Early surgical treatments for movement disorders date to the late nineteenth century. Victor Horseley was one such early innovator and used cortical motor strip resection to address athetosis and tremor (Gabriel and Nashold 1998). This approach proved undesirable, as interruption of the pyramidal pathways caused weakness and intolerable long-term side effects. Later, in the 1930s and 1940s, partial cordectomies were introduced to reduce tremor (Walker 1949), and cerebral pedunculotomies were utilized for the relief of choreoathetosis and hemiballismus. Morbidity was again unacceptably high and included such problems as motor weakness, loss of pain and temperature sensation, impairment of sexual function, and urinary incontinence (Putnam 1938; 1940; Brown and Walsh 1954). By the 1950s, physicians had largely abandoned pyramidal tract lesioning and researchers were turning to new techniques.
Attempts were next made to address basal ganglia structures by directly targeting them. Meyer, who earlier reported on anterior caudate resections as part of his 1940 series on the treatment of postencephalitic tremor (Meyers 1940), was an early proponent of more targeted lesions as a surgical approach. The transventricular removal of the anterior two thirds of the caudate in one of his patients abolished her tremor, a target suggested by a previous observation of Browder’s during a lobectomy (Kopell et al 2006). Long-term follow-up of this caudate resection revealed that the patient had developed unacceptable hyperkinetic movements. The open transventricular approach was abandoned (Speelman and Bosch 1998).
Perhaps the greatest impact on the field was with the introduction of frame-based surgery. Spiegel and Wycis introduced stereotactically guided surgery in the late 1940s (Spiegel et al 1947; Gildenberg 2001). With this tool, mortality decreased, especially when compared to open craniotomy procedures (Tasker 2004).
Two decades following the introduction of the head frame, the literature was expanded by the further results of Spiegel and Wycis as well as those of Hassler, Cooper, and others. These detailed 3 potential surgical targets for the treatment of Parkinsonian tremor: the midbrain peduncles (Guiot and Pecker 1949; Walker 1952) and tegmentum (Siris 1942; Wycis and Spiegel 1952); the thalamus (Cooper et al 1958); and the pallidofugal pathways (Meyers 1942; 1951). Later in the same decade, JL Pool attempted to treat a patient’s depression by implanting an electrode into her caudate nucleus but did not follow up with a detailed pre- and postoperative analysis (Pool 1954).
Irving Cooper and Rolf Hassler introduced neurosurgical targeting based on the given brain area’s anatomy and functional connectivity (Das et al 1998). Part of this rationale emerged after Cooper accidentally ligated the anterior choroidal artery of a patient with Parkinson disease and observed a dramatic improvement in the patient’s motor symptoms (Cooper 1954b). Following similar observations, Cooper introduced reversible “chemotomy” of the pallidum as a preoperative screening tool. In this procedure, he temporarily “deactivated” the globus pallidus by injecting procaine prior to resection of the anterior choroidal artery (Cooper 1954a; 1956). Cooper later added a double lumen catheter to this technique and also applied it to targets such as the thalamus. He was able to visualize the target area by injecting dye or alternatively was able to focally destroy an area by alcohol injection or cryotherapy (Cooper 1955; Hardy and Bertrand 1965). Radiofrequency sampling and microelectrode stimulation were later introduced as methods of functional localization and aided dramatically in defining target accuracy (Andy et al 1963; Yamashiro and Tasker 1990). Heat, electricity, ultrasound, and focal gamma radiation would all be employed in coming decades (Meyers 1958; Meyers et al 1959; Thurel et al 1961; Duma et al 1998; Okun et al 2001).
Inspired by the work of Hassler and Reichert in the 1950s and 1960s, a number of surgeons began to ablate the ventrolateral thalamus at the nucleus ventralis intermedius and ventralis oralis anterior/ventralis oralis posterior nuclei (VOA-VOP) to relieve tremor and rigidity and to address more complex symptoms such as choreoathetosis, hemiballismus, and the cardinal symptoms of Parkinson disease (Hassler et al 1960; Redfern 1989). Hassler observed that stimulation of the pallidum could elicit dystonia at low frequencies whereas, conversely, he observed improvement at higher stimulation frequencies (Vercueil et al 2002). As smaller and better-placed surgical lesions became more feasible, new targets emerged including the centromedian nucleus, the posterior limb of the internal capsule, the subthalamic nucleus, and deep cerebellar nuclei (Markham and Rand 1963; Zervas et al 1968; Yamashiro and Tasker 1990). The better localization of old targets also improved overall outcome.
Though lesioning was effective, problems were increasingly evident with this approach. The technique was irreversible and produced static side effects and frequent morbidity, especially when lesions were applied bilaterally (Ghika et al 1999; Favre et al 2000; Higuchi and Iacono 2003). Surgery fell out of favor with the introduction of levodopa therapy for Parkinson disease in the late 1960s. Surgical therapy would re-emerge in the early 1980s after it was realized that long-term exposure to levodopa could result in debilitating on-off fluctuations and dyskinesia (Narabayashi et al 1984).
The modern era of ablation and stimulation is considered by many to have begun following Benabid’s observation that high frequency stimulation during lesion localization resulted in a reduction of Parkinson disease tremor (Benabid et al 1987; 1991). His finding that prolonged or chronic thalamic stimulation could persistently suppress tremor resulted in a new technique referred to as chronic deep brain stimulation (DBS). It was then adapted and applied as a treatment for movement disorders and pain. The implantation of deep brain stimulators provided an alternative to lesioning that was reversible, adjustable, and allowed for bilateral treatment, without the degree of cognitive and mood effects seen with stereotactic lesioning (Schuurman et al 2000; Pahwa et al 2001; Bittar et al 2005).