Journal Club · Functional Neurosurgery

VIM vs PSA DBS for Essential Tremor

The one randomized, double-blind trial that pits the thalamus against the territory just beneath it.

A trainee reading of Barbe and colleagues’ crossover trial — the only Class I comparison of ventral intermediate nucleus (VIM) and posterior subthalamic area (PSA) stimulation — and why the result points less to a winning nucleus than to a shared fiber tract.

Barbe Neurology 2018 Class I evidence Randomized, double-blind, crossover n = 13 ET

Why This Trial

The ventral intermediate nucleus (VIM) of the thalamus has been the default DBS target for essential tremor (ET) for three decades. Yet two persistent problems — stimulation-resistant proximal and intention tremor, and tolerance requiring escalating amplitudes with dysarthria and ataxia — have kept interest alive in a target a few millimeters deeper: the posterior subthalamic area (PSA), encompassing the caudal zona incerta and the prelemniscal radiation. The unifying idea is anatomical: the dentatorubrothalamic tract (DRT), the cerebellar outflow that carries the tremor signal, ascends from the PSA into the VIM — so the two targets may be the same fiber bundle engaged at two different heights. Most comparative data are retrospective; this journal club is built around the one randomized trial that isolates the cleanest version of the question.

One-Line Takeaway

In the only randomized, blinded comparison, PSA and VIM stimulation suppressed tremor equally, but PSA achieved it at significantly lower amplitude — a tie on efficacy, a win on efficiency — consistent with both targets engaging the same dentatorubrothalamic tract at different depths.

Part I

The Trial — Barbe et al. (Neurology 2018)

1.A Design That Controls For Lead Placement

This is the only Class I evidence comparing the two targets, and its design is its great strength. Rather than implanting two cohorts at two targets (where lead-placement variability confounds everything), the investigators implanted a single lead per side spanning both targets: at least one contact on the intercommissural line (the VIM contact) and at least one below it in the PSA. Each patient served as their own control, with VIM and PSA stimulation differing only by which contact on the same lead was active.

After implantation, all patients ran 3 months on the neutral (intercommissural-line) contact, then were randomized into a double-blind crossover: 2 months of PSA stimulation followed by 2 months of VIM stimulation, or the reverse, with patients and the assessing physician blinded to which target was active. After unblinding, settings could be optimized freely to a 12-month follow-up. Blinded physicians scored the Tremor Rating Scale (TRS) — the primary endpoint was TRS total score after 2 months of each condition — and an independent blinded physician re-rated all videotapes.

2.What They Found

  • Both targets worked. From a baseline TRS total of 47.4, scores fell to 17.2 with PSA and 23.8 with VIM stimulation.
  • Tremor control: statistically equivalent, with a non-significant trend favoring PSA. The PSA-vs-VIM difference of −6.9 points did not reach significance (P=0.086) — PSA was not proven superior, but was at least as good.
  • Efficiency: PSA wins. Clinical benefit was achieved at significantly lower stimulation amplitude in the PSA (VIM mean 5.88 ± 2.1 mA vs a lower PSA amplitude; P=0.006).
  • Quality of life (QUEST) improved significantly under PSA stimulation.
  • Side effects (including ataxia on the ICARS) were comparable between targets within the therapeutic range; period/carry-over effects complicate the ataxia comparison.
  • Patient preference: the majority of patients remained on PSA-DBS at 1 year.
The Conclusion, Carefully Stated The trial provides Class I evidence that PSA-DBS is not significantly different from VIM-DBS in tremor suppression, but achieves clinical benefit at lower stimulation amplitudes — “at least as effective as, but possibly more efficient than” VIM-DBS. Lower amplitude buys a wider therapeutic window and longer battery life.
Limitations Very small n (13); single center; the VIM and PSA contacts sit only millimeters apart on one lead, so current spread overlaps and the comparison arguably samples one fiber system at two depths (which is also conceptually the point); 1-year follow-up is too short to adjudicate long-term habituation; framed around equivalence/efficiency rather than powered to detect a modest efficacy difference.
Part II

Putting Barbe In Context

3.The Supporting (Retrospective) Literature

The randomized result sits on top of a larger, mostly retrospective literature that points the same direction, and which belongs in discussion rather than as the evidentiary backbone:

  • A 2025 network meta-analysis (~52 studies) found comparable limb-tremor reduction between VIM and PSA, with PSA showing an edge for head tremor and lower dysarthria risk — and, importantly, that both targets lose roughly half their benefit over ~10 years, tempering the older claim that the PSA does not habituate. (Recent; not yet replicated.)
  • Influential retrospective work (Sandvik 2012; Plaha 2008) first argued that the best contact in nominally “VIM” patients frequently lands in the PSA, and that caudal zona incerta stimulation controls the proximal, intention, cerebellar, and Holmes tremor that VIM handles poorly. Hypothesis-generating, not randomized.
Two Tremor Caveats Worth Teaching First, VIM stimulation classically controls distal limb tremor better than proximal tremor, and head and vocal tremor are often refractory to unilateral stimulation — exactly the gap PSA/cZI is invoked to fill. Second, delayed loss of efficacy (tolerance/habituation) emerging weeks to months after VIM-DBS, with tremor reappearing despite repeated reprogramming, is a real and counseling-relevant phenomenon — and the pooled data suggest it is not unique to VIM.

4.The Idea The Field Is Converging On

The defensible modern interpretation is that VIM and PSA are two levels of the same cerebello-thalamic (DRT) pathway, that the benefit of what we call “VIM-DBS” is frequently delivered by the deepest contact sitting in the PSA, and that the practical future is DRT tractography-guided targeting rather than atlas coordinates. The honest caveats: the randomized evidence is a single 13-patient trial, intraoperative PSA physiology is harder to interpret than VIM, and habituation affects both targets.

Questions For The Table

  • If a single lead can place contacts in both VIM and PSA, is “target choice” really a choice of depth along one tract?
  • Does the lower-amplitude advantage of PSA matter enough to change your default trajectory?
  • For a patient with prominent proximal, intention, or head tremor, does the retrospective signal justify favoring the PSA?
  • How would DRT tractography change your targeting, and what would you want validated before adopting it?
  • Is a 13-patient crossover enough to move practice, or only enough to justify the next trial?
  • The only randomized, blinded comparison shows equivalent tremor control with lower required amplitude for PSA — a tie on tremor, a win on efficiency.
  • Because both contacts lie on one lead, the trial supports the DRT-as-common-substrate model: the choice may be how deep along the same tract to stimulate.
  • PSA/cZI is most attractive for proximal, intention, and head tremor, and for its energy efficiency and lower dysarthria risk.
  • Habituation affects both targets; counsel patients that delayed loss of efficacy is possible.
  • The field is moving toward tractography-guided DRT targeting rather than atlas coordinates — on still-thin randomized evidence.

Selected References

  1. Barbe MT, Reker P, Hamacher S, et al. DBS of the PSA and the VIM in essential tremor: A randomized, double-blind, crossover trial. Neurology. 2018;91(6):e543–e550.
  2. Aghajanian S, Naeimi A, Mohammadifard F, et al. Comparative efficacy of PSA and VIM deep brain stimulation in essential tremor: a network and multilevel meta-analysis. Neurosurg Rev. 2025;48(1):745. Recent; verify before final publication.
  3. Sandvik U, Koskinen LO, Lundquist A, Blomstedt P. Thalamic and subthalamic deep brain stimulation for essential tremor: where is the optimal target? Neurosurgery. 2012;70(4):840–845. Discussion context.
  4. Plaha P, Khan S, Gill SS. Bilateral stimulation of the caudal zona incerta nucleus for tremor control. J Neurol Neurosurg Psychiatry. 2008;79(5):504–513. Discussion context.