- Design
- pooled individual patient data analysis of six randomised trials, with logistic regression and nomogram development
- Population
- 15,954 patients with STEMI given weight-adjusted tenecteplase; 169 intracranial haemorrhages
- Primary outcome
- intracranial haemorrhage after fibrinolysis
- Effect
- aOR 1.02 per mmHg systolic above 110 (95% CI 1.01-1.03); age effect 1%/year at 30 mg vs ~7%/year at 50 mg; female sex aOR 1.49 (1.06-2.11)
Individual patient data from six randomised tenecteplase trials — ASSENT-2, ASSENT-3, ASSENT-3 PLUS, ASSENT-4 PCI, STREAM and STREAM-2, all using weight-adjusted dosing and the same exclusion above 180/110 mmHg — were pooled to model intracranial haemorrhage. Among 15,954 patients, 169 bled intracranially. Above a systolic threshold of 110 mmHg, each additional 1 mmHg carried a 2% relative increase in risk (adjusted OR 1.02, 95% CI 1.01-1.03, P < 0.0001), with the curve steepening sharply past 160 mmHg. Age mattered, but only through dose: risk rose about 1% per year of age at 30 mg and close to 7% per year at 50 mg. Female sex carried 49% higher odds (adjusted OR 1.49, 95% CI 1.06-2.11) and each 5 kg less body weight 16% (adjusted OR 1.16, 95% CI 1.03-1.30). The model discriminated reasonably (C index 0.718, 95% CI 0.68-0.76) and is presented as a nomogram.
What this changes is the treatment of blood pressure in the minutes before lysis. The exclusion threshold of 180/110 mmHg is widely read as a pass-fail gate, so a patient at 172 mmHg is lysed as though the pressure were irrelevant. On this data that patient carries roughly a 60% higher relative risk than one at 150 mmHg, and the pressure is something you can act on while the drug is being drawn up.
That matters most where pharmacoinvasive treatment is the default rather than the fallback, which describes most of India outside the large metros. The small, older, female patient who presents hypertensive is the one this model singles out — and the one in whom a lower tenecteplase dose band and five minutes spent on the pressure are most worth having.
- Treat systolic pressure actively before the bolus rather than checking it against 180/110 once.
- Confirm the weight band before drawing up; the dose-age interaction means an overestimated weight in an older patient is the dangerous error.
- Flag low body weight and female sex as risk amplifiers when consenting for lysis.
- Document the systolic pressure at the time of administration, not at triage.
- Where transfer for primary PCI is feasible within guideline times, this is another argument for it in the high-risk phenotype.
Why it matters
The 180/110 exclusion is being read as a gate to pass rather than a number to lower.
Don't overread it
No trial has randomised patients to a lower tenecteplase dose; dose reduction in the elderly is a model-derived suggestion, not proven practice.
The statistics, in plain English
These are odds ratios from a regression model built on trial data, so they describe association within a selected population, not a randomised comparison of dose strategies. A C index of 0.718 means the model ranks two patients correctly about 72% of the time — useful for identifying who is high risk, not precise enough to promise an individual their exact probability. The female sex estimate (95% CI 1.06-2.11) only just excludes 1.0 and rests on a small number of events, so treat it as a signal rather than a settled magnitude.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for cardiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free