| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Click here to read this article as a CME activity
a Royal Prince Alfred Hospital, Sydney, New South Wales, Australia
b Monash University Department of Epidemiology and Preventive Medicine, Melbourne, Victoria, Australia
c The Austin Hospital, Melbourne, Victoria, Australia
d Royal North Shore Hospital, Sydney, New South Wales, Australia
e Prince of Wales Hospital, Sydney, New South Wales, Australia
f Novo Nordisk Pharmaceuticals Pty Ltd, Baulkham Hills, New South Wales, Australia
g Emergency and Trauma Centre, The Alfred Hospital, Melbourne, Victoria, Australia
Accepted for publication June 26, 2007.
* Address correspondence to Dr Phillips, Monash University, Epidemiology and Preventive Medicine, Haemostasis Registry, Alfred Hospital, Melbourne, Victoria 3004, Australia (Email: louise.phillips{at}med.monash.edu.au).
| Adult cardiac surgery:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
Methods: The ANZHR collects retrospective and contemporaneous data on the use of rFVIIa in patients with critical bleeding from hospitals throughout Australia and New Zealand. Participating centers commit to the collection of data on all patients without hemophilia treated with rFVIIa, which limits bias and prevents the reporting of only positive or anecdotal experiences.
Results: At September 2006, the cardiac surgical cohort comprised 304 patients (43%) of a total of 695 cases reported to the ANZHR from 46 hospitals. The 304 cases date from January 2001. The median patient age was 66 years (interquartile range [IQR], 53 to 75 years), and 73% were men. After administration of rFVIIa, all blood product usage was significantly reduced. Patients received a median dose of 93 µg/kg (IQR, 82 to 102), and 85% of patients received a single dose. The documented response rate to a single dose of rFVIIa was 84%, of which 23% reported cessation of bleeding and 61% reported a reduction in bleeding. Patients received a median volume of 6 U of red blood cells before rFVIIa treatment. The median reduction in red blood cells after the rFVIIa dose compared with before was 4 U. Response was reduced in patients with a lower baseline hemoglobin, coagulopathy (determined by international normalized ratio, fibrinogen, and platelets), the number of red blood cell units transfused before rFVIIa, advanced age, more complex operations, hypothermia, and acidosis. Responders had a significantly reduced mortality (p < 0.001). The percentage of patients alive at 28 days was 95% if bleeding ceased after rFVIIa, 86% if bleeding reduced, and 60% for nonresponders. A 7% adverse event rate attributed as "probably" or "possibly" associated with rFVIIa was reported with a 4% reported thromboembolic event rate.
Conclusions: Recombinant FVIIa is a potential rescue therapy in severe uncontrollable critical bleeding after cardiac operations. The observed response rate was high, and response was associated with improved mortality. There was an observed reduction in blood product usage after rFVIIa. The adverse event rate reported was similar to documented adverse event rates in complex cardiac surgical patients. In the absence of randomized controlled trials, this registry provides a basis for understanding current clinical practice with rFVIIa in cardiac surgical procedures.
| Mr Jankelowitz discloses that he has a financial relationship with Novo Nordisk Pharmaceuticals Pty Ltd.
|
| This article has been selected for the open discussion forum on the CTSNet Web Site: http://www.ctsnet.org/sections/newsandviews/discussions/index.html
|
Recombinant activated factor VII (rFVIIa; NovoSeven, Novo Nordisk, Bagsvaerd, Denmark) is effective for the prevention or treatment of bleeding in patients with coagulation inhibitors [1–3]. In Australia and New Zealand, rFVIIa is licensed for this indication and congenital factor VII deficiency. It has been extensively used amongst patients with hemophilia [2, 3]. Its mechanism of action is complex, but the supraphysiologic levels of FVIIa can act in a tissue factor (TF)–dependent manner, forming TF:rFVIIa complexes at the site of injury, activating coagulation factors and platelets, and ultimately, the generation of a "burst" of thrombin and fibrin deposition, or to a lesser extent in a TF independent manner, by directly binding platelets and generating thrombin by direct activation of factor X [4, 5].
In recent years, an increasing number of case reports [6–8] and case series [9–15] have reported on the use of rFVIIa in salvage situations associated with persistent bleeding in cardiac operations. In many cases, control of critical bleeding is possible with surgical hemostasis and the replacement of blood products to support the underlying coagulopathy. In a small number of cases (< 3.5%) [16], this is not possible despite appropriate surgical intervention.
Patients with uncontrolled critical bleeding and coagulopathy have significant mortality, despite standard replacement of coagulation factors and surgical intervention [17, 18]. The transfusion of blood and blood products has been correlated with increased morbidity, including multiorgan failure, acute respiratory distress syndrome (ARDS), transfusion-related acute lung injury (TRALI), immune modulation, and infection [19, 20], as well as death [21]. This cohort of patients is often associated with increased costs to the health care system in the form of excessive blood and blood products and extended stays in the intensive care unit (ICU) and hospital [22, 23].
To date, reports of use of rFVIIa in cardiac surgical procedures are mostly case reports or series or retrospective chart reviews [24]. Only two randomized controlled studies have been published in cardiac surgery. The first was a pilot study that explored prophylactic use in adults undergoing complex cardiac operations [25], and the second examined prophylactic use in neonates [26]. Novo Nordisk is currently conducting a phase 2 randomized controlled study in adults, and an investigator-led study investigating salvage use of rFVIIa in complex cardiac operations is also ongoing.
Groups from both the United States [27] and Europe [28] have published recommendations on the use of rFVIIa in perioperative coagulopathic bleeding in cardiothoracic surgical procedures. Although some individual centers in Australia have described guidelines that are specific for cardiac surgery [14], consensus guidelines in Australia focus on the surgical management of persistent hemorrhage, the medical management of coagulopathy, and the salvage use of rFVIIa in general and are not specific for cardiac operations [29, 30]. Inadequate hemostasis in cardiac operations is inherently different (eg, due to cardiopulmonary bypass and the ability to directly observe the bleeding source) from other areas of critical bleeding, such as from trauma or obstetric bleeding, and thus requires a different approach [14].
The Australian and New Zealand Haemostasis Registry (ANZHR) was established by the Department of Epidemiology and Preventive Medicine at Monash University, Melbourne, Australia to collect information on patients without hemophilia treated with rFVIIa throughout Australia and New Zealand. The development of the ANZHR has enabled the systematic collection of data from a large cohort of patients undergoing cardiac operations complicated by critical bleeding and treated with rFVIIa. This article describes the safety, indications, dosing, reported efficacy, mortality and outcomes from cardiac surgical cases included in the Haemostasis Registry in the period January 2001 to September 2006 from institutions across Australia and New Zealand.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Patients
The Haemostasis Registry began receiving data in May 2005 but includes cases dating back to January 2001. Forty-six hospitals from all states and territories of Australia and from New Zealand are currently contributing data to the registry, of which there are 21 cardiac surgery centers. Participating hospitals undertake to provide information to the registry on all patients without hemophilia treated with rFVIIa at that hospital, and compliance is audited to ensure that complete data sets are received. Local investigators identify eligible patients after treatment with rFVIIa through pharmacy or blood bank records, depending on the protocol at each hospital. Patients were included in this study if the primary cause of bleeding was associated with cardiac operations as reported by the local investigator. In this registry, "cardiac surgery" includes all surgical procedures on the heart or on major thoracic vessels requiring cardiopulmonary bypass.
Data
Standardized data forms were completed by trained ANZHR coordinators at each center. The following data were obtained for the Haemostasis Registry:
In addition, prescribing clinicians were asked to assess patient response by indicating whether bleeding had stopped or decreased (responders), or was unchanged or had increased (nonresponders) after the use of rFVIIa. Clinicians were also asked to report all adverse events within 28 days of treatment with rFVIIa and to give an assessment of the probability that these events were linked to the administration of rFVIIa by assigning one of the following relationships: not linked, unlikely to be linked, possibly linked, probably linked, or definitely linked. The ANZHR data collection forms and the Data Dictionary containing definitions are available at the Haemostatis Registry Web site (www.med.monash.edu.au/epidemiology/traumaepi/haemostasis.html).
Statistical Analysis
Because almost none of the variables measured conformed to normal distributions, nonparametric statistics were used throughout for consistency of reporting and analysis. Coagulation indicators and blood product usage before and after rFVIIa dose were compared in individuals by using Wilcoxon matched-pairs signed ranks tests. Univariate and multivariate analyses were performed to assess the relationship between the outcomes of interest (response and mortality, or both) and age, sex, patient weight, size of dose, temperature, pH, hemoglobin, hematocrit, platelet and fibrinogen levels, PT/INR, APTT, type of operation; number of RBCs, FFP, cryoprecipitate, and platelet units before dose; time to dose from bleeding onset, and place of administration. Where necessary, continuous variables were categorized according to clinical indicators. Analysis by 
2 was used to identify associations between categoric baseline variables and the outcomes. Multivariate analysis was used to identify significant associations with each outcome. A binary logistic regression procedure was used. Variables demonstrating a univariate association of p 
0.10 with either outcome were included in the multivariate models. For each model, the odds ratio and 95% confidence interval (CI) are provided.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Demographics
Most patients were male (73%). The median age of patients was 66 years (range, 0 to 89). Nine patients were younger than 17 years old and for this study were excluded from all subsequent analyses. The mean (SD) age of the remaining patients was 63 (15) years. Most patients were treated at hospitals in the states of New South Wales (47%) or Victoria (45%).
Type of Operation
A description of the types of cardiac operations involved in registry cases is summarized in Table 1. A total of 48 (15%) patients had redo operations.
|
Dose
The 293 patients received 343 doses of rFVIIa: 6 patients received three consecutive doses, 38 patients received two doses, and the remaining patients (85%) receiving a single dose. The median dose was 92.3 µg/kg (interquartile range [IQR], 82 to 103 µg/kg; range, 9 to 141 µg/kg). The median time from onset of bleeding to dose administration was 4 hours (IQR, 1 to 8 hours; range, 0 to 66 hours). Before treatment with rFVIIa, 35% of patients were returned to the operating theater in an attempt to control blood loss, including 12 patients who were returned more than once before treatment with rFVIIa. After treatment with an initial dose of rFVIIa, 35 patients were returned to the operating theater for bleeding control.
Blood Products
Transfusion of all blood products was significantly reduced (Table 2). A comparison of units of RBCs received before and after the initial dose of rFVIIa in each individual showed a significant reduction (p < 0.001). Similar significant reductions were seen in individuals in the amount of FFP (p < 0.001), platelets (p < 0.001), cryoprecipitate (p < 0.001), and total blood products (p < 0.001). After the administration of rFVIIa, most patients (88%) received less than 6 U of RBCs. The median reduction in RBC transfusion was 4 U (IQR, 0 to 8 U; range, –21 to 40).
|
Temperature and pH
Temperature was not recorded for 83 patients at the time of the initial rFVIIa dose. Of the remaining 210 cases, 36 (17%) were hypothermic (temperature < 35.0°C). The pH was not documented for 84 patients at the time of the initial rFVIIa dose. Of the remaining 209 patients, 109 (52%) were acidotic (pH < 7.35) at the time of the initial rFVIIa dose, including 18 patients (10%) who were moderately acidotic (pH, 7.2 to 7.05) and 2 patients who were severely acidotic (pH < 7.05).
Laboratory Indicators
Details of coagulation indicators before and after the initial dose of rFVIIa are provided in Table 3. Prothrombin time was prolonged (>15.5 seconds) in 156 of 214 patients (73%) before administration of rFVIIa. Similarly, INR was prolonged (>1.5) in 118 of 235 patients (50%) before administration of rFVIIa.
|
Response and Outcome
Recombinant FVIIa was considered to have decreased or stopped bleeding in 245 of 297 (82%) of uses where efficacy was reported. Of the 256 first doses of rFVIIa where efficacy was reported, 215 doses (84%) resulted in the reduction or cessation of bleeding. The reported response rate for the second rFVIIa dose was 25 of 35 doses (71%) and for the third dose, 5 of 6 patients (83%) responded to treatment. In the 41 patients reported as not responding to the initial dose of rFVIIa, 6 later had surgical bleeding identified in a return to the operating room.
Patients whose bleeding was considered to have stopped or decreased were less likely to die than those whose bleeding was unchanged (2
2 = 22.3; p < 0.001, see Figure 1). Patients undergoing re-do operations showed no significant difference in the rate of response to rFVIIa, but were more likely to die (OR, 2.1; 95% CI, 1.0 to 4.4).
|
Adverse Events
Details of adverse events are given in Table 4. A total of 161 adverse events were reported in 130 (44%) patients within 28 days of receiving rFVIIa. Of these 161, 127 (79%) were considered to be unlikely to be linked (n = 53) or not linked (n = 74) to the administration of rFVIIa. A total of 22 (7%) were considered to be either probably or possibly linked to the administration of rFVIIa, of which 13 were thromboembolic in nature. No patients were considered to have died of adverse events related to the use of rFVIIa.
|
Mortality demonstrated an association with female sex (p = 0.020), low pH (p < 0.001), type of cardiac operation (other type vs CABG or valve, or both; p = 0.011), place of administration (operating theater vs ICU; p = 0.096), and high APTT (> 80 seconds; p = 0.004).
Multivariate Analyses
Variables showing a univariate association (p < 0.1) with either of the outcome variables were entered into the multivariate analyses, the results of which are provided in Table 5. Hematocrit level was not included because it showed a strong colinearity with hemoglobin level (Spearman correlation p = 0.981). Sex, age, hemoglobin level, category of cardiac operation, and pH were all found to be associated with death when taking into account place of administration, PT/INR, APTT, and number of RBC units. Women were more likely to die (OR, 3.23; 95% CI, 1.05 to 9.90) than men. Patients aged 75 years or older were more likely to die than those younger than 35 years (OR, 35.4, 95% CI, 2.49 to 503.74). Patients with a pH < 7.20 were significantly more likely to die (OR, 7.33, 95% CI, 1.61 to 33.55) than those with higher pH. Patients undergoing operations in the "other" category were more likely to die than those undergoing isolated CABG (OR, 7.03, 95% CI, 1.04 to 47.41).
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
One of the real difficulties with this project was the lack of truly objective measures of efficacy that are applicable to the wide variety of situations in which rFVIIa is used. Chest tube output, although applicable to those patients who have left the operating theater, is not universally applicable to all cardiac surgical cases. Return to the operating theater, likewise, is not a universal outcome measure, although returns to the operating theater may indicate some instances of lack of effect. The number of units of blood products is unreliable due to inconsistencies in practice across hospitals and because of issues such as refilling after bleeding has ceased. The measure of response we used is subjective, but it is highly reproducible, and the strong correlation between this measure and mortality, although obvious, goes against any suggestion of systematic bias in the subjective measure of efficacy used.
Incomplete data records remain an issue in some areas; however, all patients had data on outcomes, blood products, dosage information, age, and sex. The percentage of missing data values ranges from 0% to 48% (fibrinogen level after dose 1) and was particularly evident in laboratory results: 31 patients had no laboratory results for the period after treatment with rFVIIa. Some hospitals regularly report only PT or INR (and not both). Temperature and pH at the time of dose were missing in 28% and 29% of cases, respectively. Laboratory results were not taken at all after rFVIIa treatment in some patients, particularly if bleeding had ceased or if the patient had died or was close to death.
The patients recorded in the Haemostasis Registry are reflective of typical cardiac surgical patients, frequently older men, who are undergoing a wide range of cardiac procedures ranging from simple revascularization to more complex valvular and aortic procedures. Before the use of rFVIIa, patients had received blood product support in an attempt to control hemorrhage, and 102 (35%) had already had surgical reexploration.
This group had a high response rate of 84% and, more important, responders had an improved mortality. The bleeding response rate after cardiac operations is higher than that seen in other critical bleeding settings (data submitted for publication), emphasizing the potential value of rFVIIa in this group. Why this is so is open to speculation.
Most patients received a single rFVIIa dose of about 90 µg/kg, consistent with the most commonly used dose in patients with hemophilia with inhibitors. No specific guidelines exist in Australia for dosing in bleeding after cardiac operations. Given the heterogeneity in cases, treatment protocols, blood products, and the triggers for treatment, it is difficult to compare outcomes among patient groups. The patients who received rFVIIa before RBC transfusion are an interesting group, but the lack of difference between this group and the cardiac group in general probably simply reflects a growing awareness of the importance of early control of excessive hemorrhage.
Recent evidence has suggested efficacy with lower doses in bleeding after cardiac operations [31, 32], but the number of patients who received lower doses in our sample was too small to provide meaningful data. Our findings demonstrate a statistically significant decrease in the number of all types of blood products received after the administration of rFVIIa.
It has been shown both theoretically [33] and clinically in this study, that low pH is associated with a reduction in response to rFVIIa. Ideally, rFVIIa should be given before patients become acidotic. Further work is necessary to enable prediction of which patients are likely to become acidotic and to determine whether a reversal of acidosis relates to improved response to rFVIIa.
An associated adverse event rate was reported in 7% of patients, of which 4% were thromboembolic in nature. This event rate is comparable with a recent systematic review of rFVIIa in cardiac surgery [24] and also to the adverse event rate reported in a meta-analysis of the first 13 Novo Nordisk–sponsored randomized controlled trials with rFVIIa in coagulopathic bleeding [34]. In addition, the adverse event rates do not appear higher than those reported in other complex cardiac surgical procedures that require significant transfusion [24, 35] where rFVIIa was not used. Of importance was that no cases of microvascular graft stenosis occurred, a hypothetical complication that has been of concern to clinicians [36].
This large case series has limitations, which include the absence of a control group and the heterogeneity of bleeding contexts, but is strengthened by the intention to capture all patients at participating hospitals and data collection by a nonclinical independent research group.
Given the nature of the database, firm conclusions regarding patient selection, dosing guidelines, and adjuvant treatment cannot be drawn. Response to rFVIIa was improved in patients who had received less RBC transfusion. This study showed reduced pH was associated with death, but whether this identifies a poor risk group or highlights the need to attempt to normalize pH before treatment with rFVIIa is unclear.
In the absence of large, international, randomized controlled trials, this observational study provides insights into the current clinical practice of rFVIIa use in critical bleeding in cardiac surgery.
Participating Hospitals
Australia—Australian Capital Territory: The Canberra Hospital; New South Wales: Liverpool Hospital, Prince of Wales Hospital, Royal North Shore Hospital, Royal Prince Alfred Hospital, St Vincent’s Private Hospital, and St Vincent’s Public Hospital; Queensland: Princess Alexandra Hospital, and The Prince Charles Hospital; South Australia: Flinders Medical Centre; Tasmania: Royal Hobart Hospital; Victoria: Alfred Hospital, Austin Hospital, Geelong Hospital, Knox Private Hospital, Monash Medical Centre, and St Vincent’s Hospital; Western Australia: Fremantle Hospital and Sir Charles Gairdner Hospital.
New Zealand—Auckland City Hospital and Wellington Hospital.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Tatoulis, S. Theodore, M. Meswani, R. Wynne, C. Hon-Yap, and N. Powar Safe use of recombinant activated factor VIIa for recalcitrant postoperative haemorrhage in cardiac surgery Interactive CardioVascular and Thoracic Surgery, September 1, 2009; 9(3): 459 - 462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Baker Suction, Salvage, Sutures, and Potions: Blood Management Post-Aprotinin Seminars in Cardiothoracic and Vascular Anesthesia, June 1, 2009; 13(2): 122 - 126. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Shander and M. Javidroozi Invited Commentary Ann. Thorac. Surg., March 1, 2008; 85(3): 844 - 844. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
