- Email: [email protected]
Relationship Between Transfusion of Blood Products and the Incidence of Thrombotic Complications in Neonates and Infants Undergoing Cardiac Surgery
Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
Contents lists available at ScienceDirect
HOSTED BY
journal homepage: www.jcvaonline.com
Original Article
Relationship Between Transfusion of Blood Products and the Incidence of Thrombotic Complications in Neonates and Infants Undergoing Cardiac Surgery David Faraoni, MD, PhD, FCCPn,1, Sirisha Emani, PhD†, Erin Halpin, MSN, RN‡, Rachel Bernier, MPH‡, Sitaram Emani, MD†, James A. DiNardo, MD, FAAP‡, Juan C. Ibla, MD‡ *
Department of Anesthesia and Pain Medicine, The Hospital for Sick Children, University of Toronto, Toronto, Canada † Department of Cardiovascular Surgery, Boston Children’s Hospital, Harvard Medical School, Boston, MA ‡ Department of Anesthesiology, Peri-operative and Pain Medicine, Boston Children’s Hospital, Harvard Medical School, Boston, MA
Objectives: The authors hypothesized that transfusion of blood products in neonates and infants undergoing high-risk cardiac surgery in the absence of intraoperative coagulation monitoring increases the risk of thrombotic complications. Design: Prospective observational study. Setting: Neonates and infants undergoing cardiac surgery at a tertiary pediatric center. Participants: Neonates weighing 42.5 kg and infants r 12 months of age undergoing elective cardiac surgery with cardiopulmonary bypass were included in this prospective observational study. Intervention: None. Measurements and Results: Demographic data, surgical characteristics, transfusion data, and coagulation parameters (thromboeslatography and thromboelastometry) were collected. Logistic regression analysis was performed to identify potential determinants of postoperative thrombotic complication. Among the 138 neonates and infants included in the study, 12 (9%) developed a postoperative thrombotic complication. Unadjusted logistic regression analysis confirmed that the number and volume of blood products transfused was associated significantly with the increased incidence of thrombotic complication (odds ratio: 2.78, 95% confidence interval: 1.30-5.94, p ¼ 0.008). This association persisted after adjustment for patient’s age, the need for deep hypothermic cardiac arrest, and bypass time (odds ratio: 2.23, 95% confidence interval: 1.024.87, p ¼ 0.044). The number of blood products transfused was associated with a significant increase in parameters of clot amplitudes measured at cardiac intensive care unit admission, while no difference was reported when measured after the administration of protamine. Conclusions: This prospective observational study reports a significant association between transfusion of blood products in neonates and young infants undergoing cardiac surgery and an increased incidence of thrombotic complications in the absence of intraoperative coagulation monitoring. & 2017 Elsevier Inc. All rights reserved. Key Words: congenital heart disease; neonates; cardiac surgery; coagulopathy; bleeding
The authors have no conflict of interest. 1 Address reprint requests to Dr. David Faraoni, MD, PhD, FCCP, Department of Anesthesia and Pain Medicine, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. E-mail address: [email protected] (D. Faraoni).
NEONATES AND INFANTS undergoing cardiac surgery with cardiopulmonary bypass (CPB) are at high risk for major perioperative bleeding, and regularly require the transfusion of
http://dx.doi.org/10.1053/j.jvca.2017.04.039 1053-0770/& 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
2
large volumes of allogeneic blood products. Excessive postoperative bleeding in neonates after CPB has been shown to be independently associated with an increased incidence of major postoperative adverse events.1 Over the past decades, a variety of perioperative strategies have been used to improve perioperative management in this high-risk population. Most recently, attention has been directed toward the development of transfusion algorithms with point-of-care (POC)-testing used to guide patient-specific blood product administration.2,3 Studies in adult cardiac surgical patients clearly demonstrate the efficacy of such POC-based algorithms in reducing the need for allogenic blood products transfusion, decreasing the incidence of major postoperative bleeding, and reducing costs.4,5 To date, only 1 prospective study in children undergoing cardiac surgery has demonstrated the efficacy of a POC-based transfusion algorithm.6 Neonates and infants undergoing cardiac surgery also are at higher risk for postoperative thrombotic complications.7 In a recent retrospective analysis of the Health Care Cost and Use Project (HCUP) Kid’s Inpatient Database (KID), the authors reported that neonates and infants o12 months of age were at higher risk to develop a thrombotic complication in the perioperative period of cardiac surgery.8 Considering that blood product transfusion has been shown to contribute to the increased risk of thrombotic complication after adult cardiac surgery in a dose-dependent fashion,9 the authors hypothesized that algorithm-based transfusion of blood products in neonates and infants undergoing high-risk cardiac surgery in the absence of coagulation monitoring could increase the risk of thrombotic complication through the development of a pro-coagulant profile. Methods The study design is prospective, observational study at a single institution. The authors’ study protocol was approved by the institutional review board at Boston Children’s Hospital (P00016625), and recorded on clinicaltrial.gov (NCT02410473). Written informed consent was obtained from the parents/guardians of each patient. Neonates weighing greater than 2.5 kg and infants equal to or less than 12 months of age undergoing elective cardiac surgery with CPB were eligible for this prospective study. Neonates and infants undergoing emergent procedure and/or those deemed to be in a moribund condition (American Society of Anesthesiology Physical Status [ASA PS 5]) were excluded from the study. No specific interventions were required for this study. Anesthesia and CPB management were standardized per departmental protocols. The CPB circuit was primed with 1 unit of packed red blood cells (RBCs) and 1 unit of fresh frozen plasma (FFP). Target hematocrit during CPB was 430%. Following termination of CPB non-surgical, microvascular bleeding was treated with 1 to 2 units of volume reduced (20-30 mL/U) or non-volume reduced (40-50 mL/U) platelet concentrates. If microvascular bleeding continued 1 to
2 units of cryoprecipitate (20-30 mL/U) were transfused. If bleeding persisted despite these maneuvers, the sequence of platelets followed by cryoprecipitate transfusion was repeated. No FFP was transfused following CPB. Recombinant activated factor VII (rFVIIa) was administered at the discretion of the attending surgeon and anesthesiologist in cases where bleeding was deemed to be refractory to standard therapy. Target hematocrit following CPB was 40% in cyanotic patients and 35% in non-cyanotic patients. This algorithm has been utilized at the authors’ institution for 10 years, and its use is supported by both clinical efficacy and laboratory investigation of the coagulation defects found to exist in the authors’ neonatal and infant patient population.10 The authors’ objective was to understand the relationship between transfusion of blood products and the incidence of thrombotic complication with their current transfusion protocol. Thrombotic complication was defined as any arterial or venous thrombosis diagnosed between Cardiac Intensive Care Unit (CICU) admission and hospital discharge. This included clinically symptomatic events (shunt thrombosis, stroke, or limb ischemia) as well as clinically occult thromboses detected by imaging studies (echocardiography, catheterization, or ultrasonography). Location of thrombosis and interval between surgery and thrombosis event were recorded. Demographic data and surgical characteristics were recorded: gender, age, diagnosis, date of surgery, type of procedure. Duration of the surgery was defined as the time between skin incision and the last surgical stitch. The authors recorded CPB characteristics (prime volume, CPB duration, aortic clamp duration, and duration of deep hypothermic cardiac arrest [DHCA]). Blood product transfusion was defined as any intraoperative exposure to RBCs, FFP, cryoprecipitate, and platelet concentrates (during or outside CPB). Administration of rFVIIa also was recorded. Blood samples were obtained at 3 different time points: after induction of anesthesia and placement of arterial line (Baseline), 3 minutes after protamine administration (postprotamine), and upon arrival to CICU admission (CICU admission). Each blood sample consisted of: 1 citrated tube (1.8 mL) used for coagulation analysis (TEG and ROTEM). Because ROTEM currently is not integrated in the authors’ standard bleeding management strategy, the results obtained from the ROTEM analysis (EXTEM and FIBTEM) were not communicated to any caregivers. Bleeding management during the study period for all patients enrolled in the study were based on current available practices using standard coagulation assays. The authors recorded the following ROTEM parameters on EXTEM and FIBTEM: clotting time (CT [min]), angle (α [degree]), clot formation time (CFT [min]), maximum clot firmness (MCF [mm]), clot amplitudes measured after 10 minutes (A10, mm) and 20 minutes (A20, mm), the clot lysis index at 30 min (LI30, %) which corresponds to the percentage of remaining clot stability in relation to the MCF value 30 minutes after the clotting time, maximal lysis (ML, %), and lysis onset time (LOT, min). In addition, the authors recorded the following TEG parameters: clotting time (r, sec), angle (degree), and maximal amplitude (MA, mm).
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
Statistical Analysis Continuous variables were tested for normality with the Shapiro-Wilk test. Data are presented as median and interquartile range (25th percentile to 75th percentile) or mean and standard deviation accordingly. Categorical variables are expressed as number and percentage (%). For continuous variables, univariable comparisons were performed using the Mann–Whitney U test or t test based on the distribution. For categorical, chi-square was used for comparison between group. Logistic regression analysis also was performed to identify potential determinants of postoperative thrombotic complication. The overall area under the receiver operating characteristic (ROC) curve (AUC) was used as a measure of discrimination between those associated with and without thrombotic complication. Results are expressed as area under the ROC curve with 95% confidence intervals (95% CI), Younden criterion, sensitivity, and specificity. The p values less than 0.05 were considered statistically significant for all tests. Statistical analyses were performed using STATA version 14.2 for Mac OS (StataCorp, College Station, TX). Results Among the 138 neonates and infants included in the study, 12 (9%) developed a postoperative thrombotic complication. Patients with a thrombotic complication had lower preoperative SpO2 (81% [78-86] v 95% [88-100], p o 0.001), had longer CPB time (204 min [149-284] v 150 min [112-187], p ¼ 0.010), and required DHCA more frequently (10 [83%] v 41 [33%], p o 0.001). Those patients also had prolonged chest closure time (138 min [111-173] v 75 min [65-95], p o 0.001), and had increased chest tube output during the first 24 postoperative hours (37 mL/kg [25-59] v 20 mL/kg [13-27], p ¼ 0.001) (Table 1). Coagulation parameters (ROTEM and TEG) are summarized in Table 2. Parameters of clot amplitudes were significantly higher in neonates and children with a thrombotic complication at CICU admission compared with patients without thrombosis, while no differences were observed at baseline and post-protamine. Incidence and volume of blood products transfused in neonates and infants with and without thrombotic complication are summarized in Table 3. Patients with thrombotic complications were more frequently exposed to any blood products (11% v 0%, p ¼ 0.042), but also to a larger number of blood products, as well as larger volumes. The highest incidence of thrombosis was observed in neonates and infants who received a combination of platelets (PLT) þ cryoprecipitates (Cryo) þ rFVIIa (30%), compared with those who received a combination of PLT þ Cryo (14%), or PLT alone (3%) (p ¼ 0.009). Unadjusted logistic regression analysis confirmed that the number of blood products transfused was associated significantly with an increased incidence of thrombotic complication (odds ratio: 2.78, 95% CI: 1.30-5.94, p ¼ 0.008). This association persisted after adjustment for patient
3
Table 1 Demographic Characteristics of Neonates and Infants With and Without Thrombotic Complications Variables
Thrombosis (n ¼ 12)
Age (days) 14 (6-85) Weight (kg) 3.3 (3.1-3.8) Height (cm) 50 (49-52) Male (%) 6 (50) ASA PS (%) 3 1 (8) 4 11 (92) 5 0 (0) Preoperative SpO2 (%) 81 (78-86) Heart disease (%) TOF 1 (8) Double ventricle 6 (50) Single ventricle 5 (42) Previous Surgery (%) 4 (33) CBP time (min) 204 (149-284) Cross-clamp time (min) 109 (65-142) DHCA (%) 10 (83) Lowest temperature (ºC) 18 (18-22) Chest closure time (min) 138 (111-173) Chest tube output (mL/kg) 37 (25-59) Re-exploration for bleeding (%) 1 (8)
No Thrombosis (n ¼ 126) 52 3.7 53 75
(6-130) (3.1-5.4) (50-60) (60)
26 97 1 95
(21) (78) (1) (88-100)
35 60 31 16 150 99 41 28 75 20 3
(28) (48) (24) (13) (112-187) (72-120) (33) (22-28) (65-95) (13-27) (2)
p Value
0.677 0.132 0.122 0.522 0.542
o0.001 0.244
0.052 0.010 0.627 o0.001 0.001 o0.001 0.001 0.240
NOTE. Data are presented as median and interquartile range (25th percentile to 75th percentile), or number and percentage (%). Abbreviations: ASA PS, American Society of Anesthesiology Physical Status; CPB, cardiopulmonary bypass; DHCA, deep hypothermic cardiac arrest; TOF, Tetralogy of Fallot.
age, the need for DHCA, and CPB time (odds ratio: 2.23, 95% CI: 1.02-4.87, p ¼ 0.044). Results of the ROC analyses are reported in Table 4. Clot amplitude (MCF) 422 mm measured on FIBTEM at CICU admission offered the best predictive accuracy (AUC: 0.829, 95% CI: 0.624-0.914) to predict the incidence of thrombotic complication, when compared with the clot amplitude (MCF) 469 mm measured on EXTEM (AUC: 0.691, 95% CI: 0.636-0.783) or clot amplitude (MA) 467 mm measured on TEG (AUC: 0.770, 95% CI: 0.706-0.951). Figure 1 represents the association between the number of blood products transfused, and changes in ROTEM parameters. The number of blood products transfused was associated with a significant increase in parameters of clot amplitudes measured at CICU admission, while no difference was reported post-protamine when the parameters were within the normal range. Discussion This prospective observational study reports an association between transfusion of blood products in neonates and young infants undergoing cardiac surgery with CPB and an increased incidence of thrombotic complications. With use of a transfusion algorithm without intraoperative coagulation monitoring, neonates and young infants frequently were treated with cryoprecipitate, platelets, or rFVIIa, alone or in combination. In this context, transfusion of coagulation products led to a
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
4
Table 2 Coagulation Data in Neonates and Infants With and Without Thrombotic Complications
Table 3 Post-bypass Transfusion and Bleeding Data in Neonates and Infants With and Without Thrombotic Complications
Variables
Variables
TEG - R (sec) Baseline Post-protamine CICU admission TEG - Angle (deg) Baseline Post-protamine CICU admission TEG - MA (mm) Baseline Post-protamine CICU admission EXTEM - CT (sec) Baseline Post-protamine CICU admission EXTEM - CFT (sec) Baseline Post-protamine CICU admission EXTEM - Angle (deg) Baseline Post-protamine CICU admission EXTEM - MCF (sec) Baseline Post-protamine CICU admission FIBTEM - MCF (sec) Baseline Post-protamine CICU admission
Thrombosis (n ¼ 12)
No Thrombosis (n ¼ 126)
p Value
Transfusion
Thrombosis (n ¼ 12)
p Value
105 33
12 (11%) 0 (0%)
0.042
7 (6-8) 7 (6-8) 9 (6-9)
6 (5-7) 7 (6-8) 7 (6-8)
0.013 0.375 0.162
67 (64-73) 60 (48-72) 69 (67-75)
71 (66-75) 60 (54-64) 65 (59-70)
0.092 0.476 0.064
62 (53-67) 49 (45-55) 68 (65-75)
60 (54-64) 51 (46-55) 61 (54-67)
0.629 0.809 0.003
59 (53-71) 97 (84-108) 68 (61-74)
60 (55-68) 91 (79-101) 71 (64-81)
0.810 0.375 0.458
Any transfusion Yes No Multiple products None PLT Cryo PLT þ Cryo PLT þ Cryo þ rFVIIa Postoperative RBC (%) Postoperative RBC (mL/kg) Cell saver (mL/kg) Cryoprecipitate (mL/kg) Platelets (mL/kg)
82 (60-105) 183 (86-241) 67 (50-96)
89 (74-121) 198 (147-291) 95 (72-142)
0.336 0.321 0.024
NOTE. Data are presented as median and interquartile range (25th percentile to 75th percentile). Abbreviations: Cryo, cryoprecipitates; PLT, platelets; RBC, red blood cells; rFVIIa, recombinant human activated factor VII.
75 (72-78) 58 (51-73) 79 (75-80)
72 (67-75) 58 (49-66) 72 (64-76)
0.066 0.279 0.003
60 (56-67) 49 (41-63) 69 (61-76)
59 (54-62) 45 (40-51) 61 (52-68)
0.434 0.230 0.021
18 (12-32) 13 (8-16) 25 (17-28)
13 (10-17) 10 (8-11) 14 (10-19)
0.032 0.153 o0.001
NOTE. Data are presented as median and interquartile range (25th percentile to 75th percentile). Abbreviations: CT, clotting time; CFT, clot formation time; MA, maximum amplitude; MCF, maximal clot firmness.
pro-coagulant status, observed after the patients were admitted to the CICU, which might contribute to the development of further thrombotic complications. In neonates and young infants undergoing cardiac surgery with CPB, excessive blood loss often results from the development of a perioperative coagulopathy, which can be triggered by several factors such as contact between blood and non-endothelial surfaces, unfractionated heparin anticoagulation, hypothermia, the immaturity of the hemostatic system, and the higher degree of bypass hemodilution.11,12 Rapid detection of this coagulopathy should allow early goal-directed hemostatic therapy, aiming at preventing postoperative bleeding, reducing morbidity, mortality, and costs. This goaldirected therapy is best obtained through the development of specific algorithms, which are applied after identification of abnormal bleeding.13 Viscoelastic tests increasingly are used in adults and children undergoing cardiac and non-cardiac surgery because those tests allow for a global assessment of clot formation and clot stability using whole blood, and provide the treating physicians with important information
33 33 0 62 10
0 1 0 8 3
(0%) (3%) (0%) (14%) (30%)
0.009
0 (0-9)
0 (0-8)
0.044
41 (27-75) 25 (13-35)
15 (5-23) 0 (0-12)
o0.001 o0.001
17 (12-39)
5 (0-17)
0.002
within 5 to 10 minutes.14 Standard coagulation assays were not designed to monitor perioperative coagulation. The results of standard coagulation assays require 30 to 45 minutes, which decrease their accuracy when used to guide the administration of blood products in the presence of active bleeding.15 Algorithms have been designed specifically for the pediatric population.2,3,6 In a recent prospective randomized study, Nakayama et al reported that ROTEM-guided early hemostatic intervention by rapid intraoperative correction of EXTEM-A10 and FIBTEM-A10 reduced blood loss and red cell transfusion requirements after CPB, and reduced critical care duration in pediatric cardiac surgical patients.6 In another recent retrospective case-control study, Kane et al confirmed that intraoperative TEG also reduced platelet and cryoprecipitate transfusions without increasing the incidence of postoperative complications.16 While studies to date have examined the relationship between transfusion, bleeding, and outcomes, no pediatric studies have looked at the potential relationship between transfusion and the incidence of thrombotic complications. While a dose-dependent relationship between transfusion and Table 4 Predictive Accuracy and Cutoff of Coagulation Parameters Measured a CICU Admission to Predict Thrombosis in Neonates and Children Undergoing Cardiac Surgery Variables
AUC
95% CI
EXTEM - MCF (mm) 0.691 0.484-0.897 TEG - MA (mm) 0.770 0.706-0.951 FIBTEM – MCF (mm) 0.829 0.624-0.914
Cutoff Sensitivity Specificity 469 467 422
0.636 0.727 0.700
0.783 0.750 0.844
NOTE. Data are presented as area under the receiver operating curve (AUC) and 95% CI, cutoff corresponding to the Younden criterion, sensitivity, and specificity. Abbreviations: CI, confidence interval; CICU, cardiac intensive care unit; MA, maximum amplitude; MCF, maximal clot firmness.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
5
Fig 1. Maximal clot amplitudes measured on EXTEM (A) and FIBTEM (B) at baseline, post-protamine, and CICU admission in children that did not receive any blood products (Black), those who received platelets only (dark grey), platelets þ cryoprecipitate (light grey), or platelets þ cryoprecipitate þ recombinant human activated factor VII (white).
deep vein thrombosis following adult cardiac surgery has been reported,9 no study has examined specifically how excessive transfusion influences the development of thrombotic complications. In this study, the authors report a relationship between the number of products transfused and the incidence of thrombotic complications. This relationship remained after adjustment for variables related to surgical complexity and duration. The authors’ study also confirmed that the group of patients exposed to rFVIIa presented the highest incidence of thrombotic complication. In a recent retrospective study of children who received rFVIIa in conjunction with cardiac surgery, the authors found that pediatric patients with postCPB bleeding who received rFVIIa were an estimated 3.9 times (95% CI, 2.6-5.9) more likely to develop a thrombotic complication when compared with propensity-matched controls. Perioperative administration of rFVIIa was associated with increased length of CICU and overall hospital stay.17 Interestingly, the authors observed that coagulation variables obtained 3 minutes after administration of protamine were, for the most part, within the normal range. The prolonged EXTEM CFT in both groups post-protamine administration is consistent with the defect in platelet-fibrinogen mediated clot polymerization that forms the basis of the authors’ transfusion protocol. Nonetheless, the results suggest that in some instances over aggressive initiation of the authors’ transfusion protocol in the absence of POC test results led to creation of a hypercoagulable state at the time of CICU admission. The study presents some major limitations. When assessing the relationship between transfusion and the incidence of thrombotic complication, one could argue that neonates and infants who received more blood products underwent more complex procedures, and that the relationship could be explained by other comorbidities or postoperative complications associated with bleeding, transfusion, and/or surgery itself. If the authors’ study was not designed to look at those multiple aspects, the use of coagulation monitoring at different time-points allowed them to confirm the pro-coagulant profile in children who received multiple coagulation products. Further prospective studies are needed to confirm that POC-based transfusion could decrease the magnitude of
the pro-coagulant status, as well as the incidence of thrombotic complication. Despite the use of standardized transfusion protocol at the authors’ institution, some inter-physician difference could not be excluded completely. In conclusion, this study reports an association between transfusion of blood products in neonates and young infants undergoing cardiac surgery with CPB and an increased incidence of thrombotic complications. With use of a transfusion algorithm without intraoperative coagulation monitoring, neonates and young infants frequently were treated with cryoprecipitate, platelets, or rFVIIa, alone or in combination. It is quite possible that transfusion occurred in the absence of a ROTEM or TEG confirmed coagulopathy or that an existing coagulopathy was overtreated. This raises the obvious concern that the presence of a hypercoagulable state upon CICU arrival led to an increase in thrombotic complications. In addition, this study raises the question as to how microvascular bleeding, perceived to exist by the treating team of surgeons and anesthesiologists, should be treated in the absence of POC testing confirmation of a coagulopathic state. Further prospective studies are needed to determine whether a POC-based transfusion algorithm can reduce the incidence of a hypercoagulable state and the incidence of thrombotic complications. Acknowledgments The authors thank Michelle Mulone and Gregory Owren for their help during the study. D.F. has received the 2016 SABMHaemonetics Research Starter Grant Award, from the Society for the Advancement of Patient Blood Management (www. SABM.org). This study was also founded in part by the Goldwin Foundation (S.E.), and Trailblazer Award, Department of Anesthesiology, Perioperative and Pain Medicine, Boston Children’s Hospital (J.C.I.). References 1 Guzzetta NA, Allen NN, Wilson EC, et al. Excessive postoperative bleeding and outcomes in neonates undergoing cardiopulmonary bypass. Anesth Analg 2015;120:405–10.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
6
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
2 Romlin B, Wåhlander H, Berggren H, et al. Intraoperative thromboelastometry is associated with reduced transfusion prevalence in pediatric cardiac surgery. Anesth Analg 2011;112:30–6. 3 Faraoni D, Willems A, Romlin BS, et al. Development of a specific algorithm to guide haemostatic therapy in children undergoing cardiac surgery: A single-centre retrospective study. Eur J Anaesthesiol 2015;32: 320–9. 4 Weber C, Görlinger K, Meininger D, et al. Point-of-care testing: A prospective, randomized clinical trial of efficacy in coagulopathic cardiac surgery patients. Anesthesiology 2012;117:531–47. 5 Karkouti K, Callum J, Wijeysundera D, et al. Point-of-care hemostatic testing in cardiac surgery: A stepped-wedge clustered randomized controlled trial. Circulation 2016;134:1152–62. 6 Nakayama Y, Nakajima Y, Tanaka KA, et al. Thromboelastometry-guided intraoperative haemostatic management reduces bleeding and red cell transfusion after paediatric cardiac surgery. Br J Anaesth 2015;114: 91–102. 7 Giglia T, Massicotte M, Tweddell J, et al. Prevention and treatment of thrombosis in pediatric and congenital heart disease: A scientific statement from the American Heart Association. Circulation 2013;128:2622–703. 8 Faraoni D, Gardella KM, Odegard KC, et al. Incidence and predictors for postoperative thrombotic complications in children with surgical and nonsurgical heart disease. Ann Thorac Surg 2016;102:1360–7. 9 Ghazi L, Schwann TA, Engoren MC, et al. Role of blood transfusion product type and amount in deep vein thrombosis after cardiac surgery. Thromb Res 2015;136:1204–10.
10 Hornykewycz S, Odegard KC, Castro RA, et al. Hemostatic consequences of a non-fresh or reconstituted whole blood small volume cardiopulmonary bypass prime in neonates and infants. Paediatr Anaesth 2009;19:854–61. 11 Osthaus W, Boethig D, Johanning K, et al. Whole blood coagulation measured by modified thrombelastography (ROTEM) is impaired in infants with congenital heart diseases. Blood Coagul Fibrinolysis 2008;19: 220–5. 12 Arnold P. Treatment and monitoring of coagulation abnormalities in children undergoing heart surgery. Paediatr Anaesth 2011;21:494–503. 13 Savan V, Willems A, Faraoni D, et al. Multivariate model for predicting postoperative blood loss in children undergoing cardiac surgery: A preliminary study. Br J Anaesth 2014;112:708–14. 14 Perez-Ferrer A, Vicente-Sanchez J, Carceles-Baron MD, et al. Early thromboelastometry variables predict maximum clot firmness in children undergoing cardiac and non-cardiac surgery. Br J Anaesth 2015;115: 896–902. 15 Segal J, Dzik W, Transfusion Medicine/Hemostasis Clinical Trials Network. Paucity of studies to support that abnormal coagulation test results predict bleeding in the setting of invasive procedures: An evidence-based review. Transfusion 2005;45:1413–25. 16 Kane LC, Woodward CS, Husain SA, et al. Thromboelastography–does it impact blood component transfusion in pediatric heart surgery? J Surg Res 2016;200:21–7. 17 Downey L, Brown ML, Faraoni D, et al. Recombinant factor VIIa is associated with increased thrombotic complications in pediatric cardiac surgery patients. Anesth Analg 2017;124:1431–6.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
Contents lists available at ScienceDirect
HOSTED BY
journal homepage: www.jcvaonline.com
Original Article
Relationship Between Transfusion of Blood Products and the Incidence of Thrombotic Complications in Neonates and Infants Undergoing Cardiac Surgery David Faraoni, MD, PhD, FCCPn,1, Sirisha Emani, PhD†, Erin Halpin, MSN, RN‡, Rachel Bernier, MPH‡, Sitaram Emani, MD†, James A. DiNardo, MD, FAAP‡, Juan C. Ibla, MD‡ *
Department of Anesthesia and Pain Medicine, The Hospital for Sick Children, University of Toronto, Toronto, Canada † Department of Cardiovascular Surgery, Boston Children’s Hospital, Harvard Medical School, Boston, MA ‡ Department of Anesthesiology, Peri-operative and Pain Medicine, Boston Children’s Hospital, Harvard Medical School, Boston, MA
Objectives: The authors hypothesized that transfusion of blood products in neonates and infants undergoing high-risk cardiac surgery in the absence of intraoperative coagulation monitoring increases the risk of thrombotic complications. Design: Prospective observational study. Setting: Neonates and infants undergoing cardiac surgery at a tertiary pediatric center. Participants: Neonates weighing 42.5 kg and infants r 12 months of age undergoing elective cardiac surgery with cardiopulmonary bypass were included in this prospective observational study. Intervention: None. Measurements and Results: Demographic data, surgical characteristics, transfusion data, and coagulation parameters (thromboeslatography and thromboelastometry) were collected. Logistic regression analysis was performed to identify potential determinants of postoperative thrombotic complication. Among the 138 neonates and infants included in the study, 12 (9%) developed a postoperative thrombotic complication. Unadjusted logistic regression analysis confirmed that the number and volume of blood products transfused was associated significantly with the increased incidence of thrombotic complication (odds ratio: 2.78, 95% confidence interval: 1.30-5.94, p ¼ 0.008). This association persisted after adjustment for patient’s age, the need for deep hypothermic cardiac arrest, and bypass time (odds ratio: 2.23, 95% confidence interval: 1.024.87, p ¼ 0.044). The number of blood products transfused was associated with a significant increase in parameters of clot amplitudes measured at cardiac intensive care unit admission, while no difference was reported when measured after the administration of protamine. Conclusions: This prospective observational study reports a significant association between transfusion of blood products in neonates and young infants undergoing cardiac surgery and an increased incidence of thrombotic complications in the absence of intraoperative coagulation monitoring. & 2017 Elsevier Inc. All rights reserved. Key Words: congenital heart disease; neonates; cardiac surgery; coagulopathy; bleeding
The authors have no conflict of interest. 1 Address reprint requests to Dr. David Faraoni, MD, PhD, FCCP, Department of Anesthesia and Pain Medicine, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. E-mail address: [email protected] (D. Faraoni).
NEONATES AND INFANTS undergoing cardiac surgery with cardiopulmonary bypass (CPB) are at high risk for major perioperative bleeding, and regularly require the transfusion of
http://dx.doi.org/10.1053/j.jvca.2017.04.039 1053-0770/& 2017 Elsevier Inc. All rights reserved.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
2
large volumes of allogeneic blood products. Excessive postoperative bleeding in neonates after CPB has been shown to be independently associated with an increased incidence of major postoperative adverse events.1 Over the past decades, a variety of perioperative strategies have been used to improve perioperative management in this high-risk population. Most recently, attention has been directed toward the development of transfusion algorithms with point-of-care (POC)-testing used to guide patient-specific blood product administration.2,3 Studies in adult cardiac surgical patients clearly demonstrate the efficacy of such POC-based algorithms in reducing the need for allogenic blood products transfusion, decreasing the incidence of major postoperative bleeding, and reducing costs.4,5 To date, only 1 prospective study in children undergoing cardiac surgery has demonstrated the efficacy of a POC-based transfusion algorithm.6 Neonates and infants undergoing cardiac surgery also are at higher risk for postoperative thrombotic complications.7 In a recent retrospective analysis of the Health Care Cost and Use Project (HCUP) Kid’s Inpatient Database (KID), the authors reported that neonates and infants o12 months of age were at higher risk to develop a thrombotic complication in the perioperative period of cardiac surgery.8 Considering that blood product transfusion has been shown to contribute to the increased risk of thrombotic complication after adult cardiac surgery in a dose-dependent fashion,9 the authors hypothesized that algorithm-based transfusion of blood products in neonates and infants undergoing high-risk cardiac surgery in the absence of coagulation monitoring could increase the risk of thrombotic complication through the development of a pro-coagulant profile. Methods The study design is prospective, observational study at a single institution. The authors’ study protocol was approved by the institutional review board at Boston Children’s Hospital (P00016625), and recorded on clinicaltrial.gov (NCT02410473). Written informed consent was obtained from the parents/guardians of each patient. Neonates weighing greater than 2.5 kg and infants equal to or less than 12 months of age undergoing elective cardiac surgery with CPB were eligible for this prospective study. Neonates and infants undergoing emergent procedure and/or those deemed to be in a moribund condition (American Society of Anesthesiology Physical Status [ASA PS 5]) were excluded from the study. No specific interventions were required for this study. Anesthesia and CPB management were standardized per departmental protocols. The CPB circuit was primed with 1 unit of packed red blood cells (RBCs) and 1 unit of fresh frozen plasma (FFP). Target hematocrit during CPB was 430%. Following termination of CPB non-surgical, microvascular bleeding was treated with 1 to 2 units of volume reduced (20-30 mL/U) or non-volume reduced (40-50 mL/U) platelet concentrates. If microvascular bleeding continued 1 to
2 units of cryoprecipitate (20-30 mL/U) were transfused. If bleeding persisted despite these maneuvers, the sequence of platelets followed by cryoprecipitate transfusion was repeated. No FFP was transfused following CPB. Recombinant activated factor VII (rFVIIa) was administered at the discretion of the attending surgeon and anesthesiologist in cases where bleeding was deemed to be refractory to standard therapy. Target hematocrit following CPB was 40% in cyanotic patients and 35% in non-cyanotic patients. This algorithm has been utilized at the authors’ institution for 10 years, and its use is supported by both clinical efficacy and laboratory investigation of the coagulation defects found to exist in the authors’ neonatal and infant patient population.10 The authors’ objective was to understand the relationship between transfusion of blood products and the incidence of thrombotic complication with their current transfusion protocol. Thrombotic complication was defined as any arterial or venous thrombosis diagnosed between Cardiac Intensive Care Unit (CICU) admission and hospital discharge. This included clinically symptomatic events (shunt thrombosis, stroke, or limb ischemia) as well as clinically occult thromboses detected by imaging studies (echocardiography, catheterization, or ultrasonography). Location of thrombosis and interval between surgery and thrombosis event were recorded. Demographic data and surgical characteristics were recorded: gender, age, diagnosis, date of surgery, type of procedure. Duration of the surgery was defined as the time between skin incision and the last surgical stitch. The authors recorded CPB characteristics (prime volume, CPB duration, aortic clamp duration, and duration of deep hypothermic cardiac arrest [DHCA]). Blood product transfusion was defined as any intraoperative exposure to RBCs, FFP, cryoprecipitate, and platelet concentrates (during or outside CPB). Administration of rFVIIa also was recorded. Blood samples were obtained at 3 different time points: after induction of anesthesia and placement of arterial line (Baseline), 3 minutes after protamine administration (postprotamine), and upon arrival to CICU admission (CICU admission). Each blood sample consisted of: 1 citrated tube (1.8 mL) used for coagulation analysis (TEG and ROTEM). Because ROTEM currently is not integrated in the authors’ standard bleeding management strategy, the results obtained from the ROTEM analysis (EXTEM and FIBTEM) were not communicated to any caregivers. Bleeding management during the study period for all patients enrolled in the study were based on current available practices using standard coagulation assays. The authors recorded the following ROTEM parameters on EXTEM and FIBTEM: clotting time (CT [min]), angle (α [degree]), clot formation time (CFT [min]), maximum clot firmness (MCF [mm]), clot amplitudes measured after 10 minutes (A10, mm) and 20 minutes (A20, mm), the clot lysis index at 30 min (LI30, %) which corresponds to the percentage of remaining clot stability in relation to the MCF value 30 minutes after the clotting time, maximal lysis (ML, %), and lysis onset time (LOT, min). In addition, the authors recorded the following TEG parameters: clotting time (r, sec), angle (degree), and maximal amplitude (MA, mm).
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
Statistical Analysis Continuous variables were tested for normality with the Shapiro-Wilk test. Data are presented as median and interquartile range (25th percentile to 75th percentile) or mean and standard deviation accordingly. Categorical variables are expressed as number and percentage (%). For continuous variables, univariable comparisons were performed using the Mann–Whitney U test or t test based on the distribution. For categorical, chi-square was used for comparison between group. Logistic regression analysis also was performed to identify potential determinants of postoperative thrombotic complication. The overall area under the receiver operating characteristic (ROC) curve (AUC) was used as a measure of discrimination between those associated with and without thrombotic complication. Results are expressed as area under the ROC curve with 95% confidence intervals (95% CI), Younden criterion, sensitivity, and specificity. The p values less than 0.05 were considered statistically significant for all tests. Statistical analyses were performed using STATA version 14.2 for Mac OS (StataCorp, College Station, TX). Results Among the 138 neonates and infants included in the study, 12 (9%) developed a postoperative thrombotic complication. Patients with a thrombotic complication had lower preoperative SpO2 (81% [78-86] v 95% [88-100], p o 0.001), had longer CPB time (204 min [149-284] v 150 min [112-187], p ¼ 0.010), and required DHCA more frequently (10 [83%] v 41 [33%], p o 0.001). Those patients also had prolonged chest closure time (138 min [111-173] v 75 min [65-95], p o 0.001), and had increased chest tube output during the first 24 postoperative hours (37 mL/kg [25-59] v 20 mL/kg [13-27], p ¼ 0.001) (Table 1). Coagulation parameters (ROTEM and TEG) are summarized in Table 2. Parameters of clot amplitudes were significantly higher in neonates and children with a thrombotic complication at CICU admission compared with patients without thrombosis, while no differences were observed at baseline and post-protamine. Incidence and volume of blood products transfused in neonates and infants with and without thrombotic complication are summarized in Table 3. Patients with thrombotic complications were more frequently exposed to any blood products (11% v 0%, p ¼ 0.042), but also to a larger number of blood products, as well as larger volumes. The highest incidence of thrombosis was observed in neonates and infants who received a combination of platelets (PLT) þ cryoprecipitates (Cryo) þ rFVIIa (30%), compared with those who received a combination of PLT þ Cryo (14%), or PLT alone (3%) (p ¼ 0.009). Unadjusted logistic regression analysis confirmed that the number of blood products transfused was associated significantly with an increased incidence of thrombotic complication (odds ratio: 2.78, 95% CI: 1.30-5.94, p ¼ 0.008). This association persisted after adjustment for patient
3
Table 1 Demographic Characteristics of Neonates and Infants With and Without Thrombotic Complications Variables
Thrombosis (n ¼ 12)
Age (days) 14 (6-85) Weight (kg) 3.3 (3.1-3.8) Height (cm) 50 (49-52) Male (%) 6 (50) ASA PS (%) 3 1 (8) 4 11 (92) 5 0 (0) Preoperative SpO2 (%) 81 (78-86) Heart disease (%) TOF 1 (8) Double ventricle 6 (50) Single ventricle 5 (42) Previous Surgery (%) 4 (33) CBP time (min) 204 (149-284) Cross-clamp time (min) 109 (65-142) DHCA (%) 10 (83) Lowest temperature (ºC) 18 (18-22) Chest closure time (min) 138 (111-173) Chest tube output (mL/kg) 37 (25-59) Re-exploration for bleeding (%) 1 (8)
No Thrombosis (n ¼ 126) 52 3.7 53 75
(6-130) (3.1-5.4) (50-60) (60)
26 97 1 95
(21) (78) (1) (88-100)
35 60 31 16 150 99 41 28 75 20 3
(28) (48) (24) (13) (112-187) (72-120) (33) (22-28) (65-95) (13-27) (2)
p Value
0.677 0.132 0.122 0.522 0.542
o0.001 0.244
0.052 0.010 0.627 o0.001 0.001 o0.001 0.001 0.240
NOTE. Data are presented as median and interquartile range (25th percentile to 75th percentile), or number and percentage (%). Abbreviations: ASA PS, American Society of Anesthesiology Physical Status; CPB, cardiopulmonary bypass; DHCA, deep hypothermic cardiac arrest; TOF, Tetralogy of Fallot.
age, the need for DHCA, and CPB time (odds ratio: 2.23, 95% CI: 1.02-4.87, p ¼ 0.044). Results of the ROC analyses are reported in Table 4. Clot amplitude (MCF) 422 mm measured on FIBTEM at CICU admission offered the best predictive accuracy (AUC: 0.829, 95% CI: 0.624-0.914) to predict the incidence of thrombotic complication, when compared with the clot amplitude (MCF) 469 mm measured on EXTEM (AUC: 0.691, 95% CI: 0.636-0.783) or clot amplitude (MA) 467 mm measured on TEG (AUC: 0.770, 95% CI: 0.706-0.951). Figure 1 represents the association between the number of blood products transfused, and changes in ROTEM parameters. The number of blood products transfused was associated with a significant increase in parameters of clot amplitudes measured at CICU admission, while no difference was reported post-protamine when the parameters were within the normal range. Discussion This prospective observational study reports an association between transfusion of blood products in neonates and young infants undergoing cardiac surgery with CPB and an increased incidence of thrombotic complications. With use of a transfusion algorithm without intraoperative coagulation monitoring, neonates and young infants frequently were treated with cryoprecipitate, platelets, or rFVIIa, alone or in combination. In this context, transfusion of coagulation products led to a
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
4
Table 2 Coagulation Data in Neonates and Infants With and Without Thrombotic Complications
Table 3 Post-bypass Transfusion and Bleeding Data in Neonates and Infants With and Without Thrombotic Complications
Variables
Variables
TEG - R (sec) Baseline Post-protamine CICU admission TEG - Angle (deg) Baseline Post-protamine CICU admission TEG - MA (mm) Baseline Post-protamine CICU admission EXTEM - CT (sec) Baseline Post-protamine CICU admission EXTEM - CFT (sec) Baseline Post-protamine CICU admission EXTEM - Angle (deg) Baseline Post-protamine CICU admission EXTEM - MCF (sec) Baseline Post-protamine CICU admission FIBTEM - MCF (sec) Baseline Post-protamine CICU admission
Thrombosis (n ¼ 12)
No Thrombosis (n ¼ 126)
p Value
Transfusion
Thrombosis (n ¼ 12)
p Value
105 33
12 (11%) 0 (0%)
0.042
7 (6-8) 7 (6-8) 9 (6-9)
6 (5-7) 7 (6-8) 7 (6-8)
0.013 0.375 0.162
67 (64-73) 60 (48-72) 69 (67-75)
71 (66-75) 60 (54-64) 65 (59-70)
0.092 0.476 0.064
62 (53-67) 49 (45-55) 68 (65-75)
60 (54-64) 51 (46-55) 61 (54-67)
0.629 0.809 0.003
59 (53-71) 97 (84-108) 68 (61-74)
60 (55-68) 91 (79-101) 71 (64-81)
0.810 0.375 0.458
Any transfusion Yes No Multiple products None PLT Cryo PLT þ Cryo PLT þ Cryo þ rFVIIa Postoperative RBC (%) Postoperative RBC (mL/kg) Cell saver (mL/kg) Cryoprecipitate (mL/kg) Platelets (mL/kg)
82 (60-105) 183 (86-241) 67 (50-96)
89 (74-121) 198 (147-291) 95 (72-142)
0.336 0.321 0.024
NOTE. Data are presented as median and interquartile range (25th percentile to 75th percentile). Abbreviations: Cryo, cryoprecipitates; PLT, platelets; RBC, red blood cells; rFVIIa, recombinant human activated factor VII.
75 (72-78) 58 (51-73) 79 (75-80)
72 (67-75) 58 (49-66) 72 (64-76)
0.066 0.279 0.003
60 (56-67) 49 (41-63) 69 (61-76)
59 (54-62) 45 (40-51) 61 (52-68)
0.434 0.230 0.021
18 (12-32) 13 (8-16) 25 (17-28)
13 (10-17) 10 (8-11) 14 (10-19)
0.032 0.153 o0.001
NOTE. Data are presented as median and interquartile range (25th percentile to 75th percentile). Abbreviations: CT, clotting time; CFT, clot formation time; MA, maximum amplitude; MCF, maximal clot firmness.
pro-coagulant status, observed after the patients were admitted to the CICU, which might contribute to the development of further thrombotic complications. In neonates and young infants undergoing cardiac surgery with CPB, excessive blood loss often results from the development of a perioperative coagulopathy, which can be triggered by several factors such as contact between blood and non-endothelial surfaces, unfractionated heparin anticoagulation, hypothermia, the immaturity of the hemostatic system, and the higher degree of bypass hemodilution.11,12 Rapid detection of this coagulopathy should allow early goal-directed hemostatic therapy, aiming at preventing postoperative bleeding, reducing morbidity, mortality, and costs. This goaldirected therapy is best obtained through the development of specific algorithms, which are applied after identification of abnormal bleeding.13 Viscoelastic tests increasingly are used in adults and children undergoing cardiac and non-cardiac surgery because those tests allow for a global assessment of clot formation and clot stability using whole blood, and provide the treating physicians with important information
33 33 0 62 10
0 1 0 8 3
(0%) (3%) (0%) (14%) (30%)
0.009
0 (0-9)
0 (0-8)
0.044
41 (27-75) 25 (13-35)
15 (5-23) 0 (0-12)
o0.001 o0.001
17 (12-39)
5 (0-17)
0.002
within 5 to 10 minutes.14 Standard coagulation assays were not designed to monitor perioperative coagulation. The results of standard coagulation assays require 30 to 45 minutes, which decrease their accuracy when used to guide the administration of blood products in the presence of active bleeding.15 Algorithms have been designed specifically for the pediatric population.2,3,6 In a recent prospective randomized study, Nakayama et al reported that ROTEM-guided early hemostatic intervention by rapid intraoperative correction of EXTEM-A10 and FIBTEM-A10 reduced blood loss and red cell transfusion requirements after CPB, and reduced critical care duration in pediatric cardiac surgical patients.6 In another recent retrospective case-control study, Kane et al confirmed that intraoperative TEG also reduced platelet and cryoprecipitate transfusions without increasing the incidence of postoperative complications.16 While studies to date have examined the relationship between transfusion, bleeding, and outcomes, no pediatric studies have looked at the potential relationship between transfusion and the incidence of thrombotic complications. While a dose-dependent relationship between transfusion and Table 4 Predictive Accuracy and Cutoff of Coagulation Parameters Measured a CICU Admission to Predict Thrombosis in Neonates and Children Undergoing Cardiac Surgery Variables
AUC
95% CI
EXTEM - MCF (mm) 0.691 0.484-0.897 TEG - MA (mm) 0.770 0.706-0.951 FIBTEM – MCF (mm) 0.829 0.624-0.914
Cutoff Sensitivity Specificity 469 467 422
0.636 0.727 0.700
0.783 0.750 0.844
NOTE. Data are presented as area under the receiver operating curve (AUC) and 95% CI, cutoff corresponding to the Younden criterion, sensitivity, and specificity. Abbreviations: CI, confidence interval; CICU, cardiac intensive care unit; MA, maximum amplitude; MCF, maximal clot firmness.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
5
Fig 1. Maximal clot amplitudes measured on EXTEM (A) and FIBTEM (B) at baseline, post-protamine, and CICU admission in children that did not receive any blood products (Black), those who received platelets only (dark grey), platelets þ cryoprecipitate (light grey), or platelets þ cryoprecipitate þ recombinant human activated factor VII (white).
deep vein thrombosis following adult cardiac surgery has been reported,9 no study has examined specifically how excessive transfusion influences the development of thrombotic complications. In this study, the authors report a relationship between the number of products transfused and the incidence of thrombotic complications. This relationship remained after adjustment for variables related to surgical complexity and duration. The authors’ study also confirmed that the group of patients exposed to rFVIIa presented the highest incidence of thrombotic complication. In a recent retrospective study of children who received rFVIIa in conjunction with cardiac surgery, the authors found that pediatric patients with postCPB bleeding who received rFVIIa were an estimated 3.9 times (95% CI, 2.6-5.9) more likely to develop a thrombotic complication when compared with propensity-matched controls. Perioperative administration of rFVIIa was associated with increased length of CICU and overall hospital stay.17 Interestingly, the authors observed that coagulation variables obtained 3 minutes after administration of protamine were, for the most part, within the normal range. The prolonged EXTEM CFT in both groups post-protamine administration is consistent with the defect in platelet-fibrinogen mediated clot polymerization that forms the basis of the authors’ transfusion protocol. Nonetheless, the results suggest that in some instances over aggressive initiation of the authors’ transfusion protocol in the absence of POC test results led to creation of a hypercoagulable state at the time of CICU admission. The study presents some major limitations. When assessing the relationship between transfusion and the incidence of thrombotic complication, one could argue that neonates and infants who received more blood products underwent more complex procedures, and that the relationship could be explained by other comorbidities or postoperative complications associated with bleeding, transfusion, and/or surgery itself. If the authors’ study was not designed to look at those multiple aspects, the use of coagulation monitoring at different time-points allowed them to confirm the pro-coagulant profile in children who received multiple coagulation products. Further prospective studies are needed to confirm that POC-based transfusion could decrease the magnitude of
the pro-coagulant status, as well as the incidence of thrombotic complication. Despite the use of standardized transfusion protocol at the authors’ institution, some inter-physician difference could not be excluded completely. In conclusion, this study reports an association between transfusion of blood products in neonates and young infants undergoing cardiac surgery with CPB and an increased incidence of thrombotic complications. With use of a transfusion algorithm without intraoperative coagulation monitoring, neonates and young infants frequently were treated with cryoprecipitate, platelets, or rFVIIa, alone or in combination. It is quite possible that transfusion occurred in the absence of a ROTEM or TEG confirmed coagulopathy or that an existing coagulopathy was overtreated. This raises the obvious concern that the presence of a hypercoagulable state upon CICU arrival led to an increase in thrombotic complications. In addition, this study raises the question as to how microvascular bleeding, perceived to exist by the treating team of surgeons and anesthesiologists, should be treated in the absence of POC testing confirmation of a coagulopathic state. Further prospective studies are needed to determine whether a POC-based transfusion algorithm can reduce the incidence of a hypercoagulable state and the incidence of thrombotic complications. Acknowledgments The authors thank Michelle Mulone and Gregory Owren for their help during the study. D.F. has received the 2016 SABMHaemonetics Research Starter Grant Award, from the Society for the Advancement of Patient Blood Management (www. SABM.org). This study was also founded in part by the Goldwin Foundation (S.E.), and Trailblazer Award, Department of Anesthesiology, Perioperative and Pain Medicine, Boston Children’s Hospital (J.C.I.). References 1 Guzzetta NA, Allen NN, Wilson EC, et al. Excessive postoperative bleeding and outcomes in neonates undergoing cardiopulmonary bypass. Anesth Analg 2015;120:405–10.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039
6
D. Faraoni et al. / Journal of Cardiothoracic and Vascular Anesthesia ] (]]]]) ]]]–]]]
2 Romlin B, Wåhlander H, Berggren H, et al. Intraoperative thromboelastometry is associated with reduced transfusion prevalence in pediatric cardiac surgery. Anesth Analg 2011;112:30–6. 3 Faraoni D, Willems A, Romlin BS, et al. Development of a specific algorithm to guide haemostatic therapy in children undergoing cardiac surgery: A single-centre retrospective study. Eur J Anaesthesiol 2015;32: 320–9. 4 Weber C, Görlinger K, Meininger D, et al. Point-of-care testing: A prospective, randomized clinical trial of efficacy in coagulopathic cardiac surgery patients. Anesthesiology 2012;117:531–47. 5 Karkouti K, Callum J, Wijeysundera D, et al. Point-of-care hemostatic testing in cardiac surgery: A stepped-wedge clustered randomized controlled trial. Circulation 2016;134:1152–62. 6 Nakayama Y, Nakajima Y, Tanaka KA, et al. Thromboelastometry-guided intraoperative haemostatic management reduces bleeding and red cell transfusion after paediatric cardiac surgery. Br J Anaesth 2015;114: 91–102. 7 Giglia T, Massicotte M, Tweddell J, et al. Prevention and treatment of thrombosis in pediatric and congenital heart disease: A scientific statement from the American Heart Association. Circulation 2013;128:2622–703. 8 Faraoni D, Gardella KM, Odegard KC, et al. Incidence and predictors for postoperative thrombotic complications in children with surgical and nonsurgical heart disease. Ann Thorac Surg 2016;102:1360–7. 9 Ghazi L, Schwann TA, Engoren MC, et al. Role of blood transfusion product type and amount in deep vein thrombosis after cardiac surgery. Thromb Res 2015;136:1204–10.
10 Hornykewycz S, Odegard KC, Castro RA, et al. Hemostatic consequences of a non-fresh or reconstituted whole blood small volume cardiopulmonary bypass prime in neonates and infants. Paediatr Anaesth 2009;19:854–61. 11 Osthaus W, Boethig D, Johanning K, et al. Whole blood coagulation measured by modified thrombelastography (ROTEM) is impaired in infants with congenital heart diseases. Blood Coagul Fibrinolysis 2008;19: 220–5. 12 Arnold P. Treatment and monitoring of coagulation abnormalities in children undergoing heart surgery. Paediatr Anaesth 2011;21:494–503. 13 Savan V, Willems A, Faraoni D, et al. Multivariate model for predicting postoperative blood loss in children undergoing cardiac surgery: A preliminary study. Br J Anaesth 2014;112:708–14. 14 Perez-Ferrer A, Vicente-Sanchez J, Carceles-Baron MD, et al. Early thromboelastometry variables predict maximum clot firmness in children undergoing cardiac and non-cardiac surgery. Br J Anaesth 2015;115: 896–902. 15 Segal J, Dzik W, Transfusion Medicine/Hemostasis Clinical Trials Network. Paucity of studies to support that abnormal coagulation test results predict bleeding in the setting of invasive procedures: An evidence-based review. Transfusion 2005;45:1413–25. 16 Kane LC, Woodward CS, Husain SA, et al. Thromboelastography–does it impact blood component transfusion in pediatric heart surgery? J Surg Res 2016;200:21–7. 17 Downey L, Brown ML, Faraoni D, et al. Recombinant factor VIIa is associated with increased thrombotic complications in pediatric cardiac surgery patients. Anesth Analg 2017;124:1431–6.
Please cite this article as: Faraoni D, et al. (2017), http://dx.doi.org/10.1053/j.jvca.2017.04.039