The terrible disability inflicted by primary and secondary dystonia inspired surgical intervention in attempts to modify the associated movement disorders that strike previously normal patients out of the blue in the case of primary dystonia, or that are superimposed on a pre-existing neurological deficit in the case of secondary dystonia.

OBJECTIVETo describe the outcomes in our first 40 microelectrode-guided thalamotomies for parkinsonian tremor.METHODSTwenty-four left-sided and 16 right-sided thalamotomies were performed between October 1984 and January 1996; the mean follow-up period was 35.8 months (range, 1-152 mo). The results were evaluated retrospectively and semiquantitatively by a disinterested observer (MNL) and correlated with the quality of the microelectrode recording and the number and size of radiofrequency lesions made. The first 20 and second 20 procedures were evaluated separately.RESULTSAt the last follow-up, the Unified Parkinson's Disease Rating Scale showed no or virtually no tremor in the upper limb in 75% of patients or in the lower limb in 73% of patients. No significant persistent complications were found. These results were achieved at the expense of having to repeat the procedure on 11 sides (in 5 because of technical problems and in 6 for no obvious reason). Total or nearly total abolition of tremor occurred after the first procedure in 40% of the first 20 operations and in 65% of the second 20. Eight of the first 20 procedures and 2 of the second 20 failed for technical reasons. Lesions were made larger in the second 20 procedures than in the first 20. With the use of an electrode with a 1.1 x 3-mm bare tip for 60 seconds, it seems that lesions had to be created at 60 degrees C or more to produce a successful result.CONCLUSIONThalamotomy with microelectrode recording is an effective procedure with which to treat tremor in patients with Parkinson's disease and may involve fewer complications than conventional techniques. The procedure appears to involve a learning curve.

Many amputees have a sense of their missing ‘phantom’ limb1,2,3. Amputation can alter the representation of the body's surface in the cerebral cortex4,5,6,7,8,9,10,11,12,13,14 and thalamus15,16, but it is unclear how these changes relate to such phantom sensations. One possibility is that, in amputees who experience phantom sensations, the region of the thalamus that originally represented the missing limb remains functional and can give rise to phantom sensations even when some thalamic ‘limb’ neurons begin to respond to stimulation of other body regions. Here we use microelectrode recording and microstimulation during functional stereotactic mapping of the ventrocaudal thalamus in amputees to determine both the responses of the neurons to stimulation of the skin and the perceptual effects of electrical activation of these neurons. Thalamic mapping revealed an unusually large thalamic stump representation, consistent with the findings from animal experiments. We also found that thalamic stimulation in amputees with a phantom limb could evoke phantom sensations, including pain, even in regions containing neurons responsive to tactile stimulation of the stump. These findings support the hypothesis that the thalamic representation of the amputated limb remains functional in amputees with phantoms.

I t is particularly gratifying to review the advances in knowledge concerning neuropathic pain syndromes over the last 20 years and Bud Craig's article is a further attempt to define the anatomic and physiologic substrate for one of the most puzzling and distressing pain syndromes, that induced by stroke. He provides evidence to suggest that stroke-induced pain results from selective loss of thermal sensibility pathways, taking with them the inhibition of pain induced by cold and the disinhibition of cold-evoked burning pain. That is, the lateral lamina I spinothalamocortical pathway to parietal insular cortex is disrupted and the polymodal nociceptive activity in the medial lamina I spinothalarnocortical pathway to anterior cingulate cortex is disinhibited. This proposal is supported by the observations of multiple observers that the critical feature seen in strokeinduced pain is loss of pain and temperature sensing. And Craig also reviews evidence for cold-induced inhibition of pain. As a clinician who has been interested in the neuropathic pain syndromes and specifically stroke-induced pain, I have reviewed Bud Craig's paper with great interest, but I have taken the track of studying the pain in the light of clinical characteristics. I am not so convinced as he is that the quality of pain described by stroke patients is all that different from the quality of pain described by patients with neuropathic pain of peripheral or cord origin, all of whom share three common elements in varying degrees: causalgic, dysesthetic, spontaneous pain; neuralgia-like pain; and evoked pain meaning allodynia and hyperpathia. This has always raised in my mind the possibility that the mechanisms might be similar

Studies of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced parkinsonism in monkeys 1,2 suggest that excessive inhibitory outflow from the internal segment of the globus pallidus (GPi) suppresses the motor thalamus, which reduces activation of the cerebral cortex motor system, resulting in the slowness and poverty of movement of Parkinson's disease (PD) 3 . This hypothesis is supported by reports of high rates of spontaneous neuronal discharges and hypermetabolism in GPi (ref. 4–7) and impaired activation of the supplementary motor area (SMA) and dorsolateral prefrontal regions 8,9 in PD patients. Furthermore, lesion or chronic high-frequency electrical (likely inactivating) stimulation of GPi (ref. 10–14) is associated with marked improvements in akinesia and rigidity, and the impaired activation of SMA is reversed when the akinesia is treated with dopamine agonists 15 . To test whether improvement in motor function with pallidal surgery can be attributed to increased activity in premotor cortical regions, we assessed the changes in regional cerebral blood flow (rCBF) and parkinsonian symptoms during disruption of GPi activity with high-frequency stimulation delivered through implanted brain electrodes. Positron emission tomography (PET) revealed an increase in rCBF in ipsi-lateral premotor cortical areas during GPi stimulation, which improved rigidity and bradykinesia. These results suggest that disrupting the excessive inhibitory output of the basal ganglia reverses parkinsonism, via a thalamic relay, by activation of brain areas involved in the initiation of movement.

Deep brain stimulation (DBS) and thalamotomy are both capable of abolishing tremor. However, no technique is perfect and if thalamotomy proves inadequate so that tremor recurs, presumably because of suboptimal lesion location, the only option is to repeat the thalamotomy. With DBS all that has been necessary to date is to change the parameters of stimulation. Similarly with complications such as the ''cerebellar'' ones and paraesthesiae. If these occur after thalamotomy one can only wait and hope that they will subside and they do not always do so. With DBS, changing the parameters in the authors' patients has so far been successful in eliminating them.DBS, like thalamotomy is very effective for controlling tremor in Parkinson's disease (PD) and essential tremor (ET) and for improving dexterity in ET, but both techniques are less useful for the control of dopa dyskinesia, Parkinsonian rigidity, or impaired dexterity in PD, though DBS may be better than thalamotomy for the latter condition. On the other hand, both DBS and thalamotomy are very effective in improving dexterity in PD and ET may depend upon the fact that in PD bradykinesia is a major component, whereas in ET only the tremor is. The advantages of DBS over thalamotomy have to be weighed against the peculiar risks of DBS and of course, its cost.