Assistant Professor of Biomedical Engineering Bucknell University Lewisburg, Pennsylvania, United States
Introduction:: Deep brain stimulation (DBS) is a neurosurgical procedure where electrodes are implanted in the brain to modulate specific targets with electricity. Stimulation of the subthalamic nucleus (STN) and globus pallidus internus (GPi) have shown to improve motor function in Parkinson’s disease (PD) [1,2]. Gait disturbances are particularly common in PD, creating a therapeutic need due to declining mobility and quality of life as the disease progresses [3]. While some symptoms of PD, such as tremor, rigidity, and bradykinesia, respond well to DBS at either the STN or GPi, others involving cognition, mood, and behavior respond differently to STN and GPi DBS [2]. A generalized targeting approach may therefore not be ideal for every patient due to the diversity of their symptoms. For example, a better clinical outcome for gait impairment might be achieved through stimulation of different regions within or around the STN [4]. The position of the active electrode contact is often used as a measure of stimulation location to determine ideal stimulation sites [4]. Volume of tissue activation (VTA) modeling can also be used to measure stimulation location by estimating the spatial extent of stimulation [5,6]. The VTA is a three-dimensional representation of stimulation in the brain that enables the investigation of how stimulation spreading from the active contact relates to specific brain structures and clinical outcomes [5]. This may facilitate more personalized treatment.
Materials and Methods:: Data from 40 advanced PD patients who received bilateral STN DBS at the University of Michigan were used in this retrospective study. For 72 implants, the location of therapeutic stimulation was calculated using patient-specific VTA models built from imaging data and stimulation settings [5]. This individualized modeling approach was taken to create a more accurate representation of stimulation location for each patient. The VTA was used to quantify STN and external (non-STN) activation in different regions (Figure 1A). Stepwise regression was used to evaluate associations between stimulation location and gait improvement, which was calculated from the MDS-Unified Parkinson’s Disease Rating Scale. Gait, freezing of gait (FoG), postural stability, and total gait (the sum of these) were evaluated. Two regression models were created for comparison: one based on STN and external activation in six directions (12 predictors total) and the other based on active contact position in three directions (3 predictors total). This was done for each symptom. Implants grouped by stimulation location were then compared based on gait symptom improvement using Kruskal-Wallis tests. The ratios between STN activation in opposing directions (±x, ±y, ±z) were calculated and sorted in ascending order. This was done to separate implants with stimulation centered around the STN (middle 20%). To correct for multiple comparisons, the Benjamini-Hochberg false discovery rate control procedure was used. Significance was defined at a p-value of less than 0.05 after corrections. Electrode position (relative to the STN centroid) was also examined to compare to the VTA analyses.
Results, Conclusions, and Discussions:: Stepwise regression between VTA location and symptom improvement showed statistically significant positive associations between anterior STN activation and gait improvement (r = 0.26, p = 0.03) and total gait improvement (r = 0.31, p = 0.01) (Figure 1B). A significant positive association between anterior external activation and FoG improvement (r = 0.48, p = 0.02) was also found despite most patients having complete improvement. After grouping the implants based on stimulation location, Kruskal-Wallis tests showed significant differences between majority anterior STN activation and majority posterior STN activation in terms of FoG improvement (p = 0.03) and total gait improvement (p = 0.02) (Figure 1C). In contrast to VTA location, stepwise regression between active contact distance from the STN centroid and symptom improvement did not show any significant associations. Kruskal-Wallis tests also did not show any significant differences between active contacts located in opposite halves of the STN in terms of symptom improvement after grouping the implants. The dorsolateral, motor portion of the STN is commonly targeted to treat the cardinal symptoms of PD (tremor, rigidity, and bradykinesia) with DBS [6]. It is important to consider this deviation from the traditional target region alongside possible side effects. Greif et al. found that more anterior active contacts relative to the midpoint of the left STN were associated with phonemic verbal fluency decline [7]. Conversely, Dafsari et al. reported that anterior, medial, and ventral active contacts were related to better non-motor outcomes [8]. Using VTA-based analyses, McIntyre et al. found that postural stability improved with decreased stimulation in the anterior, non-motor portions of the STN [9], but Dembek et al. found that side effects in general occurred with more posterior and ventral stimulation [6]. Discrepancies in these findings may be explained by the different methods used to measure stimulation location. Overall, this study demonstrates how the VTA can provide more information about stimulation location than active contact position alone and highlights the importance of patient- and symptom-specific DBS targeting. Future work can investigate other symptoms of PD (motor and non-motor) in a similar manner to determine optimal stimulation sites.
Acknowledgements (Optional): : The authors would like to thank the patients included in this study for their voluntary participation and Parag G. Patil, M.D., Ph.D. and Kelvin L. Chou, M.D. from the University of Michigan for clinical data.
References (Optional): : [1] K.A. Follett, D. Torres-Russotto, Deep brain stimulation of globus pallidus interna, subthalamic nucleus, and pedunculopontine nucleus for Parkinson’s disease: Which target?, Parkinsonism & Related Disorders. 18 (2012) S165–S167. https://doi.org/10.1016/S1353-8020(11)70051-7.
[2] S. Mirza, U. Yazdani, R. Dewey, N. Patel, R.B. Dewey, S. Miocinovic, S. Chitnis, Comparison of Globus Pallidus Interna and Subthalamic Nucleus in Deep Brain Stimulation for Parkinson Disease: An Institutional Experience and Review, Parkinson’s Disease. 2017 (2017). https://doi.org/10.1155/2017/3410820.
[3] R.G. Cury, N. Pavese, T.Z. Aziz, J.K. Krauss, E. Moro, Gaps and roadmap of novel neuromodulation targets for treatment of gait in Parkinson’s disease, Npj Parkinson’s Disease. 8 (2022). https://doi.org/10.1038/s41531-021-00276-6.
[4] E.L. Johnsen, N. Sunde, P.H. Mogensen, K. Østergaard, MRI verified STN stimulation site - Gait improvement and clinical outcome, European Journal of Neurology. 17 (2010) 746–753. https://doi.org/10.1111/j.1468-1331.2010.02962.x.
[5] K.A. Malaga, J.T. Costello, K.L. Chou, P.G. Patil, Atlas-independent, N-of-1 tissue activation modeling to map optimal regions of subthalamic deep brain stimulation for Parkinson disease, NeuroImage: Clinical. 29 (2021) 102518. https://doi.org/10.1016/j.nicl.2020.102518.
[6] T.A. Dembek, J. Roediger, A. Horn, P. Reker, C. Oehrn, H.S. Dafsari, N. Li, A.A. Kühn, G.R. Fink, V. Visser-Vandewalle, M.T. Barbe, L. Timmermann, Probabilistic sweet spots predict motor outcome for deep brain stimulation in Parkinson disease, Annals of Neurology. 86 (2019) 527–538. https://doi.org/10.1002/ana.25567.
[7] T.R. Greif, A. Askari, A. Cook Maher, P.G. Patil, C. Persad, Anterior lead location predicts verbal fluency decline following STN-DBS in Parkinson’s disease, Parkinsonism & Related Disorders. 92 (2021) 36–40. https://doi.org/10.1016/j.parkreldis.2021.10.012.
[8] H.S. Dafsari, J.N. Petry-Schmelzer, K. Ray-Chaudhuri, K. Ashkan, L. Weis, T.A. Dembek, M. Samuel, A. Rizos, M. Silverdale, M.T. Barbe, G.R. Fink, J. Evans, P. Martinez-Martin, A. Antonini, V. Visser-Vandewalle, L. Timmermann, Non-motor outcomes of subthalamic stimulation in Parkinson’s disease depend on location of active contacts, Brain Stimul. 11 (2018) 904–912. https://doi.org/10.1016/J.BRS.2018.03.009.
[9] C.C. McIntyre, S.J. Richardson, A.M. Frankemolle, G. Varga, A.M. Noecker, J.L. Alberts, Improving postural stability via computational modeling approach to deep brain stimulation programming, Proc. Annu. Int. Conf. IEEE Eng. Med. Biol. Soc. EMBS. (2011) 675–676. https://doi.org/10.1109/IEMBS.2011.6090151.