- Email: [email protected]
Recombinant Activated Factor VII Increases Stroke in Cardiac Surgery: A Meta-analysis
Recombinant Activated Factor VII Increases Stroke in Cardiac Surgery: A Meta-analysis Martin Ponschab, MD,* Giovanni Landoni, MD,* Giuseppe Biondi-Zoccai, MD,† Elena Bignami, MD,* Elena Frati, MD,* Davide Nicolotti, MD,* Fabrizio Monaco, MD,* Federico Pappalardo, MD,* and Alberto Zangrillo, MD* Objectives: Recombinant activated factor VII (rFVIIa) is used in various surgical procedures to reduce the incidence of major blood loss and the need for re-exploration. Few clinical trials have investigated rFVIIa in cardiac surgery. The authors performed a meta-analysis focusing on the rate of stroke and surgical re-exploration. Design: Meta-analysis. Setting: Hospitals. Participants: A total of 470 patients. Interventions: None. Measurements and Main Results: Four investigators independently searched PubMed and conference proceedings including backward snowballing (ie, scanning of reference of retrieved articles and pertinent reviews) and contacted international experts. A total of 470 patients (254 receiving rFVIIa and 216 controls) from 6 clinical trials (2 randomized, 3 propensity matched, and 1 case matched) were included in the analysis. The use of rFVIIa was associated with an in-
creased rate of stroke (12/254 [4.7%] in the rFVIIa group v 2/216 [0.9%] in the control arm, odds ratio [OR] ⴝ 3.69 [1.1-12.38], p ⴝ 0.03) with a nonsignificant reduction in rate of surgical re-exploration (13% v 42% [OR ⴝ 0.27 (0.04-1.9), p ⴝ 0.19]). The authors observed a trend toward an increase of overall perioperative thromboembolic events (19/254 [7.5%] in the rFVIIa group v 10/216 [5.6%] in the control arm [OR ⴝ 1.84 (0.82-4.09), p ⴝ 0.14]). No difference in the rate of death was observed. Conclusions: The administration of rFVIIa in cardiac surgery patients could result in a significant increase of stroke with a trend toward a reduction of the need for surgical re-exploration. The authors do not recommend routine use in cardiac surgery patients. rFVIIa may be considered with caution in patients with refractory life-threatening bleeding. © 2011 Elsevier Inc. All rights reserved.
C
et al, safety concerns already had been raised, but the literature was inconclusive and did not focus on adverse effects in the area of cerebrovascular events like stroke. These references out of various clinical settings in conjunction with the previous results of Zangrillo et al12 led the authors to the decision to scrupulously review and scan scientific articles on this topic. A rising number of references and notices in literature currently being published showed the necessity to conduct a systematic review of data pooled from existing trials limited to the cardiac surgery setting. Considering the hypothesis that the potential beneficial hemostatic effects could be counterbalanced by adverse effects, the authors performed a meta-analysis and comprehensive review of the current literature to determine the potential impact of rFVIIa on surgical re-exploration and the risk of adverse arterial vascular effects, namely stroke complications.
OAGULATION FACTORS like recombinant activated factor VII (rFVIIa) (NovoSeven; Novo Nordisk A/S, Bagsvaerd, Denmark) recently gained special interest because they interact in their natural physiologic environment of coagulation cascade once being activated. As soon as the first reports on rFVIIa were published on the successful resuscitation and survival of combat victims, rFVIIa increasingly was used off-label in refractory bleeding associated with trauma1 and cardiac surgery.2-8 Because rFVIIa exerts its pharmacologic action by inducing thrombin generation on locally activated platelets and contributes to the formation of a stabilized fibrin clot at the site of vessel injury,9 safety concerns have been raised on the potential danger of severe thromboembolic events.10,11 Recently, safety and efficacy data published by Gill et al2 in a randomized controlled trial showed a reduced rate of patients undergoing surgical revision and a reduction in overall mediastinal drainage blood loss as well as transfusion requirements with a trend toward a higher incidence of critical severe adverse events. A meta-analysis published in 2009 by Zangrillo et al12 at the authors’ institution reported a nonsignificant reduction in surgical re-exploration in patients receiving rFVIIa and a nonsignificant increase in the rate of perioperative stroke in these patients. At the time of the publication of the work of Zangrillo
From the *Department of Anesthesia and Intensive Care, Università Vita-Salute San Raffaele, Milan, Italy; and †Division of Cardiology, University of Modena and Reggio Emilia, Modena, Italy. Address reprint requests to Giovanni Landoni, MD, Department of Cardiothoracic Anesthesia and Intensive Care, Istituto Scientifico San Raffaele, Via Olgettina 60, Milan 20132, Italy. E-mail: landoni. [email protected] © 2011 Elsevier Inc. All rights reserved. 1053-0770/2505-0010$36.00/0 doi:10.1053/j.jvca.2011.03.004 804
KEY WORDS: bleeding, cardiac surgery, FVIIa, recombinant factor VII, stroke, surgical revision, anesthesia
METHODS
Search Strategy Four trained investigators independently searched for clinical trials in PubMed (updated December 1, 2010). The search strategy was established by Biondi-Zoccai et al13 and is available in Appendix 1. In addition to the special search strategy mentioned earlier, conference proceedings (2008-2010) from the European Society of Anaesthesia, International Anesthesia Research Society, American Society of Anesthesiologists, European Society of Intensive Care Medicine, Society of Cardiovascular Anesthesiologists, European Association of Cardiothoracic Anaesthesiologists, International Society of Thrombosis and Haemostasis, and American College of Chest Physicians congresses were scanned in order to detect further abstracts or recent scientific articles of relevance. In addition, the authors systematically applied backward snowballing (ie, scanning of references of retrieved articles and pertinent reviews) and contacted international experts for further studies. No language restriction was enforced, and non–English-language articles were translated before further analysis.
Journal of Cardiothoracic and Vascular Anesthesia, Vol 25, No 5 (October), 2011: pp 804-810
RECOMBINANT ACTIVATED FACTOR VII
805
Table 1. Description of the 6 Studies Included in the Systematic Review on rFVIIa Versus Control With the Number of Patients and Interventions
Journal
Year
rFVIIa (g/kg) First Dose (Mean ⫾ Standard Deviation)
Diprose
Br J Anaesth
2005
90 (0)
No
Karkouti
Transfusion
2005
Tritapepe Von Heymann Gelsomino
Crit Care Med Crit Care Med Eur J Cardiothorac Surg
2007 2005 2008
Gill
Circulation
2009
7 pts: 62 (13) 44 pts: 37 (9) 70 (0) 60 (median) 18 (9-16) (median; interquartile) 35 pts: 40 69 pts: 80
First Author
No. of Patients Receiving a Further Dose
Cardiac Surgery Procedures
Study Design
No. of Patients in the rFVIIa Group
No. of Patients in the Control Group
Randomized
10
10
14
Complex noncoronary cardiac surgery Varied procedures
Propensity-matched
51
51
4 10 Nonspecified
Aortic dissection Varied procedures Varied procedures
Propensity-matched Case-matched Propensity-matched
23 26 40
23 24 40
No
Varied procedures
Randomized
104
68
254
216
Total Abbreviation: pts, patients.
Study Selection A total of 208 references obtained from database and literature searches first were examined independently at the title/abstract level by 4 investigators. Included studies had to compare the use of rFVIIa versus placebo or standard treatment; the authors only included scientific publications relating to adult cardiac surgical patients. Exclusion criteria were nonhuman experimental or animal studies, duplicate publications (in this case only the article reporting the longest follow-up was included), missing outcome data, and no case matching. In case of missing outcome data in studies, original authors were contacted and asked for additional information. Four investigators selected studies for the final analysis by independently assessing compliance to the selection criteria (Table 1).
Data Abstraction and Study Characteristics Baseline, procedural, and outcome data were abstracted independently by 4 investigators, with divergences resolved by consensus and are listed in Table 1. Specifically, the authors extracted study design, population, clinical setting, rFVIIa dosage, and treatment duration. At least 2 separate attempts at contacting original authors were made in case of missing data. The primary endpoint of this analysis was to determine the safety of rFVIIa in terms of stroke (as per author definition) and the efficacy of rFVIIa in terms of the reduction of surgical revision for bleeding. Secondary endpoints included death, the transfusion of blood products, and the incidence of overall venous and arterial vascular complications (ie, myocardial infarction, stroke, and deep venous thrombosis as per author definition).
Data Analysis and Synthesis Binary outcomes from individual studies were analyzed and pooled in order to compute individual odds ratios (ORs) with pertinent 95% confidence intervals according to the fixed-effect Peto method in case of homogenous data (inconsistency ⬍50%); and the random-effect DerSimonian-Laird method in case of statistical heterogeneity (inconsistency ⬍50%).13,14 Statistical heterogeneity (ie, the variability in effect estimates from study to study) and inconsistency (a more robust estimate of the variability in effect estimates from study to study, which is largely independent of the number of included studies) were measured using the Cochran Q test (which is largely comparable to the chi-square test, and, when yielding p values ⬍0.10, suggests significant statistical heterogeneity) and I2, respectively.15 According to Higgins et al,16 I2 values around 25%, 50%, and 75% were considered as representing low, moderate, and severe statistical inconsistency, respec-
tively. Specifically, OR ⬍1 suggests that the first comparator is associated with a reduced incidence of the event, whereas OR ⬎1 suggests that the 2nd comparator is associated with an increased incidence of the event. Unadjusted p values are reported throughout. The risk of smallstudy bias (including publication bias) was assessed by a visual inspection of funnel plots (ie, the graphic display of effect estimates and study precision, which enables appraisal of the potential asymmetry in metaanalysis datasets, often caused by the selective lack of publication of small negative studies, ie, publication bias).16 Statistical significance was set at the 2-tailed 0.05 level for hypothesis testing and at 0.10 for heterogeneity testing. Computations were performed with SPSS 11.0 (SPSS, Chicago, IL) and RevMan 4.2 (a freeware available from The Cochrane Collaboration).14 This study was performed in accordance with The Cochrane Collaboration and the Quality of Reporting of Meta-Analyses guidelines. RESULTS
Database searches, snowballing, and contacts with experts yielded a total of 208 citations. From these 208 citations, 187 nonpertinent titles or abstracts had to be excluded, and 21 studies were retrieved in complete form. A total of 15 studies were further excluded because they turned out to be review articles,17,18 they were noncase matched19 or had nonrandomized experimental design,20-27 they studied a pediatric population,28,29 they did not report outcome data,30 or duplicate publication.31 Six eligible clinical trials were identified (all of them recently published between 2005 and 2009), which were included in the final analysis2-7 (Table 1). Study Characteristics The 6 included clinical trials were comprised of 470 patients (254 receiving rFVIIa and 216 controls) (Table 1). One trial was conducted in the setting of noncoronary artery surgery,4 1 in aortic dissections,7 and the other 4 studies were involved in various cardiac surgery procedures.2,3,5,6 All authors studied the hemostatic properties of rFVIIa and reported on vascular complications and/or surgical re-exploration. rFVIIa dosage varied across studies (Table 1), between 18 and 70 g/kg being given in repeatable dosage and 90 g/kg used as a single dose. One study used rFVIIa prophylactically,4 whereas the other 5 used rFVIIa in patients with refractory bleeding (definitions of refractory bleeding varied among authors). Five studies reported a
806
PONSCHAB ET AL
Fig 1.
Individual and pooled estimates for stroke. (Color version of figure is available online.)
reduction in bleeding after rFVIIa administration,2-4,6,7 whereas 1 study5 proved that rFVIIa was safe but not incrementally efficacious over conventional hemostatic therapy. Some of the studies included in this meta-analysis appeared of poor internal validity, with 2 that were randomized2,3 and 1 with a multicenter design.2 A clear definition of stroke (new focal neurologic deficit, either transient but present ⬎24 hours or permanent), was present only in 1 study.2 Quantitative Data Synthesis Safety Overall analysis showed that the use of rFVIIa was associated with an increased rate of stroke (12/254 [4.7%] in the rFVIIa v 2/216 [0.9%] in the control arm (Fig 1), OR ⫽ 3.69 [1.1-12.38], p for effect ⫽ 0.03, p for heterogeneity ⫽ 0.92, I2 ⫽ 0% with 470 patients included in 6 studies, number needed to harm ⫽ 26). A nonsignificant increase was found in the rate of overall perioperative vascular events; the authors included myocardial infarction, stroke, and deep venous thrombosis as per author definition (19/254 [7.5%] in the rFVIIa v 10/216 [5.6%] in the control arm, OR ⫽ 1.84 [0.82-4.09], p for effect ⫽ 0.14, p for heterogeneity ⫽ 0.84, I2 ⫽ 0% with 470 included patients in 6 studies) (Fig 2).
Fig 2.
Because myocardial infarctions and the incidence of deep venous thrombosis did not show any statistically significant difference, it meant that the overall effect of the nonsignificant increase of arterial and venous vascular events was driven primarily by the increase in the events of stroke. In the 5 studies that provided this information, 5 of 214 (2.3%) myocardial infarctions were noted in the rFVIIa group versus 8 of 176 (4.5%) in the placebo group (OR ⫽ 0.59 [0.2-1.79], p for effect ⫽ 0.35, p for heterogeneity ⫽ 0.89, I2 ⫽ 0% with 390 patients included). Mortality was similar in the 2 groups (33/ 254 [13%] in the rFVIIa group and 26/216 [12%] in the control group [OR ⫽ 1.14 (0.65-2.01), p for effect ⫽ 0.65, p for heterogeneity ⫽ 0.88, I2 ⫽ 0% with 470 patients included in 5 studies]). Efficacy for Surgical Revision, Bleeding, and Blood Product Transfusion The present meta-analysis shows a nonsignificant reduction in the rate of surgical re-exploration (23/180 [13%] in the rFVIIa group v 59/142 [42%] in the control group [OR ⫽ 0.27 (0.04-1.9), p for effect ⫽ 0.19, p for heterogeneity ⬍0.001, I2 ⫽ 87% with 322 patients included in 4 studies]) (Fig 3). The authors cannot report on the length of intensive care or hospital
Individual and pooled estimates for thromboembolic events. (Color version of figure is available online.)
RECOMBINANT ACTIVATED FACTOR VII
Fig 3.
807
Individual and pooled estimates for surgical re-exploration. (Color version of figure is available online.)
stay because only 1 study provided sufficient data. The inspection of the funnel plots for all these endpoints did not disclose any evidence of small-study/publication bias, even if such appraisal appeared underpowered by the small set of studies and limited by differences in study design. All of the 6 studies included in this meta-analysis reported a reduced amount of blood loss or a reduction of blood product use but used incomparable performance data of measurements in their documentation. Four2,4-6 of the 6 trials reported on the reduction of blood transfusion requirements versus control, whereas 2 studies3,7 only reported on significant reductions in blood being transfused after the use of rFVIIa versus the time before the application of rFVIIa. Gill et al2 only documented the overall transfusion requirements, whereas the other 3 studies4-6 split the results on red blood cell units, platelet units, and fresh frozen plasma units being transfused. Reduced hourly blood loss was reported in 2 studies3,7 for the first hour before rFVIIa compared with blood loss during the first hour after rFVIIa, whereas Gelsomino et al6 reported on hourly bleeding over 24 hours after rFVIIa. Gelsomino et al6and von Heymann et al5 showed the overall blood loss over 24 hours after rFVIIa versus control, whereas Gill et al2 reported on 4-hour blood loss via mediastinal drainage versus control. Because of the different ways of documentation outlined previously, the treatment groups could not be compared with the control groups in matters of absolute numbers of blood product requirements or in matters of blood loss over a definite time. DISCUSSION
The present authors performed an updated meta-analysis of pooled data from 6 studies and showed that the use of rFVIIa in cardiac surgery was associated with a significant increase in the occurrence of perioperative stroke while presenting a nonsignificant reduction in the need for surgical re-exploration. Beyond that the authors established a comprehensive review of evidence of various clinical settings recently being published focusing on undesirable vascular events. Because the use of rFVIIa proved to interact as a coagulation-promoting pharmacologic substance, safety concerns were raised on the induction of the coagulation cascade causing severe vascular events. However, each study focusing on the
possible potential side effects after the application of rFVIIa failed to substantiate these objections. Although an extensive literature search was performed in this meta-analysis, the authors could only identify 6 randomized or propensity-matched or case-matched studies, included in the present data collection. Most publications on the use of rFVIIa in various surgical fields were performed as case series with the lack of a proper control group. The inclusion of these 6 studies allowed the present authors, for the first time, to identify a statistically significant increase in the occurrence of stroke after the administration of rFVIIa in the setting of cardiac surgery. In fact, some recent meta-analyses and reviews12,23,27,32 already suggested that the use of rFVIIa could be dangerous, but the results did not reach statistical significance, mainly because of the small sample size. The present results could be compared with the meta-analysis of Ranucci et al,33 which included 7 randomized controlled trials performed in major surgical procedures. In their study, no differences in terms of thromboembolic complications and mortality rate were observed, but the study did not focus on stroke events. In a subgroup analysis, Ranucci showed a significant effect of treatment in reducing the number of patients being exposed to allogeneic packed red blood cell transfusions only after a minimum dosage of 50 g/kg (137/211 [64.9%] v 108/158 [68.4%] in the control arm; fixed-effects OR, 0.43; 95% confidence interval, 0.23-0.78; p ⫽ 0.006). In contrast to this study, the present meta-analysis included only cardiac surgical patients. Reporting on a different study population of patients with spontaneous intracerebral hemorrhage, a recent trial by Diringer et al34 showed a dose-dependent significant increase of arterial thromboembolic events in the group of patients receiving 80 g/kg of rFVIIa (OR ⫽ 2.14, p ⫽ 0.031). The study group identified risk factors for arterial thromboembolism after the use of rFVIIa, namely increased age, signs of ischemia at baseline (cardiac or cerebral), and the prior use of platelet inhibitors. Another meta-analysis of 22 randomized controlled studies in nonhemophilic patients reported a reduction in the need for blood transfusion in patients receiving rFVIIa, with a trend toward an increased risk of arterial thrombosis.32 A recently published retrospective review of 69 patients receiving off-label rFVIIa identified 12 patients (17%) with
808
thromboembolic adverse events deemed to be related to drug exposure.27 Observed arterial thromboembolic adverse events consisted almost exclusively of ischemic stroke in mixed surgical approaches, mainly in cardiac surgery settings. A systematic review of rFVIIa in cardiac surgery comparing 6 studies including 298 patients was performed by Warren et al17 in 2008 and included 2 studies on pediatric surgical procedures, which is in contrast to the present study population in which pediatric surgical patients were excluded. The authors of this review reported a 2.6% incidence of stroke of a total of 5.3% thromboembolic events in the adult study group. This meta-analysis included 2 randomized controlled trials, 3 propensity-matched, and 1 case-matched study of a total of 470 patients presenting at the highest perioperative risk profile for adverse outcomes. In fact, this study provided the most comprehensive and thorough comparison of rFVIIa versus control in patients undergoing cardiac surgery that currently exists. Safety issues about the use of a procoagulant drug such as rFVIIa were raised repeatedly in the medical literature, and the debate of the use of rFVIIa in cardiac surgery was controversial apart from the indication of refractory bleeding.35,36 The main concern was that hemostatic drugs may induce thromboembolic events.10,11 In the present analysis, a significant increase in the occurrence of perioperative stroke was found, and pooled data showed a high level of heterogeneity throughout all studies being included. The finding of a significant increase of stroke was again confirmed by the results of Karkouti et al23 in a retrospective, non– case-matched study on 503 patients receiving rFVIIa, with a 12% incidence of stroke. An Australian study group retrospectively (non– case matched) compared 2 different medical registries (Australian and New Zealand Hemostasis Registry and Australian Society of Cardiac and Thoracic Surgeons Database) in the field of cardiac surgery. They showed increased mortality and length of hospital stay in the rFVIIa group as expected. No association with an increased incidence of stroke (OR ⫽ 1.0, p ⫽ 0.994) was founded although reporting a 6.9% incidence of stroke in the rFVIIa group (Australian and New Zealand Hemostasis Registry) versus 4.0% in the control group (Australian Society of Cardiac and Thoracic Surgeons Database).19 Various concerns on the use of rFVIIa in the specific setting of cardiac surgery led to the decision of the Canadian Consensus Conference on the emerging role of rFVIIa in on-pump cardiac surgery in 2006, which was published in 2007, that according to the lack of evidence, rFVIIa should be considered only in cardiac surgical patients who develop refractory hemorrhage (grade of recommendation 2C).18 When compared with the present authors’ previous meta-analysis published in 2009,12 the present work added 1 large randomized controlled study,2 which showed, for the first time in cardiac surgery, a statistically significant difference in the incidence of stroke. The authors’ previous work evidenced only a nonsignificant reduction in the rate of surgical re-exploration (10/76 [13%] in the rFVIIa group v 42/74 [57%] in the control group [OR ⫽ 0.25 (0.01-7.01), p ⫽ 0.42]) and a trend toward an increase in the rate of perioperative stroke (8/150 [5%] in the rFVIIa v 2/148 [1.4%] in the control arm [OR ⫽ 3.17 (0.8312.10), p ⫽ 0.09]).
PONSCHAB ET AL
Finally, the present results have to be compared with a very recently published meta-analysis by Levi et al37 in 2010 including 35 randomized clinical trials with a total number of 4,468 patients in various clinical settings. They substantiated a significantly increased risk of arterial (5.5% v 3.2%, p ⫽ 0.003) but not venous thromboembolic events. Further analysis of overall arterial thromboembolic events showed a significantly higher frequency of coronary thromboembolic events and a trend toward an increased rate of cerebrovascular thromboembolic complications, namely cerebral infarction. Even if the present authors were not the first to observe a significantly increased incidence of stroke in patients receiving rFVIIa limited to the cardiac surgery setting, they wanted to highlight the importance of the fact that adverse reactions may not be considered unless meticulous attention and efforts are made in follow-up assessments.27 The present authors want to encourage authors as well as reviewers to focus even more on intent-to-treat principles to evaluate the results of randomized controlled trials because the inclusion of patients following the protocol as planned contains a potential number of bias underestimating adverse effects of the treatment groups.38 An interesting point is represented by the wide range of rFVIIa dosage of the studies included in this meta-analysis. Even if most authors and reviews recommended a dosage of 90 g/kg for refractory bleeding in the setting of cardiac surgery,39 the authors suggested, following Welsby et al,40 that single, low-dose boluses of rFVIIa could be effective in this population; 20 to 40 g/kg, repeated if clinically indicated, may emerge as an appropriate dosing regimen in cardiac surgery patients. LIMITATIONS
A major limitation of this study was that not all of the studies included were performed as randomized controlled trails. Although no evidence of small-study bias was found, the reader should be aware of this problem as well as other well known limitations of systematic reviews and meta-analysis.41 Another important limitation of this meta-analysis (and obviously of all the studies conducted in this setting) was the absence of adequate follow-up. Another limitation of this study was the fact that the way to describe significant reduction of blood product use was different in the 6 studies being included in the present meta-analysis, and the authors cannot compare the treatment groups with control groups in matters of absolute numbers of blood product requirements. Nevertheless, all of the 6 studies included in the present meta-analysis reported on the reduction of blood product use or the reduction of blood loss. It should be added, as a limitation of the present study that a clear definition of stroke (new focal neurologic deficit, either transient but present ⬎24 hours or permanent) was present only in 1 study.2 Finally, the sample size of the 6 articles included in this analysis was small, accounting for a total of 470 patients (254 receiving rFVIIa and 216 controls). CONCLUSIONS
The administration of rFVIIa could result in a significant increase of stroke with a nonsignificant reduction of need for surgical re-exploration. While awaiting results of adequately powered randomized controlled trials, the present authors do
RECOMBINANT ACTIVATED FACTOR VII
809
not recommend widespread or prophylactic use of rFVIIa in adult cardiac surgery patients. Besides use of rFVIIa under controlled study conditions, recombinant FVIIa may be considered with caution in patients with refractory life-threatening bleeding after conventional methods to restore hemostasis and treatment like re-warming and surgical re-approach have failed. ACKNOWLEDGMENTS The authors thank Giulio Cozzuto, RN; Isotta Virzo, RN; Cristiano Chiappa, RN; Giuseppe Giardina, RN; Lara Castelnuovo, RN; Marco
Costantini; and Mariano Fichera for the careful data entry and revision of the article. APPENDIX 1. SEARCH STRATEGY FOR PUBMED DEVELOPED ACCORDING TO BIONDI-ZOCCAI ET AL13
((cardiac OR heart) AND (surgery OR bypass) AND (“rFVIIa” OR (recombinant AND factor AND (activated AND “VII”) OR “VIIa”))) AND ((clinical[Title/Abstract] AND trial[Title/Abstract]) OR clinical trials[MeSH Terms] OR clinical trial[Publication Type] OR random*[Title/Abstract] OR random allocation[MeSH Terms] OR therapeutic use[MeSH Subheading]).
REFERENCES 1. Wade CE, Eastridge BJ, Jones JA, et al: Use of recombinant factor VIIa in US military casualties for a five-year-period. J Trauma 69:344-352, 2010 2. Gill R, Herbertson M, Vuylsteke A, et al: Safety and efficacy of recombinant activated factor VII. A randomized placebo-controlled trial in the setting of bleeding after cardiac surgery. Circulation 320:21-27, 2009 3. 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 45:2634, 2005 4. Diprose P, Herbertson MJ, O’Shaughnessy D, et al: Activated recombinant factor VII after cardiopulmonary bypass reduces allogeneic transfusion in complex non-coronary cardiac surgery: Randomized double-blind placebocontrolled pilot study. Br J Anaesth 95:596-602, 2005 5. von Heymann C, Redlich U, Jain U, et al: Recombinant activated factor VII for refractory bleeding after cardiac surgery—A retrospective analysis of safety and efficacy. Crit Care Med 33:2241-2246, 2005 6. Gelsomino S, Lorusso R, Romagnoli S, et al: Treatment of refractory bleeding after cardiac operations with low-dose recombinant activated factor VII (NovoSeven): A propensity score analysis. Eur J Cardiothorac Surg 33:64-71, 2008 7. Tritapepe L, De Santis V, Vitale D, et al: Recombinant activated factor VII for refractory bleeding after acute aortic dissection surgery: A propensity score analysis. Crit Care Med 35:1685-1690, 2007 8. Van de Garde EMW, Bras LJ, Heijmen RH, et al: Low-dose recombinant factor VIIa in the management of uncontrolled postoperative hemorrhage in cardiac surgery patients. J Cardiothorac Vasc Anesth 20:573-575, 2006 9. Monroe DM, Hoffman M, Oliver JA, et al: Platelet activity of high dose factor VIIa is independent of tissue factor. Br J Haematol 99:542-547, 1997 10. Pang DA: Probable right atrial thrombus immediately after recombinant activated factor VII administration. Br J Anaesth 99:221225, 2007 11. Lichtman AD, Carullo V, Minhaj M, et al: Case 6 —2007: Massive intraoperative thrombosis and death after recombinant activated factor VII administration. J Cardiothorac Vasc Anesth 21:897902, 2007 12. Zangrillo A, Mizzi A, Biondi-Zoccai G, et al: Recombinant activated factor VII in cardiac surgery: A meta-analysis. J Cardiothorac Vasc Anesth 23:34-40, 2009 13. Biondi-Zoccai GG, Agostoni P, Abbate A, et al: A simple hint to improve Robinson and Dickersin’s highly sensitive PubMed search strategy for controlled clinical trials. Int J Epidemiol 34:224225, 2005 14. Fleiss JL: The statistical basis of meta-analysis. Stat Methods Med Res 2:121-145, 1993 15. Higgins JP, Green S:The Cochrane Handbook for Systematic Reviews of Interventions 4.2.5. The Cochrane Library, Issue 3, 2005. Chichester, UK, John Wiley & Sons, Ltd, 2005
16. Higgins JP, Thompson SG, Deeks JJ, et al: Measuring inconsistency in meta-analyses. BMJ 327:557-560, 2003 17. Warren O, Mandal K, Hadjianastassiou V, et al: Recombinant activated factor VII in cardiac surgery: A systematic review. Ann Thorac Surg 83:707-714, 2007 18. Karkouti K, Beattie WS, Crowther MA, et al: The role of recombinant factor VIIa in on-pump cardiac surgery: Proceedings of the Canadian Consensus Conference. Can J Anaesth 54:573-582, 2007 19. Mitra B, Phillips L, Cameron PA, et al: The safety of recombinant factor VIIa in cardiac surgery. Anaesth Intensive Care 38:671-677, 2010 20. Karkouti K, Yau MT, Riazi S, et al: Determinants of complications with recombinant factor VIIa for refractory blood loss in cardiac surgery. Can J Anaesth 53:802-809, 2006 21. McCall P, Story DA, Karapillai D: Audit of factor VIIa for bleeding resistant to conventional therapy following complex cardiac surgery. Can J Anaesth 53:926-933, 2006 22. Diprose P, Gill R: The potential role for recombinant factor VIIa in cardiac surgery. Br J Cardiol (Acute Interv Cardiol) 11:77-79, 2004 23. Karkouti K, Beattie WS, Arellano R, et al: Comprehensive Canadian review of the off-label use of recombinant activated factor VII in cardiac surgery. Circulation 118:331-338, 2008 24. Bowman LJ, Uber WE, Stroud MR, et al: Use of recombinant activated factor VII concentrate to control postoperative hemorrhage in complex cardiovascular surgery. Ann Thorac Surg 85:1669-1677, 2008 25. Dunkley S, Phillips L, McCall P, et al: Recombinant activated factor VII in cardiac surgery: Experience from the Australian and New Zealand Haemostasis Registry. Ann Thorac Surg 85:836-844, 2008 26. Isbister J, Phillips L, Dunkley S, et al: Recombinant activated factor VII in critical bleeding: Experience from the Australian and New Zealand Haemostasis Register. Intern Med J 38:156-165, 2008 27. Hsia CC, Zurawska JH, Tong MZY, et al. Recombinant activated factor VII in the treatment of non-haemophilia patients: Physician under-reporting of thromboembolic adverse events. Transfus Med 19:43-49, 2009 28. Tobias JD, Simsic JM, Weinstein S, et al: Recombinant factor VIIa to control excessive bleeding following surgery for congenital heart disease in pediatric patients. J Intensive Care Med 19:270-273, 2004 29. Ekert H, Brizard C, Eyers R, et al: 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: Controlled study of rFVIIa and standard haemostatic replacement therapy. Blood Coag Fibrinolysis 17:389-395, 2006 30. Essam MA: Prophylactic administration of recombinant activated factor VII in coronary revascularization surgery. Int J Anesthesiol 13:10, 2007 31. Romagnoli S, Bevilacqua S, Gelsomino S, et al: Small-dose recombinant activated factor VII (NovoSeven) in cardiac surgery. Anesth Analg 102:1320-1326, 2006
810
32. Hsia CC, Chin-Yee IH, McAlister VC: Use of recombinant activated factor VII in patients without hemophilia: A meta-analysis of randomized control trials. Ann Surg 248: 61-68, 2008 33. Ranucci M, Isgró G, Soro G, et al: Efficacy and safety of recombinant activated factor VII in major surgical procedures: Systematic review and meta-analysis of randomised clinical trials. Arch Surg 143:296-304, 2008 34. Diringer MN, Skolnick BE, Mayer SA, et al: Thromboembolic events with recombinant activated factor VII in spontaneous intracerebral hemorrhage. Stroke 41:48-53, 2010 35. Karkouti K, Beattie WS: Pro: The role of recombinant factor VIIa in cardiac surgery. J Cardiothorac Vasc Anesth 22:779-782, 2008 36. Al-Ruzzeh S, Ibrahim K, Navia JL: Con: The role of recombinant factor VIIa in the control of bleeding after cardiac surgery. J Cardiothorac Vasc Anesth 22:783-785, 2008
PONSCHAB ET AL
37. Levi M, Levy MD, Andersen HF, et al: Safety of recombinant activated factor VII in randomized clinical trials. N Engl J Med 363:1791-1800, 2010 38. Hardy JF, Belisle S, van der Linden P: Efficacy and safety of recombinant factor VII in cardiac surgical patients. Curr Opin Anaesthesiol 22:95-99, 2009 39. Richardson A, Herbertson M, Gill R: The role of recombinant activated factor VII in cardiac surgery. HSR Proc Intensive Care and Cardiovascular Anesthesia 1:9-12, 2009 40. Welsby IJ, Monroe DM, Lawson JH, et al: Recombinant activated factor VII and the anaesthetist. Anaesthesia 60:1203-1212, 2005 41. Biondi-Zoccai GG, Agostoni P, Abbate A: Parallel hierarchy of scientific studies in cardiovascular medicine. Ital Heart J 4:819-820, 2003
creased rate of stroke (12/254 [4.7%] in the rFVIIa group v 2/216 [0.9%] in the control arm, odds ratio [OR] ⴝ 3.69 [1.1-12.38], p ⴝ 0.03) with a nonsignificant reduction in rate of surgical re-exploration (13% v 42% [OR ⴝ 0.27 (0.04-1.9), p ⴝ 0.19]). The authors observed a trend toward an increase of overall perioperative thromboembolic events (19/254 [7.5%] in the rFVIIa group v 10/216 [5.6%] in the control arm [OR ⴝ 1.84 (0.82-4.09), p ⴝ 0.14]). No difference in the rate of death was observed. Conclusions: The administration of rFVIIa in cardiac surgery patients could result in a significant increase of stroke with a trend toward a reduction of the need for surgical re-exploration. The authors do not recommend routine use in cardiac surgery patients. rFVIIa may be considered with caution in patients with refractory life-threatening bleeding. © 2011 Elsevier Inc. All rights reserved.
C
et al, safety concerns already had been raised, but the literature was inconclusive and did not focus on adverse effects in the area of cerebrovascular events like stroke. These references out of various clinical settings in conjunction with the previous results of Zangrillo et al12 led the authors to the decision to scrupulously review and scan scientific articles on this topic. A rising number of references and notices in literature currently being published showed the necessity to conduct a systematic review of data pooled from existing trials limited to the cardiac surgery setting. Considering the hypothesis that the potential beneficial hemostatic effects could be counterbalanced by adverse effects, the authors performed a meta-analysis and comprehensive review of the current literature to determine the potential impact of rFVIIa on surgical re-exploration and the risk of adverse arterial vascular effects, namely stroke complications.
OAGULATION FACTORS like recombinant activated factor VII (rFVIIa) (NovoSeven; Novo Nordisk A/S, Bagsvaerd, Denmark) recently gained special interest because they interact in their natural physiologic environment of coagulation cascade once being activated. As soon as the first reports on rFVIIa were published on the successful resuscitation and survival of combat victims, rFVIIa increasingly was used off-label in refractory bleeding associated with trauma1 and cardiac surgery.2-8 Because rFVIIa exerts its pharmacologic action by inducing thrombin generation on locally activated platelets and contributes to the formation of a stabilized fibrin clot at the site of vessel injury,9 safety concerns have been raised on the potential danger of severe thromboembolic events.10,11 Recently, safety and efficacy data published by Gill et al2 in a randomized controlled trial showed a reduced rate of patients undergoing surgical revision and a reduction in overall mediastinal drainage blood loss as well as transfusion requirements with a trend toward a higher incidence of critical severe adverse events. A meta-analysis published in 2009 by Zangrillo et al12 at the authors’ institution reported a nonsignificant reduction in surgical re-exploration in patients receiving rFVIIa and a nonsignificant increase in the rate of perioperative stroke in these patients. At the time of the publication of the work of Zangrillo
From the *Department of Anesthesia and Intensive Care, Università Vita-Salute San Raffaele, Milan, Italy; and †Division of Cardiology, University of Modena and Reggio Emilia, Modena, Italy. Address reprint requests to Giovanni Landoni, MD, Department of Cardiothoracic Anesthesia and Intensive Care, Istituto Scientifico San Raffaele, Via Olgettina 60, Milan 20132, Italy. E-mail: landoni. [email protected] © 2011 Elsevier Inc. All rights reserved. 1053-0770/2505-0010$36.00/0 doi:10.1053/j.jvca.2011.03.004 804
KEY WORDS: bleeding, cardiac surgery, FVIIa, recombinant factor VII, stroke, surgical revision, anesthesia
METHODS
Search Strategy Four trained investigators independently searched for clinical trials in PubMed (updated December 1, 2010). The search strategy was established by Biondi-Zoccai et al13 and is available in Appendix 1. In addition to the special search strategy mentioned earlier, conference proceedings (2008-2010) from the European Society of Anaesthesia, International Anesthesia Research Society, American Society of Anesthesiologists, European Society of Intensive Care Medicine, Society of Cardiovascular Anesthesiologists, European Association of Cardiothoracic Anaesthesiologists, International Society of Thrombosis and Haemostasis, and American College of Chest Physicians congresses were scanned in order to detect further abstracts or recent scientific articles of relevance. In addition, the authors systematically applied backward snowballing (ie, scanning of references of retrieved articles and pertinent reviews) and contacted international experts for further studies. No language restriction was enforced, and non–English-language articles were translated before further analysis.
Journal of Cardiothoracic and Vascular Anesthesia, Vol 25, No 5 (October), 2011: pp 804-810
RECOMBINANT ACTIVATED FACTOR VII
805
Table 1. Description of the 6 Studies Included in the Systematic Review on rFVIIa Versus Control With the Number of Patients and Interventions
Journal
Year
rFVIIa (g/kg) First Dose (Mean ⫾ Standard Deviation)
Diprose
Br J Anaesth
2005
90 (0)
No
Karkouti
Transfusion
2005
Tritapepe Von Heymann Gelsomino
Crit Care Med Crit Care Med Eur J Cardiothorac Surg
2007 2005 2008
Gill
Circulation
2009
7 pts: 62 (13) 44 pts: 37 (9) 70 (0) 60 (median) 18 (9-16) (median; interquartile) 35 pts: 40 69 pts: 80
First Author
No. of Patients Receiving a Further Dose
Cardiac Surgery Procedures
Study Design
No. of Patients in the rFVIIa Group
No. of Patients in the Control Group
Randomized
10
10
14
Complex noncoronary cardiac surgery Varied procedures
Propensity-matched
51
51
4 10 Nonspecified
Aortic dissection Varied procedures Varied procedures
Propensity-matched Case-matched Propensity-matched
23 26 40
23 24 40
No
Varied procedures
Randomized
104
68
254
216
Total Abbreviation: pts, patients.
Study Selection A total of 208 references obtained from database and literature searches first were examined independently at the title/abstract level by 4 investigators. Included studies had to compare the use of rFVIIa versus placebo or standard treatment; the authors only included scientific publications relating to adult cardiac surgical patients. Exclusion criteria were nonhuman experimental or animal studies, duplicate publications (in this case only the article reporting the longest follow-up was included), missing outcome data, and no case matching. In case of missing outcome data in studies, original authors were contacted and asked for additional information. Four investigators selected studies for the final analysis by independently assessing compliance to the selection criteria (Table 1).
Data Abstraction and Study Characteristics Baseline, procedural, and outcome data were abstracted independently by 4 investigators, with divergences resolved by consensus and are listed in Table 1. Specifically, the authors extracted study design, population, clinical setting, rFVIIa dosage, and treatment duration. At least 2 separate attempts at contacting original authors were made in case of missing data. The primary endpoint of this analysis was to determine the safety of rFVIIa in terms of stroke (as per author definition) and the efficacy of rFVIIa in terms of the reduction of surgical revision for bleeding. Secondary endpoints included death, the transfusion of blood products, and the incidence of overall venous and arterial vascular complications (ie, myocardial infarction, stroke, and deep venous thrombosis as per author definition).
Data Analysis and Synthesis Binary outcomes from individual studies were analyzed and pooled in order to compute individual odds ratios (ORs) with pertinent 95% confidence intervals according to the fixed-effect Peto method in case of homogenous data (inconsistency ⬍50%); and the random-effect DerSimonian-Laird method in case of statistical heterogeneity (inconsistency ⬍50%).13,14 Statistical heterogeneity (ie, the variability in effect estimates from study to study) and inconsistency (a more robust estimate of the variability in effect estimates from study to study, which is largely independent of the number of included studies) were measured using the Cochran Q test (which is largely comparable to the chi-square test, and, when yielding p values ⬍0.10, suggests significant statistical heterogeneity) and I2, respectively.15 According to Higgins et al,16 I2 values around 25%, 50%, and 75% were considered as representing low, moderate, and severe statistical inconsistency, respec-
tively. Specifically, OR ⬍1 suggests that the first comparator is associated with a reduced incidence of the event, whereas OR ⬎1 suggests that the 2nd comparator is associated with an increased incidence of the event. Unadjusted p values are reported throughout. The risk of smallstudy bias (including publication bias) was assessed by a visual inspection of funnel plots (ie, the graphic display of effect estimates and study precision, which enables appraisal of the potential asymmetry in metaanalysis datasets, often caused by the selective lack of publication of small negative studies, ie, publication bias).16 Statistical significance was set at the 2-tailed 0.05 level for hypothesis testing and at 0.10 for heterogeneity testing. Computations were performed with SPSS 11.0 (SPSS, Chicago, IL) and RevMan 4.2 (a freeware available from The Cochrane Collaboration).14 This study was performed in accordance with The Cochrane Collaboration and the Quality of Reporting of Meta-Analyses guidelines. RESULTS
Database searches, snowballing, and contacts with experts yielded a total of 208 citations. From these 208 citations, 187 nonpertinent titles or abstracts had to be excluded, and 21 studies were retrieved in complete form. A total of 15 studies were further excluded because they turned out to be review articles,17,18 they were noncase matched19 or had nonrandomized experimental design,20-27 they studied a pediatric population,28,29 they did not report outcome data,30 or duplicate publication.31 Six eligible clinical trials were identified (all of them recently published between 2005 and 2009), which were included in the final analysis2-7 (Table 1). Study Characteristics The 6 included clinical trials were comprised of 470 patients (254 receiving rFVIIa and 216 controls) (Table 1). One trial was conducted in the setting of noncoronary artery surgery,4 1 in aortic dissections,7 and the other 4 studies were involved in various cardiac surgery procedures.2,3,5,6 All authors studied the hemostatic properties of rFVIIa and reported on vascular complications and/or surgical re-exploration. rFVIIa dosage varied across studies (Table 1), between 18 and 70 g/kg being given in repeatable dosage and 90 g/kg used as a single dose. One study used rFVIIa prophylactically,4 whereas the other 5 used rFVIIa in patients with refractory bleeding (definitions of refractory bleeding varied among authors). Five studies reported a
806
PONSCHAB ET AL
Fig 1.
Individual and pooled estimates for stroke. (Color version of figure is available online.)
reduction in bleeding after rFVIIa administration,2-4,6,7 whereas 1 study5 proved that rFVIIa was safe but not incrementally efficacious over conventional hemostatic therapy. Some of the studies included in this meta-analysis appeared of poor internal validity, with 2 that were randomized2,3 and 1 with a multicenter design.2 A clear definition of stroke (new focal neurologic deficit, either transient but present ⬎24 hours or permanent), was present only in 1 study.2 Quantitative Data Synthesis Safety Overall analysis showed that the use of rFVIIa was associated with an increased rate of stroke (12/254 [4.7%] in the rFVIIa v 2/216 [0.9%] in the control arm (Fig 1), OR ⫽ 3.69 [1.1-12.38], p for effect ⫽ 0.03, p for heterogeneity ⫽ 0.92, I2 ⫽ 0% with 470 patients included in 6 studies, number needed to harm ⫽ 26). A nonsignificant increase was found in the rate of overall perioperative vascular events; the authors included myocardial infarction, stroke, and deep venous thrombosis as per author definition (19/254 [7.5%] in the rFVIIa v 10/216 [5.6%] in the control arm, OR ⫽ 1.84 [0.82-4.09], p for effect ⫽ 0.14, p for heterogeneity ⫽ 0.84, I2 ⫽ 0% with 470 included patients in 6 studies) (Fig 2).
Fig 2.
Because myocardial infarctions and the incidence of deep venous thrombosis did not show any statistically significant difference, it meant that the overall effect of the nonsignificant increase of arterial and venous vascular events was driven primarily by the increase in the events of stroke. In the 5 studies that provided this information, 5 of 214 (2.3%) myocardial infarctions were noted in the rFVIIa group versus 8 of 176 (4.5%) in the placebo group (OR ⫽ 0.59 [0.2-1.79], p for effect ⫽ 0.35, p for heterogeneity ⫽ 0.89, I2 ⫽ 0% with 390 patients included). Mortality was similar in the 2 groups (33/ 254 [13%] in the rFVIIa group and 26/216 [12%] in the control group [OR ⫽ 1.14 (0.65-2.01), p for effect ⫽ 0.65, p for heterogeneity ⫽ 0.88, I2 ⫽ 0% with 470 patients included in 5 studies]). Efficacy for Surgical Revision, Bleeding, and Blood Product Transfusion The present meta-analysis shows a nonsignificant reduction in the rate of surgical re-exploration (23/180 [13%] in the rFVIIa group v 59/142 [42%] in the control group [OR ⫽ 0.27 (0.04-1.9), p for effect ⫽ 0.19, p for heterogeneity ⬍0.001, I2 ⫽ 87% with 322 patients included in 4 studies]) (Fig 3). The authors cannot report on the length of intensive care or hospital
Individual and pooled estimates for thromboembolic events. (Color version of figure is available online.)
RECOMBINANT ACTIVATED FACTOR VII
Fig 3.
807
Individual and pooled estimates for surgical re-exploration. (Color version of figure is available online.)
stay because only 1 study provided sufficient data. The inspection of the funnel plots for all these endpoints did not disclose any evidence of small-study/publication bias, even if such appraisal appeared underpowered by the small set of studies and limited by differences in study design. All of the 6 studies included in this meta-analysis reported a reduced amount of blood loss or a reduction of blood product use but used incomparable performance data of measurements in their documentation. Four2,4-6 of the 6 trials reported on the reduction of blood transfusion requirements versus control, whereas 2 studies3,7 only reported on significant reductions in blood being transfused after the use of rFVIIa versus the time before the application of rFVIIa. Gill et al2 only documented the overall transfusion requirements, whereas the other 3 studies4-6 split the results on red blood cell units, platelet units, and fresh frozen plasma units being transfused. Reduced hourly blood loss was reported in 2 studies3,7 for the first hour before rFVIIa compared with blood loss during the first hour after rFVIIa, whereas Gelsomino et al6 reported on hourly bleeding over 24 hours after rFVIIa. Gelsomino et al6and von Heymann et al5 showed the overall blood loss over 24 hours after rFVIIa versus control, whereas Gill et al2 reported on 4-hour blood loss via mediastinal drainage versus control. Because of the different ways of documentation outlined previously, the treatment groups could not be compared with the control groups in matters of absolute numbers of blood product requirements or in matters of blood loss over a definite time. DISCUSSION
The present authors performed an updated meta-analysis of pooled data from 6 studies and showed that the use of rFVIIa in cardiac surgery was associated with a significant increase in the occurrence of perioperative stroke while presenting a nonsignificant reduction in the need for surgical re-exploration. Beyond that the authors established a comprehensive review of evidence of various clinical settings recently being published focusing on undesirable vascular events. Because the use of rFVIIa proved to interact as a coagulation-promoting pharmacologic substance, safety concerns were raised on the induction of the coagulation cascade causing severe vascular events. However, each study focusing on the
possible potential side effects after the application of rFVIIa failed to substantiate these objections. Although an extensive literature search was performed in this meta-analysis, the authors could only identify 6 randomized or propensity-matched or case-matched studies, included in the present data collection. Most publications on the use of rFVIIa in various surgical fields were performed as case series with the lack of a proper control group. The inclusion of these 6 studies allowed the present authors, for the first time, to identify a statistically significant increase in the occurrence of stroke after the administration of rFVIIa in the setting of cardiac surgery. In fact, some recent meta-analyses and reviews12,23,27,32 already suggested that the use of rFVIIa could be dangerous, but the results did not reach statistical significance, mainly because of the small sample size. The present results could be compared with the meta-analysis of Ranucci et al,33 which included 7 randomized controlled trials performed in major surgical procedures. In their study, no differences in terms of thromboembolic complications and mortality rate were observed, but the study did not focus on stroke events. In a subgroup analysis, Ranucci showed a significant effect of treatment in reducing the number of patients being exposed to allogeneic packed red blood cell transfusions only after a minimum dosage of 50 g/kg (137/211 [64.9%] v 108/158 [68.4%] in the control arm; fixed-effects OR, 0.43; 95% confidence interval, 0.23-0.78; p ⫽ 0.006). In contrast to this study, the present meta-analysis included only cardiac surgical patients. Reporting on a different study population of patients with spontaneous intracerebral hemorrhage, a recent trial by Diringer et al34 showed a dose-dependent significant increase of arterial thromboembolic events in the group of patients receiving 80 g/kg of rFVIIa (OR ⫽ 2.14, p ⫽ 0.031). The study group identified risk factors for arterial thromboembolism after the use of rFVIIa, namely increased age, signs of ischemia at baseline (cardiac or cerebral), and the prior use of platelet inhibitors. Another meta-analysis of 22 randomized controlled studies in nonhemophilic patients reported a reduction in the need for blood transfusion in patients receiving rFVIIa, with a trend toward an increased risk of arterial thrombosis.32 A recently published retrospective review of 69 patients receiving off-label rFVIIa identified 12 patients (17%) with
808
thromboembolic adverse events deemed to be related to drug exposure.27 Observed arterial thromboembolic adverse events consisted almost exclusively of ischemic stroke in mixed surgical approaches, mainly in cardiac surgery settings. A systematic review of rFVIIa in cardiac surgery comparing 6 studies including 298 patients was performed by Warren et al17 in 2008 and included 2 studies on pediatric surgical procedures, which is in contrast to the present study population in which pediatric surgical patients were excluded. The authors of this review reported a 2.6% incidence of stroke of a total of 5.3% thromboembolic events in the adult study group. This meta-analysis included 2 randomized controlled trials, 3 propensity-matched, and 1 case-matched study of a total of 470 patients presenting at the highest perioperative risk profile for adverse outcomes. In fact, this study provided the most comprehensive and thorough comparison of rFVIIa versus control in patients undergoing cardiac surgery that currently exists. Safety issues about the use of a procoagulant drug such as rFVIIa were raised repeatedly in the medical literature, and the debate of the use of rFVIIa in cardiac surgery was controversial apart from the indication of refractory bleeding.35,36 The main concern was that hemostatic drugs may induce thromboembolic events.10,11 In the present analysis, a significant increase in the occurrence of perioperative stroke was found, and pooled data showed a high level of heterogeneity throughout all studies being included. The finding of a significant increase of stroke was again confirmed by the results of Karkouti et al23 in a retrospective, non– case-matched study on 503 patients receiving rFVIIa, with a 12% incidence of stroke. An Australian study group retrospectively (non– case matched) compared 2 different medical registries (Australian and New Zealand Hemostasis Registry and Australian Society of Cardiac and Thoracic Surgeons Database) in the field of cardiac surgery. They showed increased mortality and length of hospital stay in the rFVIIa group as expected. No association with an increased incidence of stroke (OR ⫽ 1.0, p ⫽ 0.994) was founded although reporting a 6.9% incidence of stroke in the rFVIIa group (Australian and New Zealand Hemostasis Registry) versus 4.0% in the control group (Australian Society of Cardiac and Thoracic Surgeons Database).19 Various concerns on the use of rFVIIa in the specific setting of cardiac surgery led to the decision of the Canadian Consensus Conference on the emerging role of rFVIIa in on-pump cardiac surgery in 2006, which was published in 2007, that according to the lack of evidence, rFVIIa should be considered only in cardiac surgical patients who develop refractory hemorrhage (grade of recommendation 2C).18 When compared with the present authors’ previous meta-analysis published in 2009,12 the present work added 1 large randomized controlled study,2 which showed, for the first time in cardiac surgery, a statistically significant difference in the incidence of stroke. The authors’ previous work evidenced only a nonsignificant reduction in the rate of surgical re-exploration (10/76 [13%] in the rFVIIa group v 42/74 [57%] in the control group [OR ⫽ 0.25 (0.01-7.01), p ⫽ 0.42]) and a trend toward an increase in the rate of perioperative stroke (8/150 [5%] in the rFVIIa v 2/148 [1.4%] in the control arm [OR ⫽ 3.17 (0.8312.10), p ⫽ 0.09]).
PONSCHAB ET AL
Finally, the present results have to be compared with a very recently published meta-analysis by Levi et al37 in 2010 including 35 randomized clinical trials with a total number of 4,468 patients in various clinical settings. They substantiated a significantly increased risk of arterial (5.5% v 3.2%, p ⫽ 0.003) but not venous thromboembolic events. Further analysis of overall arterial thromboembolic events showed a significantly higher frequency of coronary thromboembolic events and a trend toward an increased rate of cerebrovascular thromboembolic complications, namely cerebral infarction. Even if the present authors were not the first to observe a significantly increased incidence of stroke in patients receiving rFVIIa limited to the cardiac surgery setting, they wanted to highlight the importance of the fact that adverse reactions may not be considered unless meticulous attention and efforts are made in follow-up assessments.27 The present authors want to encourage authors as well as reviewers to focus even more on intent-to-treat principles to evaluate the results of randomized controlled trials because the inclusion of patients following the protocol as planned contains a potential number of bias underestimating adverse effects of the treatment groups.38 An interesting point is represented by the wide range of rFVIIa dosage of the studies included in this meta-analysis. Even if most authors and reviews recommended a dosage of 90 g/kg for refractory bleeding in the setting of cardiac surgery,39 the authors suggested, following Welsby et al,40 that single, low-dose boluses of rFVIIa could be effective in this population; 20 to 40 g/kg, repeated if clinically indicated, may emerge as an appropriate dosing regimen in cardiac surgery patients. LIMITATIONS
A major limitation of this study was that not all of the studies included were performed as randomized controlled trails. Although no evidence of small-study bias was found, the reader should be aware of this problem as well as other well known limitations of systematic reviews and meta-analysis.41 Another important limitation of this meta-analysis (and obviously of all the studies conducted in this setting) was the absence of adequate follow-up. Another limitation of this study was the fact that the way to describe significant reduction of blood product use was different in the 6 studies being included in the present meta-analysis, and the authors cannot compare the treatment groups with control groups in matters of absolute numbers of blood product requirements. Nevertheless, all of the 6 studies included in the present meta-analysis reported on the reduction of blood product use or the reduction of blood loss. It should be added, as a limitation of the present study that a clear definition of stroke (new focal neurologic deficit, either transient but present ⬎24 hours or permanent) was present only in 1 study.2 Finally, the sample size of the 6 articles included in this analysis was small, accounting for a total of 470 patients (254 receiving rFVIIa and 216 controls). CONCLUSIONS
The administration of rFVIIa could result in a significant increase of stroke with a nonsignificant reduction of need for surgical re-exploration. While awaiting results of adequately powered randomized controlled trials, the present authors do
RECOMBINANT ACTIVATED FACTOR VII
809
not recommend widespread or prophylactic use of rFVIIa in adult cardiac surgery patients. Besides use of rFVIIa under controlled study conditions, recombinant FVIIa may be considered with caution in patients with refractory life-threatening bleeding after conventional methods to restore hemostasis and treatment like re-warming and surgical re-approach have failed. ACKNOWLEDGMENTS The authors thank Giulio Cozzuto, RN; Isotta Virzo, RN; Cristiano Chiappa, RN; Giuseppe Giardina, RN; Lara Castelnuovo, RN; Marco
Costantini; and Mariano Fichera for the careful data entry and revision of the article. APPENDIX 1. SEARCH STRATEGY FOR PUBMED DEVELOPED ACCORDING TO BIONDI-ZOCCAI ET AL13
((cardiac OR heart) AND (surgery OR bypass) AND (“rFVIIa” OR (recombinant AND factor AND (activated AND “VII”) OR “VIIa”))) AND ((clinical[Title/Abstract] AND trial[Title/Abstract]) OR clinical trials[MeSH Terms] OR clinical trial[Publication Type] OR random*[Title/Abstract] OR random allocation[MeSH Terms] OR therapeutic use[MeSH Subheading]).
REFERENCES 1. Wade CE, Eastridge BJ, Jones JA, et al: Use of recombinant factor VIIa in US military casualties for a five-year-period. J Trauma 69:344-352, 2010 2. Gill R, Herbertson M, Vuylsteke A, et al: Safety and efficacy of recombinant activated factor VII. A randomized placebo-controlled trial in the setting of bleeding after cardiac surgery. Circulation 320:21-27, 2009 3. 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 45:2634, 2005 4. Diprose P, Herbertson MJ, O’Shaughnessy D, et al: Activated recombinant factor VII after cardiopulmonary bypass reduces allogeneic transfusion in complex non-coronary cardiac surgery: Randomized double-blind placebocontrolled pilot study. Br J Anaesth 95:596-602, 2005 5. von Heymann C, Redlich U, Jain U, et al: Recombinant activated factor VII for refractory bleeding after cardiac surgery—A retrospective analysis of safety and efficacy. Crit Care Med 33:2241-2246, 2005 6. Gelsomino S, Lorusso R, Romagnoli S, et al: Treatment of refractory bleeding after cardiac operations with low-dose recombinant activated factor VII (NovoSeven): A propensity score analysis. Eur J Cardiothorac Surg 33:64-71, 2008 7. Tritapepe L, De Santis V, Vitale D, et al: Recombinant activated factor VII for refractory bleeding after acute aortic dissection surgery: A propensity score analysis. Crit Care Med 35:1685-1690, 2007 8. Van de Garde EMW, Bras LJ, Heijmen RH, et al: Low-dose recombinant factor VIIa in the management of uncontrolled postoperative hemorrhage in cardiac surgery patients. J Cardiothorac Vasc Anesth 20:573-575, 2006 9. Monroe DM, Hoffman M, Oliver JA, et al: Platelet activity of high dose factor VIIa is independent of tissue factor. Br J Haematol 99:542-547, 1997 10. Pang DA: Probable right atrial thrombus immediately after recombinant activated factor VII administration. Br J Anaesth 99:221225, 2007 11. Lichtman AD, Carullo V, Minhaj M, et al: Case 6 —2007: Massive intraoperative thrombosis and death after recombinant activated factor VII administration. J Cardiothorac Vasc Anesth 21:897902, 2007 12. Zangrillo A, Mizzi A, Biondi-Zoccai G, et al: Recombinant activated factor VII in cardiac surgery: A meta-analysis. J Cardiothorac Vasc Anesth 23:34-40, 2009 13. Biondi-Zoccai GG, Agostoni P, Abbate A, et al: A simple hint to improve Robinson and Dickersin’s highly sensitive PubMed search strategy for controlled clinical trials. Int J Epidemiol 34:224225, 2005 14. Fleiss JL: The statistical basis of meta-analysis. Stat Methods Med Res 2:121-145, 1993 15. Higgins JP, Green S:The Cochrane Handbook for Systematic Reviews of Interventions 4.2.5. The Cochrane Library, Issue 3, 2005. Chichester, UK, John Wiley & Sons, Ltd, 2005
16. Higgins JP, Thompson SG, Deeks JJ, et al: Measuring inconsistency in meta-analyses. BMJ 327:557-560, 2003 17. Warren O, Mandal K, Hadjianastassiou V, et al: Recombinant activated factor VII in cardiac surgery: A systematic review. Ann Thorac Surg 83:707-714, 2007 18. Karkouti K, Beattie WS, Crowther MA, et al: The role of recombinant factor VIIa in on-pump cardiac surgery: Proceedings of the Canadian Consensus Conference. Can J Anaesth 54:573-582, 2007 19. Mitra B, Phillips L, Cameron PA, et al: The safety of recombinant factor VIIa in cardiac surgery. Anaesth Intensive Care 38:671-677, 2010 20. Karkouti K, Yau MT, Riazi S, et al: Determinants of complications with recombinant factor VIIa for refractory blood loss in cardiac surgery. Can J Anaesth 53:802-809, 2006 21. McCall P, Story DA, Karapillai D: Audit of factor VIIa for bleeding resistant to conventional therapy following complex cardiac surgery. Can J Anaesth 53:926-933, 2006 22. Diprose P, Gill R: The potential role for recombinant factor VIIa in cardiac surgery. Br J Cardiol (Acute Interv Cardiol) 11:77-79, 2004 23. Karkouti K, Beattie WS, Arellano R, et al: Comprehensive Canadian review of the off-label use of recombinant activated factor VII in cardiac surgery. Circulation 118:331-338, 2008 24. Bowman LJ, Uber WE, Stroud MR, et al: Use of recombinant activated factor VII concentrate to control postoperative hemorrhage in complex cardiovascular surgery. Ann Thorac Surg 85:1669-1677, 2008 25. Dunkley S, Phillips L, McCall P, et al: Recombinant activated factor VII in cardiac surgery: Experience from the Australian and New Zealand Haemostasis Registry. Ann Thorac Surg 85:836-844, 2008 26. Isbister J, Phillips L, Dunkley S, et al: Recombinant activated factor VII in critical bleeding: Experience from the Australian and New Zealand Haemostasis Register. Intern Med J 38:156-165, 2008 27. Hsia CC, Zurawska JH, Tong MZY, et al. Recombinant activated factor VII in the treatment of non-haemophilia patients: Physician under-reporting of thromboembolic adverse events. Transfus Med 19:43-49, 2009 28. Tobias JD, Simsic JM, Weinstein S, et al: Recombinant factor VIIa to control excessive bleeding following surgery for congenital heart disease in pediatric patients. J Intensive Care Med 19:270-273, 2004 29. Ekert H, Brizard C, Eyers R, et al: 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: Controlled study of rFVIIa and standard haemostatic replacement therapy. Blood Coag Fibrinolysis 17:389-395, 2006 30. Essam MA: Prophylactic administration of recombinant activated factor VII in coronary revascularization surgery. Int J Anesthesiol 13:10, 2007 31. Romagnoli S, Bevilacqua S, Gelsomino S, et al: Small-dose recombinant activated factor VII (NovoSeven) in cardiac surgery. Anesth Analg 102:1320-1326, 2006
810
32. Hsia CC, Chin-Yee IH, McAlister VC: Use of recombinant activated factor VII in patients without hemophilia: A meta-analysis of randomized control trials. Ann Surg 248: 61-68, 2008 33. Ranucci M, Isgró G, Soro G, et al: Efficacy and safety of recombinant activated factor VII in major surgical procedures: Systematic review and meta-analysis of randomised clinical trials. Arch Surg 143:296-304, 2008 34. Diringer MN, Skolnick BE, Mayer SA, et al: Thromboembolic events with recombinant activated factor VII in spontaneous intracerebral hemorrhage. Stroke 41:48-53, 2010 35. Karkouti K, Beattie WS: Pro: The role of recombinant factor VIIa in cardiac surgery. J Cardiothorac Vasc Anesth 22:779-782, 2008 36. Al-Ruzzeh S, Ibrahim K, Navia JL: Con: The role of recombinant factor VIIa in the control of bleeding after cardiac surgery. J Cardiothorac Vasc Anesth 22:783-785, 2008
PONSCHAB ET AL
37. Levi M, Levy MD, Andersen HF, et al: Safety of recombinant activated factor VII in randomized clinical trials. N Engl J Med 363:1791-1800, 2010 38. Hardy JF, Belisle S, van der Linden P: Efficacy and safety of recombinant factor VII in cardiac surgical patients. Curr Opin Anaesthesiol 22:95-99, 2009 39. Richardson A, Herbertson M, Gill R: The role of recombinant activated factor VII in cardiac surgery. HSR Proc Intensive Care and Cardiovascular Anesthesia 1:9-12, 2009 40. Welsby IJ, Monroe DM, Lawson JH, et al: Recombinant activated factor VII and the anaesthetist. Anaesthesia 60:1203-1212, 2005 41. Biondi-Zoccai GG, Agostoni P, Abbate A: Parallel hierarchy of scientific studies in cardiovascular medicine. Ital Heart J 4:819-820, 2003