One-pass deep brain stimulation of dentato-rubro-thalamic tract and subthalamic nucleus for tremor-dominant or equivalent type Parkinson’s disease

Anatomically, the Vim/DRT as the usual target region for tremor typically lies more superficially and posteriorly (with respect to the MCP reference system) than the STN target region [3] (cf. Fig. 1). Therefore, a parietal approach is required in order to allow traversing (and later modulating) both regions at the same time or differentially. In the typical approach to the STN region anteriorly, the antero-lateral and superior STN is targeted [6]. In this traditional approach, the Vim/DRT region is likely missed anteriorly, since it is located more superficial and posterior. If the traditional approach is used, this region can only occasionally be reached via an electric field that spreads in to this region posteriorly (cf. Fig. 3a, b). Typically, the target for tremor surgery (Vim/DRT) is approached by a separate trajectory anteriorly [3] that will not or only occasionally end in the posterior STN region (pSTR, cf. Fig. 1). However, our simulation studies showed that it could be safely reached on the way to the STN posteriorly. In order to achieve an optimal placement for tremor reduction, it appears that the STN region can be skewered posteriorly, while traversing the DRT in the thalamic region over this parietal route (cf. Figs. 1a; 2c, d; 3c–e). The posterior (parietal) approach can be used to reach the thalamic and the STN region at the same time over one single brain-perforating path. This approach has until today anecdotally been reported as a salvage strategy pathway to reach the STN if in the frontal region an infection has occurred (similar to our case 1) [27]. Our approach here is an evolution from this first description in a different patient group namely TD IPS. For targeting purposes, the STN is directly visualized on T2-weighted MRI sequences. We additionally utilize the DTI FT technology that we already use routinely in daily clinical practice to visualize the DRT [2‐4, 7]. Planning this approach might be demanding, since a three-point trajectory (STN, DRT, entry point) has to be found that safely enters into the target region. However, two of the points are directly visualized with the MRI technology and there will typically be some room to move in the cortical entry zone. More medial approaches will penetrate the ventricles (like in case 2, here, cf. Fig. 4).

More lateral approaches might have the danger of damaging (sub-) cortical eloquent brain regions (Wernicke’s area on the left, etc.). However, there is no literature that supports a higher bleeding risk of this proposed posterior trajectory. From the regulatory point, it is left to the surgeon’s discretion how he approaches the STN. In our two cases, the parietal approach allowed a safe placement of electrodes in both the STN and the DRT.

Some authors discuss additional risks of traversing the ventricle during DBS placement. There is a possible risk of reduced placement accuracy and an increased bleeding risk form traversing ependymal surfaces. Kramer and coworkers recently showed equal accuracy and a similarly low bleeding risk when traversing the ventricle for subthalamic DBS including the use of MER [15]. In the SANTE trial on anterior nucleus DBS for partial epilepsy, only trans-ventricular approaches were used. No symptomatic bleedings (4.5 % incidental hemorrhages on imaging) were detected in 110 cases [10]. While we are aware of possible problems, we believe that with the methods applied (macroelectrode, 15-mm rigid tip of the DBS electrode for accurate placement while traversing the ependymal surfaces; high-resolution T1W MRI to plan a safe trajectory with respect to vascular anatomy) accurate and safe placement is possible.

In the cases described here, a change of the trajectory was not necessary. One could envision that an adjustment of the trajectory might be a difficult task. Certainly an adjustment will not be just a parallel tract. There are multiple restrictions from the anatomical and functional environment. However, stimulation of the DRT teaches us that the tract can be modulated at different parts of it and will still be effective [2, 3]. In this respect, the adjustment would look more like an idealized “ice cone” with a rather fixed tip in the anterior lateral and superior STN [6] and larger adjustment possibilities in the DRT and cortical portion of the trajectory.

There are patients that have electrodes located in Forel’s field above the STN who benefit from electrodes located there with better tremor control than in traditional STN surgery (some researchers actually think that the benefit of STN-DBS stems from stimulation of the hyperdirect pathway (HDP) above the STN in Forel’s field). This occasionally happens, if the DRT region (because of anatomical variations) in the traditional anterior approach is close enough to the more proximal electrodes (Figs. 1b; 2a, b; 3a, b). In the approach chosen, we have operationalized this in two cases and ensured the accessibility of the DRT by choosing a parietal approach.

With this technique, patients with TD or EQT IPS can—depending on their age, and because of the higher incidence of neuropsychiatric side effects in pure STN DBS—receive STN DBS, DRT DBS, or gradually both. This possibly could have advantages during the course of the disease progression because of the adjustability of the therapy (STN for bradykinesia, rigidity, and tremor; DRT for tremor).

There are anecdotal reports of patients with pure thalamic DBS that later received STN-DBS in a second surgery [11, 19]. There is no indication that hints towards detrimental effects of simultaneous stimulation of both targets (DRT and STN) in these reports. The other is also known, patients with STN-DBS that later received thalamic DBS because of a later loss of tremor control over the course of the disease. According to our limited experience, the combined approach presumably would allow for a better symptom control in TD and EQT IPS.

Interestingly, we have observed a very low but nevertheless effective stimulation amplitude in both target regions. These low amplitudes correspond to a dramatic clinical improvement and a clear reduction of the LED. We are not quite sure what the reason is but it might be that modulating fiber structures from another angle than usually (in the anterior approach) might prove be superior.

We have extensively discussed the limitations of the fiber tractography approach to DBS (anatomic rendition of fibers with DTI, crossing and kissing fibers) and the limitations of the deterministic DTI technology in our previous publications [2‐4, 7]. The deterministic approach is prone to a higher subjectivity than other approaches (e.g., probabilistic tractography). The extension of a fiber tract very much depends on the software used and on a variety of other factors (region of interest, algorithm, multiple tensor models, etc.) [1]. The “anatomical truth” of a fiber tract like the DRT depends on the definition of a variety of parameters and on the imaging data it originates from (e.g., most deterministic approaches will not show a pontine decussation of the DRT). Moreover, it is difficult to determine a “true accuracy” of the deterministic approach [5, 22]. In a recent study, we found a level of agreement to be ±2.5 mm when comparing electrophysiological distance measurements with DTI. In our experience, however, the accuracy of the method is higher [18]. There are examples in the literature for detrimental results of uncritical use of deterministic tractography in conjunction with neuro-navigation while ignoring electrophysiology [13]. However, when cautiously applying this technology, it can be safely used for navigation and targeting purposes [2, 3, 5, 7].

Certain limitations, primarily owing to the case study character of this report, have to be discussed: This report shows that a one-pass threading of two target regions (STN and DRT) is possible without obvious detrimental effects to the patients. Tremor could more optimally be investigated with a more detailed tremor rating (e.g., Fahn–Marsden Tremor rating scale) and with blinded raters using video rating. We perform this in our clinical study. We have shown that tremor and, to a certain extent other IPS symptoms, can be positively influenced with stimulation of these targets. We have, however, not differentially tested the effects of singular STN or DRT stimulation via this route. This certainly would be an interesting question. The focus in the treatment of the two patients was the alleviation of clinical symptoms of IPS. It would, however, have been interesting to see to what extent dopaminergic medication could be decreased depending on the target stimulated and if stimulation of the STN and DRT over the parietal approach could be considered in much younger patients with EQT IPS that might later develop strong fluctuations. We cannot address these questions with our results here. This and the question if a combined stimulation of both regions is superior to stimulation of single structures and the effect on the patients’ quality of life is the focus of a clinical study that we have just initiated and which is currently recruiting.