Ultra-early tranexamic acid after subarachnoid haemorrhage (ULTRA): a randomised controlled trial

Post, R. et al. Lancet 2021; 397:112-18. doi:10.1016/S0140-6736(20)32518-6]

Clinical Question

  • In adults presenting with subarachnoid haemorrhage (SAH), does ultra early short term treatment with tranexamic acid improve clinical outcome at 6 months?


  • Tranexamic acid has previously been thought to be a wonder drug in terms of bleeding, but many recent negative studies looking at effectiveness of TXA haltering bleeding at various sites, suggest that its use needs to be refined (CRASH 2, WOMAN, TICH-2 etc)
  • Patients presenting with SAH are at risk of a re-bleed, which can be catastrophic; this most commonly occurs within the first 24hrs post sentinel event and it is thought that about half of these within the first 3 hours
  • Those with SAH are also prone to developing delayed cerebral ischaemia which can lead to poor clinical outcome in patients; there are concerns that TXA may potentially encourage this
  • Some studies have shown a trend towards TXA reducing rates of re-bleeds in SAH, however there are no high quality studies powered specifically to look at overall clinical outcome


  • Randomised
    • Computer generated allocation
    • Stratified by treatment centre
      • Permuted blocks within each centre
      • Random block sizes up to max of 12
  • Non blinded, other than the assessors at 6 month follow up
  • 950 patients would be needed to provide an 80% power to detect an 8.1% difference in good clinical outcome (defined here as mRS 0-3), with alpha set at 0.05
  • Intention to treat (ITT) analysis for analysing data but as treated (AT) and per protocol (PP) also to be done in original methodology given risk of protocol deviations etc being open label
  • Registered on clinicaltrials.gov


  • 24 sites in Netherlands (16 referring hospitals and 8 SAH treatment centres)
  • Data collected July 2013- July 2019


  • Inclusion:
    • ≥18yrs
    • Admitted to one of the participating centres
    • Symptoms less than 24 hrs
    • Non contrast CT confirming SAH
  • Exclusion:
    • Perimesencephalic bleeding pattern on CT AND GCS 13-15 with no loss of consciousness after onset nor focal neurological deficit on admission
    • Traumatic SAH pattern on CT
    • Ongoing treatment for DVT/PE
    • Hypercoagulability disorder
    • Pregnancy
    • Creatinine >150umol/L
    • Imminent death within 24hrs
  • Comparing baseline characteristics (Tranexamic acid vs. control)
    • Female 69% vs 66%
    • World Federation of Neurosurgical Societies (WFNS) Grade
      • I: 36% vs 40%
      • II: 20% vs 20%
      • III: 5% vs 3%
      • IV: 20% vs 20%
      • V: 19% vs 17%
    • Fisher Grade Score
      • II: 8% vs 4%
      • III: 26% vs 32%
      • IV: 66% vs 64%
    • Medication use
      • Platelet inhibitor: 13% vs 13 %
      • Anticoagulation: 3% vs 4%
      • Antihypertensives: 24% vs 23%
      • None of the above: 69% vs 69%
    • Location of aneurysm
      • Ant. circulation: 70% vs 69%
      • Post. circulation: 16% vs 16%
      • None: 14% vs 15%
    • Treatment modality
      • Endovascular: 67% vs 64%
      • Clipping: 21% vs 22%
      • None: 13% vs 14%


  • Tranexamic acid (n=480)
    • 1g bolus followed by 1g infusion every 8hrs
    • Continued until start of endovascular or surgical treatment or until a max of 24 hrs ( whichever came first)


  • Standard treatment (n=475)

Management common to both groups

  • Trial was done by deferred consent given necessity of immediate treatment if enrolled into study arm
  • Tranexamic acid was stopped if:
    • No aneurysm seen after diagnostic work up ( CT angio or digital subtraction angiography)
    • When other intracranial pathology was responsible for SAH
    • When patients or authorised representatives declined tranexamic acid
  • Standardised telephone interview at 6 months to assess modified Rankin Scale (mRS)


  • Primary outcome: (Tranexamic acid vs control)
    • Good clinical outcome at 6months, ie mRS 0-3 – no significant difference
      • 60% vs 64% OR 0.87 ( 95% CI 0.67- 1.13); aOR 0.86 ( 0.66-1.12)
  • Secondary outcomes: (Tranexamic acid vs control)
  •  Excellent outcome (mRS 0-2) at 6months – significantly lower in TXA group
      • 48% vs 56%,OR 0.74 (CI 0.57-0.96), aOR 0.73 (CI 0.57- 0.95)
  • No significant difference in:
    • Ordinal shift analysis of the mRS at 6months
      • OR 0.8 (CI 0.64-1.01), aOR 0.8 (CI 0.63-1.01)
    • All cause mortality at 6 months
      • 27% vs 24%
    • Serious adverse effects
      •  80% vs 78%
    • All re-bleedings before aneurysm treatment
      • 10% vs 14% (OR 0.71, 95% CI 0.48-1.04)
    • Delayed cerebral ischaemia
      • 23% vs 22%
    • Thomboembolic complications during endovascular treatment
      • 11% vs 13%
  • NB no differences in clinical outcome seen in primary ITT analysis vs AT or PP analysis

Authors’ Conclusions

  • The use of tranexamic acid after CT confirmation of SAH does not improve clinical outcome at 6months as measured by mRS


  • Randomised
  • Powered for clinical outcome rather than re-bleeds (seen in other studies)
  • Adjusted OR used to account for potential confounders in baseline characteristics/differences in baseline centres
  • Information can be lost with dichotomisation of mRS thus ordinal shift analysis also carried out
  • Intention to treat
  • Outcomes were also assessed “As treated” (AT) , given protocol deviations
  • Only 1% of patients were lost to follow up (5 patients in each study group)


  • Unblinded thus prone to bias
  • Convenient sample thus potential for bias
  • Multicentre, but only in one country. Reduced external validity esp with variety of access globally to speedy surgical intervention
    • This study had a median time of surgical intervention 14hrs (IQR 5-20)
  • Large number of patients (14%) recruited had no aneursyms seen on angiography which will affect power/outcomes of study
  • True efficacy of TXA potentially not caught by this study as median time from first signs and symptoms and TXA was 185mins (IQR 130-333)
  • mRS is subjective and scored by one assessor, multiple assessors in study
    • intrarater/inter rater reliability

The Bottom Line

  • This study does not show any clear clinical benefit from tranexamic acid use in SAH.
  • There may be some potential use of tranexamic acid in SAH in the future, when the patient group and timings are clearer. However, given the potential increased risk of delayed cerebral ischaemia i would not be routinely be using this in my practice

External Links


Summary author: Halah Zareian @DocHZ
Summary date: 14/4/21
Peer-review editor: @davidslessor




  • Priya Patel

    These exclusion criteria:

    – Loss of consciousness after onset
    – Focal neurological deficit on admission

    were only in reference to peri-mesenchephalic bleeds. Not entire exclusion criteria in themselves, otherwise the trial would have been extremely unhelpful.

    Although as a learning point, I can’t work out why they would exclude these bleeds. Is it because they’re mostly non-aneurysmal?

    • Halah Zareian

      Thank you for the spot and have corrected our exclusion list.

      Yes, I would assume the investigators were trying to exclude the non aneurysmal causes of perimesencephalic bleeds.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.