← Back

CSP-468 STN vs GPi

Year of Publication: 2010

Journal: New England Journal of Medicine

Link: https://doi.org/10.1056/NEJMoa0907083

PDF: https://www.nejm.org/doi/pdf/10.1056/NEJMoa0907083

Clinical Question

Does pallidal (GPi) or subthalamic (STN) deep-brain stimulation produce better 24-month motor and non-motor outcomes in advanced Parkinson's disease?

Bottom Line

In 299 patients with advanced PD randomized to bilateral GPi or STN DBS, 24-month change in UPDRS-III motor score (off-medication, on-stimulation) was similar (-11.8 GPi vs -10.7 STN; p=0.50). STN allowed greater levodopa reduction (408 vs 243 mg; p=0.02), but worsened visuomotor processing speed (p=0.03) and depression (p=0.02). Established GPi and STN as both effective with distinct non-motor profiles — target choice should incorporate cognitive and mood considerations.

        Major Points
        Phase 3 multicenter randomized blinded-assessment trial at 7 VA + 6 university hospitals (CSP-468, Follett/Weaver NEJM 2010)
N=299 adults with advanced idiopathic PD, Hoehn-Yahr ≥2 OFF, levodopa responsive, ≥3 h/d poor motor function
1:1 randomization to bilateral GPi (n=152) vs STN (n=147) DBS stratified by site and age <70 vs ≥70
Primary outcome: 24-month off-medication on-stimulation UPDRS-III change, blinded assessment
Primary outcome: GPi -11.8 vs STN -10.7; difference -1.1 (95% CI -4.3 to 2.1); p=0.50 — motor-equivalent
Levodopa-equivalent reduction greater with STN: -408 vs -243 mg/day; p=0.02
WAIS-III processing speed declined more with STN (p=0.03); BDI-II worsened with STN (p=0.02)
PDQ-39 quality of life, motor diaries, and stand-walk-sit on-medication did not differ by target
Serious AEs: 51% GPi vs 56% STN (NS); 13 deaths total (5 GPi, 8 STN)
Established GPi and STN as motor-equivalent over 24 months with distinct non-motor profiles
Practice implication: STN preferred in younger patients for medication reduction; GPi preferred with cognitive or mood vulnerability
Confirmed by subsequent PD SURG (UK) and NSTAPS (Netherlands) trials

      

Design

Study Type: Phase 3 multicenter randomized parallel-group trial with blinded outcome assessment

Randomization: 1

Blinding: Blinded motor assessment (open-label surgical arms)

Follow-up Duration: 24 months

Sample Size: 299

Analyzed: 299

Analysis: Intention-to-treat with LOCF; mixed-effects sensitivity analyses

Baseline Characteristics

Characteristic	Control	Active
N	152	147
Age mean	61.8 ± 8.7	61.9 ± 8.7
Male	87.5%	78.9%
PD duration	11.5 y	11.1 y
UPDRS-III off	41.8 ± 13.1	43.0 ± 15.0
Hoehn-Yahr off	3.3	3.4
Levodopa equivalents	1361 mg/d	1295 mg/d

Arms

Field	Control	Subthalamic DBS (STN)
N	152	147
Intervention	Bilateral globus pallidus interna DBS with programming optimization	Bilateral subthalamic nucleus DBS with programming optimization
Duration	24 months	24 months

Outcomes

Outcome	Type	Control	Intervention	P-value
Change in UPDRS-III motor score at 24 months, off-medication / on-stimulation, blinded assessment	Primary	-11.8 (GPi)	-10.7 (STN)	p=0.50
On-medication on-stimulation UPDRS-III change	Secondary	-1.2 (GPi)	+0.8 (STN)	p=0.09
Off-medication off-stimulation UPDRS-III	Secondary	-3.7 (GPi)	+2.2 (STN worsened)	p<0.001
Levodopa-equivalent reduction at 24 months	Secondary	-243 mg/day	-408 mg/day	p=0.02 (favoring STN)
Stand-walk-sit test, off medication / off stimulation	Secondary	-4.1 sec (GPi)	-0.2 sec (STN)	p=0.005 (favoring GPi)
Processing speed index (WAIS-III)	Secondary	Less decline	Greater decline	p=0.03 (favoring GPi)
Beck Depression Inventory II change	Secondary	Slight improvement	Slight worsening	p=0.02 (favoring GPi)
PDQ-39 quality of life (8 subscales)	Secondary	Improved on 6/8	Improved on 6/8	No between-target difference
Time with good motor function (motor diary)	Secondary	Improved	Improved	No between-target difference
Any serious AE	Adverse	51% (77/152)	56% (83/147)	NS
Surgical site infection	Adverse	Comparable	Comparable	Similar
Intracerebral hemorrhage	Adverse	Uncommon	Uncommon	No difference
Depression/mood worsening	Adverse	Less common	More common	p=0.02 on BDI-II
Cognitive decline (processing speed)	Adverse	Less	More	p=0.03
Falls	Adverse	Similar	Similar	No between-target difference
Speech/dysarthria	Adverse	Common	Common	Similar
Dyskinesia (stimulation-related)	Adverse	Less common	More common	STN-specific
Death	Adverse	5/152	8/147	No excess

Subgroup Analysis

No significant interaction between target and age, PD duration, or baseline motor severity. Unplanned exploratory analyses suggested older patients may derive less cognitive tolerance from STN — reinforcing the principle that cognitive/mood vulnerability should influence target selection. The processing-speed and BDI-II differences favoring GPi are modest in magnitude but cumulative and clinically meaningful over years of DBS therapy.

Criticisms

Short 24-month follow-up relative to typical 10-15 year DBS device lifetime — long-term durability and late AEs not captured
Open-label surgical arms — only motor assessment was blinded; cognitive/mood assessments not blinded
Patient population was predominantly male veterans — generalizability to broader PD populations uncertain
Programming was individualized; STN programming complexity may have favored more experienced centers
Levodopa reduction with STN is partly an assumption of equal motor benefit — in practice, patients and neurologists may prefer STN for medication-burden relief despite mood/cognitive trade-offs
Did not compare to newer directional or adaptive DBS technologies, focused ultrasound pallidotomy, or LCIG/subcutaneous apomorphine

Funding

VA Cooperative Studies Program, NINDS, Medtronic (partial)

Based on: CSP-468 STN vs GPi (New England Journal of Medicine, 2010)

Content summarized and formatted by NeuroTrials.ai.

CSP-468 STN vs GPi

(2010)

Objective

Bilateral DBS of the globus pallidus interna (GPi) vs subthalamic nucleus (STN) — to compare 24-month motor, cognitive, and mood outcomes in advanced Parkinson's disease.

Study Summary

• 24-month off-medication UPDRS-III improvement was equivalent: -11.8 GPi vs -10.7 STN; difference -1.1 (95% CI -4.3 to 2.1); p=0.50.
• STN allowed 408-mg levodopa-equivalent reduction vs 243 mg with GPi (p=0.02) — a key STN advantage.
• GPi better preserved visuomotor processing speed and mood; STN caused more decline on both (p=0.03 cognitive, p=0.02 BDI-II).
• Serious AEs occurred in 51% (GPi) and 56% (STN) — no significant between-target difference.
• Motor diaries, stand-walk-sit test (on medication), and PDQ-39 did not differ between targets.
• Established that target choice should weigh non-motor factors (cognition, mood, medication reduction), not motor efficacy alone.

Intervention

Bilateral DBS of globus pallidus interna (GPi) or subthalamic nucleus (STN) with 24-month follow-up. Blinded UPDRS-III motor assessment. Medication optimization at all time points without regard to target.

Inclusion Criteria

Adults ≥21 years with idiopathic PD, Hoehn-Yahr ≥2 off-medication, levodopa responsive, ≥3 hr/day poor motor function or disabling fluctuations/dyskinesia despite optimized medical therapy.

Study Design

Arms: Bilateral GPi DBS vs Bilateral STN DBS

Patients per Arm: GPi 152; STN 147 (N=299)

Outcome

• Primary: Off-medication on-stimulation UPDRS-III at 24 months — GPi -11.8 vs STN -10.7; diff -1.1 (95% CI -4.3 to 2.1); p=0.50
• On-medication UPDRS-III: GPi -1.2 (slight improvement) vs STN +0.8 (slight worsening); p=0.09
• Levodopa equivalents reduced more with STN: -408 vs -243 mg/day (p=0.02)
• Processing speed index (WAIS-III visuomotor subtest): worse with STN (p=0.03 between targets)
• Beck Depression Inventory II: slight improvement GPi vs worsening STN (p=0.02); Serious AEs: 51% GPi vs 56% STN (NS); 13 deaths (5 GPi, 8 STN)

Bottom Line

Major Points

Phase 3 multicenter randomized blinded-assessment trial at 7 VA + 6 university hospitals (CSP-468, Follett/Weaver NEJM 2010)
N=299 adults with advanced idiopathic PD, Hoehn-Yahr ≥2 OFF, levodopa responsive, ≥3 h/d poor motor function
1:1 randomization to bilateral GPi (n=152) vs STN (n=147) DBS stratified by site and age <70 vs ≥70
Primary outcome: 24-month off-medication on-stimulation UPDRS-III change, blinded assessment
Primary outcome: GPi -11.8 vs STN -10.7; difference -1.1 (95% CI -4.3 to 2.1); p=0.50 — motor-equivalent
Levodopa-equivalent reduction greater with STN: -408 vs -243 mg/day; p=0.02
WAIS-III processing speed declined more with STN (p=0.03); BDI-II worsened with STN (p=0.02)
PDQ-39 quality of life, motor diaries, and stand-walk-sit on-medication did not differ by target
Serious AEs: 51% GPi vs 56% STN (NS); 13 deaths total (5 GPi, 8 STN)
Established GPi and STN as motor-equivalent over 24 months with distinct non-motor profiles
Practice implication: STN preferred in younger patients for medication reduction; GPi preferred with cognitive or mood vulnerability
Confirmed by subsequent PD SURG (UK) and NSTAPS (Netherlands) trials

Study Design

Study Type: Phase 3 multicenter randomized parallel-group trial with blinded outcome assessment
Randomization: Yes
Blinding: Blinded motor assessment (open-label surgical arms)
Sample Size: 299
Follow-up: 24 months

Primary Outcome

Definition: Change in UPDRS-III motor score at 24 months, off-medication / on-stimulation, blinded assessment

Control	Intervention	HR/OR	P-value
-11.8 (GPi)	-10.7 (STN)	- (-4.3 to 2.1)	p=0.50

Limitations & Criticisms

Short 24-month follow-up relative to typical 10-15 year DBS device lifetime — long-term durability and late AEs not captured
Open-label surgical arms — only motor assessment was blinded; cognitive/mood assessments not blinded
Patient population was predominantly male veterans — generalizability to broader PD populations uncertain
Programming was individualized; STN programming complexity may have favored more experienced centers
Levodopa reduction with STN is partly an assumption of equal motor benefit — in practice, patients and neurologists may prefer STN for medication-burden relief despite mood/cognitive trade-offs
Did not compare to newer directional or adaptive DBS technologies, focused ultrasound pallidotomy, or LCIG/subcutaneous apomorphine

Citation

View Original Publication →

CSP-468 STN vs GPi

Clinical Question

Bottom Line

Major Points

Design

Baseline Characteristics

Arms

Outcomes

Subgroup Analysis

Criticisms

Funding

Ask about CSP-468 STN vs GPi