Letter: Heterogeneous Effect of Tranexamic Acid in Traumatic Brain Injury.

Abstract

To the Editor: The efficacy of tranexamic acid (TXA) in traumatic brain injury (TBI) remains inconsistent. Recently, two meta-analyses1,2 reviewed the available evidence and reported that TXA had no significant effect on mortality compared to placebo. Homogeneity is an essential factor in meta-analysis. However, there is evidence indicating that the effect of TXA may be interacted by many factors, such as the timing of TXA use, neurological severity, or short/long-term outcomes. Therefore, we reanalyzed the current evidence and hope our findings will be helpful. First, in the meta-analysis by Lawati et al,1 they used the longest follow-up timepoint during data extraction and found that the pooled outcome was nonsignificant. However, the efficacy of TXA may vary over time. For instance, in an international, randomized placebo-controlled trial (TICH-2)3 involving 2325 patients with intracerebral hemorrhage, researchers found that compared to placebo, TXA use only reduced the short-term mortality (7 d) but not the 90-d mortality rate. Furthermore, the second analysis4 of this trial also reported that the TXA reduced the risk of early but not late neurological deterioration. Similarly, in a study of subarachnoid hemorrhage, Post et al5 found that compared to placebo, TXA use was significantly associated with lower in-hospital mortality, while the long-term neurological outcome (at 6 mo) was comparable between these two groups. However, in the Lawati et al meta-analysis,1 the timepoint of death varies in all 8 included trials (such as death at 28 d, at hospital discharge, at 6-mo, or others). Therefore, there is a risk that the conclusion may be biased as the efficacy of TXA changes over time. In another review,2 2 critical trials (Rowell-2020 and Mousavinejad-2020) were missing, which may also lead to a biased conclusion. Therefore, we re-extracted the data from 8 trials6-13 to investigate the different impacts of TXA on short/long-term mortality. The pooled outcome showed that TXA significantly reduced the short-term mortality (death at 28 d or hospital discharge, Figure; risk ratio (RR) 0.91, 95% CI 0.84-0.99, P = .026) in TBI patients. However, the impact on long-term mortality remains unclear due to the lack of sufficient data.FIGURE.: Forest plot comparing TXA and placebo for mortality at 28 d or at hospital discharge.Second, the timing of TXA therapy also played an important role. One multicenter trial involving 10 096 trauma patients (CRASH-2)14 reported that TXA should be given as early as possible to bleeding trauma patients, as TXA use at a late stage is less effective and could even be harmful. In most included trials, the enrollment time after trauma was 2 to 8 h except for the Chakroun 2018 study (24 h).9 This may be the reason for the opposite trend of efficacy on mortality in Chakroun-2018’s study (Figure, red arrow). Third, the neurological severity was different within all included studies. The CRASH-3 trial11 showed that TXA leads to a substantial reduction in mortality in patients with mild to moderate TBI, while there is no benefit in severe TBI. Therefore, we excluded data of patients with severe brain injury (detailed description in the Figure legend)6-8,11,12 and found that the efficacy of TXA therapy in the mild to moderate brain injury subgroup was more significant (RR 0.86, 95% CI 0.77-0.95, P = .004). In conclusion, TXA may have a positive impact on short-term outcomes in patients with mild to moderate TBI. However, the benefit in terms of long-term outcome remains unclear, and the late use of TXA in severe TBI should be done with caution. Notes: Short-term mortality subgroup: The 28-d mortality was reported in the Perel-2012, Chakroun-2018, Roberts-2019, and Rowell-2020 trials. Mortality at hospital discharge was reported in the Yutthankasemsunt-2013 and Fakharian-2017 trials. The timepoint was unclear in the Ebrahimi-2019 (most likely at hospital discharge) and Mousavinejad-2020 trials. Enrollment time ≤ 8 hours subgroup: The enrollment time after trauma was 24 h in Chakroun2018’s trial, which was excluded from this subgroup. Mild to moderate brain injury subgroup: Patients with Glasgow Coma Scale (GCS) = 3 or with bilateral unreactive pupils were excluded in Roberts-2019’s trial. No patients were excluded in the Rowell-2020 (at least 1 reactive pupil), Fakharian-2017 (the mean GCS value at hospital arrival was 12.1), Yutthankasemsunt-2013 (GCS 4-12), and Perel-2012(the proportion of patients with severe brain injury was 33.3%) trials. The proportion of patients with severe brain injury was high in the Mousavinejad-2020 (67.5%) and Ebrahimi-2019 (58.5%) trials, and these 2 trials were excluded from this subgroup. Funding Dr Shen received funding from the Medical Health Science and Technology Project of Zhejiang Provincial Health Commission (NO. 2021KY001). Disclosures The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.