LAUSR

Freezing of gait (FOG) is a highly debilitating symptom of advanced Parkinson’s disease (PD) during which patients experience a sudden inability to take a step while walking, turning or when attempting to initiate gait, often resulting in falls (Nutt et al., 2011). FOG has a complex and heterogeneous pathophysiology whereby the impaired habitual control of gait in PD becomes vulnerable to interference from consecutive processing across motor, cognitive and limbic cortico-striatal circuits in the absence of striatal dopamine, whilst the ability to apply executive control over gait diminishes as a result of progressing extra-nigral neuropathology (Bohnen and Jahn, 2013; Lewis and Shine, 2016). Models of FOG suggest it could be due to transient over-activity in the subthalamic nucleus (STN), which leads to a surge of inhibitory projections from the globus pallidus internus and substantia nigra pars reticulata to the ‘final common pathway’ of the brainstem locomotor regions that control gait (Fig. 1A) (Lewis and Shine, 2016). The STN is indeed the main output structure of the indirect cortico-striatal pathways that are overly active in PD during the dopaminergic off state resulting in motor inhibition. The corticoSTN hyper-direct pathway further blocks motor output during instances of response conflict that likely arise more frequently in advanced PD due to worsening sensorimotor deficits (Shine et al., 2013; Nambu et al., 2015). However, the exact role of the STN in FOG still needs further elucidation. Manipulating STN deep brain stimulation (STN-DBS) frequencies therefore provides an interesting window in clarifying the modulatory role of the STN in the etiology of FOG in PD. High-frequency (120–140 Hz) STN-DBS is the treatment of choice for many patients with PD, as it provides sustained improvements of motor signs that are dopamine responsive but cannot be optimally relieved with medications due to dosage limiting side-effects (Rodríguez-Oroz et al., 2012). The effectiveness of high-frequency STN-DBS has been linked to the attenuation of tremor related (4–7 Hz) oscillations as well as pathological beta (15–30 Hz) oscillations that are prominent across the cortico-basal ganglia circuits in PD during the off state (Kühn et al., 2008; Blumenfeld et al., 2017). Unfortunately, however, high-frequency STN-DBS does not satisfactorily alleviate axial motor signs that are partly dopamine resistant, such as FOG, and may even aggravate these symptoms for some patients (Moreau et al., 2008; Xie et al., 2012). Many specialized centers therefore consider dopamine resistant FOG as an exclusion criterion for receiving high-frequency STN-DBS. Moreover, DBS in PD is currently applied using pre-set stimulation settings that remain static over time (i.e., open-loop), often requiring multiple post-operative parameter iterations to achieve the desired clinical benefits and probably limiting the effectiveness of DBS for paroxysmal symptoms, such as FOG (Udupa and Chen, 2015; Gilat et al., 2018). In this issue of Clinical Neurophysiology, Varriale and colleagues (Varriale et al., 2018) provide important evidence of how different STN-DBS settings impact on gait initiation performance in PD. In a double blind and randomized controlled manner, conditions before and after STN-DBS surgery, both ON and OFF medication and with low (80 Hz) and high (130 Hz) DBS frequency settings were compared. Most interestingly, stimulation effects were assessed during both a simple task condition (n = 19), when patients initiated gait in response to a single auditory tone, as well as in a complex task condition (n = 9) when presented with congruent and incongruent visual cues depicting go or no-go trials. The complex task condition was used to investigate patient’s ability to simultaneously process cognitive and motor information (Varriale et al., 2018), which requires consecutive processing across parallel cortico-basal ganglia circuits that converge in the STN and often leads to FOG (Lewis and Shine, 2016). Their findings showed that compared to being OFF, both dopamine and either DBS setting improved simple gait initiation with only a slight enhancement seen during lowversus high-frequency DBS (Varriale et al., 2018). During the complex task condition, however, a significant effect of stimulation was found, whereby high-frequency DBS worsened gait initiation significantly more than low-frequency DBS. This study thus provides important behavioral evidence that isolated motor programs can be passed through when applying either stimulation setting, whereas highfrequency STN-DBS impairs PD patient’s ability to simultaneously process motor and cognitive information during complex gait tasks. In contrast, low-frequency DBS helped to prevent cognitive-motor interference and likely allowed for compensatory gait control. Although the authors did not measure FOG directly, gait initiation commonly provokes FOG and serves as a good proxy for this gait complication. One possibility for the differential effect of STN-DBS on FOG, described by Varriale et al. (2018), could be overflow of stimulation