TICH-2: TXA for Spontaneous ICH?

Background: Spontaneous, non-traumatic intracerebral hemorrhage, is one of the only stroke subtypes without a proven treatment. It is not as common as ischemic stroke, representing up to 20% of all strokes, but it accounts for almost half of all stroke deaths worldwide.  Furthermore, about a quarter of intracerebral hemorrhage can be complicated by hematoma expansion which can occur up to 24 hours later and is itself associated with poor outcomes.  There have been only small trials looking at the use of tranexamic acid in this group of patients, until now.  The Tranexamic acid for hyperacute primary IntraCerebral Haemorrhage (TICH-2) trial looked to see if intravenous TXA reduces death and dependence when given within 8 hours of spontaneous ICH.

What They Did: This was an international, randomized, double-blind, placebo-controlled, parallel group, phase 3 trial of adults with intracranial hemorrhage from acute stroke across 124 hospitals (12 countries).  Patient were randomized in a 1:1 fashion to receive either 1g IV TXA bolus followed by an 8hr infusion of 1g of TXA or a matching placebo within 8 hours of symptom onset.

  • Intervention: 1g TXA in 100mL of NS infused over 10min followed by another 1g in 250mL of NS infused over 8hrs
  • Placebo: 100mL of NS infused over 10min followed by another 250mL of NS infused over 8hrs


  • Primary: Functional status at day 90, defined as an ordinal shift (see discussion section) in the modified Rankin Scale (mRS)
  • Secondary:
    • Neurological impairment at day 7 or discharge (whichever came first) assessed by NIHSS
    • Activities of daily living according to the Barthel index
    • Costs including length of hospital stay
    • Change in hematoma volume from baseline to 24hrs
    • Safety outcomes up to day 90
      • Death
      • VTE
      • Ischemic events (Stroke, TIA, MI, ACS, PAD)
      • Seizures


  • Adults with acute ICH
  • Admitted to a participating hospital within 8 hours of stroke symptom onset (or time last seen well) 


  • ICH secondary to anticoagulation, thrombolysis, trauma, or known underlying structural abnormality
  • Contraindication to TXA
  • Prestroke dependence (mRS >4)
  • Life expectancy <3 months
  • GCS <5


  • 2325 patients enrolled
    • 1161 randomized to TXA
      • 1150 received intervention
      • 1152 analyzed at day 90 ITT analysis
    • 1164 randomized to placebo
      • 1157 received
      • 1155 analyzed at day 90 ITT analysis
    • Median time from stroke onset to randomization = 3.6hrs
    • Median time from randomization to treatment = 21min
    • 36% of patients recruited within 3hrs
  • Primary Outcome – Functional Status at Day 90:
    • No difference between groups
    • aOR 0.88; 95% CI 0.76 – 1.03; p = 0.11
    • No difference when mRS dichotomized at mRS 0-3 vs 4-6 (aOR 0.82; 95% CI 0.65 – 1.03), p=0.08).
  • Hematoma Expansion at Day 2
    • TXA: 25%
    • Placebo: 29%
    • aOR 0.80; 95% CI 0.66 – 0.98; p = 0.0300
  • Mean Hematoma Volume Expansion from Baseline to 24hr (mL):
    • TXA: 3.72
    • Placebo 4.90
    • Adjusted mean difference -137mL; 95% CI -2.71 – -0.04; p = 0.0432
  • Death by Day 7:
    • TXA: 9%
    • Placebo 11%
    • OR 0.73; 95% CI 0.53 – 0.99; p = 0.0406
  • Death by Day 90:
    • TXA: 22%
    • Placebo 21%
    • HR 0.92; 95% CI 0.77 – 1.10; p = 0.37
  • No difference in neurological improvement (mean NIHSS score at day 7), 90 day functional outcomes, length of hospital stay, discharge disposition, venous thromboembolic events, or arterial occlusions


  • Asks a clinically important question
  • Multicenter, international, randomized, double-blind, placebo-controlled, parallel group trial. All the things we like to see in a trial.
  • Treatment allocation was concealed from patients, outcome assessors, and all other health-care workers involved in the trial
  • Pre-specified secondary and safety outcomes prior to start of the trial
  • Groups were balanced in baseline characteristics
  • High protocol adherence of 95% receiving all of the randomized treatment
  • Broad inclusion criteria which increases generalizability into practice


  • Most patients recruited in the UK (82%), which is a different health system than what may occur in other countries
  • Screening logs were not collected therefore no data on eligibility
  • Wide inclusion criteria also led to a heterogenous population with more severe strokes, larger hematoma volumes and a greater proportion of lobar hematomas and intraventricular hemorrhage which may have diluted any potential treatment effect


  • Some major reasons why this trial didn’t show any benefit:
    • Although it has been shown in the trauma literature that ≤3 hours is an important window period for the effectiveness of TXA, this trial chose to use 8hrs as their treatment window to maintain consistency with prior trials of TXA in traumatic ICH. Future trials should target a window of <3hrs, which is the window in which most hematoma growth occurs
    • Broad inclusion criteria is great from a generalizability standpoint, but ultimately ended up being a major limitation in this study. Future trials should exclude patients with non-survivable ICH or who won’t clinically benefit from further hematoma growth
  • Although there was no difference in the primary outcome, there may be some signal in all the noise. There were reductions in early death by day 7 and hematoma expansion, but both of these should be taken as hypothesis generating at this time, as they may be random occurrences alone
  • I have never understood the fear of venous thromboembolism with TXA as this is an anti-fibrinolytic agent not a prothrombotic agent. This study once again shows no increase in VTEs in the TXA group.  (CAVEAT: The MATTERS trial did have a higher VTE rate in the TXA group, but the higher injury burden in this group is most likely what caused the increase in thrombotic events, NOT the TXA itself.)
  • An important point is patients with SBP ≤170mmHg did better than patients with SBP >170mmHg. Blood pressure lowering has been the only intervention to date to improve functional outcomes in patients with ICH, therefore early blood pressure control is a key factor to improve patient outcomes (See the REBEL EM ATACH-2 post HERE).
  • An ordinal shift analysis assesses differences in the distribution of treatment groups across the full range of an outcome scale (i.e. mRS of 0 – 6) as opposed to a dichotomous outcome of mRS≤2 vs mRS>2. This accounts for the most severe patients, without the expectation of unrealistic and dramatic improvements to be called a successful study. Another reason ordinal shift analysis can be useful is it allows physicians to detect clinically meaningful shifts in the middle of the mRS spectrum.  The major drawback of this is the difficulty in expressing the treatment effect measured in terms that are meaningful to patients and clinicians. Calculating the number needed to treat is possible, but requires much more statistical hocus pocus to achieve.  Another drawback of this technique is do reductions in mRS at the extremes of scale carry equal weight?  For example, is a reduction in mRS from 6 to 5 as good as a mRS reduction from 3 to 2?

Author Conclusion: “Functional status 90 days after intracerebral haemorrhage did not differ significantly between patients who received tranexamic acid and those who received placebo, despite a reduction in early deaths and serious adverse events. Larger randomized trials are needed to confirm or refute a clinically significant treatment effect.”

Clinical Take Home Point: In this trial TXA was given >3hrs after stroke onset, patients had more severe strokes, and larger hematoma volumes (>60mLs) than prior studies and therefore there is no surprise in the fact that there was no difference in 90 day functional outcomes in patients with spontaneous ICH receiving TXA compared to placebo. TXA cannot be recommended at this time in clinical practice for spontaneous ICH based on the results of this trial, however some hypothesis generating outcomes included reductions in early death by day 7 and hematoma expansion. Hopefully future research will try and answer these questions.


  1. Sprigg N et al. Tranexamic Acid for Hyperacute Primary IntraCerebral Haemorrhage (TICH-2): An International Randomised, Placebo-Controlled, Phase 3 Superiority Trial. Lancet 2018. PMID: 29778325

For More Thoughts on This Topic Checkout:

Post Peer Reviewed By: Rick Pescatore, DO (Twitter: @Rick_Pescatore) and Anand Swaminathan, MD (Twitter: @EMSwami)

Cite this article as: Salim Rezaie, "TICH-2: TXA for Spontaneous ICH?", REBEL EM blog, July 30, 2018. Available at: https://rebelem.com/tich-2-txa-for-spontaneous-ich/.

Like this article?

Share on Facebook
Share on Twitter
Share on Linkdin
Share via Email

Want to support rebelem?