13Invasive Neuromodulation (DBS, Cortical, Invasive VNS)

FWhy go inside the brain at all?

Chronic tinnitus is increasingly understood as a network problem rather than an ear problem. After auditory deprivation, central structures raise their gain and a distributed circuit — auditory cortex, thalamus, parahippocampus and limbic hubs — locks into a self-sustaining pattern that the brain reads as sound [2013]. Non-invasive tools such as repetitive transcranial magnetic stimulation and bimodal devices try to nudge this circuit from the outside and are covered in an earlier chapter.

Invasive neuromodulation asks a different question: if the network is the problem, can a permanently implanted electrode reach a node that surface stimulation cannot, and hold it in a quieter state? The appeal is precision and continuous, titratable delivery. The cost is a craniotomy or an implanted pulse generator, with all the attendant surgical and ethical weight that places on a non-life-threatening symptom [2011].

TImplanted auditory-cortex stimulation

The earliest invasive attempts grew out of imaging studies that localised tinnitus-related hyperactivity to the contralateral auditory cortex. De Ridder and colleagues used a functional-imaging target to place an epidural or subdural electrode over the auditory cortex, driven by an implanted pulse generator [2008].

Results were instructive rather than triumphant. Pure-tone (narrowband) tinnitus responded best, sometimes dramatically, whereas noise-like tinnitus responded poorly. Benefit often faded over months, prompting attempts at burst stimulation and, later, recognition that targets such as the parahippocampal area might matter as much as the cortex itself [2011]. Auditory-cortex stimulation remains a small case-series technique, not a standard offering.

CDeep brain stimulation: borrowed targets and a Phase I trial

The first human signals came serendipitously: patients implanted for movement disorders or other indications sometimes reported that their tinnitus changed when the stimulator was on. This pointed attention to deep targets within the auditory and striatal-limbic systems [2015].

The most deliberate human work targets the caudate nucleus — specifically a body region (the so-called area LC) implicated in gating tinnitus loudness. A formal Phase I safety trial of caudate DBS demonstrated that the procedure was feasible and that stimulation could modulate tinnitus in some participants, while underscoring that this remains strictly investigational [2020]. Thalamic and inferior-collicular targets have been proposed largely from animal and modelling work; human evidence there is thin [2015].

TInvasive vagus-nerve stimulation paired with tones

A mechanistically distinct strategy uses an implanted cervical vagus-nerve stimulator, not to suppress the network continuously, but to drive targeted plasticity. The principle, established in a rodent model, is that pairing a brief burst of vagal stimulation with specific tones releases neuromodulators (acetylcholine, noradrenaline) that re-tune the auditory cortical map away from the tinnitus frequency [2011].

Translated to humans, De Ridder and colleagues reported a small case series in which paired VNS-and-tones produced clinically meaningful improvement in roughly half of treated patients [2014]. A subsequent double-blind randomised pilot by Tyler and colleagues found the therapy safe but with only a subgroup showing benefit, and the controlled effect on handicap scores was modest [2017]. This is the same logic that the later, non-invasive transcutaneous and bimodal devices exploit without surgery.

CRisk, evidence level and the ethics of last resort

Every technique here carries the risks of intracranial or neck surgery: haemorrhage, infection, hardware failure, lead migration, and for VNS, hoarseness, cough and bradyarrhythmia. None has Level I evidence for tinnitus; the literature is case series and a single small randomised pilot [2017].

Because tinnitus is not life-threatening, the bar for implanting hardware is high. Candidates should have severe, validated, refractory distress, intact psychological screening, realistic expectations, and enrolment in a research protocol with ethics oversight. The honest counselling message is that invasive neuromodulation is experimental, that benefit is partial and may wane, and that the patient is accepting real surgical risk for an uncertain gain [2013].