Off Episodes in Parkinson’s Disease: Continuous Apomorphine Infusion or Staying with Rescue Injections
A patient whose rescue injections work when he can administer them — the real problem is administering them during the exact episodes they are meant to treat.
The pen was on the kitchen table and Walter “Walt” B., 68, could not get it assembled. His wife described the scene at the visit, and it is the actual reason he is here — not that the drug has stopped working, but that the twenty to forty minutes each morning when he most needs it are the same twenty to forty minutes his hands cannot manage the device that delivers it. He spent thirty years as an auto mechanic and still has a 1967 Mustang in pieces in the garage, a project he gets to most afternoons, once his medication has caught up with him. He has had Parkinson’s disease for nine years, managed on oral carbidopa/levodopa and pramipexole, and started on-demand apomorphine injections eighteen months ago specifically for the ‘off’ episodes his oral regimen couldn’t fully cover. The problem isn’t that the injections don’t work — when he can get one in reliably, they do, often within fifteen to twenty minutes — it’s that using the pen has become its own obstacle during the exact episodes it’s meant to treat. His worst ‘off’ time is the twenty to forty minutes before his first oral dose takes effect each morning, when his hands shake too much to reliably assemble the pen and dial the correct dose, and by his own account he has under-dosed or missed the injection site more than once trying to self-administer while already bradykinetic. He has also never fully tolerated the nausea that comes with each dose despite pretreatment with trimethobenzamide, and reports skipping doses some mornings specifically to avoid it — which defeats the medication’s purpose on exactly the days he needs it most.
That combination — real efficacy undermined by a delivery method that depends on the same fine motor control the disease has already taken from him during the window he needs it most — is a specific, mechanical mismatch, not a question of whether apomorphine itself works for him. The continuous infusion pump studied in the 12-week TOLEDO trial delivers the same drug, the same D1/D2 receptor agonism, through a device his wife can set up once each morning rather than something he has to execute correctly mid-episode. TOLEDO’s own patients, like Walt, had ‘off’ time inadequately controlled by oral therapy alone; the trial measured a real, significant reduction in daily ‘off’ time and a corresponding increase in good ‘on’ time compared with placebo infusion. Whether that translates specifically to solving his morning-assembly problem, versus simply shifting his nausea and site-reaction burden onto a different schedule, is what nobody has established yet. The drug class itself is not in dispute here; the route is.
Clinic, reviewing eighteen months of rescue-pen use
Walt’s problem isn’t the drug — it’s that the pen requires the exact fine motor control his worst ‘off’ window has already taken from him. TOLEDO enrolled 107 patients with ‘off’ time inadequately controlled by oral therapy, essentially his profile, and found a real reduction in daily ‘off’ time and a real increase in good ‘on’ time — ‘on’ time without troublesome dyskinesia — with continuous infusion versus placebo. A pump his wife sets up once in the morning removes the assembly step entirely.
I agree the pump removes the mechanical failure point — that part isn’t in dispute. But before anyone moves toward it, I want his current antiemetic explicitly reconciled. Apomorphine, in any formulation, is contraindicated with 5-HT3 antagonists — ondansetron and related agents — because of reported profound hypotension and loss of consciousness. If anyone along the way switched him to one of those for the nausea without flagging it against his apomorphine, that’s the thing that has to surface before we touch the delivery method at all.
And I wouldn’t assume his current trimethobenzamide pretreatment should simply carry over unchanged into a continuous-delivery plan. The apomorphine labeling has it as a pre-treatment started three days ahead and generally continued no longer than about two months after initiation — Walt is eighteen months in. That isn’t a small drift: trimethobenzamide carries its own extrapyramidal-symptom risk and is documented to increase somnolence, dizziness, and falls, which is a real thing to be adding to a patient whose problem is already unsteadiness during ‘off’ periods.
His nausea-driven skipping and his morning-assembly failure overlap but aren’t identical problems, and I’d rather not solve one by assuming it fixes the other. I’d propose a two-week trial adjusting his first oral dose earlier, run in parallel with — not instead of — scheduling his pump education visits, since the antiemetic reconciliation raised a moment ago needs to happen before either delivery method is escalated safely.
Agreed: a two-week trial of an earlier first oral dose, run in parallel with scheduling pump education visits so no time is lost; explicit chart reconciliation confirming trimethobenzamide, not a 5-HT3 antagonist, as his antiemetic of record.
Not agreed: whether his nausea alone — independent of the timing question — is reason enough to move to the pump now, since continuous subcutaneous delivery may itself reduce peak-concentration-driven nausea compared with a bolus injection. The Neurologist thinks so; the Pharmacologist wants to see whether that's genuinely the delivery profile at work or simply nausea that has been undertreated all along.