bmj - spinal inright cervical sympathectomy 5 77 m 11 claudication 3 months atheroma. stenosis...

Journal of Neurology, Neurosurgery, and Psychiatry 1983 ;46:478-484

Spinal cord stimulation in peripheral vascular diseaseRC TALLIS,* LS ILLIS, EM SEDGWICK, C HARDWIDGE, JS GARFIELD

From the Wessex Neurological Centre and the Southampton General Hospital, Southampton, UK

SUMMARY The results of ten patients with severe, intractable symptoms of arterial disease receiv-ing spinal cord stimulation are reported. Six out of ten patients showed clinical improvement.Three of five patients with severe rest pain obtained complete or very marked relief and one oftwo patients with moderate rest pain in the legs obtained complete relief. The mean claudicationdistance in the ten patients increased from 65 to 212 metres during epidural stimulation of thespinal cord. Exercise tolerance as measured on a bicycle ergometer increased by 61%. Thesechanges were associated with small increases in cutaneous and muscle blood flow. In thosepatients who responded clinically, the improvements seen were maintained as long as spinal cordstimulation was continued. There was no clinical response to transcutaneous (placebo) stimula-tion and four patients did not respond in any way to spinal cord stimulation. The improvementsseen are unlikely to be due to either the natural history of the disease or to a placebo effect. Theeffect is probably due to antidromic stimulation of the central processes of the first order sensoryneurons. It is suggested on the basis of animal studies that this effect may be mediated by releaseof prostaglandins as well as indirectly via pain relief.

Spinal cord stimulation was first introduced into clin-ical practice for the treatment of intractable pain.'Since Cook & Weinstein2 reported its therapeuticbenefit in multiple sclerosis, it has also been used inthe rehabilitation of patients with chronic neurologi-cal disease. The employment of epidural elecrrodes3makes spinal cord stimulation a relatively minorprocedure since the electrodes can be introduced viaan epidural needle. Most workers are agreed thatthe main benefits of spinal cord stimulation are inbladder function46 and it is these changes that havebeen most carefully documented.

Less attention has been paid to other influencesthat spinal cord stimulation may have upon the aut-onomic nervous system. The fact that patientsreceiving spinal cord stimulation usually describe asensation of warmth, especially in the lower limbs,suggested that spinal cord stimulation may alterblood flow in the skin. Patients with dependent dis-colouration due to stagnant hypoxia associated withpoor peripheral circulation in multiple sclerosisshow improvement in the colour of dependentlimbs. Friedman7 noted a thermographically detect-

*Present address and address for reprint requests: Dr RC Tallis,Senior Lecturer in Geriatric Medicine, Royal Liverpool Hospital,Liverpool, L69 3BX, UK

Received 25 November 1982Accepted 5 January 1983

able increase in blood flow in the region of analgesiaduring spinal cord stimulation for pain. Cook et a18described pain relief and healing of ulcers in ninepatients with vascular insufficiency in the limbswhich in many cases had failed to respond to con-ventional therapy. These changes could berepeatedly cycled by starting or stopping stimula-tion. However, few measurements have been madeof blood flow during spinal cord stimulation. Moredetailed studies have been done in the case of trans-cutaneous stimulation with conflicting results beingreported.9- 2

In view of these observations, and our ownexperience of a patient with multiple sclerosis andarteriographically demonstrated peripheral vascularinsufficiency who reported decreased claudicationpain during the period of spinal cord stimulation,despite increased mobility, we studied the effects ofspinal cord stimulation upon cutaneous blood flowin the lower limbs of patients receiving this treat-ment for chronic neurological disease.We were able to demonstrate an increase in

cutaneous blood flow in these patients'3 and there-fore felt justified in embarking upon a trial of spinalcord stimulation in peripheral vascular insufficiency.The purpose of this paper is to present preliminaryobservations on 10 such patients who receivedepidural spinal cord stimulation. In seven of thesepatients this was preceded by a period of trans-

478

Protected by copyright.

on February 15, 2021 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.46.6.478 on 1 June 1983. Dow

nloaded from

http://jnnp.bmj.com/

Spinal cord stimulation in peripheral vascular disease

cutaneous stimulation which we believe acts only as

a placebo in this context. In an eighth patient trans-cutaneous stimulation immediately followed spinalcord stimulation. Two patients received spinal cordstimulation alone.

Methods

PATIENTS:All patients were referred from the Department of Vascu-lar Surgery at the Royal South Hampshire Hospital wherethey had been under the care of either Mr J Webster, or

Mr AD Chant. In all cases the diagnosis was clear andarteriographically proven, symptoms were severe and dis-abling and had not responded to conventional surgical andmedical therapy. Clinical details of the patients studied are

given in Table 1. All patients had been instructed to give

up smoking at the onset of their illness and four had com-

plied with this absolutely. Six had been unable to give upsmoking entirely but their smoking pattern had been stablefor many months prior to the period of stimulation andremained constant throughout this. In particular, the tim-ing of cigarettes in relation to investigations was constantthroughout the period of study.

METHODS OF ASSESSMENTClaudication distance: Claudication distance was measuredseveral times prior to and during each period of stimulationat fixed times of day over a given course at a fixed walkingspeed. In cases 9 and 10 claudication distance was assessedby means of a treadmill set at 2 km per hour. Severe restpain prevented complete assessment of case 7.Exercise tolerance: Exercise tolerance was measured on a

bicycle ergometer, the patient being required to maintain100 watts. The time to the point at which this was no

Table 1 Patients with peripheral vascular disease receiving spinal cord stimulation

Case Age Sex Duraton of Symptoms at time of Period for which Result of last Previous treatmentIschaemic spinal cord symptoms had been angiogramsymptoms (yr) stimulation stable prior to

stimulation

1 67 M 2 Rest pain 5 months Diffuse atheroma Lumbar sympathectomyClaudication Ulcer with stenosis in Diabetic controlled on

femoral tree insulin2 69 M 2 Claudication 1 year Diffuse atheroma Diffuse disease: surgery

not appropriate3 69 M 2 Rest pain 6 months Superficial femoral Left femoro-popliteal

Claudication arteries blocked bypassRight femoro-popliteal

bypass4 34 F 10 Claudication 9 months Severe generalised Left lumbar

Rest pain Buerger's disease sympathectomyPain right hand Right lumbar

sympathectomyRight Gritti-Stokes

amputationRight cervicalsympathectomy

5 77 M 11 Claudication 3 months Atheroma. Stenosis Bilateral lumbarIntermittent rest pain le¶t common iliac sympathectomiesSevere nocturnal

rest pain6 71 M 10 Claudication 18 months Occluded left femoro- Right femoro-

popliteal by-pass popliteal by-passAtheromatous left Left femoro-superficial femoral popliteal by-passwith block. Occludedpopliteal.

7 62 M 3Y2 Ulcer 3 months Atheroma. Occlusion Left femoro-Rest pain of superficial popliteal by-passClaudication femorals.

8 60 M 11 Claudication 5 months Occlusion left iliac Aorto-iliacNocturnal rest and superficial endarterectomy

pain femorals. Right femoro-popliteal by-pass

External iliac toexternal iliac cross-over graft

9 61 M 13 Claudication 3-4 years Narrowing of aorta Right sapheno-and iliac vessels. femoralBlocked right endarterectomysuperficial femoral Right femoro-poplitealand popliteal artery. by-pass

10 57 M 7 Claudication 4 months Iliacs blocked. No Aortic onlay graft.flow in left Thrombectomy frompopliteal. aortic graft

479



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Tallis, Illis, Sedgwick, Hardwidge, Garfield

longer possible was recorded. A mean of three measure-

ments was obtained on at least two occasions prior tostimulation and during each period of stimulation. Theseverity of rest pain prevented this from being measured incases 1 and 7.Cutaneous blood flow: Cutaneous blood flow after a halfhour period of rest with the patient supine was measured ineight patients by clearance of injected xenon 133 dissolvedin saline." Sites just below the head of the fibula and on

the dorsum of the feet were used. In two patients (cases 7and 10) rest pain made it impossible to make measure-ments in both legs so the leg more affected by symptomswas examined. Case 4 had had a mid-thigh amputation ofthe right leg.Muscle blood low: This was measured by xenon clearance,the blood flow being recorded in the calf muscles at restand during exercise with a specially designed standard calfexerciser. Only three out of the 10 patients were able tomaintain exercise for a sufficiently long period for a mean-ingful clearance curve to be obtained (cases 2, 3 and 4).Resting muscle blood flow was recorded in two furtherpatients (cases 6 and 10).

After selection for the trial the patients underwent fullassessment. This was repeated on admission to hospitaland the following day electrodes were put in place. Sevenof the patients had ten days of transcutaneous stimulationand the assessment was repeated. Electrodes were thenchanged over from transcutaneous stimulation to spinalcord stimulation and after a further ten days of stimulationthe assessment was again repeated. In one patient spinalcord stimulation immediately preceded transcutaneousstimulation and in two patients spinal cord stimulationalone was given because they were in a clinically criticalstate and a ten day delay before spinal cord stimulationcould not be ethically justified.

PROCEDURE:Epidural spinal cord stimulation: The method of insertingepidural electrodes through an epidural needle for spinalcord stimulation has been described.'"5 The electrodeswere positioned so that the patient experienced a pleasanttingling sensation extending symmetrically down into thelegs when the stimulator was switched on. The stimulatordelivers a continuous train of square wave pulses of 200 itsduration at approximately 8 mA'6 and at a frequency ofabout 33 Hz and was switched on continuously during theperiod of the trial.Transcutaneous Stimulation: In eight patients the responseto spinal cord stimulation was compared with that to a

comparable period of transcutaneous stimulation deliveredby means of surface slectrodes placed over the lumbarspine. It was intended that transcutaneous stimulationshould act as a placebo. No indication was given to thepatient as to which electrodes would be expected to pro-duce an effect. Nevertheless, it was not possible to carryout a completely blind trial because different stimulatorsensations are associated with surface electrodes (localtingling) and epidural electrodes (paraesthesiae andwarmth throughout the lower limbs).

Results

CLINICAL RESPONSE TO EPIDURAL SPINAL CORDSTIMULATION (table 2)Of five patients with severe rest pain occurring nightand day, three obtained complete or very markedrelief and two obtained no relief. One of twopatients with moderate rest pain in the leg obtainedcomplete relief but rest pain persisted in her hands.

Table 2 Clinical response to spinal cord stmulation (SCS)

Patient Rest pain* Claudication distance Other clinical changes Subsequent progress

Before SCS Before SCS Permanent FoUlow-up SubsequentcourseSCS SCS sdmulation (months)

1. +++ - 58 ± 5 320 + 80 Indolent ulcer healed. YES 26 Improvement maintained exceptImproved Improved colour of foot sporadic rest pain

2. - - 120 + 21 372 ± 92 Improved colour of foot YES 15 Further improvement. ClaudicationImproved at 900 metres

3. +++ + 16 ± 3 366 t 208 Improved colour of foot. YES 12 Further improvement, able to walkImproved Reported restoration of 3 miles

sweating in feet4. Leg + - 98 ± 22 606 ± 190 None YES 6 Further improvement

Hand ++ + + Improved5. +++ + 8 ± 2 24 ± 6 None NO 1 week Died of ruptured aortic aneurysm

Improved6. - - .50 + 11 83 6 None NOT 2 Improvement continues

Improved YET7. +++ +++ 13 ± 15 12 ± 8 Temporary improvement of NO 9 Left below knee amputation

No change colour of foot8. + + 133 + 35 142 ± 12 None NO 9 Claudication. Indolent ulcer

No change9. - - 100 ± 20 126 t 25 None NO 1 No change

No change10. +++ +++ 55 t 7 69 ± 8 None NO 1 Being considered for amputation

No change* Rest pain: (+) Sporadic; (++) Every night; (+++) Day and night; (-) No rest pain.

480



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



Of three patients with indolent ulcers, one healedwithin a fortnight of stimulation and one was notimproved. The third patient (case 5) who had hadremarkable relief from rest pain during the period ofspinal cord stimulation died of a ruptured aorticaneurysm one week after starting. The duration ofhis stimulation was too short to allow for healing ofulcers. In four out of seven patients with strikingdiscolouration of the feet, there was an improve-ment in colour though in one this was only tempor-ary. One patient reported restoration of sweating inthe feet. In four out of the ten patients there was avery striking increase in the claudication distance. Intwo others there was a moderate but significantincrease in claudication distance which, in one ofthese patients, was sufficient to enable him to goshopping with his wife, for the first time for years. Infour patients there was no significant change inclaudication distance. In one of these patients,although there was no reduction in claudication dis-tance, rest pain and dependent rubor became worse.The mean pre-stimulation claudication distance was65 + 45 metres and that during spinal cord stimula-tion was 212 + 147 metres. This difference was

significant (0.025 > p > 0-01).

CLINICAL RESPONSE TO TRANSCUTANEOUS

STIMULATIONSeven patients received transcutaneous stimulationbefore going on to epidural stimulation of the spinalcord. None of these patients showed any clinicalimprovement during a two week period of trans-cutaneous stimulation. Assessment of their claudica-tion distance and exercise tolerance on the bicycleergometer showed no change from the pre-stimulation state.

EXERCISE ENDURANCE ON THE BICYCLEERGOMETERExercise tolerance increased from 3929 1961joules to 6328 + 3421 joules (61% increase) butthis was not statistically significant (0-10 > p >0-05).

BLOOD FLOW STUDIES (tables 3 and 4)Skin blood flow was measured in eight out of the 10patients who responded clinically. Skin blood flowmeasurements were made at two sites on each leg.For a variety of reasons blood flow studies on theright legs are incomplete and only the results for theleft legs and feet are given. There was a significant

Table 3 Effect ofspinal cord stimulation (SCS) on blood flow in peripheral vascular diseasePatients respondingMean and standard deviation given

Test Units n Before SCS p *

Skin blood flow:Left foot ml/min/lOOG 4 5-45 6-90 0-02

+ 1-13 + 1-64Left leg ml/min/lOOG 4 3-33 3-23 NS

± 1-86 + 0-70Muscle blood flow:

Rest ml/min/lOOG 4 1-10 2-20 0-05+ 0-68 + 1-13

Exercise ml/min/lOOG 4 3-68 5-24 NS+ 1-45 + 2-93

*p calculated by paired t test

Table 4 Effect ofspinal cord stimulation (SCS) on blood flow in peripheral vascular diseaseNon-respondersMean and standard deviation given

Test Units n Before SCS p *

Skin blood flow:Left foot ml/min/lOOG 4 2-40 1-68 NS

+ 3-50 + 1-71Left Leg ml/min/lOOG 4 1-33 1-38 NS

+ 0-40 ± 1-31Muscle blood flow:

Rest ml/min/lOOG 1 1-2 1-2 NSExercise ml/min/lOOG - Not done

*p calculated by paired t test

481



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


482

increase in blood flow in the left foot of the clinicallyresponding patients but not at the site on the leftlegs just below the head of the fibula. The non-

responders showed no significant change in bloodflow.

Muscle blood flow at rest and during exercise was

measured in four of the responding patients but onlyincomplete studies in one of the non-responders are

available. Of the responding patients there was a

statistically significant doubling of muscle blood flowat rest. During exercise an increase in blood flowwas recorded but this did not reach statisticalsignificance. Only one non-responding patient was

studied and showed no change at rest.In summary, those patients who had a favourable

clinical response also showed an improvement infour of the six objective tests of exercise toleranceand blood flow, while of the patients who did notrespond clinically there were no changes in theassessment tests with the exception of the bicycleergometer test where there was a slight deteriora-tion in performance. We conclude that the clinicalimprovement seen in these patients cannot be attri-buted to a placebo response but it is probable thatsome beneficial physiological change occurred in theperipheral vascular system.

Discussion

The patients selected for spinal cord stimulation allhad severe symptoms and had either failed torespond to conventional surgical and medical treat-ment or were unsuitable for such treatment. Five outof ten patients had a sufficiently marked clinicalresponse to justify permanent stimulation. A sixthpatient had had a good response but died suddenlyof a ruptured aortic aneurysm in the epidural stimu-lation period. It was possible to demonstrate moder-ate increases in skin and/or muscle blood flow andthese were statistically significant in the skin of thefoot and in the case of resting muscle blood flow.Muscle blood flow during exercise showed an

increase which did not reach statistical significance.In all patients the improvements obtained initially

were maintained as long as permanent stimulationcould be continued. In two of the patients receivingpermanent stimulation, spinal cord stimulation hadto be discontinued for technical reasons and bothdeteriorated clinically. Repair of the spinal cordstimulation system so that a stimulator sensation was

once more obtained in the legs at once restored theimprovement that had been obtained. In one patientreceiving permanent stimulation, who haddeveloped gangrene in the hand, an area not beingstimulated at the time, insertion of electrodes toproduce a stimulator sensation in the hand did not


reverse the gangrene. This experience is consistentwith Cook's observation8 that once gangrene has setin, spinal cord stimulation has little to offer.

In interpreting the response to a new form oftreatment one must take into account the naturalhistory of the disease and the possibility of placeboeffect. It is well known that intermittent claudicationtends to improve with time17 18 but this usuallyoccurs over a period of months rather than days, andmoreover tends to occur in the early stages of theillness. It is usual for patients with severe persistentrest pain to remain stable or to get worse.

Transcutaneous stimulation might reasonablyhave been expected to produce similar symptomaticrelief if the patients were responding to a placeboeffect. Both spinal cord stimulation and trans-cutaneous stimulation techniques involved elec-trodes, wires and stimulator controls and both pro-duced a sensation. Care was taken to ensure that thepatients could not infer which method was supposedto work. Measurement of current in the epiduralspace during transcutaneous stimulation showed itto be less than with spinal cord stimulation by afactor of 2,000. It would therefore appear imposs-ible to stimulate the spinal cord by transcutaneousstimulation using the currents and voltages avail-able. No symptomatic or objectively measuredimprovement occurred with transcutaneous stimula-tion. The duration of the improvement seen and itsdependency upon continuing spinal cord stimulationis against a placebo effect. The improvements notedin blood flow and the other objectively measuredchanges make a placebo effect less likely.Our findings with transcutaneous stimulation are

contrary to those of Canosa et al'9 who reportedimprovement in patients with peripheral vasculardisease during transcutaneous stimulation over athree month period.

MECHANISMSSeveral mechanisms may be invoked to account forthe improvements seen:1. Pain relief. Pain relief, a well documented effectof spinal cord stimulation, may release the reflexvasoconstriction that is known to occur in responseto pain. This might break a vicious circle by whichpain leads to vasoconstriction and vasoconstrictionleads to further pain.2. Sympathetic nervous system: Very small changesin blood flow may be critical in determining whetherskin is viable and also in influencing claudicationdistance. The xenon clearance method of blood flowmeasurement has well known sources of error andvariation and must, at best, be considered a ratherinsensitive measure of detecting clinically significantalterations in circulation. However, a statistically



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



significant increase in muscle blood flow and in skinblood flow of the foot, but not of the lower leg, wasdetected in response to spinal cord stimulation. Ifthe clinical improvement is due to increased bloodflow, as well as to pain control, then it seemsunlikely that the sympathetic nervous system shouldbe involved. Whereas inhibition of alpha adrenergicvasconstrictor fibres and stimulation of cholinergicvasodilator fibres will produce respectivelyincreased blood flow in the skin and in the muscles,it is unlikely that spinal cord stimulation will pro-duce this balanced combination of effects.3. Prostaglandins and prostacyclins: Bayliss20demonstrated that stimulation of the dorsal rootsproduced hind limb vasodilatation in the cat andFoerster2' showed that dorsal root stimulation inman produced vasodilatation in the appropriatedermatome. Recently Hilton and Marshall22 havedemonstrated vasodilatation in the muscles follow-ing dorsal root stimulation and they believe that thiseffect is due to antidromic stimulation of smalldiameter fibres. Interestingly, it could not beblocked by sympathetic blocking agents but could beblocked by prostaglandin inhibitors from which theyinferred that is was mediated by the release of pros-taglandins in muscles. Whether these fibres areinvolved in the effects seen in ischaemic patientsreceiving spinal cord stimulation is not certain.There is no doubt, however, that spinal cord stimu-lation involves stimulation of the central processesof the first order sensory neurons and that there isno reason why these should not be stimulated anti-dromically as well as orthodromically.4. Local vascular control: There may be somedifficulty in understanding how neural modulationcould overcome the essentially structural problemthat occurs in obliterative arterial disease. There is,however, some evidence that physiologicalresponses to arterial insufficiency may not be opti-mal. As noted above, reflex vasoconstriction occursin response to pain and structural changes in thearterial wall may cause augmentation of vasocons-trictor responses to adrenergic stimuli.23 The intra-arterial infusion of alpha-adrenergic blockers suchas thymoxamine may produce marked improvementin the pulse plethysmograph in patients with severeperipheral vascular disease.24 Recently it has beenshown that intra-arterial prostacyclin infusion mayproduce dramatic increases in blood flow associatedwith marked clinical improvement.25 This effect hasbeen attributed to reduction in platelet coagulation,vasodilatation and in some cases the induction of theproliferation of small vessels. At the very least, itdoes indicate that even in obstructive arterial dis-ease there is a residual neuropharmacologicalreserve that may be exploited. It is possible that

483

neurally mediated release of prostacyclins resultingfrom antidromic stimulation of sensory neuronscould produce this effect.

It has been argued that a generalised vasodilata-tion would have an adverse effect on patients withsevere arterial disease, increased blood flow beingdirected towards unobstructed arteries so that bloodflow distal to the site of obstruction falls evenfurther. If, however, the ony effect of spinal cordstimulation is to reverse adverse physiologicalchanges that occur locally distal to the site of ablock, then of course vasodilatation would occuronly in those places where it was required and notelsewhere where it may lead to further diminution ofblood flow in ischaemic areas.

This work is still at a very early stage. The resultsthat we have obtained so far, both in terms of clini-cal response and objective measures of claudicationand exercise tolerance, as well as some of thechanges seen in blood flow, suggest that spinal cordstimulation may have a part to play in the manage-ment of advanced arterial disease where all otherforms of treatment have failed. Clearly this will rep-resent only a small group of patients and this groupwill be even smaller if sufferers from vascular dis-ease were able, to resist the urge to smoke.26Moreover, spinal cord stimulation will have noinfluence upon the underlying pathology and cannotbe expected to reverse gangrene. In view of theencouraging results a larger trial is underway inorder to clarify the indications for spinal cord stimu-lation and to investigate the mechanisms involved.

We are very grateful to Mr J Webster, FRCS andMr AD Chant, FRCS of the Department of Vascu-lar Surgery at the Royal South Hampshire Hospitalfor kindly referring patients. We are also grateful toMr CAC Clyne for useful advice at many stages ofthe present project. Finally, we thank Miss GillianGreen for typing the manuscript.

References

lShealy CN, Mortimer JT, Reswick, JB. Electrical inhibi-tion of pain by stimulation of the dorsal columns: pre-liminary clinical report. Anesth. Analg. (Cleve)1967;46:489-91.

2 Cook AW, Weinstein SP. Chronic dorsal column stimu-lation in multiple sclerosis: preliminary report. NTState J .Med 1973;73:2868-72.

3 Shimoji K, Kitamura H, Hezono E, Shimizu H,Okamoto K, Iwakura Y. Spinal hypalgesia andanalgesia by low frequency electrical stimulation inthe epidural space. Anesthesiology 1974;41:91-4.

4 Illis LS, Sedgwick EM, Tallis RC. Spinal cord stimula-tion in multiple sclerosis: clinical results. J NeurolNeurosurg Psychiatry 1980;43:1-14.



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



5Hawkes CH, Wyke MA, Desmond A, Bultitude MI,Kanegaonkar GS. Stimulation of the dorsal column inmultiple sclerosis. Br Med J 1980;1:889-91.

6 Read DJ, Matthews WB, Higson RH. The effect of spi-nal cord stimulation on function in patients with mul-tiple sclerosis. Brain 1980;103:803-33.

Friedman H, Nashold BS, Somjen G. Physiplogicaleffects of dorsal column stimulation. Adv Neurol1974;4:769.

8 Cook AW, Oygar A, Baggenstos P, Pacheto S, KlerigaE. Vascular disease of the extremities: electricalstimulation of the spinal cord and the posterior roots.NY State J Med 1976;76:366-8.

9 Owens S, Atkinson ER, Lees DE. Thermographic evi-dence of reduced sympathetic tone with transcutane-ous nerve stimulation. Anesthesiology 1979;50:620-5.

10 Ebershold MJ, Laws ER, Albers JW. Measurements ofautonomic function before, during and after TNS inpatients with chronic pain and control subjects. MayoClin Proc 1977;52:228-32.

" Abram SE. Increased sympathetic tone associated withtranscutaneous electrical stimulation. Anesthesiology1976;45:575-7.

12 Abram SE, Asiddao CB, Reynolds AC. Increased skintemperature during transcutaneous electrical stimula-tion. Anesth Analg (Cleve) 1980;59:22-5.

13 Tallis RC, Illis LS, Sedgwick EM, Hardwidge C, Ken-nedy K. The effect of spinal cord stimulation uponperipheral blood flow in patients with chronicneurological disease. Int Rehabil Med 1983; in press.

14 Sjersen P. Cutaneous blood flow in man studied byfreely diffusable radioactive indicators. Scand J ClinLab Invest 1966;Suppl 99:52-9.

,s Cook AW. Electrical stimulation in multiple sclerosis.Hosp Pract 1976;2:51-8.

16 Jobling DT, Tallis RC, Sedgwick EM, Illis LS. Elec-tronic aspects of spinal cord stimulation in multiplesclerosis. Med Biol Eng Comput 1980;18:48-56.

7 Hughson WG, Mann JI, Tibbs DJ, Woods HF, WaltonI. Intermittent claudication: factors determining out-come. Br Med J 1978;1:1377-81.

18 Jonason T, Jonzon B, Ringquist I, Oman-Rydberg A.Effect of physical training on different categories ofpatients with intermittent claudication. Acta MedScand 1979;206:253-8.

19 Canosa F, Dooley DM, Boyle E. Vascular reactions topre and post transcutaneous electrical stimulation ofthe spinal cord and nerve roots. In "Proceedings of theSixth International Symposium on External Control ofHuman Extremities". Yugoslav Committee for Elec-tronics and Automation, Belgrade. 1978;585-601.

20 Bayliss W. On the origin from the spinal cord of vas-odilator fibres of the hind limb. J Physiol (Lond)1901;26:173-209.

21 Foerster 0. The dermatomes in man. Brain1933;56: 1-38.

22 Hilton SM, Marshall JM. Dorsal root vasodilatation incat skeletal muscle. J Physiol (Lond) 1980;299:277-88.

23 Folkow B. The haemodynamic consequences of adap-tive structural changes of the resistance vessels inhypertension. Clin Sci 1971 ;41: 1-12.

24 Rose SS. The use and abuse of vasodilator drugs. VasDis 1971;4:(2);67-82.

25 Szczeklic A, Nizankowski R, Shawinski S, Szczeklik J,Gluszko P, Gryglewski RJ. Successful therapy ofadvanced arteriosclerosis obliterans with prostacyc-lins. Lancet 1979;1:1111-14.

26 Jamieson CW, Clyne CAC. The fate of the ischaemiclimb. In: Yu PN, Goodwyn JF, eds. Progress in Car-diology 1978;7:219-28.

484



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


bmj - spinal inright cervical sympathectomy 5 77 m 11 claudication 3 months atheroma. stenosis...

Documents