The starting point is an empirical fact. Unlike bradykinesia and rigidity, which scale with the dopaminergic deficit and improve predictably with levodopa, parkinsonian tremor's magnitude is not closely tied to dopaminergic denervation and does not respond as readily or reliably to dopamine-based medication (Hallett, Parkinsonism Relat Disord, 2012). Fishman catalogued this among the "paradoxical" features of the tremor: a complex pharmacologic basis in which levodopa is clearly effective for many patients, yet the response is neither uniform nor dose-linear (Fishman, Mov Disord, 2008). The variability is patterned rather than random — dopaminergic response tends to be relatively poorer in tremor-dominant Parkinson's than in the akinetic-rigid subtype (Abusrair et al., Tremor Other Hyperkinet Mov, 2022) — but at the level of the individual patient it remains difficult to predict.

This has a counseling consequence worth setting early. A patient whose slowness and stiffness melt away on levodopa reasonably expects the tremor to follow, and its persistence can read as treatment failure or non-adherence when it is neither. Framing the tremor at the outset as a partly separate problem — one that may respond well, may respond incompletely, and occasionally barely responds — prevents both the patient's disappointment and the clinician's temptation to escalate dopaminergic therapy past the point of benefit chasing a symptom that was never going to yield to it. It is also worth noting the countervailing reassurance: a tremor-dominant presentation, precisely the phenotype most likely to disappoint on levodopa, is broadly associated with slower overall progression, so a stubbornly dopamine-resistant tremor is not a marker of aggressive disease.

Before a tremor is called levodopa-resistant, a prior question must be settled: is it genuinely resistant, or only apparently so? Nonnekes and colleagues formalized this distinction with the concept of pseudoresistance — features that are actually dopamine-sensitive but that falsely appear resistant, or appear to lose responsiveness, for reasons unrelated to true pharmacological failure (Nonnekes et al., Mov Disord, 2016). The common contributors are practical: insufficient dosing, impaired gastrointestinal absorption or delayed gastric emptying, poor adherence, protein-timing interactions, and simple under-recognition of a partial response.

The clinical consequence is direct: an apparently refractory tremor warrants a deliberate check for pseudoresistance before the label sticks. Confirming that the patient has actually received an adequate, well-absorbed, well-timed dopaminergic trial — and excluding reversible contributors — is the step that separates a tremor that needs regimen optimization from one that genuinely needs escalation beyond dopamine. Skipping this step is a common and consequential error: it risks both prematurely labeling a treatable tremor as refractory and exposing the patient to the added burden of adjuncts that a simpler dose adjustment would have made unnecessary.

When resistance is genuine, the most illuminating evidence for why comes from functional imaging. Comparing patients whose tremor responded to dopaminergic medication with those whose did not, Dirkx and colleagues showed that dopaminergic medication reduces tremor by inhibiting tremulous activity in the pallidum and thalamus, and that interindividual differences in this circuitry underlie the clinical heterogeneity of the response (Dirkx et al., Brain, 2019). The finding reframes the paradox: parkinsonian tremor is not uniformly dopamine-independent, but its dopaminergic control is exerted through a specific pallido-thalamic node whose engagement varies between patients.

This dovetails with the broader circuit model, in which a largely non-dopaminergic cerebello-thalamo-cortical loop generates tremor amplitude while the basal ganglia act as a dopaminergic trigger (Helmich, Curr Opin Neurol, 2020). A tremor weighted toward the cerebellar amplifier would be expected to respond less to a dopaminergic intervention aimed upstream — a coherent account of the dopamine-resistant phenotype, and of why the same drug transforms one patient's tremor and barely touches another's.

The honest summary is that the tremor-specific evidence base is limited. As a recent review states plainly, the pathophysiology of Parkinson's tremor remains incompletely understood, and there is a lack of clinical trials specifically addressing its pharmacological treatment (Pirker et al., J Parkinsons Dis, 2023). Within that constraint, the evidence supports a clear hierarchy.

Two points deserve emphasis. First, dopamine agonists offer no consistent antitremor advantage over levodopa, so the older intuition that agonists are preferentially "anti-tremor" is not well supported (Pirker et al., J Parkinsons Dis, 2023). Second, anticholinergics, historically the go-to for tremor, deliver a smaller effect than levodopa and carry a cognitive and autonomic burden that restricts them to a narrow patient group.

That second point carries a historical irony worth appreciating. Anticholinergics were the mainstay of Parkinson's treatment in the pre-levodopa era, valued specifically for tremor; the arrival of levodopa displaced them, and the subsequent recognition of their cognitive risks — particularly relevant in an often older, sometimes cognitively vulnerable population — has pushed them to the margins. The net effect is that the drug class once considered the antitremor treatment is now a cautious, selective option, while levodopa, never designed with tremor foremost in mind, is the most effective pharmacological agent available for it. This inversion is a useful reminder that the tremor's pharmacology has been mapped largely by observation and repurposing rather than by tremor-powered trial design, which is exactly why the evidence base carries the caveats it does (Pirker et al., J Parkinsons Dis, 2023).

These strands converge on a practical sequence. Because levodopa remains the most effective agent and because pseudoresistance is common, the rational first move for a troublesome tremor is to optimize the dopaminergic trial — adequate dose, confirmed adherence, attention to absorption and timing — rather than to reach immediately for an adjunct or to declare the tremor refractory (Nonnekes et al., Mov Disord, 2016; Armstrong & Okun, JAMA, 2020). Only when a genuine optimization has failed does the addition of an anticholinergic or propranolol, or referral for procedural evaluation, become the reasoned next step.

Framed this way, "unpredictable levodopa response" is less a dead end than a decision node: distinguish pseudoresistance from true resistance, optimize what dopamine can do, and escalate deliberately toward the non-dopaminergic and procedural options that target the tremor's cerebellar amplifier directly.

Because the history alone is an unreliable guide to whether a tremor is truly dopamine-resistant, formal assessment of responsiveness is useful before committing to escalation. The acute levodopa (or apomorphine) challenge — scoring tremor in the practically-defined "off" state and again at peak "on" after a supra-threshold dose — provides a more objective read of the dopaminergic ceiling than a patient's impression across a variable day. It is worth remembering that tremor is state-dependent in a way that can mislead: it is amplified by anxiety, cognitive load, and the clinic setting itself, so an isolated observation may over- or under-represent the usual burden.

A further nuance complicates the on/off picture. Rest tremor may re-emerge or fluctuate with the medication cycle, and in some patients tremor is prominent precisely in the "off" state and responds to dopaminergic top-up, while in others it persists across the cycle irrespective of the dopaminergic level. Characterizing which pattern a given patient shows — off-period tremor that tracks the dopaminergic trough, versus a fixed tremor indifferent to it — is often more informative than a single global judgment of "responsive" or "not," and it directly shapes whether the lever to pull is dopaminergic optimization or something beyond it (Nonnekes et al., Mov Disord, 2016; Armstrong & Okun, JAMA, 2020).

Once pseudoresistance has been excluded and dopaminergic therapy genuinely optimized, a tremor that remains disabling enters a defined escalation pathway. The oral options are the adjuncts already noted — an anticholinergic in a suitable, younger, cognitively intact patient, or propranolol — each offering incremental benefit rather than a reliable solution (Pirker et al., J Parkinsons Dis, 2023). Clozapine has evidence as an effective agent for otherwise refractory parkinsonian tremor, but its mandatory hematologic monitoring and sedative burden confine it to selected, closely supervised cases — it is a second- or third-line consideration, not a routine one.

The most effective interventions for a genuinely dopamine-resistant, disabling tremor are procedural. VIM-targeted therapies — thalamic deep brain stimulation and MRI-guided focused ultrasound — act on the cerebello-thalamo-cortical node where tremor amplitude is generated, which is precisely why they succeed where more dopamine does not (Helmich, Curr Opin Neurol, 2020). The circuit logic makes the escalation coherent rather than arbitrary: a tremor that resists dopamine is, in effect, telling you that its dominant driver lies in the non-dopaminergic amplifier, and the therapeutic target should follow the physiology. Recognizing that early — rather than cycling indefinitely through dopaminergic adjustments — is what spares a patient months of disabling tremor before an effective option is offered (Abusrair et al., Tremor Other Hyperkinet Mov, 2022).

Several questions remain genuinely open and should be represented as such. There is no validated way to predict, at the individual level, which patient's tremor will be dopamine-responsive before a trial, though the imaging correlates of responsiveness are becoming clearer (Dirkx et al., Brain, 2019). The tremor-specific pharmacological evidence base is thin, with few dedicated randomized trials, so much of the guidance rests on subgroup analyses and expert synthesis rather than tremor-powered studies (Pirker et al., J Parkinsons Dis, 2023). And the contribution of non-dopaminergic systems to tremor — and whether they represent tractable, tremor-specific drug targets — is recognized but unresolved. The defensible clinical stance is that levodopa is the right first test and the right thing to optimize, that a non-response should trigger a pseudoresistance check before escalation, and that a genuinely dopamine-resistant tremor is a circuit problem for which VIM-targeted therapies, not more dopamine, are the logical endpoint.