389
Neurophysiol Clin (1990) 20, 389-398 © Elsevier, Paris
Review article
Deep brain stimulation in the treatment of chronic pain in m a n : where and w h y ? J Gybels, R Kuper s Department o f Neurology and Neurosurgery, University o f Leuven, U Z Gasthuisberg, Herestraat 49 B-3000 Leuven, Belgium (Received 26 June 1990; accepted 2 August 1990)
Summary - in current clinical practice, two brain structures are stimulated for the relief of chronic pain, namely the somatosensory thal~mic nuclei (VPL-VPM) and the periventricfilar and p&iaqueductal gray matter (PVG-PAG). Whereas gtimulation of the V P L - V P M is almost exclusively used for the treatment of deafferentation pain, stimulation of the P V G - P A G is mostly used in cases of nociceptive pain. We present our results of V P L - V P M stimulation in 36 patients with deafferentation pain. Initial pain relief was obtained in 61% of patients. To-day, after a mean follow-up of more t h a n 4 years, 30% are still pain free. This success rate was found to be lower than the m e a n reported success r a t e ' o f 5"~%, based on a survey of the world literature. U p o n reviewing the literature, it was apparent that the reported success rates vary considerably between different authors. Some tentative explanations are given for this large discrepancy in success rate. The mechanisms by which electrical stimulation of the V P L - V P M suppresses deafferentation pain remain to be elucidated. Recent clinical and experimental findings suggest that a dopaminergic m e c h a n i s m might be involved. brain stimulation / pain
R~sum~ - Stimulation c~r~brale profonde dans les d6uleurs chroniques: off et pourquoi? En pratique clinique courante, deux structures c&~brales sont stimtll~es p o u r traiter la douleur chronique : les noyaux sensitifs du thalamus ( V P L - VPM) et la substance grise p&iventriculaire et pOriaqueducale (PVG-PA G). La stimulation du noyau V P L - V P M est appliquOe clans le traitement de la douleur de d~saff&entation, tandis que la stimulation du P V G - P A G est presque exclusivement utilis~e p o u r soulager la douleur nociceptire. N o u s pr~sentons nos r~sultats de la stimulation du VPL- V P M chez 36 malades, souffrant de douleur de d~saffdrentation. Un succbs initial a OtO obtenu chez 61% des malades. Actuellement, aprbs un suivi d'une dur~e moyenne de 4 ans, 30% de ces interventions sont consid~r6es c o m m e Otant un suceOs. Ce r~sultat est moins bon que celui relevd clans la litt&ature mondiale. Il f a u t ajouter qu "on note des variations importantes des taux de succbs entre les diff&ents auteurs. N o u s d o n n o n s une esquisse d'explication de cette variabifit~ clans les rdsultats. Les m~canismes par lesquels la stimulation 6lectrique des noyaux sensitifs du thalamus diminue la douleur restent ~ Oclaircir. Certaines donn~es exp&imentales et cliniques laissent penser qu 'un m&'anisme dopaminergique peut 6tre impliqud. stimulation c6r6brale / douleur
390
J Gybels, R Kupers
Introduction
Although the possibility of producing analgesi a by electrical stimulation of the human brain has already been reported in the late fifties (Heath and Mickle, 1960), the real interest in brain stimulation for the treatment of chronic pain in man only arose at the end the sixties. Important stimuli for this sudden interest were the proposal of the gate control theory by Melzack and Wall (1965) and Reynolds' discovery of stimulation-produced analgesia in the animal (1969). The discovery by Reynolds that electrical stimulation of the rat midbrain can produce profound analgesia without the concurrent administration of drugs opened new horizons for the treatment of chronic pain. It marked the beginning of a new era in neurosurgical pain treatment by offering an alternative for the previous destructive and irreversible procedures. The subsequent discovery, some years later, of the narrow relationship between stimulation-produced analgesia and the endogenous opioid system, offered a theoretical basis for the understanding of how stimulation of the midbrain may suppress pain. However, it was not until 1973 that the first clinical applications of periventricular gray stimulation in man appeared (Richardson and Akil, 1977). A second line of interest arose from the gate control theory. Although historically older than the discovery of stimulation-produced analgesia, the importance of the gate control theory as a conceptual framework for deep brain stimulation (DBS) was highlighted by Reynolds' discovery. It formed the first important incentive for the use of stimulation procedures in the treatment of chronic pain. According to the gate control theory, stimulation of large diameter fibers is capable of inhibiting nociceptive information. The first successful clinical applications of this theory were periph eral nerve stimulation (Wall and Sweet, 1967), and dorsal column stimulation (Shealy et al, 1967). Since in certain clinical syndromes, eg after brachial plexus avulsion, anesthesia dolorosa etc, there is a lack of primary afferent fibers in the peripheral nerve or dorsal column, it was suggested that in those cases, supraspinal stimulation at the level of the somatosensory system might offer an alternative substrate for activating the lemniscal system. Although the first publication on somatosensory thalamic stimulation in the treatment of deafferentation pain only appeared in 1973 (Hosobuchi et al), it had already been practiced since the early sixties by Mazars et a! (1979), ie even before the proposal of the gate control theory. Their theoretical framework was the theory of Head and Holmes (1911) according to which pain might be the consequence of an imbalance between protopathic and epicritic sensory functioning. Stimulation of the thalamic sensory relay nuclei would presumably increase the epicritic component and hence inhibit the noxious protopathic inflow. Although welcomed with greht enthusiasm by those neurosurgeons who had a profound interest in pain research and therapy, DBS was not always as successful as was initially hoped. The clinical results did not always fit with the experimental findings. This was in part due to the fact that in the animal, acute experimental pain was investigated, while in clinical practice, chronic pain had to be beaten (it was ap-
DBS for treatment of chronic pain
391
parently much easier to produce a prolonged tail-flick in the rat than to alleviate chronic pain in man). It is only recently that DBS has been studied in animal models of chronic pain (Kupers et al, 1988; Lombard, 1989).Moreover, large discrepancies were noted between the results of different neurosurgical groups, some reporting great successes, others almost complete failures. Although DBS for the treatment of chronic pain has not become a real routine intervention, it is now accepted that it can offer an alternative in those cases in which all other treatment modalities have failed. However, knowledge as to how electrical brain stimulation suppresses chronic pain is still fragmentary. In the first part of this paper we will summarize clinical data on DBS ; in the second part we will explore the possible mechanisms by which DBS alleviates pain due to lesions o f the nervous system.
Clinical results
In current clinical practice, ~two brain structures are stimulated for the relief of chronic pain, namely the primary' somatosensory relay nuclei of the thalamus and their afferent and efferent pathways (VPL-VPM), and the periventricular and periaqueductal gray matter ( P V G - P A G ) . Whereas stimulation of the somatosensory thalamus is mainly used for the treatment of deafferentation pain, P V G - P A G stimulation is mostly used in cases of nociceptive pain (Gybels and Sweet, 1989). However, some authors also report favorable results of V P L - V P M stimulation in nociceptive pain (Tsubokawa et al, 1984). Stimulation of both P V G - P A G and V P L - V P M targets is sometimes performed in those cases in which there is a mixture of somatic and neurogenic pain (eg in low back pain). In this paper we will discuss some of our results of V P L - V P M stimulation in patients with deafferentation pain. We will limit ourselves to the results o f those patients who were frequently seen for consultation and in which a thorough evaluation of the therapeutic result on long-term follow-up could be obtained. This clinical material comprises, 36 patients with deafferentation pain. The mean age at the time o f surgery was 52. Mean duration of the pain complaints was 8.5 years, ranging f r o m less than one year up to 28 years. Table I gives an overview of the treatments the patients underwent before they came to us for the implantation of a deep brain electrode. As can be seen, almost all patients took opioids to relieve their pain and m a n y of them were addicted. Before deciding to implant a deep brain electrode, patients were seen by a psychiatrist to confirm the organic origin of the pain complaints. Table II briefly summarizes the results obtained in the 36 cases. Stimulation target was VPLV P M or internal capsule, except for one case in which the paraventricular" nucleus of the thalamus was stimulated. We made a distinction between the initial success and the success at the latest follow-up.
392
J Gybels, R Kupers
The mean duration of the follow-up was 4 years. Initial pain relief was obtained in 22 patients, ie 61%. Actually, 11 of them are still pain-free, which indicates a success rate at the latest follow-up of 30°70. Only one of them takes some light analgesics, the others are completely drug-free. Our worst results were obtained in postherpetic neuralgia (0 success out of 5), thalamic pain (1 success out of 4) and phantom pain (1 success out of 4). These results are in sharp contrast with those described by Siegfried (1983) who reported 12 successes out of 14 cases with postherapeutic neuralgia and 4 successes out o f 5 with thalamic pain. It is interesting to note that in our series, 4 initial successes out of 5 were reported in the phantom pain group, but 3 of them were only partially relieved of their pain (50 to 80%). In these 3 patients, the effect of stimulation faded away over time. It was a general trend that the patients with an initial high pain relief score obtained the best results at longterm follow-up. We reviewed the literature on DBS up to 1988 (Gybels and Sweet, 1989). From this survey it appears that 57% out of a total of 1 539 patients benefitted from DBS, a success rate which is remarkably higher than our own success rate of 30%. When we look at the individual series, it is striking that the results vary considerably between the different authors. To give an example, in the treatment of cancer pain by P V G - P A G stimulation, Groth e t a l reported a success rate of 85% (22 out of 26 patients), Lazorthes a success rate of 47°70 (17 out of 36 patients), while we only obtained a success rate of 14°70 (1 patient out of 7) (Gybels and Sweet, 1989). What is the reason for this large discrepancy in results ? Is it patient selection, the qualification of what is called a success or the thoroughness with which the follow-up studies have been performed ? Perhaps the most plausible explanation for the discrepancy in success rates could be that different structures were being stimulated. We tested this hypothesis by means of autopsy data obtained in some of our patients treated by P V G - P A G stimulation for cancer pain. Comparison of our autopsy data (Gybels e t al, 1980) with 5 cases reported by Boivie and Meyerson (1982) shows that in our less favorable series the electrode entered the P V G more posteriorly. Another explanation which has been put forward to explain the variable results of VPL-VPM stimu-
Table I. List of treatments applied to the patients with deafferenation pain (n = 36) prior to implantation of a deep brain electrode. Treatment
No
Treatment
No
Opioid medication Non-opioid medication Nerve infiltrations Neurosurgery Orthopedic surgery (esp amputation)
33 36 15 20 8
Sympathetic blocks Psychotherapy/behavioural therapy Dorsal column stimulation TENS Acupuncture Physiotherapy
3 2 13 8 4 5
DBS for treatment of chronic pain
393
lation concerns the particular design o f the electrode tip. Indeed, favorable results have been reported by Mazars et al (1979) in the treatment o f deafferentation pain. These authors attribute their high success rate to the particular design o f their electrodes. However, using M a z a r s ' electrodes, target coordinates and stimulation parameters in 2 o f our patients did not result in pain relief. The same result has been reported by others (eg Meyerson personnal communication). Some authors state that the success rate o f P V G - P A G stimulation can rise by a preliminary m o r p h i n e test (Hosobuchi, 1982). Prior to surgery, morphine, up to 30 m g IV is administered to the patient. W h e n the patient reports dose-dependent gradual pain relief which is reversed by the administration o f naloxone, h e / s h e is considered as a g o o d candidate for P V G - P A G stimu-lation. Not everyone accepts the value o f this test (Arn6r et al, 1986; Y o u n g and C h a m b i , 1987) and o u r own experience has also been disappointing. All the a b o v e - m e n t i o n e d considerations m a y for some part explain the observed differences in success rates but, however, they remain tentative. As long as we remain devoid o f controlled multicentre studies in which standardized protocols are used and in which success is evaluated by an independent observer, we c a n n o t give a definitive answer to the question as to why success rates v a r y so widely a m o n g different neurosurgeons..' Table III shows the effect o f DBS in relation to the pain syndrome. As can be seen, success rates range f r o m 27 to 85% (Gybels and Sweet, 1989). The best results were obtained in p o s t c o r d o t o m y and low back pain, the worst results in paraplegia and thalamic pain. J In a recent study, the effect o f m o t o r cortex stimulation in the control o f thalamic pain was investigated ( T s u b o k a w a et al, 1990). A c c o r d i n g to these authors, the effect o f thalamic stimulation in the control o f deafferentation pain due to lesions in
Table II. Results of DBS (VPL-VPM, CI) stimulation in deafferenation pain (Gybels et al, 1990).
Table III. Results of DBS (VPL-VPM) stimulation in deafferentation and low back pain.
Diagnosis
n
Diagnosis
Thalamic pain Facial anesthesia dolorosa Brachial plexus avulsion Phantom pain Spinal cord lesion Postcordotomy pain Postradicotomy pain Postzoster neuralgia Failed disc surgery Sudeck's atrophy
5 6 7 4 5 1 1 5 1 1 36
Success
1 2 3 1 2 0 0 0 1 1
(3) (4) (4) (4) (3) (0) (1) (1) (l) (1)
11 (22)
Numbers between brackets indicate initial results.
n
% Success
Amputation/phantom limb Brachial plexus avulsion Thalamic pain Anesthesia dolorosa Paraplegic pain Postcordotomy pain Spinal cord/peripheral nerve Postherpetic neuralgia Causalgia
87 65 70 106 32 20
67 55 27 42 34 85
69 47 26
72 57 69
Low back pain
254
79
394
J Gybels, R Kupers
the central nervous system is not satisfactory. In 25 of such patients they implanted electrodes for m o t o r cortex stimulation. The reported success rate after 7 months of foUow-up was 75%, which is r e m a r k a b l y h i g h . As already mentioned, the success of DBS at short-term follow-up is considerably higher than at long-term follow-up. Notwithstanding this, the therapeutic success often lasts for several years. In our experience with DBS we have observed patients who have been pain-free for more than 10 years. All of them still use their stimulator, although often less frequently than at the beginning, since they know that if stimulation is stopped, the initial pain with its very real characteristics will slowly reappear. In the evaluation of the success rate of DBS, two important issues should not be forgotten. Firstly, as criterion for success, a pain relief score of 50% or more is often used. Of course this is an arbitrary criterion and one m a y wonder what the functional worth of 50% pain relief is since it implies that a certain number of patients who are classified as successes still experience pain, albeit less than before the implantation of the electrode. Although it is not possible to give exact numbers, we can state that in the group of 57% patients who were classified as a success, only a minority were completely pain-free. A second issue to bear in mind is that most of the patients who undergo implantation of a deep brain electrode, have already undergone all other possible treatment such as opioids, antidepressants, physiotherapy, psychotherapy, nerve infiltrations, orthopaedic surgery, ablative neurosurgery, peripheral nerve stimulation, dorsal column stimulation.., in other words, DBS is mostly used in patients with a long medical and " t h e r a p e u t i c " history and in which the chance for success is certainly smaller than in an average population. This certainly reduces the success rate of DBS. All too often, DBS is considered as a last resort that can be tried when all other treatment modalities have failed and only few efforts have been undertaken to find specific indications for the use of it.
Possible mechanisms in pain relief by stimulation of the somatosensory thalamus For stimulation of the somatosensory thalamus to be successful, two important conditions must be fulfilled. Firstly, somatosensory thalamic stimulation is especially effective and is almost exclusively used for the treatment of deafferentation or neurogenic pain. Stimulation of the somatosensory thalamus in the treatment of nociceptive pain is usually ineffective although some favorable results have been reported (Tsubokawa et al, 1984). A second criterion which must be fulfilled is that the stimulation induces paresthesias in the painful zone. However, the inverse is not true : the production of paresthesias i n t h e painful zone is not a guarantee for successful pain relief. In our clinical experience we have observed patients in w h o m the paresthesias covered very well the painful zone without any effect on their pain. Let us add some peculiar observations which, although seemingly anecdotal, underline the complexity of the mechanisms by which stimulation suppresses chronic pain. In one case, the
DBS for treatment of chronic pain
395
patient reported complete pain relief after simple placement of the electrode, without stimulation. This patient remained pain-free until the stimulator was turned on : the pain came back and stimulation even worsened it. Another patient with causalgic pain complained of a burning sensation together with a permanent sensation of warmth. Stimulation in this patient suppressed the pain but the sensation of warmth in the hand persisted. A third patient with deafferentation pain reported that the stimulation induced " a pleasant sensation of warmth which pushed the pain to the background". The mechanism by which VPL-VPM stimulation suppresses chronic pain is not clear. Probably it is not due to the activation of an endogenous opioid system since the analgesic effect of VPL-VPM stimulation is not reversed by naloxone. In a study performed by Tsubokawa et al (1984), changes in thq l~vel of beta-endorphin in the ventricular fluid after thalamic stimulation were measured. It was found that after thalamic stimulation, beta-endorphin levels were more than twice the resting level. However, no significant differences in beta-endorphin levels were found between patients who reported complete pain relief and patients who only obtained partial pain relief. Moreover, a far more pronounced rise in betfi-endorphin levels was found after stimulation of the periaqueductal gray. From experimental work on monkeys (Gerhart et al, 1983), it appears that stimulation of the somatosens0ry thalamic nuclei can strongly inhibit spinothalamic tract (STT) neurons. Responses evoked by noxious as well as innocuous mechanical and thermal stimuli were suppressed by VPL-VPM stimulation. This has lead to the suggestion that the neural correlate of thalamic stimulation produced pain~suppression might lie in its capacity to inhibit STT neurons. However, one must be cautious with such an explanation. The study by Gerhart et a! (1983) was performed in intact monkeys. It has been shown in man (Loeser et al, 1968) and in the animal (Lombard and Larabi, 1983) that lesions in the nervous system can induce hyperactivity in the dorsal horn. It is not because VPL-VPM stimulation can suppress activity in STT neurons, induced by a brief noxious stimulus, that can also suppress the abnormal neural activity often seen after deafferentation. Furthermore, it is known that the effect of brain stimulation can dramatically~change after lesions in the nervous system (Hodge et al, 1983). There are however some experimental data indicating that hyperactivity in the cat trigeminal medullary subnucleus caudalis, induced by retrogasserian trigeminal rhizotomy, can be inhibited by VPL stimulation (Tsubokawa et al, 1985). Since no significant descending projections from the VPL-VPM are known, the question remains as to how stimulation of this target might suppress STT neurons. Anatomical studies have shown that STT neurons not only project to the thalamus but that they also send axon collaterals to the periaqueductal gray (Price et al, 1978) and nucleus raphe magnus (Giesler et al, 1981). Since stimulation of these latter structures may inhibit STT neurons (Willis, 1988), it might be that VPL stimulation antidromically activates the descending inhibitory pathways in the raphe magnus and periaqueductal gray. This brings us to the question of the neural basis of this VPL-VPM induced excitation of raphe-spinal neurons. One can hypothesize that,
396
DBS for treatment of chronic pain
since the analgesic effect of somatosensory thalamic stimulation is not reversed by the administration of naloxone, mechanisms other than opioid must be at the basis of VPL-induced excitation of raphe-spinal neurons. It has been suggested that a dopaminergic mechanism might be involved (Tsubokawa et al, 1982). This'hypothesis has recently been strengthened by clinical observations (Hosobuchi, 1990). This author investigated the effect of an antidopaminergic (alpha-methylphdopa) in 7 patients with periaqueductal gray electrodes and in 7 patients with somatosensory thalamic electrodes. It was found that neither alpha-methyldopa nor placebo had any effect on P A G induced analgesia. In contrast, in 6 of the 7 patients treated with somatosensory thalamic stimulation, it was found that the stimulation-induced paresthesias ceased 1.5 to 2 min after administration of the antidopaminergic agent. In 3 of these patients, the deafferentation pain reappeared. When paresthesias returned, the pain again ceased. The phenomenon of stimulation tolerance, which is often seen after sustained stimulation of VPL-VPM target, has also been linked with the involvement of a dopaminergic mechanism. In an electrophysiological study, it has been shown that after long sustained stimulation of the VPL, the facilitatory effect on raphe-spinal neurons completely disappears, a phenomenon which could be reversed by the systemic administration of LL-dopa (Tsubokawa et al, 1982). In clinical practice also, it has been demonstrated that stimulation tolerance can largely be reduced by the concurrent administration o f L-dopa (Tsubokawa et al, 1982). The major problem when one tries to explain clinical pain relief on the basis of electrophysiological findings is the large difference in the duration of both phenomena. While the experimentally shown inhibition is in the order of milliseconds, the observed clinical pain relief after VPL-VPM stimulation can last for hours and occasionally days or even weeks. It implies that the experimentally observed inhibition of nociceptive neurons is inadequate to explain the long lasting pain relief which can be induced by VPL-VPM stimulation.
Acknowledgments The authors are indebted to P De Sutter and M Feytons-Heeren for expert technical assistance. Some aspects of this work were supported by the FGWO (grant 3.0031.87) of Belgium.
References Arn6r S, Lindblom U, Meyerson B (1986) Letters to the editor. Pain 24, 117-118 Boivie J, Meyerson BA (1982) A correlative anatomical and clinical study of pain suppression by deep brain stimulation. Pain 13, 113-126
Gerhart KD, Yezierski RP, Fang ZR, Willis WD (1983) Inhibition of primate spinothalamic tract neurons by stimulation in ventral posterior lateral (VPLc) thalamic nucleus: possible mechanisms. J Neurophysiol 49, 406-423
DBS for treatment of chronic pain Giesler G J, Yezierski RP, Gerhart KD, Willis WD (1981) Spinothalamic tract neurons that project to medial a n d / o r lateral thalamic nuclei: evidence for a physiologically novel population of spinal cord neurons. J Neurophysiol 46, 1285-1308 Gybels J, Dora R, Cosyns P (1980) Electrical stimulation of the central gray for pain relief in human : autopsy data. Acta Neurochir (suppl) 30, 259-268 Gybels J, Kupers R, Van Calenbergh F (1990) Physiological approach to management of pain. In: Recent Achievements in Restorative Neurology, vol 3 (Dimitrijevic MR, Wall PD, Lindblomu, eds) Karger, Basel, 7 9 - 8 6 Gybels JM, Sweet W H (1989) Neurosurgical treatment o f persistent pain. Karger, Basel, 442 pp Head H, Holmes G (1911) Sensory disturbances from cerebral lesions. Brain 34, 102-254 Heath RG, Mickle W A (1960) Evaluation of seven years experience with depth electrode studies in human patients. In: Electrical Studies on the Unanesthetized Brain, eds (Ramey ER, O'Doherty LDS, eds) Hoeber, New York, 214 Hodge, C J, Apkarian AV, Owen MP, Hanson BS (1983) Changes in the effects of stimulation of locus coeruleus and nucleus raphe magnus following dorsal rhizotomy. Brain Res 288, 325-329 Hosobuchi Y (1982) Analgesia induced by brain stimulation with chronically implanted electrodes. In: Operative Neurosurgical Techniques: Indications, Methods and Results vol 2 (Schmidek HH, Sweet LWH, eds). Grune & Stratton, New York, 981-991 Hosobuchi Y (1990) Alpha-methyldopa blocks the anagesic effect of sensory thalamic stimulation in humans. Pain (suppl) 5, $274 Hosobuchi Y, Adams JE, Rutkin B (1973) Chronic thalamic stimulation for the con-
397
trol of facial anesthesia dolorosa. Arch Neurol 29, 158-161 Kupers RC, Vos BPJ, Gybels JM (1988) Stimulation of the nucleus paraventricularis thalami suppresses scratching and biting behaviour of arthritic rats and exerts a powerful effect on tests for acute pain. Pain 32, 115-125 Lazorthes Y (1979) European study on deep brain stimulation. R6sum6 of the 3rd European workshop on electrical neurostimulation. Medtronic, Paris (in mimeo)" ¢ Loeser JD, Ward A A , White LE (1968) Chronic deafferenation of human spinal cord neurons. J Neurosurg 29, 48-50 Lombard MC (1989) Le syndrome de d6saff6rentation chez le rat: approches comportementales, 61ectrophysiologiques et pharmacologiques. Th6se, Paris Lombard MC, Larabi Y (1983) Electrophysiological study of cervical clorsal horn cells in partially deafferented rats. A d v Pain Res Ther 5, 147-154 Mazars G, M6rienne L, Ciolocca C (1979) Comparative study of electrical stimulation of posterior thalamic nuclei, periaqueductal gray, and other midline mesencephalic structures in man. A d v Pain Res Ther 3, 541-546 Melzack R, Wall PD (1965) Pain mechanisms: a new theory. Science 150, 971-978 Price DD, Hayes RL, Ruda M, Dubner R • / (1978) Spatia(and temporal transformations of input to spinothalamic tract neurons and their relation to somatic sensations. J Neurophysiol 41, 933-947 Reynolds DV (1969) Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 164, 444-445 Richardson DE, Akil H (1977) Pain reduction by electrical brain stimulation in man. II. Chronic self-administration in the periventricular gray matter. J Neurosurg 47, 184-194 Shealy CN, Mortimer JT, Reswick J (1967)
398
J Gybels, R Kupers
Electrical inhibition of pain by stimulation of the dorsal column: preliminary clinical reports. Anesth Analg 46, 489-491 Siegfried J (1983) Long-term results of electrical stimulation in the treatment of pain by means of implanted electrodes (epidural spinal cord and deep brain stimulation). In: Pain Therapy (Rizzi R, Visentin M, eds). Elsevier, Amsterdam, 463 -475 Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T (1985) Deafferentation pain and stimulation of the thalamic sensory relay nucleus: clinical and experimental study. Appl Neurophysiol 48, 166-171 Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S (1990) Motor cortex stimulation for control of thalamic pain. Pain (suppl) 5, $491 Tsubokawa T, Yamamoto T, Katayama Y, Hirayama T, Sibuya H (1984) Thalamic relay nucleus stimulation for relief of in-
tractable pain. Clinical results and betaendorphin immuno-reactivity in the cerebrospinal fluid. Pain 18, 115-126 Tsubokawa T, Yamamoto T, Katayama Y, Moriyasu N (1982) Clinical results and physiological basis of thalamic relay nucleus stimulation for relief of intractable pain with morphine tolerance. Appl Neurophysiol 45, 143-155 Wall PD, Sweet WH (1967) Temporary abolition of pain in man. Science 155, 108-109 Willis WD (1988) Anatomy and physiology of descending control of nociceptive responses of dorsal horn neurons: comprehensive review. Prog Brain Res 77, 1-29
Young RF, Chambi VI (1987) Pain relief by electrical stimulation of the periaqueductal and periventricular gray matter. Evidence for a non-opioid mechanism. J Neurosurg 66, 364-371
Neurophysiol Clin (1990) 20, 389-398 © Elsevier, Paris
Review article
Deep brain stimulation in the treatment of chronic pain in m a n : where and w h y ? J Gybels, R Kuper s Department o f Neurology and Neurosurgery, University o f Leuven, U Z Gasthuisberg, Herestraat 49 B-3000 Leuven, Belgium (Received 26 June 1990; accepted 2 August 1990)
Summary - in current clinical practice, two brain structures are stimulated for the relief of chronic pain, namely the somatosensory thal~mic nuclei (VPL-VPM) and the periventricfilar and p&iaqueductal gray matter (PVG-PAG). Whereas gtimulation of the V P L - V P M is almost exclusively used for the treatment of deafferentation pain, stimulation of the P V G - P A G is mostly used in cases of nociceptive pain. We present our results of V P L - V P M stimulation in 36 patients with deafferentation pain. Initial pain relief was obtained in 61% of patients. To-day, after a mean follow-up of more t h a n 4 years, 30% are still pain free. This success rate was found to be lower than the m e a n reported success r a t e ' o f 5"~%, based on a survey of the world literature. U p o n reviewing the literature, it was apparent that the reported success rates vary considerably between different authors. Some tentative explanations are given for this large discrepancy in success rate. The mechanisms by which electrical stimulation of the V P L - V P M suppresses deafferentation pain remain to be elucidated. Recent clinical and experimental findings suggest that a dopaminergic m e c h a n i s m might be involved. brain stimulation / pain
R~sum~ - Stimulation c~r~brale profonde dans les d6uleurs chroniques: off et pourquoi? En pratique clinique courante, deux structures c&~brales sont stimtll~es p o u r traiter la douleur chronique : les noyaux sensitifs du thalamus ( V P L - VPM) et la substance grise p&iventriculaire et pOriaqueducale (PVG-PA G). La stimulation du noyau V P L - V P M est appliquOe clans le traitement de la douleur de d~saff&entation, tandis que la stimulation du P V G - P A G est presque exclusivement utilis~e p o u r soulager la douleur nociceptire. N o u s pr~sentons nos r~sultats de la stimulation du VPL- V P M chez 36 malades, souffrant de douleur de d~saffdrentation. Un succbs initial a OtO obtenu chez 61% des malades. Actuellement, aprbs un suivi d'une dur~e moyenne de 4 ans, 30% de ces interventions sont consid~r6es c o m m e Otant un suceOs. Ce r~sultat est moins bon que celui relevd clans la litt&ature mondiale. Il f a u t ajouter qu "on note des variations importantes des taux de succbs entre les diff&ents auteurs. N o u s d o n n o n s une esquisse d'explication de cette variabifit~ clans les rdsultats. Les m~canismes par lesquels la stimulation 6lectrique des noyaux sensitifs du thalamus diminue la douleur restent ~ Oclaircir. Certaines donn~es exp&imentales et cliniques laissent penser qu 'un m&'anisme dopaminergique peut 6tre impliqud. stimulation c6r6brale / douleur
390
J Gybels, R Kupers
Introduction
Although the possibility of producing analgesi a by electrical stimulation of the human brain has already been reported in the late fifties (Heath and Mickle, 1960), the real interest in brain stimulation for the treatment of chronic pain in man only arose at the end the sixties. Important stimuli for this sudden interest were the proposal of the gate control theory by Melzack and Wall (1965) and Reynolds' discovery of stimulation-produced analgesia in the animal (1969). The discovery by Reynolds that electrical stimulation of the rat midbrain can produce profound analgesia without the concurrent administration of drugs opened new horizons for the treatment of chronic pain. It marked the beginning of a new era in neurosurgical pain treatment by offering an alternative for the previous destructive and irreversible procedures. The subsequent discovery, some years later, of the narrow relationship between stimulation-produced analgesia and the endogenous opioid system, offered a theoretical basis for the understanding of how stimulation of the midbrain may suppress pain. However, it was not until 1973 that the first clinical applications of periventricular gray stimulation in man appeared (Richardson and Akil, 1977). A second line of interest arose from the gate control theory. Although historically older than the discovery of stimulation-produced analgesia, the importance of the gate control theory as a conceptual framework for deep brain stimulation (DBS) was highlighted by Reynolds' discovery. It formed the first important incentive for the use of stimulation procedures in the treatment of chronic pain. According to the gate control theory, stimulation of large diameter fibers is capable of inhibiting nociceptive information. The first successful clinical applications of this theory were periph eral nerve stimulation (Wall and Sweet, 1967), and dorsal column stimulation (Shealy et al, 1967). Since in certain clinical syndromes, eg after brachial plexus avulsion, anesthesia dolorosa etc, there is a lack of primary afferent fibers in the peripheral nerve or dorsal column, it was suggested that in those cases, supraspinal stimulation at the level of the somatosensory system might offer an alternative substrate for activating the lemniscal system. Although the first publication on somatosensory thalamic stimulation in the treatment of deafferentation pain only appeared in 1973 (Hosobuchi et al), it had already been practiced since the early sixties by Mazars et a! (1979), ie even before the proposal of the gate control theory. Their theoretical framework was the theory of Head and Holmes (1911) according to which pain might be the consequence of an imbalance between protopathic and epicritic sensory functioning. Stimulation of the thalamic sensory relay nuclei would presumably increase the epicritic component and hence inhibit the noxious protopathic inflow. Although welcomed with greht enthusiasm by those neurosurgeons who had a profound interest in pain research and therapy, DBS was not always as successful as was initially hoped. The clinical results did not always fit with the experimental findings. This was in part due to the fact that in the animal, acute experimental pain was investigated, while in clinical practice, chronic pain had to be beaten (it was ap-
DBS for treatment of chronic pain
391
parently much easier to produce a prolonged tail-flick in the rat than to alleviate chronic pain in man). It is only recently that DBS has been studied in animal models of chronic pain (Kupers et al, 1988; Lombard, 1989).Moreover, large discrepancies were noted between the results of different neurosurgical groups, some reporting great successes, others almost complete failures. Although DBS for the treatment of chronic pain has not become a real routine intervention, it is now accepted that it can offer an alternative in those cases in which all other treatment modalities have failed. However, knowledge as to how electrical brain stimulation suppresses chronic pain is still fragmentary. In the first part of this paper we will summarize clinical data on DBS ; in the second part we will explore the possible mechanisms by which DBS alleviates pain due to lesions o f the nervous system.
Clinical results
In current clinical practice, ~two brain structures are stimulated for the relief of chronic pain, namely the primary' somatosensory relay nuclei of the thalamus and their afferent and efferent pathways (VPL-VPM), and the periventricular and periaqueductal gray matter ( P V G - P A G ) . Whereas stimulation of the somatosensory thalamus is mainly used for the treatment of deafferentation pain, P V G - P A G stimulation is mostly used in cases of nociceptive pain (Gybels and Sweet, 1989). However, some authors also report favorable results of V P L - V P M stimulation in nociceptive pain (Tsubokawa et al, 1984). Stimulation of both P V G - P A G and V P L - V P M targets is sometimes performed in those cases in which there is a mixture of somatic and neurogenic pain (eg in low back pain). In this paper we will discuss some of our results of V P L - V P M stimulation in patients with deafferentation pain. We will limit ourselves to the results o f those patients who were frequently seen for consultation and in which a thorough evaluation of the therapeutic result on long-term follow-up could be obtained. This clinical material comprises, 36 patients with deafferentation pain. The mean age at the time o f surgery was 52. Mean duration of the pain complaints was 8.5 years, ranging f r o m less than one year up to 28 years. Table I gives an overview of the treatments the patients underwent before they came to us for the implantation of a deep brain electrode. As can be seen, almost all patients took opioids to relieve their pain and m a n y of them were addicted. Before deciding to implant a deep brain electrode, patients were seen by a psychiatrist to confirm the organic origin of the pain complaints. Table II briefly summarizes the results obtained in the 36 cases. Stimulation target was VPLV P M or internal capsule, except for one case in which the paraventricular" nucleus of the thalamus was stimulated. We made a distinction between the initial success and the success at the latest follow-up.
392
J Gybels, R Kupers
The mean duration of the follow-up was 4 years. Initial pain relief was obtained in 22 patients, ie 61%. Actually, 11 of them are still pain-free, which indicates a success rate at the latest follow-up of 30°70. Only one of them takes some light analgesics, the others are completely drug-free. Our worst results were obtained in postherpetic neuralgia (0 success out of 5), thalamic pain (1 success out of 4) and phantom pain (1 success out of 4). These results are in sharp contrast with those described by Siegfried (1983) who reported 12 successes out of 14 cases with postherapeutic neuralgia and 4 successes out o f 5 with thalamic pain. It is interesting to note that in our series, 4 initial successes out of 5 were reported in the phantom pain group, but 3 of them were only partially relieved of their pain (50 to 80%). In these 3 patients, the effect of stimulation faded away over time. It was a general trend that the patients with an initial high pain relief score obtained the best results at longterm follow-up. We reviewed the literature on DBS up to 1988 (Gybels and Sweet, 1989). From this survey it appears that 57% out of a total of 1 539 patients benefitted from DBS, a success rate which is remarkably higher than our own success rate of 30%. When we look at the individual series, it is striking that the results vary considerably between the different authors. To give an example, in the treatment of cancer pain by P V G - P A G stimulation, Groth e t a l reported a success rate of 85% (22 out of 26 patients), Lazorthes a success rate of 47°70 (17 out of 36 patients), while we only obtained a success rate of 14°70 (1 patient out of 7) (Gybels and Sweet, 1989). What is the reason for this large discrepancy in results ? Is it patient selection, the qualification of what is called a success or the thoroughness with which the follow-up studies have been performed ? Perhaps the most plausible explanation for the discrepancy in success rates could be that different structures were being stimulated. We tested this hypothesis by means of autopsy data obtained in some of our patients treated by P V G - P A G stimulation for cancer pain. Comparison of our autopsy data (Gybels e t al, 1980) with 5 cases reported by Boivie and Meyerson (1982) shows that in our less favorable series the electrode entered the P V G more posteriorly. Another explanation which has been put forward to explain the variable results of VPL-VPM stimu-
Table I. List of treatments applied to the patients with deafferenation pain (n = 36) prior to implantation of a deep brain electrode. Treatment
No
Treatment
No
Opioid medication Non-opioid medication Nerve infiltrations Neurosurgery Orthopedic surgery (esp amputation)
33 36 15 20 8
Sympathetic blocks Psychotherapy/behavioural therapy Dorsal column stimulation TENS Acupuncture Physiotherapy
3 2 13 8 4 5
DBS for treatment of chronic pain
393
lation concerns the particular design o f the electrode tip. Indeed, favorable results have been reported by Mazars et al (1979) in the treatment o f deafferentation pain. These authors attribute their high success rate to the particular design o f their electrodes. However, using M a z a r s ' electrodes, target coordinates and stimulation parameters in 2 o f our patients did not result in pain relief. The same result has been reported by others (eg Meyerson personnal communication). Some authors state that the success rate o f P V G - P A G stimulation can rise by a preliminary m o r p h i n e test (Hosobuchi, 1982). Prior to surgery, morphine, up to 30 m g IV is administered to the patient. W h e n the patient reports dose-dependent gradual pain relief which is reversed by the administration o f naloxone, h e / s h e is considered as a g o o d candidate for P V G - P A G stimu-lation. Not everyone accepts the value o f this test (Arn6r et al, 1986; Y o u n g and C h a m b i , 1987) and o u r own experience has also been disappointing. All the a b o v e - m e n t i o n e d considerations m a y for some part explain the observed differences in success rates but, however, they remain tentative. As long as we remain devoid o f controlled multicentre studies in which standardized protocols are used and in which success is evaluated by an independent observer, we c a n n o t give a definitive answer to the question as to why success rates v a r y so widely a m o n g different neurosurgeons..' Table III shows the effect o f DBS in relation to the pain syndrome. As can be seen, success rates range f r o m 27 to 85% (Gybels and Sweet, 1989). The best results were obtained in p o s t c o r d o t o m y and low back pain, the worst results in paraplegia and thalamic pain. J In a recent study, the effect o f m o t o r cortex stimulation in the control o f thalamic pain was investigated ( T s u b o k a w a et al, 1990). A c c o r d i n g to these authors, the effect o f thalamic stimulation in the control o f deafferentation pain due to lesions in
Table II. Results of DBS (VPL-VPM, CI) stimulation in deafferenation pain (Gybels et al, 1990).
Table III. Results of DBS (VPL-VPM) stimulation in deafferentation and low back pain.
Diagnosis
n
Diagnosis
Thalamic pain Facial anesthesia dolorosa Brachial plexus avulsion Phantom pain Spinal cord lesion Postcordotomy pain Postradicotomy pain Postzoster neuralgia Failed disc surgery Sudeck's atrophy
5 6 7 4 5 1 1 5 1 1 36
Success
1 2 3 1 2 0 0 0 1 1
(3) (4) (4) (4) (3) (0) (1) (1) (l) (1)
11 (22)
Numbers between brackets indicate initial results.
n
% Success
Amputation/phantom limb Brachial plexus avulsion Thalamic pain Anesthesia dolorosa Paraplegic pain Postcordotomy pain Spinal cord/peripheral nerve Postherpetic neuralgia Causalgia
87 65 70 106 32 20
67 55 27 42 34 85
69 47 26
72 57 69
Low back pain
254
79
394
J Gybels, R Kupers
the central nervous system is not satisfactory. In 25 of such patients they implanted electrodes for m o t o r cortex stimulation. The reported success rate after 7 months of foUow-up was 75%, which is r e m a r k a b l y h i g h . As already mentioned, the success of DBS at short-term follow-up is considerably higher than at long-term follow-up. Notwithstanding this, the therapeutic success often lasts for several years. In our experience with DBS we have observed patients who have been pain-free for more than 10 years. All of them still use their stimulator, although often less frequently than at the beginning, since they know that if stimulation is stopped, the initial pain with its very real characteristics will slowly reappear. In the evaluation of the success rate of DBS, two important issues should not be forgotten. Firstly, as criterion for success, a pain relief score of 50% or more is often used. Of course this is an arbitrary criterion and one m a y wonder what the functional worth of 50% pain relief is since it implies that a certain number of patients who are classified as successes still experience pain, albeit less than before the implantation of the electrode. Although it is not possible to give exact numbers, we can state that in the group of 57% patients who were classified as a success, only a minority were completely pain-free. A second issue to bear in mind is that most of the patients who undergo implantation of a deep brain electrode, have already undergone all other possible treatment such as opioids, antidepressants, physiotherapy, psychotherapy, nerve infiltrations, orthopaedic surgery, ablative neurosurgery, peripheral nerve stimulation, dorsal column stimulation.., in other words, DBS is mostly used in patients with a long medical and " t h e r a p e u t i c " history and in which the chance for success is certainly smaller than in an average population. This certainly reduces the success rate of DBS. All too often, DBS is considered as a last resort that can be tried when all other treatment modalities have failed and only few efforts have been undertaken to find specific indications for the use of it.
Possible mechanisms in pain relief by stimulation of the somatosensory thalamus For stimulation of the somatosensory thalamus to be successful, two important conditions must be fulfilled. Firstly, somatosensory thalamic stimulation is especially effective and is almost exclusively used for the treatment of deafferentation or neurogenic pain. Stimulation of the somatosensory thalamus in the treatment of nociceptive pain is usually ineffective although some favorable results have been reported (Tsubokawa et al, 1984). A second criterion which must be fulfilled is that the stimulation induces paresthesias in the painful zone. However, the inverse is not true : the production of paresthesias i n t h e painful zone is not a guarantee for successful pain relief. In our clinical experience we have observed patients in w h o m the paresthesias covered very well the painful zone without any effect on their pain. Let us add some peculiar observations which, although seemingly anecdotal, underline the complexity of the mechanisms by which stimulation suppresses chronic pain. In one case, the
DBS for treatment of chronic pain
395
patient reported complete pain relief after simple placement of the electrode, without stimulation. This patient remained pain-free until the stimulator was turned on : the pain came back and stimulation even worsened it. Another patient with causalgic pain complained of a burning sensation together with a permanent sensation of warmth. Stimulation in this patient suppressed the pain but the sensation of warmth in the hand persisted. A third patient with deafferentation pain reported that the stimulation induced " a pleasant sensation of warmth which pushed the pain to the background". The mechanism by which VPL-VPM stimulation suppresses chronic pain is not clear. Probably it is not due to the activation of an endogenous opioid system since the analgesic effect of VPL-VPM stimulation is not reversed by naloxone. In a study performed by Tsubokawa et al (1984), changes in thq l~vel of beta-endorphin in the ventricular fluid after thalamic stimulation were measured. It was found that after thalamic stimulation, beta-endorphin levels were more than twice the resting level. However, no significant differences in beta-endorphin levels were found between patients who reported complete pain relief and patients who only obtained partial pain relief. Moreover, a far more pronounced rise in betfi-endorphin levels was found after stimulation of the periaqueductal gray. From experimental work on monkeys (Gerhart et al, 1983), it appears that stimulation of the somatosens0ry thalamic nuclei can strongly inhibit spinothalamic tract (STT) neurons. Responses evoked by noxious as well as innocuous mechanical and thermal stimuli were suppressed by VPL-VPM stimulation. This has lead to the suggestion that the neural correlate of thalamic stimulation produced pain~suppression might lie in its capacity to inhibit STT neurons. However, one must be cautious with such an explanation. The study by Gerhart et a! (1983) was performed in intact monkeys. It has been shown in man (Loeser et al, 1968) and in the animal (Lombard and Larabi, 1983) that lesions in the nervous system can induce hyperactivity in the dorsal horn. It is not because VPL-VPM stimulation can suppress activity in STT neurons, induced by a brief noxious stimulus, that can also suppress the abnormal neural activity often seen after deafferentation. Furthermore, it is known that the effect of brain stimulation can dramatically~change after lesions in the nervous system (Hodge et al, 1983). There are however some experimental data indicating that hyperactivity in the cat trigeminal medullary subnucleus caudalis, induced by retrogasserian trigeminal rhizotomy, can be inhibited by VPL stimulation (Tsubokawa et al, 1985). Since no significant descending projections from the VPL-VPM are known, the question remains as to how stimulation of this target might suppress STT neurons. Anatomical studies have shown that STT neurons not only project to the thalamus but that they also send axon collaterals to the periaqueductal gray (Price et al, 1978) and nucleus raphe magnus (Giesler et al, 1981). Since stimulation of these latter structures may inhibit STT neurons (Willis, 1988), it might be that VPL stimulation antidromically activates the descending inhibitory pathways in the raphe magnus and periaqueductal gray. This brings us to the question of the neural basis of this VPL-VPM induced excitation of raphe-spinal neurons. One can hypothesize that,
396
DBS for treatment of chronic pain
since the analgesic effect of somatosensory thalamic stimulation is not reversed by the administration of naloxone, mechanisms other than opioid must be at the basis of VPL-induced excitation of raphe-spinal neurons. It has been suggested that a dopaminergic mechanism might be involved (Tsubokawa et al, 1982). This'hypothesis has recently been strengthened by clinical observations (Hosobuchi, 1990). This author investigated the effect of an antidopaminergic (alpha-methylphdopa) in 7 patients with periaqueductal gray electrodes and in 7 patients with somatosensory thalamic electrodes. It was found that neither alpha-methyldopa nor placebo had any effect on P A G induced analgesia. In contrast, in 6 of the 7 patients treated with somatosensory thalamic stimulation, it was found that the stimulation-induced paresthesias ceased 1.5 to 2 min after administration of the antidopaminergic agent. In 3 of these patients, the deafferentation pain reappeared. When paresthesias returned, the pain again ceased. The phenomenon of stimulation tolerance, which is often seen after sustained stimulation of VPL-VPM target, has also been linked with the involvement of a dopaminergic mechanism. In an electrophysiological study, it has been shown that after long sustained stimulation of the VPL, the facilitatory effect on raphe-spinal neurons completely disappears, a phenomenon which could be reversed by the systemic administration of LL-dopa (Tsubokawa et al, 1982). In clinical practice also, it has been demonstrated that stimulation tolerance can largely be reduced by the concurrent administration o f L-dopa (Tsubokawa et al, 1982). The major problem when one tries to explain clinical pain relief on the basis of electrophysiological findings is the large difference in the duration of both phenomena. While the experimentally shown inhibition is in the order of milliseconds, the observed clinical pain relief after VPL-VPM stimulation can last for hours and occasionally days or even weeks. It implies that the experimentally observed inhibition of nociceptive neurons is inadequate to explain the long lasting pain relief which can be induced by VPL-VPM stimulation.
Acknowledgments The authors are indebted to P De Sutter and M Feytons-Heeren for expert technical assistance. Some aspects of this work were supported by the FGWO (grant 3.0031.87) of Belgium.
References Arn6r S, Lindblom U, Meyerson B (1986) Letters to the editor. Pain 24, 117-118 Boivie J, Meyerson BA (1982) A correlative anatomical and clinical study of pain suppression by deep brain stimulation. Pain 13, 113-126
Gerhart KD, Yezierski RP, Fang ZR, Willis WD (1983) Inhibition of primate spinothalamic tract neurons by stimulation in ventral posterior lateral (VPLc) thalamic nucleus: possible mechanisms. J Neurophysiol 49, 406-423
DBS for treatment of chronic pain Giesler G J, Yezierski RP, Gerhart KD, Willis WD (1981) Spinothalamic tract neurons that project to medial a n d / o r lateral thalamic nuclei: evidence for a physiologically novel population of spinal cord neurons. J Neurophysiol 46, 1285-1308 Gybels J, Dora R, Cosyns P (1980) Electrical stimulation of the central gray for pain relief in human : autopsy data. Acta Neurochir (suppl) 30, 259-268 Gybels J, Kupers R, Van Calenbergh F (1990) Physiological approach to management of pain. In: Recent Achievements in Restorative Neurology, vol 3 (Dimitrijevic MR, Wall PD, Lindblomu, eds) Karger, Basel, 7 9 - 8 6 Gybels JM, Sweet W H (1989) Neurosurgical treatment o f persistent pain. Karger, Basel, 442 pp Head H, Holmes G (1911) Sensory disturbances from cerebral lesions. Brain 34, 102-254 Heath RG, Mickle W A (1960) Evaluation of seven years experience with depth electrode studies in human patients. In: Electrical Studies on the Unanesthetized Brain, eds (Ramey ER, O'Doherty LDS, eds) Hoeber, New York, 214 Hodge, C J, Apkarian AV, Owen MP, Hanson BS (1983) Changes in the effects of stimulation of locus coeruleus and nucleus raphe magnus following dorsal rhizotomy. Brain Res 288, 325-329 Hosobuchi Y (1982) Analgesia induced by brain stimulation with chronically implanted electrodes. In: Operative Neurosurgical Techniques: Indications, Methods and Results vol 2 (Schmidek HH, Sweet LWH, eds). Grune & Stratton, New York, 981-991 Hosobuchi Y (1990) Alpha-methyldopa blocks the anagesic effect of sensory thalamic stimulation in humans. Pain (suppl) 5, $274 Hosobuchi Y, Adams JE, Rutkin B (1973) Chronic thalamic stimulation for the con-
397
trol of facial anesthesia dolorosa. Arch Neurol 29, 158-161 Kupers RC, Vos BPJ, Gybels JM (1988) Stimulation of the nucleus paraventricularis thalami suppresses scratching and biting behaviour of arthritic rats and exerts a powerful effect on tests for acute pain. Pain 32, 115-125 Lazorthes Y (1979) European study on deep brain stimulation. R6sum6 of the 3rd European workshop on electrical neurostimulation. Medtronic, Paris (in mimeo)" ¢ Loeser JD, Ward A A , White LE (1968) Chronic deafferenation of human spinal cord neurons. J Neurosurg 29, 48-50 Lombard MC (1989) Le syndrome de d6saff6rentation chez le rat: approches comportementales, 61ectrophysiologiques et pharmacologiques. Th6se, Paris Lombard MC, Larabi Y (1983) Electrophysiological study of cervical clorsal horn cells in partially deafferented rats. A d v Pain Res Ther 5, 147-154 Mazars G, M6rienne L, Ciolocca C (1979) Comparative study of electrical stimulation of posterior thalamic nuclei, periaqueductal gray, and other midline mesencephalic structures in man. A d v Pain Res Ther 3, 541-546 Melzack R, Wall PD (1965) Pain mechanisms: a new theory. Science 150, 971-978 Price DD, Hayes RL, Ruda M, Dubner R • / (1978) Spatia(and temporal transformations of input to spinothalamic tract neurons and their relation to somatic sensations. J Neurophysiol 41, 933-947 Reynolds DV (1969) Surgery in the rat during electrical analgesia induced by focal brain stimulation. Science 164, 444-445 Richardson DE, Akil H (1977) Pain reduction by electrical brain stimulation in man. II. Chronic self-administration in the periventricular gray matter. J Neurosurg 47, 184-194 Shealy CN, Mortimer JT, Reswick J (1967)
398
J Gybels, R Kupers
Electrical inhibition of pain by stimulation of the dorsal column: preliminary clinical reports. Anesth Analg 46, 489-491 Siegfried J (1983) Long-term results of electrical stimulation in the treatment of pain by means of implanted electrodes (epidural spinal cord and deep brain stimulation). In: Pain Therapy (Rizzi R, Visentin M, eds). Elsevier, Amsterdam, 463 -475 Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T (1985) Deafferentation pain and stimulation of the thalamic sensory relay nucleus: clinical and experimental study. Appl Neurophysiol 48, 166-171 Tsubokawa T, Katayama Y, Yamamoto T, Hirayama T, Koyama S (1990) Motor cortex stimulation for control of thalamic pain. Pain (suppl) 5, $491 Tsubokawa T, Yamamoto T, Katayama Y, Hirayama T, Sibuya H (1984) Thalamic relay nucleus stimulation for relief of in-
tractable pain. Clinical results and betaendorphin immuno-reactivity in the cerebrospinal fluid. Pain 18, 115-126 Tsubokawa T, Yamamoto T, Katayama Y, Moriyasu N (1982) Clinical results and physiological basis of thalamic relay nucleus stimulation for relief of intractable pain with morphine tolerance. Appl Neurophysiol 45, 143-155 Wall PD, Sweet WH (1967) Temporary abolition of pain in man. Science 155, 108-109 Willis WD (1988) Anatomy and physiology of descending control of nociceptive responses of dorsal horn neurons: comprehensive review. Prog Brain Res 77, 1-29
Young RF, Chambi VI (1987) Pain relief by electrical stimulation of the periaqueductal and periventricular gray matter. Evidence for a non-opioid mechanism. J Neurosurg 66, 364-371
