Telephone: 44-121-697-8391; Fax: 44 121 697 8310
Efficacy and safety of repeated perioperative doses of recombinant factor VIIa in liver transplantation†
Article first published online: 20 JUL 2005
Copyright © 2005 American Association for the Study of Liver Diseases
Volume 11, Issue 8, pages 973–979, August 2005
How to Cite
Lodge, J. P. A., Jonas, S., Jones, R. M., Olausson, M., Mir-Pallardo, J., Soefelt, S., Garcia-Valdecasas, J. C., McAlister, V. and Mirza, D. F. (2005), Efficacy and safety of repeated perioperative doses of recombinant factor VIIa in liver transplantation. Liver Transpl, 11: 973–979. doi: 10.1002/lt.20470
See Editorial on Page 872
Other members of the rFVIIa OLT Study Group include Massimo Malagó, Silvio Nadalin, James Garden, Nigel Heaton, Jorge Ortiz de Urbina, Mikel Gastaca, Bo-Göran Ericzon, Andrew Watts, Stephen Chung, Gary Levy, Peter Metrakos, Jean Marty, and Frank Lagneau.
- Issue published online: 20 JUL 2005
- Article first published online: 20 JUL 2005
- Manuscript Accepted: 5 APR 2005
- Manuscript Received: 17 DEC 2004
Patients undergoing orthotopic liver transplantation (OLT) have excessive blood loss during surgery that requires blood transfusions, leading to increased postoperative morbidity and mortality. We studied the efficacy and safety of activated recombinant factor VII (rFVIIa) in reducing transfusion requirements in OLT. This multicenter, randomized, double-blind, placebo-controlled trial enrolled patients undergoing OLT because of cirrhosis (Child-Turcotte-Pugh class B or C). Patients received a repeated intravenous bolus regimen of rFVIIa 60 or 120 μg/kg or placebo. The primary efficacy endpoint was the total number of red blood cell (RBC) units transfused during the perioperative period. A total of 182 patients were analyzed for efficacy and 183 for safety. No significant effect of rFVIIa was observed on the number of RBC units transfused or intraoperative blood loss compared with the placebo group. A significantly higher number of patients in the rFVIIa study groups avoided RBC transfusion. Administration of rFVIIa but not placebo restored the preoperative prolonged prothrombin time to normal value during surgery. Patients receiving rFVIIa and placebo did not experience a significant difference in rate of thromboembolic events. Additionally, there was no statistically significant effect of rFVIIa treatment on hospitalization rate, total surgery time, and the proportion of patients undergoing retransplantation. In conclusion, use of rFVIIa during OLT significantly reduced the number of patients requiring RBC transfusion. There was no increase in thromboembolic events with rFVIIa administration compared with placebo. (Liver Transpl 2005;11:973–979.)
Despite advances in surgical techniques and medical management, the majority of patients undergoing orthotopic liver transplantation (OLT) require varying amounts of transfused blood or hemostatic support because of coagulation abnormalities. Even though emphasis has been placed on reducing blood loss during OLT, this procedure still carries a risk of excessive blood loss, which is associated with a higher risk of mortality,1–5 postoperative multiorgan dysfunction,6 and reduced graft survival.5
Activated recombinant factor VII (rFVIIa; NovoSeven®, Novo Nordisk A/S, Bagsvaerd, Denmark) is a hemostatic agent approved for use in hemophilia patients. The action of rFVIIa is localized to the site of vascular injury, where tissue factor is expressed and activated platelets are found. In pharmacological doses, rFVIIa produces localized thrombin generation and fibrin clot formation.7–9
In a small, single-center study, rFVIIa was found to be safe in patients undergoing OLT. Patients who received rFVIIa 80 μg/kg prior to surgery required fewer transfusions than matched historical controls from the same institution.10 Additionally, rFVIIa administration enhanced thrombin generation in a localized, time-dependent manner and did not lead to systemic coagulation or fibrinolysis.11
A multicenter, placebo-controlled trial investigating a single dose of rFVIIa (20, 40, or 80 μg/kg) did not demonstrate any significant effect of rFVIIa administration on total red blood cell (RBC) transfusion requirements or blood loss due to insufficient dose.12 Based on this previous trial, the current study was designed and conducted administering a higher dose of rFVIIa at regular intervals during the surgical procedure.
Materials and Methods
Study Design and OLT Procedure
The aim of this multicenter, randomized, placebo-controlled, double-blind, exploratory trial was to evaluate the hemostatic efficacy of repeated doses of rFVIIa 60 or 120 μg/kg in patients undergoing OLT. Patients were randomized to 1 of 3 parallel study arms, receiving a repeated intravenous bolus regimen of either rFVIIa 60 or 120 μg/kg or placebo.
The first bolus dose was administered within 10 minutes of the first skin incision. Bolus doses were repeated every 2 hours during OLT until approximately 30 minutes prior to the expected reperfusion of the transplanted liver. A final bolus dose was administered at the completion of wound closure.
OLT was conducted in accordance with institutional standard procedures. The OLT procedure was evaluated in total and was divided into preanhepatic, anhepatic, and postanhepatic phases. Prophylactic administration of hemostatic agents was not permitted during OLT but could be carried out during the procedure when clinically indicated. Hemostatic agents, including antifibrinolytic therapies at routine dosages, were administered according to hospital procedures. Transfusion of blood products during surgery was restricted to the following guidelines. RBC transfusions were allowed when hematocrit levels fell below 25%. Fresh-frozen plasma transfusions were permitted when the international normalized ratio was >1.5 or activated partial thromboplastin time was >1.5 times control. Platelet concentrate administration was permitted when platelet counts fell below 30,000/mm2. Cryoprecipitate or fibrinogen concentrate administrations were permitted when fibrinogen levels were <1.0 g/L. Fresh- frozen plasma, plasma concentrate, cryoprecipitate, and fibrinogen concentrate transfusions could also be administered as indicated by thromboelastogram tracings. The protocol was approved by the local ethics committee of each participating center, and each patient signed a written informed consent.
Patients were eligible for enrollment if they were scheduled to undergo OLT because of cirrhosis, defined as Child-Turcotte-Pugh class B or C,13 and were ≥18 years of age. Exclusion criteria included previous liver transplantation, split liver transplantation, scheduled multiorgan transplantation, scheduled living related-donor transplantation, renal insufficiency requiring dialysis, documented coagulation disorders, and documented history or presence of portal vein thrombosis.
Donor Liver Data
Donor liver data were collected regarding donor-recipient blood group match, donor age and gender, cold and warm ischemia time, qualitative evaluation of the donor liver (prior to and after reperfusion), surgical method, surgical complications that increased bleeding risk, and the use of an autologous salvage transfusion system.
Efficacy and Safety Assessments
The primary efficacy endpoint was the number of units of allogeneic and autologous RBC transfused during the perioperative period, defined as the time during the OLT procedure and in the 24-hour period postsurgery.
The secondary endpoints were measured during the perioperative period: units of fresh-frozen plasma, milliliters of platelet concentrate, blood loss, and the crystalloids or colloids administered. Other secondary endpoints were changes in hematocrit from baseline to completion of surgery, and from baseline to 24 hours after wound closure; the number of patients receiving treatment with antifibrinolytic drugs during transplantation; time of surgery; length of intensive care unit stay; and length of hospital stay.
The safety analysis consisted of a physical examination, assessment of vital signs, and a 12-lead electrocardiogram performed on the day of treatment prior to trial product administration and 3 days postsurgery. A physical examination was also performed at 24 hours and day 7 postsurgery. Vital signs were recorded at the conclusion of surgery (wound closure) and at day 3 postsurgery. Doppler ultrasonography of hepatic vessels was performed at 24 hours, day 3, and day 7 postsurgery to detect evidence of thromboembolism. Additional Doppler ultrasonography studies were performed in the event of suspected hepatic thrombosis.
The frequency and type of adverse events (AEs), including thromboembolic events and bleeding complications, were recorded.
The analyses completed at the central laboratory were plasma factor VII clotting activity and prothrombin time (PT). Local hospital laboratories performed hematocrit, hemoglobin, platelet count, and screening analyses. Blood samples were collected to determine plasma factor VII clotting activity and PT at the following time points: on the day of treatment prior to trial product administration; 15 minutes after product administration; at 1-hour intervals during OLT until the completion of surgery; and at 24 hours, day 3, and day 7 postsurgery.
The primary efficacy endpoint was the total number of RBC units transfused during the perioperative period. The sample size calculation was based on y = log(1 + x), where x is RBC transfusion requirements from previous rFVIIa data in OLT patients. We assumed the mean and intracenter standard deviation of y were 2.48 and 0.83, respectively. Based on 80% power to detect a significant difference (P < 0.05, 2-sided) of 40% reduction between the rFVIIa 120-μg/kg and the placebo study group, 60 patients were required for each study group. All statistical analyses were performed for the intent-to-treat population, defined as those patients completing at least the preanhepatic phase of OLT and receiving at least 1 dose of study product.
The primary analysis was performed using a nonparametric trend analysis (Jonckheere-Terpstra test).14 Jonckheere-Terpstra tests were also used to analyze total perioperative transfusion requirements of fresh-frozen plasma, crystalloid or colloid fluid replacement volume, surgical drain volume, calculated intraoperative blood loss volume, total surgery time, total intensive care unit days, and total hospital days. Fisher's exact test was used to compare study groups with respect to the number of patients requiring transfusion products and antifibrinolytic drugs. Changes in hematocrit values and coagulation-related parameters were evaluated with F tests. The total number of patients experiencing AEs related to thrombosis, hemorrhagic complications, or mortality was compared between study groups using Fisher's exact test. Separate analyses were performed for thrombosis of hepatic vessels, all episodes of thrombosis excluding those of hepatic vessels, and all episodes of thrombosis combined. Two-sided tests using the 5% critical level were used for all statistical calculations.
Fourteen investigative sites enrolled patients scheduled for OLT from August 2001 to September 2003. The study groups were comparable with respect to patient demographics and baseline characteristics (Table 1). A flow diagram of patient enrollment and randomization is shown in Figure 1. A total of 219 patients were screened for the study, 10 of whom were determined to be screening failures. Of the remaining 209 patients, 26 were withdrawn because of failure to meet inclusion / exclusion criteria on the day of treatment. One patient received trial product but did not complete the preanhepatic phase of surgery. The efficacy analysis, therefore, included 182 patients and the safety analysis included 183 patients. Of the 182 patients in the intent-to-treat population, 172 completed the observation period. The majority of patients (57%) received 3 doses of study product, with 30 (49%) receiving placebo, 41 (65%) receiving 60 μg/kg rFVIIa, and 33 (57%) receiving 120 μg/kg rFVIIa. Only 1 patient received 6 doses of 60 μg/kg rFVIIa.
|Placebo (n = 61)||rFVIIa 60 μg/kg (n = 63)||rFVIIa 120 μg/kg (n = 58)|
|Age, mean (SD) years||52.3 (11.5)||53.3 (11.2)||52.6 (9.2)|
|Gender, n (%)|
|Male||33 (54)||36 (57)||41 (71)|
|Female||28 (46)||27 (43)||17 (29)|
|aPTT, seconds, mean (SD)||61.8 (18.8)||57.7 (19.1)||64.9 (27.2)|
|Fibrinogen, g/L mean (SD)||2.1 (1.0)||2.2 (1.4)||2.2 (1.2)|
|Platelet count × 109/L, mean (SD)||120.1 (76.9)||122.0 (89.0)||107.7 (75.6)|
|Primary OLT indication, n (%)|
|Hepatitis C cirrhosis||11 (18)||15 (24)||15 (26)|
|Hepatitis B cirrhosis||1 (2)||1 (2)||4 (7)|
|Alcoholic cirrhosis||19 (31)||18 (29)||17 (29)|
|Primary biliary cirrhosis||8 (13)||10 (16)||5 (9)|
|Secondary biliary cirrhosis||0 (0)||1 (2)||0 (0)|
|Other||22 (36)||18 (29)||18 (31)|
|Previous abdominal surgery, n (%)||29 (48)||36 (57)||20 (34)|
|Child-Turcotte-Pugh score, n (%)|
|A||1 (2)||0 (0)||0 (0)|
|B||32 (52)||37 (59)||38 (67)|
|C||28 (46)||26 (41)||20 (33)|
Donor Liver and Surgery Characteristics
No notable differences between study groups were found with respect to surgical variables, cold and warm ischemia times of the graft, operation duration, surgical method, or the use of an autologous transfusion system.
No statistically significant difference (P > 0.05) was found for any parameter of transfusion requirements (Table 2). However, a significantly higher number of patients in each rFVIIa study group avoided RBC transfusions when compared with the placebo group (Fig. 2). The study groups did not differ significantly with regard to the number of patients requiring antifibrinolytic agents, the requirements for crystalloid and colloid replacement therapy, or blood loss. Intensive care unit stays in the placebo, rFVIIa 60-μg/kg, and rFVIIa 120-μg/kg study groups were 3.0, 3.5, and 3.0 days, respectively (P > 0.05). Hospitalization days in the placebo, rFVIIa 60-μg/kg, and rFVIIa 120-μg/kg study groups were 17.0, 19.0, and 22.0, respectively (P > 0.05).
|Placebo (n = 61†)||rFVIIa 60 μg/kg (n = 62‡)||rFVIIa 120 μg/kg (n = 56§)|
|Median total RBCs, U (range)||8.2 (1.5–100.0)||7.0 (0.0–76.5)||6.3 (0.0–76.4)|
|Median total FFP, U (range)||11.0 (0.0–87.2)||9.4 (0.0–62.4)||11.9 (0.0–52.8)|
|Median PC, mL (range)||141.7 (0.0–1953.0)||81.8 (0.0–2640.0)||170.6 (0.0–886.0)|
|Other transfusion product (eg, CP, FC), n (%)||17 (28)||23 (37)||24 (43)|
No notable differences were observed between study groups with respect to AEs or serious AEs. The most commonly reported AEs included hyperglycemia, constipation, intraoperative hemorrhage, and renal impairment. A total of 59 serious AEs occurred in 45 patients, including 6 deaths (Table 3). There were 27 thromboembolic events in 25 patients.
|Placebo (n = 62) n (%)||rFVIIa 60 μg/kg (n = 63) n (%)||rFVIIa 120 μg/kg (n = 58) n (%)|
|Number of patients experiencing serious AEs||12 (19)||17 (27)||16 (28)|
|Thromboembolic events||6 (10)||12 (19)||7 (12)|
|Blood and lymphatic system disorders||0 (0)||0 (0)||1 (2)|
|Cardiac disorders||1 (2)||5 (8)||3 (5)|
|Congenital, familial, and genetic disorders||0 (0)||0 (0)||1 (2)|
|Gastrointestinal disorders||1 (2)||0 (0)||1 (2)|
|Hepatobiliary disorders||5 (8)||3 (5)||5 (9)|
|Immune system disorders||1 (2)||3 (5)||0 (0)|
|Infections and infestations||0 (0)||0 (0)||1 (2)|
|Injury, poisoning, and procedure complications||4 (6)||3 (5)||3 (5)|
|Investigations||0 (0)||0 (0)||1 (2)|
|Metabolism and nutrition||0 (0)||0 (0)||1 (2)|
|Nervous system disorders||0 (0)||1 (2)||0 (0)|
|Renal and urinary disorders||1 (2)||2 (3)||0 (0)|
|Respiratory, thoracic, and mediastinal disorders||1 (2)||1 (2)||2 (3)|
|Vascular disorders||1 (2)||2 (3)||0 (0)|
|Deaths||1 (2)||1 (1)||2 (3)|
There was no statistically significant effect of rFVIIa treatment observed for the proportion of patients experiencing thromboembolic events. Thromboembolic AEs were observed in 10% of patients in the placebo group, in 19% of the rFVIIa 60-μg/kg group and in 12% of the rFVIIa 120-μg/kg group. The number of thromboembolic events found in this trial is comparable with other reports in the literature.10–12
The mean PT was similar between study groups prior to trial product administration. After product administration, the PT decreased to normal value in the rFVIIa 60-μg/kg and rFVIIa 120-μg/kg study groups, but not in the placebo group. There was no difference in the PT between the rFVIIa 60-μg/kg and rFVIIa 120-μg/kg study groups, and the 120-μg/kg study group had a reduced PT level toward normal range with the last dose. Mean PT levels returned to baseline levels in the rFVIIa study groups at 7 hours from the first dose (Fig. 3). The plasma factor VII clotting activity level increased proportionally with rFVIIa dose and approached baseline levels within 5 hours of the first rFVIIa dose (Fig. 4).
We report the preliminary experience with multiple perioperative doses of rFVIIa in patients with a Child-Turcotte-Pugh class of B or C undergoing OLT. Although there were no statistically significant differences found between placebo and rFVIIa with respect to the primary endpoints, a number of patients receiving rFVIIa avoided RBC transfusions completely.
The coagulation values observed in the study population were outside the range for healthy volunteers; such abnormal coagulation values are consistent with the critical state and impaired hemostatic function of patients with severe liver disease. Recombinant FVIIa has demonstrated the ability to improve hemostasis and was investigated in this patient population in an attempt to reverse the coagulopathy associated with cirrhotic disease that leads to excessive blood loss during OLT.10
The study protocol defined a significant effect on blood loss as a 40% reduction in RBC transfusion requirements. The improvements in surgical procedures and the extensive training received by surgeons have led to improved clinical outcomes, including reductions in bleeding rates.15 These improvements have made a 40% reduction in RBC transfusion a difficult goal to achieve. In our study, we found a 15% reduction of RBC transfusion for patients receiving the 60-μg/kg dose and a 23% reduction for patients in the 120-μg/kg group compared with that of the placebo group. This finding was not sufficient to reach statistical significance, as the study was powered to show a 40% reduction of RBC transfusion. The avoidance of transfusions during OLT is a clinically relevant goal, as excessive transfusion requirements are associated with adverse outcomes, including higher mortality,1–3, 5, 6 perioperative multiorgan dysfunction,1, 6 and reduced graft survival.2, 5 Additionally, multiorgan failure and adult respiratory distress syndrome were reduced in blunt trauma patients when rFVIIa was administered and compared with patients who received a placebo.16 A reduction in the number and rate of transfusions required during OLT also lowers the possible risk of blood-borne disease transmission and deleterious immunomodulation associated with transfusion therapy.6, 17–19
Although a previous trial of single-dose rFVIIa failed to demonstrate an effect on any parameter of blood loss or transfusion requirements,12 we found in this study that a significantly higher number of patients treated with rFVIIa avoided transfusions compared with those receiving placebo. This effect might have been due to the higher-dose and multiple-dose regimens employed in our study. Our trial was limited by the parameters of achieving a 40% reduction in RBC transfusion requirements. The doses in this study were selected based on previous data in OLT patients10, 11 and a computer-generated pharmacokinetic simulation,20 which led to the choice of 120-μg/kg dosing.
To fully understand the utility of rFVIIa and its hemostatic effect in reducing the associated complications of blood transfusions, it will be important to consider the costs and benefits of its administration. A complete pharmacoeconomic analysis that includes an examination of data points such as the rate and costs of complications (e.g., multiorgan dysfunction), mortality, morbidity, length of hospitalization, and number of intensive care unit days, will need to be considered together with the cost of rFVIIa. This phase II trial was not designed to include the collection of data for this type of analysis. However, additional phase III studies will evaluate this information for its clinical usefulness.
Preliminary safety of rFVIIa in OLT patients has been demonstrated in previous trials10–12 and was confirmed by our findings. No significant differences between study groups were observed in the incidence of AEs, including thromboembolic AEs. There was no statistically significant effect of rFVIIa treatment on hospitalization rate, total surgery time, and the proportion of patients undergoing retransplantation.
The administration of rFVIIa may be useful because it results in localized thrombin generation and does not lead to systemic coagulation or fibrinolysis.7–9 It might therefore be an alternative to other agents currently used for blood loss management. However, trials to establish the minimal effective dosing regimen and pharmacoeconomics are needed.
Additional trials are warranted to confirm whether rFVIIa administration offers a clinical hemostatic benefit in selected patients undergoing OLT, as it has been shown to be beneficial in cirrhotic bleeding patients not undergoing surgery21 and in the small-scale study of patients undergoing OLT.10 The selection of patients should include those with a high risk of previous upper abdominal surgery, high preoperative Model for End-Stage Liver Disease (MELD) score, retransplantation, portal vein thrombosis, and renal impairment. This trial found that the highest RBC transfusion reduction occurred in the 120-μg/kg group. Therefore, future trial design should also consider the selection of patients who have a high risk of bleeding due to a nonsurgical cause and a repeated dosing schedule of 120 μg/kg to account for the high clearance of rFVIIa. Additionally, avoidance of RBC transfusion needs to be examined, as this is a more meaningful measure of overall outcomes as well as a pharmacoeconomic analysis.
A significant number of patients receiving rFVIIa avoided RBC transfusions completely, although rFVIIa treatment did not show a significant effect on the number of RBC transfusions required during OLT. There was no increase in thromboembolic events with rFVIIa administration compared with placebo. Further trials are needed to complete a pharmacoeconomic analysis and to confirm a potential hemostatic benefit of rFVIIa in patients undergoing OLT.
- 14Nonparametric statistical methods. New York: John Wiley & Sons; 1973., .
- 16A randomized, placebo-controlled, double-blind study to investigate the efficacy and safety of rFVIIa as adjunctive therapy for control of bleeding in patients with severe blunt trauma: a re-analysis following the exclusion of early (<48 hours) deaths. Paper presented at the 25th Annual Meeting of Intensive Care and Emergency Medicine, Brussels, Belgium, March 21–25, 2005., , , , , , et al.