ultrasound guidance has allowed stel-
late ganglion block to be performed without vascular or nerve
injury [12,13]. Ultrasound can precisely locate the blood vessels
(carotid, vertebral and thyroid vessels), nerves (cervical nerve
roots, phrenic, recurrent laryngeal and divisions of the brachial
plexus) together with other important soft tissues (longus colli
muscle, esophagus, trachea, mediastinum and thyroid gland) [10].
In the current study, forty patients, were treated using PRF of
the stellate ganglion for control of chronic neuropathic pain syn-
drome, there is significant reduction of VAS, morphine consump-
tion and pregabalin consumption after the block as compared
with pre block values.
These findings agreed with Kim et al. studying Twelve CRPS
patients who underwent PRF on the cervical sympathetic chain .
The pain intensity decreased significantly at 1 week after the pro-
cedure. In his study, overall, 91.7% of patients experienced at least
moderate improvement [12].
Imani et al., studying 14 patients with CRPS underwent stellate
ganglion block divided into 2 equal groups according whether flu-
oroscopic or ultrasound guided procedure found a meaningful sta-
tistical difference among patients of any group in terms of the pain
intensity (before the block until six month after the block) [14].
However, in the present analysis, there is no significant differ-
ence between the fluoroscopic guided group and the ultrasound
guided group regarding the pain scores and the morphine con-
sumption dose before and after the procedure and all over the
study period, agreeing with Imani et al., which stated no meaning-
ful statistical difference between the patients of ultrasound and
fluoroscopic guidance groups, in terms of the pain intensity (from
one week to six months after block) [14].
In the current study, the procedure time was significantly lower
in the ultrasound guided block compared with the fluoroscopic
guided block. The results of this study showed shorter procedure
time in ultrasound guidance may be due to different views taken
by fluoroscopy with and without dye.
Also the incidence of hematoma formation and hoarseness of
voice was lower in the ultrasound guided block group. Agreeing
with Lee et al. who stated that the ultrasound technique improves
the safety of the procedure through the direct visualization of the
related soft tissue structures that are not visualized with fluo-
roscopy [15].
The source of such Hematoma could be due to injury of the thy-
roid tissue or thyroid vessels [16,17]. Kapral et al. compared sonar
guided stellate block with blind technique and found that good
block with sonar guidance with 100% success rate and lower inci-
dence of hematoma with 3 patients in blind group and none in US
group [16]. During Fluoroscopic technique, the patient neck is
hyperextended by small pillow under the shoulder to stretch the
esophagus in the midline behind the trachea away from the needle
path. However, the risk of esophageal penetration and conse-
quently mediastinal infection and emphysema is still a problem
[10]. US easily identify the esophagus and thus will be of great
value in cases of esophageal diverticulum [10].
PRF on the cervical sympathetic chain therefore appears to be a
valid option for the management of CRPS of the upper extremities,
and the incorporation of ultrasound can increase the ease and
safety of this procedure [12].
Limitations of the current study is the inability to assess long
term effects of the treatment, as the follow up period was limited
to 3 month post procedure.
To sum up, the present results support the opinion that sympa-
thetic blockade in stellate ganglion using pulsed radiofrequency is
of reasonable efficacy in treating chronic neuropathic pain whether
Table 1
Demographic characteristics.
Variable Group F (n = 20) Group U (n = 20)
Age (year) 44.3 ± 8.3 46.1 ± 11.4
Male/female ratio 6/14 7/13
Body weight (kg) 67.6 ± 7.4 72.5 ± 7.2
Type of neuropathic pain
Post mastectomy 12 11
CRPS 6 8
Phantom pain 2 1
Data are represented as mean ± standard deviation, ratio or number.
*
p < .05 in comparison between both groups.
Table 2
Duration of the procedure (min).
Duration of the procedure (min) 15.5 ± 3.2 11.7 ± 2.4
*
Data are represented as mean ± SD.
*
p < .05 in comparison between both groups.
74 M.H. Shaaban et al. / The Egyptian Journal of Radiology and Nuclear Medicine 49 (2018) 71–75
fluoroscopic or ultrasound guided, yet ultrasound takes consider-
ably less time with less complications.
Conflict of interest
We the authors, here, admit that there is no conflict of interest
encountered during the study done in this research article.
References
[1] Raj Prithvi, Erdine Serdar. In: Pain-relieving procedures: the illustrated guide.
1st ed. Published 2012 by John Wiley and Sons Ltd.; 2012. p. 207–17 [chapter
13: Interventional pain procedures in the Neck].
[2] Gibbs GF, Drummond PD, Finch PM, Phillips JK. Unravelling the
pathophysiology of complex regional pain syndrome: focus on
sympathetically maintained pain. Clin Exp Pharmacol Physiol
2008;35:717–24.
[3] Erdine Seradar. Sympathetic blocks of the head and neck. In: Raj P Prithvi, Lou
Leland, Erdine Seradar, Staats Peter S, Waldman Steven D, Racz Gabor,
Hammer Michael, Niv David, Ruiz-lopez Ricardo, Heavner James E, editors.
Interventional pain management, image-guided procedures. 2nd ed. Saunders
El Sevier; 2008. p. 115–25 [chapter 7].
[4] Day M. Sympathetic blocks: the evidence [review] [published correction
appears in pain pract, 2008; 18(4): 335-336]. Pain Pract 2008;8:98–109.
[5] Wiliams PL. Gray’s anatomy. 38th ed. New York: Churchil Livingstone; 1995.
[6] Waldman SD, editor. Interventional pain management. Philadolphia: WB
Saunders Company; 2001.
[7] Chua Nicholas HL, Vissers Kris C, Sluijter Menno E. Pulsed radio frequency
treatment in interventional pain management: mechanisms and potential in
dications – a review. Acta Neurochir 2011;153:763–71.
[8] Soledad Cepeda M, Carr Daniel B, Lau Joseph. Local anaesthetic sympathetic
blockade for complex regional pain syndrome. Cochrane Rev (January);2010.
[9] Manchikanti L. The role of radio frequency in the management of complex
regional pain syndrome. Curr Rev Pain 2000;4:437–44.
[10] Narouze S, Vydyanathan A, Patel N. Utrasound-guided stellate ganglion block
successfully prevented esophageal puncture. Pain Phys 2007;10:747–52.
[11] Narouze SN. Ultrasound – guided interventional procedures in pain
management: evidence-based medicine. Reg Anesth Pain Med 2010;35
(2):55–8.
[12] Kim E, Yoo W, Kim Y, Park H. Ultrasound-guided pulsed radiofrequency
treatment of the cervical sympathetic chain for complex regional pain
syndrome. A retrospective observational study. Medicine 2017;96:1. e5856.
[13] Narouze S. Ultrasound-guided stellate ganglion block: safety and efficacy. Curr
Pain Headache Rep 2014;18:424.
[14] Imani F, Hemati K, Rahimzadeh P, RezaKazemi M, Hejazian K. Effectiveness of
stellate ganglion block under fuoroscopy or ultrasound guidance in upper
extremity CRPS. J Clin Diagn Res 2016;10:9–12.
[15] Lee M, Kim K, Song J, Jung H, Lim H, et al. Minimal volume of anaesthetic
required for an ultrasound guided stellate ganglion block. Pain Med
2012;13:1381–8.
[16] Kapral S, Krafft P, Gosch M, Fleischmann D, Weinstabl C. Ultrasound imaging
for stellate ganglion block: direct visualization of puncture site and local
anesthetic spread. Reg Anesth 1995;20:323–8.
[17] Higa K, Hirata K, Hirota K, Nitahara K, Shono S. Retropharyngeal hematoma
after stellate ganglion block. Anesthesiology 2006;105:1238–45.
Table 3
The median VAS score values in the two studied groups.
Group/time Before the block 24 h after 1 week after 1 month after 3 month after
Group F (n = 20) 8.5 (7, 9) 1 (0, 2)
*
2 (1, 4)
*
3 (2, 4)
*
3 (2, 5)
*
Group U (n = 20) 8 (8, 9) 1 (0, 3)
*
2 (1, 3)
*
4 (1, 5)
*
4 (2, 5)
*
Data are represented as median (IQ).
*
P < .05 in comparison to before block value in the same group.
Table 4
Mean daily morphine consumption, in mgs, in the two studied groups.
Group/time Before the block 1 week after 1 month after 3 month after
Group F (n = 20) 75 ± 12.4 20 ± 4.3
*
30 ± 6.4
*
35 ± 7.8
*
Group U (n = 20) 80 ± 10.6 25 ± 5.4
*
30 ± 4.4
*
35 ± 6.7
*
Data are represented as mean ± SD.
*
p < .05 in comparison to before block value in the same group.
Table 5
Mean daily pregabalin consumption in the two studied groups.
Group/time Before block 1 week after 1 month after 3 month after
Group F (n = 20) (mg) 375 ± 30.5 120 ± 15.7
*
145 ± 30.5
*
165 ± 35.8
*
Group U (n = 20) (mg) 405 ± 35.7 135 ± 20.8
*
175 ± 45.6
*
190 ± 40.6
*
Data are represented as mean ± SD.
*
p < .05 in comparison to before block value in the same group.
Table 6
Side effects and complications.
Side effect/group Group F (n = 20) Group U (n = 20)
Pain at puncture site 8 (40%) 11 (55%)
Temporary horner syndrome 17 (85%) 16 (80%)
Hematoma 2 (10%) 0
Hoarseness of voice 1 (5%) 0
Data are represented as number (percentage).
M.H. Shaaban et al. / The Egyptian Journal of Radiology and Nuclear Medicine 49 (2018) 71–75 75
