Technological Advances in Deep Brain Stimulation

Interleaving

Interleaving stimulation has been available on the Medtronic Activa range since 2008 and allows delivery of independent stimulation programs to two contacts in the quadripolar DBS lead with different values for voltage and pulse width but with the same frequency (Fig. 1). This allows finer control over the stimulation field by enabling stimulation of adjacent targets with different energies and thus potentially help maximise therapeutic effects whilst minimizing undesired side effects. However, despite case reports [5–7] and small case series [8] highlighting benefit, overall no conclusive benefits or large scale studies have proven this method of stimulation to be superior to standard strategies. On the contrary, increased battery drainage and inability to change the frequency of stimulation limits this technique [1].

Electrode contact design

Boston Scientific (Boston Scientific Corporation, Natick, Massachusetts, USA) have European CE approval (2014) for their Vercise DBS lead which houses 8 contacts (compared with the traditional quadripolar electrodes) spanning 15.5 mm (Fig. 2) in length to allow for greater stimulation options and the ability to potentially target multiple nuclei along the trajectory [9]. The system is unique in allowing independent current settings for each of the eight contacts on the lead referred to as ‘multi-independent current controlled (MICC)’ stimulation and also accommodates stimulation with pulse widths below 60 microseconds and the capability to offer independent frequency adjustments to either side of the brain, or in separate areas along a single lead. Computer modelling studies have shown that this approach can achieve superior activation of target neuron fibres with decreased activation of bordering eloquent fibres such as the internal capsule [10]. To address these queries, there are two on-going trials investigating the Vercise system in patients with PD. (i) The VANTAGE trial is investigating if the flexibilities in variation of stimulation parameters at different electrode contacts would improve outcome whilst decreasing side effects. Preliminary results in 40 patients at six months have shown significant improvement in UPDRS III of 62.2% (P <  0.0001), which was sustained at one year [11]. (ii) The CUSTOM-DBS trial involved 15 patients who were previously implanted for Parkinson’s Disease. The subjects were programmed unilaterally at test pulse width of 30 microseconds and compared with control of 60 microseconds with both the subjects and evaluating neurologists blinded to the settings. Overall, an improved therapeutic window was noted with 30 microseconds (P = 0.0009) suggesting that it may lead to improved side effect profile. Efficacy mapped by UPDRS-III and quantitative outcomes were not different between the two parameters, though as the shortened pulse width required less electrical energy it was postulated that there may be advantageous energy settings [12]. Long term outcomes from both trials are awaited to confirm sustainabilityof outcomes.

Whilst both interleaving and multi-source stimulation represent progress in shaping the field of stimulation, as both emit a spherical field, neither method offers true field sculpting/shaping and thus represents a limitation that looks likely to be surpassed by more novel designs of current steering.

Current steering

The mainstream of currently implanted DBS systems are engineered using annular electrode contacts which produce a spherical field of stimulation. On the contrary, target nuclei for stimulation are not spherical so whilst increasing the stimulation can deliver therapy to a target for a sub-optimally placed DBS lead, this will also increase the risk of side effects to bordering structures. It is estimated that stimulation related side effects can occur in up to 15–30% of patients [13]. Novel electrode contact designs are in development which allow true steering and sculpting of the field and should enable an improved therapeutic window with reduced side effects:

(i) Aleva Neurotherapeutics (Lausanne, Switzerland) have reported their pilot results from their novel quadripolar lead, ‘directStim’ which is compatible with all DBS systems. Each ring electrode is split into three independent angular electrodes that can deliver the stimulation at 0°, 120° and 240° directions. A double-blind pilot study utilising the directStim electrode compared the effect of multidirectional stimulation to omnidirectional stimulation. Testing was performed intraoperatively in the STN of 11 patients with Parkinson’s disease and in the ventral intermediate nucleus (VIM) of the thalamus of two other subjects with essential tremor. At the trajectory chosen for implantation of the definitive electrode, the current threshold window between positive and side effects, defined as the therapeutic window was recorded. All but one patient showed a benefit of directional stimulation compared to omnidirectional and a best direction of stimulation was observed in all the patients. Compared to omnidirectional direction, the therapeutic window in the best direction was 41.3% wider (P = 0.037), with the current threshold producing meaningful therapeutic effect 43% lower than in omnidirectional stimulation (P = 0.002). The observed side effects related to direction of stimulation were consistent with the anatomical location of surrounding structures [14].

(ii) St. Jude Medical (St Paul, Minnesota, USA) are due to launch their new multidirectional lead called Infinity in 2015 which consists of a cylindrical ring electrode at each end of a quadripolar lead. The middle two contacts are sectorised into 3 directional electrodes. Stimulation to each electrode will be independent and adjustable (Fig. 3).

(iii) Sapiens (Eindhoven, The Netherlands), recently acquired by Medtronic have developed the SureStim lead which consists of 40 fully configurable (Fig. 4), evenly distributed contacts on the lead which can be activated in specific groups. This allows for very precise sculpting of the field of stimulation from lateral and rotational movements to conforming to complex shapes (Fig. 5), thus theoretically having the advantage of being able to precisely target sub-structures e.g. motor component of the STN and also allow for suboptimal placement. An additional advantage of the system is the ability to measure Local Field Potentials (LFP) intra-operatively [1].

This lead has been trialled in the ‘First Acute in Man Evaluation’ (FAME) FiH study on safety and performance of the lead. Eight patients with Parkinson’s disease scheduled for STN-DBS had a SureStim lead temporarily inserted during regular surgery for stimulation. Stimulation in ring mode (12 electrodes in 3 consecutive rings) and steering mode (4 electrodes,blinded, random) (Fig. 6) and LFP recordings were performed on all electrodes. The study demonstrated equivalency relative to commercial lead in the ring mode and proof of concept was demonstrated in the steering mode with statistically significant effect of steering on onset of therapeutic effects and side-effects. LFP recordings provided clinically relevant information for placement validation and programming guidance [15].

Current steering electrodes represent true innovation, especially when combined with other closed-loop technologies (LFPs, electrochemical detection) and looks set to represent the next generation of DBS leads. This has potential advantages over traditional DBS leads by being more versatile, precise and efficient, thus allowing an improved therapeutic window whilst causing fewer side effects and complications. Other advantages include reduced unnecessary reoperation due to lead misplacement or migration and decreased power requirements leading to improved IPG lifespan.

Constant current versus constant voltage

Constant current devices, as opposed to constant-voltage devices, represent an innovation in the way that electricity is delivered to targets. Stimulation fields produced by traditional voltage-driven devices are susceptible to changes in size caused by changing tissue impedance potentially providing a less stable field of stimulation which may have clinical implications. Constant-current devices automatically adjust the voltage as necessary for different impedances to maintain the same electrical current. The distinction may be important, as it is the electrical current delivered by the DBS pulse that determines the activation of neural elements [16, 17]. The varying levels of tissue impedance, especially in the days/weeks following implantation [18], imply that a constant-current solution may be more attractive than constant-voltage, although clinically this is yet to be proven [19]. Whilst new companies to the market offer only constant-current devices, the Medtronic Activa range is able to deliver both constant-voltage and constant-current stimulation.

A prospective, randomised, multi-centre trial described the use of constant-current STN-DBS in 136 patients and confirmed good quality ‘on’ time in the stimulation group compared with a control group in whom stimulation was delayed for 3 months after DBS implantation [20]. However, though this study noted a safe and efficacious role for constant current stimulation, it did not provide a direct comparison with voltage-driven stimulation. A more recent non-randomised study comparing constant current versus constant voltage pallidal stimulation in primary or segmental dystonia noted a statistically significant improved outcome in the BFMDRS-M (Burke_Fahn_Marsden Dystonia Rating Scale) score (P = 0.0324) and for the BFMDRS-D (P = 0.0478) in favour of the constant current cohort at 12 months [21]. However, whilst theoretical evidence lends support to constant current devices over constant voltage ones, clinical data from a head-to-head randomised controlled trial is necessary to validate this.

Scheduled stimulation

Classic DBS for movement and neuropsychiatric disorders is delivered using a chronic and continuous stimulation paradigm. Okun et al. [22] have provided proof of concept that scheduled intermittent stimulation has the potential to improve motor and vocal tics in Tourette syndrome. Whilst in their study continuous stimulation outperformed scheduled stimulation, the latter paradigm may be useful in paroxysmal conditions in personalising the therapy to the frequency and duration of the particular disorder and may also improve battery life. Larger and longer follow-up studies however are required to verify benefit of this technique, which may be of limited use in traditional movement disorders that are persistent and continuous in nature.

Adaptive stimulation

Symptoms in patients with movement disorders even post-DBS implantation are subject to fluctuations which may be related to underlying pathology of the disease, bioavailability of oral medications, or variable psychological/cognitive states. Despite this dynamism, traditional DBS systems deliver stimulation continuously and are set by the physician according to clinical response on a trial and error basis, also referred to as open loop DBS (OLD). Limitations to OLD include lack of responsiveness to patients’ needs, fixed therapeutic window, increased battery usage and reduced battery life, repeated hospital visits for stimulation adjustment, and ultimately a less optimal and more expensive therapy. Thus the ability to record a biomarker of physiological activity and adjusting the delivery of electricity based on the demand, termed ‘close loop/adaptive DBS’ is very appealing.

Closed loop DBS (CLD) utilises this intelligent biofeedback mechanism in order to adjust the delivery of stimulation. The DBS biomarkers that have been employed can either be internal brain biomarkers (e.g. beta bands in the STN, neurotransmitters, electro-corticography) or external biomarkers which measure change of physical behaviour recorded with wearable sensors (e.g. accelerometers) [23].

A recent proof-of-concept study focussed on the use of a brain–computer interface in order to detect LFPs directly from the STN of PD patients, thereby timing the stimulation exactly to when it is required. Monopolar stimulation at 130 Hz was applied in 8 patients in whom β-wave activity was detected. A significant improvement in UPDRS was found in adaptive DBS as opposed to continuous DBS with half the energy expenditure required in standard DBS [24].

Activa Primary Cell + Sensing (Brain Radio) developed by Medtronic represents a significant milestone in the journey towards closed-loop stimulation. It is capable of detecting, measuring and recording brain signals (LFP), acting as biomarker to allow for adjustment of stimulation. A limited number of devices were released in 2013 for specific research programmes, the results of which are currently awaited.

Alternative electrochemical biomarkers for CLD involve detection of neurotransmitters such as dopamine and serotonin as has been demonstrated in animal models [25]. In humans it has been shown that the impact effect associated with acute VIM-DBS implantation in tremor is not only related to microlesioning effect, but also to immediate adenosine neuro-chemical release [26]. This would suggest that integration of neuro-chemical feedback to a closed-loop system could prove beneficial, though translation into clinical practice is likely to be some time away.

Robust clinical proof for of the benefits of chronic stimulation of closed-loop technology remains to be established. Nevertheless, it has the potential to offer several advantages over open loop DBS including efficiency and efficacy of therapy, reducing IPG changes or recharging, eliminating lengthy start-up periods for programming and adjustment, reducing hospital visits, providing a personalised treatment, and making parameters setting automatic and adaptive [23].

Rechargeable IPGs

Until recently IPGs were non-rechargeable with a lifespan of about 3–5 years depending on the use. This relatively short duration can lead to problems such as unpredictable expiration and recurrent surgeries to replace depleted batteries. The latter is associated with a three times greater risk of infection than de novo implantation of the IPG [27]. Additionally, the relatively larger design of such batteries can cause discomfort and cosmetic concerns, especially in the paediatric patients. To overcome these issues, the first rechargeable DBS device (Medtronic Activa Rechargeable Cell) was introduced in 2008: smaller, lighter and intended to function for 9 years. Recharging occurs through an induction-powering process which utilises an external power source to induce a flow of current into the battery by means of the current flowing into an external inductor [28]. Though the lifespan of the rechargeable device is significantly improved, the recharging system is bulky and requires regular charging, typically at least once a week, to prevent depletion. However, if the charge is allowed to deplete below the minimum level required to support telemetry, in the case of the Activa RC it will shut down and become unresponsive, termed over-discharge. It can be over discharged twice but if it occurs a third time, the IPG has to be surgically replaced. Hence, active involvement of the patient and carer is a necessity for this device. Transient recharging problems are the most common complications (36% of cases). Overall, patient/carer satisfaction with this device remains high and adverse events associated with it low[29, 30].

Brio (St. Jude Medical) is currently the smallest rechargeable IPG on the market with a displaceable volume of 18cc. It has the added advantage of an enhanced antenna system which allows a deeper implant of up to 2.5 cm, offering the potential of increased patient comfort and better cosmetic profile. Like the Medtronic RC, it also houses an integrated protection system against over discharge, however, if a completely depleted battery is not recharged within 30–90 days, it has to be surgically replaced to resume stimulation.

Alternative recharging technology of Vercise, introduced by the Boston Scientific, with up to 25 years of battery life and less frequent recharging need, is also likely to increase the uptake of this technology [11]. In an effort to reduce battery depletion damage, Boston Scientific has introduced the Zero-Volttrademark technology. This features a titanium anode which unlike copper does not dissolve at 0 V, thus guaranteeing the battery is not damaged in the case of depletion. This impacts neither performance nor capacity whilst minimising the risk of premature device failure. To strengthen this, the manufacturers assert that the warranty is not void in case of depletion.

Wireless communication

The Vercise system (Boston Scientific Ltd) is the first and currently the only system to offer wireless recharging, allowing patients to be active while recharging. Further, adjustments to stimulation is possible at a range of 45 cm via a wireless remote control that connects to the IPG, allowing for greater patient comfort.

St. Jude Medical, in 2015, are in the process of launching an IPG with wireless Bluetooth communication ability. This will be accessed and programmed using an iPAD mini (clinician) or iPod touch (patient) with pre-loaded software that is upgradeable. Electronic hardware and software design has lagged behind in comparison to other technological industries (e.g. mobile phones), thus this represents a much needed step to develop technology that is easy to use, intuitive and versatile.

A potential advantage of the wireless communication system would be the ability to allow physicians to program systems remotely in response to performance data uploaded via secure networks to the healthcare providers. Additionally, in the intra-operative phase, when performing test stimulation or impedance measurements, non-contact testing could be performed from a distance potentially reducing the risk of infection or disturbance to the surgeon. This could also prove an advantage performing telemetry in patients with complex movement disorders (e.g. paediatric dystonia). Challenges of remote communication, namely security, immunity to interference, reliability and ability for a back-up method will determine the ultimate success of this method.

Miniaturised and wireless powering

Despite progress in technology which has paved the way for devices at the scale of a millimetre or less “microimplants”, miniaturisation of the power source for DBS, has remained a challenge. RNS Neurostimulator is a skull mounted device developed by NeuroPace for the use in epilepsy. A miniaturised skull mounted IPG for DBS would remove the morbidity (e.g. pain and infection) associated with tunnelling currently necessary for placing the IPG in a chest/ abdominal pocket. Wireless powering would advance this to the next level. Although wireless powering has been demonstrated, energy transfer beyond superficial depths in tissue has so far been limited by large coils (at least a centimetre in diameter) unsuitable for a microimplant. Recent developments in midfield powering [31] are able to create a high-energy density region deep in tissue inside of which the power-harvesting structure can be made extremely small. Unlike conventional near-field (inductively coupled) coils, for which coupling is limited by exponential field decay, a patterned metal plate is used to induce spatially confined and adaptive energy transport through propagating modes in tissue. It has been demonstrated that power can be transferred to a deep-tissue (>5 cm) microimplants for both complex electronic function and physiological stimulation [32]. Theoretically this should enable new generations of implantable systems without bulky charging systems or extension cables that can be integrated into the body at reduced cost and risk.

Advances in imaging

Over the last decade there has been a shift from primarily clinical and physiological mapping targeting to primarily image guided direct targeting of nuclei. This has evolved with imaging techniques such as susceptibility weighted imaging which better delineates the border between the STN and substantia nigra compared to standard images [38]. Even the normally MRI-indistinguishable VIM has been directly targeted using a novel WAIR sequence [39]. Diffusion Tensor Imaging (DTI) has been used to create probabilistic stimulation atlas based on the proximity of effective DBS leads to white matter tracts [40]. This combined with DTI’s ability to defining subregions of the DBS target and it’s projections to subcortical targets [41] and associated projections neurons has allowed parcellation of target nuclei into functional zones [42] and may form the basis for future targeting reducing the need for detailed intra-operative physiologicaltesting.

Frameless stereotaxy

Frameless implantation of DBS lead with NexFrame (Medtronic) which utilises optical tracking has been available for a number of years with accuracy rates initially reported to being comparable to frame-based techniques [43]. Despite this, there has been a distinct lack of uptake for this technology. Moreover recent studies have demonstrated overall lower targetting accuracy for the frameless system especially in the medial-lateral direction despite improved euclidean accuracy compared with frame based approaches [44]. Advances on this system include the StarFix MTP system which is a custom-built, miniature stereotactic platform mounted onto bone anchored fiducial markers which serve as both image reference points and anchors for a custom-made stereotactic frame. It has the advantage over NexFrame in that the trajectory planning is built into the platform. The disadvantage however is that it takes 3 days for the manufacture of customised stereotactic platform during which period the patients are sent home with fiducials in-situ. The targeting error with this system was reported to be on average 1.99 mm (SD 0.9) [45].

Intra-operative imaging

Advances in imaging allowing direct targeting have paved the way for intra-operative radiological control for lead implantation with intraoperative CT (iCT) or MRI (iMRI). Current iCT systems available include the O-arm (Medtronic, Minneapolis, MN), a flat-panel cone-beam CT unit [46] and more recently CereTom (Neurologica Corporation, Danvers, MA) a fan-beam portable system both used for 3D lead position verification. Holloway et al reported an accuracy of final lead position of 2.04 mm using the O-arm which was not significantly different from normal CT-based registration at 2.16 mm [47]. Burchiel et al. reported a mean trajectory deviation error of 1.24 mm in 60 implanted patients using CereTom with the NexFrame system [48]. The portability of iCT, cost and adaptability for other uses (pedicle screw cannulation in spine fixation) makes this, rather than iMRI, a more attractive option for intra-operative lead verification. As iCT verification requires fusion with pre-operatively acquired MRI, one drawback is that error in accuracy may be introduced as result of the fusion [49].

iMRI provides real time immediate and definite control for lead implantation and removes the inaccuracies associated with CT fusion. A number of centres [2] have combined iMRI with skull mounted devices (e.g. ClearPoint MRI, Fig. 9) with very good reported accuracy rates for final lead position of 0.6±0.5 mm from intended target, no serious adverse events and clinical outcome comparable to frame based techniques [50].

Regardless of the modality used, there are clear advantages for intra-operative image verification of lead position (i) Immediate real time verification of lead position, thus allowing immediate re-positioning if needed (ii) shorter operative times and improved patient safety by removing the need to transport the patient outside the operative theatre (iii) Monitoring for intracranial complications e.g. haemorrhage.

Robotics

The use of a robot to implant DBS electrodes was inherent to the original technique introduced by Prof A.L. Benabid at the dawn of DBS. There is now a resurgence of interest in robotics. Robot assisted implantation of DBS with accuracy similar to frame-based methods have recently been reported. The Neuromate robot (Renishaw PLC) (Fig. 10) is used for DBS implantation with von Langsdorff et al. recently reporting a mean in-vivo accuracy in 17 patients of 0.86 mm (±0.32 mm SD) [51], comparable with frame-based procedures (assuming a perfectly calibrated frame). The ROSA^™ (Fig. 11) has a robotic arm with six degrees of freedom and is capable of non-invasive laser measurement for patient registration allowing frameless intervention, though currently is not accurate enough for DBS. Lefranc et al have reported their experiences with frame-based, iCT verified, robot assisted DBS implantations [52]. Despite these advances, robot assisted implanations currently still requires the procedure to be frame-based to maintain accuracy. This has been confirmed with in-vitro studies which have demonstrated superior accuracy rates in favour of frame-based robot assisted compared with frameless robot assisted surgery [53]. Thus the next challenge in this field will be to remove the need for a stereotactic frame whilst maintaining accuracy of lead position.