Usefulness of Intraoperative Continuous Infusion of Tranexamic Acid during Emergency Surgery for Type A Acute Aortic Dissection

Abstract

Purpose: We investigated the influence of intraoperative continuous tranexamic acid (TA) infusion on the amount of blood transfusion required in emergency surgery for type A acute aortic dissection.

Methods: The study was based on the data of 55 consecutive patients who underwent surgery for type A acute aortic dissection. The patients were divided into 2 groups for comparison: Group T, consisting of 26 patients who received intraoperative continuous infusion of TA, and Group N, consisting of 29 patients who did not receive TA infusion during the surgery.

Results: The mean amounts of blood transfusion required during and after surgery were compared between the 2 groups: they were 10.5 ± 8.7 and 16.2 ± 10.0 units of mannitol-adenine-phosphate-added red cell concentrate, 9.3 ± 8.6 and 17.1 ± 10.0 units of fresh frozen plasma, and 20.4 ± 12.2 and 29.7 ± 14.9 units of platelet concentrate, respectively, in Groups T and N. Thus, the amount of each of these blood products required was significantly reduced in Group T.

Conclusions: During emergency surgery for type A acute aortic dissection, continuous infusion of TA resulted in a significant reduction in the amount of blood transfusion required.

Introduction

Stanford type A acute aortic dissection is one of the diseases requiring emergency surgery. In acute aortic dissection, a large number of thrombi are rapidly formed in the false lumen immediately after onset, which activates the fibrinolytic system. Moreover, the surgery is performed using extracorporeal circulation with haparin, under hypothermic conditions. Both use of extracorporeal circulation and hypothermia are known to activate the fibrinolytic system. Thus, perioperative bleeding is a major problem in surgery for acute aortic dissection, often necessitating blood transfusion in large amounts during surgery. Aprotinin, which had been used to reduce intraoperative bleeding, was reported to increase the hospital mortality,¹⁾ and the drug had to be withdrawn from the market in 2007. Then, tranexamic acid (TA), which has been used as a hemostatic drug for a long time, began attracting attention again as an alternative. In the present study, we examined the usefulness of TA administered as a continuous infusion during surgery for acute aortic dissection.

Materials and Methods

We retrospectively studied data from 55 consecutive adult patients aged 18 years or older who underwent surgery for acute aortic dissection (Stanford type A) at our hospital between April 2008 and April 2010. The mean age of the patients was 67.3 ± 12.2 years. There were 24 men and 31 women. The patients were divided into 2 groups for comparison: Group T, consisting of 26 patients who received intraoperative continuous infusion of TA (mean age, 67.2 ± 12.6 years; 13 men and 13 women), and Group N, consisting of 29 patients who did not receive intraoperative TA infusion (mean age, 67.3 ± 12.1 years; 11 men and 18 women). (Surgery was performed between April 2009 and April 2010 in Group T and between April 2008 and May 2009 in Group N. From April to May 2009, TA was infused only when a certain anesthesiologist was in attendance during the surgery.) In all patients, the chest was opened through a median sternotomy under general anesthesia, and implantation of the blood vessel prosthesis was performed with the support of cardiopulmonary bypass under cardiac arrest. In principle, hemi-arch replacement was performed. However, if the entry tear was confirmed to be located at or below the level of the aortic arch, partial or total arch replacement was selected as necessary, based on the need for resection of the entry tear. For the proximal side, in case of the dissection reaching the coronary ostia or of concomitant severe aortic regurgitation, the Bentall procedure was performed as the procedure of first choice. Prior to the cardiopulmonary bypass, heparin was administered at the smaller dose of 400 units per kilogram body weight or 9000 units per square meter body surface area, to maintain the target activated clotting time (ACT) of 400 s or over. In all patients, cardiopulmonary bypass was established after the median sternotomy, by inserting a perfusion cannula into the true lumen of the ascending aorta and a return cannula into the right atrium under ultrasound guidance. Then, cooling was started. When the bladder temperature reached 28°C, circulation was arrested. The aorta was incised, and selective anterograde cerebral perfusion was performed under direct vision. At the end of the cardiopulmonary bypass, heparin was neutralized with protamine to achieve a target ACT of 150 s or lower. After the start of surgery, infusion of TA was started at the dose of 16 mg/kg/hr before incision of the pericardium and discontinued at the end of surgery. However, the maximum dose was set at 1000 mg/hr. Groups T and N were compared in respect of the following items: preoperative test results, past history, total amounts of blood transfusion during and after the surgery, duration of stay at the intensive care unit (ICU), and duration of hospital stay. The cut-off value of hematocrit (Ht) required to initiate transfusion of RCC was determined to be <24% during surgery and <30% postoperatively. In addition, fresh frozen plasma was administered at a ratio of 1:1 against the transfused red cell concentrate (The final dose was determined taking into consideration the circulating plasma volume). Administration of platelet concentrates was indicated when the platelet count was ≤50000/µL. The results were expressed as mean ± standard deviation. For the statistical analysis, the student t-test or χ² test was performed using the SPSS. P <0.05 was considered to indicate a statistically significant difference.

Results

The preoperative patient background characteristics are shown in Table 1. No significant differences between the two groups were observed in respect of the age, sex, or preoperative test results on Ht and platelets. Furthermore, there were no significant differences in the previous medical history between the two groups. The surgical procedure performed was ascending aorta replacement in 15 patients in Group T and 21 patients in Group N, partial arch replacement in 3 patients in Group T and 2 patients in Group N, total arch replacement in 6 patients in Group T and 6 patients in Group N, and the Bentall methods in 2 patients in Group T and no patient in Group N (Table 2). The mean amounts of blood transfusion during and after the surgery were compared: they were 10.5 ± 8.7 and 16.2 ± 10.0 units of red cell concentrate (P = 0.029), 9.3 ± 8.6 and 17.1 ± 10.0 units of fresh frozen plasma (P = 0.0018), and 20.4 ± 12.2 and 29.7 ± 14.9 units of platelet concentrate (P = 0.015), respectively, in Group T and Group N. Thus, the total amounts of all the blood products transfused were significantly reduced in Group T (Fig. 1).

Table 1.

Patient characteristics

	T group	N group	P
age	67.2 ± 12.6	67.3 ± 12.1	(NS)
male	13/26	11/29	(NS)
Ht (%)	34.66 ± 6.22	35.18 ± 5.28	(NS)
Plt (*104/dl)	16.85 ± 8.62	15.96 ± 7.00	(NS)
Hypertension	21/26	24/29	(NS)
Diabetes	2/26	1/29	(NS)
Hyperlipidemia	1/26	1/29	(NS)
Renal failure	1/26	0/29	(NS)

NS: not significant

Table 2.

Operative and post-operative data

	T group	N group	P
Operative procedures
▪ Ascending aorta replacement	15	21	(NS)
▪ Pertial arch replacement	3	2	(NS)
▪ Total arch replacement	6	6	(NS)
▪ Bentall methods	2	0	(NS)
Operation time (min)	324.4 ± 72.7	298.8 ± 77.6	(NS)
CPB time (min)	192.4 ± 52.7	166.2 ± 48.2	(NS)
ICU stay (day)	4.0 ± 3.0	5.6 ± 4.4	(NS)
Hospital stay (day)	30.7 ± 33.1	30.1 ± 19.6	(NS)

CPB: cardiopulmonary bypass; ICU: intensive care unit; NS: not significant

Fig. 1.

Amount of blood transfusion. RCC; red cell concentrate: FFP; fresh frozen plasma: Pc; platelet concentrate.

In order to evaluate the hemostasis time for oozing blood after controlling intraoperative surgical bleeding, the time from the discontinuation of artificial heart-lung machines until the end of surgery was determined and compared between the two groups. It was 80.3 ± 17.1 min in Group T and 97.1 ± 33.2 min in Group N (P = 0.025). It was confirmed that the time from the discontinuation of artificial heart-lung machines until the end of surgery was significantly shorter in Group T, suggesting a decrease in hemostasis time. Next, the total volume from drains placed in the pericardium and anterior mediastinum postoperatively was measured as a reference value and the two groups were compared. The volume was 2018 ± 1550 cc in Group T and 3533 ± 3098 cc in Group N (P = 0.029). As the effusion in the pericardium was also included in the calculated volume, these figures do not necessarily represent the correct postoperative bleeding volume. However, these figures may provide important data suggesting decreased postoperative blood loss in Group T (Fig. 2).

Fig. 2.

The time of chest closure; the time from the finish of cardiopulmonary bypass until the end of surgery. Total amount of chest tube drainage; the total volume from drains placed in the pericardium and mediastinum postoperatively.

Comparison of the duration of surgery, cardiopulmonary bypass, stay in the ICU, and hospital stay revealed no significant differences between the two groups (Table 2). None of the patients in Group T, but 1 patient in Group N died in hospital (the patient dying of rupture of anastomotic aneurysm). Other major complications were observed in 10 of 26 patients in Group T (cerebral infarction in 4 patients, bleeding necessitating a repeat thoracotomy in 3 patients, intracerebral bleeding, seizure and mediastinitis in 1 patient each) and 15 of 29 patients in Group N (cerebral infarction in 7 patients, bleeding necessitating a repeat thoracotomy in 6 patients, mediastinitis in 1 patient). No statistically significant differences were observed in the incidence of complications between the two groups (Table 3).

Table 3.

Post-operative complications

	T group	N group	P
Cerebral infarction	4/26	7/29	NS
Cerebral bleeding	1/26	0/29	NS
PMI	0/26	0/29	-
Graft infection	0/26	0/29	-
Mediastinitis	1/26	1/29	NS
Re exploration	3/26	6/29	NS
Seizure	1/26	0/29	NS
Hospital death	0/26	1/29	NS

PMI: peri-operative myocardial infarction; NS: not significant

Discussion

In patients acute aortic dissection, a large blood clot volume is known to form rapidly at the false lumen immediately after the onset, thereby enhancing fibrinolytic activity. This can be regarded as one of the reasons why a large-volume transfusion is needed during surgery for acute aortic dissection. Moreover, several studies have reported that administration of an anti-fibrinolytic agent is effective for prevention of bleeding in patients undergoing surgery using extracorporeal circulation. Therefore, we conducted the present study, considering that an anti-fibrinolytic agent, TA, might be particularly effective in emergency surgery for acute aortic dissection using extracorporeal circulation.

The fibrinolytic system refers to the phenomenon in which insoluble fibrin clots formed in blood vessels via the activity of the blood coagulation system are dissolved into soluble peptide fragments by plasmin, a fibrinolytic enzyme. Plasmin is normally present in blood as an inactive precursor, plasminogen. For onset of the activity of the fibrinolytic system, activation of plasminogen by plasminogen activators (PAs) plays an important role. There are two types of PAs, namely, tissue-type PA and urokinase-type PA. Several factors including hormones, cytokines, and cell growth factors are thought to be involved in the production of PAs. Moreover, the activity of PAs is regulated by type-1 PA inhibitor (PAI-1) which is a specific inhibitory protein of PAs. The blood clot dissolution process involving PAs, PAI-1 and plasmin is collectively referred to as the fibrinolysis system.

The amount of tissue-PA released into circulating blood is known to increase during extracorporeal circulation.²⁾ The increase in tissue-PA release begins immediately after the initiation of extracorporeal circulation, and subsequently, the production of plasmin increases more than 100 fold.³⁾ The blood concentration of PAI-1, which is a primary endogenous tissue-PA antagonist mainly synthesized in the liver, does not increase during extracorporeal circulation. Once extracorporeal circulation has been completed, the blood concentration of PAI-1 increases up to 15 fold within 2 hours, and thereby serves to lower the activated tissue-PA concentration, and eventually decreases the production of plasmin. Therefore, the most appropriate strategy for administration of an anti-fibrinolytic agent is considered to be continuous infusion during extracorporeal circulation.

A few small-scale studies conducted in the late 1980s showed that aprotinin, an inhibitor of fibrinolysis and natural protease inhibitor originating from bovine lung, dramatically reduced the amount of blood transfusion required in cardiac surgery performed under extracorporeal circulation.^4–6) However, from 2006, several reports showed that aprotinin increased the incidence of complications despite reducing the frequency/amount of intraoperative bleeding.^1,7,8) The Blood Conservation Using Antifibrinolytics in a Randomized Trial (BART) conducted in Canada involved 2331 patients scheduled to undergo high-risk cardiac surgery. In this trial, aprotinin reduced the frequency of perioperative bleeding as compared to TA and ε-aminocaproic acid; however, the mortality of aprotinin group at 30 days after surgery was higher than TA and e-aminocaproic acid groups.⁸⁾ Based on these results, aprotinin was recalled and became unavailable.

As a consequence of the unavailability of aprotinin, TA, which is an inhibitor of fibrinolysis historically used as a hemostatic drug, began to attract attention again.

TA is a substance with molecular weight of 157.21 and molecular formula of C8H15NO2. The structure that occupies the-lysine binding site of plasminogen is considered to exert an antifibrinolytic effect by acting as a lysine analog binding to the lysine-binding sites of plasmin and plasminogen. As a result, an antifibrinolysis system action will be revealed by checking adsorption to the fibrin of plasminogen.

TA is said to inhibit fibrinolysis at a serum concentration of 10 µg/mL.

The available intraoperative administration route for TA is limited to intravenous injection. However, lysine analogs, which are small water-soluble molecules, are considered to be rapidly distributed to extravascular fluid compartments before being taken up by various cells and tissues.^9,10)

The optimal dose of TA has not been clearly established yet. In previous reports, various doses were applied, including continuous intravenous infusion at the dose of 1 mg/kg/hr after injection at the dose of 10 mg/kg for 20 min,^11,12) intravenous injection at the dose of 100 mg/kg before the start of extracorporeal circulation,¹³⁾ and injection at the dose of 1 g once followed by continuous intravenous infusion at the dose of 400 mg/hr, and administration of 500 mg of TA via the perfusion fluid for extracorporeal circulation.¹⁴⁾

In a study on the association between doses of continuous infusion of TA and the blood concentrations during extracorporeal circulation, Noreen, et al.¹⁵⁾ found that a high blood concentration of TA (≥125 µg/mL) was maintained with a loading dose of 32 mg/kg followed by continuous infusion at 16 mg/kg/h. Based on this report, we determined the dosing regimen. However, since hemodynamics are often unstable at the initial stage of surgery for an acute aortic dissection due to conditions such as cardiac tamponade, we decided to perform continuous infusion at 16 mg/kg/h only, rather than starting with a loading dose, considering potential risks such as shock caused by rapid infusion of TA at the loading stage. However, we are aware that the blood concentration may be lowered to some extent under such circumstances.

Residual persistence of lysine analogs, including TA, after neutralization of heparin is theoretically considered to increase the risk of occurrence of thromboembolism, such as deep venous thrombosis, pulmonary embolism, stroke and myocardial infarction. However, several reviews and meta-analyses have revealed the lack of any association between intraoperative administration of TA and thromboembolism.^16–18)

In the present comparative study, no increase in the incidence of thromboembolism was observed in Group T. On the other hand, there is also a report that administration of a large dose of TA may cause nonischemic convulsions.¹⁹⁾ Thus, we consider that further studies are needed in the future for establishing the optimal dose of TA.

There was 1 death in hospital postoperatively (aneurysmal rupture at the anastomotic site) in the N group in which TA was not administered. Since it is known that the fibrinolysis system does not cause aneurysms to rupture at the anastomotic site, we conclude that there was no association between the observed death and the administration of TA.

Cerebral infarction was observed in several patients (4 patients in the T group and 8 patients in the N group). Patients who presented with neurological symptoms postoperatively and had an infraction detected by diagnostic imaging were diagnosed as having cerebral infarction. Among them, 1 patient in the T group and 3 patients in the N group had a major infarction severe enough to decrease ADL after discharge. In other patients, the symptoms had disappeared by the time of discharge. Since the incidence of cerebral infarction was low in the T group, we concluded that administration of TA at the dose given in this study does not increase the incidence of cerebral infarction. The main causes of cerebral infarction were assumed to be inclusion of air during selective cerebral perfusion and/or plaque rupture. Further modifications and improvements are needed to resolve technical problems.

The time of chest closure was longer in the group N than in the group T. However, the total operative time was longer in the group T than in the group N, although the difference was not statistically significant. This was probably related to the differences in the surgical procedures used in these two groups. Patients undergoing ascending aortic replacement with minimal operative time accounted for 58% of patients in the group T and 72% of patients in the group N. On the other hand, the proportion of patients undergoing total arch replacement or replacement by the Bentall procedure, for which the operative time is longer, was 31% in the group T and 21% in the group N. This may explain a longer operative time in the group T, although the proportion of surgical procedures did not statistically differ between the two groups.

The major limitation of the present comparative study was its retrospective design. Because Group T was not operated upon during the same period, but later than Group N, technical improvements in the surgical procedures may have led to the reduction in the amounts of blood transfusion required in this group. In order to overcome these limitations, prospective randomized studies need to be conducted in the future. Because no consensus has been arrived at on the optimal dose of TA, we also consider that comparative studies using different doses are necessary.

In surgeries of the type A acute aortic dissection, intraoperative continuous infusion of TA was associated with significant reduction in the amount of blood transfusion required. Furthermore, no increase in the incidence of complications associated with infusion of TA was observed either. Intraoperative continuous infusion of TA may be useful for surgeries of the thoracic aorta.

Disclosure Statement

Kun Tae Ahn and other co-authors have no conflict of interest.