The Effect of Intravenous Administration of Active Recombinant Factor VII on Postoperative Bleeding in Cardiac Valve Reoperations; A Randomized Clinical Trial

Cardiac surgeries are among the most complicated surgical procedures worldwide with significant use of blood products. Postoperative bleeding and blood transfusion are among their major clinical challenges especially in redo surgeries. Therefore, we have to find more blood salvaging strategies. The cause of post-cardiac surgery bleeding can be either surgical or nonsurgical (1-4). Microvascular diffuse bleeding after cardiac surgery is still considered as an important and common problem leading to the use of antifibrinolytic drugs; however, the efficacy of these drugs is not always the same (1-10).

Disturbances of the hemostatic system are more prominent in redo cardiac surgeries (6-9) including more consumption of blood products, which could be associated with significant morbidity and mortality. Hence, the effort to obtain safe and effective hemostatic condition can improve patient outcome (3, 4, 11-18). Among the many strategies to reduce blood transfusion, recently, active recombinant factor VII (rFVIIa) has been proposed (19-22). Some studies claimed it effective in reducing postoperative bleeding, e.g. in trauma patients (23-26). In cardiac surgery, although promising, rFVIIa is still "off-label" and not among FDA-approved indications, mandating more clinical trials (4, 14-29).

One of the outcome measures for considering the efficacy of rFVIIa is to assess its use as a prophylactic agent (5, 28). However, there is not enough evidence to support prophylactic use of rFVIIa in cardiac surgery and its use still remains unclear, mandating randomized controlled trials to establish the efficacy of prophylactic rFVIIa in cardiac surgery (30).

In this randomized double-blinded clinical trial, we assessed the effect of using rFVIIa on the amount of postoperative bleeding in redo cardiac valve surgeries in adult patients.

The study design and its clinical steps were approved by the Ethics Committee of Research Deputy of Shahid Beheshti University of Medical Sciences, Tehran, Iran. Afterwards, throughout the study, the participants entered the study after written informed consent was explained for them and read and signed by them. In this randomized double-blinded clinical trial, 36 patients entered the study during a 12-month period and randomly allocated to either "rFVIIa" or "placebo" groups. Randomization of the patients into two groups was done using randomization table. All the patients underwent redo cardiac valve surgery by the same surgeon. All participants were in the age range of 20 to 70 years and their redo operation included cardiopulmonary bypass. The following conditions were considered as exclusion criteria:

• - Underlying renal disease,

• - Underlying cerebral vascular disease,

• - Use of heparin or anti-platelet drugs up to the time of operation,

• - History of deep vein thrombosis,

• - Underlying hematologic disorders,

• - Any carotid plaque creating significant stenosis (i.e. stenosis > 75%),

• - Underlying pulmonary disorders affecting the spirometry indices significantly, and

• - Emergent or urgent surgery.

To ensure the blinding process in administration of rFVIIa (AryoSeven, AryoGen Biopharma, Tehran, Iran) or placebo, a nurse in recovery room prepared the solution and brought it to operation room. Therefore, none of the colleagues was aware of the combination of the solutions. All the patients were visited the night before surgery by the same colleague. During this visit, the patients were informed of the study and its conditions and that they were free to leave the study due to any cause at any stage. The morning of the surgery day, the patients entered the study.

The day of the operation and after entering the operating room, each patient was monitored by five-lead electrocardiography, pulse oximetry, invasive blood pressure monitoring through left radial artery, and cerebral oximetry. After inducing general anesthesia, central venous pressure was monitored through a right internal jugular central venous catheter (CVC). All the patients were anesthetized using a combination of 1 to 3 mg/kg of sodium thiopental, 5 µg/kg of fentanyl, and 0.2 mg/kg of cisatracurium. Laryngoscopy and intubation was performed after three minutes and endotracheal tube was fixed.

All the patients were operated using the same technique including femoral cannulation for arterial and venous perfusion. Afterwards, with sternal opening, the adhesions were relieved and cardiopulmonary bypass was started after heparinization with 300 mg/kg of intravenous heparin, checked with an Activated Coagulation Time (ACT) > 480 seconds. After valve replacement and closure of the cardiac chambers and gaining a nasal temperature of 36.5℃, weaning from cardiopulmonary bypass was performed. After stabilization of the hemodynamic status and full mechanical ventilation, weaning was gradually completed and heparin reversal was done with a full dose of protamine considering the heparin rebound effect. Surgical control of bleeding (including control of all suture lines for potential surgical bleeding) was done. At this time, the study drug was administered as the protocol: 40 µg/kg of rFVIIa as a 5-mL solution in rFVIIa group and 5 mL of sterile water in placebo group. Both solutions were administered through CVC line. Afterwards, the sternum was closed using steel wires. The patient was transferred to cardiac ICU and extubated after full recovery and hemodynamic stability. None of the patients in any of the groups received any systemic or topical hemostatic agent. Moreover, none of them received further or repeated doses of rFVIIa.

The primary measured outcomes were the volume of postoperative bleeding, incidence of reoperation, blood transfusion, and potential thromboembolic events. Therefore, during the postoperative period, the following variables were monitored: volume of chest tube drainage; need for administration of blood products based on the study protocols for blood transfusion; prothrombin time (PT); partial thromboplastin time (PTT); hemoglobin and hematocrit; platelet count; and international normalized ratio (INR).

All the above mentioned variables were monitored during the first 24 hours of surgery while bleeding was checked at third, 12th, and 24th hour after surgery. The patients were screened for thromboembolic events by physical examination of the entire body. All patients received blood and blood products after bleeding. A target hematocrit of 27% was maintained in the postoperative period. Indications for transfusion of fresh frozen plasma were INR > 1.5 or plasma activated PTT > 50 seconds; the indication for platelet transfusion was platelets < 100 × 10⁹ cells/L.

Heparin infusion was begun in all patients six hours after ICU entrance to preserve PTT around 60 seconds. Furthermore, transthoracic echocardiography was done for cardiac valve assessment in all patients.

The majority of our patients were female (63%) with a female to male ratio of 63:37; however, the two groups had no statistically significant difference regarding sex ratio. Mean age was 49.4 ± 16.1 years and mean weight was 66.6 ± 10.9 kg. In addition, there was no statistical difference between the two groups regarding age and weight. The difference between the two groups regarding blood products transfusion was statistically significant; three patients in rFVIIa group needed blood products (16.67%) while 13 patients in the placebo group received blood products (72.23%) (P < 0.01).

During the first three hours after surgery, mean chest tube blood drainage was 62.8 ± 8 mL in rFVIIa and 79.4 ± 12 mL in placebo groups (P = 0.71). During the first 12 hours after surgery, mean chest tube blood drainage was 160.6 ± 56 mL in rFVIIa and 293.3 ± 92 mL in placebo groups (P = 0.22). Finally, during the first 24 hours after surgery, mean chest tube blood drainage was 315 ± 177 mL in rFVIIa and 557 ± 168 mL in placebo groups (P = 0.03). The two groups were similar regarding mean of preoperative PT; however, mean of postoperative PT was significantly lower in rFVIIa group compared with the placebo group (13.8 ± 1.8 vs. 15.8 ± 3 seconds; P = 0.02). However, there was no statistically significant difference between two groups in mean of preoperative and postoperative PTT and INR.

The mean of surgery preoperative platelet count was not statistically different between two groups; However, the mean of postoperative platelet count was higher in rFVIIa group compared with placebo group (5551 ± 875 vs. 4191 ± 856; P = 0.03). Nonetheless, there was no statistically significant difference between the two groups regarding the mean of preoperative and postoperative hematocrit or hemoglobin levels. Although two groups were similar regarding ICU stay, the mean time to extubation was longer in the placebo group compared with rFVIIa group (352 ± 57 vs. 287 ± 46 minutes; P = 0.003). No thromboembolic complications including potential central nervous system-related or deep venous embolic events were noticed during postoperative clinical follow-ups.

Postoperative bleeding after redo cardiac surgery is an important potential complication with resource utilization. Replacement of bleeding with blood products is associated with increased morbidity and mortality. Hence, any safe intervention to preserve hemostasis status without systemic activation of the coagulation cascade is appreciated. However, a substantial body of evidence has demonstrated a strong association between transfusion of blood and its products with adverse outcomes, leading us to a number of blood salvaging strategies for decreasing the need for extra blood transfusion.

The results of this study demonstrated that administration of single dose of 40 µg/kg of intravenous rFVIIa in patients undergoing redo cardiac valve surgery could be effective in reducing the volume of postoperative chest tube drainage and need to blood transfusion, with no thromboembolic event. Prophylactic use of rFVIIa is suggested to be benefit in redo cardiac surgeries when the risk of coagulopathic bleeding and need for transfusion is high. In addition, some of the lab indices for coagulation were improved in the rFVIIa group.

Thus far, prophylactic use of rFVIIa in redo cardiac valve surgery has not been assessed in prospective trails. At first, rFVIIa was approved for treating hemorrhagic events in hemophilic patients in 1999 and now its indications are extended to handle and even prevention of bleeding in a wide spectrum of surgical procedures (31). There is no consensus and guideline regarding dosage, time, or number of administration; therefore, physicians have individualized the decision on administrating rFVIIa in each patient (31, 32).

In our study, the trend in the amount of bleeding measured as "chest tube drainage" demonstrated the difference between two groups increase by time. In one study, administration of rFVIIa could decrease bleeding in patients on extracorporeal membrane oxygenation (33). Therefore, the results of our study were in concordance with some other studies in which rFVIIa was used as a prophylactic agent, although none were in redo cardiac surgeries. In addition, as mentioned in other studies, rFVIIa could be used both as a prophylactic agent and as a "rescue therapy" in order to treat refractory bleeding; however, there are some controversies on that regard (1, 5, 10, 17, 18, 28, 29, 34-39). Gill et al. performed a randomized placebo-controlled trial in those with bleeding after cardiac surgery; they observed more critical adverse events in the rFVIIa groups (40 and 80 µg/kg) than in placebo group although it was statistically insignificant. They stated that some of patients in rFVIIa groups were older and had longer on-pump duration with more transfusions before randomization (40).

von Heymann et al. used rFVIIa (60-80 µg/kg) as the last resort in 24 adults with refractory bleeding despite extensive use of blood products and medications after a median of 14 hour from surgery; they observed no additional efficacy over usual hemostatic therapy (41). It might be partly interpreted that early administration of drug could be more effective in hemostatic enhancement than it is used as the rescue treatment after optimal conventional methods. Some studies reported better outcome when it was given early in the course of blood loss in comparison to when it was administrated as salvage medication (42-47). In one cohort study, it was shown that in cardiac surgery, the response rate to control of bleeding and thromboembolic events were similar with either ≤ 40 µg/kg or ≥ 100 µg/kg doses (4, 48).

The most concern on rFVIIa safety is potential to inappropriate clotting in coronary, grafts, and cerebral arteries; it looks to be dose dependent and higher in older subjects (5, 49). Reported thromboembolic complications after use of rFVIIa have been shown to be outside of the cardiac surgery setting; however, retrospective studies, and RCT studies failed to show increased rate of thromboembolic events in treated patients (50, 51). Roberts et al. in a study on the safety profile of rFVIIa across all specialties in over 400000 cases (mostly, patients with hemophilia) expressed the risk of serious adverse events to be < 1% (52). Chapman et al. retrospectively analyzed the safety profile and efficacy of rFVIIa used for bleeding following cardiac surgery in 236 patients and compared the results with that of 213 controls. Results showed a reduction in the requirement of additional blood products in rFVIIa group and no significant difference in the incidence in thromboembolic events or acute renal failure between two groups (53).

Repeated cardiac surgeries pose high morbidity risks, e.g. bleeding needing aggressive replacement; however, sometimes rescue therapy is not effective, making the benefits of prophylactic rFVIIa more than its risks. According to the study of Karkouti et al. the benefits of using rFVIIa might outweigh its risks; Chapman et al. mentioned similar findings (53, 54).

Although the prophylactic role of rFVIIa has not been determined in other studies (5, 30), our study demonstrated the efficacy of rFVIIa, in controlling postoperative bleeding in redo cardiac surgeries, predominantly by assessment of subsequent blood loss as determined by chest tube drainage and transfusion requirement. Therefore, it seems that rFVIIa is a potent agent in controlling postoperative bleeding in redo cardiac valve surgeries, especially when rFVIIa is used as a prophylactic agent. However, like other studies (55), we could not conclude whether rFVIIa could decrease mortality in our patients.

Comparative studies of rFVIIa in adult cardiac surgery in nine separate studies had inconsistent outcomes in means of blood transfusion although most of them had retrospective design with a wide range of rFVIIa dosage (10-200 µg/kg). The similar finding was that patients had experienced major bleeding and required significant blood transfusion before administrating rFVIIa. Our study suggested that its prophylactic use was associated with less postoperative bleeding and transfusion requirement.

In most studies on cardiac surgeries, rFVIIa was used after extensive use of blood products, aprotinin, and other antifibrinolytic agents was ineffective to stop bleeding. The massive blood transfusion and suboptimal hemodynamic status can increase the risk of myocardial infarction, stroke, and multiple end-organ failure. Hence, some of this concern might be related to critical patients’ condition. Indeed in patients with refractory life-threatening bleeding after failure of conventional hemostasis methods, any efforts might be worthless, mandating some more prophylactic methods including prophylactic use of rFVIIa (1, 2, 5, 20, 23, 44, 56). The administration of rFVIIa is considered as "not unreasonable for the management of intractable nonsurgical bleeding that is unresponsive to routine hemostatic therapy after cardiac procedures using CPB" based on the common Clinical Practice Guideline declared by The Society of Thoracic Surgeons and The Society of Cardiovascular Anesthesiologists in 2007 (57).

Another finding of the study was decreased time for extubation in rFVIIa group, which could be related to the effects of increased blood products usage in control group. This finding could be attributable to the adverse effects of blood products on lung function (58-61). There is a general agreement to restrict blood transfusion in cardiac surgery; therefore, the benefits of rFVIIa might outweigh the risks, especially if it is given before the onset of vicious cycle of bleeding and transfusion (62). The cost and associated risks of both massive blood transfusion and re-exploration seems to overweigh that of administrating rFVIIa alone (4, 23, 24, 39). Nevertheless, rFVIIa should be viewed as a life-saving agent not a complete replacement for other lifesaving medical or surgical interventions; in fact, but rFVIIa is a complement to the previous array for control of bleeding.

The present study has some limitations that need to be considered. It was a small single-center study that could not individually reach to final decision and needed more complementary studies. This study did not assess the cost-effectiveness of rFVIIa. Moreover, because of budget limitations, this study did not measure the level of factor before its intravenous administration. We tried to modify confounding factors as much as possible while there might be some clinical variables such as cardiac or hepatic function that were not evaluated in our study.

This study had some strength too. This study was a prospective, placebo-controlled, randomized trial included a homogeneous patient population who were at high risk of bleeding, and this allowed assessing the effectiveness of rFVIIa with greater precision. Moreover, the dosage and time of rFVIIa therapy were similar among all patients. The study was without intervention of other hemostatic agents; therefore, it could show the cause-effect relationship between rFVIIa and blood loss.

The early administration of rFVIIa in the setting of redo cardiac surgeries could be an effective effort to decrease postoperative nonsurgical bleeding safely. Further studies with a large population of cardiac surgery patients are suggested to confirm its safety and efficacy.