Thrombotic Risk of Recombinant Factor Seven in Pediatric Cardiac Surgery: A Single Institution Experience

Thrombotic Risk of Recombinant Factor Seven in Pediatric Cardiac Surgery: A Single Institution Experience Todd J. Karsies, MD, Kathleen K. Nicol, MD, Mark E. Galantowicz, MD, FACS, Julie A. Stephens, MS, and Bryce A. Kerlin, MD PEDIATRIC CARDIAC

Divisions of Critical Care and Pediatric Hematology/Oncology/BMT, Department of Laboratory Medicine, and The Heart Center, Nationwide Children’s Hospital, Columbus; Center for Clinical and Translational Research, The Research Institute at Nationwide Children’s Hospital, Columbus; Departments of Pathology, Surgery, and Pediatrics, The Ohio State University College of Medicine, Columbus; and Center for Biostatistics, The Ohio State University, Columbus, Ohio

Background. Recombinant activated factor seven (rFVIIa) is increasingly being used as a hemostatic adjunct in pediatric cardiac surgery. We evaluated the thrombotic safety profile of rFVIIa in pediatric congenital heart disease (CHD) surgery. Methods. This was a retrospective matched case-control study over six years at a single institution. Patients who received rFVIIa after CHD surgery were matched to controls based on age, diagnosis, and procedure. We compared thrombosis, hemorrhage, transfusions, length of stay, and repeat procedures between groups. Results. Twenty-five patients received rFVIIa (mean dose: 70 mcg/kg); 50 controls were matched. There was no significant difference in the rate of thrombosis between patients who received rFVIIa and controls (8% vs 4%).

After rFVIIa, there was a significant reduction in transfusion volume (median 77.1 mL/kg vs 14.6 mL/kg; p < 0.001) as well as a significant decrease in hemorrhagic chest tube output (8.3 ⴞ 1.6 mL/kg/hour vs 1.4 ⴞ 0.3 mL/kg/hour; mean ⴞ standard error of the mean; p < 0.001). No difference was seen in intensive care unit or hospital length of stay or mortality between patients receiving rFVIIa and controls. Conclusions. The rFVIIa therapy did not increase thrombotic complications when used as rescue therapy after CHD surgery but did appear to decrease bleeding complications in this small cohort.

S

risk may be due to a variety of factors, including proteinlosing enteropathy, endothelial dysfunction, and abnormalities in clotting factors even without protein loss [6, 7]. These abnormalities appear most pronounced in the first 1 to 2 days after surgery and are not unique to singleventricle physiology [8, 9]. Recombinant activated factor VII, or rFVIIa (NovoSeven; Novo Nordisk, Princeton, NJ), is a prohemostatic agent approved by the Food and Drug Administration in 1999 to treat bleeding in patients with hemophilia and inhibitors to factor VIII or IX; it has subsequently gained additional approval for the treatment of factor VII deficiency. Since its introduction, rFVIIa has been widely used as an off-label hemostatic adjuvant therapy. In several case series, some with cardiac surgery patients, it has been reported to reduce blood transfusions, chest tube output, and other bleeding complications [1, 10 –15]. To date there have been few randomized controlled studies looking at rFVIIa use in off-label indications [14, 16 –18]. Despite its increasing use, there have been few

evere postoperative bleeding is a major source of morbidity and mortality in pediatric patients undergoing cardiac surgery and cardiopulmonary bypass for correction of congenital heart defects (CHD). Blood loss and blood product transfusion can exceed 100 mL/kg, and bleeding requiring surgical reexploration occurs in 1% of all pediatric cardiac surgery patients [1]. Many times there is no identifiable source of bleeding, and the bleeding may be prolonged despite standard therapy with blood products. In addition to problems directly related to blood loss such as need for surgery, hypovolemic shock, and decreased oxygen delivery, these patients are also at increased risk for infectious complications due to the potential exposure to large numbers of blood products as well as other complications such as transfusion-related acute lung injury [2– 4]. In addition to increased risk of bleeding, patients with CHD are also at increased risk of thrombotic complications. This is especially common in patients with singleventricle physiology both prior to and after Fontan repair, with the reported incidence of thromboembolic events ranging from 3% to 20% [5, 6]. This thrombosis

Accepted for publication Nov 9, 2009. Address correspondence to Dr Kerlin, 700 Children’s Dr, ED 583A, Columbus, OH 43205; e-mail: [email protected].

© 2010 by The Society of Thoracic Surgeons Published by Elsevier Inc

(Ann Thorac Surg 2010;89:570 –7) © 2010 by The Society of Thoracic Surgeons

This article has been selected for the open discussion forum on the CTSNet Web Site: http://www.ctsnet. org/sections/newsandviews/discussions/index.html

0003-4975/10/$36.00 doi:10.1016/j.athoracsur.2009.11.023

Abbreviations and Acronyms aPTT ⫽ Activated partial thromboplastin time AS/AI ⫽ Aortic stenosis and (or) insufficiency CHD ⫽ Congenital heart disease DIC ⫽ Disseminated intravascular coagulation DKS ⫽ Damus-Kaye-Stansel d-TGA ⫽ d-transposition of the great arteries DVT ⫽ Deep venous thrombosis FFP ⫽ Fresh frozen plasma HIT ⫽ Heparin-induced thrombocytopenia HLHS ⫽ Hypoplastic left heart syndrome IVC ⫽ Inferior vena cava LV ⫽ Left ventricle NCH ⫽ Nationwide Children’s Hospital PA ⫽ Pulmonary artery PRBC ⫽ Packed red blood cells PT ⫽ Prothrombin Time rFVIIa ⫽ Recombinant activated factor seven RV ⫽ Right ventricle TAPVR ⫽ Total anomalous pulmonary venous return

studies of any kind in the setting of pediatric congenital heart surgery, and controversy remains regarding both its safety and effectiveness in this population [19, 20]. Given the potential increased risk for thrombosis found in CHD patients, there is concern that the routine use of rFVIIa in pediatric patients undergoing cardiac surgery will result in an unacceptably high rate of thrombotic events. In recent studies, there is disagreement as to whether there is an increased rate of thrombotic complications in pediatric CHD patients who receive rFVIIa postoperatively, although these studies were designed primarily to examine the efficacy of rFVIIa [19, 20]. In light of these controversies, the primary objective of this study was to examine the thrombotic safety profile of rFVIIa administered to CHD surgery patients at our institution. To do so, we retrospectively assessed the incidence of clinically significant thrombosis in patients who did or did not receive rFVIIa who were matched for well-recognized risk factors for CHD-associated thrombosis (age, CHD diagnosis, and type of repair) [6, 8, 21]. Secondary outcomes included assessments of mortality, nonthrombotic morbidity, length of stay, and efficacy.

Patients and Methods The study protocol was approved by the Institutional Review Board of the Nationwide Children’s Hospital (NCH), Columbus, Ohio (IRB05-00503). Written informed consent was waived due to the retrospective study design.

Patients Patients who received rFVIIa were identified from a hospital pharmacy database. All patients who received rFVIIa either intraoperatively or within 1 week after

KARSIES ET AL PEDIATRIC RECOMBINANT FACTOR VII THERAPY

571

surgery for congenital heart disease during the period from January 1, 1999 to December 31, 2005 were considered for inclusion. Patients were excluded if they had a diagnosis of hemophilia, a diagnosis of congenital factor VII deficiency, were greater than 21 years old, or had previously received rFVIIa. Controls from the same time period were selected and matched for age, diagnosis, and type of procedure in order to control for known epidemiologic determinants of thrombotic risk in CHD. To be considered a match for age, a control needed to be within 1 month if less than 1 year old, within 6 months if between 1 year and 5 years old, within 1 year if between 5 years and 10 years old, and within 2 years if greater than 10 years old. If no age-matched controls were found, then 2 or 3 of the closest age matches were selected. This only occurred with patients less than 1 year old, and in all cases the age was within 3 months.

Standard Institutional Cardiopulmonary Bypass Regimen During the study time frame, saline-primed bypass circuits were preferred for patients greater than 5 kg and were primed with blood for children less than 5 kg. Some Jehovah’s Witness patients weighing less than 5 kg were saline primed. Although the records were incomplete, we estimate that about 50% of both cases and controls were blood primed. Blood transfused in the primed circuit or during cardiopulmonary bypass was not included in the transfusion requirement analysis. Bypass was performed using the Jostra HL20 (Maquet Cardiopulmonary AG, Hirrlingen, Germany) with either the Terumo Capiox SX10R or SX18R oxygenators for children or the Terumo Baby Rx for infants (Terumo Cardiovascular Systems, Ann Arbor, MI). The patients were heparinized during bypass using reversal with protamine sulfate at the end of the case. The patients also received aprotinin during bypass: loading dose 10,000 KIU/kg; pump dose 10,000 KIU/kg; and maintenance dose 10,000 KIU/kg/hour. Aprotinin was discontinued at the end of the case and not typically given in the postoperative setting.

Standard Institutional rFVIIa Usage In our institution, rFVIIa is typically used as rescue therapy for bleeding observed in the operating room (OR) or the intensive care unit (ICU) that is refractory to standard therapy with blood products. The typical transfusion practice for management of bleeding without an identifiable surgical source is as follows. A total of 5 to 10 mL/kg of apheresis platelets and 10 mL/kg of fresh frozen plasma are empirically given for excessive bleeding risking hemodynamic compromise. If there is hemodynamic instability or laboratory evidence of anemia, then 15 to 20 mL/kg of packed red blood cells are also transfused. Cryoprecipitate (1 unit/10 kg) may be given based on abnormal fibrinogen levels. If bleeding persists despite normal temperature, pH, and calcium, these transfusions are typically repeated. If laboratory values such as prothrombin time (PT), activated partial thromboplastin time (aPTT), fibrinogen, or platelet count are available, specific blood products may be chosen based on these

PEDIATRIC CARDIAC

Ann Thorac Surg 2010;89:570 –7

572

KARSIES ET AL PEDIATRIC RECOMBINANT FACTOR VII THERAPY

results. The addition of rFVIIa is considered at this point for refractory bleeding. Dosage is at the discretion of the treating physician, although the most commonly used dose is 90 micrograms per kilogram (mcg/kg). The dose may be repeated and additional blood products given if bleeding continues.

Definitions

PEDIATRIC CARDIAC

Significant bleeding was defined as bleeding requiring transfusion or causing hemodynamic compromise. Significant thrombosis was defined as any venous or arterial thrombosis, or ischemic stroke. Readmission was defined as an unplanned admission to the hospital for any reason within 3 months after initial discharge.

Data Collection Data were collected from patient charts, as well as OR, perfusion, pharmacy, and blood bank records. Data collected included the following: date of birth; congenital heart disease (CHD) diagnosis; other diagnoses; previous surgeries; admission weight; date of hospital admission, discharge, and surgery; any history of thrombosis or bleeding, including preoperative anticoagulation; preoperative medications; preoperative and postoperative laboratory results; blood product usage; bypass times; rFVIIa usage and dosing; postoperative bleeding or thrombotic complications, including any diagnostic testing, anticoagulation, or additional procedures; chest tube output; and readmission within 3 months. These data were entered into a database and deidentified. The risk adjustment in congenital heart surgery (RACHS-1)

Table 1. Breakdown of Diagnoses and Procedures Procedure/Diagnosis Procedure: Arterial switch Ross or Ross-Konno TAPVR repair Comprehensive stage 2a Pulmonary valve replacement Otherb Diagnosis: AS/AI d-TGA TAPVR HLHS Otherc

Number of Patients (%)

Number of Controls (%)

6 (24) 6 (24) 3 (12) 3 (12) 2 (8) 5 (20)

9 (18) 7 (14) 6 (12) 9 (18) 6 (12) 13 (26)

7 (28) 6 (24) 3 (12) 3 (12) 6 (24)

9 (18) 11 (22) 6 (12) 10 (20) 14 (28)

a Comprehensive stage 2 includes bidirectional Glenn as well as aortic b arch reconstruction/Damus-Kaye-Stansel; other procedures include central shunt, right ventricle (RV) (or left ventricle [LV]) to pulmonary artery conduit, aortic valve replacement, bilateral bidirectional Glenn, c and heart transplant; other diagnoses include pulmonary stenosisinsufficiency, tricuspid atresia, complex LV outflow tract obstruction, tetralogy of Fallot with pulmonary atresia, l-TGA, double-chamber RV, anomalous coronary arteries, and congenital cardiomyopathy.

AS/AI ⫽ aortic stenosis with aortic insufficiency; d-TGA ⫽ d-transposition of the great arteries; HLHS ⫽ hypoplastic left heart syndrome; TAPVR ⫽ total anomalous pulmonary venous return.

Ann Thorac Surg 2010;89:570 –7

Table 2. Demographics and Laboratory Dataa Variables Age (months) Male, n (%) Previous bleeding, n (%) Previous thrombosis, n (%) RACHS-1 score Bypass time (minutes) Aortic cross-clamp time (minutes) Preoperative platelet count (thousands) Postoperative platelet count (thousands) Postoperative bleeding, n (%)b Mortality, n (%)

All Patients

rFVIIa

Control

56 ⫾ 9 43 (57) 3 (4) 3 (4) 3.6 ⫾ 1.3 217 ⫾ 10 105 ⫾ 8

50 ⫾ 15 56 ⫾ 11 13 (52) 30 (60) 2 (8) 1 (2) 1 (4) 2 (4) 3.7 ⫾ 1.1 3.6 ⫾ 1.3 228 ⫾ 12 214 ⫾ 13 127 ⫾ 14 98 ⫾ 10

296 ⫾ 12

278 ⫾ 20

306 ⫾ 15

144 ⫾ 7

162 ⫾ 15

134 ⫾ 7

35 (47)

25 (100)

10 (21)

3 (4)

1 (4)

2 (4)

Values are mean ⫾ standard error of the mean unless otherwise b p ⬍ 0.001. stated; a

RACHS-1 ⫽ risk adjustment for congenital heart surgery; recombinant activated factor seven.

rFVIIa ⫽

scores were calculated for all patients to allow comparison of surgical risk between patients [21]. The RACHS-1 score is a consensus-based tool for risk stratification of pediatric patients undergoing congenital heart surgery that was developed using a large database of pediatric congenital heart patients and allows for comparison of surgical outcomes while taking into account baseline risks. It has been shown in additional studies to be an effective tool for predicting risk of mortality in pediatric congenital heart surgery [22].

Statistical Analysis Statistical analysis was performed using SAS 9.1 (SAS Institute, Inc, Cary, NC). The Fisher exact test was used to compare all categoric variables except for the rate of thrombosis and gender between the cases and controls. The Wilcoxon rank sum test was used to compare the total transfusion requirement, length of stay in the ICU, and the total length of stay. Otherwise a Student t test was used for comparison of continuous variables such as baseline, operative, and postoperative characteristics between groups. Transfusion requirements pretreatment and posttreatment were compared using a signed rank test. Chest tube output pretreatment and posttreatment was compared using a paired t test and between groups using a Student t test. A p value less than 0.05 was considered significant. Continuous variables are reported as mean ⫾ standard error of the mean) unless otherwise stated.

Results Fifty-nine patients received rFVIIa at NCH during the study time frame. Of these 34 were CHD surgery patients, 9 of whom were excluded due either to presence of a known clotting factor deficiency or because they re-

Fig 1. Comparison of transfusion requirements before and after rFVIIa dose in patients who received rFVIIa after CHD surgery. (*p ⬍ 0.001; **p ⫽ 0.002; bars displayed as median ⫾ interquartile range; Cryo ⫽ cryoprecipitate; FFP ⫽ fresh frozen plasma; pRBC ⫽ packed red blood cells; rFVIIa ⫽ recombinant activated factor seven.)

ceived rFVIIa after a procedure other than CHD surgery. A total of 25 patients received rFVIIa and were included; 50 controls were matched after searching 1,165 records for appropriate matches. Surgical procedures and diagnoses for all patients are shown in Table 1. There were no significant differences in baseline or intraoperative characteristics between groups, including age, history of thrombosis or bleeding complications, preoperative anticoagulation, RACHS-1 score, bypass and cross-clamp time, and postoperative platelet count (Table 2). The mean rFVIIa dose (with standard deviation) was 70 (⫾ 28) mcg/kg. Most patients (64%) received a single dose of 90 mcg/kg, although some patients received a lower initial dose. A second dose was given to 28% of patients, and 68% of patients received at least one dose, if not both doses, in the OR. All patients who received rFVIIa had significant bleeding complications, while 22% of controls had bleeding complications (p ⬍ 0.001). There were no significant differences between patients who received rFVIIa and controls in rate of thrombosis, need for readmission, or mortality. Significant thrombotic complications were seen in 2 patients from each group (8% rFVIIa vs 4% controls; p ⫽ 0.597). Of the rFVIIa patients, 20% were readmitted, while 15% of controls were readmitted (p ⫽ 0.519). The mortality rate in both groups was identical (4%). Furthermore, there were no significant differences in the need for additional procedures due to bleeding (16% rFVIIa vs 6% controls; p ⫽ 0.213) or thrombosis (0% rFVIIa vs 2% controls; p⫽0.999) and no difference in the ICU or hospital length of stay between groups. A total of 2 rFVIIa patients and 5 controls required delayed chest closure (p ⫽ 0.999). No patients who received rFVIIa required postoperative extracorporeal membrane oxy-

KARSIES ET AL PEDIATRIC RECOMBINANT FACTOR VII THERAPY

573

genation (ECMO), while 3 controls did. All controls requiring ECMO did have significant hemorrhagic complications, 1 had thrombotic complications, and 2 ultimately died. Neither of the rFVIIa patients who had thrombotic complications required additional procedures due to their thrombosis, although both did receive systemic anticoagulation. The first of these patients had a diagnosis of tricuspid atresia that had been palliated with a modified Blalock-Taussig (BT) shunt. She presented with a thrombosed BT shunt and underwent takedown of the BT shunt with placement of a central shunt. She received 2 doses of rFVIIa in the OR for bleeding despite massive transfusion. Her postoperative course was complicated by shock requiring reexploration of her chest as well as right iliac artery thrombosis, portal venous thrombosis, as well as ischemic stroke. Her evaluation for hypercoaguable state revealed undetectable protein C levels; she was treated with a heparin drip followed by aspirin therapy. The other rFVIIa patient with postoperative thrombosis underwent orthotopic heart transplantation. She required ECMO preoperatively, and this course was complicated by severe disseminated intravascular coagulation (DIC). She also received rFVIIa in the OR. The postoperative course was complicated by continued DIC as well as a left leg deep venous thrombosis and inferior vena cava thrombus. She also required a heparin drip that was transitioned to subcutaneous enoxaparin therapy. Her hypercoaguable evaluation did not reveal any abnormalities. Two patients in the control group also had serious thrombotic complications. The first had d-transposition of the great arteries with ventricular septal defect (VSD)

Fig 2. Comparison of total postoperative transfusion volume for patients who received recombinant activated factor seven (rFVIIa) after congenital heart disease surgery and matched controls, with controls further stratified into those with significant bleeding and those without significant postoperative bleeding. (p ⫽ not significant for all comparisons; bars displayed as median ⫾ interquartile range).

PEDIATRIC CARDIAC

Ann Thorac Surg 2010;89:570 –7

574

KARSIES ET AL PEDIATRIC RECOMBINANT FACTOR VII THERAPY

PEDIATRIC CARDIAC

and underwent arterial switch and VSD closure. His immediate postoperative period was complicated by shock, return to the OR on postoperative day 1 for aortic arch reconstruction, and ECMO after this second surgery. His ECMO course was complicated by significant bleeding, including significant hemorrhagic chest tube output as well as bleeding from ECMO cannula sites. On postoperative day 7, he was found to have a large left ventricular thrombus and ultimately arrested and died on postoperative day 8. He did not have a hypercoaguable evaluation performed due to being anticoagulated and on ECMO. The other control who had thrombotic complications also had d-transposition of the great arteries and underwent arterial switch. Her course was complicated by poor coronary blood flow in the OR requiring reopening sternotomy and exploration of the coronary arteries. Sternal closure was on postoperative day 3, and follow-up echocardiogram the next day revealed a large right atrial thrombus. She did not have a hypercoaguable evaluation but was started on heparin and transitioned to coumadin for discharge. For patients who received rFVIIa, there were statistically significant decreases in total transfusion requirements through the remainder of the study period for all blood products after the dose (Fig 1). Median total packed red blood cells transfusion volume decreased from 38.4 mL/kg to 10.1 mL/kg, median platelet volume from 9.8 mL/kg to 0 mL/kg, median fresh frozen plasma volume from 13.1 mL/kg to 0 mL/kg, median cryoprecipitate volume from 3.6 mL/kg to 0 mL/kg, and median total combined transfusion volume from 77.7 mL/kg to 14.6 mL/kg (p ⬍ 0.001 for all comparisons except cryoprecipitate, p ⫽ 0.002). When compared with the subset of controls who bled significantly (22%), there was no difference in the total transfusion requirement for patients receiving rFVIIa (median total transfusion volume 93.2 mL/kg for rFVIIa vs 108.3 mL/kg for controls; p ⫽ 0.225) (Fig 2). Hemorrhagic chest tube output was assessed as the main hemorrhagic endpoint. Average hourly chest tube output was recorded for the pre-rFVIIa period (variable duration from the end of the operative procedure to rFVIIa dosage), the 12 hours immediately post-rFVIIa, and the first 12 postoperative hours (control group). Patients who received rFVIIa intraoperatively were excluded from this analysis. Chest tube output for the rFVIIa group decreased dramatically, with output quickly falling to the level of nonbleeding controls (Fig 3). Mean hemorrhagic chest tube output prior to rFVIIa was 8.3 (⫾ 1.6) mL/kg/hour, mean output for the first 12 hours after rFVIIa was 1.1 (⫾ 0.3) mL/kg/hour, and mean output for the first 12 hours postoperative for controls was 1.1 (⫾ 0.1) mL/kg/hour (p ⬍ 0.001 for prerFVIIa vs post-rFVIIa; p ⫽ 0.002 for pre-FVIIa vs controls; p ⫽ 0.827 for post-rFVIIa vs control).

Comment In this single institution retrospective study, rFVIIa therapy was not associated with increased thrombotic adverse events when used as rescue therapy for bleeding after congenital heart disease surgery. As in previous

Ann Thorac Surg 2010;89:570 –7

Fig 3. Comparison of hemorrhagic chest tube output after congenital heart disease surgery in patients who received recombinant activated factor seven (rFVIIa) (both predose as well as the first 12 hours after the dose) and matched controls (first 12 hours postoperative). (*p ⬍ 0.001; **p ⫽ 0.002; ***p ⫽ 0.827; bars displayed as mean ⫾ standard error of the mean).

studies, rFVIIa did appear to improve hemorrhage control. It is becoming clear from multiple institutional case series and studies that rFVIIa is effective for the control of postsurgical hemorrhage in CHD surgery patients [1, 11–13, 19]. Concern remains that this treatment may be associated with thrombotic adverse events [19, 20]. This study attempts to address this concern systematically by matching cases for known CHD-associated thrombosis risk: age, diagnosis, and repair. There are several limitations to this study, including its small sample size which significantly limits our ability to assess a difference in thrombotic rate between the case and control groups. The thrombotic rate in the case group was twice that in the control group (8% vs 4%). In a larger sample this may become a statistically significant signal. Additionally, rFVIIa did not appear to increase the mortality rate, readmission rate, or length of stay. Again, these findings are limited by the small sample size. Furthermore, due to sampling limitations, we were not able to identify control patients who had severe bleeding in addition to matched thrombosis risk. Ideally a thorough analysis of both safety and efficacy would match for all four of these parameters in a large randomized sample. Although heparin-induced thrombocytopenia (HIT) is rare in children, it may be most common in those undergoing cardiac surgery [23, 24]. It is possible that HIT contributed to the development of thrombosis in some of these patients. However, none of the patients were tested for HIT. It would be prudent to include an analysis for HIT in any prospective study of rFVIIa in this population as a potential confounding thrombotic risk marker. As with HIT assays, there were not uniform coagulation studies in the case and control groups to help define either bleeding or thrombosis risk, and how rFVIIa treatment may or may not have altered these parameters. The secondary analysis of bleeding control was consistent with previous studies; however there were some im-

portant differences. First, we were able to demonstrate a significant decrease in hemorrhagic chest tube output and transfusion requirement after rFVIIa therapy. In contrast, Ekert and colleagues [20] did not demonstrate any effect on bleeding complications in their study. There are several possible explanations for this difference. First, the mean rFVIIa dose used was higher in our study so this difference in effect on bleeding could be dose-dependent. However, many studies have demonstrated a significant reduction in bleeding with rFVIIa doses of 30 to 40 mcg/kg, making this explanation less likely [16, 17, 19]. Also, our study included a higher proportion of patients undergoing high-risk procedures, including single-ventricle palliation. It may be that our population had a higher risk of bleeding complications and thus represented patients more likely to benefit from rFVIIa therapy. In this study no patient required surgical reexploration due to thrombosis, despite having a higher dosage than other similar studies that showed higher rates of thrombosis. For example, Agarwal and colleagues [19] reported a thrombosis rate of 25% in their study of rFVIIa in pediatric CHD surgery. Over 20% of their patients who received rFVIIa required surgical reexploration due to thrombosis, although their study was not designed to evaluate thrombotic risk. However, in this series, we did not have any patient who received rFVIIa and required postoperative ECMO, while in Agarwal and colleagues study 50% of the patients who received rFVIIa were on ECMO. Both studies had a similar breakdown of diagnoses and procedures so the reasons for the difference in ECMO rate are unclear and may just be reflective of institutional practice variation. Given the significant thrombotic risk associated with ECMO therapy, this may explain the higher rate of thrombotic complications seen in their study [25, 26]. While a formal cost analysis was not part of this study, we did evaluate ICU and hospital length of stay as surrogate markers for cost and found no difference between patients who received rFVIIa and those who did not. Ekert and colleagues [20] did evaluate the prophylactic use of rFVIIa in pediatric CHD surgery, but this was not done in a population with a high bleeding risk so it is difficult to draw conclusions from this study. In addition, our analysis did not evaluate transfusion complications, either short-term or long-term, so we cannot comment about their impact on ultimate cost. Furthermore, implementing early rFVIIa therapy as has been advocated in some trauma literature or even using rFVIIa as initial therapy for postoperative bleeding may dramatically reduce or eliminate the need for transfusions altogether [27]. Additional studies are needed to evaluate the effect of early rFVIIa use after CHD surgery. Prospective studies are needed to determine the optimal dosing, duration of therapy, and criteria for use after CHD surgery. In summary, rFVIIa therapy did not increase thrombotic complications in this series of children, when used as rescue therapy after congenital heart disease surgery, but does appear to decrease bleeding complications in this population. Prospective studies are needed to fully explore the safety and efficacy of rFVIIa in the CHD surgery population.

KARSIES ET AL PEDIATRIC RECOMBINANT FACTOR VII THERAPY

575

Special thanks to Vince Olshove for his assistance with the intraoperative and perfusion records, to Katherine McGlumphy for her help in obtaining transfusion records, and to Jennifer McCormack for her assistance in chart reviews.

References 1. Razon Y, Erez E, Vidne B, Katz J, Tamari H, Dagan O. Recombinant factor VIIa (NovoSeven®) as a hemostatic agent after surgery for congenital heart disease. Paediatr Anaesth 2005;15:235– 40. 2. Szekely A, Cserep Z, Sapi E, et al. Risks and predictors of blood transfusion in pediatric patients undergoing open heart operation. Ann Thorac Surg 2009;87:187–97. 3. The Society of Thoracic Surgeons Blood Conservation Guidelines Task Force. Perioperative blood transfusion and blood conservation in cardiac surgery: the Society of Thoracic Surgeons and the Society of Cardiovascular Anesthesiologists clinical practice guideline. Ann Thorac Surg 83(5 suppl):S27– 86. 4. Triulzi D. Transfusion-related acute lung injury: current concepts for the clinician. Anesth Analg 2009;108:770 – 6. 5. Odegard K, McGowan F, Zurakowski D, et al. Coagulation factor abnormalities in patients with single-ventricle physiology immediately prior to the Fontan procedure. Ann Thorac Surg 2002;73:1770 –7. 6. Kaulitz R, Ziemer G, Rauch R, et al. Prophylaxis of thromboembolic complications after the Fontan operation (total cavopulmonary anastamosis). J Thorac Cardiovasc Surg 2005;129:569 –75. 7. Binotto M, Maeda N, Lopes A. Evidence of endothelial dysfunction in patients with functionally univentricular physiology before completion of the Fontan operation. Cardiol Young 2005;15:26 –30. 8. Procelewska M, Kolcz J, Januszewska K, Mroczek T, Malec E. Coagulation abnormalities and liver function after hemiFontan and Fontan procedures—the importance of hemodynamics in the early postoperative period. Eur J Cardiothorac Surg 2007;31:866 –72. 9. Odegard K, Zurakowski D, Hornykewycz S, et al. Evaluation of the coagulation system in children with two-ventricle congenital heart disease. Ann Thorac Surg 2007;83:1797– 804. 10. Martinowitz U, Michaelson M on behalf of the Israeli Multidisciplinary rFVIIa Task Force. Guidelines for the use of recombinant activated factor VII (rFVIIa) in uncontrolled bleeding: a report by the Israeli Multidisciplinary rFVIIa Task Force. J Thromb Haemost 2005;3:640 – 8. 11. Karkouti K, Beattie WS, Wijeysundera DN, et al. Recombinant factor VIIa for intractable blood loss after cardiac surgery: a propensity score-matched case-control analysis. Transfusion 2005;45:26 –34. 12. Egan J, Lammi A, Schell D, Gillis J, Nunn G. Recombinant activated factor VII in paediatric cardiac surgery. Intensive Care Med 2004;30:682–5. 13. Tobias J, Sinsic J, Weinstein S, Schechter W, Kartha V, Michler R. Recombinant factor VIIa to control excessive bleeding following surgery for congenital heart disease in pediatric patients. J Intensive Care Med 2004;19:270 –3. 14. Franchinia M, Manzatob F, Salvagnoc GL, Lippic G. Potential role of recombinant activated factor VII for the treatment of severe bleeding associated with disseminated intravascular coagulation: a systematic review. Blood Coagul Fibrinolysis 2007;18:589 –93. 15. Sachs B, Delacy D, Green J, et al. Recombinant activated factor VII in spinal surgery: a multicenter, randomized, double-blind, placebo-controlled, dose-escalation trial. Spine 2007;32:2285–93. 16. Mayer S, Brun N, Begrup K, et al. Recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med 2005;352:777– 85.

PEDIATRIC CARDIAC

Ann Thorac Surg 2010;89:570 –7

576

KARSIES ET AL PEDIATRIC RECOMBINANT FACTOR VII THERAPY

PEDIATRIC CARDIAC

17. Pugliese F, Ruberto F, Summonti D, et al. Activated recombinant factor VII in orthotopic liver transplantation. Transplant Proc 2007;39:1883–5. 18. Rizoli S, Boffard K, Riou B, et al. Recombinant activated factor VII as an adjunctive therapy for bleeding control in severe trauma patients with coagulopathy: subgroup analysis from two randomized trials. Crit Care 2006;10:R178. 19. Agarwal H, Bennett J, Churchwell K, et al. Recombinant factor seven therapy for postoperative bleeding in neonatal and pediatric cardiac surgery. Ann Thorac Surg 2007;84:161–9. 20. Ekert H, Brizard C, Eyers R, Cochrane A, Henning R. Elective administration in infants of low dose recombinant activated factor VII (rFVIIa) in cardiopulmonary bypass surgery for congenital heart disease does not shorten time to chest closure or reduce blood loss and need for transfusions: a randomized, double-blind, parallel group, placebo-controlled study of rFVIIa and standard haemostatic replacement therapy versus standard haemostatic replacement therapy. Blood Coagul Fibrinolysis 2006;17:389 –95. 21. Jenkins K, Gauvreau K, Newburger J, Spray T, Moller J, Iezzoni L. Consensus-based method for risk adjustment for surgery for congenital heart disease. J Thorac Cardiovasc Surg 2002;123:110 – 8.

Ann Thorac Surg 2010;89:570 –7

22. Al-Radi O, Harrell F, Calderone C, et al. Case complexity scores in congenital heart surgery: a comparative study of the Aristotle basic complexity score and the risk adjustment in congenital heart surgery (RACHS-1) system. J Thorac Cardiovasc Surg 2007;133:865–74. 23. Mullen P, Wessel D, Thomas K, et al. The incidence and implications of anti-heparin-platelet factor 4 antibody formation in a pediatric cardiac surgical population. Anesth Analg 2008;107:371– 8. 24. Schmugge M, Risch L, Huber A, Benn A, Fischer J. Heparininduced thrombocytopenia-associated thrombosis in pediatric intensive care patients. Pediatrics 2002;109:e10. 25. Muntean W. Coagulation and anticoagulation in extracorporeal membrane oxygenation. Artific Organs 1999;23:979 – 83. 26. Riccabona M, Kuttnig-Haim M, Dacar D, et al. Venous thrombosis in and after extracorporeal membrane oxygenation: detection and follow-up by color Doppler sonography. Eur Radiol 1997;7:1383– 6. 27. Spinella P, Perkins J, McLaughlin D, et al. The effect of recombinant activated factor VII on mortality in combatrelated casualties with severe trauma and massive transfusion. J Trauma 2008;64:286 –94.

INVITED COMMENTARY When deciding on a therapeutic agent, a physician is essentially performing a risk-benefit analysis. The decision-making process becomes more complicated in cases of off-label use because physicians find themselves in a precarious territory. Although off-label uses are sometimes the result of the lengthy regulatory process despite published data supporting a favorable risk-benefit ratio, unfortunately, many off-label uses lack the safety profile inherent in the regulatory approval. In off-label use, the risks or benefits are generally “perceived” rather than “established,” and thus, they are prone to optimism bias, which is a general tendency to be over-optimistic about the outcomes of the planned action. As such, growing off-label use of recombinant activated factor VII (rFVIIa) as a hemostatic agent in general patient populations warrants extra caution. In their study, Karsies and colleagues [1] evaluated some measures of safety and efficacy of rFVIIa in bleeding pediatric patients undergoing congenital heart disease (CHD) operations in their institution, with special focus on thrombotic events. The investigators compared 25 CHD surgical patients who received an average of 70 ␮g/kg rFVIIa with 50 patients matched for their thrombosis risk who did not receive rFVIIa. Although transfusion volume and chest tube drainage were significantly lower in the rFVIIa-treated patients, the difference in the incidence of thrombotic events between the rFVIIa-treated (8%) and untreated (4%) patients did not reach statistical significance. For clarification, although this is branded as a “casecontrol” study, the design conforms to that of a cohort study in which the study arms are defined by exposure (receiving rFVIIa) rather than the outcome (developing thrombotic complications). This design in conjunction with the relative low incidence of the primary outcome (only 2 patients with thrombotic complications in each arm) leads to the main limitation of the study: The sample size is too small, rendering the study essentially

underpowered to reliably draw a conclusion on the safety of rFVIIa in this context. Unfortunately, a true casecontrol design in which patients who presented with thrombotic complications are compared with a matched group of patients who did not have these complications is still unlikely to be adequately powered to unveil the role of rFVIIa due to the small proportion of patients receiving rFVIIa in this population. Another concern is the comparability of the study arms: The study groups were matched only for the known risk factors of thrombotic complications, being age, diagnosis, and type of procedure. As a result, although all rFVIIatreated patients had substantial bleeding, only one-fifth of the control arm experienced significant bleeding complications. Moreover, the provided data indicate that the study arms differed in certain other baseline aspects, and lack of statistical significance here is not completely reassuring given the small sample size. Although the authors argue that sampling limitations prevented them from identifying a control group adequately matched for bleeding, the rather large number of 1165 cases who did not receive rFVIIa and the fact that the authors managed to match 2 controls for each rFVIIa-treated case suggest that other more adequate matching schemes (perhaps even in 1:1 ratio) might have been worth considering. In conclusion, this study undoubtedly adds valuable data to the currently limited (and case-report dominated) information available on the off-label use of rFVIIa to control bleeding in general and in children undergoing operations for CHD in particular. The reported effect of the treatment on reducing transfusion volume and blood loss is impressive and indicates great potentials for improving patients’ outcomes. Nonetheless, the safety issue remains unresolved, because the current study demonstrates a higher (albeit statistically nonsignificant) incidence of thrombotic complications in rFVIIa-treated patients. It is up to future, better-powered studies to