Antiplatelet Strategy After Cryptogenic Stroke: A Regimen Choice Just Outside the Trial Window
A single patient, thirty-six hours into a mild cryptogenic stroke with an incidental PFO already judged more likely bystander than cause. The disagreement isn't about whether to close anything — it's about which antiplatelet regimen, and for how long, to build his secondary prevention around now that he's arrived just past the window the trials were actually built on.
P.N., a 57-year-old man, teaches shop safety and woodworking at the local high school and still keeps a small workbench going in his garage most weekends, though his knees have made him give up standing at it for long stretches. He has hypertension and type 2 diabetes, both treated for years but never quite at target — his home blood pressure log runs closer to 138/86 than his doctor would like, and his last A1c was 7.4%. He has never smoked and has no history of stroke or TIA.
On a Monday evening, while grading shop-safety quizzes at the kitchen table, he noticed his right hand felt clumsy holding the pen and a couple of words came out garbled when he read a sentence aloud to his wife. He chalked it up to a long day and went to bed. The clumsiness was still faintly there the next morning, but he taught his classes as usual, attributing it to having slept on his arm wrong. A colleague mentioned around midday that his speech had sounded a little off in the teachers' lounge. By Wednesday morning — roughly thirty-six hours after the hand symptoms began — his wife noticed him fumbling for words describing his weekend plans and drove him straight to the emergency department.
His exam showed mild expressive word-finding difficulty and a subtle right pronator drift (NIHSS 2); MRI showed a small acute cortical infarct in the left frontal operculum. Carotid imaging and a 30-day cardiac monitor were unrevealing, and an agitated-saline echocardiogram found a PFO with a moderate right-to-left shunt at rest and no atrial septal aneurysm. His RoPE score — five points, with his hypertension and diabetes each costing him a point and his age at 57 earning only two — placed the stroke in the range where a PFO is estimated to be causally responsible only about a third of the time, well short of the anatomic and score thresholds most closure trials use to identify patients likely to benefit. The stroke and structural-heart team judged his hypertension and diabetes the more probable driver and did not recommend closure; adding an anticoagulant was never really on the table either, since two dedicated trials in ESUS — NAVIGATE ESUS and RE-SPECT ESUS — found no advantage of oral anticoagulation over aspirin in exactly this population — and NAVIGATE ESUS was stopped early with significantly more major bleeding on the anticoagulant. What was actually left to decide, once he arrived thirty-six hours after a mild, non-disabling deficit rather than the roughly twelve-to-twenty-four hours the major short-course DAPT trials were built around, was which regimen — a short dual-antiplatelet course or a single agent from the start — and which drug to build the years of secondary prevention ahead around.
Deciding a regimen thirty-six hours out
Start dual therapy now — aspirin plus clopidogrel, twenty-one days. POINT and CHANCE, the trials that established this, enrolled within twelve to twenty-four hours of onset, and he's thirty-six hours out, but the biology being treated is the front-loaded recurrence risk after a minor ischemic event, and that risk doesn't have a hard switch at hour twenty-four. It decays; it doesn't vanish. He meets every other criterion those trials used — mild deficit, no indication for anticoagulation, no bleeding history. I'd rather extend a well-supported window by half a day than withhold a treatment whose mechanism is still fully in play.
I'll say plainly that this is extrapolation, not a labeled indication for a patient presenting at thirty-six hours. I'm comfortable with it because the biology argues for continuity, not because a guideline says so.
I'd have started aspirin alone. POINT and CHANCE didn't just test a shorter interval — they only ever enrolled patients inside it, and in POINT the two-drug arm produced significantly more major bleeding than aspirin alone. We're not just borrowing the efficacy signal by extending the window; we're extending an intervention whose safety was only ever measured in patients who started it within a day of onset, into a starting point nobody has direct safety data for.
I take the point that recurrence risk doesn't switch off at hour twenty-four. But nobody has measured bleeding risk in patients started this late either, and that cuts both ways — we're inferring the safety holds up just as readily as we're inferring the efficacy does.
On the twenty-one-day question, I think you can defend either position — his bleeding risk is low, and the absolute cost of being wrong about a few extra hours of dual therapy is small. I'd go ahead and start it. The question I'd actually push back on is what happens at day twenty-one. Clopidogrel's overall edge over aspirin comes out of CAPRIE, where the benefit was carried by the peripheral arterial disease patients specifically — the stroke-only patients in that same trial didn't reach significance on their own. Reaching for clopidogrel here because of an aggregate result driven by a different population isn't the same kind of extrapolation as the timing question. I'd keep him on aspirin once the dual course ends.
Agreed within the visit: start dual antiplatelet therapy — aspirin 81mg plus clopidogrel 75mg daily — today, for a defined 21-day course, with a follow-up visit at day 21 to decide the continuation agent.
Not agreed, and left as an explicit branch rather than a single expectation, is which single agent takes over once the 21 days end:
The Neurologist's default — trusting CAPRIE's overall result even though the stroke-specific subgroup within that same trial wasn't independently significant.
The Primary Care Physician's and Clinical Pharmacologist's default — favoring the agent whose evidence doesn't rely on extrapolating from a subgroup he doesn't belong to.
All three treated the day-21 question as genuinely open rather than urgent — none of them made it a condition of starting therapy today. What they did agree on without any argument at all: his hypertension and diabetes control matter more to his five-year stroke risk than the choice between the two drugs, and neither one could be safely deferred until day 21 either.