Search

Object. The goal of this study was to identify the neurological characteristics of patients with poststroke pain who show a favorable response to motor cortex (MC) stimulation used to control their pain. Methods. The neurological characteristics of 31 patients treated by MC stimulation were analyzed. In 15 patients (48%), excellent or good pain control (pain reduction > 60%) was achieved for follow-up periods of more than 2 years by using MC stimulation at intensities below the threshold for muscle contraction. Satisfactory pain control was achieved in 13 (73%) of 18 patients in whom motor weakness in the painful area was virtually absent or mild, but in only two (15%) of the 13 patients who demonstrated moderate or severe weakness in the painful area (p < 0.01). Muscle contraction was inducible in the painful area in 20 patients when stimulated at a higher intensity. No such muscle response was inducible in the remaining 11 patients, no matter how extensively the authors attempted to determine appropriate stimulation sites. Satisfactory pain control was achieved in 14 (70%) of the 20 patients in whom muscle contraction was inducible, but in only one (9%) of the 11 patients in whom muscle contraction was not inducible (p < 0.01). No significant relationship was observed between pain control and various sensory symptoms, including the presence of hypesthesia, spontaneous dysesthesia, hyperpathia, and allodynia, or the disappearance of the N20 component of the median nerve—evoked somatosensory scalp potential. No significant relationship existed between the effect of MC stimulation on the pain and stimulation-induced phenomena, including paresthesia, improvement in motor performance, and attenuation of involuntary movements. Conclusions. These findings suggest that the pain control afforded by MC stimulation requires neuronal circuits that are maintained by the presence of intact corticospinal tract neurons originating from the MC. Preoperative evaluation of motor weakness of the painful area appears to be useful for predicting a favorable response to MC stimulation in the control of poststroke pain.

In order to elucidate the mechanisms of release of EAAs and their excitotoxicity in cerebral contusion, cortical contusion was produced in the rat parietal cortex, and the changes in extracellular levels of EAAs in the central and peripheral areas of contusion were investigated using microdialysis. The cortical contusion induced a rapid increase in dialysate concentration of glutamate ([Glu]d) from a baseline level of 4.6+/-2.8 microM to a maximum level of 36.3+/-12.8 microM. This elevation of glutamate was significantly attenuated by mild hypothermia (32 degrees C for 90 min, comprising 20 min before and 70 min after the injury induction) in the peripheral area of contusion (p < 0.01) but not in the central area. Histological evaluations revealed that the hypothermia reduced the necrosis volume of contusion to 38.3% of that in the normothermic control (p < 0.01). In situ administration of Co2+, an inhibitor of Co2+-dependent exocytotic release of EAAs from the nerve terminals, via the microdialysis system, also attenuated the [Glu]d elevation following cortical contusion, in the peripheral area of contusion (p < 0.01) but not in the central area. These findings indicate that cerebral contusion involves heterogeneous and complex mechanisms of EAA release into the extracellular space. The release of EAAs in the contusion core was nonsensitive to hypothermia and Co2+ administration, suggesting that such EAA release was related to primary disruption of the cell membrane or vascular wall by the physical force of the head trauma, resulting in leakage of EAAs from the metabolic pool in the cytosole or blood stream. In contrast, in the peripheral area, the effectiveness of hypothermia and Co2+ administration implied a presynaptic mechanism of EAA release, which consisted, at least in part, of Ca2+-dependent exocytotic EAA release from depolarized nerve terminals. The EAAs released in the contusion core may diffuse towards a peripheral direction and act on the postsynaptic receptors, causing neuronal depolarization. Such a diffusion-reaction process appears to induce additional release of EAAs from the depolarized nerve terminals. Hypothermia may block this diffusion-reaction process and eventually reduce the contusion volume.

The jugular bulb oxygen saturation (SjO 2 ) and end-tidal carbon dioxide (ETCO 2 ) were monitored continuously during surgery in six cases of Moyamoya disease who had demonstrated multiple episodes of transient ischemic attacks (TIAs) and/or fluctuating neurological deficits preoperatively. The arterial carbon dioxide tension (PaCO 2 ) levels were also measured repeatedly at predetermined interval. In two cases (group H), the ETCO 2 was controlled at hypercapnic levels during surgery (45.5±1.5 mmHg) and the remaining four (group N) were operated on in a normocapnic state (39.0±2.0 mmHg). The group H patients demonstrated high levels of SjO 2 ranging from 72 to 85%, indicative of excessive hyperemia. One of the group H patients demonstrated mild and transient motor weakness postoperatively. The group N patients demonstrated normal levels of SjO 2 ranging from 66 to 78%. All the patients in both groups demonstrated fluctuations in SjO 2 levels in clear positive correlation with spontaneous changes in PaCO 2 levels. The present findings indicated that: (1) Global carbon dioxide reactivity of cerebral perfusion is well preserved in patients with Moyamoya disease; and (2) hypercapnia in these patients often causes excessive hyperemia. The occurrence of postoperative neurological deficits in association with such an excessive hyperemia suggests that hyperapnia during surgery is not always beneficial. Intraoperative monitoring of SjO 2 is useful for maintaining cerebral perfusion within the optimum range.

In an attempt to clarify the neurochemical background of central post-stroke pain and to undertake a pharmacological analysis, the basic pharmacological characteristics of this intractable pain syndrome were investigated by the morphine, thiamylal and ketamine tests. In addition, the correlation between the pharmacological characteristics and the effects of chronic motor cortex stimulation therapy was examined. The study employed 39 central post-stroke pain patients who had intractable hemibody pain associated with dysesthesias, and radiologically demonstrated lesions in the thalamic area (thalamic pain, n = 25) or suprathalamic area (suprathalamic pain, n = 14). The pharmacological evaluations showed that definite pain reduction occurred in eight of the 39 cases (20.5%) by the morphine test, in 22 of the 39 cases (56.4%) by the thiamylal test, and in 11 of 23 cases (47.8%) by the ketamine test. Based on these pharmacological assessments, there was no obvious difference between thalamic and suprathalamic pain. A comparison of the long-term follow-up results of chronic motor cortex stimulation therapy revealed that thiamylal and ketamine-sensitive and morphine-resistant cases displayed long-lasting pain reduction with chronic motor cortex stimulation therapy, whereas the remaining cases did not show good results. We conclude that pharmacological classification of central post-stroke pain by the morphine, thiamylal and ketamine tests could be useful for predicting the effects of chronic motor cortex stimulation therapy. It has recently been suggested that excitatory amino acids may be involved in the development of central post-stroke pain. However, the fact that only 23 of the present 39 cases (59.0%) of thalamic and suprathalamic pain were sensitive to the thiamylal or ketamine test reflects the complex pharmacological background and the difficulties associated with treating central post-stroke pain.

Corticospinal MEP (D-wave), cortico-muscular MEP in response to direct motor cortex stimulation, and subcortico-spinal MEP (subcortical D-wave) in response to subcortical pyramidal tract stimulation were monitored in patients with brain tumors around the motor cortex who showed no obvious motor disturbance before the operation. We could not establish the definite warning criteria of the cortico-muscular MEP, because disappeared cortico-muscular MEP reappeared during the operation. We concluded that cortico-muscular MEP is a safety range monitoring, and cortico-muscular MEP can guarantee the preserved motor function if no obvious change occurred during the tumor resection. On the other hand, monitoring of the D-wave enabled the function of the corticospinal tract to be evaluated selectively. A decrease of less than 30% may indicate postoperative preservation of motor function, including transient motor disturbance with subsequent complete recovery. Monitoring of the subcortical D-wave is useful to know the distance from the pyramidal tract. Intraoperative monitoring of D-wave and subcortical D-wave allows maximum resection of brain tumors located around the motor cortex.

We analyzed the features of neuronal activities, (a) firing rate, (b) firing pattern, and (c) response to passive joint movement, recorded from the globus pallidus internus (GPi) of patients with dystonia (DT) and Parkinson’s disease (PD). The spontaneous firing rate of GPi neurons in DT tended to be lower than that in PD. The reduced GPi activity resulted in disinhibition of thalamocortical pathways, which could cause DT as a form of hyperkinesia. However, identifying this reduced GPi activity is inconsistent with alleviation of DT in patients with pallidotomy. The pathology of DT, thus, cannot be explained solely by the reduced GPi activity. Quantitative examination of pauses and bursts from interspike intervals showed that pauses and bursts were significantly higher in DT than in PD. These abnormal firing patterns are conjectured to disrupt signaling along cerebello-thalamocortical pathways, impairing motor control and resulting in DT. Thus, abnormal firing patterns may represent temporal dysfunction. Changes in the GPi included enlargement of the receptive fields and an increased number of neurons exhibiting inhibitory reactions. DT symptoms may arise because of dysfunction of the mechanism inhibiting unrequired movements. The abnormal changes in the neurons in DT may underlie spatiotemporal dysfunction.

Spinal cord stimulation (SCS) is used to treat for several types of intractable pain syndromes, but it is not commonly used for treating post-stroke pain. Therefore, in this study, we analyzed the effects of SCS in patients with post-stroke pain. Neurological imaging and pharmacological analysis using thiopental, ketamine and morphine tests were conducted for 30 patients with post-stroke pain. For all patients, percutaneous electrodes were inserted under local anesthesia, and trial stimulation was performed for 5–7 days. If the patient experienced pain relief during the trial stimulation, a pulse generator was implanted. Of the 30 patients, 26 (87%) experienced satisfactory pain relief and underwent implantation. Ten patients received dual-octad lead SCS. The stimulation induced adequate paresthesia in all 10 patients. Nine (90%) of the 10 patients experienced satisfactory pain relief and opted for pulse generator implantation in order to proceed with the treatment. Upper cervical SCS was performed for patients with pain in not only the upper and lower extremities but also the face. Upper cervical SCS provided good paresthesia coverage and was effective for such patients. We think that patients with post-stroke pain may respond favorably to dual-octad lead SCS, and advancements in devices for SCS would further improve treatment outcomes for such patients.

It has been demonstrated in experimental animals that synchronous impulses occurring in efferent axons of corticospinal tract (CST) neurons in response to stimulation of the motor cortex can be recorded from the spinal cord1. The response consists of an initial D-wave and a later sequence of volleys termed I-waves. The D-wave reflects impulses arising from direct activation of the axons of CST neurons, and is induced most efficiently with anodal stimulation of the surface of the motor cortex.1.

During recording of spinal cord evoked potentials, a monophasic positive potential is observed when a nerve impulse approaches but does not pass beyond the recording site. This type of potential has been termed a killedend evoked potential (KP). While this potential is useful for localizing the injured site clinically, no other clinical value has yet been reported for it. The KP is usually higher in amplitude than the positive wave of normal evoked potentials. The higher amplitude has been attributed to background injury currents resulting from a low transmembrane resistance at the damage site of axons. The amplitude of the KP may therefore represent a clinically available indicator of injury currents, which have been implicated in secondary processes of axonal degeneration.

It has become increasingly clear that complete surgical excision is the best therapeutic approach for the majority of intramedullary spinal cord tumours. There is, however, always a risk in the radical excision of intramedullary spinal cord tumours that a greater postoperative neurological deficit may be inflicted. Extensive dissection can cause damage to the well-functioning cord tissue and/or the anterior spinal artery and penetrating branches. In order to minimize neurological deficits occurring as a result of such damage, the somatosensory evoked potential (SEP) has often been employed as an intraoperative monitor of cord function. The SEP is, however, not an ideal monitor since it reflects primarily the conductivity of the dorsal column. There are many cases in which SEP monitoring has missed motor deficits following damage to the corticospinal tract (CST), which is more devastating than any other neurological deficits1,2. These cases emphasize the need for a direct monitor specific to the CST integrity.

We reviewed 31 cases of acute thin subdural hematomas including those reported in the literature and our own which had disappeared spontaneously within 72 hours post-trauma. The subdural hematomas in these cases were detected mostly within 60 min after injury. A significant decrease in hematoma size was demonstrated at as early as 2 hours after injury. It seems therefore that the earlier the initial CT scan was taken, the more frequently acute thin subdural hematomas which rapidly disappeared could be detected. The hematomas did not explain clinical conditions on admission. Brain swelling and cardiorespiratory failure appeared to be responsible for poor outcome. Repeated CT scans and careful observation of the neurological changes is recommended before undertaking surgical intervention in such cases.

Since 1988, we have conservatively treated 90 cases of acute thin subdural hematomas (ATSDHs). These hematomas were less than 4 mm in width and disappeared spontaneously within 2 weeks in most cases. In 7 cases (7.8%), however, chronic subdural hematomas (CSDHs) developed. Craniotomy was performed in 3 cases and irrigation through a burr hole in the other 4 cases. We analyzed the characteristics of these cases of ATSDH in which CSDHs were later formed. The results indicated that (1) most of the cases were older than the remaining cases and had sustained minor head trauma; (2) cerebral atrophy was frequently evident on initial CT scans; (3) CT scans taken at 2 weeks after trauma demonstrated unresolved clots more frequently than in the other cases; (4) the development of voluminous CSDH was first identified on CT scans at as early as 3 weeks after trauma; (5) an external membrane, but not an inner membrane, was noted on surgery at approximately 3 weeks after trauma, implying that this form of CSDH may often develop quite rapidly; (6) the arachnoid membrane beneath the CSDH was frequently non-transparent and might have acted as an inner membrane; (7) the contents of the hematomas were identical to those observed in other CSDHs. These findings suggest that a large subdural space and low intracranial pressure may delay the resolution of ATSDH, and CSDHs may be rapidly formed from the unresolved clots. In addition, it appeared that CT scans taken at 2 weeks after trauma could be important for predicting subsequent development of CSDH.

It has become increasingly clear that the acute thin subdural hematomas detected in the very early period following trauma can disappear spontaneously within a day [1–6]. We have experienced 10 cases of such hematomas since our initial reports [3]. It seems that the earlier the initial CT scan is taken, the more frequently such hematomas are encountered.

through: (1) Somatosensory evoked potential phase reversal technique; (2) DCS with a short train of 5 7 monopolar stimuli (0.5ms duration, ISI 4.1, intensity up to 20mA) at 1 2Hz. To monitor motor pathways during tumor removal, MEPs are recorded from controlateral limb muscles after either DCS and/or transcranial electrical stimulation. When approaching tumor borders, MEP monitoring is combined with periodical direct subcortical stimulation (DSS) to localize the corticospinal tract (CT) and guide resection.A more than 75% drop in transcranial MEP amplitude at the end of surgery, and a DSS thresholds lower than 3 4mA are associated to post-operative neurological worsening. However these two neurophysiological warning signs are not significantly related and do not necessarily identify the same patients, suggesting that continuous MEP monitoring and periodical subcortical mapping should be combined to minimize risks. Recently, thanks to the introduction of diffusion tensor imaging and neuronavigation, there has been an increasing interest for the correlation of fiber tractography with subcortical stimulation. This work is providing some degree of reliability in the assessment of subcortical current spreading. Preliminary data suggest that 1mA = 1mm may be an acceptable practical parameter to judge on the distance from the CT, when using subcortical mapping.

A series of eight patients in a persistent vegetative state (PVS) were subjected to chronic deep brain stimulation (DBS) for the purpose of promoting recovery from the PVS. The characteristics of the brain activity in these patients were evaluated from the late positive component of the cerebral evoked potential in response to painful stimuli (pain-related P250). While any neurological scoring system for the comatose state includes evaluations of motor reactions to painful stimuli, the pain-related P250 is unique in terms of its ability to assess the cortical responsiveness to painful stimuli directly and quantitatively without involving functions of the motor system. It was found that the pain-related P250 was more or less depressed in patients in a PVS. It was repeatedly demonstrated in four patients, however, that the pain-related P250 could be transiently increased by preceding stimulation of the mesencephalic reticular formation. Furthermore, a persistent increase in the pain-related P250 was produced in these four patients following chronic DBS of the mesencephalic reticular formation or nonspecific thalamic nuclei for more than 6 months, and this was correlated with the clinical improvements. These results imply that responsiveness at the cortical level to pain is depressed in the PVS. It also appears that some fraction of the depression may, however, be functionally produced and potentially reversible.

It has been demonstrated that a corticospinal response to direct stimulation of the motor cortex can be recorded from the lateral column of the spinal cord or the spinal epidural space [4, 5, 9, 10]. It has also been reported that identical responses can be recorded in cats and humans using transcranial brain stimulation [6, 7]. A reliable technique of transcranial brain stimulation would be of great value for monitoring pyramidal tract function, since this method does not require any operation to be performed in order to stimulate the motor cortex directly. However, there is some doubt as to whether spinal cord responses to transcranial brain stimulation really represent impulses mediated by pyramidal neurons. We previously compared the corticospinal D response to stimulation of exposed motor cortex and the spinal cord responses to transcranial brain stimulation in cats [4], and we found that spinal cord responses to transcranial brain stimulation are easily confused with responses other than the corticospinal D response.

A monophasic positive potential is recorded when a nerve impulse approaches but never passes beyond the recording site. This type of potential has been termed the killed end potential (KP). Studies in experimental animals have repeatedly suggested a clinical value for the KP in localizing the level of spinal cord lesions [3, 4, 9]. The KP has been recorded intraoperatively in patients with spinal cord lesions [7, 10, 11]. We have also demonstrated that the KP of spinospinal responses can be readily recorded preoperatively in patients with eletrodes inserted percutaneously into the spinal epidural space and indeed has a value for localizing lesions [5].

In an attempt to facilitate recovery from prolonged coma, deep brain stimulation has been applied to the pallidum, the ventroanterior thalamus, the midbrain reticular formation, and the thalamic intralaminar nucleus [1–4]. This approach is based on the anatomical and electrophysiological finding that brain arousal activity might be produced by stimulation of the ascending activating system [5,6]. Deep brain stimulation might also inhibit an active neural process within the brain stem that is involved in the production of coma [7]. However clinical results of deep brain stimulation have been quite variable and long-term follow-up results have not yet been reported [1–4, 8].

All forms of therapy, including chronic stimulation of the thalamic relay nucleus, can provide satisfactory pain control in only 20%-30% of cases of thalamic pain syndrome. In order to develop a more effective treatment for thalamic pain syndrome, we investigated the effects of stimulation of various brain regions on the burst hyperactivity of thalamic neurons recorded in cats after deafferentiation of the spinothalamic pathway. Complete, long-term inhibition of the burst hyperactivity was induced by stimulation of the motor cortex. Based on this experimental finding, we treated seven cases of thalamic pain syndrome by chronic motor cortex stimulation employing epidural plate electrodes. Excellent or good pain control was obtained in all cases without any complications or side effects. During the stimulation, an increase in regional blood flow of the cerebral cortex and thalamus, a marked rise in temperature of the painful skin regions, and improved movements of the painful limbs were observed. These results suggest that thalamic pain syndrome can be most effectively treated by chronic motor cortex stimulation.

Twelve patients with deafferentation pain secondary to central nervous system lesions were subjected to chronic motor cortex stimulation. The motor cortex was mapped as carefully as possible and the electrode was placed in the region where muscle twitch of painful area can be observed with the lowest threshold. 5 of the 12 patients reported complete absence of previous pain with intermittent stimulation at 1 year following the initiation of this therapy. Improvements in hemiparesis was also observed in most of these patients. The pain of these patients was typically barbiturate-sensitive and morphine-resistant. Another 3 patients had some degree of residual pain but considerable reduction of pain was still obtained by stimulation. Thus, 8 of the 12 patients (67%) had continued effect of this therapy after 1 year. In 3 patients, revisions of the electrode placement were needed because stimulation became incapable of inducing muscle twitch even with higher stimulation intensity. The effect of stimulation on pain and capability of producing muscle twitch disappeared simultaneously in these cases and the effect reappeared after the revisions, indicating that appropriate stimulation of the motor cortex is definitely necessary for obtaining satisfactory pain control in these patients. None of the patients subjected to this therapy developed neither observable nor electroencephalographic seizure activity.

In nine patients with acute cervical spinal cord injury, corticospinal motor evoked potentials (MEPs) elicited by either epidural stimulation at the motor cortex or transcranial stimulation were recorded during acute periods in order to obtain information concerning motor function, in comparison with other evoked potentials (SSEPs, somatosensory evoked potentials and SESPs, spinally evoked spinal cord potentials). The courses of acute cervical spinal cord injury were classified into four patterns based on the results for the multimodality evoked potentials. The first pattern, which failed to show any activity in SSEPs, SESPs, or corticospinal MEPs throughout the course of sequential recordings for 4 days, was associated with a very poor prognosis. Patients with the second pattern, which showed recovery of detectable activity only in SESPs, did not undergo improvement of paraplegia. Those with the third pattern, which showed recovery of detectable activity in corticospinal MEPs as well as SESPs, eventually acquired a capacity to walk. Those with the fourth pattern, which showed detectable activity in SSEPs, SESPs, and corticospinal MEPs, had a good prognosis and their paraplegia improved within a week. These results suggest that recordings of multi-modality evoked potentials can be useful for determining the prognosis and severity of acute cervical spinal cord injury.

Thirty-one prolonged coma patients, who continued in a comatose state for at least 2 months, were classified electrophysiologically employing EEG, a compressed spectral array of continuous EEG frequency analysis, BSR, SEP and SSEP. The prognoses of long-term follow-up (at least 8 months) in these patients were compared with the results of such electrophysiological analysis and with the neurological gradings of the prolonged coma patients. In the continuous EEG frequency analysis, 22 cases were classified as having a changeable spectrogram, nine cases as having a slow monotonous spectrogram, and no cases as having a borderline spectrogram. We also classified the changeable spectrograms into the following three patterns: no desynchronization, slight desynchronization, and desynchronization. The nine cases with a slow monotonous spectrogram revealed severe damage of the brain stem and cerebral function with multimodality evoked potentials, and most of these cases could not survive without assisted respiration. The 22 cases with a changeable spectrogram had mild damage of the brain stem and cerebral function with multimodality evoked potentials, and these cases were identical with a persistent vegetative state. Three of the cases showed a changeable spectrogram that revealed a desynchronization pattern, normal BSR and normal N20 of SEP and SSEP, recovered spontaneously from the persistent vegetative state within 6 months after electrophysiological evaluation. We conclude that not only a neurological but also an electrophysiological classification is necessary to evaluate the brain function and prognosis of prolonged coma.

Corticomyographic motor evoked potentials (MEP) activated by transcranial magnetic stimulation of the motor cortex provide clinicians with an opportunity to evaluate corticospinal motor systems quantitatively and noninvasively. Threshold, amplitude, and latency of the corticomyographic MEP, however, are variable between subjects mainly because current directions and intensities induced by magnetic stimulation cannot be determined precisely due to anatomical variations of subjects. The variability of corticomyographic MEPs has limited the use of corticomyographic MEP for evaluating mild changes in corticospinal motor function. In the present study, we used an internal standard to assess hemiplegia, expressing relative amplitude, latency, and threshold of responses on the paretic side as a function of responses elicited from the intact side (%MEP). Neurological function of paretic muscles, as determined by a muscle maneuver test (MMT), clearly correlated to %MEP threshold, amplitude, and latency. Since corticomyographic MEP are similar when recorded from symmetrical sites on two extremities of normal subjects, %MEP provided a sensitive measure of mild hemiparesis. The %MEP approach revealed abnormal MMT scores of 3 or 4 more frequently than did standard MEP approaches. %MEP amplitude was more sensitive to mild hemiparesis than %MEP latency or %MEP threshold. Since magnetic stimulation with a safe intensity range cannot reliably produce corticomyographic MEP in severely paretic muscles with MMT scores of 2 or less, the MEP appears to be most useful for evaluating mild hemiparesis. This technique should expand significantly the clinical usefulness of corticomyographic MEP in neurosurgical practice.

To clarify the mechanism of the distant electron nuclear double resonance effect, ruby Al2O3 (Cr3+) was studied by ENDOR technique in the V-band (50 GHz band). The most prominent ENDOR effect was observed at fields a little off the resonance field of electron spin resonance (ESR). To interpret this phenomenon, we propose the spin temperature model which deals with thermal equilibrium between the electron spin system and the nuclear spin system on the frame rotating at the Larmor frequency of the electron. Below the resonance field, the ENDOR effect caused by the energy flow from the nuclear spin system to the electron spin system is most prominent when the electron spin temperature is the lowest compared with the nuclear spin temperature. In the higher magnetic field than the resonance field where the electron spin temperature is negative, the ENDOR effect is caused by the electron spin deexcitation induced by the fluctuations of the effective field.

Ectopic intracranial retinoblastomas arc rare. These tumors usually occur in the pineal, parasellar. or suprasellar regions several years after the successful treatment of ocular retinoblastomas with no evidence of direct extension or distant metastasis. We report here a case of ectopic retinoblastoma occurring within the third ventricle. The tumor was surgically excised by the transventricular approach. Ectopic retinoblastomas exhibit greater differentiation than one would expect to observe in a metastatic lesion of this tumor. The distinction of ectopic retinoblastomas and metastasis from ocular retinoblastomas is important, since ectopic retinoblastomas, unlike metastasis, can be successfully managed by intensive therapy including radical excision.

The modulatory effects of diencephalic stimulation on the activities of raphe-spinal neurons were studied extracellularly in cats. Among 240 raphe neurons recorded, 57 neurons were activated antidromically by stimulation of the cervical dorsolateral funiculus. These raphe-spinal neurons were found in the caudal raphe nuclei, i.e., the raphe magnus (43 neurons), raphe obscurus (11), raphe pallidus (2), and raphe pontis (1). All of them responded to innocuous and/or noxious peripheral mechanical stimuli with a broad receptive field. The activities of the majority of these neurons were facilitated by trains of pulse stimulation of the rostral periaqueductal gray and the thalamic relay nucleus but not of the thalamic center median nucleus. The facilitation of firing persisted for more than 3 min after the cessation of train pulse stimulation when the stimulation was applied at 20 Hz for 5 to 30 s. This facilitation was not affected by decortication of the sensorimotor area bilaterally. The facilitatory response to periaqueductal gray stimulation was markedly suppressed by systemic administration of naloxone. On the other hand, that of the thalamic relay nucleus stimulation was found to be unaffected. Based on these findings, the mechanisms of pain relief by stimulation of the rostral periaqueductal gray and thalamic relay nucleus reported in human intractable pain appear to relate, at least partly, to the activation of raphe-spinal neurons. However, the paths to raphe-spinal neurons of stimuli from the periaqueductal gray and the thalamic relay nucleus are thought to be independent from each other based on the different effects of naloxone.

Deep brain stimulation (thalamic relay nucleus, periaqueductal gray and internal capsule) was applied to various cases of intractable pain, and the resulting degree of pain reduction and alteration in β-endorphin immunoreactivity in the cerebrospinal fluid (CSF) were compared. The following results were obtained. 1. (1) The studies on intractable pain revealed that the levels of β-endorphin immunoreactivity in the CSF were lower than those in the control group. 2. (2) Thalamic relay nucleus stimulation proved effective not only for deafferentation pain, but also for somatogenic pain. No relationship was, however, noted between pain reduction and the rate of increase of β-endorphin immunoreactivity in the CSF. 3. (3) The incidence of stimulation tolerance following prolonged stimulation of the thalamic relay nucleus can be reduced to a minimum by administration of l -DOPA. It is concluded that the increase in β-endorphin in the CSF is not the direct and major cause of pain reduction during treatment by thalamic relay nucleus stimulation. It may be assumed that neuronal facilitation on the monoaminergic descending pain inhibitory system plays a role in reducing pain.