Spinal cord stimulation has been successfully used for more than three decades since Shealy and colleagues, inspired by the gate theory , introduced it into clinical practice . However, although its detailed mechanism of action remains unclear, spinal cord stimulation has been employed in several longstanding painful conditions, including failed back surgery syndrome, angina pectoris, complex regional pain syndromes (types I and II), phantom pain and ischaemic limb pain [3-9]. One mechanism of peripheral neuropathic pain is peripheral nervous ischaemia (e.g. diabetic neuropathy).
Increase in sympathetic tone induced by ischaemic pain tends to be highest at the source of the pain . Changes in blood flow correlated to the pain relief effect of spinal cord stimulation are therefore most likely to occur within the same nervous segment. In human beings, the blood flow changes that are the most accessible to measurement within a nervous segment are in the skin. The purpose of this study was to test the hypothesis whether changes in cutaneous blood flow in the same nervous segment may serve as a marker for the effectiveness of stimulation-induced pain relief. Spinal cord stimulation electrodes are positioned in a way that the stimulation is felt in the location of the pain. At the same time, a stimulus-related sensation is perceived in the skin. It was this skin area that was chosen for blood flow measurements.
The effects of spinal cord stimulation on peripheral blood flow have been studied in different types of chronic pain conditions. The microcirculation in patients with peripheral vascular disease improves during spinal cord stimulation [11,12]. However, its limb-saving effect in patients with inoperable leg ischaemia remains controversial [13,14]. The mechanism of action of increased microcirculation is still not fully understood and remains an issue of debate among researchers and clinicians. Local factors, such as acidosis and accumulation of metabolic products, are likely to favour microvascular vasoconstriction, which may be removed by blockade of the sympathetic nerves. On the other hand, the dynamic equilibrium will swing towards vasodilatation through reduced sympathetic tone. An important point has been raised by a recent study in patients with implanted spinal cord stimulation for peripheral vascular disease. It showed that the level of spinal cord stimulation in the epidural space produces changes in capillary blood flow, with a tendency to decrease when the stimulator is placed above T10 but to increase when placed at the lower level, T12 . The present study has been designed to evaluate any changes in skin blood flow (SBF) caused by spinal cord stimulation in relation to mechanisms of neuropathic pain in patients without peripheral vascular impairment.
Approval from the Guy's and St Thomas' Hospitals' Ethics Committee was obtained before the study, and after explanation, the patients were asked to sign a written informed consent form. Twelve patients with implanted spinal cord stimulators for the management of neuropathic pain of at least 2 yr duration participated in the study. They were experiencing pain relief rated between good and very good since implantation of spinal cord stimulation, as documented in the medical notes (Table 1). The stimulators were implanted epidurally between T10 and T12. Seven patients were suffering from pain caused by failed back surgery syndrome, two patients from radiculopathy, one from arachnoiditis, one from spastic hemiplegia and one from phantom limb pain.
In order to evaluate microcirculatory blood flow in patients suffering from neuropathic pain, laser Doppler perfusion scanning was used as a direct method for selective measurement of changes in peripheral SBF . This technique can discriminate between the superficial and deeper layers, therefore making exclusive measurements of SBF.
Experiments were conducted at the Pain Research laboratory of Guy's Hospital. All patients were asked to switch their spinal cord stimulator off the evening before the day of the experiment. They were asked to come in the early morning. After achieving steady state for 30 min, a laser Doppler perfusion scan was performed over skin areas to which pain was projected except in patients with phantom limb pain where the measurement was taken from the stump. Skin blood flow measurements were taken at baseline with the spinal cord stimulation off. Patients were then asked to switch the stimulator on; measurements were repeated at the time the spinal cord stimulation was switched on and at 10, 30 and 120 min later. The laser Doppler perfusion scanning was performed over the area of the skin corresponding to the evoked stimulation. Room temperature was kept stable between 20 and 22°C. Skin blood flow was measured with an LDI laser Doppler imager (Moor Instruments, Axminster, UK) on an infrared wavelength of 780 nm. FIGURE
Using repeated-measures analysis of variance, the effect of time (0, 10, 30, 120 min) on subjects' mean SBF was investigated (i.e. the within-subject effect). When individuals are measured repeatedly over time, their measurements are typically autocorrelated; where necessary, any lack of independence between measurements needs to be adjusted for . Therefore, when required, the Greenhouse-Geisser correction factor  was applied to the degrees of freedom in the analysis of variance. The correction factor has the effect of reducing the magnitude of the degrees of freedom before statistical testing. The level of statistical significance was set at P < 0.05.
Skin blood flow, measured as relative units with the laser Doppler perfusion scanning system, did not show any significant change during the 2 h of the experiment in 11 of 12 patients studied. The pain in individual patients is shown in Table 1.
There was no significant difference between the four time points (0, 10, 30, 120 min) in subjects' mean skin blood flow (P = 0.190). Overall means (SD) were 82.3 (13.64) units at 0 min, 79.5 (14.02) units at 10 min, 79.3 (8.09) units at 30 min and 87.1 (11.60) units at 120 min after the spinal cord stimulation was switched on. One patient showed a significant change in SBF over time, with an increase from a baseline relative units value of 53.0 (SD ± 2.6) to 53.2 (±2.9), 54.5 (±2.8) and 107.6 (±2.6) at 10, 30 and 120 min, respectively.
One patient showed an increase in SBF. He was suffering from failed back surgery syndrome following a lumbar laminectomy. The pain was localized in the lower back and radiated down the right leg. The spinal cord stimulator was implanted 3 yr after spinal surgery and provided him with very good relief. He stopped most analgesic drugs and returned to work. However, his pain history did not differ substantially from those of the other patients entered in the study.
The success of spinal cord stimulation is presently assessed mostly by subjective measures such as the visual analogue scale, McGill or quality of life questionnaires. Physical performance indicators are used as slightly more objective measures. The present experiments have been performed to test the hypothesis whether stimulation-induced changes in SBF can be used as an additional, objective measure for the efficacy of this form of treatment.
Laser Doppler scanning - rather than single-point blood flow measurement - was used in order to avoid the risk of choosing a non-representative point on the skin. The area of projected pain and the stimulation was chosen for the reason outlined above. The exposure time of 2 h following activation of the stimulator was adopted because this is the maximum time interval during which a pain-relieving effect can be expected when using a transcutaneous electrical nerve stimulator.
The majority of our patients experienced neuropathic pain following primary afferent injury. In this situation, there is extensive sprouting of sympathetic nerve fibres into the dorsal root ganglion . These connections could be expected to have a bearing on cutaneous blood flow via the axon reflex pathway. We have examined the vascular response to spinal cord stimulation projected into the painful area of patients suffering from neuropathic pain caused by trauma to spinal roots or peripheral nerves. Our study showed that spinal cord stimulation did not influence skin perfusion in the area to which the stimulus-related sensation was projected in all but one patient suffering from neuropathic pain, up to 2 h after the spinal cord stimulation was switched on. However, all patients found spinal cord stimulation effective in controlling their chronic pain. It appears that changes in blood flow are not a major factor in the pathophysiology of our patients.
Other mechanisms of pain control through spinal cord stimulation include stimulation of descending inhibitory pathways either directly or through long loop reflexes including the nucleus raphe magnus system. Our results from patients with no ischaemic vascular disease are consistent with previous findings using infrared thermography, capillaroscopy techniques and laser Doppler flowmetry in patients suffering from neuropathic pain [20,21]. In contrast, Ghajar and Miles detected an increase in transcutaneous oxygen tension in the affected limb in four of five patients with ischaemic vascular disease .
We thank Dr Philip Sedgwich, Department of Public Health Sciences, St George's Hospital Medical School, London, UK, for his contribution to the statistical analysis.
1. Melzack R, Wall PD. Pain mechanisms: a new theory. Science
2. Shealy CN, Mortimer JT, Reswick JB. Electrical inhibition of pain by stimulation of the dorsal columns: preliminary clinical report. Anesth Analg
3. Broseta J, Barbera J, de Vera JA, et al.
Spinal cord stimulation in peripheral arterial disease. J Neurosurg
4. De la Porte, van de Kelft E. Spinal cord stimulation in failed back surgery syndrome. Pain
5. Krainick JU, Thoden U, Riechert T. Pain reduction in amputees by long term spinal cord stimulation. Long-term follow up study over 5 years. J Neurosurg
6. Kumar K, Toth C, Nath RK. Spinal cord stimulation is effective in reflex sympathetic dystrophy. Neurosurgery
7. Kemler MA, Barendse GA, van Kleef M, et al.
Spinal cord stimulation in patients with chronic reflex sympathetic dystrophy. N Engl J Med
8. Kumar K, Toth C, Nath RK, Laing P. Epidural spinal cord stimulation for treatment of chronic pain: some predictors of success. A 15-year experience. Surg Neurol
9. North RB, Kidd DH, Zahurak M, James CS, Long DM. Spinal cord stimulation for chronic, intractable pain: experience over two decades. Neurosurgery
10. McLachlan EM, Janig W, Devor M, Michaelis M. Peripheral nerve injury triggers noradrenergic sprouting within dorsal root ganglia. Nature
11. Jacobs MJ, Jorning PJG, Joshi SR, Kitslaar PJ, Slaaf DW, Reneman RS. Epidural spinal cord electrical stimulation improves microvascular blood flow in severe limb ischaemia. Ann Surg
12. Claeys LGY. Improvement of microcirculatory blood flow under epidural spinal cord stimulation in patients with nonreconstructible peripheral arterial occlusive disease. Artif Organs
13. Kumar K, Toth C, Nath RK, Verma AK, Burgess JJ. Improvement of limb circulation in peripheral vascular disease using epidural spinal cord stimulation: a prospective study. J Neurosurg
14. Klomp HM, Spincemaille GH, Steyerberg EW, Habbema JD, van Urk H. Spinal cord stimulation in critical limb ischaemia: a randomised trial, ESES Study Group. Lancet
15. Ghajar AW, Miles JB. The differential effect of the level of spinal cord stimulation on patients with advanced peripheral vascular disease in the lower limbs. Br J Neurosurg
16. Essex TJH, Byrne PO. A laser Doppler scanner for imaging blood flow in skin. J Biomed Engl
17. Everitt BS. Analysis of longitudinal data. Beyond MANOVA. Br J Psych
18. Geisser S, Greenhouse S. An extension of Box's results on the use of the F distribution in multivariate analysis. Ann Math Stat
19. McCance AJ, Thompson PA, Forfar JC. Increased cardiac sympathetic nervous activity in patients with unstable coronary heart disease. Eur Heart J
20. Devulder J, Duprez D, De Laat M, Rolly G. Epidural spinal cord stimulation does not improve microvascular blood flow in neuropathic pain. Angiology
21. Kemler MA, Barendse GA, van Kleef M, Egbrink MG. Pain relief in complex regional pain syndrome due to spinal cord stimulation does not depend on vasodilation. Anesthesiology