- Email: [email protected]
Tranexamic acid in open aortic aneurysm surgery: a randomised clinical trial
British Journal of Anaesthesia, xxx (xxx): xxx (xxxx) doi: 10.1016/j.bja.2019.08.028 Advance Access Publication Date: xxx Clinical Investigation
CLINICAL INVESTIGATION
Tranexamic acid in open aortic aneurysm surgery: a randomised clinical trial Fabrizio Monaco1, Pasquale Nardelli1, Laura Pasin2, Gaia Barucco1, Cristina Mattioli1, Nora Di Tomasso1, Giuseppe Dalessandro1, Giuseppe Giardina1, Giovanni Landoni1,3,*, Roberto Chiesa3,4 and Alberto Zangrillo1,3 1
Department of Anesthesia and Intensive Care, IRCCS San Raffaele Scientific Institute, Milan, Italy, 2Department of
Anesthesia and Intensive Care, Ospedale S. Antonio, Padova, Italy, 3Department of Anesthesia and Intensive Care, VitaSalute San Raffaele University, Milan, Italy and 4Department of Vascular Surgery, IRCCS San Raffaele Scientific Institute, Milan, Italy *Corresponding author. E-mail: [email protected]
Abstract Background: Bleeding and transfusions affect mortality in aortic surgery. Although tranexamic acid significantly reduced bleeding in multiple settings, its role in major vascular surgery was never studied. The aim of this study was to determine if tranexamic acid reduces blood loss in open abdominal aortic aneurysm (AAA) surgery. Methods: A total of 100 patients undergoing elective open AAA repair were randomised to receive tranexamic acid (a loading dose of 500 mg and a continuous infusion of 250 mg h1) or placebo. The primary outcome was intraoperative blood loss, and the secondary outcomes were the number of patients receiving red blood cells, occurrence of thromboembolic events, and mortality. Data were analysed using the intention-to-treat principle. Results: Fifty patients were randomised into each group. Median (inter-quartile range) intraoperative blood loss was 400 (300e1050) ml in the tranexamic acid group vs 500 (360e1000) ml in the placebo group (P¼0.44). Transfusion rate was seven/50 (14%) in the tranexamic group vs 12/50 (24%) in the placebo group (P¼0.20). No thrombosis was recorded. In a post hoc analysis, postoperative blood loss was reduced in the tranexamic group both at 4 h (60 [40e80] ml vs 100 [60e140] ml, P<0.001) and 24 h (180 [120e275] vs 275 [190e395] ml, P¼0.003) after surgery. At 1 yr, three patients were dead, all in the placebo group (P¼0.24) and all after 28 days. Conclusions: Tranexamic acid did not reduce intraoperative blood loss or blood transfusions in open AAA repair, although it may reduce postoperative blood loss without increasing adverse effects. Clinical trial registration: NCT02335359. Keywords: aortic surgery; bleeding; mortality; randomised clinical trial; tranexamic acid; transfusion
Editor’s key points This study investigated whether tranexamic acid could reduce intraoperative blood loss when compared with placebo in open abdominal aortic aneurysm (AAA) repair. Although there was no difference in intraoperative blood loss, the study shows a reduction in postoperative blood loss in the tranexamic acid group.
The study was underpowered to show differences in safety features of tranexamic acid or mortality, but the data suggest no increase in adverse effects. A larger study that is sufficiently powered to investigate adverse effects and mortality for tranexamic acid in open AAA surgery is warranted.
Editorial decision: 22 August 2019; Accepted: 22 August 2019 © 2019 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]
1
2
-
Monaco et al.
Abdominal aortic aneurysm (AAA) is a common condition, affecting almost 3.5% of the general population between 65 and 74 yr old.1 Although use of endovascular repair (EVAR) has increased during the past decade and is now recommended in patients with suitable anatomy, open surgical repair is still the treatment of choice in patients with long life expectancy or in those unsuitable for EVAR.2 However, this lifesaving procedure is burdened by a list of serious complications, which can ultimately lead to death. Among these, bleeding and blood transfusions play a major role, affecting mortality and morbidity.3 Bleeding is multifactorial in vascular surgery: after surgical haemostasis is achieved, microvascular bleeding must be addressed and corrected, starting with the regulation of the fibrinolytic system.4 Tranexamic acid is a synthetic derivative of the amino acid lysine. It exerts an antifibrinolytic effect by means of a reversible block of plasminogen bind on the surface of fibrin, thus inhibiting the conversion of plasminogen to plasmin, preventing fibrin degradation5 and platelets activation.6 Tranexamic acid introduction was a major element of novelty in various surgical and medical fields, contributing to reduce mortality in trauma patients with significant haemorrhage,7 in post-partum haemorrhage,8 and to reduce perioperative bleeding in several surgical settings9,10 including cardiac surgery.11,12 However, sporadic reports suggested that acute thrombosis may represent a rare yet serious complication after tranexamic acid administration.13e15 This circumstance is particularly felt in vascular surgerydwhich is per se prone to thrombosis. This risk accounts for the fact that the effects of tranexamic acid were never investigated before in major vascular surgery. We therefore performed the first randomised clinical trial to assess if tranexamic acid reduces blood loss in patients undergoing open AAA repair.
Methods Trial design We performed a single-centre, double-blinded, parallel-group, randomised clinical trial to test the effects of intraoperative treatment with tranexamic acid on perioperative blood losses in patients undergoing open AAA repair. The study started in February 2015 after IRCCS San Raffaele Scientific Institute Ethical Committee approval was obtained in December 2014 and was registered at ClinicalTrials.gov (NCT02335359) and EudraCT (2014-001456-39). The study was conceived according to the principles of the Declaration of Helsinki,16 and the present report follows the guidelines for reporting parallels group randomised trials CONSORT (Consolidated Standards of Reporting Trials) 2010.17 The last patient was enrolled on October 30, 2017. One year follow-up was completed in November 2018.
Participants Patients older than 50 yr undergoing open AAA surgical repair who provided written informed consent were eligible for the trial. Patients undergoing urgent or emergent surgery, noncollaborating or psychiatric patient, patients with known history of allergy to tranexamic acid, seizures, acute venous or arterial thrombosis, fibrinolytic conditions because of consumptive coagulopathy, severe renal insufficiency (defined
as estimated glomerular filtration rate below 30 ml min1 1.73 m2), haematuria, and patients with ocular disturbances, including blurred vision, poor sight, or altered colour perception were excluded from the trial.
Intervention Patients randomised to the intervention group received tranexamic acid (Ugurol®; Rottapharm S.p.A., Milan, Italy; 0.5 g 5 ml1 vial) according to the following protocol: a loading dose of 500 mg of tranexamic acid diluted in 100 ml saline was intravenously infused slowly 20 min before surgery, and a continuous intravenous infusion of tranexamic acid was then administered at a rate of 250 mg h1 (2.5 ml h1, using nondiluted tranexamic acid contained in the vials) from surgical incision until skin closure. Patients randomised to the control group received placebo (saline) with identical volumes and rates of infusion.
Outcomes The primary outcome was intraoperative blood loss calculated as the sum of blood volume aspired during surgery and blood volume absorbed in gauzes (at the end of the surgery, all gauzes used were weighted and net weight of the gauzes was subtracted from the total weight). The secondary outcomes consisted of patients receiving packed red blood cells (during surgery and until hospital discharge), occurrence of thromboembolic events up to 28 days after surgery, and mortality 28 days and 1 yr after surgery. A phone follow-up was performed 28 days and 1 yr after surgery. If loss to follow-up by telephone occurred, to establish vital status at 1 yr different attempts were made, including contacting the patient’s general practitioner, contacting the city or municipality, and sending a letter to the home address of the patient. A post-hoc analysis to assess if tranexamic acid influences postoperative blood loss 4 and 24 h after surgery was also performed. Postoperative blood loss was measured as the quantity of blood in the periaortic abdominal drainage tube bag. This measure was routinely performed by ward nurses every 2 h for postoperative AAA patients.
Perioperative care Perioperative care was carried out routinely by a multidisciplinary team, including 14 vascular surgeons (five first operators) and 10 anaesthesiologists. All patients underwent a full preoperative work-up, including standard laboratory tests, chest X-ray scan, and electrocardiography. Pulmonary function was further investigated in patients with known history of severe obstructive pulmonary disease or severe smoking history. Non-aspirin antiplatelet agents were discontinued before surgery according to current guidelines,18 whereas aspirin was continued throughout the perioperative period. Warfarin was stopped 5 days before surgery, whereas novel oral anticoagulants were suspended 48 h before surgery. Bridging with low weight molecular heparin was considered in patients with increased thromboembolic risk. To prevent deep venous thrombosis and pulmonary thromboembolism, all patients received 4000 units of low molecular weight heparin once a day, from the evening before surgery until 4 weeks after surgery.
Tranexamic acid in open AAA surgery
Anaesthetic management A thoracic epidural catheter (usually placed between T8 and T10) was placed in all patients as they did not present contraindications. The epidural catheter was inserted via a midline approach using an 18 G Tuohy epidural needle. General anaesthesia was induced with intravenous administration of propofol (1.5e2.5 mg kg1) and fentanyl (1e2 mg kg1) or sufentanil (0.1e0.2 mg kg1). General anaesthesia was maintained with volatile anaesthetics (desflurane or sevoflurane) and analgesia was maintained with intermittent i.v. boluses of fentanyl, sufentanil, or continuous infusion of remifentanil. Neuromuscular block was achieved and maintained with cis-atracurium or rocuronium. Unfractioned heparin (5000 UI) was administered before aortic crossclamping. Additional heparin was administered if activated clotting time (ACT) did not reach 200 s. Heparin was fully reverted with protamine after declamping. A cell saver machine was available for all open AAA surgery, but was only used in selected cases involving moderate to severe surgical bleeding, upon operating surgeon judgement. Red blood cells were transfused if haemoglobin level decreased to less than 80 g L1 or, in the event of severe unresponsive hypotension, less than 100 g L1 after all salvaged blood was reinfused, following an institutional protocol that is supported by recent data.19 Plasma was transfused if severe bleeding occurred after surgical haemostasis was completed if the international normalised ratio (INR) was >1.5, the activated partial thromboplastin time (aPTT) was >40 s, or if both conditions were met.
Surgical management The surgical technique used for open AAA repair was described in detail in a previous study.20 A midline incision from the xiphoid to the symphysis pubis was performed to obtain adequate exposure. In selected cases involving local abdominal conditions, a bilateral subcostal laparotomy was preferred. The posterior retroperitoneum was incised to expose the infrarenal aorta. After systemic heparinisation, proximal and distal clamps were placed. The aneurysm was incised and thrombus was removed. Proximal and distal endto-end anastomosis was performed using a running polypropylene suture. A Dacron graft with a diameter ranging from 12 to 24 mm was used. In the case of aorto-iliac aneurysm, a bifurcated graft was used and distal anastomoses were performed to the common or external iliac arteries. After unclamping, the presence of femoral pulses was checked. Accurate surgical haemostasis was performed. No locally applied haemostatic agents were used. The aneurysmal sac was closed over the graft, and the retroperitoneal adipose tissue was used to separate the proximal anastomosis from the duodenum. Periaortic drainage was always positioned before closing the retroperitoneum with an absorbable suture.
Postoperative care Patients were routinely extubated in the operative room after surgery anddafter a short observation period in the recovery roomddischarged to the surgical ward. ICU admission was only considered in patients experiencing severe haemodynamic instability or major surgical complications. Patient undergoing open AAA repair at our institution receive an enhanced recovery after surgery (ERAS) approach, which has
-
3
been adopted since 2012.21 In uncomplicated courses, the surgical drainage was removed the day after surgery, together with the urinary catheter. Early oral fluid intake and mobilisation were encouraged a few hours after surgery. A multidisciplinary team including nurses and physiotherapists helped the patients to start walking on day 0. Epidural catheter was usually removed on the third postoperative day.
Sample size Sample size calculations were based on a two-sided alpha error of 0.05 and 80% power. We expected a reduction in intraoperative blood losses around 34% in patients treated with tranexamic acid, as reported in the literature in other surgical settings.22e24 The mean intraoperative blood loss for open AAA surgery at our centre was 800 (250e3100) ml, according to a previous study that covered an 18 yr period.20 On the basis of these data, a sample size of 48 patients per group was deemed necessary, and we decided to enrol 100 patients in our trial to account for possible protocol deviation.
Randomisation and blinding Randomisation sequence was created by permuted block randomisation with a block size of 20 and a 1:1 allocation generated by a computer. The allocation sequence was prepared by an independent operator not otherwise involved in the trial and was concealed by opaque sequentially numbered sealed envelopes. Patients were screened for eligibility and were asked to sign an informed consent document to participate in the trial on the day before surgery. Patients were randomised in the trial after they entered the surgical theatre, performed by dedicated study personnel in a separate environment. Patients were randomly allocated to the intervention or the placebo group by assigning them the sequentially numbered envelope with the lowest numeration. Trial participants, care providers, and data collectors were blinded to group assignment. The treatment was prepared by a nurse who was not involved in the study or in the care of the patient: the nurse opened the sealed envelope and prepared the treatment. The bottle with the loading dose and the syringe for continuous i.v. infusion were labelled only with the acronym of the trial and the number of randomisation of the patient; bottles containing tranexamic acid and those containing placebo were indistinguishable (tranexamic acid is a colourless substance indistinguishable from saline). After that, the envelope was resealed.
Statistical analysis Data were stored electronically and analysed using SPSS Statistics software version 23 (IBM Corp., Armonk, NY). The distribution of continuous data was tested for normality with the ShapiroeWilk test and parametric or non-parametric tests used as appropriate. A P-value <0.05 was considered statistically significant. Continuous variables were compared using ttest when normally distributed or ManneWhitney U-test when non-normally distributed. Categorical values were compared using a two-tailed c2 test with Yates correction, except when a small size is required, in which case Fisher’s exact test was used. Two-sided significance tests were used in all analyses. Data are presented as mean (standard deviation, SD) or median (inter-quartile range) if not otherwise indicated. Categorical variables are presented as n (%). The estimated
4
-
Monaco et al.
effect size was assessed with Cohen’s kappa statistic for continuous variables with normal distribution, Cliff’s d for continuous variables with non-normal distribution, and absolute risk difference for categorical variables with 95% of confidence intervals (CI). To account for residual imbalance in demographic, medical, or surgical characteristics between the two groups, a multivariable regression analysis with a stepwise method was performed. All preoperative data with a P-value <0.1 in univariable analysis were entered into the model. The b coefficient with 95% CI was reported for all potential confounding variables associated with the outcome. Data analysis was by intention to treat, according to a pre-established analysis plan. No imputation for missing data was done.
randomisation, and none were lost to follow-up. The randomisation flowchart as per CONSORT 2010 Guidelines is shown in Figure 1. The majority of patients were male (93/100, 93%) and median age was 70 (64e75) yr. The two groups were balanced in terms of patient characteristics, revised cardiac risk index, renal function, or perioperative medical therapydincluding anticoagulation and antiplatelet with the exception of there being more female patients in the tranexamic acid group, which consequently had a lower BMI, and a slightly higher baseline INR in the tranexamic acid group (Table 1). The ratio of patients receiving tube and bifurcated grafts did not vary between groups. Use of cell saver was also similar between the two groups (Table 2). Haematocrit variations during and after surgery are shown in Supplementary Figure S1.
Results A total of 100 patients undergoing open AAA repair between March 2015 and October 2017 were randomised in the trial. Fifty patients were randomised to receive tranexamic acid and 50 patients to receive placebo. All randomised patients received their assigned treatment and were analysed for the primary outcome. All patients were contacted 1 yr after
Primary analyses The median (inter-quartile range) intraoperative blood loss (Table 2) was 400 (300e1050) ml in the tranexamic acid group vs 500 (360e1000) ml in the placebo group (P¼0.44). The rate of patients receiving red blood cell transfusion was seven/50 (14%) in the tranexamic acid group vs 12/50 (24%) in the
Fig 1. Study flowchart as per CONSORT guidelines. CONSORT, Consolidated Standards of Reporting Trials.
Tranexamic acid in open AAA surgery
-
5
Table 1 Baseline characteristics. Inter-quartile range (IQR) is presented as (Q1 [first quartile] e Q3 [third quartile]). AAA, abdominal aortic aneurysm; ACE, angiotensin converting enzyme; aPTT, activated partial thromboplastin time; ARB, angiotensin II receptor blockers; CABG, coronary artery bypass graft; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; INR, international normalised ratio; PTCA, percutaneous transluminal coronary angioplasty; SD, standard deviation; TIA,: transient ischaemic attack.
Age (yr), median (IQR) Male sex, n (%) BMI (kg m2), mean (SD) ASA physical status, n (%) 1 2 3 4 Revised cardiac risk index, n (%) 1 predictor 2 predictors 3 predictors Preoperative laboratory tests Creatinine (mg dl1), mean (SD) Haematocrit (%), mean (SD) Platelet count (109 ml1), mean (SD) aPTT (s), mean (SD) INR, median (IQR) Medical history, n (%) Smoking history COPD Hypertension Diabetes mellitus Dyslipidaemia Atrial fibrillation Ischaemic heart disease Previous stroke or TIA Previous vascular surgery Previous cardiac surgery Previous AAA repair surgery Previous PTCA or CABG, n (%) CKD, n (%) Stage 1 Stage 2 Stage 3 Preoperative medical therapy, n (%) Beta blockers ACE inhibitors ARBs Diuretics Ca2þ channel blockers Statins Aspirin Dual antiplatelet therapy Anticoagulant therapy
Tranexamic acid group (n¼50)
Control group (n¼50)
69 (63e75) 43 (86) 25.3 (3.33)
71 (65e74) 50 (100) 26.2 (2.64)
0 14 (28) 35 (70) 1 (2)
0 13 (26) 37 (74) 0
32 (64) 17 (34) 1 (2)
31 (62) 18 (36) 1 (2)
0.97 (0.31) 41 (3.8) 195 (54) 29 (3) 1.08 (1.03e1.15)
1.03 (0.27) 41 (3.6) 208 (59) 29 (2) 1.05 (1.00e1.10)
31 (62) 11 (22) 36 (72) 6 (12) 20 (40) 1 (2) 15 (30) 3 (6) 4 (8) 5 (10) 0 12 (24)
39 (78) 3 (6) 37 (74) 4 (8) 22 (44) 1 (2) 16 (32) 4 (8) 7 (14) 3 (6) 2 (4) 13 (26)
13 (26) 29 (58) 8 (16)
14 (28) 27 (54) 9 (18)
20 (40) 12 (24) 15 (30) 9 (18) 14 (28) 33 (66) 29 (58) 2 (4) 2 (4)
20 (40) 17 (34) 16 (32) 10 (20) 15 (30) 27 (54) 34 (68) 4 (8) 2 (4)
placebo group (P¼0.20). We did not find differences in the rate of thromboembolic events (including acute myocardial infarction) or acute kidney injury between the tranexamic acid and the placebo groups. We did not record any occurrence of pulmonary embolism, bowel infarction, or seizures. No death at 28 days was recorded in the present trial, whereas at the 1 yr follow-up, three patientsdall belonging to the control groupddied (0/50 vs 3/50, P¼0.24).
Secondary analyses Median (inter-quartile range) postoperative blood loss was reduced in the tranexamic group 4 h after surgery (60 [40e80] ml vs 100 [60e140] ml, P<0.001) and 24 h after surgery (180 [120e275] ml vs 275 [190e395] ml, P¼0.003) (Table 3). A
multivariable regression analysis was performed to account for residual imbalance. After correction for age, BMI, INR, smoking history, and chronic obstructive pulmonary disease (COPD), tranexamic acid administration was the only significant predictor (tranexamic acid protective) of bleeding (b¼e37 ml; 95% CI, e58 to e15; P¼0.001) (Supplementary Table S1) and 24 h (b¼e79 ml; 95% CI, e134 to e24 ml; P¼0.002) 4 h after surgery (Supplementary Table S2). A reduction of overall bleeding (from beginning of surgery to 24 h after surgery) was also observed in the tranexamic acid group, although it was not statistically significant (620 [450e1275] ml vs 920 [570e1220] ml, P¼0.13). Only one patient, belonging to the placebo group, required surgical reintervention (Table 3). Postoperative admission to the ICU was necessary in two/50 patients (4%) in the tranexamic acid group and
6
-
Monaco et al.
Table 2 Intraoperative data. Inter-quartile range (IQR) is presented as (Q1 [first quartile] e Q3 [third quartile]).
Surgical time (min), mean (SD) Clamping time (min), mean (SD) Tube graft, n (%) Bifurcated graft, n (%) Use of cell saver, n (%) Intraoperative blood loss (ml), median (IQR) Number of patients receiving blood products, n (%) Red blood cells, n (%) Fresh frozen plasma, n (%)
SD,
standard deviation.
Tranexamic acid group (n¼50)
Control group (n¼50)
P-value
137 (52) 36 (13) 40 (80) 10 (20) 16 (32) 400 (300e1050) 2 (4) 1 (2) 1 (2)
138 (49) 37 (14) 44 (88) 6 (12) 21 (42) 500 (360e1000) 3 (6) 2 (4) 1 (2)
0.45 0.34 0.41 0.41 0.41 0.44 0.65 0.56 0.99
two/50 patients (4%) in the control group (P¼0.99) because of intraoperative haemodynamic instability in two cases (one in each group) and major surgical complications in the other two cases (one in each group). All patients were discharged to the ward within the first postoperative day.
Discussion In this randomised report investigating tranexamic acid in major vascular surgery, we found no reduction in intraoperative blood loss in patients undergoing open AAA repair, although we observed a reduction in postoperative bleeding. No adverse or thrombotic event was reported in the present trial. Although modern-era vascular surgery is going towards endovascular procedures for treatment of abdominal aorta disease, current guidelines report that open surgery is still the treatment of choice in patients with long life expectancy or in those unsuitable for EVAR.2 Life-threatening haemorrhage in aortic aneurysm repair is a feared perioperative complication independently associated with increased mortality.25 Perioperative bleeding in patients undergoing open AAA repair relies on the fragility of the diseased aortic tissues and on
coagulation disorders. The coagulation impairment in this setting is multifactorial: an excessive fibrinolysis is frequently observed in patients with abdominal aortic disease, possibly because of the exposure of blood to thromboplastin within atherosclerotic plaques.26 In addition, intraoperative surgical bleeding, haemodilution, and consumption coagulopathy further worsen the blood loss. Previous literature demonstrated how the application of the aortic cross clamp induces an elevation of the fibrinolytic function and a reduction in clotting factors activity in AAA open repair.27 Excessive bleeding leads to blood transfusions, which are associated with worse postoperative outcomes. In fact, two large observational studies proved the relationship between perioperative transfusions and increased 30 day morbidity and mortality rates in major vascular surgery.28,29 In addition, allogenic blood is a limited resource and its use should be reserved to those situations where it is absolutely necessary. Thus, any strategy or drug with haemostatic properties capable of reducing bleeding and transfusion requirements without increasing adverse events may have a significant impact on daily clinical practice. However, in spite of being low-cost, widely available, and well renowned for its haemostatic properties, tranexamic acid was never previously
Table 3 Postoperative data and outcome variables. Inter-quartile range (IQR) is presented as (Q1 [first quartile] e Q3 [third quartile]). ARD, absolute risk difference; CI, confidence interval; RBC, red blood cells; RIFLE, Risk, Injury, Failure, Loss of kidney function, Endstage kidney disease; SD, standard deviation. Tranexamic acid group (n¼50) Blood loss 0e4 h after surgery (ml), median (IQR) 60 (40e80) Blood loss 0e24 h after surgery (ml), median (IQR) 180 (120e275) Blood loss from the beginning of surgery to 24 h 620 (450e1275) after surgery (ml), median (IQR) Patients receiving RBC in the postoperative period, 6 (12) n (%) Patients receiving RBC in the intraoperative period, 7 (14) postoperative period, or both, n (%) Postoperative ICU admission, n (%) 2 (4) Acute myocardial infarction, n (%) 2 (4) Acute kidney injury according to RIFLE criteria, n (%) Risk 8 (16) Injury 2 (4) Failure 0 Thrombosis, n (%) 0 Seizures, n (%) 0 Bleeding requiring reintervention, n (%) 0 Postoperative hospital stay (days), mean (SD) 6 (1.5) 30-day mortality, n (%) 0 1-yr mortality, n (%) 0
Control group (n¼50)
Estimated effect size Cliff’s d ± 95 % CI or ARD ± 95 % CI
P-value
100 (60e140) 275 (190e395) 920 (570e1220)
e0.41 (e0.19 to e0.59) e0.37 (e0.14 to e0.55)
<0.001 0.002 0.13
10 (20)
0.08 (e0.06 to 0.22)
0.28
12 (24)
0.1 (e0.05 to 0.25)
0.20
2 (4) 2 (4)
0.99 0.99
7 0 4 0 0 1 6 0 3
0.99 0.49 0.12 0.99 0.99 0.32 0.99
(14) (8)
(2) (1.2) (6%)
0.02 (e0.3 to 0.07) 0.05 (0.16 to e0.27) 0 0.06 (e0.01 to 0.13)
0.24
Tranexamic acid in open AAA surgery
investigated in major vascular surgery, because of sporadic reports of thrombotic events and stroke related to its use. Recently, the Aspirin and Tranexamic Acid for Coronary Artery Surgery (ATACAS) trial11 proved that tranexamic acid administration did not increase the risk of death or thrombotic complications in a population of more than 5000 cardiac surgery patients. Thromboembolic complications did not occur in the present trial, even if it was underpowered to detect safety issues. The ATACAS trial reported a higher incidence of postoperative benign seizures in the tranexamic acid group, whereas we did not observe any seizures in the present trial.11 As seizures are believed to be dose related in this setting, it is important to note that the ATACAS trial administered a dose of tranexamic acid five-to seven-fold higher than that used in the present trial,11 although the greater fluid shift and cardiopulmonary bypass-related haemodilution may partially explain the higher dosing in cardiac surgery. The dose of tranexamic acid in the present trial (an average of 1g in total) was determined on the basis of previous evidence in other settings,8,9 and of our previous clinical experience, to obtain a balance between the anti-fibrinolytic effects and the avoidance of the mentioned neurological complications. Although the dose-related sideeffects of tranexamic acid must be kept in mind, further studies are required to assess the effects of higher doses of tranexamic acid in this setting. The present trial is the first double-blind randomised trial to investigate the effects of tranexamic acid in patients undergoing major vascular surgery and one of the first ever to report complete 1 yr follow-up data. In our study, a reduction of postoperative blood losses was detected, whereas no intraoperative effect was observed in patients receiving tranexamic acid. We speculate that this is because intraoperative bleeding is mostly secondary to tissue dissection and aortic manipulation and requires surgical haemostasis, whereas postoperative bleeding is, for the most part, associated with microvascular bleeding. In this setting, tranexamic acid administered during surgery may decrease fibrin degradation and prevent postoperative microvascular bleeding. Previous clinical pharmacology studies showed how tranexamic acid is present in the blood in therapeutic concentrations up to 5 h after i.v. administration.5 Therefore, it seems reasonable that its anti-haemorrhagic properties are present in the postoperative period, particularly during the first 4 h after operation. The present study has several limitations and strengths. The sample size calculation was based on a historical estimation of intraoperative blood loss during AAA surgery that was higher compared with reported values in the present study. The trial was further underpowered to assess safety outcomes; therefore, even if in our study we did not observe thrombosis or mortality in the tranexamic group, no strong conclusions could be drawn from the present work. However, our findings are in agreement with recent literature in that tranexamic acid does not increase the risk of thromboembolic events.7,8,11 We also reported a reduction of postoperative bleeding in patients receiving intraoperative tranexamic acid. This exploratory post-hoc analysis, however, was based on drainage content. Although using drainage content to estimate postoperative bleeding was part of our pragmatic approach, this method does not offer a precise measure of how much blood was lost. Further trials should specifically focus on postoperative bleeding,
-
7
including measurements of blood content in postoperative drainages. This was the first randomised study ever performed focusing on tranexamic acid in the setting of open AAA repair. Moreover, the present study also reported a complete 1 yr follow-up (with a trend towards a lower mortality in the tranexamic acid group), which is rarely reported even in other, more extensively studied settings. Interestingly, the mean intraoperative blood loss in the overall study population was less than 500 ml per patient, which is much lower than those reported in previous literature. For example, Ferrante and colleagues30 in a recent study on AAA open repair reported a mean estimated intraoperative blood loss of 900 ml. Even in our centre, a previous retrospective study reported a high transfusion rate in the period 1993e2010.20 These differences could be explained by the improvement achieved in our centre, which is nowadays an international referral for aortic surgery. The important surgical learning curve observed in our centre dramatically reduced intraoperative blood loss in open AAA at our centre in the past decade and prevented us from finding a statistically significant reduction in intraoperative bleeding. In fact, although a 20% lower intraoperative blood loss and a 33% lower overall blood loss were observed in the present trial, these differences were not statistically significant. As intraoperative blood loss was lower than anticipated, the study was possibly underpowered to assess the primary outcome, and should therefore be considered a pilot trial. The same applies to perioperative transfusions, which were a lowlikelihood event in the present trial. In fact, although the transfusion rate was 14% in the tranexamic acid group vs 24% in the placebo group, no statistical significance was reached, possibly because the study was underpowered to detect it. These data suggest that a multimodal approach to blood sparing in major vascular surgery should at least consider tranexamic acid. In conclusion, the use of tranexamic acid in major vascular surgery did not reduce intraoperative blood loss, although it might have had a beneficial effect on postoperative bleeding. Further large, multicentric studiesdincluding the ongoing 10 000 patients PeriOperative ISchemic Evaluation-III trial31dwill help us understand the role of tranexamic acid in major clinical outcomes and survival in the perioperative setting of highrisk noncardiac surgery.
Authors’ contributions Study concept: LP Study design: GL, AZ Data acquisition: GG, GD Data analysis: PN, FM Data interpretation: CM, NDT, RC Drafting of the manuscript: PN, FM Revising manuscript: LP, GB, CM, NDT, GD, GG, GL, RC, AZ All authors approved the final version to be published; agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved
Acknowledgements We thank all the staff of Vascular Surgery Operating Rooms at IRCCS San Raffaele Hospital for their constant help and effort, making this study possible.
8
-
Monaco et al.
Declarations of interests The authors declare they have no conflicts of interests.
Funding
13.
14.
Rottapharm Biotech S.r.l. (an unrestricted grant).
Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2019.08.028.
15.
16.
References 1. Grondal N, Sogaard R, Lindholt JS. Baseline prevalence of abdominal aortic aneurysm, peripheral arterial disease and hypertension in men aged 65e74 years from a population screening study (VIVA trial). Br J Surg 2015; 102: 902e6 2. Wanhainen A, Verzini F, Van Herzeele I, et al. European Society for Vascular Surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal aorto-iliac artery aneurysms. Eur J Vasc Endovasc Surg 2019; 57: 8e93 3. Davidovic LB, Maksic M, Koncar I, et al. Open repair of AAA in a high volume center. World J Surg 2017; 41: 884e91 4. Skagius E, Siegbahn A, Bergqvist D, Henriksson AE. Fibrinolysis in patients with an abdominal aortic aneurysm with special emphasis on rupture and shock. J Thromb Haemost 2008; 6: 147e50 5. Verstraete M. Clinical application of inhibitors of fibrinolysis. Drugs 1985; 29: 236e61 6. Soslau G, Horrow J, Brodsky I. Effect of tranexamic acid on platelet ADP during extracorporeal circulation. Am J Hematol 1991; 38: 113e9 7. CRASH-2 trial collaborators, Shakur H, Roberts I, Bautista R, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010; 376: 23e32 8. WOMAN Trial Collaborators. Effect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post-partum haemorrhage (WOMAN): an international, randomised, doubleblind, placebo-controlled trial. Lancet 2017; 389: 2105e16 9. Crescenti A, Borghi G, Bignami E, et al. Intraoperative use of tranexamic acid to reduce transfusion rate in patients undergoing radical retropubic prostatectomy: double blind, randomised, placebo controlled trial. BMJ 2011; 343: d5701 J. 10. Colomina MJ, Koo M, Basora M, Pizones J, Mora L, Bago Intraoperative tranexamic acid use in major spine surgery in adults: a multicentre, randomized, placebo-controlled trial. Br J Anaesth 2017; 118: 380e90 11. Myles PS, Smith JA, Forbes A, et al. Tranexamic acid in patients undergoing coronary-artery surgery. N Engl J Med 2017; 376: 136e48 12. Casati V, Sandrelli L, Speziali G, Calori G, Grasso MA, Spagnolo S. Hemostatic effects of tranexamic acid in elective thoracic aortic surgery: a prospective,
17.
18. 19.
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
randomized, double-blind, placebo-controlled study. J Thorac Cardiovasc Surg 2002; 123: 1084e91 Gerstein NS, Kelly SP, Brierley JK. Yet another tranexamic acid-related thrombotic complication. J Cardiothorac Vasc Anesth 2016; 30. e21e2 Gupta PN, Mullamalla UR, Sabin P, Vellappan P. Acute MI in a young hypertensive woman: could it be due to tranexamic acid? BMJ Case Rep 2013; 2013. bcr2013009979 Taparia M, Cordingley FT, Leahy MF. Pulmonary embolism associated with tranexamic acid in severe acquired haemophilia. Eur J Haematol 2002; 68: 307e30 World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 2013; 310: 2191e249 Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010; 340: c332 Oprea AD, Popescu WM. Perioperative management of antiplatelet therapy. Br J Anaesth 2013; 111. i3e17 Møller A, Nielsen HB, Wetterslev J, et al. Low vs high hemoglobin trigger for transfusion in vascular surgery: a randomized clinical feasibility trial. Blood 2019; 133: 2639e50 Chiesa R, Tshomba Y, Psacharopulo D, et al. Open repair for infrarenal AAA: technical aspects. J Cardiovasc Surg 2012; 53: 119e31 Pasin L, Nardelli P, Landoni G, et al. Enhanced recovery after surgery program in elective infrarenal abdominal aortic aneurysm repair. J Cardiovasc Surg 2019; 60: 369e74 Henry DA, Carless PA, Moxey AJ, et al. Anti-fibrinolytic use for minimising perioperative allogeneic blood transfusion. Cochrane Database Syst Rev 2011; 3: CD001886 Ker K, Prieto-Merino D, Roberts I. Systematic review, meta-analysis and meta-regression of the effect of tranexamic acid on surgical blood loss. Br J Surg 2013; 100: 1271e9 Dunn CJ, Goa KL. Tranexamic acid: a review of its use in surgery and other indications. Drugs 1999; 57: 1005e32 Lederle FA, Freischlag JA, Kyriakides TC, et al. Open versus Endovascular Repair (OVER) Veterans Affairs Cooperative Study Group. Outcomes following endovascular vs open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009; 302: 1535e42 Gertler JP, Cambria RP, Brewster DC, et al. Coagulation changes during thoracoabdominal aneurysm repair. J Vasc Surg 1996; 24: 936e43 Adam DJ, Ludlam CA, Ruckley CV, Bradbury AW. Coagulation and fibrinolysis in patients undergoing operation for ruptured and nonruptured infrarenal abdominal aortic aneurysms. J Vasc Surg 1999; 30: 641e50 Obi AT, Park YJ, Bove P, et al. The association of perioperative transfusion with 30-day morbidity and mortality in patients undergoing major vascular surgery. J Vasc Surg 2015; 61: 1000e9 Osborne Z, Hanson K, Brooke BS, et al. Variation in transfusion practices and the association with perioperative adverse events in patients undergoing open abdominal aortic aneurysm repair and lower extremity arterial bypass in the vascular quality initiative. Ann Vasc Surg 2018; 46: 1e16
Tranexamic acid in open AAA surgery
30. Ferrante AMR, Moscato U, Snider F, Tshomba Y. Controversial results of the Revised Cardiac Risk Index in elective open repair of abdominal aortic aneurysms: retrospective analysis on a continuous series of 899 cases. Int J Cardiol 2019; 277: 224e8
-
9
31. PeriOperative ISchemic evaluation-3 trial (POISE-3). ClinicalTrials.gov Identifier: NCT03505723. Available from: https://clinicaltrials.gov/ct2/show/record/NCT03505723. [Accessed 29 September 2019]
Handling editor: C. Boer
CLINICAL INVESTIGATION
Tranexamic acid in open aortic aneurysm surgery: a randomised clinical trial Fabrizio Monaco1, Pasquale Nardelli1, Laura Pasin2, Gaia Barucco1, Cristina Mattioli1, Nora Di Tomasso1, Giuseppe Dalessandro1, Giuseppe Giardina1, Giovanni Landoni1,3,*, Roberto Chiesa3,4 and Alberto Zangrillo1,3 1
Department of Anesthesia and Intensive Care, IRCCS San Raffaele Scientific Institute, Milan, Italy, 2Department of
Anesthesia and Intensive Care, Ospedale S. Antonio, Padova, Italy, 3Department of Anesthesia and Intensive Care, VitaSalute San Raffaele University, Milan, Italy and 4Department of Vascular Surgery, IRCCS San Raffaele Scientific Institute, Milan, Italy *Corresponding author. E-mail: [email protected]
Abstract Background: Bleeding and transfusions affect mortality in aortic surgery. Although tranexamic acid significantly reduced bleeding in multiple settings, its role in major vascular surgery was never studied. The aim of this study was to determine if tranexamic acid reduces blood loss in open abdominal aortic aneurysm (AAA) surgery. Methods: A total of 100 patients undergoing elective open AAA repair were randomised to receive tranexamic acid (a loading dose of 500 mg and a continuous infusion of 250 mg h1) or placebo. The primary outcome was intraoperative blood loss, and the secondary outcomes were the number of patients receiving red blood cells, occurrence of thromboembolic events, and mortality. Data were analysed using the intention-to-treat principle. Results: Fifty patients were randomised into each group. Median (inter-quartile range) intraoperative blood loss was 400 (300e1050) ml in the tranexamic acid group vs 500 (360e1000) ml in the placebo group (P¼0.44). Transfusion rate was seven/50 (14%) in the tranexamic group vs 12/50 (24%) in the placebo group (P¼0.20). No thrombosis was recorded. In a post hoc analysis, postoperative blood loss was reduced in the tranexamic group both at 4 h (60 [40e80] ml vs 100 [60e140] ml, P<0.001) and 24 h (180 [120e275] vs 275 [190e395] ml, P¼0.003) after surgery. At 1 yr, three patients were dead, all in the placebo group (P¼0.24) and all after 28 days. Conclusions: Tranexamic acid did not reduce intraoperative blood loss or blood transfusions in open AAA repair, although it may reduce postoperative blood loss without increasing adverse effects. Clinical trial registration: NCT02335359. Keywords: aortic surgery; bleeding; mortality; randomised clinical trial; tranexamic acid; transfusion
Editor’s key points This study investigated whether tranexamic acid could reduce intraoperative blood loss when compared with placebo in open abdominal aortic aneurysm (AAA) repair. Although there was no difference in intraoperative blood loss, the study shows a reduction in postoperative blood loss in the tranexamic acid group.
The study was underpowered to show differences in safety features of tranexamic acid or mortality, but the data suggest no increase in adverse effects. A larger study that is sufficiently powered to investigate adverse effects and mortality for tranexamic acid in open AAA surgery is warranted.
Editorial decision: 22 August 2019; Accepted: 22 August 2019 © 2019 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]
1
2
-
Monaco et al.
Abdominal aortic aneurysm (AAA) is a common condition, affecting almost 3.5% of the general population between 65 and 74 yr old.1 Although use of endovascular repair (EVAR) has increased during the past decade and is now recommended in patients with suitable anatomy, open surgical repair is still the treatment of choice in patients with long life expectancy or in those unsuitable for EVAR.2 However, this lifesaving procedure is burdened by a list of serious complications, which can ultimately lead to death. Among these, bleeding and blood transfusions play a major role, affecting mortality and morbidity.3 Bleeding is multifactorial in vascular surgery: after surgical haemostasis is achieved, microvascular bleeding must be addressed and corrected, starting with the regulation of the fibrinolytic system.4 Tranexamic acid is a synthetic derivative of the amino acid lysine. It exerts an antifibrinolytic effect by means of a reversible block of plasminogen bind on the surface of fibrin, thus inhibiting the conversion of plasminogen to plasmin, preventing fibrin degradation5 and platelets activation.6 Tranexamic acid introduction was a major element of novelty in various surgical and medical fields, contributing to reduce mortality in trauma patients with significant haemorrhage,7 in post-partum haemorrhage,8 and to reduce perioperative bleeding in several surgical settings9,10 including cardiac surgery.11,12 However, sporadic reports suggested that acute thrombosis may represent a rare yet serious complication after tranexamic acid administration.13e15 This circumstance is particularly felt in vascular surgerydwhich is per se prone to thrombosis. This risk accounts for the fact that the effects of tranexamic acid were never investigated before in major vascular surgery. We therefore performed the first randomised clinical trial to assess if tranexamic acid reduces blood loss in patients undergoing open AAA repair.
Methods Trial design We performed a single-centre, double-blinded, parallel-group, randomised clinical trial to test the effects of intraoperative treatment with tranexamic acid on perioperative blood losses in patients undergoing open AAA repair. The study started in February 2015 after IRCCS San Raffaele Scientific Institute Ethical Committee approval was obtained in December 2014 and was registered at ClinicalTrials.gov (NCT02335359) and EudraCT (2014-001456-39). The study was conceived according to the principles of the Declaration of Helsinki,16 and the present report follows the guidelines for reporting parallels group randomised trials CONSORT (Consolidated Standards of Reporting Trials) 2010.17 The last patient was enrolled on October 30, 2017. One year follow-up was completed in November 2018.
Participants Patients older than 50 yr undergoing open AAA surgical repair who provided written informed consent were eligible for the trial. Patients undergoing urgent or emergent surgery, noncollaborating or psychiatric patient, patients with known history of allergy to tranexamic acid, seizures, acute venous or arterial thrombosis, fibrinolytic conditions because of consumptive coagulopathy, severe renal insufficiency (defined
as estimated glomerular filtration rate below 30 ml min1 1.73 m2), haematuria, and patients with ocular disturbances, including blurred vision, poor sight, or altered colour perception were excluded from the trial.
Intervention Patients randomised to the intervention group received tranexamic acid (Ugurol®; Rottapharm S.p.A., Milan, Italy; 0.5 g 5 ml1 vial) according to the following protocol: a loading dose of 500 mg of tranexamic acid diluted in 100 ml saline was intravenously infused slowly 20 min before surgery, and a continuous intravenous infusion of tranexamic acid was then administered at a rate of 250 mg h1 (2.5 ml h1, using nondiluted tranexamic acid contained in the vials) from surgical incision until skin closure. Patients randomised to the control group received placebo (saline) with identical volumes and rates of infusion.
Outcomes The primary outcome was intraoperative blood loss calculated as the sum of blood volume aspired during surgery and blood volume absorbed in gauzes (at the end of the surgery, all gauzes used were weighted and net weight of the gauzes was subtracted from the total weight). The secondary outcomes consisted of patients receiving packed red blood cells (during surgery and until hospital discharge), occurrence of thromboembolic events up to 28 days after surgery, and mortality 28 days and 1 yr after surgery. A phone follow-up was performed 28 days and 1 yr after surgery. If loss to follow-up by telephone occurred, to establish vital status at 1 yr different attempts were made, including contacting the patient’s general practitioner, contacting the city or municipality, and sending a letter to the home address of the patient. A post-hoc analysis to assess if tranexamic acid influences postoperative blood loss 4 and 24 h after surgery was also performed. Postoperative blood loss was measured as the quantity of blood in the periaortic abdominal drainage tube bag. This measure was routinely performed by ward nurses every 2 h for postoperative AAA patients.
Perioperative care Perioperative care was carried out routinely by a multidisciplinary team, including 14 vascular surgeons (five first operators) and 10 anaesthesiologists. All patients underwent a full preoperative work-up, including standard laboratory tests, chest X-ray scan, and electrocardiography. Pulmonary function was further investigated in patients with known history of severe obstructive pulmonary disease or severe smoking history. Non-aspirin antiplatelet agents were discontinued before surgery according to current guidelines,18 whereas aspirin was continued throughout the perioperative period. Warfarin was stopped 5 days before surgery, whereas novel oral anticoagulants were suspended 48 h before surgery. Bridging with low weight molecular heparin was considered in patients with increased thromboembolic risk. To prevent deep venous thrombosis and pulmonary thromboembolism, all patients received 4000 units of low molecular weight heparin once a day, from the evening before surgery until 4 weeks after surgery.
Tranexamic acid in open AAA surgery
Anaesthetic management A thoracic epidural catheter (usually placed between T8 and T10) was placed in all patients as they did not present contraindications. The epidural catheter was inserted via a midline approach using an 18 G Tuohy epidural needle. General anaesthesia was induced with intravenous administration of propofol (1.5e2.5 mg kg1) and fentanyl (1e2 mg kg1) or sufentanil (0.1e0.2 mg kg1). General anaesthesia was maintained with volatile anaesthetics (desflurane or sevoflurane) and analgesia was maintained with intermittent i.v. boluses of fentanyl, sufentanil, or continuous infusion of remifentanil. Neuromuscular block was achieved and maintained with cis-atracurium or rocuronium. Unfractioned heparin (5000 UI) was administered before aortic crossclamping. Additional heparin was administered if activated clotting time (ACT) did not reach 200 s. Heparin was fully reverted with protamine after declamping. A cell saver machine was available for all open AAA surgery, but was only used in selected cases involving moderate to severe surgical bleeding, upon operating surgeon judgement. Red blood cells were transfused if haemoglobin level decreased to less than 80 g L1 or, in the event of severe unresponsive hypotension, less than 100 g L1 after all salvaged blood was reinfused, following an institutional protocol that is supported by recent data.19 Plasma was transfused if severe bleeding occurred after surgical haemostasis was completed if the international normalised ratio (INR) was >1.5, the activated partial thromboplastin time (aPTT) was >40 s, or if both conditions were met.
Surgical management The surgical technique used for open AAA repair was described in detail in a previous study.20 A midline incision from the xiphoid to the symphysis pubis was performed to obtain adequate exposure. In selected cases involving local abdominal conditions, a bilateral subcostal laparotomy was preferred. The posterior retroperitoneum was incised to expose the infrarenal aorta. After systemic heparinisation, proximal and distal clamps were placed. The aneurysm was incised and thrombus was removed. Proximal and distal endto-end anastomosis was performed using a running polypropylene suture. A Dacron graft with a diameter ranging from 12 to 24 mm was used. In the case of aorto-iliac aneurysm, a bifurcated graft was used and distal anastomoses were performed to the common or external iliac arteries. After unclamping, the presence of femoral pulses was checked. Accurate surgical haemostasis was performed. No locally applied haemostatic agents were used. The aneurysmal sac was closed over the graft, and the retroperitoneal adipose tissue was used to separate the proximal anastomosis from the duodenum. Periaortic drainage was always positioned before closing the retroperitoneum with an absorbable suture.
Postoperative care Patients were routinely extubated in the operative room after surgery anddafter a short observation period in the recovery roomddischarged to the surgical ward. ICU admission was only considered in patients experiencing severe haemodynamic instability or major surgical complications. Patient undergoing open AAA repair at our institution receive an enhanced recovery after surgery (ERAS) approach, which has
-
3
been adopted since 2012.21 In uncomplicated courses, the surgical drainage was removed the day after surgery, together with the urinary catheter. Early oral fluid intake and mobilisation were encouraged a few hours after surgery. A multidisciplinary team including nurses and physiotherapists helped the patients to start walking on day 0. Epidural catheter was usually removed on the third postoperative day.
Sample size Sample size calculations were based on a two-sided alpha error of 0.05 and 80% power. We expected a reduction in intraoperative blood losses around 34% in patients treated with tranexamic acid, as reported in the literature in other surgical settings.22e24 The mean intraoperative blood loss for open AAA surgery at our centre was 800 (250e3100) ml, according to a previous study that covered an 18 yr period.20 On the basis of these data, a sample size of 48 patients per group was deemed necessary, and we decided to enrol 100 patients in our trial to account for possible protocol deviation.
Randomisation and blinding Randomisation sequence was created by permuted block randomisation with a block size of 20 and a 1:1 allocation generated by a computer. The allocation sequence was prepared by an independent operator not otherwise involved in the trial and was concealed by opaque sequentially numbered sealed envelopes. Patients were screened for eligibility and were asked to sign an informed consent document to participate in the trial on the day before surgery. Patients were randomised in the trial after they entered the surgical theatre, performed by dedicated study personnel in a separate environment. Patients were randomly allocated to the intervention or the placebo group by assigning them the sequentially numbered envelope with the lowest numeration. Trial participants, care providers, and data collectors were blinded to group assignment. The treatment was prepared by a nurse who was not involved in the study or in the care of the patient: the nurse opened the sealed envelope and prepared the treatment. The bottle with the loading dose and the syringe for continuous i.v. infusion were labelled only with the acronym of the trial and the number of randomisation of the patient; bottles containing tranexamic acid and those containing placebo were indistinguishable (tranexamic acid is a colourless substance indistinguishable from saline). After that, the envelope was resealed.
Statistical analysis Data were stored electronically and analysed using SPSS Statistics software version 23 (IBM Corp., Armonk, NY). The distribution of continuous data was tested for normality with the ShapiroeWilk test and parametric or non-parametric tests used as appropriate. A P-value <0.05 was considered statistically significant. Continuous variables were compared using ttest when normally distributed or ManneWhitney U-test when non-normally distributed. Categorical values were compared using a two-tailed c2 test with Yates correction, except when a small size is required, in which case Fisher’s exact test was used. Two-sided significance tests were used in all analyses. Data are presented as mean (standard deviation, SD) or median (inter-quartile range) if not otherwise indicated. Categorical variables are presented as n (%). The estimated
4
-
Monaco et al.
effect size was assessed with Cohen’s kappa statistic for continuous variables with normal distribution, Cliff’s d for continuous variables with non-normal distribution, and absolute risk difference for categorical variables with 95% of confidence intervals (CI). To account for residual imbalance in demographic, medical, or surgical characteristics between the two groups, a multivariable regression analysis with a stepwise method was performed. All preoperative data with a P-value <0.1 in univariable analysis were entered into the model. The b coefficient with 95% CI was reported for all potential confounding variables associated with the outcome. Data analysis was by intention to treat, according to a pre-established analysis plan. No imputation for missing data was done.
randomisation, and none were lost to follow-up. The randomisation flowchart as per CONSORT 2010 Guidelines is shown in Figure 1. The majority of patients were male (93/100, 93%) and median age was 70 (64e75) yr. The two groups were balanced in terms of patient characteristics, revised cardiac risk index, renal function, or perioperative medical therapydincluding anticoagulation and antiplatelet with the exception of there being more female patients in the tranexamic acid group, which consequently had a lower BMI, and a slightly higher baseline INR in the tranexamic acid group (Table 1). The ratio of patients receiving tube and bifurcated grafts did not vary between groups. Use of cell saver was also similar between the two groups (Table 2). Haematocrit variations during and after surgery are shown in Supplementary Figure S1.
Results A total of 100 patients undergoing open AAA repair between March 2015 and October 2017 were randomised in the trial. Fifty patients were randomised to receive tranexamic acid and 50 patients to receive placebo. All randomised patients received their assigned treatment and were analysed for the primary outcome. All patients were contacted 1 yr after
Primary analyses The median (inter-quartile range) intraoperative blood loss (Table 2) was 400 (300e1050) ml in the tranexamic acid group vs 500 (360e1000) ml in the placebo group (P¼0.44). The rate of patients receiving red blood cell transfusion was seven/50 (14%) in the tranexamic acid group vs 12/50 (24%) in the
Fig 1. Study flowchart as per CONSORT guidelines. CONSORT, Consolidated Standards of Reporting Trials.
Tranexamic acid in open AAA surgery
-
5
Table 1 Baseline characteristics. Inter-quartile range (IQR) is presented as (Q1 [first quartile] e Q3 [third quartile]). AAA, abdominal aortic aneurysm; ACE, angiotensin converting enzyme; aPTT, activated partial thromboplastin time; ARB, angiotensin II receptor blockers; CABG, coronary artery bypass graft; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; INR, international normalised ratio; PTCA, percutaneous transluminal coronary angioplasty; SD, standard deviation; TIA,: transient ischaemic attack.
Age (yr), median (IQR) Male sex, n (%) BMI (kg m2), mean (SD) ASA physical status, n (%) 1 2 3 4 Revised cardiac risk index, n (%) 1 predictor 2 predictors 3 predictors Preoperative laboratory tests Creatinine (mg dl1), mean (SD) Haematocrit (%), mean (SD) Platelet count (109 ml1), mean (SD) aPTT (s), mean (SD) INR, median (IQR) Medical history, n (%) Smoking history COPD Hypertension Diabetes mellitus Dyslipidaemia Atrial fibrillation Ischaemic heart disease Previous stroke or TIA Previous vascular surgery Previous cardiac surgery Previous AAA repair surgery Previous PTCA or CABG, n (%) CKD, n (%) Stage 1 Stage 2 Stage 3 Preoperative medical therapy, n (%) Beta blockers ACE inhibitors ARBs Diuretics Ca2þ channel blockers Statins Aspirin Dual antiplatelet therapy Anticoagulant therapy
Tranexamic acid group (n¼50)
Control group (n¼50)
69 (63e75) 43 (86) 25.3 (3.33)
71 (65e74) 50 (100) 26.2 (2.64)
0 14 (28) 35 (70) 1 (2)
0 13 (26) 37 (74) 0
32 (64) 17 (34) 1 (2)
31 (62) 18 (36) 1 (2)
0.97 (0.31) 41 (3.8) 195 (54) 29 (3) 1.08 (1.03e1.15)
1.03 (0.27) 41 (3.6) 208 (59) 29 (2) 1.05 (1.00e1.10)
31 (62) 11 (22) 36 (72) 6 (12) 20 (40) 1 (2) 15 (30) 3 (6) 4 (8) 5 (10) 0 12 (24)
39 (78) 3 (6) 37 (74) 4 (8) 22 (44) 1 (2) 16 (32) 4 (8) 7 (14) 3 (6) 2 (4) 13 (26)
13 (26) 29 (58) 8 (16)
14 (28) 27 (54) 9 (18)
20 (40) 12 (24) 15 (30) 9 (18) 14 (28) 33 (66) 29 (58) 2 (4) 2 (4)
20 (40) 17 (34) 16 (32) 10 (20) 15 (30) 27 (54) 34 (68) 4 (8) 2 (4)
placebo group (P¼0.20). We did not find differences in the rate of thromboembolic events (including acute myocardial infarction) or acute kidney injury between the tranexamic acid and the placebo groups. We did not record any occurrence of pulmonary embolism, bowel infarction, or seizures. No death at 28 days was recorded in the present trial, whereas at the 1 yr follow-up, three patientsdall belonging to the control groupddied (0/50 vs 3/50, P¼0.24).
Secondary analyses Median (inter-quartile range) postoperative blood loss was reduced in the tranexamic group 4 h after surgery (60 [40e80] ml vs 100 [60e140] ml, P<0.001) and 24 h after surgery (180 [120e275] ml vs 275 [190e395] ml, P¼0.003) (Table 3). A
multivariable regression analysis was performed to account for residual imbalance. After correction for age, BMI, INR, smoking history, and chronic obstructive pulmonary disease (COPD), tranexamic acid administration was the only significant predictor (tranexamic acid protective) of bleeding (b¼e37 ml; 95% CI, e58 to e15; P¼0.001) (Supplementary Table S1) and 24 h (b¼e79 ml; 95% CI, e134 to e24 ml; P¼0.002) 4 h after surgery (Supplementary Table S2). A reduction of overall bleeding (from beginning of surgery to 24 h after surgery) was also observed in the tranexamic acid group, although it was not statistically significant (620 [450e1275] ml vs 920 [570e1220] ml, P¼0.13). Only one patient, belonging to the placebo group, required surgical reintervention (Table 3). Postoperative admission to the ICU was necessary in two/50 patients (4%) in the tranexamic acid group and
6
-
Monaco et al.
Table 2 Intraoperative data. Inter-quartile range (IQR) is presented as (Q1 [first quartile] e Q3 [third quartile]).
Surgical time (min), mean (SD) Clamping time (min), mean (SD) Tube graft, n (%) Bifurcated graft, n (%) Use of cell saver, n (%) Intraoperative blood loss (ml), median (IQR) Number of patients receiving blood products, n (%) Red blood cells, n (%) Fresh frozen plasma, n (%)
SD,
standard deviation.
Tranexamic acid group (n¼50)
Control group (n¼50)
P-value
137 (52) 36 (13) 40 (80) 10 (20) 16 (32) 400 (300e1050) 2 (4) 1 (2) 1 (2)
138 (49) 37 (14) 44 (88) 6 (12) 21 (42) 500 (360e1000) 3 (6) 2 (4) 1 (2)
0.45 0.34 0.41 0.41 0.41 0.44 0.65 0.56 0.99
two/50 patients (4%) in the control group (P¼0.99) because of intraoperative haemodynamic instability in two cases (one in each group) and major surgical complications in the other two cases (one in each group). All patients were discharged to the ward within the first postoperative day.
Discussion In this randomised report investigating tranexamic acid in major vascular surgery, we found no reduction in intraoperative blood loss in patients undergoing open AAA repair, although we observed a reduction in postoperative bleeding. No adverse or thrombotic event was reported in the present trial. Although modern-era vascular surgery is going towards endovascular procedures for treatment of abdominal aorta disease, current guidelines report that open surgery is still the treatment of choice in patients with long life expectancy or in those unsuitable for EVAR.2 Life-threatening haemorrhage in aortic aneurysm repair is a feared perioperative complication independently associated with increased mortality.25 Perioperative bleeding in patients undergoing open AAA repair relies on the fragility of the diseased aortic tissues and on
coagulation disorders. The coagulation impairment in this setting is multifactorial: an excessive fibrinolysis is frequently observed in patients with abdominal aortic disease, possibly because of the exposure of blood to thromboplastin within atherosclerotic plaques.26 In addition, intraoperative surgical bleeding, haemodilution, and consumption coagulopathy further worsen the blood loss. Previous literature demonstrated how the application of the aortic cross clamp induces an elevation of the fibrinolytic function and a reduction in clotting factors activity in AAA open repair.27 Excessive bleeding leads to blood transfusions, which are associated with worse postoperative outcomes. In fact, two large observational studies proved the relationship between perioperative transfusions and increased 30 day morbidity and mortality rates in major vascular surgery.28,29 In addition, allogenic blood is a limited resource and its use should be reserved to those situations where it is absolutely necessary. Thus, any strategy or drug with haemostatic properties capable of reducing bleeding and transfusion requirements without increasing adverse events may have a significant impact on daily clinical practice. However, in spite of being low-cost, widely available, and well renowned for its haemostatic properties, tranexamic acid was never previously
Table 3 Postoperative data and outcome variables. Inter-quartile range (IQR) is presented as (Q1 [first quartile] e Q3 [third quartile]). ARD, absolute risk difference; CI, confidence interval; RBC, red blood cells; RIFLE, Risk, Injury, Failure, Loss of kidney function, Endstage kidney disease; SD, standard deviation. Tranexamic acid group (n¼50) Blood loss 0e4 h after surgery (ml), median (IQR) 60 (40e80) Blood loss 0e24 h after surgery (ml), median (IQR) 180 (120e275) Blood loss from the beginning of surgery to 24 h 620 (450e1275) after surgery (ml), median (IQR) Patients receiving RBC in the postoperative period, 6 (12) n (%) Patients receiving RBC in the intraoperative period, 7 (14) postoperative period, or both, n (%) Postoperative ICU admission, n (%) 2 (4) Acute myocardial infarction, n (%) 2 (4) Acute kidney injury according to RIFLE criteria, n (%) Risk 8 (16) Injury 2 (4) Failure 0 Thrombosis, n (%) 0 Seizures, n (%) 0 Bleeding requiring reintervention, n (%) 0 Postoperative hospital stay (days), mean (SD) 6 (1.5) 30-day mortality, n (%) 0 1-yr mortality, n (%) 0
Control group (n¼50)
Estimated effect size Cliff’s d ± 95 % CI or ARD ± 95 % CI
P-value
100 (60e140) 275 (190e395) 920 (570e1220)
e0.41 (e0.19 to e0.59) e0.37 (e0.14 to e0.55)
<0.001 0.002 0.13
10 (20)
0.08 (e0.06 to 0.22)
0.28
12 (24)
0.1 (e0.05 to 0.25)
0.20
2 (4) 2 (4)
0.99 0.99
7 0 4 0 0 1 6 0 3
0.99 0.49 0.12 0.99 0.99 0.32 0.99
(14) (8)
(2) (1.2) (6%)
0.02 (e0.3 to 0.07) 0.05 (0.16 to e0.27) 0 0.06 (e0.01 to 0.13)
0.24
Tranexamic acid in open AAA surgery
investigated in major vascular surgery, because of sporadic reports of thrombotic events and stroke related to its use. Recently, the Aspirin and Tranexamic Acid for Coronary Artery Surgery (ATACAS) trial11 proved that tranexamic acid administration did not increase the risk of death or thrombotic complications in a population of more than 5000 cardiac surgery patients. Thromboembolic complications did not occur in the present trial, even if it was underpowered to detect safety issues. The ATACAS trial reported a higher incidence of postoperative benign seizures in the tranexamic acid group, whereas we did not observe any seizures in the present trial.11 As seizures are believed to be dose related in this setting, it is important to note that the ATACAS trial administered a dose of tranexamic acid five-to seven-fold higher than that used in the present trial,11 although the greater fluid shift and cardiopulmonary bypass-related haemodilution may partially explain the higher dosing in cardiac surgery. The dose of tranexamic acid in the present trial (an average of 1g in total) was determined on the basis of previous evidence in other settings,8,9 and of our previous clinical experience, to obtain a balance between the anti-fibrinolytic effects and the avoidance of the mentioned neurological complications. Although the dose-related sideeffects of tranexamic acid must be kept in mind, further studies are required to assess the effects of higher doses of tranexamic acid in this setting. The present trial is the first double-blind randomised trial to investigate the effects of tranexamic acid in patients undergoing major vascular surgery and one of the first ever to report complete 1 yr follow-up data. In our study, a reduction of postoperative blood losses was detected, whereas no intraoperative effect was observed in patients receiving tranexamic acid. We speculate that this is because intraoperative bleeding is mostly secondary to tissue dissection and aortic manipulation and requires surgical haemostasis, whereas postoperative bleeding is, for the most part, associated with microvascular bleeding. In this setting, tranexamic acid administered during surgery may decrease fibrin degradation and prevent postoperative microvascular bleeding. Previous clinical pharmacology studies showed how tranexamic acid is present in the blood in therapeutic concentrations up to 5 h after i.v. administration.5 Therefore, it seems reasonable that its anti-haemorrhagic properties are present in the postoperative period, particularly during the first 4 h after operation. The present study has several limitations and strengths. The sample size calculation was based on a historical estimation of intraoperative blood loss during AAA surgery that was higher compared with reported values in the present study. The trial was further underpowered to assess safety outcomes; therefore, even if in our study we did not observe thrombosis or mortality in the tranexamic group, no strong conclusions could be drawn from the present work. However, our findings are in agreement with recent literature in that tranexamic acid does not increase the risk of thromboembolic events.7,8,11 We also reported a reduction of postoperative bleeding in patients receiving intraoperative tranexamic acid. This exploratory post-hoc analysis, however, was based on drainage content. Although using drainage content to estimate postoperative bleeding was part of our pragmatic approach, this method does not offer a precise measure of how much blood was lost. Further trials should specifically focus on postoperative bleeding,
-
7
including measurements of blood content in postoperative drainages. This was the first randomised study ever performed focusing on tranexamic acid in the setting of open AAA repair. Moreover, the present study also reported a complete 1 yr follow-up (with a trend towards a lower mortality in the tranexamic acid group), which is rarely reported even in other, more extensively studied settings. Interestingly, the mean intraoperative blood loss in the overall study population was less than 500 ml per patient, which is much lower than those reported in previous literature. For example, Ferrante and colleagues30 in a recent study on AAA open repair reported a mean estimated intraoperative blood loss of 900 ml. Even in our centre, a previous retrospective study reported a high transfusion rate in the period 1993e2010.20 These differences could be explained by the improvement achieved in our centre, which is nowadays an international referral for aortic surgery. The important surgical learning curve observed in our centre dramatically reduced intraoperative blood loss in open AAA at our centre in the past decade and prevented us from finding a statistically significant reduction in intraoperative bleeding. In fact, although a 20% lower intraoperative blood loss and a 33% lower overall blood loss were observed in the present trial, these differences were not statistically significant. As intraoperative blood loss was lower than anticipated, the study was possibly underpowered to assess the primary outcome, and should therefore be considered a pilot trial. The same applies to perioperative transfusions, which were a lowlikelihood event in the present trial. In fact, although the transfusion rate was 14% in the tranexamic acid group vs 24% in the placebo group, no statistical significance was reached, possibly because the study was underpowered to detect it. These data suggest that a multimodal approach to blood sparing in major vascular surgery should at least consider tranexamic acid. In conclusion, the use of tranexamic acid in major vascular surgery did not reduce intraoperative blood loss, although it might have had a beneficial effect on postoperative bleeding. Further large, multicentric studiesdincluding the ongoing 10 000 patients PeriOperative ISchemic Evaluation-III trial31dwill help us understand the role of tranexamic acid in major clinical outcomes and survival in the perioperative setting of highrisk noncardiac surgery.
Authors’ contributions Study concept: LP Study design: GL, AZ Data acquisition: GG, GD Data analysis: PN, FM Data interpretation: CM, NDT, RC Drafting of the manuscript: PN, FM Revising manuscript: LP, GB, CM, NDT, GD, GG, GL, RC, AZ All authors approved the final version to be published; agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved
Acknowledgements We thank all the staff of Vascular Surgery Operating Rooms at IRCCS San Raffaele Hospital for their constant help and effort, making this study possible.
8
-
Monaco et al.
Declarations of interests The authors declare they have no conflicts of interests.
Funding
13.
14.
Rottapharm Biotech S.r.l. (an unrestricted grant).
Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.bja.2019.08.028.
15.
16.
References 1. Grondal N, Sogaard R, Lindholt JS. Baseline prevalence of abdominal aortic aneurysm, peripheral arterial disease and hypertension in men aged 65e74 years from a population screening study (VIVA trial). Br J Surg 2015; 102: 902e6 2. Wanhainen A, Verzini F, Van Herzeele I, et al. European Society for Vascular Surgery (ESVS) 2019 clinical practice guidelines on the management of abdominal aorto-iliac artery aneurysms. Eur J Vasc Endovasc Surg 2019; 57: 8e93 3. Davidovic LB, Maksic M, Koncar I, et al. Open repair of AAA in a high volume center. World J Surg 2017; 41: 884e91 4. Skagius E, Siegbahn A, Bergqvist D, Henriksson AE. Fibrinolysis in patients with an abdominal aortic aneurysm with special emphasis on rupture and shock. J Thromb Haemost 2008; 6: 147e50 5. Verstraete M. Clinical application of inhibitors of fibrinolysis. Drugs 1985; 29: 236e61 6. Soslau G, Horrow J, Brodsky I. Effect of tranexamic acid on platelet ADP during extracorporeal circulation. Am J Hematol 1991; 38: 113e9 7. CRASH-2 trial collaborators, Shakur H, Roberts I, Bautista R, et al. Effects of tranexamic acid on death, vascular occlusive events, and blood transfusion in trauma patients with significant haemorrhage (CRASH-2): a randomised, placebo-controlled trial. Lancet 2010; 376: 23e32 8. WOMAN Trial Collaborators. Effect of early tranexamic acid administration on mortality, hysterectomy, and other morbidities in women with post-partum haemorrhage (WOMAN): an international, randomised, doubleblind, placebo-controlled trial. Lancet 2017; 389: 2105e16 9. Crescenti A, Borghi G, Bignami E, et al. Intraoperative use of tranexamic acid to reduce transfusion rate in patients undergoing radical retropubic prostatectomy: double blind, randomised, placebo controlled trial. BMJ 2011; 343: d5701 J. 10. Colomina MJ, Koo M, Basora M, Pizones J, Mora L, Bago Intraoperative tranexamic acid use in major spine surgery in adults: a multicentre, randomized, placebo-controlled trial. Br J Anaesth 2017; 118: 380e90 11. Myles PS, Smith JA, Forbes A, et al. Tranexamic acid in patients undergoing coronary-artery surgery. N Engl J Med 2017; 376: 136e48 12. Casati V, Sandrelli L, Speziali G, Calori G, Grasso MA, Spagnolo S. Hemostatic effects of tranexamic acid in elective thoracic aortic surgery: a prospective,
17.
18. 19.
20.
21.
22.
23.
24. 25.
26.
27.
28.
29.
randomized, double-blind, placebo-controlled study. J Thorac Cardiovasc Surg 2002; 123: 1084e91 Gerstein NS, Kelly SP, Brierley JK. Yet another tranexamic acid-related thrombotic complication. J Cardiothorac Vasc Anesth 2016; 30. e21e2 Gupta PN, Mullamalla UR, Sabin P, Vellappan P. Acute MI in a young hypertensive woman: could it be due to tranexamic acid? BMJ Case Rep 2013; 2013. bcr2013009979 Taparia M, Cordingley FT, Leahy MF. Pulmonary embolism associated with tranexamic acid in severe acquired haemophilia. Eur J Haematol 2002; 68: 307e30 World Medical Association. World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 2013; 310: 2191e249 Schulz KF, Altman DG, Moher D, CONSORT Group. CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ 2010; 340: c332 Oprea AD, Popescu WM. Perioperative management of antiplatelet therapy. Br J Anaesth 2013; 111. i3e17 Møller A, Nielsen HB, Wetterslev J, et al. Low vs high hemoglobin trigger for transfusion in vascular surgery: a randomized clinical feasibility trial. Blood 2019; 133: 2639e50 Chiesa R, Tshomba Y, Psacharopulo D, et al. Open repair for infrarenal AAA: technical aspects. J Cardiovasc Surg 2012; 53: 119e31 Pasin L, Nardelli P, Landoni G, et al. Enhanced recovery after surgery program in elective infrarenal abdominal aortic aneurysm repair. J Cardiovasc Surg 2019; 60: 369e74 Henry DA, Carless PA, Moxey AJ, et al. Anti-fibrinolytic use for minimising perioperative allogeneic blood transfusion. Cochrane Database Syst Rev 2011; 3: CD001886 Ker K, Prieto-Merino D, Roberts I. Systematic review, meta-analysis and meta-regression of the effect of tranexamic acid on surgical blood loss. Br J Surg 2013; 100: 1271e9 Dunn CJ, Goa KL. Tranexamic acid: a review of its use in surgery and other indications. Drugs 1999; 57: 1005e32 Lederle FA, Freischlag JA, Kyriakides TC, et al. Open versus Endovascular Repair (OVER) Veterans Affairs Cooperative Study Group. Outcomes following endovascular vs open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009; 302: 1535e42 Gertler JP, Cambria RP, Brewster DC, et al. Coagulation changes during thoracoabdominal aneurysm repair. J Vasc Surg 1996; 24: 936e43 Adam DJ, Ludlam CA, Ruckley CV, Bradbury AW. Coagulation and fibrinolysis in patients undergoing operation for ruptured and nonruptured infrarenal abdominal aortic aneurysms. J Vasc Surg 1999; 30: 641e50 Obi AT, Park YJ, Bove P, et al. The association of perioperative transfusion with 30-day morbidity and mortality in patients undergoing major vascular surgery. J Vasc Surg 2015; 61: 1000e9 Osborne Z, Hanson K, Brooke BS, et al. Variation in transfusion practices and the association with perioperative adverse events in patients undergoing open abdominal aortic aneurysm repair and lower extremity arterial bypass in the vascular quality initiative. Ann Vasc Surg 2018; 46: 1e16
Tranexamic acid in open AAA surgery
30. Ferrante AMR, Moscato U, Snider F, Tshomba Y. Controversial results of the Revised Cardiac Risk Index in elective open repair of abdominal aortic aneurysms: retrospective analysis on a continuous series of 899 cases. Int J Cardiol 2019; 277: 224e8
-
9
31. PeriOperative ISchemic evaluation-3 trial (POISE-3). ClinicalTrials.gov Identifier: NCT03505723. Available from: https://clinicaltrials.gov/ct2/show/record/NCT03505723. [Accessed 29 September 2019]
Handling editor: C. Boer