1.1

Introduction

Complex regional pain syndrome (CRPS) can occur in an extremity after minor injury, fracture, surgery, peripheral nerve insult or spontaneously and is characterised by spontaneous pain, changes in skin temperature and colour, oedema, and motor disturbances . Pathophysiology is likely to involve peripheral and central components, and neurological and inflammatory elements , incidence has been estimated as 26.2 per 100,000 person-years in the Netherlands, with fracture accounting for 44% of cases . There is no consistent approach to treatment with a wide variety of specialists involved. A systematic review with meta-analysis supported the use of spinal cord stimulation in patients with CRPS type I (Grade A evidence) although this is too costly a procedure for most CRPS patients . Medical management includes opioids, anticonvulsants, corticosteroids and bisphosphonates. Sympathetically maintained pain may occur in CRPS type I and interventional management, including sympathetic plexus blockades, have been described with varied success . A systematic review of 21 Randomised Controlled Trials (RCT) concluded that there were no significant analgesic effects by sympathetic suppressing agents . However, a more recent Cochrane review found only two small RCTs meeting the inclusion criteria and concluded that there was a lack of evidence to support the use of local anaesthetic sympathetic blockade for CRPS . It can be logically argued that clinicians should not deliberately limit their therapeutic options unnecessarily, particularly in long-term, complex and established cases where single modality treatment employed in isolation is unlikely to offer the optimal degree of improvement for the patient. In all likelihood, a considered combination of therapeutic, pharmacological and invasive interventional procedures may have the greatest probability of achieving this aim.

The International Association for the Study of Pain (IASP) supports the use of sympathetic blocks to reduce sympathetic nervous system overactivity and relieve CRPS symptoms . Educational reviews promote stellate ganglion blockade (SGB) as beneficial . Recently, the stellate ganglion has been implicated in generating a cascade of physiological events mediated by nerve growth factor (NGF) leading to the symptoms associated with hot flashes, posttraumatic stress disorder and CRPS . It was suggested that applying local anaesthetic to the stellate ganglion would reverse this cascade. In this case report, we describe the use of stellate ganglion block to reduce overactivity of the sympathetic nervous system. This case supports the notion that SGB is preferable to upper limb intravenous regional anaesthetic block (IVRB) for refractory index finger pain associated with CRPS.

1.2

Case report

A forty-year-old, right hand dominant horsewoman presented to Accident & Emergency one week after twisting her left index finger in the bridle of a bolting horse. The finger was swollen and bruised around the proximal interphalangeal (PIP) joint and there was discernable tenderness in the metacarpophalangeal (MCP) joint and the second metacarpal. The collateral ligaments demonstrated no abnormal laxity, capillary refill was normal and there was no bony injury on X-ray. Neighbour strapping failed to adequately support the distal phalanx and therefore a decision had been made to place the digit in a mallet splint for five weeks. At this time, there was no reference made to any rehabilitation advice offered.

Seven weeks post-injury, the finger remained painful, swollen and stiff and an internal referral was made to the Plastic, Reconstructive and Hand Surgery department. Case notes revealed that a ring block with 2% lidocaine and 0.5% chirocaine was performed on the digit and obtained “nearly full range” (unknown whether this was active or passive range). Benefit was transient and additional specialist hand physiotherapy achieved little, due to the limitations of the patient experiencing extreme pain with mobilisation. An ultrasound scan of the muscles of the left hand had showed “no clear evidence of any significant flexor tendon injury, nor abnormality of the pulleys. There is possible thickening of the A3 pulley on the radial side which could indicate previous soft tissue injury, but this is a subtle abnormality”. Given the lack of progress impeded by suboptimal pain control, a referral was made to the Pain Management Service. Maintenance of the patient’s unaffected function was encouraged and best function of the impaired area within reasonable tolerance of discomfort was promoted.

The patient presented to the Pain Management Service 3 months after twisting her left index finger. The patient’s past medical history included a fall from a horse 8 years earlier, sustaining a spinal injury around the T8/9 level, which required immobilisation, hospitalisation and investigation. No fracture was detected upon X-ray of cervical, thoracic and lumbar regions. MRI of the cervical spine was unremarkable, though a scoliosis centering on T5 and concave to the right was noted. However, the medical records confirmed that the patient had presented shortly after with “a cold sweaty left hand. Pulses and tone normal. Small muscle atrophy. Grip power reduced. C8/T1 sensation reduced. Co-ordination normal”. Lower brachial plexus nerve conduction studies were reported as normal. Amitriptyline 25 mg had been commenced at that time. The problem appeared to resolve as there was little further reference to this in further medical record entries. As a consequence of the spinal insult, the patient had chronic back and thoracic wall pain for which she had in the past received repeat intercostal blocks with little clinical benefit before commencing gabapentin. Due to incomplete bladder emptying, the patient self-catherised four to five times daily and was prone to recurrent urinary tract infections. The patient had dyspareunia and recurrent abdominal pain. This ultimately and coincidentally culminated in the surgical implantation of a sacral cord stimulator under the care of urologists.

The patient was seen in the CRPS clinic of the Pain Management Service by the clinical physiotherapy specialist. Examination of the left index finger revealed a shiny, red index finger, which was reduced in local temperature, and held rigid in an extended position. There was a degree of hypervigilance from the patient regarding exposure of the finger to potential stimulus. There was a localised increase in perspiration but no trophic change. A small swelling was apparent in the radial aspect of the proximal third of the digit. Extreme pain prevented any active or passive movement of the MCP, PIP or distal interphalangeal (DIP) joints with obvious allodynia within the digit and a sense of dysaesthesia within the palmar tissues over the second metacarpal. A clinical diagnosis of CRPS was suspected based upon the observation above of sensory signs and symptoms, autonomic and motor abnormalities in keeping with reported common clinical characteristics . It was considered that the fall from a horse 8 years earlier and the subsequent spinal trauma and its associated reported symptomology may have generated central sensitisation predisposing the patient to neuropathic and sympathetically mediated pain resulting from future injuries, such as twisting the left index finger in the bridle of the bolting horse. However, it is acknowledged that both the specific precipitating injury and the splinted immobilisation as part of initial treatment could have created and contributed to the peripheral digital nerve disturbance subsequently encountered. The previous spinal trauma may be an interesting but entirely coincidental finding being hard to prove or disprove within the case study.

On presentation, the patient’s medication was: gabapentin 600 mg tds, codeine/acetaminophen 30/500 × 2 bd, sustained release morphine 10 mg/5 mL for breakthrough thoracic pain only on an approximately once per month basis, amlodipine 5 mg od and trimethoprim 200 mg as required for urinary tract infections. After primary multidisciplinary team (MDT) case discussion between the pain management clinicians, initial medicines management included an increase of gabapentin to 2400 mg daily total and an increase in codeine/acetaminophen to a qds regime, to try and provide more optimal pain control and exert a greater normalisation of the neuropathic symptomology. Unfortunately there was no improvement in patient reported pain reduction or findings on examination. This may be related to the complex multipathophysiology of CRPS, given that there is some suggestion that CRPS is not strictly neuropathic in nature and that peripheral inflammatory and immune mechanisms play a part. The service has had limited and disappointing past experience with the use of oral steroids in cases of refractory CRPS. Only very short duration benefit has been derived with a significant flare of symptoms upon graded reduction of the reducing course when prescribed in conjunction with our rheumatology colleagues. We did not in this case pursue this line of treatment.

A trial of acupuncture was declined by the patient because of a previously unfavourable and unconnected experience with acupuncture. The patient agreed to a series of left upper limb IVRBs using a mixture of 20 mL 0.5% lidocaine, 100 mg of bretylium and 20 mg ketorolac, under sterile technique and with intravenous access via left hand using a 22 gauge cannulae. A total of three blocks were conducted at monthly intervals. The aim of these blocks was to deplete noradrenaline at the axon terminals of the postganglionic sympathetic efferent axons, which would reduce sympathetic activity to the finger. Guanethidine is often considered the mainstay for IVRB and is both widely accessible and frequently used in the clinical setting. Guanethidine is rapidly taken up from the venous system by the peripheral nerves, having an immediate action locally upon the sympathetic nerves during the block. A more delayed action is achieved within the dorsal root ganglion within the spinal cord over the subsequent twenty-four hours post-block providing greater potential sympathetic blockade. However, a synthesis of evidence from 21 RCTs on its efficacy was not conclusive . Thus, it was decided to use bretylium based on clinical experience within our Pain Management Service and on the evidence from a small number of trials that it may provide some relief . Physiotherapeutic mobilisation techniques were performed to the joints and tissues affected under block conditions within patient tolerance of discomfort. Between 30 to 50% of range was achieved in all affected joints under supervision from the physiotherapist. However, notwithstanding comprehensive advice and exercise instruction, the patient was unable to adhere to the finger exercise regime at home despite understanding it’s importance and being highly motivated; the pain was simply too overwhelming. As a result, pain continued to present a significant barrier to recovery and the finger remained functionless.

Following MDT review including the patient, a left SGB was proposed. Three blocks were given at 8, 10 and 13 months after the initial injury under local anaesthesia and sterile conditions and taking an anterior approach with a 22 gauge Yale needle, with the patient in a supine position. A solution of 10 mL 0.25% bupivicaine and 80 mg of methylprednisolone was used (NB: the service has since moved to using non-particulate steroid solutions). The rationale for undertaking the SGB was to reduce the overactivity of the sympathetic nervous system by blocking activity in paravertebral ganglion that project to the finger. Physiotherapy was repeated under block conditions to maximise joint and tissue mobility.

The first SGB achieved excellent relief in the MCP & DIP but the PIP remained exceedingly painful. The reason for this is unclear. Ninety degrees of movement was obtained at the MCP joint and twenty degrees at the others when passively mobilised by the therapist. This could then be actively achieved through repetition by the patient by the end of the session.

The second SGB achieved significant overall reductions in pain, improved colour, reduced local finger perspiration and demonstrated full DIP range, ninety degrees at PIP and eighty degrees at the MCP. The third SGB added further improvement with normalisation of colour (except in extremes of external temperature), reduced swelling and pain and almost full active finger flexion to achieve a closed fist.

Upon clinic review some 3 months after the third SGB, the patient reported negligible pain, full finger function and displayed a full range of movement. Early morning index finger stiffness was present but the patient was immensely satisfied with the outcome in what had been a complex and protracted case.

1.3

Discussion

This case demonstrates the need for practitioners from all disciplines to be able to identify the clinical characteristics of CRPS to instigate immediate treatment. Diagnosis can be difficult, with over diagnosis resulting from undue emphasis placed on pain disproportionate to an inciting event despite the absence of other symptoms. CRPS can be missed when subtle symptoms are not recognised. It took over one year before negligible pain, full finger function and full range of movement was achieved, even at the hands of pain specialists. Because the pathogenesis is complex and multiple mechanisms have a role, treatment needs to be tailored to each individual patient and in a multidisciplinary setting. This case highlights the need for a combination of interventional techniques and sustained physiotherapy (both supervised and self-directed) to achieve restoration of full function of the extremity.

SGB may be overlooked as a viable treatment option because of little published evidence on its use for CRPS, despite educational reviews, which promote SGB as being beneficial . In this case, SGB was only used when more common interventions failed. In 2001, a meta-analysis of 21 RCTs evaluating sympathetic suppressing agents for CRPS included only one RCT on stellate ganglion block, which was conducted in 1983. The RCT found no differences in pain, hyperpathia, allodynia, vasomotor disturbances, trophic changes, oedema and limited motion at the end of SGB treatment and at one month and three months follow-up when compared to intravenous guanethidine block . Since then, there has been one additional clinical trial that examined the effect of three SGB’s performed at weekly intervals in 25 patients with CRPS in one hand . Ten patients had complete symptom relief, nine had partial relief and six had no relief. Skin perfusion, measured by Laser Doppler fluxometry, was greater in the affected hand when compared to the normal hand for patients reporting complete symptom relief. It was also found that symptoms lasting greater than 16 weeks before SGB intervention reduced the efficacy of SGB therapy. Our case highlights the need for an RCT to examine the efficacy of SGB compared to IVRB.

The physiological rationale for SGB efficacy in CRPS is strong. The SG innervates the ispilateral head, neck, upper thorax and arms. Traditionally it was thought that SGB caused sympathetic inhibition of these regions resulting peripheral vasodilatation . More recently, evidence suggests that the noxious insult results in CRPS, which is centrally mediated involving a cascade of events mediated by increased levels of NGF in the stellate ganglion . Sympathetic sprouting with elevated levels of brain and peripheral noradrenaline follow and these substances are known to be associated with symptoms of CRPS. It has been hypothesised that the efficacy of SGB is a result of the action of local anaesthetic at reducing NGF and thus inhibiting this physiological cascade . Studies using the pseudorabies virus have identified neural connections between the SG and brain areas associated with body temperature and regional pain . A SGB could be supplemented with a sympathetic block specifically to the upper Thoracic spine to ensure as much sympathetic fibre activity to the affected region is suppressed as possible however this would inevitably increase the risks associated with procedure performance.

SGB is not without risk, although significant complication such as vertebral artery injection, subarachnoid block, pneumothorax, phrenic nerve block or recurrent laryngeal block are rare. Our case highlights that SGB can produce complete symptom relief even with a delay of more than 16 weeks. Ideally the procedures would be performed with a higher frequency over a shorter timescale to attempt to achieve the most positive outcome however this can be significantly challenging within an NHS facility of finite resources and a compulsory target-driven system to balance the aspirational with the essential. The service has since developed to allow more rapid access for such refractory cases. We recommend that SGB should be considered earlier in the patient’s care pathway as it compared favourably when compared to upper limb IVRA.