- Email: [email protected]
Reduced perioperative blood loss in children undergoing craniosynostosis surgery using prolonged tranexamic acid infusion: a randomised trial
British Journal of Anaesthesia, 122 (6): 760e766 (2019) doi: 10.1016/j.bja.2019.02.017 Advance Access Publication Date: 2 April 2019 Paediatric Anaesthesia
PAEDIATRIC ANAESTHESIA
Reduced perioperative blood loss in children undergoing craniosynostosis surgery using prolonged tranexamic acid infusion: a randomised trial Christian Fenger-Eriksen1,*, Alexander D’Amore Lindholm1, Sven Erik Nørholt2, Gorm von Oettingen3, Mona Tarpgaard4, Lisbeth Krogh1, Niels Juul1, Anne Mette Hvas5,6 and Mads Rasmussen1 1
Department of Anaesthesiology, Aarhus University Hospital, Denmark, 2Department of Oral and Maxillofacial Surgery, Aarhus University Hospital, Denmark, 3Department of Neurosurgery, Aarhus University Hospital, Denmark, 4Department of Anaesthesia, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark, 5Department of Clinical Biochemistry, Centre for Haemophilia and Thrombosis, Aarhus University Hospital, Denmark and 6Department of Clinical Medicine, Aarhus University, Denmark
*Corresponding author. E-mail: [email protected]
Abstract Background: Tranexamic acid (TXA) reduces intraoperative blood loss and transfusion during paediatric craniosynostosis surgery. Additional reduction of postoperative blood loss may further reduce exposure to allogeneic blood products. We studied the effect of combined intra- and postoperative TXA treatment on postoperative blood loss in children. Methods: Thirty children admitted for craniosynostosis surgery were randomised to combined intra- and postoperative TXA treatment or placebo. The primary endpoint was postoperative blood loss. Secondary endpoints included total blood loss, transfusion requirements, and clot stability evaluated by tissue plasminogen activator-stimulated clot lysis assay. Results: TXA reduced postoperative blood loss by 18 ml kg1 (95% confidence interval 8.9) and total blood loss from a mean of 52 ml kg1 (standard deviation [SD]; 20) ml kg1 to 28 (14) ml kg1 (P<0.001). Intraoperative red blood cell (RBC) and fresh frozen plasma (FFP) transfusions were reduced in the treatment group from RBC 14.0 (5.2) ml kg1 to 8.2 (5.1) ml kg1 (P¼0.01) and from FFP 13.0 (6.3) ml kg1 to 7.8 (5.9) ml kg1 (P¼0.03). Postoperative RBC transfusion median was 5 (inter-quartile range [IQR] 0e6) ml kg1 in the placebo group and 0 (0e5.7) ml kg1 in the TXA group. Resistance to lysis was higher in the treatment group (P<0.001). Conclusions: Combined intra- and postoperative tranexamic acid treatment reduced postoperative and overall blood loss and transfusion requirements. Improved clot stability represents a possible mechanism for blood loss reduction. Keywords: antifibrinolytics; blood transfusion; coagulation; craniosynostosis; fibrinolysis; general surgery; haemorrhage; paediatric
Editorial decision date: 03 February 2019; Accepted: 3 February 2019 © 2019 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]
760
Tranexamic acid infusion during craniofacial surgery
Editor’s key points There is no standardisation for tranexamic acid administration to reduce perioperative bleeding in paediatric craniosynostosis surgery. This study investigated whether intra- and postoperative administration of tranexamic acid reduces bleeding and transfusion requirements compared with placebo. Children who received tranexamic acid showed increased clot resistance against fibrinolysis. Despite the small sample size, tranexamic acid administration was associated with reduced postoperative and overall blood loss and resulted in decreased transfusion requirements.
Craniosynostosis (CS) is a condition where one or more cranial sutures fuses prematurely, causing inability of the cranium to continue normal growth. The incidence is estimated to be around one per 2000 children,1 and if left untreated, abnormalities of head shape, increased intracranial pressure, seizures, and developmental delay may occur.2 Surgical correction within the first year of life is frequently required. The surgical procedure is often associated with a high amount of blood loss, both during surgery and in the postoperative period where a significant drainage loss frequently occurs. Total blood loss during the procedure may even exceed the total blood volume of the infant, and therefore normally requires transfusion with allogeneic blood products.3,4 The short- and long-term effects of blood products transfused in children are not known. Although allogeneic transfusions are considered safe, a number of immune- or non-immune-mediated complications such as haemolytic reactions, acute lung injury, coagulopathy, and infectious disease transmission may harm.5 Recently, increased attention has focused on reducing the exposure to allogeneic blood products in both adults and children.6e8 The antifibrinolytic drug tranexamic acid (TXA) competitively decreases the activation of plasminogen to plasmin and suppresses fibrinolysis by inhibiting plasminogen and the binding of plasmin to fibrin. Two prospective, randomised, placebo-controlled trials investigated the effects of TXA during surgical correction of CS. In both trials, transfusion requirements were reduced after treatment with TXA during surgery.9,10 So far, no international guidelines exist on TXA administration for CS surgery. Some centres administer TXA as a single bolus during surgery, although the dose range used is wide and not standardised.11 As blood loss in the postoperative period often reaches significant levels, it may be speculated whether prolonged TXA infusion could also further improve treatment and reduce blood loss in the postoperative phase. The primary aim of this prospective, randomised, double blind, placebo-controlled trial was to evaluate the effects of a protocol combining intra- and postoperative TXA treatment on postoperative and total blood loss in children undergoing craniofacial surgery. We hypothesised that a treatment protocol combining intra- and postoperative TXA would decrease the postoperative and total amount of blood loss during and after CS surgery.
-
761
Patients and methods Study design The study was an investigator-initiated, single-centre, prospective, double-blind, placebo-controlled, randomised clinical trial conducted in accordance with the Note for Guidance on Good Clinical Practice (GCP) (CPMP/ICH/135/95) and the Declaration of Helsinki. The study was monitored by the GCPUnit at Aarhus University Hospital, Denmark and approved by the Danish Medicines Agency (Eudra-CT 2015-000545-22), The Regional Committee on Health Research Ethics (reference 1-10-72-65-15), and the Danish Data Protection Agency (reference 1-16-02-441-15). The manuscript was prepared in accordance with the Consort statement and registered in a public database (https://www.clinicaltrialsregister.eu/ctrsearch/trial/2015-000545-22/DK).
Randomisation and inclusion/exclusion criteria We screened all children (ASA classification score 1/2) undergoing CS surgery at the Department of Neurosurgery, Aarhus University Hospital, Aarhus, Denmark. The exclusion criteria were as follows: any known bleeding disorders, reduced renal function, history with unexplained seizures, and known allergy for TXA. Randomisation was achieved using a closed envelope principle. The children were allocated to receive either TXA (bolus dose of 10 ml kg1 injected before the first surgical incision, followed by 8 h continuous infusion of 3 ml kg1 h1) or an equivalent volume of placebo (isotonic saline, B. Braun, Frederiksberg, Denmark). The choice of TXA dose was based on a pharmacokinetic study in a similar study population.12 The study medicine was prepared and administered by a person not involved in the project or in the anaesthesia procedure; thus, the study staff were blinded to infusion of the drug vehicle. Allocation concealment was sustained until patient recruitment and data collection, transfusion requirement data, and laboratory analyses were completed.
Endpoints The primary endpoint was postoperative blood loss measured in drainage canisters from the end of surgery and during the following 24 h. The drainage canisters were placed during skin closure at the end of surgery. Pre-specified secondary endpoints were: blood loss during surgery, transfusion requirements, haemoglobin concentrations, volume fluid infusion, duration of surgery, and standard coagulation analyses (international normalised ratio [INR], activated partial prothrombin time [APTT], platelet count, and fibrinogen concentrations). Clot stability and fibrinolysis were investigated by the fibrin clot lysis assay, thromboelastometry, and coagulation factor XIII (FXIII) concentrations.
Surgery, anaesthesia, and monitoring Surgery was performed by a team consisting of a neurosurgeon and a maxillofacial surgeon. Anaesthetic management was performed by two dedicated paediatric consultant anaesthetists. The surgical procedures included uni- or bilateral fronto-orbital advancement, total cranial vault reconstruction, and single strip craniotomies. All surgeries were
762
-
Fenger-Eriksen et al.
performed via a bicoronal flap incision and no surgical interventions were conducted endoscopically. General anaesthesia was induced with sevoflurane and maintained with sevoflurane or propofol combined with remifentanil, fentanyl, or both. All children were monitored with ECG, pulse oximetry, invasive BP measurement, and central venous pressure via a catheter placed in the subclavian vein. To minimise postoperative blood loss, a 15 elevation of the head and proper pain treatment were given to all children. All children were successfully extubated immediately after the end of surgery and subsequently transferred to the neurointensive care unit for postoperative care. Although some centres have introduced TXA as standard treatment during CS surgery, this has, until the finalisation of this study, not been the case at our institution.
Fluid therapy, measurements of blood loss, and transfusion strategy Blood loss during surgery was estimated at the end of the procedure based on surgical aspiration and by weighing sponges from the field. Information on postoperative blood loss collected in drainage canisters until 24 h after termination of surgery was retrieved from ICU charts. Products transfused perioperatively consisted of red blood cells (RBC), fresh frozen plasma (FFP), and platelet concentrate. During the operation, no specific transfusion protocol was applied. Instead, an intraoperative decision to administer blood products or further isotonic saline was based on an overall assessment of haemoglobin concentration, current blood loss, central venous pressure, and BP measurements at the discretion of the anaesthetist. The transfusion limit in the postoperative period was haemoglobin concentrations below 5.0 mmol L1. Basic fluid infusion included isotonic saline; 4 ml kg1 for the first 10 kg of body weight þ 2 ml kg1 for every kg >10 kg þ 1 ml for every kg >20 kg. As an addition to the clinical estimation of blood loss, a calculated estimation of blood loss was performed using the following formula, that has previously been used to calculate blood loss after paediatric CS surgery:9
Estimated blood volume loss (ml kg1) (EBV)¼ERCVlost (ml)/ [weight (kg)haematocritpreoperative/100], (1) where ERCVlost is the calculated RBC volume based on preand postoperative haematocrit concentrations. Blood volume was calculated as 75 ml kg1 in children <1 yr and 80 ml kg1 in children >1 yr.
Blood sampling and processing All blood samples were drawn from an already placed arterial line into VenoJect® tubes (Terumo Europe, Leuven, Belgium, containing trisodium citrate 3.2%) for coagulation analyses or ethylenediaminetetraacetic acid-coated tubes (BD, Franklin Lakes, NJ, USA) for haematological parameters. Haematocrit, INR, APTT, platelet count, and fibrinogen values were measured before surgery and the morning of the first postoperative day. Tubes for fibrin clot lysis analysis were centrifuged for 25 min at 4000 rpm, 4 C, and platelet-free plasma was frozen at 80 C.
Standard coagulation tests and haematology APTT, fibrinogen, and INR were assessed on a CS-5100 coagulation analyser (Siemens Healthcare Products, Marburg
Germany) using APTT reagent (TriniCLOT, Tcoag, Bray, Ireland), Dade Thrombin Reagent (Siemens) for determination of fibrinogen concentrations according to the Clauss method, and prothrombin reagent (Prothrombin Owren, Medirox AB, Tystberga, Sweden), respectively.
Clot stability: thromboelastometry and FXIII Whole blood coagulation profiles were recorded in parallel using thromboelastometry (ROTEM® MiniCup system, Instrumentation Laboratory, Munich, Germany) with or without the addition of tissue plasminogen activator (tPA). In brief, pre-warmed (37 C) ROTEM® plastic cups were loaded with citrated whole blood 150 ml and buffer 10 ml (Hepes 20 mM, NaCl 150 mM) or tPA 10 ml (Calbiochem, Merck, Darmstadt, Germany # 612200). The coagulation process was activated by the addition of low dose tissue factor (final concentration one:50 000) and re-calcified by the addition of CaCl2 200 mM, 20 ml. All analyses were processed in duplicate for a minimum of 45 min. FXIII antigen concentrations were measured in citrated plasma using ACL TOP analyser and Hemosil factor XIII antigen reagent (#0020201300, Instrumentation Laboratory, Bedford, MA, USA).
Clot stability: fibrin clot lysis assays Citrated platelet-poor plasma (70 ml) was added in duplicate to € lndal, Sweden) was the wells. Phospholipid 4 mM (Rossix, Mo mixed with HEPES buffer 10 ı`l (20 nm, NaCl 150 mM, pH 7.4, Ampliqon, Odense, Denmark) and recombinant tissue factor (Siemens Healthcare, Marburg, Germany; final dilution one:5000) before addition to the well. Afterwards, tPA 20 ml (final concentration 116 ng ml1) constituted with PBS buffer with bovine serum albumin (Sigma-Aldrich, Schnelldorf, Germany) was added. Finally, HEPES buffer 20 ml in calcium was added to each well. The absorbency was read every minute at 405 nm for 90 min at 37 C, using the Victor microplate reader (Perkin Elmer, Singapore). The following parameters were calculated: a) lysis time (time from maximum clot absorbance to time where a 50% reduction in absorbency has occurred being an indicator of fibrinolytic potential) and b) area under the curve (AUC) (reflects the balance between clot formation and clot lysis).
Statistical analysis Sample size calculation was based on internal data (n¼29) that documented a postoperative blood loss during CS surgery of 30 ml kg1 (standard deviation [SD]: 23 ml kg1). We aimed to reduce postoperative blood loss by 85% to 5 ml kg1 according to the study by Dadure and colleagues.10 Consequently, such a reduction in postoperative blood loss was considered to be clinically relevant. Assuming a two-sided a level of 0.05 and a study power of 0.80, a sample size of 15 patients in each group was required to detect an 85% reduction in postoperative blood loss. In the case of non-normal distribution of data, unequal variances, or both, the non-parametric ManneWhitney U-test rank sum test was used. Haematological and coagulation parameters were evaluated by means of a repeated measurements analysis of variance test and alternatively a KruskaleWallis one way analysis of variance on ranks if a normality test failed with Bonferroni corrections. Data are presented as mean (SD) and median (inter-quartile range [IQR]) where appropriate. SigmaPlot version 11.0 (Systat
Tranexamic acid infusion during craniofacial surgery
-
763
Fig 1. Consolidated standards of reporting trials (CONSORT) diagram showing steps of the patient recruitment and randomisation process.
Software, San Jose, CA, USA) was used for statistical analysis and graph preparation.
Results During the period from September 2015 until January 2018, 39 patients were screened for eligibility; 30 proceeded to randomisation, fulfilled the protocol requirements, and were included in the final patient sample (Fig. 1). Nine patients were ineligible for the study because they failed to meet the inclusion criteria (n¼1), declined participation (n¼6), or for other reasons (n¼2). Hence, the final study group consisted of 30 children (13 girls and 17 boys), ASA classification 1/2. The average age and weight for the entire study group was 1.4 yr (range, 0.35e9.1) and 10.7 kg (range, 6.1e39), with no statistical difference between the groups. The type of CS was equally distributed and the duration of surgery was similar in the two groups (Table 1). No adverse effects (anaphylaxis, seizures, and clinical thromboembolic events) were observed in either group.
Blood loss and transfusion requirements The primary endpoint, postoperative blood loss, was 31 ml kg1 (SD 14 ml kg1) in the placebo group and 13 ml kg1 (SD 7 ml kg1) in the TXA treatment group, with a significant difference (P<0.001). Absolute postoperative bleeding reduction was 18 ml kg1 (95% confidence interval 8.9) favouring TXA treatment. Intraoperative blood loss during surgery was reduced from 20 ml kg1 (IQR 13e28 ml kg1) to 10 ml kg1 (IQR
range 8e14 ml kg1) in the treatment group, without reaching statistical significance (P¼0.06) (Fig. 2). Calculated blood loss was 20 ml kg1 (IQR 13e28 ml kg1) in the placebo group and 11 ml kg1 (IQR 8e14 ml kg1; P¼0.03) in the TXA group (Table 2). Intraoperative RBC transfusion was 14 ml kg1 (SD 5.2 ml kg1) in the placebo group and 8.2 ml kg1 (SD 5.1 ml kg1; P¼0.01) in the TXA group. During the same period, children in the placebo group received FFP 13 ml kg1 (SD 6.3), and children in the TXA group 7.8 ml kg1 (SD 5.9; P¼0.03). In the
Table 1 Patient characteristics and surgery related parameters. Syndrome/other includes Crouzon/Carpenter/Alport syndromes. SD, standard deviation Tranexamic acid
Placebo
n¼15
n¼15
Age, median (range) 8.6 (4.2e110) months Weight, median 9.2 (6.9e39) (range) kg Type of craniosynostosis (n) - Plagiocephaly 3 - Scaphocephaly 4 - Trigonocephaly 4 - Syndrome/other 4 Duration of surgery, 139 (36) mean (SD) min
8.2 (4.9e22.8) 9.7 (6.6e12.5)
2 5 4 4 139 (36)
764
-
Fenger-Eriksen et al.
Clot stability and fibrinolysis FXIII concentrations decreased significantly from 0.72 U ml1 to 0.58 U ml1 on the first day after surgery without any differences between the groups. Preoperative and postoperative maximum clot firmness was equal between the groups, although a trend towards a better preservation of maximum clot firmness (MCF) was noted in the TXA group with 52 mm (SD 5.3) vs 47.5 mm (SD 5.7). Acceleration of fibrinolysis by tPA further reduced MCF at all measurements, without any differences between the groups. An increased potential of fibrin formation was noticed in both groups, as evidenced by increased AUC and prolonged lysis time, on the first postoperative day as compared with preoperative values. Fibrinolysis in the TXA group seemed more resistant as evidenced by a significant prolonged lysis time (P<0.001) and a trend towards improved AUC (1848 [IQR 1469e2130] vs 954 [IQR 653e1236]) as compared with placebo at the first postoperative day (Table 4).
Fig 2. Blood loss (ml kg1) during operation, the following 24 h and total in children receiving placebo or tranexamic acid. Data given as 5th/95th percentile.
Discussion following 24 h, children in the placebo group received RBC transfusion 5 ml kg1 (IQR 0e6 ml kg1) and 0 ml kg1 (IQR 0e5.7 ml kg1) in the TXA group, however, without reaching statistical significance (P¼0.41). No FFP was transfused in the postoperative period and only one child in the TXA group received an intraoperative (4 ml kg1) platelet transfusion (Table 2).
Laboratory parameters Preoperative platelet count, APTT, INR, haemoglobin, and fibrinogen concentrations were all within the reference interval and there were no differences between the two groups (Table 3). Haemoglobin concentrations and platelet count on the first postoperative day were expectedly lower in both groups without reaching statistical significance and still within the normal reference interval. The APTT was significantly shortened in the placebo group on the first postoperative day from 27 s (IQR: 24e29) to 25 s (IQR: 23e26), although without any clinical significance. Fibrinogen concentrations on the first postoperative day increased significantly (P¼0.01) in the TXA group but not in the placebo group.
In this randomised clinical trial, both postoperative and total blood loss was lower among children who received TXA compared with placebo. The present study thus validates two current randomised controlled clinical trials, which also reported reduced blood loss and transfusion requirements in paediatric CS surgery.10,11 Notably, total blood loss in our study population was much lower in the placebo group (median 52 ml kg1) compared with the previous study by Goobie and colleagues9 (mean 119 ml kg1) and Dadure and colleagues10 (mean 66 ml kg1). These very different blood loss rates may be explained by potential differences in patient characteristics, surgical or anaesthetic management, or it may illustrate the difficulties in correct estimation of blood loss in small children accurately. Similarly, RBC transfusion differs substantially between the published trials. Intraoperatively, children receiving TXA received 33 ml kg1,10 1.6 ml kg111 as compared with 14 ml kg1 in our study. To our knowledge, this is the first study to combine laboratory measurements of clot stability and fibrinolysis with the effect from TXA treatment in this study population. The positive effect of TXA on blood loss seems to be driven by a more lysis-resistant clot, as evaluated by two different methodologies, fibrin clot lysis assay and thromboelastometry. Both
Table 2 Transfusion requirements, total fluid infusion, and calculated blood loss for children undergoing craniofacial surgery. N¼15 in each group. *P<0.001 compared with placebo. IQR, inter-quartile range; SD, standard deviation
Red blood cells, mean (SD) ml kg1 - Intraoperative - Postoperative Fresh frozen plasma, mean (SD) ml kg1 - Intraoperative - Postoperative Total fluid infusion - Intraoperative, mean (SD) ml kg1 - Postoperative, median (IQR) ml kg1 Calculated blood loss, median (IQR) ml kg1
P-value
Tranexamic acid
Placebo
n¼15
n¼15
8.2 (5.1) 0 (0e5.7)
14.1 (5.2) 5 (0e6)
0.006* 0.41
7.8 (5.9) 0
13 (6.3) 0
0.028* e
27 (6.4) 49 (42e61) 11 (8e14)
30 (8.3) 47 (33e65) 20 (13e28)
0.34 0.71 0.034*
Tranexamic acid infusion during craniofacial surgery
-
765
Table 3 Haemoglobin concentrations and standard coagulation parameters before and after (1. postoperative day) surgery. *P<0.001 compared with value before operation. APTT, activated partial thromboplastin time; INR, international normalised ratio; IQR, inter-quartile range; SD, standard deiation Preoperative
1
Haemoglobin, mean (SD) mmol l Platelet count, mean (SD) 109 l1 APTT, median (IQR) s INR, mean (SD) s Fibrinogen mean ± SD / median (IQR) mmol L1
24 h Postoperative
Tranexamic acid
Placebo
Tranexamic acid
Placebo
n¼15
n¼15
n¼15
n¼15
6.3 (0.62) 300 (67) 28 (26e32) 1.2 (0.1) 6.1 (4.9e7.6)
6.1 (0.48) 337 (101) 28 (27e29) 1.2 (0.1) 5.6 (4.5e7.4)
5.8 (0.36) 261 (75) 27 (24e29) 1.3 (0.1) 8.3 (1.8)*
5.8 (0.49) 263 (87) 25 (23e26)* 1.3 (0.2) 7.8 (2.6)
assays used tissue factor as the activator and tPA to unmask ongoing fibrinolysis. Using these assays, TXA has previously been shown to significantly reduce facilitated fibrinolysis, although the study by Tang and colleagues13 did not report any clinical data regarding blood loss. Fibrinogen concentrations and ROTEM MCF in the present study were higher on the first postoperative day in the TXA group, and there were no differences in FXIII concentrations. This probably reflects reduced lysis, reduced blood loss, reduced consumption, and therefore preserved fibrinogen reserves. The ROTEM MiniCup methodology used in this study implies a huge advantage of smaller blood sampling volumes. Values are not consistent with conventional ROTEM14 which in the present study was of minor importance as inter-group differences were measured. Moreover, resistance to fibrinolysis was investigated by a recent validated dynamic clot lysis assay providing detailed information about fibrin formation and lysis.15 No established reference interval of this assay exists for children, however the test of difference between the two groups is not affected by this limitation. The TXA dosing regimen for CS surgery has not been defined. In this study, TXA was administered as a loading dose of 10 mg kg1 followed by an of infusion 3 mg kg1 h1 for 8 h, which is lower than reported elsewhere. Although blood loss and transfusion requirements were significantly reduced, it may be speculated if higher doses would decrease blood loss even further. However, higher doses may lead to undesirable side-effects such as seizures and thromboembolic events. In the present study, no adverse or serious adverse events were
observed, although the small sample size did not allow any conclusions on the general safety of the drug in these children. As we compared TXA against placebo, it is difficult to differentiate the effects of the intra- and postoperative TXA treatment on postoperative/total blood loss. Thus, it is possible that intraoperative administration of TXA also reduces postoperative blood loss and that extending the infusion into the postoperative period only has limited influence on total blood loss. In the previous study by Goobie and colleagues,9 intraoperative TXA also significantly reduced postoperative blood loss, while this was not the case in the study by Dadure and colleagues.10 The primary limitation is the small sample size which is, nonetheless, almost similar to the number of patients included in the previous two randomised studies.10,11 However, the small standard errors and the highly significant P-value of the primary outcome parameter may indicate that our study is adequately powered to make conclusions on the use of TXA to reduce blood loss. Further, our study was conducted at a single centre which may limit its generalisability to centres that use different approaches to surgery, anaesthesia, and transfusion of blood products. The data on blood loss and transfusion may be subject to inaccuracy. During surgery, blood loss is difficult to estimate precisely because of the small size of the children and the imprecision of estimating blood loss from canisters, surgical drape, and sponges combined with quite large amounts of washing saline fluid used. Blood loss after surgery was more accurately measured in drains, although some exudation may be present and the amount of
Table 4 Parameters of clot stability. *Significantly different (P<0.05) from preoperative. ySignificantly different (P<0.05) from preoperative. AUC, area under curve; FXIII, coagulation factor XIII; IQR, inter-quartile range; MCF, maximum clot firmness; SD, standard deviation; tPA, tissue plasminogen activator Preoperative
ROTEM Thromboelastometry MCF, mean (SD) þtPA, median (IQR) Fibrin clot lysis assay AUC, median (IQR) Lysis time, median (IQR) FXIII (U ml1), mean (SD)
24 h Postoperative
Tranexamic acid
Placebo
Tranexamic acid
Placebo
49.4 (4.8) 46.5 (41e49)
50.4 (6.6) 47.5 (43e50)
52 (5.3) 51.5 (49e53)
47.5 (5.7) 47 (43e49)
533 (477e622) 679 (633e772) 0.72 (0.1)
484 (384e680) 690 (504e891) 0.71 (0.08)
1848 (1469e2130)* 2070 (1856e2498)* 647* 0.58 (0.1)y
954 (653e1236)* 1070 (776e1531)*y 0.57 (0.12)y
766
-
Fenger-Eriksen et al.
blood loss overestimated. However, the impact of these limitations is minimised because of the randomised study design. Overall, the current study and the available randomised trials collectively suggest that TXA is highly effective in reducing blood loss and transfusion requirements during CS surgery. Further actions may potentially reduce blood loss even more in these children (i.e. implementation of goaldirected coagulation management strategy.)16 In conclusion, combined intra- and postoperative TXA treatment, at a relatively low dose regimen (bolus of 10 mg kg1 followed by infusion of 3 mg kg1 h1 for 8 h) reduces postoperative and overall blood loss and transfusion requirements. TXA improves clot stability as evidenced by increased resistance against fibrinolysis.
Author’s contributions Study conception and design: CFE, MT, NJ, AMH, MR. Patient inclusion and surgery/anaesthesia procedure: MR, GvO, SEN, LK, NJ. Data acquisition and analysis: ADL, AMH, CFE, MR, AL. Manuscript draft: ADL, CFE, AMH, MR. Revising: all authors.
Acknowledgements The authors are grateful for excellent laboratory assistance from Vivi Mogensen and Mai Stenulm Veirup, Department of Clinical Biochemistry, Aarhus University Hospital, Denmark. ROTEM equipment was provided free of charge from Instrumentation Laboratory, TEM International, Germany.
Declaration of interest The authors declare that they have no conflicts of interest.
Funding Novo Nordisk Fonden, Denmark (NNF14OC0011787) and Health Research Fund of Central Denmark Region to MR.
References 1. French LR, Jackson I, Melton LJ. A population-based study of craniosynostosis. J Clin Epidemiol 1990; 43: 69e73 2. Lekovic GP, Bristol RE, Rekate HL. Cognitive impact of craniosynostosis. Semin Pediatr Neurol 2004; 11: 305e10
3. White N, Marcus R, Dover S, et al. Predictors of blood loss in fronto-orbital advancement and remodeling. J Craniofac Surg 2009; 20: 378e81 4. Williams GD, Ellenbogen RG, Gruss JS. Abnormal coagulation during pediatric craniofacial surgery. Pediatr Neurosurg 2001; 35: 5e12 5. Hendrickson JE, Hillyer CD. Noninfectious serious hazards of transfusion. Anesth Analg 2009; 108: 759e69 6. Goobie SM. A blood transfusion can save a child’s life or threaten it. Paediatr Anaesth 2015; 12: 1182e3 bert PC, Hutchison JS, et al. Transfusion 7. Lacroix J, He strategies for patients in pediatric intensive care units. N Engl J Med 2007; 356: 1609e19 8. Vamvakas EC, Blajchman MA. Blood still kills: six strategies to further reduce allogeneic blood transfusion-related mortality. Transfus Med Rev 2010; 24: 77e124 9. Goobie SM, Meier PM, Pereira LM, et al. Efficacy of tranexamic acid in pediatric craniosynostosis surgery. Anesthesiology 2011; 114: 862e71 10. Dadure C, Sauter M, Bringuier S, et al. Intraoperative tranexamic acid reduces blood transfusion in children undergoing craniosynostosis surgery. a randomised doubleblind study. Anesthesiology 2011; 114: 856e61 11. Nishijima DK, Monuteaux MC, Faraoni D, et al. Tranexamic acid use in US children’s hospitals. J Emerg Med 2016; 50: 868e874.e1 12. Goobie SM, Meier PM, Sethna NF, et al. Population pharmacokinetics of tranexamic acid in paediatric patients undergoing craniosynostosis surgery. Clin Pharmacokinet 2013; 52: 267e76 13. Tang M, Wierup P, Rea CJ, Ingerslev J, Hjortdal V, Sørensen B. Temporal changes in clot lysis and clot stability following tranexamic acid in cardiac surgery. Blood Coagul Fibrinolysis 2017; 28: 295e302 14. Haas T, Spielmann N, Dillier C, Cushing M, Siegmund S, Kru¨ger B. Comparison of conventional ROTEM® cups and pins to the ROTEM® cup and pin mini measuring cells (MiniCup). Scand J Clin Lab Invest 2015; 75: 470e5 15. Neergaard-Petersen S, Mogensen VB, Veirup MS, Grove EL, Kristensen SD, Hvas AM. Fibrin clot lysis assay: establishment of a reference interval. Thromb Res 2018; 167: 9e11 16. Innerhofer P, Fries D, Mittermayr M, et al. Reversal of trauma-induced coagulopathy using first-line coagulation factor concentrates or fresh frozen plasma (RETIC): a single-centre, parallel-group, open-label, randomised trial. Lancet Haematol 2017; 4: e258e71 Handling editor: C. Boer
PAEDIATRIC ANAESTHESIA
Reduced perioperative blood loss in children undergoing craniosynostosis surgery using prolonged tranexamic acid infusion: a randomised trial Christian Fenger-Eriksen1,*, Alexander D’Amore Lindholm1, Sven Erik Nørholt2, Gorm von Oettingen3, Mona Tarpgaard4, Lisbeth Krogh1, Niels Juul1, Anne Mette Hvas5,6 and Mads Rasmussen1 1
Department of Anaesthesiology, Aarhus University Hospital, Denmark, 2Department of Oral and Maxillofacial Surgery, Aarhus University Hospital, Denmark, 3Department of Neurosurgery, Aarhus University Hospital, Denmark, 4Department of Anaesthesia, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark, 5Department of Clinical Biochemistry, Centre for Haemophilia and Thrombosis, Aarhus University Hospital, Denmark and 6Department of Clinical Medicine, Aarhus University, Denmark
*Corresponding author. E-mail: [email protected]
Abstract Background: Tranexamic acid (TXA) reduces intraoperative blood loss and transfusion during paediatric craniosynostosis surgery. Additional reduction of postoperative blood loss may further reduce exposure to allogeneic blood products. We studied the effect of combined intra- and postoperative TXA treatment on postoperative blood loss in children. Methods: Thirty children admitted for craniosynostosis surgery were randomised to combined intra- and postoperative TXA treatment or placebo. The primary endpoint was postoperative blood loss. Secondary endpoints included total blood loss, transfusion requirements, and clot stability evaluated by tissue plasminogen activator-stimulated clot lysis assay. Results: TXA reduced postoperative blood loss by 18 ml kg1 (95% confidence interval 8.9) and total blood loss from a mean of 52 ml kg1 (standard deviation [SD]; 20) ml kg1 to 28 (14) ml kg1 (P<0.001). Intraoperative red blood cell (RBC) and fresh frozen plasma (FFP) transfusions were reduced in the treatment group from RBC 14.0 (5.2) ml kg1 to 8.2 (5.1) ml kg1 (P¼0.01) and from FFP 13.0 (6.3) ml kg1 to 7.8 (5.9) ml kg1 (P¼0.03). Postoperative RBC transfusion median was 5 (inter-quartile range [IQR] 0e6) ml kg1 in the placebo group and 0 (0e5.7) ml kg1 in the TXA group. Resistance to lysis was higher in the treatment group (P<0.001). Conclusions: Combined intra- and postoperative tranexamic acid treatment reduced postoperative and overall blood loss and transfusion requirements. Improved clot stability represents a possible mechanism for blood loss reduction. Keywords: antifibrinolytics; blood transfusion; coagulation; craniosynostosis; fibrinolysis; general surgery; haemorrhage; paediatric
Editorial decision date: 03 February 2019; Accepted: 3 February 2019 © 2019 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]
760
Tranexamic acid infusion during craniofacial surgery
Editor’s key points There is no standardisation for tranexamic acid administration to reduce perioperative bleeding in paediatric craniosynostosis surgery. This study investigated whether intra- and postoperative administration of tranexamic acid reduces bleeding and transfusion requirements compared with placebo. Children who received tranexamic acid showed increased clot resistance against fibrinolysis. Despite the small sample size, tranexamic acid administration was associated with reduced postoperative and overall blood loss and resulted in decreased transfusion requirements.
Craniosynostosis (CS) is a condition where one or more cranial sutures fuses prematurely, causing inability of the cranium to continue normal growth. The incidence is estimated to be around one per 2000 children,1 and if left untreated, abnormalities of head shape, increased intracranial pressure, seizures, and developmental delay may occur.2 Surgical correction within the first year of life is frequently required. The surgical procedure is often associated with a high amount of blood loss, both during surgery and in the postoperative period where a significant drainage loss frequently occurs. Total blood loss during the procedure may even exceed the total blood volume of the infant, and therefore normally requires transfusion with allogeneic blood products.3,4 The short- and long-term effects of blood products transfused in children are not known. Although allogeneic transfusions are considered safe, a number of immune- or non-immune-mediated complications such as haemolytic reactions, acute lung injury, coagulopathy, and infectious disease transmission may harm.5 Recently, increased attention has focused on reducing the exposure to allogeneic blood products in both adults and children.6e8 The antifibrinolytic drug tranexamic acid (TXA) competitively decreases the activation of plasminogen to plasmin and suppresses fibrinolysis by inhibiting plasminogen and the binding of plasmin to fibrin. Two prospective, randomised, placebo-controlled trials investigated the effects of TXA during surgical correction of CS. In both trials, transfusion requirements were reduced after treatment with TXA during surgery.9,10 So far, no international guidelines exist on TXA administration for CS surgery. Some centres administer TXA as a single bolus during surgery, although the dose range used is wide and not standardised.11 As blood loss in the postoperative period often reaches significant levels, it may be speculated whether prolonged TXA infusion could also further improve treatment and reduce blood loss in the postoperative phase. The primary aim of this prospective, randomised, double blind, placebo-controlled trial was to evaluate the effects of a protocol combining intra- and postoperative TXA treatment on postoperative and total blood loss in children undergoing craniofacial surgery. We hypothesised that a treatment protocol combining intra- and postoperative TXA would decrease the postoperative and total amount of blood loss during and after CS surgery.
-
761
Patients and methods Study design The study was an investigator-initiated, single-centre, prospective, double-blind, placebo-controlled, randomised clinical trial conducted in accordance with the Note for Guidance on Good Clinical Practice (GCP) (CPMP/ICH/135/95) and the Declaration of Helsinki. The study was monitored by the GCPUnit at Aarhus University Hospital, Denmark and approved by the Danish Medicines Agency (Eudra-CT 2015-000545-22), The Regional Committee on Health Research Ethics (reference 1-10-72-65-15), and the Danish Data Protection Agency (reference 1-16-02-441-15). The manuscript was prepared in accordance with the Consort statement and registered in a public database (https://www.clinicaltrialsregister.eu/ctrsearch/trial/2015-000545-22/DK).
Randomisation and inclusion/exclusion criteria We screened all children (ASA classification score 1/2) undergoing CS surgery at the Department of Neurosurgery, Aarhus University Hospital, Aarhus, Denmark. The exclusion criteria were as follows: any known bleeding disorders, reduced renal function, history with unexplained seizures, and known allergy for TXA. Randomisation was achieved using a closed envelope principle. The children were allocated to receive either TXA (bolus dose of 10 ml kg1 injected before the first surgical incision, followed by 8 h continuous infusion of 3 ml kg1 h1) or an equivalent volume of placebo (isotonic saline, B. Braun, Frederiksberg, Denmark). The choice of TXA dose was based on a pharmacokinetic study in a similar study population.12 The study medicine was prepared and administered by a person not involved in the project or in the anaesthesia procedure; thus, the study staff were blinded to infusion of the drug vehicle. Allocation concealment was sustained until patient recruitment and data collection, transfusion requirement data, and laboratory analyses were completed.
Endpoints The primary endpoint was postoperative blood loss measured in drainage canisters from the end of surgery and during the following 24 h. The drainage canisters were placed during skin closure at the end of surgery. Pre-specified secondary endpoints were: blood loss during surgery, transfusion requirements, haemoglobin concentrations, volume fluid infusion, duration of surgery, and standard coagulation analyses (international normalised ratio [INR], activated partial prothrombin time [APTT], platelet count, and fibrinogen concentrations). Clot stability and fibrinolysis were investigated by the fibrin clot lysis assay, thromboelastometry, and coagulation factor XIII (FXIII) concentrations.
Surgery, anaesthesia, and monitoring Surgery was performed by a team consisting of a neurosurgeon and a maxillofacial surgeon. Anaesthetic management was performed by two dedicated paediatric consultant anaesthetists. The surgical procedures included uni- or bilateral fronto-orbital advancement, total cranial vault reconstruction, and single strip craniotomies. All surgeries were
762
-
Fenger-Eriksen et al.
performed via a bicoronal flap incision and no surgical interventions were conducted endoscopically. General anaesthesia was induced with sevoflurane and maintained with sevoflurane or propofol combined with remifentanil, fentanyl, or both. All children were monitored with ECG, pulse oximetry, invasive BP measurement, and central venous pressure via a catheter placed in the subclavian vein. To minimise postoperative blood loss, a 15 elevation of the head and proper pain treatment were given to all children. All children were successfully extubated immediately after the end of surgery and subsequently transferred to the neurointensive care unit for postoperative care. Although some centres have introduced TXA as standard treatment during CS surgery, this has, until the finalisation of this study, not been the case at our institution.
Fluid therapy, measurements of blood loss, and transfusion strategy Blood loss during surgery was estimated at the end of the procedure based on surgical aspiration and by weighing sponges from the field. Information on postoperative blood loss collected in drainage canisters until 24 h after termination of surgery was retrieved from ICU charts. Products transfused perioperatively consisted of red blood cells (RBC), fresh frozen plasma (FFP), and platelet concentrate. During the operation, no specific transfusion protocol was applied. Instead, an intraoperative decision to administer blood products or further isotonic saline was based on an overall assessment of haemoglobin concentration, current blood loss, central venous pressure, and BP measurements at the discretion of the anaesthetist. The transfusion limit in the postoperative period was haemoglobin concentrations below 5.0 mmol L1. Basic fluid infusion included isotonic saline; 4 ml kg1 for the first 10 kg of body weight þ 2 ml kg1 for every kg >10 kg þ 1 ml for every kg >20 kg. As an addition to the clinical estimation of blood loss, a calculated estimation of blood loss was performed using the following formula, that has previously been used to calculate blood loss after paediatric CS surgery:9
Estimated blood volume loss (ml kg1) (EBV)¼ERCVlost (ml)/ [weight (kg)haematocritpreoperative/100], (1) where ERCVlost is the calculated RBC volume based on preand postoperative haematocrit concentrations. Blood volume was calculated as 75 ml kg1 in children <1 yr and 80 ml kg1 in children >1 yr.
Blood sampling and processing All blood samples were drawn from an already placed arterial line into VenoJect® tubes (Terumo Europe, Leuven, Belgium, containing trisodium citrate 3.2%) for coagulation analyses or ethylenediaminetetraacetic acid-coated tubes (BD, Franklin Lakes, NJ, USA) for haematological parameters. Haematocrit, INR, APTT, platelet count, and fibrinogen values were measured before surgery and the morning of the first postoperative day. Tubes for fibrin clot lysis analysis were centrifuged for 25 min at 4000 rpm, 4 C, and platelet-free plasma was frozen at 80 C.
Standard coagulation tests and haematology APTT, fibrinogen, and INR were assessed on a CS-5100 coagulation analyser (Siemens Healthcare Products, Marburg
Germany) using APTT reagent (TriniCLOT, Tcoag, Bray, Ireland), Dade Thrombin Reagent (Siemens) for determination of fibrinogen concentrations according to the Clauss method, and prothrombin reagent (Prothrombin Owren, Medirox AB, Tystberga, Sweden), respectively.
Clot stability: thromboelastometry and FXIII Whole blood coagulation profiles were recorded in parallel using thromboelastometry (ROTEM® MiniCup system, Instrumentation Laboratory, Munich, Germany) with or without the addition of tissue plasminogen activator (tPA). In brief, pre-warmed (37 C) ROTEM® plastic cups were loaded with citrated whole blood 150 ml and buffer 10 ml (Hepes 20 mM, NaCl 150 mM) or tPA 10 ml (Calbiochem, Merck, Darmstadt, Germany # 612200). The coagulation process was activated by the addition of low dose tissue factor (final concentration one:50 000) and re-calcified by the addition of CaCl2 200 mM, 20 ml. All analyses were processed in duplicate for a minimum of 45 min. FXIII antigen concentrations were measured in citrated plasma using ACL TOP analyser and Hemosil factor XIII antigen reagent (#0020201300, Instrumentation Laboratory, Bedford, MA, USA).
Clot stability: fibrin clot lysis assays Citrated platelet-poor plasma (70 ml) was added in duplicate to € lndal, Sweden) was the wells. Phospholipid 4 mM (Rossix, Mo mixed with HEPES buffer 10 ı`l (20 nm, NaCl 150 mM, pH 7.4, Ampliqon, Odense, Denmark) and recombinant tissue factor (Siemens Healthcare, Marburg, Germany; final dilution one:5000) before addition to the well. Afterwards, tPA 20 ml (final concentration 116 ng ml1) constituted with PBS buffer with bovine serum albumin (Sigma-Aldrich, Schnelldorf, Germany) was added. Finally, HEPES buffer 20 ml in calcium was added to each well. The absorbency was read every minute at 405 nm for 90 min at 37 C, using the Victor microplate reader (Perkin Elmer, Singapore). The following parameters were calculated: a) lysis time (time from maximum clot absorbance to time where a 50% reduction in absorbency has occurred being an indicator of fibrinolytic potential) and b) area under the curve (AUC) (reflects the balance between clot formation and clot lysis).
Statistical analysis Sample size calculation was based on internal data (n¼29) that documented a postoperative blood loss during CS surgery of 30 ml kg1 (standard deviation [SD]: 23 ml kg1). We aimed to reduce postoperative blood loss by 85% to 5 ml kg1 according to the study by Dadure and colleagues.10 Consequently, such a reduction in postoperative blood loss was considered to be clinically relevant. Assuming a two-sided a level of 0.05 and a study power of 0.80, a sample size of 15 patients in each group was required to detect an 85% reduction in postoperative blood loss. In the case of non-normal distribution of data, unequal variances, or both, the non-parametric ManneWhitney U-test rank sum test was used. Haematological and coagulation parameters were evaluated by means of a repeated measurements analysis of variance test and alternatively a KruskaleWallis one way analysis of variance on ranks if a normality test failed with Bonferroni corrections. Data are presented as mean (SD) and median (inter-quartile range [IQR]) where appropriate. SigmaPlot version 11.0 (Systat
Tranexamic acid infusion during craniofacial surgery
-
763
Fig 1. Consolidated standards of reporting trials (CONSORT) diagram showing steps of the patient recruitment and randomisation process.
Software, San Jose, CA, USA) was used for statistical analysis and graph preparation.
Results During the period from September 2015 until January 2018, 39 patients were screened for eligibility; 30 proceeded to randomisation, fulfilled the protocol requirements, and were included in the final patient sample (Fig. 1). Nine patients were ineligible for the study because they failed to meet the inclusion criteria (n¼1), declined participation (n¼6), or for other reasons (n¼2). Hence, the final study group consisted of 30 children (13 girls and 17 boys), ASA classification 1/2. The average age and weight for the entire study group was 1.4 yr (range, 0.35e9.1) and 10.7 kg (range, 6.1e39), with no statistical difference between the groups. The type of CS was equally distributed and the duration of surgery was similar in the two groups (Table 1). No adverse effects (anaphylaxis, seizures, and clinical thromboembolic events) were observed in either group.
Blood loss and transfusion requirements The primary endpoint, postoperative blood loss, was 31 ml kg1 (SD 14 ml kg1) in the placebo group and 13 ml kg1 (SD 7 ml kg1) in the TXA treatment group, with a significant difference (P<0.001). Absolute postoperative bleeding reduction was 18 ml kg1 (95% confidence interval 8.9) favouring TXA treatment. Intraoperative blood loss during surgery was reduced from 20 ml kg1 (IQR 13e28 ml kg1) to 10 ml kg1 (IQR
range 8e14 ml kg1) in the treatment group, without reaching statistical significance (P¼0.06) (Fig. 2). Calculated blood loss was 20 ml kg1 (IQR 13e28 ml kg1) in the placebo group and 11 ml kg1 (IQR 8e14 ml kg1; P¼0.03) in the TXA group (Table 2). Intraoperative RBC transfusion was 14 ml kg1 (SD 5.2 ml kg1) in the placebo group and 8.2 ml kg1 (SD 5.1 ml kg1; P¼0.01) in the TXA group. During the same period, children in the placebo group received FFP 13 ml kg1 (SD 6.3), and children in the TXA group 7.8 ml kg1 (SD 5.9; P¼0.03). In the
Table 1 Patient characteristics and surgery related parameters. Syndrome/other includes Crouzon/Carpenter/Alport syndromes. SD, standard deviation Tranexamic acid
Placebo
n¼15
n¼15
Age, median (range) 8.6 (4.2e110) months Weight, median 9.2 (6.9e39) (range) kg Type of craniosynostosis (n) - Plagiocephaly 3 - Scaphocephaly 4 - Trigonocephaly 4 - Syndrome/other 4 Duration of surgery, 139 (36) mean (SD) min
8.2 (4.9e22.8) 9.7 (6.6e12.5)
2 5 4 4 139 (36)
764
-
Fenger-Eriksen et al.
Clot stability and fibrinolysis FXIII concentrations decreased significantly from 0.72 U ml1 to 0.58 U ml1 on the first day after surgery without any differences between the groups. Preoperative and postoperative maximum clot firmness was equal between the groups, although a trend towards a better preservation of maximum clot firmness (MCF) was noted in the TXA group with 52 mm (SD 5.3) vs 47.5 mm (SD 5.7). Acceleration of fibrinolysis by tPA further reduced MCF at all measurements, without any differences between the groups. An increased potential of fibrin formation was noticed in both groups, as evidenced by increased AUC and prolonged lysis time, on the first postoperative day as compared with preoperative values. Fibrinolysis in the TXA group seemed more resistant as evidenced by a significant prolonged lysis time (P<0.001) and a trend towards improved AUC (1848 [IQR 1469e2130] vs 954 [IQR 653e1236]) as compared with placebo at the first postoperative day (Table 4).
Fig 2. Blood loss (ml kg1) during operation, the following 24 h and total in children receiving placebo or tranexamic acid. Data given as 5th/95th percentile.
Discussion following 24 h, children in the placebo group received RBC transfusion 5 ml kg1 (IQR 0e6 ml kg1) and 0 ml kg1 (IQR 0e5.7 ml kg1) in the TXA group, however, without reaching statistical significance (P¼0.41). No FFP was transfused in the postoperative period and only one child in the TXA group received an intraoperative (4 ml kg1) platelet transfusion (Table 2).
Laboratory parameters Preoperative platelet count, APTT, INR, haemoglobin, and fibrinogen concentrations were all within the reference interval and there were no differences between the two groups (Table 3). Haemoglobin concentrations and platelet count on the first postoperative day were expectedly lower in both groups without reaching statistical significance and still within the normal reference interval. The APTT was significantly shortened in the placebo group on the first postoperative day from 27 s (IQR: 24e29) to 25 s (IQR: 23e26), although without any clinical significance. Fibrinogen concentrations on the first postoperative day increased significantly (P¼0.01) in the TXA group but not in the placebo group.
In this randomised clinical trial, both postoperative and total blood loss was lower among children who received TXA compared with placebo. The present study thus validates two current randomised controlled clinical trials, which also reported reduced blood loss and transfusion requirements in paediatric CS surgery.10,11 Notably, total blood loss in our study population was much lower in the placebo group (median 52 ml kg1) compared with the previous study by Goobie and colleagues9 (mean 119 ml kg1) and Dadure and colleagues10 (mean 66 ml kg1). These very different blood loss rates may be explained by potential differences in patient characteristics, surgical or anaesthetic management, or it may illustrate the difficulties in correct estimation of blood loss in small children accurately. Similarly, RBC transfusion differs substantially between the published trials. Intraoperatively, children receiving TXA received 33 ml kg1,10 1.6 ml kg111 as compared with 14 ml kg1 in our study. To our knowledge, this is the first study to combine laboratory measurements of clot stability and fibrinolysis with the effect from TXA treatment in this study population. The positive effect of TXA on blood loss seems to be driven by a more lysis-resistant clot, as evaluated by two different methodologies, fibrin clot lysis assay and thromboelastometry. Both
Table 2 Transfusion requirements, total fluid infusion, and calculated blood loss for children undergoing craniofacial surgery. N¼15 in each group. *P<0.001 compared with placebo. IQR, inter-quartile range; SD, standard deviation
Red blood cells, mean (SD) ml kg1 - Intraoperative - Postoperative Fresh frozen plasma, mean (SD) ml kg1 - Intraoperative - Postoperative Total fluid infusion - Intraoperative, mean (SD) ml kg1 - Postoperative, median (IQR) ml kg1 Calculated blood loss, median (IQR) ml kg1
P-value
Tranexamic acid
Placebo
n¼15
n¼15
8.2 (5.1) 0 (0e5.7)
14.1 (5.2) 5 (0e6)
0.006* 0.41
7.8 (5.9) 0
13 (6.3) 0
0.028* e
27 (6.4) 49 (42e61) 11 (8e14)
30 (8.3) 47 (33e65) 20 (13e28)
0.34 0.71 0.034*
Tranexamic acid infusion during craniofacial surgery
-
765
Table 3 Haemoglobin concentrations and standard coagulation parameters before and after (1. postoperative day) surgery. *P<0.001 compared with value before operation. APTT, activated partial thromboplastin time; INR, international normalised ratio; IQR, inter-quartile range; SD, standard deiation Preoperative
1
Haemoglobin, mean (SD) mmol l Platelet count, mean (SD) 109 l1 APTT, median (IQR) s INR, mean (SD) s Fibrinogen mean ± SD / median (IQR) mmol L1
24 h Postoperative
Tranexamic acid
Placebo
Tranexamic acid
Placebo
n¼15
n¼15
n¼15
n¼15
6.3 (0.62) 300 (67) 28 (26e32) 1.2 (0.1) 6.1 (4.9e7.6)
6.1 (0.48) 337 (101) 28 (27e29) 1.2 (0.1) 5.6 (4.5e7.4)
5.8 (0.36) 261 (75) 27 (24e29) 1.3 (0.1) 8.3 (1.8)*
5.8 (0.49) 263 (87) 25 (23e26)* 1.3 (0.2) 7.8 (2.6)
assays used tissue factor as the activator and tPA to unmask ongoing fibrinolysis. Using these assays, TXA has previously been shown to significantly reduce facilitated fibrinolysis, although the study by Tang and colleagues13 did not report any clinical data regarding blood loss. Fibrinogen concentrations and ROTEM MCF in the present study were higher on the first postoperative day in the TXA group, and there were no differences in FXIII concentrations. This probably reflects reduced lysis, reduced blood loss, reduced consumption, and therefore preserved fibrinogen reserves. The ROTEM MiniCup methodology used in this study implies a huge advantage of smaller blood sampling volumes. Values are not consistent with conventional ROTEM14 which in the present study was of minor importance as inter-group differences were measured. Moreover, resistance to fibrinolysis was investigated by a recent validated dynamic clot lysis assay providing detailed information about fibrin formation and lysis.15 No established reference interval of this assay exists for children, however the test of difference between the two groups is not affected by this limitation. The TXA dosing regimen for CS surgery has not been defined. In this study, TXA was administered as a loading dose of 10 mg kg1 followed by an of infusion 3 mg kg1 h1 for 8 h, which is lower than reported elsewhere. Although blood loss and transfusion requirements were significantly reduced, it may be speculated if higher doses would decrease blood loss even further. However, higher doses may lead to undesirable side-effects such as seizures and thromboembolic events. In the present study, no adverse or serious adverse events were
observed, although the small sample size did not allow any conclusions on the general safety of the drug in these children. As we compared TXA against placebo, it is difficult to differentiate the effects of the intra- and postoperative TXA treatment on postoperative/total blood loss. Thus, it is possible that intraoperative administration of TXA also reduces postoperative blood loss and that extending the infusion into the postoperative period only has limited influence on total blood loss. In the previous study by Goobie and colleagues,9 intraoperative TXA also significantly reduced postoperative blood loss, while this was not the case in the study by Dadure and colleagues.10 The primary limitation is the small sample size which is, nonetheless, almost similar to the number of patients included in the previous two randomised studies.10,11 However, the small standard errors and the highly significant P-value of the primary outcome parameter may indicate that our study is adequately powered to make conclusions on the use of TXA to reduce blood loss. Further, our study was conducted at a single centre which may limit its generalisability to centres that use different approaches to surgery, anaesthesia, and transfusion of blood products. The data on blood loss and transfusion may be subject to inaccuracy. During surgery, blood loss is difficult to estimate precisely because of the small size of the children and the imprecision of estimating blood loss from canisters, surgical drape, and sponges combined with quite large amounts of washing saline fluid used. Blood loss after surgery was more accurately measured in drains, although some exudation may be present and the amount of
Table 4 Parameters of clot stability. *Significantly different (P<0.05) from preoperative. ySignificantly different (P<0.05) from preoperative. AUC, area under curve; FXIII, coagulation factor XIII; IQR, inter-quartile range; MCF, maximum clot firmness; SD, standard deviation; tPA, tissue plasminogen activator Preoperative
ROTEM Thromboelastometry MCF, mean (SD) þtPA, median (IQR) Fibrin clot lysis assay AUC, median (IQR) Lysis time, median (IQR) FXIII (U ml1), mean (SD)
24 h Postoperative
Tranexamic acid
Placebo
Tranexamic acid
Placebo
49.4 (4.8) 46.5 (41e49)
50.4 (6.6) 47.5 (43e50)
52 (5.3) 51.5 (49e53)
47.5 (5.7) 47 (43e49)
533 (477e622) 679 (633e772) 0.72 (0.1)
484 (384e680) 690 (504e891) 0.71 (0.08)
1848 (1469e2130)* 2070 (1856e2498)* 647* 0.58 (0.1)y
954 (653e1236)* 1070 (776e1531)*y 0.57 (0.12)y
766
-
Fenger-Eriksen et al.
blood loss overestimated. However, the impact of these limitations is minimised because of the randomised study design. Overall, the current study and the available randomised trials collectively suggest that TXA is highly effective in reducing blood loss and transfusion requirements during CS surgery. Further actions may potentially reduce blood loss even more in these children (i.e. implementation of goaldirected coagulation management strategy.)16 In conclusion, combined intra- and postoperative TXA treatment, at a relatively low dose regimen (bolus of 10 mg kg1 followed by infusion of 3 mg kg1 h1 for 8 h) reduces postoperative and overall blood loss and transfusion requirements. TXA improves clot stability as evidenced by increased resistance against fibrinolysis.
Author’s contributions Study conception and design: CFE, MT, NJ, AMH, MR. Patient inclusion and surgery/anaesthesia procedure: MR, GvO, SEN, LK, NJ. Data acquisition and analysis: ADL, AMH, CFE, MR, AL. Manuscript draft: ADL, CFE, AMH, MR. Revising: all authors.
Acknowledgements The authors are grateful for excellent laboratory assistance from Vivi Mogensen and Mai Stenulm Veirup, Department of Clinical Biochemistry, Aarhus University Hospital, Denmark. ROTEM equipment was provided free of charge from Instrumentation Laboratory, TEM International, Germany.
Declaration of interest The authors declare that they have no conflicts of interest.
Funding Novo Nordisk Fonden, Denmark (NNF14OC0011787) and Health Research Fund of Central Denmark Region to MR.
References 1. French LR, Jackson I, Melton LJ. A population-based study of craniosynostosis. J Clin Epidemiol 1990; 43: 69e73 2. Lekovic GP, Bristol RE, Rekate HL. Cognitive impact of craniosynostosis. Semin Pediatr Neurol 2004; 11: 305e10
3. White N, Marcus R, Dover S, et al. Predictors of blood loss in fronto-orbital advancement and remodeling. J Craniofac Surg 2009; 20: 378e81 4. Williams GD, Ellenbogen RG, Gruss JS. Abnormal coagulation during pediatric craniofacial surgery. Pediatr Neurosurg 2001; 35: 5e12 5. Hendrickson JE, Hillyer CD. Noninfectious serious hazards of transfusion. Anesth Analg 2009; 108: 759e69 6. Goobie SM. A blood transfusion can save a child’s life or threaten it. Paediatr Anaesth 2015; 12: 1182e3 bert PC, Hutchison JS, et al. Transfusion 7. Lacroix J, He strategies for patients in pediatric intensive care units. N Engl J Med 2007; 356: 1609e19 8. Vamvakas EC, Blajchman MA. Blood still kills: six strategies to further reduce allogeneic blood transfusion-related mortality. Transfus Med Rev 2010; 24: 77e124 9. Goobie SM, Meier PM, Pereira LM, et al. Efficacy of tranexamic acid in pediatric craniosynostosis surgery. Anesthesiology 2011; 114: 862e71 10. Dadure C, Sauter M, Bringuier S, et al. Intraoperative tranexamic acid reduces blood transfusion in children undergoing craniosynostosis surgery. a randomised doubleblind study. Anesthesiology 2011; 114: 856e61 11. Nishijima DK, Monuteaux MC, Faraoni D, et al. Tranexamic acid use in US children’s hospitals. J Emerg Med 2016; 50: 868e874.e1 12. Goobie SM, Meier PM, Sethna NF, et al. Population pharmacokinetics of tranexamic acid in paediatric patients undergoing craniosynostosis surgery. Clin Pharmacokinet 2013; 52: 267e76 13. Tang M, Wierup P, Rea CJ, Ingerslev J, Hjortdal V, Sørensen B. Temporal changes in clot lysis and clot stability following tranexamic acid in cardiac surgery. Blood Coagul Fibrinolysis 2017; 28: 295e302 14. Haas T, Spielmann N, Dillier C, Cushing M, Siegmund S, Kru¨ger B. Comparison of conventional ROTEM® cups and pins to the ROTEM® cup and pin mini measuring cells (MiniCup). Scand J Clin Lab Invest 2015; 75: 470e5 15. Neergaard-Petersen S, Mogensen VB, Veirup MS, Grove EL, Kristensen SD, Hvas AM. Fibrin clot lysis assay: establishment of a reference interval. Thromb Res 2018; 167: 9e11 16. Innerhofer P, Fries D, Mittermayr M, et al. Reversal of trauma-induced coagulopathy using first-line coagulation factor concentrates or fresh frozen plasma (RETIC): a single-centre, parallel-group, open-label, randomised trial. Lancet Haematol 2017; 4: e258e71 Handling editor: C. Boer