Tranexamic Acid Dosing for Cardiac Surgical Patients With Chronic Renal Dysfunction: A New Dosing Regimen

Angela Jerath, FRCPC, FANZCA, MBBS, BSc; Qi Joy Yang, MSc; K. Sandy Pang, PhD; Nikita Looby, MSc; Nathaly Reyes-Garces, MSc; Tijana Vasiljevic, BSc; Barbara Bojko, PhD; Janusz Pawliszyn, PhD; Duminda Wijeysundera, PhD, FRCPC; W. Scott Beattie, PhD, FRCPC; Terrence M. Yau, PhD, FRCSC, MD, MSc, BA; Marcin Wąsowicz, FRCPC, PhD, MD


Anesth Analg. 2018;127(6):1323-1332. 

In This Article


This study confirmed that plasma TXA levels were elevated in proportion to the severity of CRD and the reduction in renal CL. This is most pronounced among advanced CRD stages 3–5 when TXA maintenance infusions were used. We have recommended a simple adjustment strategy to the BART dosing regimen to minimize costly drug overdosing, accumulation, and potential toxic effects of TXA. We also identified that single TXA bolus dosing in stages 1 and 2 CRD was associated with a rapid decline in plasma levels to subtherapeutic concentrations, without much risk of toxicity. This is important during prolonged surgeries where repeat dosing may be warranted or an infusion is preferable. Bolus dosing among stages 4 and 5 CRD provided near therapeutic plasma TXA levels with effective antifibrinolysis for several hours.

Although the primary aim was not to assess clinical outcomes, we revealed an alarmingly high seizure rate (8%) in advanced CRD. Postoperative seizures typically affect <1% of cardiac surgical patients and are associated with 10-fold higher incidence of delirium, stroke, mechanical ventilation, and mortality.[13] Risk-adjusted analyses demonstrate that independent predictors of early seizures include elderly patients (>75 years of age), preoperative cardiac arrest, preoperative neurological disease (cerebrovascular disease, seizure disorder, alcohol abuse, brain tumor, multiple sclerosis), Acute Physiology And Chronic Health Evaluation II scores >20, redo surgery, peripheral vascular disease, open left heart and aortic operations, long CPB times, preoperative CRD, and high TXA doses.[13,14] The recent large multicenter trial reported by Myles et al[9] provides further evidence demonstrating a link between TXA and seizures. This trial showed that seizures were 7 times more likely in patients receiving similar TXA bolus doses (50–100 mg/kg) in comparison to placebo. Seizures secondary to dialysis disequilibrium syndrome are unlikely given this is exceedingly rare and commonly occurs after commencing new-onset hemodialysis.[23,24] Dialysis disequilibrium syndrome is characterized by neurological changes such as restlessness, acute confusion, headache, and seizures whose etiology remains unknown.[25] Animal data have demonstrated that the mechanism of TXA-induced seizures is through central inhibition of glycine and glycine and gamma-aminobutyric acid subtype a receptors.[26,27] Although this mechanism is plausible, other factors such as cerebral embolic load from open cardiac procedures, blood–brain barrier leak, cerebral inflammation, and preoperative CRD would likely interplay in raising central TXA levels and inducing seizures.[28,29] Currently, we lack human data assessing the correlation between plasma and cerebrospinal fluid TXA levels and seizure induction. However, TXA dosing appears to be an important variable with several studies demonstrating a dose–response relationship.[13,27,30]

Our data support the higher mortality rate seen among cardiac patients with CRD.[4,5] Previous work showed that even mild CRD (creatinine, 130–140 mmol/L) was significantly associated with higher mortality.[31,32] Renal function further deteriorates postsurgery secondary to altered hemodynamics, surgical trauma, systemic inflammatory response leading to altered renal vasoreactivity, local hypoxia, and oxidative injury causing medullary congestion.[33] This would further reduce the CL of renally excreted drugs such as TXA.

Our findings are supported by Andersson et al[17] who administered intravenous TXA to 28 nonsurgical patients with varying CRD levels. They identified that plasma TXA levels were highest in patients with the poorest renal function. This study formed the basis of current licensed Food and Drug Administration dosing recommendations of short-term TXA use among hemophiliacs undergoing dental extraction. The dosing of cardiac surgery patients is very different, given the presence of CPB, altered distribution volumes, transfusion, and intense inflammatory response. Data in cardiac surgery are limited to a single study performed by Fiechtner et al[34] who assessed plasma TXA levels in 21 patients where only 4 had preoperative CRD (serum creatinine, 2.3–6.4 g/dL). The TXA regimen consisted of 10 mg/kg bolus dose and 1 mg/kg/h maintenance infusion for up to 2 hours in the ICU. The plasma TXA levels were significantly raised in CRD in comparison to normal renal function patients. Analyses of these 4 CRD patients led to recommending a TXA regimen of 5.4 mg/kg bolus dose, 50 mg in the CPB prime and reduced maintenance infusion by 25% (3.75 mg/kg/h) in patients with serum creatinine 141–291 mmol/L, 50% (2.5 mg/kg/h) serum creatinine 291–583 mmol/L and 75% (1.25 mg/kg/h) when creatinine exceeded 583 mmol/L. Although our findings corroborated elevated plasma TXA levels in CRD patients, we recommend an alternative dosing reduction regime. Our recommendations are based on the pharmacokinetic interpretation of a larger CRD population using the KDOQI classification, studying patients for longer CPB times and assessing 2 different dosing regimens. Our results demonstrated a reduction in the volume of distribution and CL of TXA among CRD patients (Table 1). This resulted in high plasma TXA concentrations in accordance with the severity of CRD. In advanced CRD, plasma levels were 3–4 higher than the threshold required to achieve maximal antifibrinolysis (100 mg/L). There has been no previous work that has clearly shown that plasma TXA levels rise with increasing severity of CRD as defined by the KDOQI guidelines among cardiac surgical patients.

We acknowledge that our study is not without limitations. First, the clinical outcomes are not powered in the sample size. Given that this is primarily a pharmacokinetic study, we did not aim to demonstrate a clinical difference but the data provide a useful trend and foundation for conducting a larger clinical trial. Our sample size is in keeping with other pharmacokinetic studies.[15,20–22] Although fewer patients were recruited in certain CRD stages, we believe that valuable information had been ascertained to meet our study objectives. Second, only clinically evident postoperative seizures were documented. Routine EEG monitoring for subclinical seizure activity is not standard practice in ours and many other cardiac centers. Thus, the true incidence of seizures may be even higher. Given our findings, cardiac centers may wish to consider the role of routine EEG monitoring in cardiac surgical patients who are at high risk of postoperative seizures. Third, many centers may use ε-aminocaproic acid for antifibrinolysis, which has also been associated with seizure activity. Although this agent is structurally similar to TXA and undergoes predominant renal elimination, further studies are needed to recommend the optimal dosing regimen. Fourth, we acknowledge that even though the BART dose is a commonly used regimen, our low-risk bolus dose may not be unanimously accepted. However, this bolus dose was used in the recent TXA trial conducted by Myles et al.[9] Fifth, we accept that in vivo validation of our dosing scheme has not been performed. However, given that the study findings corroborate our previous work and protein binding of TXA is low, this TXA pharmacokinetic model for CRD and dosing recommendations are likely to be robust.[18] Sixth, we acknowledge that the plasma TXA threshold of 100 mg/L used in this study is based on in vitro and theoretical considerations to provide near-maximal antifibrinolysis. There is no sensitive point-of-care test to assess individual patient response to TXA or degree of fibrinolysis. However, our bolus dose was used in the ATACAS study, which remains the largest trial assessing the safety and efficacy of TXA. Last, our model did incorporate CPB priming volume and returned blood volume at the end of CPB. However, we acknowledge that TXA levels may still be affected by sudden blood loss, blood transfusion and use of a cell saver. These factors may impact whether the proposed dosing scheme will achieve therapeutic plasma TXA levels.

This study is the first to formally conduct pharmacokinetic analysis and make simple dosing recommendations specifically for CRD patients undergoing cardiac surgery that can be widely adopted. Further work includes examining TXA dosing for other clinical areas. Off-label TXA is commonly used to manage bleeding in trauma patients (Clinical Randomization of an Antifibrinolytic in Significant Haemorrhage [CRASH] 2 trial), severe menorrhagia, hereditary angioedema, and complex noncardiac surgeries (major orthopedics, maxillofacial, prostate, hepatobiliary).[35,36] Despite widespread use, informed kinetic analysis and dosing schemes are lacking in these areas. This study will form a solid basis for further work optimizing dosing regimens beyond cardiac surgical patients.