Spinal cord stimulation in multiple sclerosis:

Transcription

1 N E ME,,EDIC.AL W r LIBfl,,ARY ; { t,_'t- f@>l 3RY 321 E. 1 5 th < a j4iouyai odttgy, Neurosulrgery, and Psyhiatry, 198, 43, 1-14 Spinal ord stimulation in multiple slerosis: linial results L S ILLIS. E M SEDGWICK, AND R C TALLIS From the Wesse Neurologial Centre, Southiampton General Hospital, Southampton MAY S U M M A R Y Clinial results of spinal ord stimulation by means of epidural eletrodes are reported in 19 patients with multiple slerosis. On temporary stimulation with perutaneous eletrodes, signifiant improvement in mobility ourred in 27.7% of 18 patients and the same number showed improved sensory funtion. Only one of 13 patients with severe upper limb ataia improved. The major response, both in terms of the perentage of patients responding and the etent of the responses seen was in bladder funtion: 75% of 16 patients with bladder symptoms improved and seven of the 11 patients with severe bladder disturbane (Kurtke grade 3 or more) improved. Four of these seven patients had before and after ystometry and 3 showed redued detrusor hyperrefleia. Altogether, 1 patients had a worthwhile linial response in one or more aspets of the disease and of these, nine have so far gone on to permanent stimulation. Medium-term results (up to two years) show that, with one eeption, improvement in bladder funtion has been maintained as long as stimulation has been ontinued and at least 5% of improvement in mobility has been maintained. A favourable response depends not upon the fat of stimulation but upon the type of stimulation reeived. This, along with other evidene, indiates that the response is not aused either by a plaebo effet or by the natural flutuation of the disease. The use of spinal ord stimulation (SCS) in linial pratie developed from physiologial advanes. Melak and Wall's (1965) theory of the gate ontrol of pain provided a rational basis for a trial of spinal ord stimulation to alleviate intratable pain. Suessful therapeuti results were first reported by Shealy et al. (1967). The appliation of this proedure to multiple slerosis was the result of observations made by Cook and Weinstein (1973) when treating a multiple slerosis patient for distressing and intratable bakahe. Spinal ord stimulation relieved the pain but, in addition, there was marked improvement in voluntary motor ativity. This report passed unnotied and ignored in this ountry until one of us (LSI) visited New York and observed the striking effets of SCS. Beause of our interest in the effets of partial lesions on the entral nervous system and the effet of repetitive stimulation (Illis, 1969, 1973), we repeated Cook's work, on- Addressforreprint requests: Dr LS Illis, Wesse Neurologial Centre, Southampton General Hospital, Southampton S9 4XY. Aepted 23 April 1979 firmed his results, and presented the first report of objetive neurophysiologial hanges (Illis et al., 1976). Abbate et al. (1977) reported linial and urodynami improvement in bladder dysfuntion in multiple slerosis patients, and there have been reports of improvement with SCS in other diseases, inluding spinal ord injury (Rihardson and MLone, 1978; Campos et al., 1978), erebral palsy (Walt and Pani, 1978), and spasmodi tortiollis (Gildenberg, 1977). At a reent international workshop devoted to this topi (Sith International Symposium on Eternal Cntrol of Human Etremities) si entres reported results. All demonstrated improvement in patients with SCS and all entres but one onfirmed the benefiial effet of SCS. In this paper we report our eperiene of 32 studies of SCS in 19 patients with multiple slerosis over a period of two and a half years. Results obtained in patients with other neurologial diseases will be reported elsewhere. Neurophysiologial hanges assoiated with SCS will be presented separately (Sedgwik et al., 198). 1 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

2 2 ~2 *o I. ; -o It..E 4 P., q. M -D EH t! L. 1- r~.= )2 2'...4) ).)... 2, '.., en 2 ) ) ~~~C: ~ ~~E. ~~~ ~~~~2 ~ e2.- 2 ~~~ Zo ~~~~ 22 o o O ~. O" U o S4 2r2:2. E a' C a. u l E to._ 2 22 o 22 O v. n >2 a L S Illis, E M Sedgwik, and R C Tallis i -.2. C I= ~ ~ ~ Cl -, ) (A) ~~ ~~~~~ 2. i j I ~ 2. ). E ] ]s- $... Zo Z.2 >. 34, on be 6 es 2 LLi n 2 to._ ) L. J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

3 Spinal ord stimulation in multiple slerosis: linial results >~~~~~~~~ ~~~~~~j ~~~~~ ~~~~~~~~~~~~~~~~~- 6 *-~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4)>~~~~~~~~~~~~~~~~ 4.~~~~~~~~~~ ~ ~ ~ ~ ~ ~ E Q~~~~~~E~~~~~~ 4 4) ~~~~~~~~~~ ~~~~~~~~~~;-. ~ o~~~~~~~~~~~~~- 4-44~~~~~~~~..&~~~~~~~~- ~~~~* ~~~~~~~4 ~~~~ ~~~ 4 >-.5E>4U) C :3. C6, r. 1. _~~~~~~. ~ ~~~~~~ O4-2-Yt. ) 4~ 64 ) ) ) 4. ed. 4 K )E-4e4E ). 4.US **to* ~=4M4 C1 7 )~~~~ 2 O 4C '> >ed >.%. N. ~ ~ o > < o E O. Oo44 44) 41) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C. )U.OU- 4)~~~~~~~~~~~~~.U Id 4) 4) a 44~~. >4 >4 >4 E 41 4~~~~~~~Et: ~~ ~ ~ ~ ~ ~ ~ i.c >4~~~~~. 4-..U- - _ ~~~~~ o 4 ~~~~~~~~~~~~~~~~ >U. ht 4- C,~. 4 3 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

4 4.9 g... o2 l, I,e u 11 >4)4 o :p teo I. A lo rd.2s '-~~~~~~~~~ ~~~~.2 ~,2 o*-e.44 X O 4 >4 A._ %n o r- L S Illis, E M Sedgwik, and R C Tallis u > > o) C C 4) C Vf ~D > X.2* O > v >, 2 _ ;,, * - 4 4) Q 4 _4._ 4)4)._ C ) _4 N D Q4 O S. * > 4) 44.2 ;.2 - o o )Z R D ṭ C = C._ U WA W) i- J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

5 Spinal ord stimulation in multiple slerosis: linial results 4 * o 4._ ) COa ow o to > 4 D Co~~~~ o a ~~2u~ ~ 4 ~.~~~4 8. ~~ ~~24~~~~~o4 S a CR.C ~ E~.2 r- LO on mg 4 8 o ' t.o Z.2E (A 5 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from [J. LA j on 19 June 218 by guest.

6 6 Patients and methods All patients gave informed onsent to the proedure after its nature and possible haards had been eplained and disussed with them and their relatives. Clinial data are summarised in Table 1. All patients but one fulfilled the riteria of Shumaher et al. (1965), and in addition 17 patients had abnormal visual evoked potentials and 1 had abnormal brainstem evoked potentials. One patient (RT) did not fulfil all the riteria of Shumaher et al. She had a progressive spasti quadriparesis with bladder involvement, normal myelogram and CSF, abnormal ervial evoked potentials but normal visual evoked potentials. The most probable linial diagnosis is multiple slerosis. All patients had been observed for at least one year before stimulation, and had been linially stable or steadily deteriorating for at least si months at the time of treatment, with the eeption of one patient (GC, table 1). Patients with flutuating symptoms, doubtful diagnosis, interurrent infetion, or who were unable to understand the eperimental proedure were eluded. Otherwise no attempt was made to selet patients. This was a deliberate poliy as we ould not predit whih patients would respond. We inluded patients with both long and short durations of illness, with a signifiant degree of disability, and with a variety of manifestations of multiple slerosis. Seven patients were virtually onfined to a wheelhair. For pratial and ethial reasons it has not been possible to have a parallel group of mathed ontrol subjets, still less to have "blind" patients or assessing physiians. The patients, however, served as their own ontrols, all but one were known to us for at least one year before stimulation, and Table 2 Funtional groups aording to the Kurtke disability sale Pyramidalfuntions Normal 1 Abnormal signs without disability 2 Minimal disability 3 Mild or moderate paraparesis or hemiparesis; severe monoparesis 4 Marked paraparesis or hemiparesis; moderate quadiparesis; or monoplegia 5 Paraplegia, hemiplegia, or marked quadriparesis 6 Quadriplegia V Unknown Sensory funtions Normal I Vibration or figure-writing derease only 2 Mild derease in touh or pain; moderate derease in position, vibration or disrimination 3 Marked hyposensitivity (not omplete) 4 Analgesia or anaesthesiato groin; hemianaesthesia or hemianalgesia 5 Analgesia and anaesthesia to nek V Unknown L S Illis, E M Sedgwik, and R C Tallis seven had been followed personally sine the onset of their multiple slerosis (two to five years). Many patients have had at least two periods of SCS separated by three to nine months, and their linial states before and after stimulation were ompared on eah oasion. In addition, linial observations were orrelated with measurement of objetive physiologial parameters. METHODS OF ASSESSMENT All patients were graded aording to the Kurtke disability sale (Kurtke, 1961, table 2). Repeated linial and neurophysiologial (Sedgwik et al., 1979) assessment as an inpatient was arried out for at least one week before stimulation. Urodynami studies were performed before and during stimulation in nine patients. Residual urine was measured, ystometry was performed with retal subtration (infusing at 5ml per minute), and a reord was obtained of the urethral pressure profile in aordane with the method of Brown and Wikham (1969). The apparatus used was a Devies reorder with an Elomati transduer EM75. Cerebrospinal fluid analysis was made before and towards the end of stimulation using standard laboratory tests in si patients. Full blood ount, sedimentation rate, blood urea and eletrolytes, mirosopy and ulture of urine, radiographs of hest and spine were arried out routinely. Two patients with severe urinary problems had intravenous pyelography, miturating ystogram, and ystosopy (Mr J Jenkins) performed to assess any possible strutural disturbane. Intake and output harts were kept on three patients before stimulation and these patients were kept to the same fluid input during stimulation. Cerebellarfuntions Normal I Abnormal signs without disability 2 Mild ataia 3 Moderate trunal or limb ataia 4 Severe ataia all limbs 5 Unable to perform o-ordinated movements due to ataia V Unknown X is used after -3 when weakness of grade 3 or more interferes with testing. Bowel and bladderfuntions Normal 1 Mild hesitany, urgeny or retention 2 Moderate hesitany, urgeny, retention or rare urinary inontinene 3 Frequent inontinene 4 In need of almost onstant atheterisation but with intat bladder sensation; severe bowel retention and/or inontinene 5 Lak of sensation and ontrol of bowel and bladder funtion V Unknown J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

7 Spinal ord stimulation in multiple slerosis: linial results METHOD OF STIMULATION The proedure was arried out with normal sterile tehniques. With the patient prone, on an X-ray table, 1.5% yloaine was injeted subutaneously in the lower thorai area to produe loal anaesthesia. Under fluorosopi ontrol an epidural needle (we have used various sies) was introdued in the interspinous spae to reah the epidural spae in the midline. The stylet of the needle was removed, and it was asertained that there was no leakage of CSF. An eletrode was passed through the needle pointing rostrally and advaned under fluorosopi ontrol to mid or high thorai levels and positioned in the midline in the epidural spae. The needle was removed and a seond eletrode was introdued in the same way, so that the two were positioned in the midline about one to three vertebral bodies apart. In some patients a third (reording) eletrode was plaed in position. The eletrodes were fied to the skin with either sutures or Steri-strips and anteroposterior and lateral radiographs were obtained to doument their position. The eletrodes were then onneted to a reeiver, usually with the positive eletrode rostral. The loop antenna from the stimulator was plaed over the reeiver and a sterile dressing was applied to seure the eletrodes and the reeiver. Patients were told that they may or may not feel a sensation of stimulation. Some were told they may feel nothing at all but others realised from talking to previous patients that the stimulator sensation aimed for was a symmetrial tingling into both legs. The reason for hoosing this as optimum stimulation was based on our initial eperiene (Illis et al., 1976). Beause of the diffiulty in plaing eletrodes and beause of frequent eletrode slippage, optimum stimulation was not always ahieved even after several attempts. Stimulation was arried out at 33 H with 2 ys width eletri pulses at a voltage adjusted by the patient to give a pleasant warm tingling sensation. The urrent requirements of 11 patients were measured aurately. Patients were stimulated until 1 days of satisfatory stimulation had been given (this usually took two to three weeks), and the eletrodes were then removed. Patients were seen at least monthly for follow-up. Those who had a satisfatory response were offered permanent stimulation, and were readmitted three to si months after initial stimulation for a repeat of perutaneous stimulation. If a onsistent response was obtained, the eletrodes were implanted subutaneously and onneted to a subutaneous reeiver. Our three most reent patients have had a permanent implant immediately after a suessful trial of perutaneous stimulation without the three to si month gap. In our first three patients permanent eletrodes were implanted via laminetomy and stithed to the dura mater to prevent movement but we have subsequently dispensed with this and have not observed an inreased tendeny to slipping. The reeiver was usually plaed in the right anterior aillary line. Initially we used Davis and Gek platinum tipped eletrodes but these have now been withdrawn and we now use Avery or Medtroni eletrodes and other equipment (Avery Co, Farmingdale, USA: Medtroni, Shirley Lodge, Slough SL3 8QY). Results TYPE OF STIMULATION Three patients had surfae stimulation for three days with utaneous eletrodes in the mid dorsal areas and two patients had eletrodes inserted and onneted but with no batteries in the apparatus for two days. There was no hange in these patients. Nine patients had stimulation sensation in the hest, shoulder, or into one leg only. There was no linial hange in these patients with the eeption of one who had inreased spasms (reversible). Siteen patients had symmetrial sensation into both legs, and 11 showed linial response (table 3). Some of the patients who had Table 3 Clinial response with different types of stimulation on initial stimulation indiating that response is related to the type not the fat of stimulation Patient I CP 5 2DS 3 3SE 5 4 NE - 5EM 1 6GC - 7JM Number of Kurtke grades of improvement (most responsive funtion) Symmetrial stimulation Other types ofstimulation into legs (hest/shoulder/one leg) 8OB* 9SBK* 2 1 NP* _ 11 CF I 12 MR* 1 13 AM* 2 14 MW* 15 GM* 16 RT 3 17 HS* I 18 JL 3 19 SBT Improvement 11 No lian-e 5 Worse *Patient aware of projeted stimulation. -Stimulation not obtained. Inreased spasms (reversible) J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

8 8 shown no linial response with unilateral stimulation responded to symmetrial stimulation. There was no differene in response between patients who were aware of the projeted sensation and those who were not. One patient (CP) had a good response to stimulation and went onto permanent implant. After three months of ontinuous improvement he deteriorated to his pre-stimulation state linially and physiologially, and the stimulator sensation beame loalised to a unilateral distribution on the hest wall. Radiography indiated that one eletrode had slipped 5 mm from the midline. When this was replaed he again showed linial and physiologial improvement. CLINICAL RESPONSE ON INITIAL STIMULATION We have studied 19 patients reeiving temporary stimulation with perutaneous eletrodes. The aim was to give 1 days ontinuous symmetrial stimulation but beause of tehnial diffiulties, in partiular eletrode slippage, this was not always ahieved. Table 1 gives the response to initial (temporary) stimulation. Eletrodes were plaed between C6 and C7, and T9 and TIO vertebral levels, and there was no relationship between eletrode position and linial response within L S Illis, E M Sedgwik, and R C Tallis these limits. Levels of the spinal ord outside this region were not stimulated systematially. Figure 1 indiates the hanges seen on initial stimulation in terms of Kurtke grading. As an be seen from table 1 and fig 1, the major benefit is in bladder funtion, both in terms of the number of patients and the degree of improvement. Out of 18 patients with mobility problems, signifiant improvement ourred in five (27.7%). Thirteen patients had definite upper limb ataia (Kurtke grade 2 or more), and only one of these had improvement in this disability. Of 18 patients with sensory impairment, five (27.7%) showed improvement. Bladder symptoms were present in 16 patients and 12 (75%) showed improvement. Improvements suffiient to produe a hange in life style, inluding redued dependeny on others, were seen even in patients onfined or virtually onfined to a wheelhair-for eample, SBK and AM-and in patients who had been permanently atheterised and with no bladder sensation-for eample SE and JL. Further details of response of bladder dysfuntion are given in Table 4. Of 11 patients with severe bladder disturbane (Kurtke grade 3 or more) nine improved by at least one Kurtke grade and si by two or more grades. A seventh patient who improved initially by only one grade sub- J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from Fig 1 Response to initial stimulation. on 19 June 218 by guest.

9 Spinal ord stimulation in multiple slerosis: linial results Table 4 Response of bladder dysfuntion to spinal ord stimulation Patient Duration Nature ofbladder symptoms Cvstometry Response to stimulation of bladder symptoms I CP 15 mo Absent bladder sensation. Epressing bladder Detrusor Normal bladder sensation and ontrol restored. two hourly day and night in an attempt to arrefleia Improvement maintained for nine months and then prevent inontinene. deteriorated to pre-stimulation state. 2 DS 9 mo Impaired bladder sensation. Urgeny. Voiding - Normal bladder sensation and ontrol. No hourly in attempt to prevent inontinene but frequeny or urgeny. Improvement maintained up often inontinent. to present, ie 24 months from permanent stimulation. 3 SE 1 yr Catheterised one year for retention of urine. - Normal bladder sensation and ontrol. Catheter Attempts to remove atheter unsuessful. removed. Reently (24 months after permanent Absent bladder sensation. stimulation) has developed mild frequeny and urgeny and oasional inontinene. 6 GC 5 yr Mild urgeny and frequeny with very oasional inontinene at time of stimulation. - No hange. (Symmetrial stimulation never ahieved-see Table 3). 7 JM Unertain Mild, rather variable urgeny and frequeny without inontinene. Detrusor hyperrefleia Urgeny less obvious. No ystometri hange. 8 OB 8 mo Moderate urgeny with rare urge inontinene. - Urgeny less obvious. 9 SBK 3 yr Severe urgeny and frequeny and frequent Detrusor Normal frequeny and urgeny. No longer urge inontinene. Spending night and day on hyperrefleia inontinent. (See tet) Dispensed with bed pan. a small bed pan. Cystometri improvement. 1 NP 9 yr Absent bladder sensation. Catheterised for two years. Unsuessful bladder nek resetion. - No hange. (Symmetrial stimulation never ahieved.) 11 CF Unertain Mild symptoms: urgeny, frequeny, hesitany. Normal Slight improvement in hesitany and urgeny. 12 MR 4 yr Impaired bladder sensation for one year. Catheterised three months for urge inontinene. Detrusor hyperrefleia Restored bladder sensation. Severe urgeny and frequeny, atheter ould not be removed. 13 AM 4 yr Severe frequeny, urgeny. Frequent urge Detrusor Progressive improvement over several months until inontinene. Living with urine bottle to hand. hyperrefleia now (si months after starting permanent stimula- Oasionally unaware of voiding. tion) bladder funtion normal. Cystometri improvement. 15 GM 5 yr Urgeny and frequeny of variable severity. Wearing urosheath. D-trusor hyperrefleia No hange. 16 RT 1 yr Frequeny, urgeny, urge inontinene, - Normal bladder sensation and ontrol. No urinary hesitany. Frequent urinary infetion. infetion. 17 HS 5 yr Moderate frequeny and urgeny. Oasional urge inontinene. Noturia 2-3. Normal Urgeny greatly reduad. No urge inontinene. No noturia. 18 JL 5 yr Catheterised for four years for frequeny and Detrusor Normal bladder sensation. Cath_ter removed. urge inontinene. No bladder sensation for hyperrefleia three years. 19 SBT 1 yr Severe frequeny and urge inontinene. Wet Detrusor Urgeny redued and dry all the tine. No noturia nearly all the time. Oasional hesitany. hyperrefleia but still has frequany by day. Cystometri Severe noturia. improvement. sequently showed further improvement. Of the nine patients who had before and after urodynami studies, seven had detrusor hyperrefieia and two had normal findings. Of the seven patients with detrusor hyperrefleia, four showed linial A noo- E 5- t- o no U Fig 2 1 Desre to miturte IInontinent 5- improvement and in three this was very striking. These showed a ystometri improvement with dereased tendeny to premature ontrations and onsequently inreased bladder apaity (fig 2). Before stimulation this patient (AM) had a first 1"Desire to miturate I~~~~~~~ ] ~~~~~~~I nontinent 1 2 3b Before s*nulation volume (ml) During stinulation vdune(ml) Detrusor or hyperrefleia: improvement of ystometrogram during stimulation. 9 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

10 1 sensation of the desire to void, assoiated with a sharp rise in bladder pressure and rapidly followed by voiding, after about 12 ml had been instilled. During stimulation, the desire to void was not felt until over 2 ml had been instilled. The ontrations whih then followed were of low amplitude, and voiding ould be voluntarily ontrolled until over 4 ml had been instilled. No hanges were noted in the urethral pressure profile. As shown in fig 3, the response of bladder symptoms to SCS seems to orrelate to some etent with their duration. In the ase of the 11 patients with severe symptoms, there seems to be less hane of a good response with symptoms greater than five years duration. The severity of bladder symptoms does not appear to be a bar to improvement. Overall, 1 of 19 patients (52.6%) showed a worthwhile linial response in terms of alteration of life and of dependeny on others. All 1 were inluded among the 16 patients who reeived symmetrial stimulation; so of the patients reeiving symmetrial stimulation 62.5% had a worthwhile linial response. Anything less than this was onsidered insuffiient to justify offering permanent stimulation through implanted eletrodes. Figure 4 shows the Kurtke grade for bladder, motor, and sensory funtion of the nine patients who at the time of writing have gone on to reeive permanent stimulation through implanted eletrodes. Note the stability during the preeding nine months. LA (u > 1 a8 E7 _ 6 4' 3.Q 2 a1 1 Fig 3 Influene of duration of bladder symptoms on response to SCS in 11 patients with severe symptoms (Kurtke grade 3 or worse). *a- - 1 QJ e N, L S Illis, E M Sedgwik, and R C Tallis 12 4/ Sensory Motor 12 Temp. Months pre-stimulation stim Fig 4 Response to initial stimulation in nine patients going on to permanent stimulation. CURRENT REQUIREMENTS In 15 patients the urrent required to produe a stimulator sensation was determined. Of the 13 patients reeiving bilateral symmetrial tingling sensation the urrent requirement was ma (range ma), while si, in whih other sensations were pereived, hose a urrent of ft 1. ma (range 8-36 ma). Four patients appear in both groups as they had "bad" and "good" sensations at different times. The differene between the groups was not statistially signifiant as judged by the Wiloon rank sum test. RESPONSE TO PERMANENT STIMULATION After initial stimulation the eletrodes were withdrawn. Those patients who had a good response to initial stimulation were offered permanent stimulation. Beause of the problems of eletrode slippage, the first three patients had permanent stimulation arried out by suturing the eletrodes to the dura mater without atually punturing the latter. There were no problems or morbidity assoiated with this but the proedure inluded laminetomy and we now, therefore, arry out the implantation with the eletrodes free in the epidural spae. All implant operations were arried out by Mr John Garfield. The response to initial stimulation is a good indiation of the response with permanent stimulation (usually arried out three to si months later). Figure 5 shows the hanges in mean Kurtke grades of patients who had permanent stimulation, ompared with the mean Kurtke grade si months before stimulation. It indiates (a) the pre-stimulation period of stability; (b) the J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

11 Spinal ord stimulation in multiple slerosis: linial results Injtil Permarent stimubtion- stimulaton a ' 3) EC) Bladder Motor +4 - lisensory Fig 5 Change in mean Kurtke grade of patients with permanent stimulation ompared with pre-stimulation mean. response to initial stimulation; () the return towards pre-stimulation levels after essation of temporary stimulation; (d) the onsistent response to a seond period of stimulation; and (e) the long-term (up to 24 months) effet with permanent stimulation. Of the patients with permanent implants, with one eeption, improvement in bladder funtion has been maintained for as long as stimulation has ontinued, and improvement in mobility has either been maintained or dereased to about 5% of initial improvement. The one eeption is a patient who had a dramati improvement with spinal ord stimulation but after several months deteriorated to the pre-stimulated state linially and physiologially at the same time as the stimulator sensation hanged. Radiology showed that one eletrode had moved 5 mm from the midline. When this was replaed, stimulator sensation and linial and physiologial hanges reverted to the improved state, and this was maintained for a further two months. He subsequently deteriorated progressively and is now more or less in the prestimulation state eept that mobility is worse. We think that the present deterioration is the result of an eaerbation of multiple slerosis. TIME RELATIONSHIP OF SCS TO CLINICAL RESPONSE Figure 6 indiates the time relationships of SCS for initial and permanent stimulation in terms of response (first response and peak) and deay to pre-stimulation level. There appears to be no relationship between the timing of initial and peak B response, or between initial response and deay time. The onset of the response to SCS ours from between si hours and three days though the peak response may not be reahed until eight weeks after the start of SCS. The deay to prestimulation level after stopping SCS begins from one day to eight weeks, and patients reah their pre-stimulation level from one day to si months. In one patient bladder symptoms never returned to the pre-stimulation level. CSF STUDIES Cerebrospinal fluid was obtained from si patients before and about 1 days after the onset of initial stimulation. One patient showed an inrease in lymphoytes from to 12/mm3 but no other patient showed a ellular reation. In no ase was there a signifiant alteration of total protein or perentage of IgG. Disussion From the work desribed here and from the work of others listed in the introdution and summarised in the reent international workshop on eternal ontrol of human etremities, there is no doubt that spinal ord stimulation, as first desribed by Cook, produes worthwhile improvement in patients with hroni multiple slerosis. The improvement surpasses that produed by any other urrent method, but long-term effets are still being studied. It is also lear that bladder dysfuntion is the 11 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

12 12 ii 12 1i tn i i an 1E - C' F) n 3 ' S 4 : * I >Ai4 12 *1 *1 I i i * * * E L S Illis, E M Sedgwik, and R C Tallis * (12 a I. *I * I" Initial stimul ation Permanent stimulation Fig 6 Time relations of response to spinal ord stimulation (from start or end of symmetrial stimulation). manifestation of multiple slerosis whih responds best to spinal ord stimulation. This is espeially relevant beause, as Miller et al. (1965) have pointed out, bladder disturbane is the most important single fator whih determines a patient's admission to hospital. Our urodynami studies are inomplete but in all patients with severe symptoms who had a signifiant linial improvement and who had before and after urodynami studies there was a ystometri improvement. In five patients, with moderate or mild symptoms and mild symptomati improvement, there was no urodynami improvement. Two of these patients had normal ystometry before and after stimulation. The improvements seen in these five ases may have been due to better o-ordination between detrusor ontration and sphinter relaation. With the urodynami methods we have available we would not be able to demonstrate hanges in this. The disrepany between symptomati and ystometri improvement was also found by Abbate and others (1977) with spinal ord stimulation in multiple slerosis and has been noted in other onditions, as, for instane, in the treatment of bladder symptoms aused by lumbar spondylosis by lumbar laminetomy (Sharr et al., 1976). It has been suggested that these results ould be the result of a plaebo response and that to demonstrate otherwise would require a doubleblind trial. A double-blind trial was not pratial as the patients always knew when they were being stimulated, and there are ethial objetions to inserting and leaving eletrodes in patients without stimulating them. It is unlikely that patients would give their informed onsent to suh a proedure. Even without a double-blind trial, however, there is evidene whih disounts a plaebo effet. There is no doubt from the results presented that neurologial improvement ours in assoiation with SCS. Improvement did not preede SCS as might be epeted if hospitalisation and motivation were the auses of the improvement. The patients had been neurologially stable for at least nine months before SCS so any spontaneous improvement would be unlikely to oinide with SCS in one patient, let alone in 1 out of 19 patients. It is even less likely that a further epsiode of J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

13 Spinal ord stimulation in multiple slerosis: linial results improvement would oinide with the seond period of stimulation unless the improvements and the stimulation were ausally related. A number of features emerge from the present study, all of whih disount the possibility that the response is due to plaebo effet. In this series 75% of patients showed an improvement in bladder funtion and this is onsiderably more than any reported plaebo effet, as well as being greater than any reported improvement with other therapy. The two patients who were "stimulated" without batteries and the three who had surfae stimulation did not improve even though they reeived the same treatment in other respets. When SCS was begun they all responded. Moreover, our results show that it was not the mere fat of stimulation whih produed improvement but a definite type of stimulation namely that whih produes bilateral sensation into the legs (table 3). Some patients had inappropriate radiular sensation and did not improve but when their eletrodes were adjusted to produe bilateral tingling in the legs, improvement began. One patient who responded well and had a permanent implant began to deteriorate after four months when his stimulation sensation altered. Radiology showed that one eletrode had moved 5 mm from the midline. After this eletrode was replaed he improved again. There are also neurophysiologial hanges assoiated with SCS whih annot be attributed to a plaebo effet (Sedgwik et al., 1978, 1979). The neurophysiologial responses with the possible eeption of the ontingent negative variation are not under voluntary ontrol, and it is diffiult to see how they ould be plaebo mediated. The ontingent negative variation is a ortial evoked response whih an be altered by the patient's motivation. Contingent negative variation responses in our patients were not altered by SCS (Sedgwik et al., 1979). All these observations make it impossible to eplain the patients' improvement on the basis of a plaebo effet. If, however, it was shown that physiotherapy or "motivation" or any other type of therapy ould produe omparable and onsistent results with related neurophysiologial hanges then we would have to reonsider this onlusion. The safety of methods used for SCS has been disussed fully (Jobling et al., 1978). The average power dissipation between eletrodes is less than 5 mw whih is too low to ause haard. Interferene effets and stimulator malfuntion are not a problem sine SCS is not, unlike ardia paing, a life support system. Eletrohemial reations at the stimulating eletrodes are a potential ause of danger through release of noious substanes, loal ph hange, and orrosion of eletrodes with resultant inrease in urrent density. We have found no evidene for this and have not observed orrosion of eletrodes removed from patients. Nevertheless, there is a need for easeless vigilane, partiularly when eperimenting with unusual parameters of stimulation. Apart from minor skin hanges at the site of entry through the skin, we have enountered no problems of infetion. One of our patients, implanted elsewhere, had a minor infetion, and to be on the safe side we removed the apparatus and replaed it some months later. Another patient (SBK) had erosion of one eletrode lead through the skin beause implantation had been too superfiial and this needed replaement. We have had no other ompliations diretly aused by SCS. There are now enough linial and physiologial data to make it lear that we are dealing with a real phenomenon. The mehanism of ation, however, remains uneplained. Spinal ord stimulation may at on different aspets of entral nervous system funtion-by altering the moleular environment and hanging ondution properties; by modifying funtional and anatomial reorganisation onsequent upon a lesion of the CNS; by altering the entral eitatory state and neurotransmitter release and altering afferent inflow. These fators at upon a nervous system whih has already reated to a partial lesion or several partial lesions. The end result of spinal ord stimulation may be to raise the level of ativity so that remaining inhibitory mehanisms an operate. The use of SCS in hroni neurologial disease always involves more than a single simple proedure and in our view it should be onfined, at present, to entres where omprehensive evaluation an be arried out, where there are the ombined resoures of a neurologist, neurophysiologist, and engineer, and where long-term, indeed permanent, follow-up is intended. There are still many unsolved problems suh as eletrode slippage, the mehanism of ation, and the long-term effets. Unless systemati studies are ontinued in entres whih have the neessary failities these problems will remain. This work was supported by the Medial Researh Counil. We aknowledge gratefully the assistane of Mr Garfield, Mr Jenkins, Dr Burrows and the staff of the Radiologial Department at the Wesse Neurologial Centre, and Mr D T Jobling. We 13 J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.

14 14 would also like to thank Miss Gillian Green for seretarial assistane. Referenes Abbate, A. D., Cook, A. W., and Attalah, N. (1977). The effet of eletrial stimulation of the thorai spinal ord on funtion of the bladder in multiple slerosis. Journal of Urology, 117, Brown, M., and Wikham, J. E. A. (1969). The urethral pressure profile. British Journal of Urology, 41, 211. Campos, R. J., Dimitrejevi, M. M., and Sharkey, P. C. (1978). Clinial evaluation of the effets of spinal ord stimulation on motor performane in patients with upper motor neurone lesions. Proeedings of the Sith International Symposium on the Eternal Control of Human Etremities, pp Yugoslav Committee for Eletronis and Automation: Belgrade. Cook, A. W., and Weinstein, S. P. (1973). Chroni dorsal olumn stimulation in multiple slerosis. Preliminary report. New York State Journal of Mediine, 73, Gildenberg, P. (1977). Treatment of spasmodi tortiollis with dorsal olumn stimulation. Ata Neurohirurgia (Wein), 24, Illis, L. S. (1969). Enlargement of spinal ord synapses after repetitive stimulation of a single posterior root. Nature, 223, Illis, L. S. (1973). Regeneration in the entral nervous system. Lanet, 1, Illis, L. S., Sedgwik, E. M., Oygar, A. E., and Sabbahi Awadalla, M. A. (1976). Dorsal olumn stimulation in the rehabilitation of patients with multiple slerosis. Lanet, 1, Jobling, D. T., Tallis, R. C., Illis, L. S., and Sedgwik, E. M. (1978). Eletroni aspets of spinal ord stimulation. In Proeedings of the Sith International Symposium on the Eternal Control of Human Etremities, pp Yugoslav Committee for Eletronis and Automation: Belgrade. Kurtke, J. F. (1961). On the evaluation of disability in multiple slerosis. Neurology (Minneapolis), 11, L S Illis, E M Sedgwik, and R C Tallis Melak, R., and Wall, P. D. (1965). Pain mehanisms: a new theory. Siene, 15, Miller, H., Simpson, C. A., and Yeates, W. K. (1965). Bladder dysfuntion in multiple slerosis. British Medial Journal, 1, Rihardson, R. R., and MLone, D. G. (1978). Perutaneous epidural neurostimulation for paraplegi spastiity. Surgial Neurology, 9, Shumaher, G. A., Beebe, G., Kibler, R. F., Kurland, L. T., Kurtke, J. F., MDowell, F., Nagler, B., Sibley, W. A., Tourtelotte, W. W., and Willmon, T. L. (1965). Problems of eperimental trials of therapy in multiple slerosis, report by the panel on the evaluation of eperimental trials of therapy in multiple slerosis. Annals of the New York Aademy of Sienes, 122, Sedgwik, E. M., Illis, L. S., Tallis, R. C., Thornton, A. R. D., Abraham, P., El-Negamy, E., Soar, J. S., and Taylor, F. M. (198). Evoked potentials and ontingent negative variation during spinal ord stimulation for multiple slerosis. Journal of Neurology, Neurosurgery, and Psyhiatry, 43, Sedgwik, E. M., Thornton, A. R. D., El-Negamy, E., Tallis, R. C., and Illis, L. S. (1978). Eletrophysiologial responses assoiated with spinal ord stimulation. In Proeedings of the Sith International Symposium on the Eternal Control of Human Etremities, pp Yugoslav Committee for Eletronis and Automation: Belgrade. Sharr, M. M., Garfield, J. S., and Jenkins, J. D. (1976). Lumbar spondylosis and neuropathi bladder: investigation of seventy-three patients with hroni urinary symptoms. British Medial Journal, 1, Shealy, C. N., Mortimer, J. T., and Reswik, J. B. (1967). Eletrial inhibition of pain by stimulation of the dorsal olumns: preliminary linial report. Anesthesia and Analgesia (Cleveland), 46, Walt, J. M., and Pani, K. C. (1978). Spinal ord stimulation in disorders of the motor system. In Proeedings of the Sith International Symposiunm on the Eternal Control of Human Etremities, pp Yugoslav Committee for Eletronis and Automation: Belgrade. J Neurol Neurosurg Psyhiatry: first published as /jnnp on 1 January 198. Downloaded from on 19 June 218 by guest.