Interventional management of intractable sympathetically mediated ...

Anaesthesia, 2011, 66, pages 699–708 doi:10.1111/j.1365-2044.2011.06765.x .....................................................................................................................................................................................................................

ORIGINAL ARTICLE

Interventional management of intractable sympathetically mediated pain by computed tomography-guided catheter implantation for block and neuroablation of the thoracic sympathetic chain: technical approach and review of 322 procedures K. Agarwal-Kozlowski,1 D. E. Lorke,2,3 C. R. Habermann,4 J. Schulte am Esch5 and H. Beck5 1 Head of Department, Centre for Palliative Care and Pain Management (T.I.P.S!), Stade, Germany 2 Professor in Anatomy, Department of Cellular Biology and Pharmacology, College of Medicine, Florida International University, Miami, FL, USA 3 Professor in Anatomy, Department of Anatomy, Faculty of Medicine and Health Sciences, UAE University, Al Ain, United Arab Emirates 4 Professor in Radiology, Department of Diagnostic and Interventional Radiology, 5 Professor in Anaesthesiology, Department of Anaesthesiology, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany

Summary

We retrospectively evaluated the safety and efficacy of computed tomography-guided placement of percutaneous catheters in close proximity to the thoracic sympathetic chain by rating pain intensity and systematically reviewing charts and computed tomography scans. Interventions were performed 322 times in 293 patients of mean (SD) age 59.4 (17.0) years, and male to female ratio 105:188, with postherpetic neuralgia (n = 103, 35.1%), various neuralgias (n = 88, 30.0%), complex regional pain syndrome (n = 69, 23.6%), facial pain (n = 17, 5.8%), ischaemic limb pain (n = 7, 2.4%), phantom limb pain (n = 4, 1.4%), pain following cerebrovascular accident (n = 2, 0.7%), syringomyelia (n = 2, 0.7%) and palmar hyperhidrosis (n = 1, 0.3%). The interventions were associated with a total of 23 adverse events (7.1% of all procedures): catheter dislocation (n = 9, 2.8%); increase in pain intensity (n = 8, 2.5%); pneumothorax (n = 3, 0.9%); local infection (n = 2, 0.6%); and puncture of the spinal cord (n = 1, 0.3%). Continuous infusion of 10 ml.h)1 ropivacaine 0.2% through the catheters decreased median (IQR [range]) pain scores from 8 (6–9 [2–10]) to 2 (1–3 [0–10]) (p < 0.0001). Chemical neuroablation was necessary in 137 patients (46.8%). We conclude that this procedure leads to a significant reduction of pain intensity in otherwise obstinate burning or stabbing pain and is associated with few hazards. . ......................................................................................................

Correspondence to: Prof. H. Beck Email: [email protected] Accepted: 30 March 2011

The management of intractable pain syndromes is complex and does not necessarily follow conventional pain management strategies (such as the World Health Organization pain ladder). The use of combined analgesics with different mechanisms of action may be inadequate in complex pain management; indeed, their 2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

use may be hampered by side effects (e.g. from opioid analgesics). Minimally invasive blockade and ⁄ or ablation of nerves seem to offer a reasonable alternative [1]. In sympathetically maintained pain, sympathetic blockade may provide temporary or permanent pain relief. Such blockade is commonly performed through 699

Æ

K. Agarwal-Kozlowski et al. Computer tomography-guided catheter implantation Anaesthesia, 2011, 66, pages 699–708 . ....................................................................................................................................................................................................................

injection around the lumbar and cervical sympathetic chains. However, repeated intermittent injections without catheters carry the risk of accidental harm to organs with each puncture; this may be devastating if it occurs in the thoracic region. The use of a continuous technique of sympathetic blockade therefore seems to offer a desirable alternative in otherwise intractable pain. The sympathetic nervous system can initiate or maintain pain as sympathetic fibres coexist with primary afferents in peripheral tissues, especially along the vasculature. Symptoms of sympathetically mediated pain include burning or stabbing pain, allodynia and ⁄ or trophic disorders. Primary sympathetic neurons are prevalent in the entire thoracic and the upper two or three lumbar segments of the spinal cord. Preganglionic fibres leave the spinal nerve via the white rami communicantes to join the sympathetic trunk, consisting of paired interconnected ganglia. Here, they behave in different ways: (i) they form synapses with neurons of the nearest ganglion; (ii) they traverse this ganglion and ascend or descend to end in another ganglion; and (iii) after traversing the nearest ganglion, they leave the sympathetic trunk as preganglionic fibres in the splanchnic nerves to end at ganglia of autonomic plexuses located in the vicinity of the thoracic aorta and its major branches. Preganglionic fibres for the head and neck emerge from the spinal cord, ascending in the sympathetic trunk and synapsing in the cervical ganglia. They supply the upper limb from upper thoracic segments (T2-6 ⁄ 7) and ascend via the sympathetic trunk to synapse in its cervicothoracic ganglion. Postganglionic fibres pass to the lower trunk of the brachial plexus. Most of the vasoconstrictor fibres for the upper limb emerge in the second and third thoracic ventral roots; the arteries can be denervated by cutting the sympathetic trunk below the third thoracic ganglion or by cutting the ventral roots of the second and third thoracic spinal nerves intradurally. Considerable anatomical variability has been observed, which may explain the high percentage of recurrences of vasomotor activity in the hand after thoracic sympathectomy.

This retrospective review describes our technique and experience of using continuous local anaesthetic blockade followed by neuroablation of the thoracic sympathetic chain in 293 patients receiving computed tomography (CT)-guided percutaneous catheter implantation as opposed to established single-shot techniques or continuous epidural blocks. The safety and efficacy of this procedure were assessed in terms of establishing the diagnosis and providing relief from pain in patients referred to us with intractable pain syndromes thought to be sympathetically independent based on previous treatment. Methods

A thorough medical assessment is routinely performed before admitting patients to our inpatient department. This involves using the German pain questionnaire, taking a detailed history, performing a physical examination (including allodynia-testing, cold-warmdiscrimination and search for dysautonomic signs and symptoms), thermographic imaging, and evaluation by a consultant in psychology; quantitative sensory testing or other procedures are only performed in very rare cases, e.g. unspecified pain in patients with co-morbidities that may have to be treated as well. Patients are admitted for interventional pain management if their pain has not responded to more conservative techniques. All patients have to undergo diagnostic blockade of the sympathetic chain using continuous administration of local anaesthetic before neuroablation is performed (Fig. 1). Patients aged < 40 years are treated with local anaesthetic for at least 7 days before neuroablation is considered. Written informed consent for CT-guided implantation of a catheter next to the thoracic sympathetic chain under propofol sedation is obtained a day before the procedure according to German law. Blood samples are analysed to verify appropriate clotting and to monitor parameters of infection. Pregnancy tests are carried out in all fertile female patients having obtained informed consent.

Figure 1 Algorithm for treatment of intractable burning ⁄ stabbing pain. Before a patient is admitted to our pain clinic, a

complete medical history is taken and psychological and physical examinations are performed which include testing for alodynia, temperature discrimination, a search for signs and symptoms of dysautonomia, thermographic imaging and completion of the German pain questionnaire; quantitative sensory testing or other procedures are only performed in very rare cases, e.g. unspecified pain with other co-morbidities. If the patient is eligible for the method described, physical examination and thermographic imaging are performed on a daily basis after implantation of the percutaneous catheter and continuous administration of local anaesthetic. Once a steady state and a sufficient decrease in pain intensity are established, the infusion is stopped to assess if pain reappears after 18 h. If the patient reports recurrence of pain during this time, neuroablation is then performed; if not, the patient is discharged and monitored by the primary physician. 700

2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

Æ

Anaesthesia, 2011, 66, pages 699–708 K. Agarwal-Kozlowski et al. Computer tomography-guided catheter implantation . ....................................................................................................................................................................................................................


701

Æ


We retrospectively analysed the charts and CT scans of all patients who underwent this procedure between 2002 and 2007. Data logs especially designed for documentation of procedures and outcome of interventional pain therapy as well as anaesthetic charts were reviewed. Review was also made of documents with numeric rating scales for the assessment of pain intensity that are routinely completed by our patients, on an hourly basis, from their admission to discharge. Concomitant medications are not altered during the course of treatment. In the event of sufficient pain relief resulting from the procedure, patients’ systemic analgesic regimes are reviewed and altered by their primary physician following discharge. Physiotherapy and psychological guidance are offered according to the patients’ co-morbidities. The technique of catheter placement for continuous blockade of the sympathetic thoracic chain has been performed and refined at our department since the late 1980s. Having identified what we believe to be the ‘ideal technique’ as well as the ‘ideal algorithm’, we have produced specific diagnostic and therapeutic algorithms to aid and standardise this approach. Procedures are performed using a multirow detector CT. A lateral scout view is assembled (slice collimation, 1 mm; 120 kV, 530 mA, field of view, 512 mm; rotation time, 0.5 s) before scanning the patient’s chest (120 kV, 240 mA, slice collimation 2.5 mm, increment 5 mm, rotation time 0.5 s). For detailed delineation of lung parenchyma and detection of a possible pneumothorax the reconstruction algorithm utilised is B80f (window width 1600, window centre 606 Hounsfield Units). For repeat needle localisation, the region of interest close to the expected needle position is scanned using a slice thickness of 2.5 mm with five slices acquired. After establishing peripheral venous cannulation and monitoring of vital signs, the patients are placed in the prone position. Depending on the underlying disease and ⁄ or dermatomes to be treated, a metal clip with a length of approximately 20 cm and diameter of 3 mm is placed 4–6 cm lateral to the spinous process of the thoracic vertebrae just before commencing CT imaging of the chest. Transverse CT imaging is performed from the upper to lower aspects of adjacent vertebrae to identify the ideal position for needle insertion without posing hazards or interfering with neighbouring structures. Distances from the skin to the anterolateral surface of the vertebra and nearby structures (e.g. the lung or large vessels) are measured. The desired direction of the Tuohy needle is then simulated on the screen (Fig. 2a) and the puncture site is defined and marked on the 702

patient’s skin with a permanent marker. The skin is then prepared with antiseptic solution before draping. Propofol is administered in small doses by a second anaesthetist to obtain light sedation while maintaining spontaneous breathing. The sedative is not infused continuously to prevent cessation of breathing. A 16-G, 120-mm Tuohy needle is then marked along its shaft with the distances derived from the CT scans. It is advanced until bony contact is established and then ‘walked’ down the surface of the vertebra. If the lung parenchyma prohibits further advancement as verified by a CT scan, the stylet is removed and 40 ml isotonic saline solution administered to detach the parietal pleura from the vertebra and allow safe passage of the needle (Fig. 2b,c). Once the correct position of the tip of the needle is confirmed by CT scan, an 18-G epidural catheter is inserted 2–3 cm past the orifice before withdrawing the needle. Accurate positioning is verified after subcutaneous tunnelling (Fig. 2d) before 10 ml contrast dye is administered to demonstrate correct positioning of the catheter tip (extravascular and extrapulmonary) and to simulate the intended path of local anaesthetic (Fig. 3). This demonstrates that the desired effect is obtained (with the catheter in its final position) by blocking the thoracic sympathetic chain instead of somatic nerves or nerve roots exiting from the spinal cord or through retrograde distribution and epidural involvement. The entire procedure takes approximately 30 min. A loading dose of 10 ml ropivacaine 1.0% is administered as soon as the patient is in the anaesthetic surveillance unit. Anaesthetic, motor and sensory radicular effects are evaluated by rating grip strength and performing cutaneous pin pricks. Contact-free infrared thermographic imaging is used to verify an increase in cutaneous blood flow and local anaesthesia [2]. A battery-driven external pump is connected to the catheter for continuous infusion of ropivacaine 0.2% at a rate of 10 ml.h)1. The dose and dose adjustments for the injected agents are chosen to avoid toxic effects; this scheme is continued for the duration of the observations in order that the conditions are comparable. Horner’s syndrome and hoarseness of the voice may develop, especially if catheters are placed above the fourth thoracic vertebra; however, they are usually well tolerated by the patient if they are warned of the risks in advance and reassured that symptoms are reversed once the infusion of local anaesthetic has stopped. If patients do not tolerate this, or if their speech ⁄ swallowing are impaired, an infusion rate of 2 ml less than the initial rate is chosen. 2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

Æ


(a)

(b)

(c)

(d)

Figure 2 (a) Measurements and sim-

ulation for needle insertion. The exact distances are measured from the metal marker to the intended point of puncture. The direction of the needle is simulated on the screen. (b) CT images showing the relationship of lung and vertebra. The needle is inserted as far as possible without damaging the lung. (c) CT images following injection of 40 ml saline 0.9%. Note the distance of the lung from the vertebra that allows the needle to pass without causing harm or pneumothorax. (d) CT images showing the catheter in place after subcutaneous placement to avoid infection.

Figure 3 CT images showing the administration of contrast

dye. Several slices are produced (not shown here) to be able to judge the contrast distribution in three dimensions.

Neuroablation of the thoracic sympathetic chain In deciding whether or not to carry out a sympathectomy, adequate indications for the procedure and the likelihood of its success must be assessed. Patients who benefit from continuous blockade of the thoracic sympathetic chain will most probably also respond to 2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

neuroablation [3, 4]. Numerous methods of disruption of the sympathetic nervous system are feasible, including surgical sympathectomy, percutaneous chemical neurolysis and radio frequency thermoablation [3]. To evaluate if the patient’s pain has been overcome by the continuous local anaesthetic infusion or if it will recur on cessation, the infusion of local anaesthetic is stopped 18 hours before neuroablation to assess the patient in the absence of the anaesthetic effects of ropivicaine. Neuroablation is only carried out if the pain intensity increases during this period. If the same kind of pain recurs in the same region as before the temporary block, neuroablation with ethanol is then performed (phenol is not approved by local authorities for neurolytic procedures in non-malignancy pain syndromes in Germany) [5, 6]. All precautions and preparations are completed as described for implantation of the catheter. After administering 0.5 ml contrast media, CT imaging of the adjacent 2 cm above and below the entrance site of the catheter is performed to validate correct positioning of the catheter tip. With the catheter still in place, 2 ml ethanol 95% mixed with contrast dye (resulting in an 80.75% solution) is injected for segments T2-4, or 5 ml of this solution for the caudal to T4 segments. Distribution of the agent is verified by CT scans. The catheter is left in place so that infusion of local anaesthetic can be resumed if pain 703

Æ


relief is inadequate, in which case the procedure is repeated the following day as described above. Once neuroablation has resulted in adequate pain relief, the catheter is removed and the patient discharged the day after the final procedure. In the event of catheter displacement, it is removed and a single-injection neuroablation technique is performed using a 12-cm 22-G spinal needle. The preparation and performance of this single-injection technique are the same as those used when inserting the needle for the continuous block procedure. Once the tip of the needle is placed in the correct position, either 2 ml or 5 ml ethanol 80.75% is injected as described above. For this study, data on procedure complications (e.g. pneumothorax, catheter dislocation, failure of treatment, infection) were collated from the patient documentation. The efficacy of the procedure was rated before treatment, throughout the course of hospitalisation, and at the end of treatment, using a numeric rating scale (range 0–10, with 0 signifying the absence of pain, and 10 maximum pain); efficacy was assessed by comparing results using the Wilcoxon test. A value of p < 0.05 was defined as being statistically significant. There was no control group as patients are referred for continuous blockade of their thoracic sympathetic chain by their primary physician only when no alternative treatment can be offered. Results

A total of 322 interventions were performed in 293 patients. Patient’s characteristics and details of the

origins of their pain are shown in Table 1. The median (IQR [range]) time for patient referral after being diagnosed with their pain syndrome was 9 (3–24 [0–218]) months; they were hospitalised for 11 (8–18 [3–55]) days and received continuous infusion of local anaesthetic to their thoracic sympathetic chain for 9 (7–15 [2–53]) days. Needle insertion for the procedures was performed from the 2nd to the 10th thoracic vertebra according to the underlying condition and the affected dermatomes. Catheters were not placed cranially to the second thoracic vertebra as neuroablation at these levels was considered to be unfeasible given the impact that the resulting Horner’s syndrome would have on the patients’ quality of life. Procedures caudal to the 10th thoracic vertebra were not required as dermatomes T11 and 12 were not affected in any of our patients. Pain intensity on the numeric rating scale was compared at admission and discharge. The characteristics of the reductions in pain scores and times taken to achieve these are demonstrated in Table 2. Pain scores were lower at discharge than on admission with a median (IQR [range]) reduction from 8 (6–9 [2–10]) to 2 (1–3 [0–10]) (p < 0.0001). One hundred and twenty-four patients (42.3%) stated that they experienced a reduction in pain intensity of more than 80%. Out of 227 (77.5%) patients who were discharged with values on the numeric rating scale equalling or less than three, a total of 70 (23.9%) patients left hospital with no pain (pain score of 0). In patients with ischaemic limb pain, the initial median (IQR [range]) scores on the numeric rating scale were lower than those found in the other groups

Table 1 Patients’ characteristics and details of pain subclassified according to underlying disease for patients undergoing

percutaneous sympathicolysis. Values are number, mean (SD) or median (IQR [range]).

Diagnosis

n

Age; years

PHN Neuralgia CRPS Facial pain Ischaemic limb pain Phantom limb pain Post-stroke pain Syringomyelia Palmar hyperhidrosis Total

103 88 69 17 7 4 2 2 1

71.5 53.2 49.7 64.0 52.3 50.8 72.5 43.5 36.0

(11.2) (15.5) (16.2) (15.9) (12.2) (6.9) (0.5) (0.5) (NA)

47 34 8 8 0 3 2 2 1

56 54 61 9 7 1 0 0 0

293

59.4 (17.0)

105

188

Males

Females

Time since diagnosis; months 22 4 7 22 24 93 12 5 2

Block only

(9–53 [1–218]) (1–8 [0–99]) (2–19 [0–103]) (4–24 [1–98]) (23–37 [15–78]) (69–105 [1–137]) (9–15 [6–18]) (4–5 [4–5]) (NA)

9 (3–24 [0–218])

Block followed by neuroablation

63 40 38 9 2 2 2 0 0

40 48 31 8 5 2 0 2 1

156

137

Time to obtain pain relief of 50%; h 203 189 213 218 303 218 171 4 4

(152–238 [2–400]) (87–220 [1–399]) (160–239 [2–399] (172–295 [79–387]) (265–341 [227–279]) (163–258 [109–299]) (157–185 [144–199]) (4–4 [3–5]) (NA)

216 (140–234 [1–400])

PHN, postherpetic neuralgia; CRPS, complex regional pain syndrome; NA, not applicable.

704


Æ


Table 2 Results of treatment

subclassified according to underlying disease in patients undergoing percutaneous sympathicolysis. Values are number or median (IQR [range]).

Diagnosis

Number

Initial NRS

End NRS

p value

PHN Neuralgia CRPS Facial pain Ischaemic limb pain Phantom limb pain Post-stroke pain Syringomyelia Palmar hyperhidrosis LA block only LA block followed by neuroablation Total

103 88 69 17 7 4 2 2 1 156 137

8 7 8 9 5 10 10 8 10 8 8

2 2 1 2 3 2 2 1 1 2 2

< 0.0001 < 0.0001 < 0.0001 0.0002 NA NA NA NA NA < 0.0001 < 0.0001

293

(6–9 [3–10]) (5–10 [4–10]) (7–9 [2–10]) (9–10 [6–10]) (4–7 [2–8]) (9–10 [6–10]) (10 [10]) (8 [8]) (NA) (6–9 [2–10]) (6–9 [2–10])

8 (6–9 [2–10])

(1–3 [0–9]) (0–5 [0–10]) (0–3 [0–8]) (1–2 [0–9]) (3–4 [2–5]) (2–4 [0–10]) (1–2 [0–3]) (1–2 [0–2]) (NA) (0–4 [0–10]) (1–3 [0–7])

2 (1–3 [0–10])

PHN, postherpetic neuralgia; CRPS, complex regional pain syndrome; NRS, numeric rating scale; NA, not applicable (sample size too small); LA, local anaesthetic.

(5 (4–7 [2–8]) vs 8 (6–9 [2–10]); p = 0.007). Moreover, in this group, the pain reduction achieved by the procedure was less (final scores of 3 (3–4 [2–5])) when compared with the other groups (2 (1–3 [0–10]); p = 0.051). It took 303 (265–341 [227–279]) h of continuous local anaesthetic infusion to reach a 50% reduction of pain intensity for patients with ischaemic limb pain, whereas patients with syringomyelia reported a comparable pain relief within only 4 (4–4 [3–5]) h (Table 1). This result is, however, based on data from only two patients with syringomyelia. Collating all data, continuous blockade of the sympathetic chain was required for a median of 216 (140–234 [1–400]) h to achieve a pain reduction of 50% and for 246 (169–267 [2–599]) h for patients to attain ‘sufficient pain reduction’. This finding is of particular interest as pain is usually classified as being sympathetically independent if it does not subside after single injection sympathetic blockade. Given that the effects of local anaesthetics typically last for 1–12 h (depending on the agent, dose and ⁄ or concentration), these patients would have been diagnosed with sympathetically independent pain, although their pain was successfully managed using our continuous technique of sympathetic blockade. Adverse events Computed tomography-guided placement of needles and catheters was viable in all cases. A total of three (0.9%) pneumothoraces occurred during 322 procedures but none required treatment as the intrapleural air volume was estimated to be less than 20 ml in every case. It is likely that these small volume pneumothoraces would not have been detected if the procedures 2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

had been performed using fluoroscopic guidance only or using subsequent chest radiography. Two cases of superficial infection were recorded (defined as cutaneous reactions with reddening of the primary puncture site), although no evidence of bacteria was found in smears or aspirates from the sites. One patient showed signs of infection, developing a fever up to 38.9 C; however, closer examination revealed a urinary infection. One case of accidental spinal cord puncture at the second thoracic vertebra occurred. Magnetic resonance imaging was carried out as recommended by the consultant neurologist and frequent neurological examinations were performed for 2 days. No neurological deficits were detected at any point up to 4 months after discharge. The catheters were removed from all patients with complications who were discharged at their earliest convenience. Their data, including pain scores, were included in our analysis. Catheter dislocation occurred nine times necessitating reimplantation or single-shot neuroablation in seven (0.41%) patients (some of the patients lost their catheter several times). At discharge, eight (2.7%) patients presenting with stabbing ⁄ burning pain of unknown origin stated that their pain intensity had increased following the procedure. Another 22 (7.5%) patients reported no change in pain intensity at discharge. Therefore, a total of 31 (10.6%) patients did not benefit from the procedure. There were no incidents of cardiovascular or pulmonary insufficiency, even if procedures were performed in an area where impact on cardiac function should be expected. Patients were able to move around 705

Æ


and undertake their normal daily routine within 2 h of implanting the catheter and initiation of sympathetic blockade or performing neuroablation. Discussion

Our report highlights the novel use of CT scanning to guide placement of percutaneous catheters to allow continuous infusion of local anaesthetic around the thoracic sympathetic chain. This technique combines the established procedures of ‘single-shot’ blockade of the sympathetic chain with the continuous infusion technique of blocking peripheral nerves or providing epidural anaesthesia using local anaesthetic. Our technique evolved due to the hazards involved with repeated needle punctures (i.e. single-shot procedures) in an anatomically dangerous area, as well as the fact that ‘single-shot’ procedures often prove ineffective due to the local anaesthetic effects’ wearing off rapidly. Our intention was to reduce pain refractory to more conservative procedures before subjecting patients to more invasive techniques (e.g. thoracoscopic sympathectomy, spinal cord stimulation etc), but with the suspicion that the time needed to treat sympathetically maintained pain may exceed the duration of effect brought about by a single exposure (injection) to local anaesthetics. Our interventional pain management techniques are primarily considered in patients with pain which is refractory to conventional treatment [7–9]. The aim of modern pain management is to move from empiric therapies to a mechanism-based approach [10]. Tests to assess sympathetically maintained pain involve trials of sympatholytic treatment as temporary relief of pain which, if successful, indicates that the sympathetic nervous system is involved. Intermittent blockade of the lumbar sympathetic chain, the stellate ganglion as well as the upper cervical ganglion [11] have become common procedures [12]. Usually, blocks are performed by (repeatedly) administering a single dose (single-shot) of local anaesthetic around the sympathetic chain. The duration of action of such blocks varies from 1 to 12 h depending on the active agent, dose and concentration used [3]. We believe that this duration of blockade may be too short and may explain failures of the technique in some patients. Our data demonstrate that continuous blockade was required for 216 (140–234 [1–400]) h to achieve a pain reduction of 50% and for 246 (169–267 [2– 599]) h for the patients to report ‘sufficient pain reduction’. This is significant as pain is usually classified as being sympathetically independent if it subsides after 706

‘single-shot’ blockade [12, 15]; by this criterion, the patients in our study would have been diagnosed as suffering from sympathetically independent pain, although their pain was, in the majority of cases, overcome by sympathetic blockade. Prolonged continuous therapy may result in decreased pain intensity and reduction or cessation of other analgesic medications. Using our technique, pain was significantly reduced in 263 (89.8%) patients. With these findings in mind, it would seem prudent to develop an algorithm for the diagnosis and treatment of sympathetically maintained pain. This should take into account the amount, concentration and type of drug injected including its duration of action. Techniques for the continuous administration of local anaesthesia using percutaneous catheters are common in anaesthesia and avoid the need for patients to be exposed to repeated single-shot injections. They also avoid the patient experiencing phases of pain as the local anaesthetic effects subside. Epidural techniques can also be used to alleviate sympathetically maintained pain [13]; however, our catheter technique has the advantage that it can be used to perform neuroablation if temporary blockade using local anaesthetic provides pain relief that returns on discontinuation of the infusion. It also avoids the adverse effects of epidural anaesthesia on motor and sensory function. Defining adequate indications and assessing the likelihood of success are the two most important factors in deciding whether or not to undertake sympathectomy for the purposes of pain relief. Patients who benefit from continuous blockade of their thoracic sympathetic chain by local anaesthetic most probably will respond to neuroablation [14]. There are numerous feasible methods of disrupting the sympathetic nervous system including surgical sympathectomy, percutaneous chemical neurolysis and radio frequency thermoablation [3]. We performed chemical neurolysis using ethanol, as the use of phenol for neurolytic procedures in non-malignancy pain syndromes is not approved by local authorities in Germany. It is unclear why some of our patients reported recurrence of pain after stopping the continuous infusion; it is possible that this results from some kind of neuromodulation or de-coupling of the efferent sympathetic and afferent sensory pathways. Steel and Burchiel [15] emphasised that sympathectomy only provides relief in the case of sympathetically maintained pain and that rigorous assessment is required before intervention to prevent treatment failure due to misdiagnosis. As success is improved by careful selection of patients and the incidence of adverse events is low, 2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

Æ


early definitive treatment (possibly including an ablative procedure) rather than expensive medical or rehabilitation therapy seems justified in patients suffering from sympathetically maintained pain syndromes. Our CT-guided technique for percutaneous needle and catheter insertion followed by confirmation of position by injection of radio-opaque contrast media produces a high success rate when compared with the landmark technique, but with low complications [16, 17]. Once sympathetically maintained pain has been confirmed, neuroablative procedures can be performed via the catheter. Several articles have been published on blockade of the thoracic sympathetic chain. Conventionally, these procedures have been performed under fluoroscopic guidance [18] and have used single-shot injection techniques [19] or catheters implanted epidurally [20]. The most common practice is still to use thoracoscopic sympathectomy [21, 22]. Our technique was not completely without complications; out of 322 interventions, pneumothoraces (n = 3) and accidental puncture of the spinal cord (n = 1) did occur. The lack of published data on adverse events following sympathetic blockade with or without radiological (fluoroscopic) guidance makes comparison of our complication rates difficult. Since the introduction of thoracoscopic sympathectomy techniques in the 1980s, only one review of the procedure’s safety has been published [23]; in this, the incidence of pneumothorax was given as 2%, though up to 75% of the patients had some residual intrathoracic gas at the end of the procedure that required drainage in 0.4–2.3% of cases; this compares with a 0.9% incidence of pneumothorax in our case series. Surgical emphysema is also a common adverse event during thoracoscopic sympathectomy, occurring in 2.7% of patients; no such events were observed in our patients. In a review of endoscopic thoracic sympathectomies performed in Japan, the risk of significant bleeding was found to be 0.3–5.3% [24]. Nine deaths were reported, due to haemorrhage (laceration of the subclavian artery, damage to the intercostal vein, unspecified bleeding), anaesthetic problems (hypoxaemia during bilateral sympathectomies) and cerebral ischaemia. The incidence of permanent Horner’s syndrome was 0.3%, probably induced by misidentification of the T2 ganglion, excessive traction on the chain or excessive use of diathermy. Another problem arising during thoracoscopic sympathectomy is that of postoperative pain requiring opioid analgesia [25] or tricyclic antidepressants [26]. None of the above events was found during our CT-guided method. 2011 The Authors Anaesthesia 2011 The Association of Anaesthetists of Great Britain and Ireland

In conclusion, our experience of blocking the thoracic sympathetic chain using infusion of local anaesthetic delivered through percutaneous catheters that have been inserted under CT guidance shows the incidence of adverse events to be low with no recorded fatalities or permanent disability. We recognise that the placement of these catheters, administration of local anaesthetics and subsequent neuroablation require adequate hospital facilities and a close interdisciplinary alliance. Nevertheless, we believe that our management presents a significant and safe alternative to conventional techniques for managing patients with sympathetically mediated pain. Further studies are desirable to identify the pathophysiological basis of our findings and how long the benefit from our technique is maintained. Acknowledgements

The authors would like to thank the reviewers for their expert guidance and Frank Kozlowski (digital artist, freelancer, Norderstedt, Germany) for his valuable aid on setting up pictures. The article is published with the written consent of the patients and approval of the local authorities for the reproduction of the CT images. No external funding and no competing interests declared. References 1 Ossipov MH, Porreca F. Challenges in the development of novel treatment strategies for neuropathic pain. Neurotherapeutics 2005; 2: 650–61. 2 Agarwal K. Medical applications. In: Vollmer M, Moellmann KP, eds. Infrared Thermal Imaging – Fundamentals, Research and Applications. Weinheim: Wiley-VCH, 2010: 535–48. 3 Boas RA. Sympathetic nerve blocks: in search of a role. Regional Anesthesia and Pain Medicine 1998; 23: 292–305. 4 Ramamurthy S, Walsh NE, Schoenfeld LS, Hoffman J. Evaluation of neurolytic blocks using phenol and cryogenic block in the management of chronic pain. Journal of Pain and Symptom Management 1989; 4: 72–5. 5 Arzneimittelkommission. 255 ⁄ 50 ⁄ 98, Phenol-haltige Sklerosierungslo¨sungen. Pharmazeutische Zeitung 1998; 143: 4322. 6 Huseboe S, Klaschik E. Palliativmedizin. Heidelberg: Springer, 2009: 220–3. 7 Chen H, Lamer TJ, Rho RH, et al. Contemporary management of neuropathic pain for the primary care physician. Mayo Clinic Proceedings 2004; 79: 1533–45.

707

Æ


8 Dworkin RH. An overview of neuropathic pain: syndromes, symptoms, signs, and several mechanisms. Clinical Journal of Pain 2002; 18: 343–9. 9 van Zundert J. Clinical reasearch in interventional pain management techniques: the clinician’s point of view. Pain Practice 2007; 7: 221–9. 10 Holdcroft A, Power I. Recent developments: management of pain. British Medical Journal 2003; 326: 635–9. 11 Elias M. Cervical sympathetic and stellate ganglion blocks. Pain Physician 2000; 3: 294–304. 12 Bernstein RM. Injections and surgical therapy in chronic pain. The Clinical Journal of Pain 2001; 17: S94–104. 13 Visser WA, Lee RA, Gielen MJM. Factors affecting the distribution of neural blockade by local anesthetics in epidural anesthesia and a comparison of lumbar versus thoracic epidural anesthesia. Anesthesia and Analgesia 2008; 107: 708–21. 14 Eldridge P. Neurosurgical techniques in the management of chronic pain. Anaesth Intensive Care 2008; 9: 65–8. 15 Steel TR, Burchiel KJ. Ablative neurosurgical techniques in the treatment of chronic pain: overview. In: Burchiel KJ, ed. Surgical Management of Pain. New York: Thieme Medical Publishers Inc., 2002: 633–46. 16 Haaga JR, Kori SH, Eastwood DW, Borowski GP. Improved technique for CT-guided celiac ganglia block. American Journal of Roentgenology 1984; 142: 1201–4. 17 Gangi A, Dietemann JL, Schultz A, Mortazavi R, Jeung MY, Roy C. Interventional radiologic procedures with CT guidance in cancer pain management. Radiographics 1996; 16: 1289–304. 18 Han KR, Kim C, Park EJ. Successful treatment of digital ulcers in a scleroderma patient with continuous bilateral

708

19

20

21

22

23

24

25

26

thoracic sympathetic block. Pain Physician 2008; 11: 91– 6. Andresen R, Radmer S, Nickel J, Fischer G, Brinckmann W. Ambulante CT-gestuetzte thorakale Sympathikusblockade als zusaetzliche Therapieoption bei komplexem regionalen Schmerzsyndrom Typ I nach Sportverletzungen. Sportverletzung und Sportschaden 2009; 23: 35–40. Clemente A, Carli F. The physiological effects of thoracic epidural anesthesia and analgesia on the cardiovascular, pulmonary and gastrointestinal systems. Minerva Anestesiologica 2008; 74: 549–63. Jeganathan R, Jordan S, Jones M, et al. Bilateral thoracoscopic sympathectomy: results and long-term follow-up. Interactive Cardiovascular and Thoracic Surgery 2008; 7: 67–70. Fibla JJ, Molins L, Mier JM, Vidal G. Effectiveness of sympathetic block by clipping in the treatment of hyperhidrosis and facial blushing. Interactive Cardiovascular and Thoracic Surgery 2009; 9: 970–2. Ojimba TA, Cameron AEP. Drawbacks of endoscopic thoracic sympathectomy. British Journal of Surgery 2004; 91: 264–9. Ueyama T, Matsumoto Y, Abe T, Yuge O, Iwai T. Endoscopic thoracic sympathicotomy in Japan. Annales Chirurgiae et Gynaecologiae 2001; 90: 200–2. Gossot D, Kabiri H, Caliandro R, Debrosse D, Girard P, Grunenwald D. Early complications of thoracic endoscopic sympathectomy: a prospective study of 940 procedures. Annals of Thoracic Surgery 2001; 71: 1116–9. Leao LEV, de Olivera R, Szulc R, Mari Jde J, Crotti PLR, Goncalves JJS. Role of video-assisted sympathectomy in the treatment of primary hyperhidrosis. Sao Paulo Medical Journal 2003; 121: 191–7.
