The Apparent Impunity of the Basal Ganglia to Therapeutic Lesioning: Clinical and Scientific Lessons

The field of Movement Disorders has benefited substantially from intense clinical and experimental interaction, which has expanded our understanding of motor control and motor manifestations associated with basal ganglia dysfunction. The classic model of the basal ganglia effectively predicted some clinical manifestations and therapeutic responses. However, models failed to explain the bulk of well‐rooted clinical observations associated with focal lesioning of the basal ganglia. ¹

Classically, the basal ganglia have been considered essential to promote acquisition and execution of automatic (habitual) movements. ² This is thought to operate essentially by allowing/facilitating the expression of selected cortically driven actions, ² , ³ , ⁴ mainly in response to recognizable stimuli, while suppressing other unwanted and less valuable stimuli for a given circumstance. ⁵ The motor areas and basal ganglia constitute the basis for the system which sustains such habitual or automatic activities. The basal ganglia are also activated to interrupt ongoing actions and switch priorities to accommodate unexpected novel and more powerful stimuli. ¹ , ⁵ More recently, the same basic concept has been extended to facilitate and mediate behavioral and emotional habits as well as habitual inhibition. ⁶

Therefore, the basal ganglia are thought to play an essential role in two basic neural processes that are intimately linked to normal and successful behaviour: the selection of appropriate actions and reinforcement learning. ⁷ , ⁸ Indeed, primates, particularly humans, have experienced major development of the basal ganglia along with the expansion of the cortex. ⁹ Thus, it has been proposed that physiologically, the basal ganglia are the phylogenetic solution to the problem of selectionism, since the cortex can only attend to one given event at a time. ⁹ Accordingly, how can such a complex and refined set of nuclei and connections be eliminated without causing deficits?

In this Viewpoint, we recapitulate and discuss the apparent impunity of the basal ganglia to lesioning particularly those made for therapeutic purposes in patients with movement disorders. We will focus on the lessons derived from non‐human primates and humans with a focal blockade or lesioning of basal ganglia structures, mainly those derived from functional neurosurgery for the treatment of Parkinson's disease (PD). The analysis is based on treatment of movement disorders with stereotaxic radiofrequency‐mediated lesions or, nowadays, with focused ultrasound thermoablation (FUS), as the observations are relatively well controlled and sufficiently large in number. ¹⁰ The dissociation between such overt clinical benefit and the seeming lack of meaningful deficit after interrupting key subcortical circuits is a long‐standing puzzle. ¹ , ¹¹

The interest in this topic is two‐fold. On the one hand, it is clinically relevant, as ablations are made with therapeutic purposes; but potential deficits are essential to consider. On the other hand, it is scientifically sound, because eliminating major output nuclei and interrupting critical circuits should reveal fundamental features about the normal function of the basal ganglia. Currently, the significant therapeutic benefit associated with focal ablation for movement disorders would suggest, as the most straightforward conclusion, that normal operation of the primate brain can take place without the aid of the basal ganglia.

Clinically, the lack of major clinical deficits with therapeutic lesions for PD contrast with the remarkable manifestations associated with dopaminergic deficits in neurodegerative disorders or with focal striatopallidal lesions (mainly putamen and globus pallidus pars interna [GPi]) causing contralateral dystonia ¹² and profound neurological manifestations of bilateral acute pallidal lesions. ¹³ , ¹⁴ Additionally, it contrasts scientifically with the bulk of data indicating the significant role of the basal ganglia in motor, behavioral and emotional domains.

Clinical Lessons

Impunity to Ablation

The therapeutic use of focal ablation of the basal ganglia dates back decades. The initial experience was blurred by methodological (ie, targeting and others) limitations and poor patient selection. In the modern era, functional neurosurgery was the revitalized via the development of the pathophysiological model of the basal ganglia and the discovery of the significant role of the subthalamic nucleus (STN) in the parkinsonian state, stemming from Crossman's lab. ¹⁵ , ¹⁶ These initial studies demonstrated experimentally that the STN's chemical ablation (fiber sparing) in the N‐methyl‐4‐phenyl‐1,2,3,6‐tetrahydropyridine (MPTP) non‐human primate was associated with a rapid and substantial reduction of akinesia/bradykinesia, rigidity, and postural tremor, which were accompanied by mild to moderate hemichorea. ¹⁷ , ¹⁸ , ¹⁹ Motor improvement was associated to a reduction of abnormal firing in the GPi ²⁰ and biochemical markers of neuronal activity in basal ganglia output nuclei. ¹⁹ These results, and the independent experience of Laitenen with pallidotomy for PD published at the same time, ²¹ led to the resurgence of functional neurosurgery for movement disorders.

Currently, it is known that lesions of the STN and GPi as well as the ventrolateral nucleus of the thalamus (VLo‐pallidal receiving area) as carried out with FUS are therapeutically beneficial with no obvious clinical deficits, aside from possible adverse events related to erroneous targeting. Similar to the effect of deep brain stimulation (DBS), the clinical benefit from subthalamotomy, pallidotomy and thalamotomy on PD motor features can be sustained in the long term. ²² , ²³ Certainly, the clinical response is somewhat variable, mainly related to specific individual factors (eg, the accuracy and precision and total volume of the lesion) but also differences in disease progression. Whether or not early reduction of neuronal pathological activity impacts disease progression is still unclear.

The clinical tolerance to the lesion is somewhat counterintuitive from the viewpoint of classic experimental neurology, which has been based on focal lesions (ie, a stroke or tumor) leading to specific clinical manifestations, and establishing precise clinicopathological correlations. ²⁴ , ²⁵ PD patients may tolerate lesioning, because the parkinsonian brain has adapted itself to operate, albeit dysfunctionally, without the aid of the habitual motor circuit. ¹ , ⁷ Activation and reliance upon basal ganglia activity occur with conscious, goal‐directed actions, which are mediated by the prefrontal cortex and rostral striatopallidal connections. Notably, dopaminergic nigrostriatal depletion occurs earliest and foremost in the caudal putamen, ²⁶ which sustains habitual motor behavior. In contrast, the rostral striatum is relatively spared and, accordingly, allows regular engagement of the goal‐directed system. The recruitment of the latter is a possible compensatory mechanism for movement in the parkinsonian state (3). However, patients exhibit clear parkinsonian signs and impaired movement parameters at baseline, indicating that compensation via recruiting different circuits, if it occurs, is insufficient. In addition, therapeutic ablations restore movement capacity to normality or near normality within minutes of impacting the appropriate target, which argues against a compensatory mechanism via recruiting different circuities. Instead, it suggests that eliminating the abnormal neuronal activity frees the motor circuit to operate more normally. Therefore, explaining the lack of significant deficits after lesioning basal ganglia output as being due to the use of alternative cerebral pathways is highly unlikely. On the other hand, compensation could also occur by relying on basal ganglia activity in the contralateral hemisphere. However, a recent monkey experiment to be discussed in the next section also makes this possibility improbable.

Deficits can be Detected

Despite the apparent impunity of the basal ganglia to focal lesions, failures in motor performance do become evident when patients are challenged with more sophisticated and demanding tasks. ²⁷ , ²⁸ , ²⁹ For example, Obeso et al described the motor performance of a man with PD in whom thermolytic lesions, first of the left STN to improve parkinsonism and after that of left GPi to treat subthalamotomy‐induced hemichorea‐ballism interrupted the basal ganglia‐thalamo‐cortical motor circuit in the left hemisphere, which in the off‐medication condition was essentially deprived of neuronal activity throughout the cortico‐basal ganglia motor loop. In this patient, they found abnormalities in learning a new task with the right hand. Implicit sequential learning in a probabilistic serial reaction time task was absent, and there was loss of reaction time superiority with the right hand in a go/no‐go task. ²⁷ Additionally, in PD patients with therapeutic subthalamotomy, it has been shown in a classic stop‐signal task that inhibitory control is impaired, as revealed by anticipation in the response, increased number of discrimination errors, and reduction in the response thresholds relative to unoperated patients with PD. ²⁹ In addition, these patients also show an increment in reactive inhibitory control, proactive inhibition, and conflict resolution relative to unoperated PD patients. All these inhibitory defects were particularly noticeable for patients treated with right hemisphere STN lesion. ²⁹ Noteworthy, these anomalies were all sub‐clinical and not evident in routine daily activity (5). These deficits are best explained by a failure of the cortex, deprived of basal ganglia input, to facilitate responses in a probabilistic context. Therefore, the basal ganglia are needed to learn and move normally under some novel circumstances.

Similarly, subtle non‐motor symptoms are encountered in PD patients treated with STN and GPi focal lesions. In that sense, unilateral right subthalamotomy is associated with a greater risk of impulsivity and disinhibition, while left subthalamotomy induces further impairment of semantic verbal fluency. ²⁹ Importantly, in PD, unilateral pallidotomy is associated with some impairment of verbal fluency by diminishing adequate and fluent word production, ³⁰ and bilateral pallidotomy causes serious complications, including severe speech disturbance;¹⁴ the latter has also been described with bilateral pallidal ischemic lesions in previously healthy subjects. ¹² Indeed, since speech is a highly automatic activity, it should be highly sensitive to basal ganglia dysfunction but has yet to receive major attention from this standpoint.

Scientific‐Experimental Lessons

According to the classic “firing rate model” of the basal ganglia, dopamine transmission modulates activity in the competing direct and indirect striatopallidal projections, which determine GPi/substantia nigra pars reticulata (SNr) neuronal output. This in turn suppresses or releases basal ganglia recipient motor thalamic regions and the superior colliculus to facilitate movement. ³¹ , ³² , ³³ Accordingly, marked reduction (ie, after blockade or lesioning) in GPi/SNr activity should be associated with exaggerated uncontrolled selection of movement. ³⁴ The critical observation here is that ablation of the GPi/SNr does not evoke dyskinetic movements. This has been the case when lesioning has been applied to normal rats, in monkeys with experimental STN lesioning, and in previously healthy humans with hemichorea‐ballism after stroke. ³⁵ , ³⁶ , ³⁷ In all these instances effective interruption of basal ganglia output did not evoke any abnormal release of movements and behavior. Similarly, PD patients treated with pallidotomy exhibit major amelioration or even abolition, not an increase, of levodopa‐induced dyskinesias and off‐dystonic postures. ³⁸ All these results, which the basal ganglia model did not predict, undoubtedly indicate that basal ganglia output and functional effects are not only signaled and conveyed via neuronal excitation/inhibition activity. ³⁹

Therefore, the firing rate has to be understood along with other physiological features, such as abnormal burst firing and coupling of oscillatory activity (mainly beta and theta band) which are known to be abnormal in the parkinsonian state. ⁴⁰ , ⁴¹ Noteworthy, STN‐DBS and GPi‐DBS showed similar effects on parkinsonian motor signs ⁴² , ⁴³ , ⁴⁴ to STN and GPi lesioning/blockade. ¹⁷ , ¹⁸ , ⁴⁵ The ablation‐lesion undoubtedly produces a drastic reduction in neuronal activity locally. Such reduction of the firing rates of abnormally increased firings and interruption of abnormal firing patterns in the STN and GPi are key mechanisms in the observed amelioration of parkinsonism.

On the other hand, variability in the anti‐parkinsonian effect has been historically recognized after lesioning either the STN or GPi ¹⁴ , ⁴⁶ and also with DBS. ⁴⁷ , ⁴⁸ In this scenario, the functional anatomy of the basal ganglia is probably fundamental to understanding some of such variability in clinical improvement observed with neurosurgical treatments for PD. Similarly, delayed responses, as described in some instances after pallidotomy ⁴⁹ and GPi‐DBS in PD patients and particularly after GPi‐DBS in dystonia, suggest a role for the functional reorganization of BG‐cortical circuites. ⁵⁰ , ⁵¹

Nevertheless, acting on all these different aspects of neuronal activity does not explain either the lack/elimination of dyskinesias or the excellent tolerance to lesioning observed in patients. Two very revealing monkey experiments probably serve best to provide an understanding of the paradox discussed here. First, Anderson's lab showed many years ago that acutely blocking GPi neuronal activity did not modify neuronal firing activity in the pallidothalamic receiving areas of the ventro‐lateral thalamus when performing a previously learned task. ⁵² Thus, thalamic activity related to the task continued to take place unaltered despite losing GPi tonic inhibitory input. More recently, Boraud's team in Bordeaux blocked the gabaergic activity of the GPi of non‐human primates, with uni‐ and bilateral microinjections of muscimol in different sessions. Monkeys had to first perform the previously learned task and subsequently had to learn a new task. Performance did not change for the learned task, but no learning took place while bilaterally pallidal activity was interrupted. ⁵³ These results are much in keeping with the patients without basal ganglia function in one hemisphere discussed above. ²⁷ Therefore, it seems that once a task/behavior has been learned, its execution does not rely upon basal ganglia output activity through the pallidothalamic output projection for performing already acquired (automatic) movements and behaviors. Moreover, bilateral GPi blockade largely rules out a possible compensatory effect by the non‐lesioned hemisphere.

Conclusions

The impact of basal ganglia focal lesioning here summarized supports that the basal ganglia are needed principally to acquire, learn, and make habitual the performance of actions and behaviors. Once a habit is created and the neuronal network associated with it has been shaped, ⁵⁴ the habitual action can occur without necessarily depending upon major basal ganglia engagement. On the other hand, the basal ganglia, particularly the hyperdirect projections, are needed to interrupt ongoing actions and switching priorities to accommodate unexpected novel circumstances.

The interruption of STN efferent activity impairs stopping and reacting to new stimuli, but non‐human primates and humans can still perform routine activities and motor tasks quite adequately. Thus, the preservation of these motor responses in the context of a basal ganglia lesion suggests that well learned stimulus‐driven responses can be executed without recourse to the basal ganglia motor loops. However, the motor regions of the basal ganglia seem to be critical when they are needed to acquire and integrate information to perform a novel motor task, as discuss for tasks that requires facilitation of appropriate responses in a probabilistic context.

Therefore, from a pathophysiological point, the detailed study of patients treated with therapeutic lesions supports the well‐known idea that it is better to stop abnormal basal ganglia activity than to allow faulty signalling to disrupt the motor system.

Of note, the impunity to lesioning is not complete, and the analysis argues against earlier claims that interrupting basal ganglia output in PD is totally unnoticeable. Clinically relevant, the highly negative impact of bilateral pallidotomy in PD ¹⁴ is still a pending issue. Whether the major neurological manifestations observed in those instances are the direct result of bilateral interruption of GPi output or related with lesioning of fibers going through the GPi and the adjacent internal capsule is unresolved. The further study of patients treated with bilateral (staged) pallido‐thalamic tractotomy by FUS ⁴⁹ could throw light into this relevant aspect.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the first draft, B. Review and Critique.

M.H.G. M.: 1A, 1B, 1C, 3A, 3B

J. U. M.‐M.: 1C, 3B

J. A. O.:1A, 1B, 1C, 3A, 3B

Disclosures

Ethical Compliance Statement: Informed patient consent was not necessary for this work, and the approval of an institutional review board was not required. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflicts of Interest: No funding was received for this work. No conflict of interest.

Financial Disclosures For Previous 12 Months: M.H.G.M. received speaker honorarium and travel grants from the Spanish Neurological Society and Abbie and speaker honorarium from Teva. M.H.G.M. is also supported by a research fellowship by Fundación Alfonso Martin Escudero. J.U.M.‐M. received speaker honoraria and reimbursement of travel expenses to attend scientific conferences from BIAL and Insightec. J.A.O.: received honorarium and travelling grants from BIAL Spain for lecturing at the Spanish Neurological Society annual meeting and received honorarium for lecturing in scientific meetings and travelling grants from Insightec Ltd and Esteve Pharmaceuticals. He has received consulting honorarium from Biogen or participating in the Adverse Events Advisory Board 2023 and consulting honorarium from Roche and Bayern for attending one Advisory Board for each. He also holds several non‐paid, non‐profit research grants from the Spanish Ministry of Education and Science, Focused Ultrasound Foundation and ASAP coalition.