Cerebellar Deep Brain Stimulation in Cerebral Palsy: Promising Early Results and a Look Forward to a Larger Clinical Trial

Cerebral palsy (CP) represents the most common physical disability of childhood, with a prevalence of 2–3.5 per 1000 live births,¹ affecting 1 of 323 children in the United States.² CP encompasses a group of disorders of movement and posture, attributed to nonprogressive disturbances that occurred in the developmental fetal or infant brain.³ Dyskinetic CP (DCP), the second most common type of CP after spastic forms, refers to subset of patients in whom dystonia and choreoathetosis are the predominant motor manifestations. Although the cause of CP is broad, most children with CP have abnormal brain magnetic resonance imaging indicative of cortical and deep gray matter damage from hypoxic ischemic encephalopathy,⁴ which may preclude or suggest decreased efficacy of standard deep brain stimulation (DBS) targets. The cerebellum has been posited as an attractive target for DCP because it is frequently spared from hypoxic ischemic damage and has shown promise in alleviating patient symptoms in early work in the 1970s and more recent case series with DBS.

Chronic cerebellar stimulation was first attempted by Cooper et al.⁵ in 1972 as a means of modulating the inhibitory potential of Purkinje cells in various neurologic conditions. Cooper subsequently reported a series of 50 patients with CP who underwent cerebellar cortical stimulation, with long-term follow-up of 28 patients showing improvements in spasticity, athetosis, speech, and overall functional status. Concerns regarding stimulator reliability, safety and surgical complications, optimal placement site and lack of quantitative measures⁶ precluded the widespread adoption of cerebellar stimulation.

We recently performed bilateral cerebellar DBS implantation, targeting the dentate nucleus (DN) and cerebellar outflow pathway, in 3 patients with DCP.⁷ All patients showed improvement in subjective motor function and objective improvement in the Burke-Fahn-Marsden Dystonia Rating Scale movement subscale, although the range of responses was variable (19%–40%). Patients experienced subjective improvement in motor function including ease of hand movements and coordination, gait, head control, speech, decreased overflow, and diminished muscle tightness.

Leveraging recent commercial availability of sensing enabled pulse generators designed for sensing during continuous stimulation, we observed prominent alpha activity in the motor DN of these patients with acquired dystonia consistent with the observed theta-alpha activity in basal ganglia in patients with isolated dystonia,⁸ in whom the magnitude of theta/alpha oscillatory activity may correlate with dystonia severity. In addition, we observed decreased variability in the tracked alpha peak, with increased stimulation in 1 patient who experienced significant improvements in speech and spasticity. Modulation of cerebellar dentate alpha activity by DBS is consistent with observed effects of DBS on supratentorial theta-alpha bands. The ability to longitudinally track local field potentials potentially correlated with motor symptom severity may open a potential pathway for closed loop stimulation in the future.

Given the preliminary evidence of clinical benefit and to further elucidate the clinical benefit of this therapy, we are launching a longer single-center phase 1/2 study of cerebellar DBS. For this pilot clinical trial, to be conducted at the University of California San Francisco, the primary end point will be the within-child difference in Caregiver Priorities & Child Health Index of Life with Disabilities (CPCHILD) quality-of-life (QOL) scale total standardized score while on intervention (effective stimulation) versus placebo (sham stimulation). Ten children and young adults with combined dystonia, choreoathetosis, and spasticity consistent with DCP secondary to hypoxic ischemic static encephalopathy will be implanted with Medtronic Percept and Medtronic SenSight electrodes (Medtronic, Minneapolis, Minnesota, USA) in bilateral dorsal (motor) DN. This is a pilot clinical trial in which 1 month of effective stimulation will be compared with 1 month of sham stimulation, administered in randomized order and cycled 3 times per participant, using an N-of-1 trial design, to mitigate the variability in clinical features in this population. Each N-of-1 trial will consist of a baseline assessment, implantation, and open-label phase to determine optimal stimulation settings, and 3 cycles of randomized effective stimulation and sham (ineffective) stimulation. In this study, we will also collect and characterize each patient’s physiologic and neuroimaging markers, which may predict hyperkinetic pathologic states, and their response to therapeutic DBS.

The rationale for using CPCHILD as the primary outcome measure is that health-related QOL measures provide a more complete picture of an individual and are complementary to specific motor assessments,⁹ and CPCHILD has been previously used in a recent trial of pallidal DBS in DCP.¹⁰ Key secondary outcome measures will include other validated health-related QOL scores and age-appropriate and condition-appropriate motor and spasticity rating as well as exploratory automatic video kinematic and gait analyses.

DBS of motor DN in DCP seems to be safe and shows preliminary evidence of clinical benefit. New chronic sensing technology may allow for determination of in vivo mechanisms of network disruption in DCP and allow for further understanding of the effects of neuromodulation on brain physiology. Our recent report addresses a gap in the literature regarding delineating the technical approach to image-based stereotactic targeting and chronic neural recording in the DN and outlines our plans for a pilot clinical trial of this therapy for patients with DCP.