Article Text
PDF
Abstract
Objective To evaluate the long-term effect of 60 Hz stimulation of the subthalamic nucleus (STN) on dysphagia, freezing of gait (FOG) and other motor symptoms in patients with Parkinson’s disease (PD) who have FOG at the usual 130 Hz stimulation.
Methods This is a prospective, sequence randomised, crossover, double-blind study. PD patients with medication refractory FOG at 130 Hz stimulation of the STN were randomised to the sequences of 130 Hz, 60 Hz or deep brain stimulation off to assess swallowing function (videofluoroscopic evaluation and swallowing questionnaire), FOG severity (stand–walk–sit test and FOG questionnaire) and motor function (Unified PD Rating Scale, Part III motor examination (UPDRS-III)) at initial visit (V1) and follow-up visit (V2, after being on 60 Hz stimulation for an average of 14.5 months), in their usual medications on state. The frequency of aspiration events, perceived swallowing difficulty and FOG severity at 60 Hz compared with 130 Hz stimulation at V2, and their corresponding changes at V2 compared with V1 at 60 Hz were set as primary outcomes, with similar comparisons in UPDRS-III and its subscores as secondary outcomes.
Results All 11 enrolled participants completed V1 and 10 completed V2. We found the benefits of 60 Hz stimulation compared with 130 Hz in reducing aspiration frequency, perceived swallowing difficulty, FOG severity, bradykinesia and overall axial and motor symptoms at V1 and persistent benefits on all of them except dysphagia at V2, with overall decreasing efficacy when comparing V2 to V1.
Conclusions The 60 Hz stimulation, when compared with 130 Hz, has long-term benefits on reducing FOG, bradykinesia and overall axial and motor symptoms except dysphagia, although the overall benefits decrease with long-term use.
Clinical trial registration NCT02549859; Pre-results.
- DBS
- STN
- Parkinson’s disease
- freezing
- gait
- dysphagia
- frequency
Statistics from Altmetric.com
Introduction
Deep brain stimulation (DBS) improves levodopa responsive cardinal symptoms, motor fluctuation, dyskinesia and medication refractory tremor in patients with Parkinson’s disease (PD).1–4 However, DBS at routinely used high frequency stimulation (HFS; 130–180 Hz) is less effective at improving the axial symptoms of postural instability, gait disorders, speech and swallowing dysfunction. The axial symptoms might even worsen over the course of 2–5 years of subthalamic nucleus (STN) stimulation, particularly in the medication on state.5–7 Freezing of gait (FOG) has been reported in many patients with bilateral STN DBS, mostly after years of HFS, and some shortly after the initiation of HFS, in either the medication off or both off and on state (see review8).
Low frequency stimulation (LFS) of 60 Hz, compared with HFS, has been reported to improve the FOG and other axial (including speech) and motor symptoms in patients with PD,9–14 mostly in short-term studies. The long-term effect of LFS has been inconsistent,9 11 15 16 with heterogeneity in patient cohorts and study designs. The impact of disease progression has not been considered in the long-term follow-up, as LFS was not compared with HFS or DBS off state. The potential carryover effect related to the order of the tests has been rarely analysed, although the axial symptoms could linger longer than the cardinal symptoms following DBS setting change.17
Recently, we found that 60 Hz stimulation improved swallowing dysfunction, along with FOG and overall axial and motor symptoms, in PD patients with FOG at HFS and medication on state.18 This is important, as dysphagia is commonly present in PD,19 and often refractory to pharmacological management,20 21 with high risk of morbidity and mortality. However, it remains unknown whether there is a long-term benefit on swallowing function beyond the 6 weeks studied,18 especially considering the potential disease progression and potential carryover effect associated with the order of DBS assessment.
Therefore, we conducted this sequence randomised, double-blind, crossover, prospective study to first confirm the previously reported initial benefits and then further assess the long-term benefits of 60 Hz stimulation compared with 130 Hz on swallowing function, FOG and overall axial and motor symptoms in PD patients with FOG at the usual bilateral 130 Hz STN stimulation and medication on state, taking disease progression and potential carryover effect into consideration.
Methods
Participants
Written informed consent was obtained from the patients. The study was listed under ClinicalTrials.gov (NCT02549859) and conducted at the University of Chicago from August 2015 to March 2017.
PD patients with FOG under optimal bilateral STN DBS at 130 Hz and medication, who were able to walk without assistance and had no clinical aspiration pneumonia, were enrolled. We initially aimed to complete 18 patients in 2 years to ensure 80% power to detect at least 25% difference at a two-tail α level of 0.05 for potential carryover effect.18 However, the study was closed prior to reaching the recruitment goals as we lagged behind the enrolment milestone due to the loss of one coordinator.
As a result, we enrolled 11 patients: 6 were new for both the initial visit 1 (V1) and follow-up visit 2 (V2) and 5 were carried over from our previous study of 7 patients who had been on 60 Hz stimulation since then (with their initial visit in the past as V118 and thus new for V2. Two patients from the previous study were not enrolled as they did not remain on 60 Hz stimulation due to worsening of tremor with LFS in one requiring return to 130 Hz, and a skin infection in the other requiring removal of his electrodes. All 11 subjects completed V1 and 10 also completed V2 after being on 60 Hz stimulation for an average of 14.5 months.
Procedures
All subjects were tested in their usual medication on state given the long assessment duration at each visit to minimise hardship on patients and variability of symptoms with the prolonged off period. At V1, each patient received three videofluoroscopic swallow (VFSS) evaluations (also called modified barium swallow or oropharyngeal motility study) in a single day under three DBS conditions (130 Hz, 60 Hz or DBS off), with other parameters unchanged and in a sequence randomised (one of the six sequences from the combinations of three DBS conditions) double-blind crossover manner, as before.18 The neurologist who programmed the DBS did not participate in any evaluation. Neither the patient nor the study team was aware of the DBS state. One certified rater rated the UPDRS-III, and another one rated the questionnaires. A team comprised of radiologists and speech pathologists conducted the VFSS studies and the Penetration-Aspiration Scale ratings as described.18 The swallowing questionnaire was completed after each swallowing study.18
The stand–walk–sit (SWS) test (measuring the number of FOG spells as well as the total time to complete the test), the FOG questionnaire (measuring a patient’s perception of walking difficulty), as well as UPDRS-III (measuring axial symptoms, bradykinesia, rigidity, tremor and overall motor symptoms) were assessed before the swallowing evaluation under each DBS condition, as described before.18 The patients were on each DBS condition for at least 30 min before the evaluation. After V1, all the patients were put on 60 Hz for at least 6 months per design until V2 to repeat the full evaluation as in V1, with a few exceptions. One patient returned to 130 Hz in 2 months due to worsening of the tremor at LFS. Another patient was only on 60 Hz for several weeks when this study was closed, and hence did not participate in the long-term V2 assessment. One patient had a hip fracture before V2 and was unable to perform the UPDRS-III exam and FOG study at V2. This patient did not complete the FOG study at V1 either due to excessive fatigue and imbalance.
Statistical analysis and outcomes
Changes in measures were analysed between each condition (60 Hz, 130 Hz and DBS off) using paired t-test within each visit (V1 or V2), and in each DBS condition across the two visits, on swallowing function (objective VFSS and subjective swallowing questionnaire), FOG severity (objective SWS study and subjective FOG questionnaire) and other motor function in UPDRS-III total score and subscores. The primary outcomes included the frequency of aspiration events in VFSS, perceived swallowing difficulty in the swallowing questionnaire and FOG severity at 60 Hz compared with 130 Hz stimulation at V2, and their corresponding changes at V2 compared with V1 at 60 Hz. The secondary outcomes included UPDRS-III total score, axial subsscore and cardinal subscore of bradykinesia, rigidity and tremor at 60 Hz compared with 130 Hz stimulation at V2 and their corresponding changes at V2 compared with V1 at 60 Hz. Changes in measures between other DBS states, including 60 Hz versus DBS off, and 130 Hz versus DBS off were also explored using a paired t-test. Bonferroni correction was applied for multiple comparisons. A multivariant linear regression model between the outcome measurements and conditions (60 Hz, 130 Hz and DBS off) adjusted by the order of DBS conditions entered into the study and the visits was also used to analyse the potential carryover effect. A two-tailed alpha level of 0.05 was taken as statistically significant for the comparisons.
Results
Participant disposition, demographics and baseline characteristics
We screened 15 patients and excluded four of them (two possible progressive supranuclear palsy, one recent aspiration pneumonia and one having no convincing FOG). The 11 enrolled subjects all completed V1 (100%) and 10 (91%) completed the V2 (figure 1), with 9 (82%) men, six self-identified as white (55%), mean age of 68.5 (SD 5.9) years old, PD duration of 14.2 (SD 5.7) years and DBS duration of 3.5 (SD 4.0) years at HFS, on normal range of voltage and pulse width, with 20 leads on monopolar and 2 on bipolar configurations, and 16 on dorsal and 6 on ventral active contacts (table 1). Over the course of the study, medication was only slightly reduced in three patients and increased in 1, and voltage was only slightly increased by 0.1–0.3 volts in three patients to maintain overall motor function.
Subject disposition flow diagram.
Demographics of the patients
Effect of LFS on aspiration, FOG and other motor symptoms at V1 and V2 and changes between V1 and V2
At V1, compared with 130 Hz, the 60 Hz stimulation reduced the frequency of aspiration events on VFSS by 79% (P<0.05) and reduced swallowing difficulty perception on swallowing questionnaire by 67% (P<0.01) (table 2). The 60 Hz stimulation also reduced the time to complete the walking (P<0.05) in SWS test and reduced the walking difficulty perception in FOG questionnaire (P<0.01) (table 2). The 60 Hz stimulation also improved the overall motor symptom scores of UPDRS-III (P<0.01), axial symptoms (P<0.001) and bradykinesia (P<0.05) (table 2). Interestingly, even compared with the DBS off state (medication on), the 60 Hz stimulation still reduced the perception of swallowing difficulty on the questionnaire by 65% (P<0.05), the number of FOG spells (P<0.05), the walking difficulty perception (P<0.05) and the overall parkinsonism in UPDRS-III (P<0.001), axial symptoms (P<0.001) and bradykinesia (P<0.01) (table 2). No significant difference was found between 130 Hz and DBS off (medication on) in any of these measures.
Motor, swallowing and FOG under different DBS conditions in medication on state at V1
At V2, after continuous LFS for an average of 14.5 months, the 60 Hz stimulation did not exhibit significant benefit in reducing the frequency of aspiration events and swallowing difficulty perception compared with the 130 Hz (table 3). However, the 60 Hz stimulation still reduced the FOG spells (P<0.05), the walking difficulty perception (P<0.01) and the overall parkinsonian symptoms in UPDRS-III (P<0.05), axial symptoms (P<0.05) and bradykinesia (P<0.05) (table 3). Interestingly, even compared with DBS off state, the LFS still reduced the overall parkinsonism in UPDRS-III (marginally, P=0.053) and bradykinesia (P<0.05) (table 3). Mild worsening in axial symptoms and perception of the gait difficulty were found at 130 Hz compared with DBS off (medication on), as possibly expected in long-term follow up.
Motor, swallowing and FOG under different DBS conditions in medication on state at V2
When comparing V2 to V1 under individual DBS conditions (table 4), the benefits on the UPDRS-III, axial symptoms, bradykinesia, FOG Q and aspiration frequency at 60 Hz became significantly or marginally less at V2, though 60 Hz still worked better than 130 Hz on all measures except the aspiration frequency (Tables 3 and 4), suggesting a persistent benefit but with decreasing efficacy after long-term LFS. There was no significant change in 130 Hz and DBS off state over the period studied (table 4).
Motor, swallowing and FOG under different DBS conditions in medication on state: comparing V2 to V1
No confounding carryover effects by the order of the DBS condition were found at each visit and across the visits on any measurements (data not shown).
Discussion
In this sequence randomised, double-blind, crossover, prospective study assessing the long-term benefits of LFS on swallowing function, FOG and other axial and motor symptoms in PD patients with FOG on routine HFS of bilateral STN DBS in medication on state, we confirmed the previously reported acute or short-term beneficial effect of LFS at V1 compared with HFS on all these symptoms.18 As the major aim of this study, we further demonstrated the long-term beneficial effect of 60 Hz stimulation compared with 130 Hz on these measures at V2 except dysphagia. In addition, the beneficial effects of LFS overall were less potent after long-term LFS. These beneficial effects of LFS also remained when compared with DBS off state at V1 and persisted to some extent at V2. We have not found any confounding carryover effect by the order of the DBS conditions tested.
Our study has unique strengths. First, this is the first study on the long-term effect of LFS on swallowing function in PD patients with FOG at HFS. Second, this is the longest follow-up study (with an average of 14.5 months as a study group) at LFS of 60 Hz on FOG and axial symptoms. Third, the potential impact of disease progression with long-term follow-up is also taken into consideration for the first time by comparing 60 Hz to 130 Hz and to the DBS off state. Fourth, the study was designed in a randomised, double-blind, crossover prospective manner, instead of retrospective chart reviews as in other long-term studies, to assess both objective and subjective measures. Fifth, the potential confounding carryover effect was also assessed, which had been rarely studied before even though it is important as the axial symptoms could persist for hours after a DBS setting change.17
Nevertheless, our study also has limitations. First, the sample size is small, which makes it difficult to determine if the lack of benefit on swallowing function at 60 Hz compared with 130 Hz after long-term LFS reflects the true loss of benefit or the lack of power. Second, we excluded patients with severe dysphagia for safety and ethical reasons that could also make it more difficult to detect subtle changes, although we did find the benefit of LFS on dysphagia compared with 130 Hz at V1. Third, we did not adjust the voltage when we switched the HFS to LFS, hence the total electric energy delivered (TEED) was reduced.22 23 This could be another possible cause of the lack of a long-term beneficial effect on swallowing function, as increase in voltage when switching from HFS to LFS could have a more profound benefit on axial symptoms as reported by Moreau et al,9 although that study was not specifically designed to test swallowing function and the DBS parameters were not exactly the same as ours. However, in a separate study, we explored the effect of LFS, with and without the adjustment of TEED, on UPDRS-III and SWS test and did not find a significant difference in the beneficial outcomes of LFS regardless of TEED adjustment.24 This finding supports more of a frequency (instead of energy) specific benefit of LFS, at least at the parameters used here. Fourth, the randomised design and the assessment of potential carryover effects and the impact of disease progression during long-term follow-up made the assessment of all DBS states in V2 in a randomised order essential. One could argue that this design partially makes V2 another acute assessment after the chronic LFS. However, without a better perceivable design to cover all, our current study makes it fair to say that the 60 Hz stimulation remains beneficial compared with HFS after long-term LFS in this patient population, which serves the aim of the study.
As in our previous study,18 this study was conducted in medication on state. Thus, we would not expect to see a significant improvement of the total motor score at HFS compared with DBS off,25 because it is well acknowledged that the improvement of all motor symptoms by DBS (at least at common HFS) is very close, or equal, to the best levodopa response.26 Similarly, there would not be a significant change in tremor score as most of the participants had medication responsive tremor, although one patient had to be switched back to 130 Hz earlier than anticipated due to worsening of tremor.
A common question of interest is who should have or would benefit from LFS. As suggested by our recent review, patients with FOG under routine HFS almost always benefit from the LFS.8 Almost all the studies with patients exhibiting FOG at HFS on examination showed a good response to LFS,9 10 13 14 18 23 regardless of medication off9 14 or on state,10 13 18 23 with9 14 23 24 or without adjusting the TEED.10 11 18 23 24 Some studies with FOG by history but not necessarily on examination could also have good response to LFS.12 15 Studies that did not specify FOG as a prerequisite or presumably had no FOG in the majority of patients might not be able to show significant benefits under LFS.16 27
Most FOG seems to occur under chronic stimulation of 2–5 years after bilateral STN DBS placement.9 10 14 15 18 One possible explanation is the worsening of the axial symptoms5–7 and the deterioration of the non-dopaminergic system as the disease progresses, particularly around 15 years after disease onset,19 which is equivalent to 2–5 years after STN DBS, given that average DBS placement is about 11–13 years after PD onset.1 3 4 The cholinergic system has been found to be associated with gait dysfunction and falls.28 29 Another possibility is the gradual escalation in the voltage over time, which could produce a relatively large stimulation volume to adversely affect axial symptoms at HFS,9 although this may not always be the case.11 13
Bradykinesia could also benefit from LFS,12 30 as it was found to worsen with axial symptoms as the disease progresses.31
The underlying mechanism for the benefits of LFS remains unclear. The FOG at HFS in our patients is less likely from overstimulation, given the usual voltage and pulse width, and the similar UPDRS-III score at 130 Hz compared with DBS off state in medication on state, particularly at V1, though axial symptoms could get worse as time goes.5–7 One possibility is the spread of the stimulation current to the PPN region (PPNr), which is close to STN.32 Functional imaging also suggests PPNr as the control region of FOG.33 HFS of STN DBS could exert unfavourable effects on these areas,34 while extreme LFS of 10 Hz–25 Hz and even 60 Hz of the PPNr improves gait and locomotion.34–37 Hence, 60 Hz probably would be more likely to produce a beneficial effect on FOG. It makes sense that ventral contacts, which are closer to PPNr, could be more likely to benefit from LFS,9 12 13 though this is not always the case.8 Overriding antikinetic neuronal oscillations in the beta band and enhancing the prokinetic gamma band activity by 60 Hz stimulation could also play a role in the beneficial effect of LFS.38 39 Speech and swallowing function are thought to be in the area surrounding PPNr,40 which is probably why they share a similar beneficial pattern as FOG.9 11 14 18
Conclusions
This is the first study examining the long-term effect of LFS on dysphagia, the longest follow-up study of 60 Hz stimulation on FOG and overall axial and motor symptoms, and the first comparing all DBS conditions in the follow-up study for possible disease progression and assessing potential carryover effect in PD patients with FOG presented at HFS of STN in medications on state. We confirmed the previously reported acute or short-term beneficial effects of LFS compared with HFS on dysphagia, FOG and other axial and motor symptoms and further demonstrated long-term beneficial effects in these symptoms except dysphagia, despite an overall decreasing benefit of LFS with the long-term use. We believe this study can have a significant impact on managing PD patients with FOG under HFS, a commonly encountered difficult symptom in STN DBS patients, and hence possibly reduce their morbidity and mortality. Further studies are underway to explore the underlying mechanism.
Acknowledgments
We thank our patients for their participation in the study.
References
Footnotes
Contributors TX: study concept and design, obtaining funding, supervision and execution, data analysis and interpretation, and drafting and revising the manuscript. LB, MP, EM, AD, JV and PW: data collection, analysis and interpretation, and revising the manuscript. BB: data collection and revising the manuscript. WK: statistical data analysis and interpretation, and revising the manuscript. DS, CZ, KM: data interpretation, and revising the manuscript. UJK: study design, data analysis and interpretation, and revising the manuscript.
Funding This study had a grant support by the Michael J. Fox Foundation for Parkinson’s Disease under the Rapid Response Innovation Award (extension) to TX.
Competing interests None declared.
Patient consent Not required.
Ethics approval The study protocol and ethics were approved by the institutional review board at the University of Chicago.
Provenance and peer review Not commissioned; externally peer reviewed.