Secondary Logo

Journal Logo

Neuroscience in Anesthesiology and Perioperative Medicine: Research Report

Use of Tranexamic Acid Is Associated with Reduced Blood Product Transfusion in Complex Skull Base Neurosurgical Procedures

A Retrospective Cohort Study

Mebel, Dmitry MD*; Akagami, Ryojo MD, MHSc, FRCSC, FAANS; Flexman, Alana M. MD, FRCPC*

Author Information
doi: 10.1213/ANE.0000000000001065

Complex skull base neurosurgical procedures are associated with significant blood loss (300–1500 mL) and often require the administration of allogeneic blood products.1,2 Skull base tumors may be associated with higher intraoperative blood loss, because these tumors may be vascular and invade into adjacent sinuses and vessels.2 In addition, tumors such as meningiomas may induce tissue plasminogen activator and cause fibrinolysis and intraoperative hemorrhage.3 Allogeneic blood transfusion during surgical procedures exposes patients to both infectious and noninfectious risks and is associated with increased morbidity and mortality.4,5 Furthermore, the risks and costs associated with allogeneic blood transfusions are increasing and highlight the need for blood conservation strategies in complex skull base neurosurgical procedures.6,7

Tranexamic acid is a synthetic derivative of the amino acid, lysine, that reversibly blocks lysine-binding sites on plasminogen molecules, thereby preventing its conversion to plasmin and inhibiting fibrin clot dissolution.8 Currently, little is known about the efficacy of tranexamic acid in neurosurgical tumors, although 2 case reports and a small, randomized clinical trial have supported a role for tranexamic acid for significant intraoperative bleeding in pediatric neurosurgical procedures.9–11 Tranexamic acid may be particularly advantageous in skull base procedures, because even small amounts of bleeding pose unique challenges, particularly near eloquent cranial base structures where coagulation should be minimized.

Tranexamic acid has a favorable safety profile, but several potential adverse effects must be considered. Although tranexamic acid has been used extensively to prevent rebleeding in patients with aneurysmal subarachnoid hemorrhage, further research suggested that this therapy increased the risk of cerebral ischemia.12 Recently, the use of tranexamic acid in the cardiac surgery population has been associated with an increased risk of seizures, although this may be related to higher dose regimens.13,14 Finally, tranexamic acid has the potential to increase the risk of thromboembolism, a major cause of morbidity and death in neurosurgical patients.15

Overall, the safety and efficacy of tranexamic acid in complex skull base neurosurgical procedures is not known. Tranexamic acid has been used for complex skull base procedures at our institution since 2006, thus presenting us with a unique opportunity to review our patient cohort before and after tranexamic acid was introduced. Our primary objective was to determine the relationship between the use of tranexamic acid and the frequency of blood product transfusion. Our secondary objective was to determine the relationship between the use of tranexamic acid and the incidence of adverse events.

METHODS

After approval from the University of British Columbia Research Ethics Board (H13-00463), we performed a retrospective cohort study. We included all patients who underwent complex skull base neurosurgical procedures by a single surgeon from January 1, 2000 (start of the neurosurgical database) to April 3, 2013 (date of ethical approval). These included resection of skull base meningiomas, acoustic neuromas resections, and nonmeningioma extraaxial posterior fossa tumors. We categorized the procedures into 3 groups based on the general surgical approach (Supplemental Digital Content, Supplemental Appendix Table 1, https://links.lww.com/AA/B277). We excluded patients with a documented preexisting coagulopathy (hypercoagulability, bleeding diathesis or history of thromboembolic disease) or a documented allergy or sensitivity to tranexamic acid. We also excluded patients who did not undergo the scheduled procedure.

We abstracted demographic information, data about the diagnosis and procedure, use of tranexamic including dose, perioperative blood transfusion, laboratory indices, and perioperative complications from the electronic and paper medical record. Our primary outcome measure was perioperative transfusion of allogeneic blood products (packed red blood cells, fresh frozen plasma, platelets, or cryoprecipitate). Our secondary outcomes were the potential complications of tranexamic acid (seizure, deep vein thrombosis, pulmonary embolism, and other thromboembolic complications). We did not collect information about estimated surgical blood loss, because this number was often missing or unreliable.

Statistical Analysis

First, we compared the group of patients who received and who did not receive tranexamic acid using a χ2 test, Fisher exact test, Student t test, simple regression, or analysis of variance, as appropriate, to understand the population who were given tranexamic acid in our cohort. We first chose to estimate the conditional treatment effect of tranexamic acid on risk of transfusion first, because the decision to administer tranexamic acid was not random and required careful adjustment for the factors that influenced this decision. Although the individual anesthesiologist may have influenced the decision to transfuse and confounded the results, we did not include the anesthesiologist as a potential predictor in our model, because well >50 individuals were involved with these cases during the study period with no single anesthesiologist caring for >5% of the cases. We looked at potential predictors of perioperative transfusion by comparing the baseline features of patients who were transfused with those who were not, including the variable of interest (tranexamic acid administration) using similar statistical tests (univariate analysis). We next built a multivariate logistic regression model based on the candidate predictors associated with the perioperative transfusion (P < 0.10) on univariate analysis. Included variables were assessed for collinearity and interactions. We assessed the performance of the model through assessments of discrimination and calibration. Model discrimination was evaluated using the area under the receiver operating characteristic curve (C-statistic). Model calibration was assessed using the Hosmer-Lemeshow goodness-of-fit test.

To understand the marginal effects of tranexamic acid on transfusion given the possibility of a future randomized trial, we also conducted a propensity score analysis. First, we created a predictive logistic regression model for receiving tranexamic acid using data after 2005 (when tranexamic acid was used) and assigned a propensity score to each patient in the database. Next, we created a logistic regression model to predict transfusion and included tranexamic acid and the propensity score as a covariate. The model performance was assessed as described earlier. We calculated an adjusted risk ratio for tranexamic acid from the adjusted odds ratio (OR) using generalized linear regression with a log link and binomial distribution. All data analysis was performed using STATA 12.1 (StataCorp, TX) and R 3.13 (R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org/. Accessed March 10, 2015).

RESULTS

We identified 547 patients from our institutional neurosurgical database. Of these patients, 28 were excluded (13 charts were unable to be accessed by medical records; 8 procedures did not fall under the specified list or date range; 6 patients had preexisting coagulopathies; and 1 surgical procedure was abandoned) leaving 519 patients in the data set for analysis. Of the total population, 245 patients received tranexamic acid and 274 patients did not. All patients who received tranexamic acid were operated on after 2006 (when tranexamic acid was introduced into clinical practice at our institution) and 48% of the patients who did not receive tranexamic acid were operated on after 2006. The patients had a variety of pathologic diagnoses, although the majority was diagnosed with either an acoustic neuroma (55%) or a meningioma (25%; see Supplemental Appendix Tables 1 and 2 for a complete list of procedures and histologic diagnoses, Supplemental Digital Content 1, https://links.lww.com/AA/B277).

The cohort of patients who received tranexamic acid had a similar demographic profile and laboratory indices to the cohort who did not (Table 1). On average, the cohort who received tranexamic acid had larger tumors (3.5 vs 2.9 cm, P < 0.001) and had longer operations (7.2 vs 6.2 hours, P < 0.001). The average total dose of tranexamic acid given was 37 (SD, 47) mg/kg. Our typical institutional practice was to give an IV bolus dose of tranexamic acid (10–25 mg/kg) followed by an infusion (5–10 mg/kg/h).

T1-29
Table 1:
Study Population and Outcomes, Stratified by Exposure to Tranexamic Acid

The overall rate of transfusion of packed red blood cells was 9.8%, and patients who were transfused received a median of 2 (range, 1–14) units of packed red blood cells (Table 1). In addition, 18 patients in our study population were transfused a total of 16 units of fresh frozen plasma, 4 units of platelets, and 2 units of cryoprecipitate intraoperatively. Overall, the rate of transfusion of any blood product (packed red blood cells, fresh frozen plasma, platelets, or cryoprecipitate) was 10.2% (Table 1). Although the rate of transfusion generally declined over time, there was a marked decrease after 2006 (Fig. 1). The majority of patients was transfused intraoperatively; however, 2 of the 51 transfused patients received blood products postoperatively. Patients who received tranexamic acid had a 6% absolute reduction in the frequency of any blood product transfusion from 13% to 7% (unadjusted OR for transfusion, 0.54; 95% confidence interval [CI], 0.30–0.98; P = 0.04). In patients who were transfused with packed red blood cells, hemoglobin levels in the recovery room were similar during the range of surgery years (spearman ρ = −0.049, P = 0.74).

F1-29
Figure 1:
Incidence of transfusion over time. Tranexamic acid was introduced in 2006.

On unadjusted analysis, male sex, preoperative hemoglobin, tumor diameter, procedural category, surgery year, and use of tranexamic acid were associated with perioperative transfusion (P < 0.05; Supplemental Digital Content 1, Supplemental Appendix Table 3, https://links.lww.com/AA/B277). We used 503 patients to construct the model because 16 patients were missing covariate data. When preoperative hemoglobin was added to the model, male sex was no longer an independent predictor and was removed from the model. Tranexamic acid was strongly associated with a reduced risk of transfusion when included in a model with preoperative hemoglobin, tumor size, and procedural category (adjusted OR, 0.32; 95% CI, 0.15–0.65, P = 0.002; Table 2). We did not identify collinearity, although there was a significant interaction between procedural category and tumor size (P = 0.011) as well as tumor size and preoperative hemoglobin (P = 0.039). Model performance was not improved by inclusion of the interaction terms, and these were not included for simplicity and because they did not affect the variable of interest. Transformation of continuous variables (preoperative hemoglobin and tumor diameter) did not significantly improve the model fit or qualitatively change the results. The c-statistic in the final model was 0.88, indicating excellent discrimination. Calibration was also good (Hosmer-Lemeshow χ2 test = 9.01, P = 0.34). Next, we investigated potential confounding by surgery year and added this variable to our model. When both surgery year and tranexamic acid were included in the model, tranexamic acid became an insignificant predictor (adjusted OR, 0.91; 95% CI, 0.33–2.54; P = 0.90); however, a significant interaction was seen between year of surgery and tranexamic acid (P = 0.026). A logistic regression model with both variables and the interaction term demonstrated that the effect of tranexamic acid appeared to be greater at earlier time points compared with later and that although tranexamic acid was associated with reduced transfusion, this effect was not significant when centered at the year 2006 (adjusted OR, 0.51; 95% CI, 0.13–1.78, P = 0.31; Supplemental Digital Content 1, Supplemental Appendix Table 4, https://links.lww.com/AA/B277). Again, model discrimination (c-statistic = 0.89) and calibration (Hosmer-Lemeshow χ2 test = 4.48, P = 0.81) were acceptable.

T2-29
Table 2:
Multivariate Model Predicting Perioperative Transfusion in Complex Skull Base Neurosurgery

We were able to create a propensity score to predict the probability of receiving tranexamic acid using the data collected after 2005 when tranexamic acid was used. The main predictors of receiving tranexamic acid in the multivariate model were tumor size, procedural category, and year of surgery (P < 0.001 for all). Tranexamic acid use remained a significant predictor of reduced transfusion when included in a logistic regression model with the propensity score as a covariate (adjusted OR, 0.27; 95% CI, 0.13–0.56; P = <0.001). Model discrimination (c-statistic = 0.68) and calibration (Hosmer-Lemeshow χ2 test = 12.57, P = 0.13) were acceptable. By using this model, tranexamic acid was associated with an adjusted risk ratio of 0.33 (95% CI, 0.18–0.61) for receiving a transfusion.

T3-29
Table 3:
Perioperative Complication Rates, Stratified by Exposure to Tranexamic Acid

Both seizure and thrombotic complications were infrequent and similar between the 2 groups (Table 3). Two patients in the tranexamic group and 6 in the control group experienced a variety of thrombotic complications, primarily deep vein thrombosis or pulmonary embolism. Seizure was relatively rare (2.7% overall) with 8 patients in the tranexamic group and 6 patients in the control group experiencing this complication. Although 7 patients (3 and 4 patients in the tranexamic and no tranexamic groups, respectively) were initially identified as having postoperative cerebral ischemia, on further review, none were attributed to a thromboembolic etiology. The total length of hospital stay was similar between the 2 groups (mean, 7 vs 8 days, P = 0.22 in the tranexamic and control groups, respectively).

DISCUSSION

In our study, we demonstrated that the administration of tranexamic acid was associated with a reduced frequency of transfusion of allogeneic blood products at our institution. Furthermore, the relationship between administration of tranexamic acid and reduced transfusion was strengthened when we adjusted for differences in tumor size and procedural category and confirmed in a propensity score analysis. Of note, we detected a significant interaction when accounting for year of surgery and cannot exclude significant historical confounding. Our study is novel, because there are very few reports in the literature regarding the use of tranexamic acid in elective neurosurgical procedures. Our results suggest that tranexamic acid is not harmful and support the need for a future randomized trial to confirm its benefits and safety.

Tranexamic acid has been shown to be effective in reducing blood transfusion in a variety of settings, including spine, major orthopedic, obstetric, urologic, and trauma patients.16–20 Tranexamic acid is an effective antifibrinolytic and acts as a plasmin inhibitor to prevent clot breakdown. The reduced risk of transfusion associated with use of tranexamic acid in our study on neurosurgical patients is supported by a recent study on pediatric neurosurgical patients.11 This trial randomized children undergoing craniosynostosis surgery to receive tranexamic acid or placebo and demonstrated reduced blood loss and a reduced risk of exposure to allogeneic blood. Furthermore, others have speculated that the benefits of tranexamic acid may extend beyond clot stabilization through suppression of proinflammatory mediators.21 Tranexamic acid may have added benefit to patients undergoing neurosurgery and potential neurologic injury, although this is currently unknown.

Our results demonstrate that transfusion rates decreased during the study period. The effect of time is an important confounder and makes our results more difficult to interpret. Although tranexamic acid was associated with reduced transfusion, this effect had a significant interaction with surgical year and did not independently predict transfusion in this model. It is possible that tranexamic acid was not associated with reduced transfusion when accounting for surgery year, although it is also likely that we were underpowered to detect a signal when accounting for changes over time. Recently, there has been a shift toward a more restrictive transfusion practice after the publication of randomized trials supporting the safety of lower transfusion triggers in critically ill patients.22,23 However, further exploratory analysis of our data demonstrated stable recovery room hemoglobin values during the study period in those who received a blood transfusion, indirectly suggesting that the transfusion trigger did not change. The changes in other aspects of anesthetic and surgical technique during the study period likely influenced the trends in transfusion. For these reasons, the effects of tranexamic acid on transfusion should be interpreted cautiously until a randomized controlled trial is completed.

The rate of adverse events observed in our cohort was similar, regardless of whether they received tranexamic acid. Although tranexamic acid has been associated with an increased rate of seizures after cardiac surgery,13,24 the rate of seizure was similar in patients who received tranexamic acid (3.3% vs 2.2% in the tranexamic acid vs control group, respectively). Similarly, the rate of other thrombotic complications was not increased in the group that received tranexamic acid, consistent with recent publications in major orthopedic and spine surgery.25,26 Although our study was not powered to detect a difference in these outcomes, the use of tranexamic acid in our study population did not appear to increase the risk of thrombosis or seizure.

Our study has several strengths. Very few studies have examined the effect and safety of tranexamic acid in elective neurosurgical procedures. To our knowledge, we are the first to describe the use of tranexamic acid in a large cohort of patients undergoing complex skull base neurosurgery. In addition, our primary end point was clinically relevant and important to patients and practitioners: transfusion of allogeneic blood products. Finally, our study results support the safety of tranexamic acid in neurosurgical patients and suggest that the theoretical risks of seizure and thrombotic complications are not increased. Should a randomized trial be conducted in the future, our propensity score analysis suggest that approximately 227 subjects would be needed in each group to determine the effect of tranexamic acid on transfusion (assuming a baseline rate of transfusion of 10%, type I error of 0.05 and type II error of 0.2, and 2-sided significance testing).

Our study has limitations given its retrospective, observational study design. The most important limitation of our data is historical confounding that may have occurred through change in clinical practice during the study period, as discussed earlier. In addition, the decision to administer tranexamic acid was not random, and the dose was not standardized; there may be other unmeasured confounders that may have biased our study results. We were unable to adjust for the heterogeneity of procedures and tumor subtypes, although we attempted to adjust for surgical approach using 3 categories. Finally, our results may not be generalizable to other centers, because the study population represents a single center and surgeon.

CONCLUSIONS

Tranexamic acid has been shown to reduce blood loss and transfusion in a variety of surgical procedures and is well tolerated with minimal side effects. We now describe the use of tranexamic acid in a large cohort of patients undergoing complex skull base neurosurgery. Our results demonstrate that tranexamic acid use is associated with reduced transfusion rates in our study population, with no apparent increase in seizure or thrombotic complications. Our data support the need for further randomized clinical trials to determine the efficacy and safety of tranexamic acid on perioperative blood loss during complex skull base neurosurgery.

DISCLOSURES

Name: Dmitry Mebel, MD.

Contribution: This author helped design the study, collect the data, analyze the data, and helped write the first draft of the manuscript.

Attestation: Dmitry Mebel attests to the integrity of the original data and the analysis reported in this manuscript and approved the final manuscript.

Name: Ryojo Akagami, MD, MHSc, FRCSC, FAANS.

Contribution: This author helped design the study, collect the data, analyze the data, and prepare the manuscript.

Attestation: Ryojo Akagami approved the final manuscript.

Name: Alana M. Flexman, MD, FRCPC.

Contribution: This author helped design the study, conduct the study, collect the data, analyze the data, and wrote the first draft of the manuscript.

Attestation: Alana M. Flexman attests to the integrity of the original data and the analysis reported in this manuscript and approved the final manuscript and is the archival author.

This manuscript was handled by: Gregory Crosby, MD.

ACKNOWLEDGMENTS

We thank Mr. Rick White (Consultant Statistician, Department of Statistics at the University of British Columbia), for his assistance with the data analysis.

REFERENCES

1. Oka H, Kurata A, Kawano N, Saegusa H, Kobayashi I, Ohmomo T, Miyasaka Y, Fujii K. Preoperative superselective embolization of skull-base meningiomas: indications and limitations. J Neurooncol. 1998;40:67–71
2. Jellish WS, Murdoch J, Leonetti JP. Perioperative management of complex skull base surgery: the anesthesiologist’s point of view. Neurosurg Focus. 2002;12:e5
3. Tsuda H, Oka K, Noutsuka Y, Sueishi K. Tissue-type plasminogen activator in patients with intracranial meningiomas. Thromb Haemost. 1988;60:508–13
4. Dodd RY. The risk of transfusion-transmitted infection. N Engl J Med. 1992;327:419–21
5. Klein HG. Allogeneic transfusion risks in the surgical patient. Am J Surg. 1995;170:21S–6S
6. Blanchette CM, Wang PF, Joshi AV, Asmussen M, Saunders W, Kruse P. Cost and utilization of blood transfusion associated with spinal surgeries in the United States. Eur Spine J. 2007;16:353–63
7. Koch CG, Li L, Duncan AI, Mihaljevic T, Cosgrove DM, Loop FD, Starr NJ, Blackstone EH. Morbidity and mortality risk associated with red blood cell and blood-component transfusion in isolated coronary artery bypass grafting. Crit Care Med. 2006;34:1608–16
8. Dunn CJ, Goa KL. Tranexamic acid: a review of its use in surgery and other indications. Drugs. 1999;57:1005–32
9. Bharath K, Bhagat H, Mohindra S. Use of tranexamic acid as a rescue measure to achieve hemostasis after massive blood loss in a pediatric neurosurgical patient. J Neurosurg Anesthesiol. 2011;23:376–7
10. Phi JH, Goobie SM, Hong KH, Dholakia A, Smith ER. Use of tranexamic acid in infants undergoing choroid plexus papilloma surgery: a report of two cases. Paediatr Anaesth. 2014;24:791–3
11. Goobie SM, Meier PM, Pereira LM, McGowan FX, Prescilla RP, Scharp LA, Rogers GF, Proctor MR, Meara JG, Soriano SG, Zurakowski D, Sethna NF. Efficacy of tranexamic acid in pediatric craniosynostosis surgery: a double-blind, placebo-controlled trial. Anesthesiology. 2011;114:862–71
12. Roos Y, Rinkel G, Vermeulen M, Algra A, van Gijn J. Antifibrinolytic therapy for aneurysmal subarachnoid hemorrhage: a major update of a cochrane review. Stroke. 2003;34:2308–9
13. Manji RA, Grocott HP, Leake J, Ariano RE, Manji JS, Menkis AH, Jacobsohn E. Seizures following cardiac surgery: the impact of tranexamic acid and other risk factors. Can J Anaesth. 2012;59:6–13
14. Royston D. Tranexamic acid in cardiac surgery: is there a cause for concern? Crit Care. 2010;14:194
15. Khaldi A, Helo N, Schneck MJ, Origitano TC. Venous thromboembolism: deep venous thrombosis and pulmonary embolism in a neurosurgical population. J Neurosurg. 2011;114:40–6
16. Ducloy-Bouthors AS, Jude B, Duhamel A, Broisin F, Huissoud C, Keita-Meyer H, Mandelbrot L, Tillouche N, Fontaine S, Le Goueff F, Depret-Mosser S, Vallet B, Susen SEXADELI Study Group. . High-dose tranexamic acid reduces blood loss in postpartum haemorrhage. Crit Care. 2011;15:R117
17. Neilipovitz DT, Murto K, Hall L, Barrowman NJ, Splinter WM. A randomized trial of tranexamic acid to reduce blood transfusion for scoliosis surgery. Anesth Analg. 2001;93:82–7
18. Rannikko A, Pétas A, Taari K. Tranexamic acid in control of primary hemorrhage during transurethral prostatectomy. Urology. 2004;64:955–8
19. Zufferey PJ, Miquet M, Quenet S, Martin P, Adam P, Albaladejo P, Mismetti P, Molliex STranexamic Acid in Hip-Fracture Surgery (THIF) Study. . Tranexamic acid in hip fracture surgery: a randomized controlled trial. Br J Anaesth. 2010;104:23–30
20. Shakur H, Roberts I, Bautista R, Caballero J, Coats T, Dewan Y, El-Sayed H, Gogichaishvili T, Gupta S, Herrera J, Hunt B, Iribhogbe P, Izurieta M, Khamis H, Komolafe E, Marrero MA, Mejia-Mantilla J, Miranda J, Morales C, Olaomi O, Olldashi F, Perel P, Peto R, Ramana PV, Ravi RR, Yutthakasemsunt S. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet. 2010;376:23–32
21. Levy JH. Antifibrinolytic therapy: new data and new concepts. Lancet. 2010;376:3–4
22. Salpeter SR, Buckley JS, Chatterjee S. Impact of more restrictive blood transfusion strategies on clinical outcomes: a meta-analysis and systematic review. Am J Med. 2014;127:124–31.e3
23. Hébert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, Tweeddale M, Schweitzer I, Yetisir E. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators, Canadian Critical Care Trials Group. N Engl J Med. 1999;340:409–17
24. Murkin JM, Falter F, Granton J, Young B, Burt C, Chu M. High-dose tranexamic acid is associated with nonischemic clinical seizures in cardiac surgical patients. Anesth Analg. 2010;110:350–3
25. Yuan C, Zhang H, He S. Efficacy and safety of using antifibrinolytic agents in spine surgery: a meta-analysis. PLoS One. 2013;8:e82063
26. Poeran J, Rasul R, Suzuki S, Danninger T, Mazumdar M, Opperer M, Boettner F, Memtsoudis SG. Tranexamic acid use and postoperative outcomes in patients undergoing total hip or knee arthroplasty in the United States: retrospective analysis of effectiveness and safety. BMJ. 2014;349:g4829

Supplemental Digital Content

© 2016 International Anesthesia Research Society