The Efficacy and Safety of Topical Tranexamic Acid

Transfusion Medicine Reviews xxx (2018) xxx–xxx

Contents lists available at ScienceDirect

Transfusion Medicine Reviews

journal homepage: www.tmreviews.com

The efﬁcacy and safety of topical tranexamic acid: A systematic review and meta-analysis

Joshua Montroy a,b, Brian Hutton a,b, Preveshen Moodley c, Nicholas A. Fergusson a, Wei Cheng a, Alan Tinmouth a, Luke T. Lavallée a,c, Dean A. Fergusson a,b,c, Rodney H. Breau a,c,⁎ a Clinical Epidemiology Program, Centre for Practice Changing Research, Ottawa Hospital Research Institute, Ottawa, ON, Canada b University of Ottawa, School of Epidemiology, Community Medicine and Preventive Medicine, Ottawa, ON, Canada c Division of Urology, Department of Surgery, The Ottawa Hospital, Ottawa, Ontario, Canada article info abstract

Available online xxxx Tranexamic acid (TXA) is an effective hemostatic agent used for the reduction of blood loss and transfusion. How- ever, the safety profile of TXA remains in question due to a potential increased risk of venous thromboembolism. Keywords: By applying TXA topically as opposed to intravenously, systemic absorption may be reduced and unwanted side- Tranexamic acid effects mitigated. The objective of our review is to investigate the efficacy and safety of topically applied Lysine analogue tranexamic acid compared to both placebo, and the intravenous administration. Cochrane Central Register of Surgery Controlled Trials, MEDLINE, Embase, ISI Web of Science, PubMed, and Clinicaltrials.gov were searched from in- Blood transfusion ception to November, 2016. We included randomized controlled trials that compared topical tranexamic acid Venous thromboembolism Topical to either placebo (or standard care) or intravenous administration, in adult patients. Surgical and non-surgical trials were included. Abstract, full-text selection, data extraction and risk of bias assessment were all performed in duplicate. In total, 67 studies involving 6,034 patients met inclusion criteria. The majority of trials evaluated orthopedic procedures. Compared to placebo, the administration of topical TXA significantly reduced the odds of receiving a blood transfusion (pooled OR 0.28, 95% CI 0.20 to 0.38; P b 0.001) and significantly reduced mean blood loss (WMD -276.6, 95% CI -327.8 to -225.4; P b 0.0001). When compared to the intravenous administration, there was no difference between the two groups in terms of transfusion requirements (pooled OR 1.03, 95% CI 0.72 to 1.46; P=0.88) or blood loss (WMD -21.95, 95% CI -66.61 to 27.71; P=0.34). There was no difference in the odds of developing a venous thromboembolic complication between the topical TXA and control groups (pooled OR=0.78, 95% CI 0.47 to 1.29; P=0.33) or the topical and intravenous groups (pooled OR=0.75, 95% CI 0.39 to 1.46; P=0.40). The topical application of TXA effectively reduces both transfusion risk and blood loss compared to placebo, without increasing thromboembolic risks. There were no major differences between topical and intravenous tranexamic acid with respect to safety and efficacy, and both were superior to placebo with regards to blood loss and transfusion requirements. Further study of the topical application is required outside of the field of orthopedics. © 2018 Elsevier Inc. All rights reserved.

Contents

Background&Rationale...... 0 Methods...... 0 . EligibilityCriteria...... 0 . LiteratureSearchStrategy...... 0 . StudySelectionProcess...... 0 . DataExtractionandRiskofBiasAssessment...... 0 . DataAnalysis...... 0 Results...... 0 . CharacteristicsofIncludedStudies...... 0 . StudyQualityandRiskofBias...... 0

⁎ Corresponding author at: Rodney H. Breau, MsC, MD, FRCSC, Surgical Oncology, Division of Urology, Associate Scientist, Ottawa Hospital Research Institute, The Ottawa Hospital/Uni- versity of Ottawa, 501 Smyth Rd, Box 222, Ottawa, ON K1H8L6, Canada. E-mail address: [email protected] (R.H. Breau). https://doi.org/10.1016/j.tmrv.2018.02.003 0887-7963/© 2018 Elsevier Inc. All rights reserved.

. EffectoftopicalTXAontransfusionrate...... 0 . Comparedtoplacebo...... 0 . Comparedtointravenousadministration...... 0 . EffectoftopicalTXAdoseontransfusionrate...... 0 . EffectoftopicalTXAonbloodloss...... 0 . Comparedtoplacebo...... 0 . Comparedtointravenousadministration...... 0 . EffectoftopicalTXAontherateofvenousthromboembolicevents...... 0 . Comparedtoplacebo...... 0 . Comparedtointravenousadministration...... 0 . EffectoftopicalTXAontheriskofdeath...... 0 . EffectoftopicalTXAontheriskofstroke...... 0 . EffectoftopicalTXAontheriskofmyocardialinfarction...... 0 . PublicationBias...... 0 Discussion...... 0 Conclusion...... 0 Funding...... 0 ConﬂictofInterestStatement...... 0 AppendixA. Supplementarydata...... 0 AppendixA. Supplementarydata...... 0 References...... 0

Background & Rationale interventions, we performed a systematic review and meta-analysis to evaluate the efficacy and safety of topically administered tranexamic Allogenic blood transfusions, although potentially life-saving, acid. may be associated with an increased risk of infection and immune reactions, which may be a major source of morbidity for patients [1,2]. Methods A number of strategies are available that limit surgical blood loss and transfusions. Use of hemostatic agents has increased over the last Our review was registered on Prospero (no. CRD42016035902). A decade [3] with antifibrinolytics emerging as popular agents in review protocol was prepared according to the PRISMA-P checklist both medical and surgical practice. Antifibrinolytic lysine analogues [21] and is available upon request. The final report was prepared in ac- have been shown to be effective in reducing blood loss during surgery cordance with the Preferred Reporting Items for Systematic Reviews and trauma as well as non-surgical diseases such as anemia related and Meta-analyses (PRISMA) [22]. to cancer [4-6]. Tranexamic acid (TXA) is a lysine analogue and is a widely used antifibrinolytic. Systemic use of TXA has been shown in Eligibility Criteria meta-analyses to be safe and effective at reducing blood loss in cardiac [7], orthopedic [8], pelvic [5], and spinal surgeries [9], among others, We included randomized controlled trials (RCTs) that compared and is listed on the World Health Organization’s List of Essentials topical tranexamic acid to either placebo (or standard care) or intrave- Medicines. nously administered tranexamic acid in adult patients. Surgical and Due to the mechanism of action of lysine analogues, there exists a non-surgical trials were included. We did not restrict inclusion by theoretical increased risk of developing venous thromboembolic (VTE) drug dose or timing of administration. Our primary outcome was the complications such as deep venous thrombosis (DVT) or pulmonary risk of blood transfusion defined as the number of patients requiring embolism (PE) [10]. In one of the largest trauma trials to date, the at least one unit of red blood cells transfused. Our secondary outcomes CRASH-2 trial (Clinical Randomisation of an Antifibrinolytic in Signifi- included blood loss and adverse events (mortality, myocardial infarc- cant Hemorrhage), the efficacy and safety of TXA was assessed in over tion (MI), stroke, and venous thromboembolic events). We did not 20,000 adult trauma patients and no increase in thromboembolic events impose any language restrictions. was reported. However the precision of the safety estimate was low and therefore a potential for increased risk of VTE was not ruled out [6]. Literature Search Strategy Many systematic reviews have examined the relationship between lysine analogues and VTE risk, with point estimates remaining imprecise A comprehensive literature search of indexed databases was con- and thus unable to rule out a potential increased risk of VTE [5,11-13]. ducted to identify all relevant studies in collaboration with an informa- Because of the theoretical risk of VTE, lysine analogues have not been tion specialist and a clinical expert in the field. We had a separate adopted as part of routine practice for most major surgeries except per- information specialist review the search strategy to check for consis- haps in cardiac and orthopedic procedures [14-16]. tency according to the PRESS framework [23]. The following databases There remains a need to assess safety and efficacy of different modes were searched: Cochrane Central Register of Controlled Trials, of TXA delivery that may maintain the benefits of systemic administra- MEDLINE, Embase, ISI Web of Science, PubMed, and Clinicaltrials.gov. tion while reducing the risk of VTE [17,18]. The plasma concentration of Grey literature was searched via Google using the same search topically applied TXA has been shown to be approximately 90% less terms. We searched from database inception to September 10th 2016 than when the medication is administered intravenously [19,20]. Con- (Appendix 1). versely, it is possible that high local tissue drug concentrations from topical application may increase the risk of adverse events. Study Selection Process A previous systematic review showed that local TXA was effective, but safety could not definitively established [12].Inrecentyears,there Citations from the literature search were collated and duplicate have been a large number of new trails published assessing the effect studies removed. Titles and abstracts were screened for inclusion by of topical TXA. Given the need for safe and effective hemostatic two independent reviewers (JM, PM or NF). Titles and abstracts deemed

Data Analysis

Studies were pooled using Cochrane Review Manager, version 5.3 (Cochrane Collaboration, Oxford, United Kingdom) and Open Meta- Analyst for Windows 10 (http://www.cebm.brown.edu/open_meta). For dichotomous outcomes, odds ratios were calculated using a random effects frequentist analysis based on the Der-Simonian Laird model and presented with accompanying 95% confidence intervals. This method allows for the inclusion of continuity corrections of 0.5 for all zero cells across outcomes, allowing us to estimate odds ratios for studies reporting no events. For outcomes with zero events, a sensitivity analysis was also performed with a fixed effects meta-analysis using Peto odds ratios. For continuous outcomes, a mean difference and 95% confidence intervals are presented. Pooled mean differences were calculated using random effects inverse variance meta-analysis. Statistical heterogeneity was assessed using the I2 statistic, as well as the Chi2 test or the Cochrane Q test, depending on the analysis method. An I2 value of N50% was considered to indicate important heterogeneity requiring additional exploration. For the Chi2 test and the Cochrane Q test, a p-value of b0.1 was deemed to indicate heterogeneity requiring additional exploration. The presence of publication bias was assessed using funnel plots if more than 6 studies were identified. Where sufficient data were available, we conducted subgroup analyses to determine whether the effect of topical TXA on blood transfusion and blood loss varied by the type of surgical procedure, as well as the baseline rate of transfusion. A sensitivity analysis was performed on outcomes with sufficient data, to determine if the effect of topical TXA varied when analyses were restricted to trials with adequate allocation concealment. A random-effects meta-regression was performed to determine if the association between topical TXA and blood transfusion rate was affected by the dose of topical TXA used in the trial.

Results

The electronic search identified 1,245 titles and abstracts after the removal of duplicates, of which 1,149 were excluded based on irrele- vancy to our study question. Of 96 full-text articles reviewed, 29 did not meet eligibility criteria and were excluded. Sixty-seven trials met eligibility representing 6,034 patients [25-75,76-91](Fig 1). Forty-two trials compared topical TXA to placebo, 14 compared topical TXA to IV TXA, and 11 were 3 arm trials comparing topical TXA to IV TXA and placebo. Eleven of the 67 articles did not report sufficient information for inclusion in our meta-analyses [25-29,35,43,44,62,79,91]. Fig 1. Selection flow diagram.

Characteristics of Included Studies potentially relevant were recorded and the full text articles obtained. The same reviewers screened the full text articles for final eligibility, Included articles were published between 1979 and 2016 from 24 with disagreements settled by consultation of a third party (RB) to different countries (Table 1). Study sample sizes ranged from 30 pa- achieve consensus. The study selection process was documented and tients to 333 patients (median = 84). Sixty-six of the studies assessed reported using a PRISMA flow diagram (Fig 1) [22]. the efficacy of topical TXA in surgical patients. Thirty trials involved knee arthroplasty, 11 hip arthroplasty, 10 cardiac surgery, four dental surgery, two otolaryngological surgery, two spinal surgery, two Data Extraction and Risk of Bias Assessment orthognathic surgery, one transurethral prostate resection, one reduction mammoplasty, one pulmonary resection, one hip and knee Two reviewers (JM, PM or NF) extracted all relevant data from the arthroplasty, and one shoulder arthroplasty. The sole study which included studies using a standardized and piloted data extraction form assessed non-surgical patients studied the efficacy of topical TXA in ep- in Microsoft Excel (version 15.0, Microsoft Corporation, Seattle, istaxis patients. Two trials consisted of two distinct topical TXA arms Washington, USA). We collected information related to publication (different doses, administration techniques) and therefore were in- traits (year of publication, journal, and authorship list), study popula- cluded twice in analysis for which they contributed data. For these trials, tion (e.g, eligibility criteria, age, gender, and comorbidities), the inter- the control group size and events were split in half. ventions compared (type of comparator, dose, and timing), study In 32 of the 66 surgical trials, topical TXA was either poured (wound design (methods, setting, sample size, and number of centres), clinical irrigation) or sprayed directly onto the operative site. In 21 trials involv- endpoints (transfusion, blood loss, death, stroke, myocardial infarction, ing hip or knee arthroplasty, topical TXA was administered via an intra- and venous thromboembolic events; endpoint definitions were also ab- articular injection. Three trials used a combination of these two tech- stracted), and risk of bias details. The risk of bias was assessed by two in- niques. Another trial assessed both these techniques in two separate dependent reviewers using the Cochrane Risk of Bias tool [24]. groups of patients. In two of the dental surgery trials, the wound was

Table 1 Characteristics of included studies

Study Host Sample Procedure Comparator Method of topical TXA Administration Transfusion Blood loss VTE risk Country Size risk reported reported reported (ies) (Y/N) (Y/N) (Y/N)

Chen (2016) Singapore 100 TKA IV Intra-articular injection Y Y Y Drosos (2016) Greece 90 TKA Placebo and N.R. Y Y Y IV Goyal (2016) Australia 168 TKA IV Intra-articular injection Y N Y Keyhani (2016) Iran 120 TKA Placebo and Intra-articular injection Y Y Y IV May (2016) USA 131 TKA IV Intra-articular injection Y Y Y North (2016) USA 139 THA IV Wound irrigation prior to closure Y Y Y Pinsornsak (2016) Thailand 60 TKA IV Peri-articular injection Y Y Y Tzatzairis (2016) Greece 120 TKA Placebo and Intra-articular injection Y Y Y IV Xie (2016) China 210 THA IV Intra-articular injection Y Y Y Zhang (2016) China 75 THA Placebo and Intra-articular injection Y Y Y IV Aggarwal (2015) India 70 TKA IV Wound irrigation prior to closure Y Y Y Aguilera (2015) Spain 150 TKA Placebo and Sprayed via syringe onto wound prior to closure Y Y N IV Ali Shah (2015) Pakistan 100 Open heart Placebo Wound irrigation prior to closure N Y N surgeries Ausen (2015) Norway 56 Bilateral reduction Placebo Poured into wound prior to closure N N N mammoplasty Digas (2015) Greece 85 TKA Placebo and Intra-articular injection Y Y Y IV Eftekharian Iran 56 Bimaxillary Placebo Wound irrigation throughout procedure Y Y Y (2015) osteotomy Gillespie (2015) USA 111 TSA Placebo Wound irrigation prior to closure Y Y Y Lin (2015) Taiwan 120 TKA Placebo Intra-articular injection Y Y Y Oztas (2015) Turkey 90 TKA Placebo and Intra-articular injection Y Y Y IV Taheriazam Iran 80 THA IV Wound irrigation N N N (2015)* Vandesande Belgium 63 THA Placebo N.R. Y Y N (2015)* Wang, C (2015) China 60 TKA Placebo Intra-articular injection Y Y Y Wang, G (2015) China 100 TKA Placebo Intra-articular injection Y Y Y Xu (2015) China 224 THA Placebo Intra-articular injection Y Y Y Yang (2015) China 80 TKA Placebo Intra-articular injection Y Y Y Emara (2014) Egypt 60 Hemiarthroplasty Placebo and Wound irrigation prior to closure Y Y Y surgeries IV Gomez-Barrena Spain 78 TKA IV Wound irrigation prior to closure and YYY (2014) intra-articular injection Hosseini (2014) Iran 71 off-pump CABG Placebo Wound irrigation prior to closure N Y Y Martin (2014) USA 100 TKA/THA Placebo Wound irrigation prior to closure Y N Y Patel (2014) USA 89 TKA IV Wound irrigation prior to closure Y N Y Sarzaeem (2014) Iran 200 TKA Placebo and Wound irrigation prior to closure (grp 2) YYY IV Intra-articular injection (grp 3) Soni (2014) India 80 TKA IV Wound irrigation prior to closure Y Y Y Wei (2014) China 333 THA Placebo and Intra-operative wound irrigation and YYY IV intra-articular injection Yue (2014) China 103 Unilateral total hip Placebo TXA soaked gauze inserted into wound YYY replacement intra-operatively and intra-articular injection Alshryda United 157 TKA Placebo Solution sprayed onto wound prior to closure Y Y Y (TRANX-K) Kingdom (2013) Alshryda United 173 Unilateral total hip Placebo Solution sprayed onto wound prior to closure Y Y Y (TRANX-H) Kingdom replacement (2013) Georgiadis (2013) USA 101 TKA Placebo Wound irrigation prior to closure Y Y Y Nouraei (2013) Iran 80 Cardiac surgery Placebo Wound irrigation prior to closure Y Y Y Sa-ngasoonsgsong Thailand 135 TKA Placebo Intra-articular injection Y Y Y (2013) Van Elst (2013)* Belgium 67 THA Placebo Intra-articular injection N N Y Abdullah (2012)* Pakistan 52 Transurethral Placebo Bladder irrigation N N N resection of the prostate Canata (2012)* N.R. 96 TKA Placebo Inﬁltration of TXA N N N Dell’Amore (2012) Italy 89 Pulmonary Placebo Wound irrigation prior to closure Y Y Y resection Maniar (2012) India 206 TKA IV Wound irrigation prior to closure Y Y Y Roy (2012) India 50 TKA Placebo Intra-articular injection Y Y Y Seo (2012) South 150 TKA Placebo and Intra-articular injection Y Y Y Korea IV Baric (2011)* Croatia 60 Elective cardiac Placebo and “Administered topically” YYN

Table 1 (continued)

Study Host Sample Procedure Comparator Method of topical TXA Administration Transfusion Blood loss VTE risk Country Size risk reported reported reported (ies) (Y/N) (Y/N) (Y/N)

surgery IV Ishida (2011) Japan 100 TKA Placebo Intra-articular injection Y Y N Kaewpradub Thailand 40 Bimaxillary Placebo Wound irrigation, intra-operatively N Y N (2011) osteotomy Kurt (2011) Turkey 100 Cardiac surgery Placebo Wound irrigation prior to closure N Y N Sa-ngasoonsgsong Thailand 48 TKA Placebo Intra-articular injection Y Y Y (2011) Girgin (2010)* Turkey 60 Non-emergent IV Wound irrigation Y Y Y coronary bypass Saberi (2010) Iran 100 Spinal surgery Placebo Wound irrigation prior to closure Y Y N Wong (2010) Canada 124 TKA Placebo Wound soaked in TXA for 5 minutes prior to YYY closure Fawzy (2009) Saudi 38 Isolated CABG Placebo Wound irrigation prior to closure Y Y N Arabia Athanasiadis Australia 30 Sinus surgery Placebo Solution sprayed into nasal cavity N N N (2007) Baric (2007) Croatia 300 Elective cardiac Placebo Wound irrigation prior to closure Y Y N surgery Abul-Azm (2006) Egypt 200 Elective open Placebo Wound irrigation prior to closure N Y N heart surgery Jabalameli (2006) Iran 56 Sinus surgery Placebo “Administered topically” NYN Yasim (2005) Turkey 300 TKA Placebo Wound irrigation prior to closure N Y N Krohn (2002) Norway 30 Spinal surgery Placebo Wound irrigation prior to closure Y Y N De Bonis (2000) Italy 40 Primary coronary Placebo Wound irrigation prior to closure Y Y N artery surgery Blinder (1999) Israel 150 Dental extraction Placebo Mouthwash rinsed for 2 minutes, 4 times/day for 4 NNN days post-op Tibbelin (1995) Sweden 68 Epistaxis Placebo TXA gel applied to nasal cavity N N N Ramstrom (1993) Denmark 93 Dental surgery Placebo Wound irrigation prior to closure and mouthwash NNN & Sweden rinsed for 2 minutes, 4 times/day for 7 days post-op Sindet-Pedersen Denmark 45 Oral surgery Placebo Wound irrigation prior to closure and mouthwash NNN (1989) & Sweden rinsed for 2 minutes, 4 times/day for 7 days post-op Gersel-Pedersen Denmark 120 Bilateral molar Placebo TXA soaked cones applied to wound prior to NNN (1979) extraction closure

• = abstract only. irrigated and the patients were given a mouthwash to use for a number methodological criterion. The majority of studies described an adequate of days post-operatively. Post-operative mouthwash alone was given in randomization technique, although two trials were deemed to be at one dental surgery trial. In the last dental surgery trial, TXA soaked high risk of bias as one trial allocated patients into groups according to cones were applied to each tooth socket. Three trials were vague in the day of the week and one trial alternately assigned patients into their method of topical TXA application describing their technique as groups [29,40]. Less than half of the trials adequately described their al- “administered topically” and “inﬁltration of TXA”. Two trials did not re- location concealment technique, with the same two trials being deemed port their technique. In the one non-surgical trial, TXA gel was applied to be at high risk of bias in this domain. to the nasal cavity of epistaxis patients. The majority of trials were deemed to have adequately blinded their participants and personnel, as well as their outcome assessment for Study Quality and Risk of Bias transfusion rate and were therefore at a low risk for both performance bias and detection bias. The majority of studies had complete data or Figure 2 summarizes the risk of bias using the Cochrane risk of bias had adequately described their reasons for missing data, with one tool. The majority of studies had an unclear risk of bias for at least one study being deemed to be at high risk of attrition bias [38]. A small

Fig 2. Risk of bias summary.

Please cite this article as: Montroy J, et al, The efficacy and safety of topical tranexamic acid: A systematic review and meta-analysis, Transfus Med Rev (2018), https://doi.org/10.1016/j.tmrv.2018.02.003 6 J. Montroy et al. / Transfusion Medicine Reviews xxx (2018) xxx–xxx portion of studies were deemed to have a low risk of reporting bias, as included in these subgroups. The sensitivity analysis using the fixed- not many studies were registered on clinicaltrials.gov or had an avail- effects Peto odds ratio, revealed similar results (Peto OR 0.32, 95% CI able study protocol. The majority of studies were found to have a low 0.27 to 0.39; P b 0.001) (Appendix 3). When the analysis was restricted risk of other potential biases. A breakdown of the risk of bias found in to the 20 trials deemed to have adequate allocation concealment, simi- each study can be found in Appendix 2. lar results were obtained (pooled OR 0.32, 95% CI 0.22 to 0.46; P b 0.001) (Appendix 3). For the subgroup analysis by baseline rate of transfusion, Effect of topical TXA on transfusion rate subgroups included baseline risks of b10%, 10-20%, 20-30%, 30-40% and N40%. With the exception of a baseline risk of b10% (pooled OR 0.47, 95% Compared to placebo CI 0.18 to 1.25; P=0.13), a significant decrease in transfusion risk was Thirty-seven trials involving 3,408 patients reported blood transfu- seen across all subgroups (Appendix 3). sion data comparing topical TXA to placebo. Compared to placebo, the administration of topical TXA significantly reduced the odds of receiving Compared to intravenous administration a blood transfusion (pooled OR 0.29, 95% CI 0.20 to 0.40; P b 0.001, I2= Twenty-four studies representing 2,154 patients reported blood 49.7%) (Fig 3). In subgroup analyses by surgery type, there was a signif- transfusion data comparing topical TXA to IV TXA. When compared to icant reduction in transfusion risk for orthopedic procedures (pooled OR the intravenous administration of TXA, there was no significant differ- 0.22, 95% CI 0.16 to 0.30; Pb0.01). There was no significant difference in ence between the two groups overall (pooled OR 1.03, 95% CI 0.72 to transfusion risk for cardiac procedures (pooled OR 0.94, 95% CI 0.60 to 1.46; P=0.88, I2=4.1%), or for orthopedic procedures (OR 0.92, 95% CI 1.46; P=0.78), orthognathic procedures (pooled OR 1.00, 95% CI 0.02 0.66 to 1.30; P=0.65). However, in the one cardiac surgery trial in- to 52.15; P=N/A), spinal procedures (pooled OR 0.15, 95% CI 0.02 to cluded in the analysis, patients who were administered topical TXA 1.31; P=0.09), and thoracic procedures (pooled OR 0.73, 95% CI 0.26 had a significantly increased odds of receiving a transfusion (OR 11.00, to 2.08; P=0.53) surgeries, although a small number of studies were 95% CI 2.00 to 60.57; P=0.006) (Fig 4). The sensitivity analysis using

Fig 3. Odds ratio (95% CI) and pooled estimates for risk of blood transfusion (topical TXA vs placebo).

Fig 4. Odds ratio (95% CI) and pooled estimates for risk of blood transfusion (topical TXA vs IV TXA). the fixed-effects Peto odds ratio, revealed similar results (Peto OR 1.04, loss between groups (WMD -22, 95% CI -67 to 28; p=0.34) (Fig 6). 95% CI 0.77 to 1.42; P =0.79) (Appendix 3). When the analysis was re- There was substantial statistical heterogeneity between trials (Chi2 = stricted to the 11 trials deemed to have adequate allocation conceal- 222.47, df = 21 (p b0.00001); I2 = 91%). When the analysis was re- ment, similar results were obtained (pooled OR 0.99, 95% CI 0.60 to stricted to the 10 trials deemed to have adequate allocation conceal- 1.61; P=0.96) (Appendix 3). ment, similar results were seen (WMD -9, 95% CI -77 to 58; P=0.78) (Appendix 3). Effect of topical TXA dose on transfusion rate Effect of topical TXA on the rate of venous thromboembolic events Thirty-two of the 36 trials that reported transfusion data comparing topical TXA to placebo were included in a univariate meta-regression Compared to placebo analysis. In four trials, we were unable to convert the dose of topical Thirty trials involving 2,845 patients reported data on the rate of ve- TXA into the appropriate mg/mL due to a lack of information provided nous thromboembolic events (either pulmonary embolism or deep vein in the related publications [38,63,67,68]. The concentration of topical thrombosis) in their topical TXA group and placebo group. Nineteen tri- TXA used in the trials ranged from 1 mg/mL to 100 mg/mL. Dosage of als reported no venous thromboembolic events in either arm of their topical TXA did not influence effectiveness of topical TXA used in the tri- trial. There was no difference in the odds of developing a venous throm- als (Coefficient = -0.002, 95% CI -0.015 to 0.011; P = 0.75). boembolic complication between the topical TXA and control groups (pooled OR=0.78, 95% CI 0.47 to 1.29; P=0.33, I2=0%) (Fig 7). In our Effect of topical TXA on blood loss sensitivity analysis using the fixed-effects Peto odds ratio, similar results were obtained (Peto OR=0.79, 95% CI 0.49 to 1.27; P=0.33, I2=0%) Compared to placebo (Appendix 3). Forty-three trials involving 3,666 patients reported blood loss data comparing topical TXA to placebo. Topical TXA resulted in a mean Compared to intravenous administration blood loss reduction of 276 mL compared to placebo (weighted mean Twenty-one trials involving 1,965 patients reported data on the rate of difference (WMD) -273, 95% CI -324. to -223; P b 0.0001) (Fig 5). venous thromboembolic events (either pulmonary embolism or deep There was substantial statistical heterogeneity between subgroups vein thrombosis) in their topical TXA group and IV groups. Ten of these (Chi2 = 275.6, df = 5 (P b0.00001); I2 = 98.2%).When the analysis trials reported no thromboembolic events in either arm. There was no sig- was restricted to the 22 trials deemed to have adequate allocation con- nificant difference in the odds of developing a venous thromboembolic cealment, similar results were seen (WMD -299, 95% CI -370 to -229; P b complication between the topical TXA and IV TXA groups (pooled OR= 0.00001) (Appendix 3). Large reductions in blood loss were reported in 0.75, 95% CI 0.39 to 1.46; P=0.40, I2=0%) (Fig 8). In our sensitivity anal- orthopedic and cardiac surgery trials, while no statistically significant ysis using the fixed-effectsPetooddsratio,similarresultswereobtained benefit was reported in the single thoracic surgery trial. (Peto OR=0.73, 95% CI 0.40 to 1.31; P=0.29, I2=0%) (Appendix 3).

Compared to intravenous administration Effect of topical TXA on the risk of death Twenty-two trials involving 1,897 patients reported blood loss data comparing topical TXA to IV TXA. When compared to the intravenous When comparing to placebo, nine trials involving 894 patients re- administration of TXA, there was no signiﬁcant difference in blood ported mortality data. Seven of the nine trials reported zero deaths in

Fig 5. Mean difference (95% CI) for estimated blood loss (topical TXA vs placebo).

each group. The administration of topical TXA did not increase the risk Effect of topical TXA on the risk of stroke of death compared to placebo (pooled OR=0.52, 95% CI 0.18 to 1.55; P=0.24, I2=0%) (Fig 9). When comparing topical to intravenous TXA, Seven trials involving 552 patients reported stroke data comparing one study reported mortality data, with one death in the topical group topical TXA to placebo. Six of the seven studies reported zero events in (2.1%), and zero in the IV group. both arms of their trial. The administration of topical TXA did not

Fig 6. Mean difference (95% CI) for estimated blood loss (topical TXA vs IV TXA).

Fig 7. Odds ratio (95% CI) and pooled estimates for the incidence of venous thromboembolic events (topical TXA vs placebo).

Fig 8. Odds ratio (95% CI) and pooled estimates for the incidence of venous thromboembolic events (topical TXA vs IV TXA). increase the risk of stroke when compared to placebo (pooled OR=0.75, comparing topical TXA to placebo, with the plots suggesting the pres- 95% CI 0.18 to 3.18; P=0.70, I2=0%) (Fig 10). Two studies involving 180 ence of small study effects favouring topical TXA. No asymmetry was patients reported stroke data comparing topical to intravenous TXA. seen in the funnel plot for thromboembolic events, suggesting low risk There was one stroke in the intravenous group (1.1%) and zero in the of bias. No asymmetry was seen in the funnel plots for the comparison topical group. of topical and intravenous TXA (Appendix 3).

Effect of topical TXA on the risk of myocardial infarction Discussion

Six trials involving 362 patients reported myocardial infarction data There is considerable interest in topical TXA, given the potential ben- comparing topical TXA to placebo. Four of the six studies reported zero efits and theoretical decreased risk of adverse events compared to sys- events in both arms of their trial. The administration of topical TXA did temic TXA. We performed a systematic review and meta-analysis not increase the risk of myocardial infarction (pooled OR=0.58, 95% CI comparing the safety and efficacy of topically applied TXA to both pla- 0.13 to 2.57; P=0.47) (Fig 11). Three studies involving 368 patients re- cebo and the intravenous application. The results of our review suggest ported myocardial infarction data comparing topical to intravenous that the topical application of tranexamic acid during surgery is highly TXA. There was no significant difference in the odds of developing an effective in reducing both blood loss and the subsequent rate of blood MI between topical and intravenous TXA groups (pooled OR=1.94, transfusions compared to placebo. However, when compared to the in- 95% CI 0.34 to 11.06; P=0.45) (Fig 12). travenous application of TXA, the topical application did not provide any further benefits in terms of blood loss reduction or a reduction in trans- Publication Bias fusion requirements. In addition, we observed no difference in the risk of thromboembolic events, stroke, myocardial infarction, or mortality Publication bias was assessed for clinical outcomes of transfusion compared to either placebo or the intravenous administration. risk, blood loss and thromboembolic events (Appendix 3). Slight asym- Thirty-seven randomized controlled trials involving 3,408 patients metry was seen in the funnel plots for transfusion risk and blood loss in our review reported on transfusion rate compared to placebo, and

Fig 9. Odds ratios (95% CI) and pooled estimates for the incidence of mortality (topical TXA vs placebo).

Fig 10. Odds ratios (95% CI) and pooled estimates for the incidence of stroke (topical TXA vs placebo). we found a 71% reduction in the odds of receiving a blood transfusion in no significant benefit of topical TXA was observed. Only two non- patients who had received TXA topically. In a systematic review and orthopedic studies compared topical to intravenous TXA [80,91].In meta-analysis of 129 randomized controlled trials involving a total of one of these studies [91] no difference in transfusion was observed, 10,488 patients treated with TXA by any form of administration, Ker while the other [80] observed a significantly higher risk of transfusion at al. reported a 30% reduction in the need for transfusion [13], a similar in the topical TXA arm. Outside of orthopedics, it is clear that more evi- magnitude of effect has been noted in reviews of pelvic and orthopedic dence is needed to determine if topical TXA is as effective as IV TXA in surgeries [5,8]. The potential for differences between the two methods reduction of blood loss and transfusion. of application was explored, with 24 studies involving 2,154 patients This review is limited by the quality of the published studies, many reporting transfusion data comparing topical TXA to IV TXA, and we of which had unclear risk of bias due to under reporting of methodolog- saw no difference in the odds of receiving a blood transfusion (pooled ical details. We found substantial statistical heterogeneity in our clinical OR 1.03, 95% CI 0.72 to 1.46; P=0.88). outcomes of blood transfusion and blood loss. This heterogeneity In theory, topical application of TXA may be safer than intravenous remained within the blood loss subgroups and in the sensitivity analy- administration because topically applied TXA results in a 90% reduction ses. There are a number of factors which could explain this heterogene- in plasma concentration compared to intravenous [19,20].Noneofthe ity. Transfusion in clinical practice can vary within healthcare systems trials included in our analysis were adequately powered to detect a dif- and even within hospitals. Considering the wide range of studies, it is ference in thromboembolic events, or other adverse events, between likely that transfusion thresholds varied from study to study. Surgical groups. Moreover, the pooled analysis also suggests there is no signifi- quality across studies is another factor which could influence the mag- cant increased risk of thromboembolic events associated with topical nitude of effect observed. With respect to blood loss, this is a difficult TXA compared to placebo (pooled OR=0.78, 95% CI 0.47 to 1.29; P= outcome to measure accurately and consistently. Differences in measur- 0.33, I2=0%). Despite these reassuring data, it is worth noting the con- ing techniques across studies could explain some of the heterogeneity fidence interval remains wide for safety estimates and only about observed. It is important to note that the variation seen in these out- two-thirds (42/67) of the studies in our review reported on thrombo- comes were in the magnitude of effect, and not the direction. Finally, embolic events. High-quality randomized controlled trials designed few trials reported data on the rate of mortality, stroke or MI, making and powered to detect differences in adverse events are needed to de- the effect estimates for these outcomes very imprecise and preclude definitively establish the safety of TXA, especially in understudied and finitive conclusions. high risk populations such as cancer patients [11]. There is no accepted standard dose of TXA for systemic or topical use Conclusion [92,93]. The studies included in our analysis used a wide range of doses of topical TXA (1mg/mL to 100mg/mL). According to our meta- The topical application of tranexamic acid effectively reduces both regression analysis, the relationship between topical TXA administra- transfusion risk and blood loss compared to placebo. This effect does tion and transfusion risk reduction was not dose-dependent, suggesting not appear to be dose-dependent. Our review suggests the lack of a re- topical TXA remains effective at small doses or low concentrations. Fur- lationship between topical TXA administration and an increased risk of ther study to identify optimal drug doses are needed, in both intrave- adverse events including thromboembolism, although more high- nous and topical applications. quality evidence from trials designed and powered to detect differences Of note, the majority of studies (about two-thirds) evaluated ortho- in safety outcomes are needed before definitive conclusions can be pedic procedures. In subgroup analysis of non-orthopedic procedures, drawn. There were no major differences between topical and

Fig 11. Odds ratios (95% CI) and pooled estimates for the incidence of myocardial infarction (topical TXA vs placebo).

Fig 12. Odds ratios (95% CI) and pooled estimates for the incidence of myocardial infarction (topical TXA vs IV TXA).

intravenous tranexamic acid with respect to safety and efficacy, and [13] Ker K, Edwards P, Perel P, Shakur H, Roberts I. Effect of tranexamic acid on surgical bleeding: systematic review and cumulative meta-analysis. BMJ 2012;344:e3054. both were superior to placebo with regards to blood loss and transfu- [14] Punjani N, Lavallée LT, Momoli F, Fergusson D, Witiuk K, Mallick R, et al. Blood trans- sion requirements. Outside of orthopedic surgery, the comparison of fusion and hemostatic agents used during radical cystectomy. Can Urol Assoc J 2013; topically administered tranexamic acid to intravenously administered 7:E275-0. https://doi.org/10.5489/cuaj.1002. [15] Jawa RS, Singer A, Mccormack JE, Huang EC, Rutigliano DN, Vosswinkel JA. tranexamic acid warrants further exploration. Tranexamic Acid Use in United States Trauma Centers: A National Survey. Am Surg 2016;82:439–47. Funding [16] Truong JL, Cyr DP, Lam-mcculloch J, Cleary SP, Karanicolas PJ. Consensus and Contro- versy in Hepatic Surgery: A Survey of Canadian Surgeons. J Surg Oncol 2014;110: 947–51. https://doi.org/10.1002/jso.23748. This research did not receive any specific grant from funding agen- [17] Ipema HJ, Tanzi MG. Use of Topical Tranexamic Acid or Aminocaproic Acid to Pre- cies in the public, commercial, or not-for-profit sectors. vent Bleeding After Major Surgical Procedures. Ann Pharmacother 2012;46: 97–107. https://doi.org/10.1345/aph.1Q383. [18] Abrishami A, Chung F, Wong J. Topical application of antifibrinolytic drugs for on- Conflict of Interest Statement pump cardiac surgery: a systematic review and meta-analysis. Can J Anaesth 2009;56:202–12. https://doi.org/10.1007/s12630-008-9038-x. fl [19] Abrishami A, Wong J, El-Beheiry H, Hasan SM, Chung F. Intra-articular application of The authors have no con icts of interest to declare. tranexamic acid for perioperative blood loss in total knee arthroplasty: a randomized controlled trial. Can J Anesth 2009;56. Appendix A. Supplementary data [20] McCormack PL. Tranexamic acid: a review of its use in the treatment of hyperfibrinolysis. Drugs 2012;72:585–617. https://doi.org/10.2165/11209070- 000000000-00000. Supplementary data to this article can be found online at https://doi. [21] Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred org/10.1016/j.tmrv.2018.02.003. reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev 2015;4:1. https://doi.org/10.1186/2046-4053-4-1. [22] Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items References for systematic reviews and meta-analyses: The PRISMA statement. Int J Surg 2010;8: 336–41. https://doi.org/10.1016/j.ijsu.2010.02.007. [1] Alter HJ, Klein HG. The hazards of blood transfusion in historical perspective. Blood [23] McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer 2008;112:2617–26. https://doi.org/10.1182/blood-2008-07-077370. Review of Electronic Search Strategies: 2015 Guideline Statement. J Clin Epidemiol [2] Mannucci PM, Levi M. Prevention and Treatment of Major Blood Loss. N Engl J Med 2016;75:40–6. https://doi.org/10.1016/j.jclinepi.2016.01.021. 2007;356:2301–11. https://doi.org/10.1056/NEJMra067742. [24] Higgins, JPT, Altman DG, Sterne, JAC (editors). Chapter 8: Assessing risk of bias in in- [3] Wright JD, Ananth CV, Lewin SN, Burke WM, Siddiq Z, Neugut AI, et al. Patterns of cluded studies. In: Higgins, JPT, Green S (editors). Cochrane Handbook for System- use of hemostatic agents in patients undergoing major surgery. J Surg Res 2014; atic Reviews of Interventions Version 5.1.0 (updated March 2011). The Cochrane 186:458–66. https://doi.org/10.1016/j.jss.2013.07.042. Collaboration. http://handbook.cochrane.org/chapter_8/8_assessing_risk_of_bias_ [4] Bennett C, Klingenberg SL, Langholz E, Gluud LL. Tranexamic acid for upper gastro- in_included_studies.html. intestinal bleeding. In: Gluud LL, editor. Cochrane Database Syst. Rev. Chichester, [25] Gersel-Pedersen N. Tranexamic acid in alveolar sockets in the prevention of alveoli- UK: John Wiley & Sons, Ltd; 2014. p. CD006640. https://doi.org/10.1002/14651858. tis sicca dolorosa. Int J Oral Surg 1979;8:421–9. CD006640.pub3. [26] Sindet-Pedersen S, Ramström G, Bernvil S, Blombäck M. Hemostatic Effect of [5] Breau RH, Kokolo MB, Punjani N, Cagiannos I, Beck A, Niznick N, et al. The effects of Tranexamic Acid Mouthwash in Anticoagulant-Treated Patients Undergoing Oral Sur- lysine analogs during pelvic surgery: a systematic review and meta-analysis. gery. N Engl J Med 1989;320:840–3. https://doi.org/10.1056/NEJM198903303201305. Transfus Med Rev 2014;28:145–55. https://doi.org/10.1016/j.tmrv.2014.05.002. [27] Ramström G, Sindet-Pedersen S, Hall G, Blombäck M, Alander U. Prevention of post- [6] CRASH-2 trial collaboratorsShakur H, Roberts I, Bautista R, Caballero J, Coats T, et al. surgical bleeding in oral surgery using tranexamic acid without dose modification of Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion oral anticoagulants. J Oral Maxillofac Surg 1993;51:1211–6. in trauma patients with significant haemorrhage (CRASH-2): a randomised, [28] Tibbelin A, Aust R, Bende M, Holgersson M, Petruson B, Rundcrantz H, et al. Effect of placebo-controlled trial. Lancet 2010;376:23–32. https://doi.org/10.1016/S0140- local tranexamic acid gel in the treatment of epistaxis. ORL J Otorhinolaryngol Relat 6736(10)60835-5. Spec 1995;57:207–9. [7] Hutton B, Joseph L, Fergusson D, Mazer CD, Shapiro S, Tinmouth A. Risks of harms [29] Blinder D, Manor Y, Martinowitz U, Taicher S, Hashomer T. Dental extractions in pa- using antifibrinolytics in cardiac surgery: systematic review and network meta- tients maintained on continued oral anticoagulant: comparison of local hemostatic analysis of randomised and observational studies. BMJ 2012;345:e5798. https:// modalities. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1999;88:137–40. doi.org/10.1136/bmj.e5798. [30] De Bonis M, Cavaliere F, Alessandrini F, Lapenna E, Santarelli F, Moscato U, et al. Top- [8] Kagoma YK, Crowther MA, Douketis J, Bhandari M, Eikelboom J, Lim W. Use of ical use of tranexamic acid in coronary artery bypass operations: a double-blind, antifibrinolytic therapy to reduce transfusion in patients undergoing orthopedic sur- prospective, randomized, placebo-controlled study. J Thorac Cardiovasc Surg 2000; gery: A systematic review of randomized trials. Thromb Res 2009;123:687–96. 119:575–80. https://doi.org/10.1016/j.thromres.2008.09.015. [31] Krohn CD, Sorensen R, Lange JE, Riise R, Bjornsen S, Brosstad F. Tranexamic acid [9] Li Z-J, Fu X, Xing D, Zhang H-F, Zang J-C, Ma X-L. Is tranexamic acid effective and safe given into the wound reduces postoperative blood loss by half in major orthopaedic in spinal surgery? A meta-analysis of randomized controlled trials. Eur Spine J 2013; surgery. Eur J Surg 2002(Suppl. 588):57–61. 22:1950–7. https://doi.org/10.1007/s00586-013-2774-9. [32] Yasim A, Aşik R, Atahan E. Effects of topical applications of aprotinin and tranexamic [10] White RH, Zhou H, Romano PS. Incidence of symptomatic venous thromboembolism acid on blood loss after open heart surgery. Anadolu Kardiyol Derg 2005;5:36–40. after different elective or urgent surgical procedures. Thromb Haemost 2010;90: [33] Abul-Azm A, Abdullah KM. Effect of topical tranexamic acid in open heart surgery. 446–55. https://doi.org/10.1160/TH03-03-0152. Eur J Anaesthesiol 2006;23:380–4. https://doi.org/10.1017/S0265021505001894. [11] Montroy J, Fergusson NA, Hutton B, Lavallée LT, Morash C, Cagiannos I, et al. The [34] Jabalameli M, Zakeri K. Evaluation of Topical Tranexamic Acid on Intraoperative Safety and Efficacy of Lysine Analogues in Cancer Patients: A Systematic Review Bleeding in Endoscopic Sinus Surgery. IJMS Iran J Med Sci 2006;31. and Meta-Analysis. Transfus Med Rev 2017. https://doi.org/10.1016/j.tmrv.2017. [35] Athanasiadis T, Beule AG, Wormald PJ. Effects of topical antifibrinolytics in endo- 03.002. scopic sinus surgery: A pilot randomized controlled trial. Am J Rhinol 2007;21: [12] Ker K, Beecher D, Roberts I. Topical application of tranexamic acid for the reduction 737–42. https://doi.org/10.2500/ajr.2007.21.3097. of bleeding. In: Ker K, editor. Cochrane Database Syst. Rev. Chichester, UK: John [36] Baric D, Biocina B, Unic D, Sutlic Z, Rudez I, Vrca VB, et al. Topical use of Wiley & Sons, Ltd; 2013. p. CD010562. https://doi.org/10.1002/14651858. antifibrinolytic agents reduces postoperative bleeding: a double-blind, prospective, CD010562.pub2. randomized study. Eur J Cardiothorac Surg 2007;31:366–71.

[37] Fawzy H, Elmistekawy E, Bonneau D, Latter D, Errett L. Can local application of hemostasis in total knee arthroplasty: a multicenter, randomized, controlled trial. Tranexamic acid reduce post-coronary bypass surgery blood loss? A randomized Arch Orthop Trauma Surg 2015;135:1017–25. https://doi.org/10.1007/s00402- controlled trial. J Cardiothorac Surg 2009;4:25. https://doi.org/10.1186/1749-8090- 015-2232-8. 4-25. [61] Ali Shah MU, Asghar MI, Siddiqi R, Chaudhri MS, Janjua AM, Iqbal A. Topical applica- [38] Wong J, Abrishami A, El Beheiry H, Mahomed NN, Roderick Davey J, Gandhi R, et al. tion of tranexamic acid reduces postoperative bleeding in open-heart surgery: myth Topical application of tranexamic acid reduces postoperative blood loss in total knee or fact? J Coll Physicians Surg Pak 2015;25:161–5 [doi:03.2015/JCPSP.161165]. arthroplasty: a randomized, controlled trial. J Bone Joint Surg Am 2010;92:2503–13. [62] Ausen K, Fossmark R, Spigset O, Pleym H. Randomized clinical trial of topical [39] Kaewpradub P, Apipan B, Rummasak D. Does tranexamic acid in an irrigating fluid tranexamic acid after reduction mammoplasty. Br J Surg 2015;102:1348–53. reduce intraoperative blood loss in orthognathic surgery? A double-blind, random- https://doi.org/10.1002/bjs.9878. ized clinical trial. J Oral Maxillofac Surg 2011;69:e186-. https://doi.org/10.1016/j. [63] Digas G, Koutsogiannis I, Meletiadis G, Antonopoulou E, Karamoulas V, Bikos C. Intra- joms.2010.11.041. articular injection of tranexamic acid reduce blood loss in cemented total knee [40] Ishida K, Tsumura N, Kitagawa A, Hamamura S, Fukuda K, Dogaki Y, et al. Intra- arthroplasty. Eur J Orthop Surg Traumatol 2015;25:1181–8. https://doi.org/10. articular injection of tranexamic acid reduces not only blood loss but also knee 1007/s00590-015-1664-8. joint swelling after total knee arthroplasty. Int Orthop 2011;35:1639–45. https:// [64] Eftekharian H, Vahedi R, Karagah T, Tabrizi R. Effect of tranexamic acid irrigation on doi.org/10.1007/s00264-010-1205-3. perioperative blood loss during orthognathic surgery: a double-blind, randomized [41] Kurt T. Evaluation of the effects of systemic or topical use of tranexamic acid and controlled clinical trial. J Oral Maxillofac Surg 2015;73:129–33. https://doi.org/10. aprotinin on the blood loss and the used amount of blood products following cardio- 1016/j.joms.2014.07.033. pulmonary bypass surgery. Anatol J Clin Investig 2011;5:116–21. [65] Gillespie R, Shishani Y, Joseph S, Streit JJ, Gobezie R. Neer Award 2015: A random- [42] Sa-ngasoongsong P, Channoom T, Kawinwonggowit V, Woratanarat P, Chanplakorn ized, prospective evaluation on the effectiveness of tranexamic acid in reducing P, Wibulpolprasert B, et al. Postoperative blood loss reduction in computer-assisted blood loss after total shoulder arthroplasty. J Shoulder Elb Surg 2015;24:1679–84. surgery total knee replacement by low dose intra-articular tranexamic acid injection https://doi.org/10.1016/j.jse.2015.07.029. together with 2-hour clamp drain: a prospective triple-blinded randomized con- [66] Lin S-Y, Chen C-H, Fu Y-C, Huang P-J, Chang J-K, Huang H-T. The Efficacy of Com- trolled trial. Orthop Rev (Pavia) 2011;3:12. https://doi.org/10.4081/or.2011.e12. bined Use of Intraarticular and Intravenous Tranexamic Acid on Reducing Blood [43] Abdullah AJA. Does topical tranexamic acid reduce post-turp hematuria: A double Loss and Transfusion Rate in Total Knee Arthroplasty. J Arthroplast 2015;30: blind randomized control trial. Urology 2012;80:S221–222. 776–80. https://doi.org/10.1016/j.arth.2014.12.001. [44] Canata GL, Casale V, Chiey A. Management of postoperative pain and bleeding in [67] Öztaş S, Öztürk A, Akalin Y, Şahin N, Özkan Y, Otuzbir A, et al. The effect of local and knee replacement. Orthop Traumatol 2012;13:S25–55. systemic application of tranexamic acid on the amount of blood loss and allogeneic [45] Dell’Amore A, Caroli G, Nizar A, Cassanelli N, Luciano G, Greco D, et al. Can topical ap- blood transfusion after total knee replacement. Acta Orthop Belg 2015;81:698–707. plication of tranexamic acid reduce blood loss in thoracic surgery? A prospective [68] Vandesande W, Haenen F. Topical use of tranexamic acid in total hip arthroplasty. A randomised double blind investigation. Heart Lung Circ 2012;21:706–10. https:// prospective randomized controlled trial. Hip Int 2015:S34. doi.org/10.1016/j.hlc.2012.06.016. [69] Wang G, Wang D, Wang B, Lin Y, Sun S. Efficacy and safety evaluation of intra- [46] Roy SP, Tanki UF, Dutta A, Jain SK, Nagi ON. Efficacy of intra-articular tranexamic acid articular injection of tranexamic acid in total knee arthroplasty operation with tem- in blood loss reduction following primary unilateral total knee arthroplasty. Knee porarily drainage close. Int J Clin Exp Med 2015;8:14328–34. Surg Sports Traumatol Arthrosc 2012;20:2494–501. https://doi.org/10.1007/ [70] Wang C, Sun Z, Liu J, Cao J-G, Li Z. Safety and efficacy of intra-articular tranexamic s00167-012-1942-5. acid injection without drainage on blood loss in total knee arthroplasty: A random- [47] Seo J-G, Moon Y-W, Park S-H, Kim S-M, Ko K-R. The comparative efficacies of intra- ized clinical trial. Int J Surg 2015;20:1–7. https://doi.org/10.1016/j.ijsu.2015.05.045. articular and IV tranexamic acid for reducing blood loss during total knee [71] Yang Y, Lv Y-M, Ding P-J, Li J, Ying-Ze Z. The reduction in blood loss with intra- arthroplasty. Knee Surg Sports Traumatol Arthrosc 2013;21:1869–74. https://doi. articular injection of tranexamic acid in unilateral total knee arthroplasty without org/10.1007/s00167-012-2079-2. operative drains: a randomized controlled trial. Eur J Orthop Surg Traumatol 2015; [48] Alshryda S, Mason J, Sarda P, Nargol A, Cooke N, Ahmad H, et al. Topical (intra-artic- 25:135–9. https://doi.org/10.1007/s00590-014-1461-9. ular) tranexamic acid reduces blood loss and transfusion rates following total hip re- [72] Drosos GI, Ververidis A, Valkanis C, Tripsianis G, Stavroulakis E, Vogiatzaki T, et al. A placement: a randomized controlled trial (TRANX-H). J Bone Joint Surg Am 2013;95: randomized comparative study of topical versus intravenous tranexamic acid ad- 1969–74. https://doi.org/10.2106/JBJS.L.00908. ministration in enhanced recovery after surgery (ERAS) total knee replacement. J [49] Alshryda S, Mason J, Vaghela M, Sarda P, Nargol A, Maheswaran S, et al. Topical Orthop 2016;13:127–31. https://doi.org/10.1016/j.jor.2016.03.007. (intra-articular) tranexamic acid reduces blood loss and transfusion rates following [73] Keyhani S, Esmailiejah AA, Abbasian MR, Safdari F. Which Route of Tranexamic Acid total knee replacement: a randomized controlled trial (TRANX-K). J Bone Joint Surg Administration is More Effective to Reduce Blood Loss Following Total Knee Am 2013;95:1961–8. https://doi.org/10.2106/JBJS.L.00907. Arthroplasty? Arch Bone Jt Surg 2016;4:65–9. [50] Georgiadis AG, Muh SJ, Silverton CD, Weir RM, Laker MW. A Prospective Double- [74] Tzatzairis TK, Drosos GI, Kotsios SE, Ververidis AN, Vogiatzaki TD, Kazakos KI. Intra- Blind Placebo Controlled Trial of Topical Tranexamic Acid in Total Knee Arthroplasty. venous vs Topical Tranexamic Acid in Total Knee Arthroplasty Without Tourniquet J Arthroplast 2013;28:78–82. https://doi.org/10.1016/j.arth.2013.03.038. Application: A Randomized Controlled Study. J Arthroplast 2016;31:2465–70. [51] Nouraei M, Baradari AG, Ghafari R, Habibi MR, Zeydi AE, Sharifi N. Decreasing blood https://doi.org/10.1016/j.arth.2016.04.036. loss and the need for transfusion after CABG surgery: a double-blind randomized [75] Zhang Y, Zhang L, Ma X, Jia Y, Wang H, Zhu Y, et al. What is the optimal approach for clinical trial of topical tranexamic acid. Turk J Med Sci 2013;43:273–8. https://doi. tranexamic acid application in patients with unilateral total hip arthroplasty? org/10.3906/sag-1206-37. Orthopade 2016;45:616–21. https://doi.org/10.1007/s00132-016-3252-y. [52] Sa-Ngasoongsong P, Wongsak S, Chanplakorn P, Woratanarat P, Wechmongkolgorn [76] Saberi H, Miri SM, Poordel Namdar M. The effects of topically applied tranexamic S, Wibulpolprasert B, et al. Efficacy of low-dose intra-articular tranexamic acid in acid on reduction of postlaminectomy hemorrhage. Tehran Univ Med J 2010;68: total knee replacement; a prospective triple-blinded randomized controlled trial. 527–33. BMC Musculoskelet Disord 2013;14:340. https://doi.org/10.1186/1471-2474-14- [77] Xu X, Li X, Liu W, Wang Z. Longtime soaking of high concentration tranexamic acid 340. in total hip arthroplasty: A prospective randomized controlled trial in 224 patients. [53] Van Elst C, Vanbiervliet J, Simon JP, Corten K. The effect of topical application of Pakistan J Med Sci 2015;31:1306–11. tranexamic acid in total hip arthroplasty through the direct anterior approach. Am [78] Aggarwal AK, Singh N, Sudesh P. Topical vs Intravenous Tranexamic Acid in Reduc- Acad Orthop Surg Annu Meet 2013. ing Blood Loss After Bilateral Total Knee Arthroplasty: A Prospective Study. J [54] Emara WM, Moez KK, Elkhouly AH. Topical versus intravenous tranexamic acid as a Arthroplast 2016;31:1442–8. https://doi.org/10.1016/j.arth.2015.12.033. blood conservation intervention for reduction of post-operative bleeding in [79] Taheriazam A, Safdari F. Using Tranexamic Acid to decrease postoperative hemor- hemiarthroplasty. Anesth Essays Res 2014;8:48–53. https://doi.org/10.4103/0259- rhage: Local versus Intravenous administration. Int Comb Meet Br Hip Soc 2015. 1162.128908. [80] Baric D, Unic D, Rudez V, Bacic-Vrca M, Planinc D, Jonjic D. Systemic usage of [55] Hosseini H, Rahimianfar AA, Abdollahi MH, Moshtaghiyoon MH, Haddadzadeh M, tranexamic acid is superior to topical: Randomized placebo-controlled trial. Interact Fekri A, et al. Evaluations of topical application of tranexamic acid on post- Cardiovasc Thorac Surg 2011:S92. operative blood loss in off-pump coronary artery bypass surgery. Saudi J Anaesth [81] Gomez-Barrena E, Ortega-Andreu M, Padilla-Eguiluz NG, Pérez-Chrzanowska H, 2014;8:224–8. https://doi.org/10.4103/1658-354X.130724. Figueredo-Zalve R. Topical Intra-Articular Compared with Intravenous Tranexamic [56] Martin JG, Cassatt KB, Kincaid-Cinnamon KA, Westendorf DS, Garton AS, Lemke JH. Acid to Reduce Blood Loss in Primary Total Knee Replacement. J Bone Joint Surg Topical Administration of Tranexamic Acid in Primary Total Hip and Total Knee Am 2014;96:1937–44. https://doi.org/10.2106/JBJS.N.00060. Arthroplasty. J Arthroplast 2014;29:889–94. https://doi.org/10.1016/j.arth.2013.10. [82] May JH, Rieser GR, Williams CG, Merkert RJ, Bauman RD, Lawless MW. The Assess- 005. ment of Blood Loss During Total Knee Arthroplasty When Comparing Intravenous [57] Sarzaeem MM, Razi M, Kazemian G, Moghaddam ME, Rasi AM, Karimi M. Comparing vs Intracapsular Administration of Tranexamic Acid. J Arthroplast 2016;31:2452–7. efficacy of three methods of tranexamic acid administration in reducing hemoglobin [83] Xie J, Ma J, Yue C, Kang P, Pei F. Combined use of intravenous and topical tranexamic drop following total knee arthroplasty. J Arthroplast 2014;29:1521–4. https://doi. acid following cementless total hip arthroplasty: a randomised clinical trial. Hip Int org/10.1016/j.arth.2014.02.031. 2016;26:36–42. https://doi.org/10.5301/hipint.5000291. [58] Wei W, Wei B. Comparison of Topical and Intravenous Tranexamic Acid on Blood [84] Chen JY, Chin PL, Moo IH, Pang HN, Tay DKJ, Chia S-L, et al. Intravenous versus intra- Loss and Transfusion Rates in Total Hip Arthroplasty. J Arthroplast 2014;29: articular tranexamic acid in total knee arthroplasty: A double-blinded randomised 2113–6. https://doi.org/10.1016/j.arth.2014.07.019. controlled noninferiority trial. Knee 2016;23:152–6. https://doi.org/10.1016/j.knee. [59] Yue C, Kang P, Yang P, Xie J, Pei F. Topical Application of Tranexamic Acid in Primary 2015.09.004. Total Hip Arthroplasty: A Randomized Double-Blind Controlled Trial. J Arthroplast [85] Maniar RN, Kumar G, Singhi T, Nayak RM, Maniar PR. Most effective regimen of 2014;29:2452–6. https://doi.org/10.1016/j.arth.2014.03.032. tranexamic acid in knee arthroplasty: a prospective randomized controlled study [60] Aguilera X, Martínez-Zapata MJ, Hinarejos P, Jordán M, Leal J, González JC, et al. Top- in 240 patients. Clin Orthop Relat Res 2012;470:2605–12. https://doi.org/10.1007/ ical and intravenous tranexamic acid reduce blood loss compared to routine s11999-012-2310-y.

[86] Goyal N, Chen DB, Harris IA, Rowden NJ, Kirsh G, MacDessi SJ. Intravenous vs Intra- [90] North WT, Mehran N, Davis JJ, Silverton CD, Weir RM, Laker MW. Topical vs Intrave- Articular Tranexamic Acid in Total Knee Arthroplasty: A Randomized, Double-Blind nous Tranexamic Acid in Primary Total Hip Arthroplasty: A Double-Blind, Random- Trial. J Arthroplast 2017;32:28–32. https://doi.org/10.1016/j.arth.2016.07.004. ized Controlled Trial. J Arthroplast 2016;31:1022–6. https://doi.org/10.1016/j.arth. [87] Patel JN, Spanyer JM, Smith LS, Huang J, Yakkanti MR, Malkani AL. Comparison of Intrave- 2015.11.003. nous versus Topical Tranexamic Acid in Total Knee Arthroplasty: A Prospective Random- [91] Girgin F, Maltepe F. Comparison of Local or Systemic Administration of Tranexamic ized Study. J Arthroplast 2014;29:1528–31. https://doi.org/10.1016/j.arth.2014.03.011. Acid on Blood Loss in Cardiac Surgery. Interact Cardiovasc Thorac Surg 2010. [88] Pinsornsak P, Rojanavijitkul S, Chumchuen S. Peri-articular tranexamic acid injection [92] Ngaage DL, Bland JM. Lessons from aprotinin: is the routine use and inconsistent in total knee arthroplasty: a randomized controlled trial. BMC Musculoskelet Disord dosing of tranexamic acid prudent? Meta-analysis of randomised and large matched 2016;17:313. https://doi.org/10.1186/s12891-016-1176-7. observational studies. Eur J Cardiothorac Surg 2010;37:1375–83. https://doi.org/10. [89] SoniA,SainiR,GulatiA,PaulR,BhattyS,RajoliSR.ComparisonBetweenIntravenousand 1016/j.ejcts.2009.11.055. Intra-articular Regimens of Tranexamic Acid in Reducing Blood Loss During Total Knee [93] Fox MA. Tranexamic Acid: how much is enough? Anesth Analg 2010;111:580–1. Arthroplasty. J Arthroplast 2014;29:1525–7. https://doi.org/10.1016/j.arth.2014.03.039. https://doi.org/10.1213/ANE.0b013e3181e2914d.