Hippocampus
We applied FreeSurfer analysis on the three different hippocampal nuclei: CA1, CA3 CA4 and the hippocampus as a whole (Figure 4): Kruskal-Wallis ANONVA, (df = 2, χ2 = 0.66, p = 0.71). Furthermore, when comparing the control ET population to the PD group, there were no differences noted in hippocampal volume following short STN-DBS (nonparametric t-test; p = 0.95). No observable trends were noted.
Lateral Orbitofrontal Cortex (OFC)
In PD patients, the lateral orbitofrontal cortex holds importance with its involvement in higher olfactory processing such as odor discrimination. Furthermore, with the convergence of gustatory and olfaction taking place in the OFC, it would be no surprise if STN-DBS induced volumetric changes in this structure; we ran FreeSurfer analysis on ipsilateral OFC volume. In the PD stimulation group, there were no significant differences found between the short, medium and long stimulation groups (Figure 4); Kruskal-Wallis ANONVA, (df = 2, χ2 = 3.44, p = 0.17). However, the long stimulation group did have a larger range of ipsilateral OFC volumes compared to short stimulation. In the ET vs PD group, there were no significant differences in volume following short stimulation; however, similarly, the ET group did have a larger range of ipsilateral OFC volume compared to the PD group (nonparametric t-test; p = 0.59).
Olfactory Bulb
Olfactory bulb is not only a site of neurogenesis, but also receives dopaminergic input; therefore, changes in its volume that have been shown to parallel with olfactory recovery should be expected with STN-DBS. Furthermore, with its role in being a phenotypical marker to signify olfactory dysfunction in PD patients, we decided to track changes in OB volume with length of stimulation to determine if this structure can be a predictive marker of olfactory recovery following stimulation. When comparing ipsilateral OB volumes (OB on the hemisphere that received the stimulation), there was a significant difference in OB volumes following short, medium, and long stimulation (Figure 5); Kruskal-Wallis ANOVA, (df = 2, χ2 = 8.62, p = 0.01). Post-hoc multiple comparisons revealed one significant difference between short and long stimulation (short vs long: p = 0.009, short vs med: p = 0.12, med vs long: p = 0.38). The long stimulation group has significantly larger volumes. When comparing the ET control group to the PD population, there was again a significant difference following short STN-DBS between the two groups (nonparametric t-test; p = 0.001).
DISCUSSION
The present study investigated the effect of varying durations of STN-DBS on cortical and sub-cortical brain areas related to olfactory processing, including the OB, in PD patients. The major findings of this study were that (1) STN-DBS in PD patients is associated with a significant impact on cortical brain areas (specifically, we observed an increase in ipsilateral white-matter volume), (2) STN-DBS does not have observable effects on sub-cortical brain areas (however, there were positive trends noted in thalamic and OFC volumes with duration of stimulation), (3) PD patients treated for longer STN-DBS exhibited significantly larger OB volumes than those treated with short and medium STN-DBS, and (4) the ET control group exhibited significantly larger OB volumes following short STN-DBS compared to PD patients (likely indicative of a lack of olfactory deficit). These data suggest a potential mechanism that could explain recent studies suggesting that STN-DBS may treat some of the non-motor symptoms of PD such as olfactory dysfunction.
With loss of dopaminergic neurons in the substantia nigra, a gray matter strucuture, along with the cognitive imparimement that has been linked to white matter pathology in PD patients, studying global brain changes is imperative to understanding the possible underlying effect and possible limitations of short, medium and long term STN-DBS (Rae et al., 2012).The current literature illustrates the regenerative capabilities of white matter connectivity and gray matter cell body survival associated with DBS. In a study that used a rat model of PD to demonstrate the effects of chronic high frequency stimulation of the STN, it was found that stimulation not only increased cell survival rates, but also restored the dopaminergic pathways between the striatum and subcortical regions of the brain (Khaindrava et al., 2011). Strucutres not normally associated with PD such as the pedunculopontine nucleus of the brainstem have also shown functional neuroplasticity following bilateral DBS in human subjects (Schweder et al., 2010). A subsequent study that looked at global brain matter changes using pre and post-operative diffusion tensor imaging of PD patients that received bilateral STN-DBS found no changes in global brain network organization, but did find changes in local subcortical brain regions related both to sensory, motor, and limbic functions, further supporting the unique neuronal plasticity associated with DBS (van Hartevelt et al., 2014). In comparison to our current study, although we similarly found no global brain changes in both total WM and total GM volumes, there were significant changes in IWM volume, suggesting that DBS can reestablish structural connectivity, which could explain associated functional connectivity responsible for the alleviation of PD symptoms extending beyond that of motor recovery.
Olfactory information is sent from the olfactory receptor neurons to the OB, which further gets sent to the piriform cortex of the primary olfactory cortex, where it sends projections that interconnect with subcortical areas such as the amygdala, hippocampus, orbitofrontal cortex, and the thalamus (Ham et al., 2016; Wilson et al., 2004). Current literature states that disturbances in olfaction that cause anosmia (the absence of smell) and hyposmia (reduced ability to smell) are correlated with decreased number of olfactory receptors, whereas olfactory dysfunction that results in loss of odor identification, discrimination and odor memory are linked to central rather than peripheral mechanisms involving the OB and subcortical brain regions mentioned (Frasnelli et al., 2010).Recovery of olfactory function would imply neurogenesis in the OB. Studies have shown the regenerative capabilities of the OB bulb by way of the subventricular zone, which is embryonic tissue containing proliferative cells that remain in adulthood located in the lateral ventricles (Lepousez et al., 2013; Lledo and Saghatelyan, 2005; Wilson et al., 2004). Furthermore, OB volume has been shown to relate not only to olfactory function, but also to duration of olfactory function loss in a cohort of non-PD patients that suffered post-traumatic or post-infectious olfactory deficits (Haehner et al., 2009). Our study showed a considerable difference in OB volume with length of stimulation, supporting that neurogenesis can in fact happen at the level of the OB that could have possibly been triggered by stimulation in the basal ganglia, which has been reported in hippocampus following DBS in rats (Chamaa et al., 2016).
Despite the demonstration of increased IWM and olfactory bulb volume with STN-DBS in PD patients, there are some limitations to our study. (1) Our sample size was only 30 PD patients, only nine of which underwent long stimulation greater than 360 days. Furthermore, our olfactory deficit control condition (ET patients) was limited to only short stimulation, and the number of subjects was small (n = 3). A larger study that includes a larger sample size of long stimulation patients would provide a clearer picture of the long term effects of STN-DBS. (2) Pre and post smell tests using the conventional UPSIT and Sniffin Sticks Test were not availiable to correlate with changes seen in OB volume with DBS. Identifying functional changes and relating those to changes in subcortical brain regions would strengthen the results found in this study. (3) We used ET patients with no known olfactory deficits as our control since they represented a different movement disorder treated with DBS. This allowed for a comparison between two neurodegenerative disorder groups; however, to normalize the measurements of the OB, using MRI's of healthy control subjects would have aided the results of this study.
Future directions of this study are not limited to but include: (1) gathering a larger sample size of PD patients that have acquired more than 360 days of stimulation, along with a larger sample of ET patients to compare to short, medium and long PD stimulation groups in order to strengthen the results of this study, (2) a sample size of brain MRI's of healthy brains to compare OB volumes of both PD and ET subjects, (3) acquiring pre and post smell tests using UPSIT and Sniffin Sticks Test to compare volumetric changes seen in cortical and subcortical brain regions to functional changes, (4) using olfactory training in PD patients that do not meet the criteria for DBS surgery, and measuring changes in OB volume pre and post training, and (5) using vertex level analysis of subcortical brain areas to inquire if there are any morphological changes in these structures. The most hopeful future direction that could stem from this study is the use of non-surgical methods to recover olfactory function in both neurodegenerative disorder groups, as well as non-neurodegenerative disorder groups that suffer from olfactory dysfunction. Furthermore, studying the neuroplasticity of the subcortical brain structures both volumetrically and morphologically could possibly open doors for future treatment methods aiding in patient recovery.
In conclusion, the significant increase in OB volume provides an anatomical explanation for published literature that has demonstrated positive olfactory responses in PD patients following STN-DBS. An increase in OB volume following STN-DBS not only supports neuronal plasticity, but also demonstrates the effects of STN-DBS on patient recovery that go beyond basal ganglia structures, allowing for a wider scope of its benefits in PD associated olfactory deficits, as well as suggesting the possible success of non-surgical alternatives such as olfactory training in PD patients not applicable for DBS, and non-PD related olfactory deficits. Additionally, the significant positive changes seen in IWM volume, and the trends observed in thalamic and OFC volumes with length of stimulation reinforce the recovery of olfactory function through the possible reestablishment of neuronal connections and structural changes in subcortical regions involved in higher olfactory processing.