Stroke is a common cause of CRPS, and between 1.5% and 61% of patients following stroke have been reported to develop CRPS.6-8 Various management modes for CRPS have been introduced but only a few techniques have been effective. The sympathetic nervous system has been implicated in the pathophysiology of CRPS, and consequently, sympathetic nerve blockade (SGB) is widely used for the treatment of CRPS.9

Classical "blind" injection of anesthetic at the C6 level has a long history, but its reliability in achieving blockade of the sympathetic trunk and stellate ganglion remains uncertain. In addition, blind injection is associated with a variety of adverse effects and complications, such as intravascular injection, formation of hematomas and temporary paralysis of the recurrent laryngeal nerve.10-13

A method of stellate ganglion blockade under ultrasonography (US) guidance was described in 1995, but only recently gained popularity. Kapral et al.12 found that compared with the blind technique, ultrasonography-guided SGB used a lower volume of local anesthetics (5 ml rather than 8 ml) and there were more rapid onsets of Horner's syndrome. Application of US for sympathetic nerve blockade has proven to be useful for the swift and accurate placement of local anesthetics on a targeted area. Thus, it might reduce the required volume of local anesthetics compared with blind technique.

The inclusion criteria were the presence of regional pain and sensory changes of hemiplegic unilateral upper extremity following a stroke; pain was associated with findings such as abnormal skin color, temperature change, abnormal sudomotor activity or edema; no distribution of the pain of a single nerve in the extremity. Seventeen subjects were excluded from this study if they had a history of upper extremity cellulitis, other autonomic nerve disease, peripheral nerve disease, bleeding disorder or severe cognitive impairment.

We therefore studied 42 patients with CRPS following stroke [19 right (45.3%) and 23 left (54.7%)] and patient signed informed consent. All 42 patients had the modified diagnostic criteria for CRPS as International Association for the Study of Pain (IASP).14 Twenty-five patients were women (59.6%) and 17 (40.4%) were men. All patients were divided randomly into US-guided SGB group (22 patients) and blind SGB group (20 patients). The mean age was 61.3±5.6 (range 34-77) years and 59.1±4.5 (range 36-71) years and the mean duration since stroke occurred were 2.8±1.1 (range 1-7) months, 2.3±0.9 (range 1-6) months, respectively (Table 1).

Kapral et al.12 reported that US-guided SGB, by comparing with the blind technique, used a lower volume of local anesthetics and proved to be more effective. In this study, the blind SGB group patients and US-guided SGB group patients received the procedure using a volume of 10 ml of 0.5% lidocaine or 5 ml of 0.5% lidocaine, respectively. SGB was performed two times with an interval of 1 week between each treatments.12

The anterior paratracheal approach was used in the procedure on the cervical sympathetic chain without fluoroscopic guidance in blind group. A 22-gauge, 5-cm needle was perpendicularly inserted to the skin until bone contact was made and then withdrawn to rest anteriorly to the precervical fascia. The patient was encouraged to lie in the supine position for 3 minute after the injection and then developed injected-sided Horner's syndrome.15

In US-guided group, procedure was performed using same needle at same posture. After passing through the skin and subcutaneous tissue and stabilizing the needle, a 3-12 MHz linear array probe of HD11-XE® (Philips, Washington, USA) was used to verify the position of the needle. The needle was shown to be aiming towards the thyroid tissue anteriorly and then the esophagus posteriorly. At this point the needle was withdrawn, reinserted obliquely, and advanced with real-time US so that the needle tip lay anterior to the longus coli muscle (anterior to C6 transverse process) (Fig. 1). After negative aspiration, 5 ml of lidocaine 0.5% were injected.

The primary outcome was evaluated by pain intensity and the amount of hand swelling. Pain intensity was evaluated by using a 10 cm visual analog scale (VAS) in which 0 represents 'no pain' and 10 represents 'the highest pain'. Determination of the volume of the hands (hand volume) was assessed by a hand volumeter. The hand volumeter set (8×13×28 cm) used was made of translucent material (acrylic), "Hand Volumeter" (Volumeter Unlimited, Phoenix, USA) and filled with room temperature water.

The volunteer was positioned beside the instrument during the procedure and the hand also positioned vertically, to avoid contact with the sides of the equipment. The hand was immersed slowly into the volumeter until it was immersed into water 2 cm above wrist crease. Additionally, the severity score for CRPS that was introduced by Harden in 2010 was evaluated as secondary outcome measure (Appendix 1).16 All these parameters were checked before procedure, after 2 weeks and 4 weeks following the SGB.

SGB significantly improved in dealing the swelling of hand in both treatment groups at 2 weeks and 4 weeks, following SGB compared to pretreatment (p-value<0.05) (Table 3). Between each group, differences of hand volume were 61.2±36.4, 67.4±27.7 at 2 weeks and 86.7±33.9, 95.7±25.8 at 4 weeks. There were no significant differences between the two groups.

Additionally, score of CRPS clinical severity between the two groups didn't significantly decrease (p-value>0.05) (Table 3).

The number of studies of SGB in the literature is limited. Cepeda published review in 2002, they concluded that there are questions as to the efficacy of local anesthetic sympathetic blockade in treating CRPS because its efficacy is based mainly on case series.9 Many imaging devices have been studied for enhancing the accuracy of SGB to prevent serious side effects. Among these, computerized tomography and MRI are considered to be time-consuming and expensive, and above of all, are impractical for most practitioners.17,18 Fluoroscopy has gained popularity for its convenience and the relatively familiar real-time imaging, but fluoroscopy still cannot prevent improper injections into other important anatomical structures, such as the esophagus, which may be accidentally punctured, leading to infection of the neck.19 Ultrasonography imaging can easily identify soft tissue around the sympathetic ganglion and this study showed that only two patients in blind group experienced short-term complications on the injected site.

This study showed improvement of hand swelling in both groups but no advantages in reduction of hand swelling comparing with blind SGB. It may be that the etiology of post-stroke hand swelling is obscure. Vasomotor dysfunction, part of the unilateral disturbance of the autonomic nervous system that may follow stroke was mentioned as a causative factor.20 However, other possible explanations for post-stroke hand swelling are impaired lymphatic or venous drainage due to loss of muscle tone and of muscle pump activity. The possibility of increased capillary permeability in paretic limbs as the cause of post-stroke hand swelling was also raised.21 It follows that the role of sympathetic vasomotor dysfunction in post-stroke hand swelling formation is unclear. Therefore, reduction of hand swelling might be affected by various factors as well as SGB in our study.