1

Amicar versus TXA

<LRH>Halanski et al.

<DOC>Original Article

<AT>The Efficacy of Amicar Vversus Tranexamic Acid in Pediatric Spinal Deformity Surgery: aA Prospective, Randomized, Double-bBlinded Pilot Study.

<AU>

Matthew A. Halanski, M.Da,.*, Jeffrey A. Cassidy, M.Db., Scott Hetzel, M.Sc., Diann Reischmann, Ph.Dd., and Nabil Hassan, M.De.

<AFN>aDepartment of Orthopaedics and Rehabilitation, University of Wisconsin/American Family Children’s Hospital, UMMF Centennial Building, 1685 Highland Avenue, Room 6170-12D, Madison, WI 53705-2281, USA

<AFN>bDepartment of Pediatric Orthopaedics, Helen DeVos Children’s Hospital, 100 Michigan NE, Grand Rapids, MI 49503, USA

<AFN>cDepartment of Biostatistics and Medical Informatics, University of Wisconsin–Madison, 4269A HSLC, 750 Highland Avenue, Madison, WI 53705, USA

<AFN>d&bull;&bull;&bull;, Grand Valley State University, A-1-178 Mackinac Hall, Allendale, MI 49401, USA

[g1]<AFN>ePediatric Critical Care Medicine and Pediatric Blood Management Program, Helen DeVos Children’s Hospital, 100 Michigan NE, Grand Rapids, MI 49503, USA

<ARTFN>Author disclosures: MAH (none), JAC (none), SH (none), DR (none), NH (none).

<COR>*Corresponding author. Department of Orthopaedics and Rehabilitation,

Corresponding Author & Reprints:

Matthew A. Halanski M.D.

University of Wisconsin/American Family Children’s Hospital,

Department of Orthopaedics and Rehabilitation

UMMF Centennial Building,

1685 Highland Avenue,

Room 6170-12D,

Madison, WI 53705-2281, USA. Tel.:

Phone: (608) -263-6208; fax:

Fax: (608)- 263-5631;

E-mail: (M.A. Halanski)

Study performed at:

Spectrum Health/Helen DeVos Children’s Hospital

Jeffrey A. Cassidy, M.D.

Helen DeVos Children’s Hospital

Pediatric Orthopaedics

100 Michigan NE
Grand Rapids, Michigan 49503

Email:

Scott Hetzel, M.S.
University of Wisconsin - Madison
Department of Biostatistics and Medical Informatics
4269A HSLC
750 Highland Ave.
Madison, WI 53705

Email:

Diann Reischmann, Ph.D.

Grand Valley State University

A-1-178 Mackinac Hall

Allendale, MI 49401

Email:

Nabil Hassan, M.D.

Helen DeVos Children’s Hospital

Pediatric Critical Care Medicine and Pediatric Blood Management Program

100 Michigan NE
Grand Rapids, Michigan 49503

Email:

IRB: Spectrum Health Research & Human Rights Committee/IRB – approval September 2, 2008.<BEGIN ABSTRACT>

Author disclosures: MAH (none), JAC (none), SH (none), DR (none), NH (none).

Abstract

Study Design:

Single-center, prospective, randomized, double-blinded trial.

Objectives:

To compare blood loss, allogenic transfusion requirements, and coagulation parameters between pediatric spinal deformity patients receiving aminocaproic acid (Amicar) or tTranexamic aAcid (TXA) during posterior spinal fusion.

Summary of Background Data:

Amicar and Tranexamic acid (TXA) have been shown to decrease blood loss in pediatric spinal deformity cases when compared with controls. The difference in efficacy between these medications in this population has not been reported.

Methods:

Enrolled patients were randomized to receive either Amicar or TXA during their scoliosis surgery. Baseline demographic and deformity comparisons were collected. Intra-operative comparisons included: estimated and calculated blood loss, number of levels instrumented, number of osteotomies, operative time, and allogenic transfusion requirements. PrePreoperative- and postoperative hemoglobin, platelets, prothrombin timePT, PTT,partial prothrombin time (PTT), international normalized ratio (INR)INR, and fibrinogen were recorded.

Results:

A total of 47Forty-seven patients were enrolled with data available for review (N = 25, Amicar,; N = 22, TXA). No difference in cohorts was found in demographics, preoperative hemoglobin, platelets, prothrombin timePT, PTT, and INRINR, initial cobbCobb angle, average number of: levels fused, patients with osteotomies and, & osteotomies, operative time, and final cobbCobb angles. Estimated blood loss was significantly less (about ~221 mlL) than the calculated blood loss in both groups (p = 0.003). Estimated blood loss (, 1,088 ml vs.s. 726 mL;( p = 0.055), and calculated blood loss (, 1,366 ml vs. 903 mL;l ( p = 0.13), trended higher in the in the Amicar group. AlthoughWhile no difference in allogenic transfusion rates (20% vs. 14%) wasere observed, average volumes transfused were significantly higher in the Amicar cohort (1,014 ml vs. 461 mlL; ) p = 0.03). The TXA cohort demonstrated a statistically significant smaller change in INR, a lower PTT, and greater fibrinogen levels post-operatively.

Conclusions:

Compared with Amicar, TXA use was associated with a lower allogenic transfusion requirement, less alteration in post-operative clotting studies, and a trend toward lower blood loss in pediatric posterior spinal fusion PSF patients.

<END ABSTRACT>

Keywords: Scoliosis; Blood loss; Anti-fibrinolytics; Amicar; Tranexamic aAcid (TXA)

<H1>Introduction

Previous reports have shown the efficacy of using aAprotinin [1], aminocaproic acid (Amicar) [g2][2-4], and tranexamic acid TXA [5-8] in decreasing blood loss in scoliosis surgery. These agents are felt thought to decrease blood loss by their anti-fibrinolytic properties. These products have been effective in cardiac surgery [9, 10], arthroplasty [11-13], and scoliosis surgery [1, 3, 4, 6-8]. Previous studies demonstrated a decreased blood loss when comparing the use of the anti-fibrinolytic agents withto a control [2, 4, 6-8] in the spinal deformity population. Amicar was found to decreased blood loss at an average of 400–-600 mL/casepatient [4], and TXA by 40% [7], when compared with controls.

The purpose of this study was to determine which anti-fibrinolytic agent used during posterior spinal fusion for scoliosis was more effective at minimizing peri-operative blood loss and need for transfusion, using a prospective, randomized, double- blinded trial. As Because previous studies have demonstrated the primary effect of Amicar to be in decreasing post-operative drainage [2], and as because the researcherswe do not routinely use post-operative wound drains, theour hypothesis was that TXA would be more effective at controlling blood loss in patients undergoing posterior spinal instrumentation and fusion than would Amicar.

<H1>Materials and Methods

The authors performed aA single- center, prospective, randomized, double-blinded comparison between TXA and Amicar was performed. After they obtained institutional review board obtaining IRB approval, they offered enrollment int the study to all patients scheduled to undergo posterior spinal fusion by our the pediatric spinal deformity service, at were offered enrollment into the study from January 2009 to December 2010. Each patient was pre-operatively evaluated by our the pediatric blood avoidance service. This service consisted of pPediatric Iintensivists and nurses with interests in minimizing pediatric patient exposure to allogenic blood products during high-risk procedures. During this evaluation, the study was explained and interested patients were provided consented. Those refusing randomization werewould be given the agent of choice of the anesthetist at the time of surgery, as was our current practice;, thus, all patients were to received Amicar or TXA, unless a contraindication for anti-fibrinolytic treatment wascould be found. Pre-operative laboratory valuess obtained included hemoglobin, platelets, prothrombin time (PT), partial prothrombin time (PTT), international normalized ratio (INR), and fibrinogen. Our The basic inclusion criteriona for recruitment was a patient with idiopathic or neuromuscular scoliosis less younger than 18 years of age, undergoing posterior spinal fusion. Medical exclusion criteria were inability to obtain a full consent, patient or parent refusal to participate, personal or family history of bleeding or clotting disorders, abnormal pre-operative clotting studies or von Willebrand screen, patients at risk for clotting as a result ofdue to long- term oral contraceptive use, and or severe immobility. Patients were routinely given back any “salvage” blood from the Ccell Ssaver unit (Haemonetics, Braintree, MA). A standard transfusion trigger of hemoglobin less than 7 or patient symptomology was used throughout the study period. A study process flowchart (See Appendix 1, Supplemental Digital Content 1) and our blood avoidance protocol (Appendix 2, Supplemental Digital Content 2) are available for review.

<H2>Method for aAssigning sSubjects to tTreatment gGroups

Upon consenting to the study, randomization was performed by a member of the blood avoidance team using a block, blinded method. At least 24 hours in advance of surgery, the BAS [g3]staff member drew an envelope and provided the pediatric staff pharmacist with a sealed envelope containing a drug-specific order set. The pharmacist prepared the study medication;, labeleding it as either “sStudy mMedication A” or “sStudy mMedication B”, and delivered it to the anesthesiologist in a fashion that wais indistinguishable in terms of which drug is was going to be used. The drug used in the study iwas blinded to all care providers (the blood avoidance team, the surgeon, the anesthesiologist, and surgical and post-operative nursing staff) throughout the study period. Only the pharmacist knew what drug had been used. If the anesthesiologist or surgeon suspected drug-related side effects, were suspected by the anesthesiologist or surgeon, the nature ofr the infusion was revealed.

<H2>Treatment rRegimen

A core group of pediatric staff pharmacists prepared mMedications for both treatment arms were prepared by a core group of pediatric staff pharmacists. All medications were prepared and secured in the pediatric satellite pharmacy. Medications were delivered by the pharmacy to the pediatric anesthesiologist beforeprior to surgery., Epsilon Amicaraminocaproic acid and TXAtranexamic acid were diluted to a final concentration of 10 mg/mlL. Whether using Amicar epsilon aminocaproic acid or TXAtranexamic acid, the first syringe contained an initial bolus dose of 100 mg/kg, which was to be given once over 30 minutes at the beginning of the operative procedure. A second syringe was prepared for study medications A and B with the same final concentration of 10 mg/mlL, and a continuous infusion was initiated to run at 10 mg/kg/hrour for the duration of the surgical procedure. At the completion of the surgery, a third prepared syringe was started. The Amicarepsilon aminocaproic acid, syringe contained the same contents as the second syringe, 10 mg/mlL, and was run at the same rate of 10 mg/kg once over the next 24 hours post operatively. (This is the routine method for Amicarepsilon aminocaproic acid administration postoperatively.) For the TXAtranexamic acid group, the third syringe contained normal saline, which was. Normal saline was infused at the same weight-based rate as calculated for Amicarthe epsilon aminocaproic acid over the next 24 hours post operatively. The TXA was discontinued after the procedure as because 90% of it is eliminated at 24 hours;, however Amicar infusion must be continued post-operatively as because it is excreted much faster, at 4 to -6 hours. The pediatric staff pharmacist notified nursing staff of the time the post-operative infusion is was to be complete. Medications were recorded in the medication administration record (MAR) as mMedication A or Medication B.

All cases procedures were performed by the same 2two, fellowship- trained, orthopaedic surgeons working simultaneous as a team. Typically, these procedures utilized used electro-cautery, subperiosteal dissection, topical hemostatic agents, and the Cell -Saver blood salvage system (Haemonetics, Braintree, MA). For the first case patient, the Aquamantis (Salient Technologies, Portsmouth, New Hamphsire) radiofrequency sealer was trialed;, however, it was inconsistently utilized used during the procedure and was therefore eliminated. Blood loss was monitored in three 3 waysmanners. Standard “estimated blood loss” was recorded from the medical record by estimates made in the operating room by anesthetist and/or surgeon looking at the wound, drapes, drain canisters, sponges, and so forth.etc. In attempts to more objectively track blood loss more objectively, we the researchers attempted to collect all blood loss into the Cell Saver system and weighed all sponges. CA “calculated blood loss” (CBL) was then determined by subtracting the volume of heparinized saline from the total volume in the Cell Saver system, and blood- soaked sponges from averaged dry sponge weight. These were then combined to give a total CBL“calculated blood loss”. In the two 2 casespatients where in whom a post-operative suction drain was used, this volume was then added to the CBLcalculated blood loss in a separate analysis. Finally, pre-operative and post-operative hemoglobin and intra-operative blood products given to the patient were recorded. Total operative time, levels fused, and number of posterior osteotomies performed were obtained from the electronic medical record. DIt should be noted that during the study period, the surgeons changed from routinely performing Ponte osteotomies to more standard facetectomies after a review failed to demonstrate any clinically significant improvement in correction [14]. Laboratory valuess were obtained, including hemoglobin, platelets, PT, PTT, INR, and fibrinogen post-operatively, on postoperative days #1, and on postoperative day #3. Complications and unexpected events associated with treatment were recorded and reported to the IRBinstitutional review board. The study period was confined to the patients’ operative hospital stay. Finally, preoperative- and post-operative Cobb angles were measured by fellowship- trained pediatric orthopaedic surgeons.

At the time of data review, cases were reviewed to select for those patients having with a diagnosis of neuromuscular scoliosis or adolescent idiopathic- (like) scoliosis. Idiopathic- like curves included patients with curves treated similar to adolescent idiopathic scoliosis (AIS), such as those related to a tethered cord or Chiari malformation, but that technically were not AIS curves. Exclusion criteria were subjectively significant kyphotic deformities, fusionless constructs, congenital scoliosis, and potentially other confounding factors (neurofibromatosis or surgeries not performed by the standard two 2 surgeons). Figure 1 shows aA flow diagram of patients enrolled can be seen in Figure 1. Orthopaedic exclusions after randomization not included in this analysis included patients with congenital scoliosis (n = 4), kyphosis (n = 1), and nurofibromatosis-1 NF-1 (n = 1), and a syndromic patient.

<H2>Statistical methods

A pre-study pPower analysis was performed to estimate the sample size needed to complete this study. Utilizing Using an the nQuery software package, [g4]based on an assumed difference in mean blood loss between the two 2 groups of 8 mLl/kg and a standard deviation in both groups of 15 mLl/kg, a sample size of 57 patients in each group would be needed for a two 2-tailed test with .05 alpha error and 80% power. Assuming an eleven 11-mlL/kg difference in the means while maintaining the other assumption, 31 patients would be needed in each group. Fisher eExact tests were utilized used tofor evaluation ofe categorical variables, and Student t T-tests were used for continuous variables. SA standard p value of < 0.05 was used to signify a significant findingce.

<H1>Results

A total of 47Forty-seven patients were enrolled with data available for review (N = 25 Amicar;, N = 22 TXA) (Table 12). Ten patients had neuromuscular scoliosis (5 with Amicar 5 and 5 with TXA 5). No differences were seen in demographics (age, sex, or weight). Similarly, no significant differences were found in preoperative hemoglobin, platelets, fibrinogen, PT, PTT, INR, or initial Cobb angles (Table 21).

Similarly the average operative time, number of levels fused, number of patients with osteotomies, average number of osteotomies performed throughout the group time per level (26 vs. 27 minutes), and final Cobb angles (18 vs. 14) were similar between the groups (. Table 12).

Estimated blood loss was significantly less than CBLthe calculated blood loss in both groups, an averaginge of 221 mlL less (p = 0.003). AThe average estimated blood loss (95% confidence interval) was 1,088 mL (7,741–-1,435 mL) [g5]mLl and 726 mL (583–-869 mL) ml (p = 0.055), whereasile CBLthe calculated blood loss was 1,366 mL (789–-1942 mL)) ml in the Amicar group and 903 mL (677–-1128 mL) ml) in the TXA group (p = 0.1). Blood loss per level was then compared using CBLthe “calculated blood loss: I,” in the Amicar group, it was 119/ per level, and was 101/ per level in the TXA group (p = 0.5). No difference in allogenic transfusion rates between groups was observed, (Amicar, 5 of /25 vs.ersus TXA, 3 of /22; (20% vs. 14%). AThe average volumes of allogenic transfusion (1,014 mL [(544–-1,484 mL]) ml vs. 461 mL [(46–-877) mlL]) were higher in the Amicar cohort (p = 0.03) (, Table 3). No differences in pre-operative hemoglobin, post-operative hemoglobin, and or change in hemoglobin were observed (, Tables 4 and 5). Although it was not the authors’While not our routine practice, at the time of closure two 2 patients in the Amicar group were felt thoughtat the time of closure to require post-operative suction drains, presumably from ongoing blood loss. The volume recorded post-operatively from these drains was 559 ml and 735 mlL. A separate analysis including these values in forthe CBLCBL wasere prerformed, which increased the CBL in the Amicar to 1,422 mlL (843–-2,001 mL) and lowered the p value to 0.09. Further sub-group analyses were performed separating neuromuscular patients from idiopathic- like patients and patients with and without Ponte osteotomies (See Table 6, Supplemental Digital Content 31).

Other laboratory markers were recorded and compared over the hospitalization. No differences were found in the hemoglobin or platelets between cohorts. The TXA cohort demonstrated a statistically significant smaller change from pre-operative INR, to postoperatively and on postoperative day #1. Similarly, a lower PTT was noticed post-operatively in the TXA cohort that normalized by post-operative day #1. Fibrinogen levels also remained significantly higher in the TXA cohort throughout the post-operative period (, Tables 4 and 5). Sub-group analyses were performed separating neuromuscular patients from idiopathic- like patients and patients with and without Ponte osteotomies (See Tables 7 and 8, Supplemental Digital Content 42 and 53).

<H1>Discussion

Tranexamic acid TXA and Amicar have been shown to decrease blood loss when compared with controls in scoliotic populations;, however, the benefits appear to occur at different time points. Sethna et al. [7] described a 40% decrease in intra-operative blood loss using TXA, whereasile Florentino-Pineda et al. [2] demonstrated no difference in intra-operative blood loss using Amicar, but did noted a significant decrease in post-operative wound drainage (drain output). As Because it was not our the current authors’ standard to use a post-operative suction drain, we they hypothesized our that the intra-operative blood loss would be less in the TXA group. Although theyWhile we did not see a significant difference in overall blood loss or transfusion rates,; statistically significant: lower allogenic transfusion volumes, smaller changes in clotting studies, and higher fibrinogen levels were noted in the TXA cohort. Looking specifically at the AIS sub-groups ([over all and, with, and without osteotomies]) (Table 6), very little difference in CBLcalculated blood loss was found between cohorts (about ~100 mLcc). Depending on how the blood loss was normalized in these cohorts (by number of levels fused or by patient weight [kilograms](kg)), the average blood loss was slightly greater or slightly less for one medication versus the other. Despite this parity in blood loss, early post-operative fibrinogen levels remained significantly higher for the TXA AIS cohort (Table 7). Given Because of the similarity between the average blood losses, the clinical relevance of these small transient differences in laboratory values may be negligible in the AIS cohorts, but might also explain the lower transfusion volumes in the TXA cohort.