Tenecteplase or Alteplase: Choosing a Fibrinolytic Inside the 4.5-Hour Window
A stroke alert inside the standard treatment window, with two guideline-acceptable fibrinolytics on the shelf and no stated preference between them — the disagreement is about which one, and about a dosing mix-up risk neither drug's own trial data addresses.
D.F., a 58-year-old man, was midway through walking a sophomore through a table-saw safety check in his woodshop classroom when he stopped mid-sentence, and the student who noticed it later told the paramedics that "Mr. F. just sounded wrong." He has taught shop for twenty-two years and by his own account has missed exactly two sick days in the last five. His only regular medication is lisinopril for hypertension diagnosed eight years ago, reasonably controlled on his last several office visits; he has never had a stroke, a seizure, or a bleeding episode, and takes no anticoagulant. He was last seen entirely normal at 9:50am, walking the same student through the same lesson. By 10:02am a colleague found him unable to move his right hand and slurring words badly enough that "safety glasses" came out as two syllables instead of four.
EMS arrived at 10:14am, called a prehospital stroke alert, and had him at the emergency department door by 10:41am — fifty-one minutes from last known well. His NIHSS on arrival is 14: dense right arm and leg weakness, a right facial droop, and expressive aphasia severe enough that he can follow one-step commands but cannot reliably name objects. Noncontrast CT shows no hemorrhage and no early large-territory hypodensity to suggest an already-completed infarct; CT angiography confirms a proximal left M1 occlusion, a vessel large enough that thrombectomy is already being arranged in parallel with the fibrinolytic decision, not instead of it. He is well inside the 4.5-hour window with more than three hours to spare, has no history of intracranial hemorrhage or recent major surgery, and his blood pressure at 148/88 needs no pretreatment lowering to clear the guideline's 185/110 threshold. The 2026 AHA/ASA acute stroke guideline lists tenecteplase 0.25mg/kg and alteplase 0.9mg/kg side by side as Class 1 options in this window, with no stated preference between them — a change from the 2018 guideline it replaced, driven by AcT and TRACE-2, the two large randomized trials that established tenecteplase's noninferiority for functional outcome in general stroke populations. His M1 occlusion adds a second consideration: EXTEND-IA TNK found better pre-thrombectomy reperfusion with tenecteplase specifically in large-vessel occlusion like his. None of which resolves the choice, which is exactly why it is being made out loud rather than by default.
In the emergency department, before the bolus is drawn up
Tenecteplase, single 0.25mg/kg bolus, capped at 25mg. He's already headed to CT for the angiogram that's going to send him to the angio suite for thrombectomy, and every extra step between now and that transfer is a step where something can go wrong — a line, a pump programmed for the wrong rate, a transport delay while an infusion finishes. One bolus and we move.
EXTEND-IA TNK showed better reperfusion before thrombectomy with tenecteplase than alteplase in exactly this kind of large-vessel-occlusion patient, and AcT and TRACE-2 have since confirmed noninferior functional outcomes in stroke populations that weren't all thrombectomy candidates. This isn't an unproven substitution anymore.
I'm not arguing the efficacy data. I'm flagging something the trials don't measure: this hospital stocks tenecteplase for two indications with two completely different weight-based dosing schemes on the same shelf. STEMI dosing runs 30 to 50mg by weight band, in 5mg increments, as a single bolus. Stroke dosing is 0.25mg per kilogram, capped at 25mg — for him, 21mg. Someone reaching for the wrong reconstituted vial or reading the wrong chart isn't a hypothetical; it's the kind of medication event root-cause reports get written about.
Alteplase has one dosing scheme in this building, for one indication. That's not nothing when the argument for tenecteplase is partly about reducing procedural risk.
You're right that the dosing-confusion risk is real, and it's worth a hard stop on the order set — a mandatory two-person verification of the vial and the calculated dose before it's drawn up, the same way we'd handle any high-alert medication with two look-alike doses in the same pharmacy. But that's a systems fix, not a reason to default back to the drug we happen to have used longer.
Two independent randomized trials, AcT in Canada and TRACE-2 in China, reached the same noninferiority conclusion in broad stroke populations, not just thrombectomy candidates — that's a stronger and more current evidence base than "alteplase has been here since 1995," which is true but isn't itself an argument about which drug helps him more today. Tenecteplase, with the verification step built into how we draw it up.
Agreed within minutes of the CT angiogram result: tenecteplase 0.25mg/kg, capped at 25mg, given as a single IV bolus, with a mandatory two-person independent verification of both the vial pulled and the calculated dose before administration — a step added specifically because of the STEMI-dosing overlap, not because anyone doubted the drug itself. He proceeded to mechanical thrombectomy immediately after.
Not agreed, and left as a standing question for the department rather than resolved for this one patient: how rigid the two-person verification step should be system-wide — a hard electronic-order-set lock that cannot be bypassed, or a soft prompt that still allows a clinician to proceed unverified in a genuine time emergency. The pharmacologist wants the hard lock; the emergency physician worries a lock with no override becomes its own delay in the rare case where every second matters more than the verification does.