Deep Brain Stimulation patient-selection pathway for Parkinson's disease — відбір кандидатів на глибоку стимуляцію мозку при Паркінсоні
Deep brain stimulation (DBS) does not treat "parkinsonism" in general — it treats idiopathic Parkinson's disease (PD) specifically. The single most powerful predictor of a good surgical outcome is a robust, sustained response to levodopa. Patients with atypical parkinsonian syndromes (progressive supranuclear palsy, multiple system atrophy, corticobasal degeneration) share superficial motor features with PD but lack the underlying dopaminergic-responsive circuitry that DBS depends on, and are considered poor candidates.
DBS of the subthalamic nucleus (STN) or globus pallidus interna (GPi) works by modulating the same basal ganglia circuitry that dopamine replacement modulates pharmacologically. As a practical rule, DBS cannot be expected to improve a motor sign that levodopa itself does not improve.
Clinical rule of thumb: any parkinsonian feature that responds to levodopa (bradykinesia, rigidity, resting tremor, gait freezing that is levodopa-responsive) is likely to respond to DBS as well. Features that do NOT respond to levodopa — postural instability, speech/swallowing difficulty, levodopa-resistant freezing — generally will not improve, and may even worsen, with stimulation.
This is why every DBS work-up begins with a structured levodopa challenge rather than a simple bedside impression: it converts "does this patient look like they have PD" into a quantitative, reproducible predictor of surgical benefit.
A patient with dramatic levodopa-responsive tremor and bradykinesia, even if fluctuating badly, is a far better DBS candidate than a patient with only mild, non-fluctuating symptoms and a weak levodopa response — the fluctuation itself is not the criterion; the underlying dopaminergic-responsiveness is.
Progressive supranuclear palsy (PSP): early falls, vertical gaze palsy, axial rigidity greater than limb rigidity, poor/absent levodopa response. DBS outcomes are poor and complication rates disproportionately high.
Multiple system atrophy (MSA): early and severe autonomic failure (orthostatic hypotension, urinary dysfunction), cerebellar or pyramidal signs, rapidly progressive course. Levodopa response is typically modest and short-lived; DBS is not indicated.
Corticobasal degeneration (CBD): asymmetric rigidity/apraxia, alien-limb phenomena, cortical sensory loss, negligible levodopa response.
Because early atypical parkinsonism can superficially resemble PD, most DBS centers require a minimum disease duration (typically 4–5 years) and repeated clinical review before final surgical candidacy is confirmed — allowing red-flag features time to emerge if present.
As Parkinson's disease progresses, the therapeutic window of oral levodopa narrows. Patients develop "on-off" fluctuations — unpredictable swings between good mobility ("on") and return of parkinsonism ("off") — and often disabling peak-dose dyskinesia. When these motor complications occur despite optimized medical therapy and can no longer be smoothed out by adjusting oral medication, DBS becomes the standard next step.
Early Parkinson's disease responds smoothly to levodopa because a large reservoir of surviving dopaminergic terminals buffers each dose, releasing dopamine steadily between administrations. As the disease progresses and terminals are lost, this buffering capacity disappears — plasma levodopa levels translate almost directly into striatal dopamine levels, and motor state begins to track the pharmacokinetic curve of each individual dose.
The result: predictable "wearing-off" before the next dose, then later unpredictable, sudden transitions between "on" (good mobility, but often complicated by dyskinesia) and "off" (return of bradykinesia, rigidity, tremor, freezing). Patients may describe "off" periods as feeling like the disease is suddenly, briefly, far more advanced than it really is.
Dyskinesia — involuntary choreiform or dystonic movements — typically appears at peak dose, reflecting pulsatile (non-physiological) striatal dopamine stimulation. Paradoxically, dyskinesia is a marker of preserved levodopa-responsiveness, not treatment failure.
Before DBS is considered, standard medical strategies should be tried: increasing dosing frequency, extended-release or continuous-infusion formulations, COMT and MAO-B inhibitors to extend levodopa half-life, and dopamine agonists to smooth striatal stimulation.
DBS is indicated specifically when these strategies are no longer sufficient — when the fluctuation burden remains disabling despite a maximally optimized regimen, or when the medication burden itself (frequent dosing, side effects, cost) has become unsustainable.
A useful clinical threshold used by many movement-disorder centers: ≥2–3 hours per day of troublesome "off" time, or dyskinesia significant enough to interfere with daily function, despite optimized pharmacotherapy.
DBS does not cure Parkinson's disease or halt its progression — it is a symptomatic therapy that restores a much larger share of "on" time with fewer fluctuations and, in most patients, allows a substantial reduction in total levodopa-equivalent daily dose.
DBS is a powerful intervention, but it is not neutral with respect to cognition and mood. Formal neuropsychological testing and psychiatric assessment are mandatory parts of every work-up, because significant dementia or uncontrolled psychiatric illness both predict poor outcomes and elevated risk of post-operative cognitive or behavioral decline.
Significant dementia is a relative contraindication to DBS for two connected reasons. First, patients with major cognitive impairment are less able to tolerate the stimulation-programming process, which requires reliable self-report over repeated visits to titrate settings safely. Second, and more importantly, DBS — particularly of the STN — carries a measurable risk of accelerating cognitive decline in patients who already have limited cognitive reserve, through a combination of surgical microlesion effect, stimulation-related changes in fronto-striatal circuits, and reduction of dopaminergic medication after surgery.
A structured neuropsychological battery (not just a bedside MMSE/MoCA) assesses executive function, processing speed, memory, and visuospatial domains, since subtle frontal-executive impairment — common even in cognitively "normal-appearing" PD patients — is itself a modest risk factor for post-operative decline.
Uncontrolled depression, anxiety, or psychosis must be identified and stabilized before proceeding. Depression is common in PD (up to 40% of patients) and, if untreated, is associated with worse quality-of-life outcomes after DBS regardless of motor improvement. Impulse-control disorders (linked to dopamine agonist therapy) and any history of suicidality require specific psychiatric clearance.
STN-DBS in particular has been associated in some patients with post-operative mood or behavioral changes (apathy, impulsivity, occasionally hypomania), thought to relate to inadvertent modulation of limbic circuits adjacent to the motor STN. A thorough pre-operative psychiatric baseline is therefore both a screening tool and a reference point for interpreting any post-operative changes.
The screening principle is symmetric: DBS is withheld not because these patients would not benefit motorically, but because the net effect on their overall function and quality of life — factoring in cognitive and psychiatric risk — may be negative even with excellent motor improvement.
The levodopa challenge test is the single most quantitative and predictive study in the DBS work-up. After an overnight withdrawal of dopaminergic medication ("off" state), the patient is scored using the UPDRS-III motor scale, given a supra-threshold dose of levodopa, and re-scored at peak benefit ("on" state). The percentage improvement is the strongest available predictor of DBS motor outcome, and also informs the choice between the two dominant surgical targets.
Protocol: dopaminergic medications withheld for ≥12 hours (overnight) → UPDRS-III motor exam performed in the practically-defined "off" state → a supra-threshold dose of immediate-release levodopa given (typically 1.5× the patient's usual morning dose) → UPDRS-III repeated at peak clinical benefit (~45–60 minutes later, confirmed by patient/examiner report of "on").
% improvement = (UPDRS_off − UPDRS_on) / UPDRS_off × 100
Interpretation: ≥30% improvement is the conventional minimum threshold used by most DBS programs; ≥50% improvement is associated with the best post-operative motor outcomes. A poor levodopa response (<30%) predicts a poor DBS response and should prompt re-examination of the diagnosis rather than proceeding to surgery — with the specific exception of levodopa-resistant tremor, which can respond well to DBS even when other signs do not.
Subthalamic nucleus (STN): the most commonly used target. Produces robust improvement in bradykinesia, rigidity and tremor, and characteristically allows the largest reduction in levodopa-equivalent daily dose after surgery — attractive for patients whose main problem is medication burden or dyskinesia driven by high drug doses. Carries a somewhat higher signal for mood/cognitive side effects in some studies, particularly in patients with baseline vulnerability.
Globus pallidus interna (GPi): produces comparable motor benefit and is directly anti-dyskinetic in its own right (useful when dyskinesia is prominent even without large dose reduction), generally allows less medication reduction, and is often preferred in patients with mild cognitive concerns, older age, or pre-existing mood disorder, where a more cognitively-neutral profile is desirable.
Target selection is therefore individualized: it weighs the levodopa challenge result together with the cognitive/psychiatric profile from Stage 3 and the patient's specific symptom priorities (dose reduction vs. maximal anti-dyskinetic effect).
Large randomized trials (e.g. the NIH COMPARE and CSP-468 studies) found STN and GPi produce similar overall motor benefit, but with STN allowing greater medication reduction and GPi carrying a more favorable neurocognitive/mood profile — making target choice a personalized decision, not a one-size-fits-all rule.
Bringing together diagnostic confirmation, fluctuation burden, cognitive/psychiatric clearance, and the quantitative levodopa challenge, the multidisciplinary DBS team reaches a final candidacy decision. Good candidates — strong levodopa responders with disabling fluctuations/dyskinesia and preserved cognition — can expect a substantial, durable reduction in daily "off" time and a meaningful reduction in medication burden, transforming quality of life even though the underlying disease continues to progress.
The multidisciplinary team (movement-disorder neurologist, neurosurgeon, neuropsychologist, psychiatrist) integrates four independent lines of evidence:
1. Diagnostic certainty — idiopathic PD, not atypical parkinsonism 2. Disabling motor fluctuations/dyskinesia despite optimized medical therapy 3. Preserved cognition and stable psychiatric status 4. A quantitatively strong levodopa challenge response (≥30%, ideally ≥50%)
No single criterion is sufficient alone — a patient can have a strong levodopa response but be excluded by significant dementia, or have preserved cognition but be excluded by a weak levodopa response suggesting an atypical process. The decision is a composite judgment, weighing expected motor benefit against cognitive/psychiatric risk for that specific patient.
In well-selected patients, randomized controlled trials and long-term registries consistently show:
• "Off" time reduced by roughly 4–6 hours per day, with a corresponding increase in good-quality "on" time without troublesome dyskinesia • Levodopa-equivalent daily dose reduced by approximately 30–50%, most pronounced with STN targeting — reducing pill burden, dosing-related side effects, and dyskinesia driven by high drug exposure • Meaningful, durable improvement in disease-specific quality-of-life measures (e.g. PDQ-39), sustained for 5 years or more in many series • Tremor, when levodopa-responsive, often shows the most dramatic and immediate improvement of any motor sign
What DBS does not do: it does not slow or halt the underlying neurodegenerative process, and axial features that are not levodopa-responsive (postural instability, some gait freezing, speech and swallowing difficulty) typically continue to progress despite stimulation — which is precisely why the pre-operative levodopa challenge, not just clinical impression, remains the decisive predictive test.
The central clinical lesson of DBS patient selection: DBS amplifies and stabilizes what levodopa can already do — it is not a rescue therapy for signs levodopa cannot touch, and not a safe option when cognitive or psychiatric vulnerability outweighs the expected motor gain.