Axial motor symptoms (such as freezing of gait, postural instability, and speech impairment) are major sources of disability in the advanced phases of Parkinson's disease (PD), given that their response to levodopa or DBS is often unsatisfactory. DBS of either STN or globus pallidus internus (GPi) is usually provided at high-frequency stimulation (HFS; >100 Hz), but recent studies have suggested that low-frequency stimulation (LFS; <100 Hz) of the STN may better control axial symptoms. Although a metaregression analysis has shown that GPi-DBS in combination with l-dopa might preserve long-term axial symptoms more efficiently than STN-DBS, no formal study has addressed the impact of LFS in patients with PD who underwent such surgical treatment, with the exception of an anecdotal report.
We conducted a retrospective chart review of PD patients who received GPi-DBS at the Toronto Western Hospital and found 6 patients who were trialed with LFS. All patients had a good outcome 6 months postsurgery, with an improvement of motor signs (−45.3 ± 11.6% of the motor score of the UPDRS), motor fluctuations, and dyskinesias (−64.7 ± 13.1% of the “complications of therapy” section score of the UPDRS). One patient was excluded because LFS was only tried for a few minutes owing to the poor control of resting tremor. The remaining 5 patients developed troublesome axial features 686 ± 375 days postsurgery, while being stimulated with standard parameters (voltage, 3.1 ± 0.5 V; pulse width: 66.0 ± 12.6 μs; frequency, 132 ± 30 Hz).
In an attempt to improve axial symptoms, patients were switched to LFS (80 and 60 Hz in 3 and 2 of them, respectively), pulse width was left unchanged, but voltage was increased to keep the energy delivered constant (3.6 ± 1.3 V). Pharmacological treatment was not changed during the follow-up period (l-dopa daily equivalents were 1,234 ± 410 mg during HFS and 1,274 ± 423 mg during LFS).
Three patients (1 female; age, 56.3 ± 5.9 years; disease duration: 19.7 ± 13 years) did not report any substantial benefit and were switched back to HFS after 164 ± 155 days. By contrast, 2 patients (1 female; age, 73.0 ± 5.7 years; disease duration: 10.5 ± 3.5 years) presented an immediate improvement of axial symptoms (speech and balance in 1 patient, gait and balance in the second) in the absence of worsening of other symptoms (Fig. 1) and were left on LFS (follow-up duration: 346 ± 208 days).
Figure 1. UPDRS total motor score (upper panel) and axial score (sum of UPDRS-III subitems for speech, arising from a chair, posture, gait, and postural stability; lower panel) of the 5 GPi DBS patients receiving LFS. All evaluations were performed on l-dopa at T0 (preop), T1 (postop, with conventional HFS), and T2 (postop with LFS). Black lines illustrate the 2 patients improving with LFS, whereas gray lines illustrate the 3 patients not reporting any substantial benefit.
We have found that LFH at either 60 or 80 Hz may be useful in a subgroup of PD patients receiving GPi-DBS. Interestingly, the 2 patients who improved had a postoperative worsening of axial symptoms, possibly meaning that LFS may lessen a detrimental effect of DBS, as described for STN-DBS patients[1, 4] and GPi-DBS patients with other movement disorders. Alternatively, LFS may be more efficient at disrupting pathological output or modulating brainstem locomotor regions. On the other hand, a role of disease progression cannot be excluded and no generalizations about mechanisms are possible given the few subjects involved.
LFS of GPi has been deemed to be clinically useful in generalized dystonia and in the parkinsonian features of Huntington's disease.Future studies systematically examining the effects of LFS in PD patients treated with GPi-DBS are needed to better explore the clinical relevance of such a stimulation paradigm.