Peri-Operative ISchemic Evaluation-3 Trial

Globally more than 200 million adults undergo major noncardiac surgery
annually. Despite the benefits of surgery, a substantial proportion of patients
suffer major complications.
The Vascular events In noncardiac Surgery patIents cOhort evaluatioN (VISION)
study (a large, international, prospective cohort study evaluating
complications after noncardiac surgery) showed that among a representative
sample of 40,000 adults >=45 years of age undergoing noncardiac surgery, 1.8%
died within the first 30 days after surgery (confidential data, not yet
published). In the VISION cohort, perioperative bleeding accounted for
approximately 25% of the deaths (confidential data, not yet published). There
is encouraging but not definitive evidence that tranexamic acid (TXA) may
prevent perioperative bleeding in noncardiac surgery. Moreover, TXA is an
anti-fibrinolytic drug for which in vivo data shows an association with
thrombus formation, and its safety in the prothrombotic noncardiac surgery
setting has not been established.
There is evidence suggesting that perioperative hypertension is associated with
cardiovascular complications. Despite some variability across practices,
routine care is to continue antihypertensive medications in the perioperative
setting in patients on chronic therapy. Despite the potential risk of
perioperative hypertension, there is also evidence that perioperative
hypotension is independently associated with all-cause death and cardiovascular
complications at 30 days after noncardiac surgery.
We previously performed two large prophylactic perioperative randomized
controlled trials (RCTs) - the PeriOperative ISchemic Evaluation (POISE) and
POISE-2 Trials - evaluating the efficacy of metoprolol, clonidine, and aspirin
(ASA). POISE and POISE-2 demonstrated the independent association between
perioperative bleeding and cardiovascular events. In these trials clinically
important hypotension was independently associated with myocardial infarction,
stroke, acute kidney injury (AKI), and death.
We will conduct a large international RCT (i.e., POISE-3) of TXA versus
placebo, and using a partial factorial design, we will evaluate a
hypotension-avoidance strategy versus a hypertension-avoidance strategy, in
patients having noncardiac surgery.

Perioperative bleeding in the setting of noncardiac surgery is strongly
associated with 30-day cardiovascular complications and mortality. Trial
evidence suggests that intravenous TXA reduces perioperative bleeding and
transfusion in orthopedic surgery; however, the data are based on small trials,
and few non-orthopedic noncardiac surgery patients have been included in
perioperative TXA trials. Moreover, the safety of TXA in the noncardiac surgery
setting (i.e., the effect on arterial and venous thrombotic events) has not
been established. There is the need for a large, adequately powered trial to
establish definitive evidence and drive subsequent practice. In particular, to
support a more extensive use in noncardiac surgery, a reliable proof of the
non-inferiority of TXA compared with placebo in terms of safety, together with
a confirmation of its efficacy in different types of noncardiac surgeries, is
required.
Although usual care is consistent with a hypertensive avoidance strategy in the
noncardiac surgery setting, there is compelling evidence that perioperative
hypotension is frequent and is associated with cardiovascular events and
mortality. There is no definitive evidence from RCTs to support the adherence
to specific perioperative BP targets, nor the adoption of strategies of
perioperative management of antihypertensive medications. Moreover, there is no
clear evidence that hemodynamics in one of the perioperative phases (i.e. pre-,
intra-, or post-surgery) has more impact on outcomes than in other phases.
There is the need to compare a hypotensive-avoidance strategy to a
hypertensive-avoidance strategy in patients undergoing noncardiac surgery to
determine the impact on major cardiovascular complications and death.

The POISE-3 Trial is an international RCT of 10,000 patients with, or at risk,
or cardiovascular disease who are randomized to TXA or placebo given
intraoperatively. Patients, health care providers, data collectors, outcome
adjudicators, and investigators (e.g., Steering Committee Members) will all be
blind to allocation to TXA or placebo. The POISE-3 Trial will utilize a partial
2x2 factorial design to randomize patients taking >=1 antihypertensive therapy
to perioperative (i.e., pre-, intra-, and post-operative phase)
hypotension-avoidance strategy vs. hypertension-avoidance strategy. Outcomes
adjudicators will be blind to treatment allocation for the partial factorial.

1. Tranexamic acid or placebo
Within 20 minutes preceding the anticipated skin incision, patients will
receive intravenous TXA or placebo (0.9% normal saline) at a loading dose of 1g
over 10 minutes, with a second 1g bolus given at the end of surgery when
closing the wound.

2. Perioperative blood pressure management strategies
The perioperative BP management strategies will have three components, (i.e.,
preoperative, intraoperative and postoperative for the first 2 days after the
day of surgery). The intervention is a perioperative BP management strategy
aimed at avoiding hypotension. The control group is a perioperative BP
management strategy aimed at avoiding hypertension (usual care). For the
preoperative BP management component, the antihypertensive medications taken by
the patients enrolled in the trial will be either their own medications, or
provided by the hospital pharmacy as per each centre*s routine practice. For
the postoperative BP management phase, the hospital pharmacy at each center
will provide the patient*s medications as per routine practice.

Perioperative TXA may increase the risk of seizure. TXA has been shown to be a
predictor for postoperative seizure in patients undergoing cardiac surgery,
independently of other possible contributing factors, like ischemia, embolic
phenomena, drugs, and hypoglycemia. High doses of TXA play a role in increasing
the risk, and thus preoperative renal dysfunction can promote the occurrence of
these events (being the drug elimination mostly renal). High doses of TXA are
usually implemented in cardiac surgery (from 50 mg/kg to more than 100 mg/kg).
In the recent ATACAS study postoperative seizures occurred in 15 patients
(0.7%) in the TXA and in 2 patients (0.1%) in the placebo group (relative risk,
7.60; 95% CI, 1.80-68.70t). No difference was found between the higher (100
mg/kg) and the lower dose (50 mg/kg), but both doses were in fact higher than
the dose that will be used in POISE-3.
In the POISE-3 trial, we will exclude patients with poor renal function (eGFR
<30 ml/min) and patients with a history of seizure disorder. Further, we will
use a standardized dose of TXA (1 g after anesthesia induction and 1 g at the
end of surgery), which corresponds to a maximum cumulative dose of 40 mg/kg in
a 50 kg patient. Any postoperative seizure will be recorded.

Patients will be administered blood pressure medication that they are already
taking chronically; the decision to administer or withhold the medication will
be advised by the trial. The patients own physician will have prescribed these
chronic medications. Thus, patients are expected to be stable under these
chronic medications both in terms of tolerance and BP control.
The trial intervention group is aimed at avoiding hypotension. However,
clinically significant hypertension will also be minimized because patients
with elevated SBP >130 mm Hg in the intervention group will be instructed to
take certain antihypertensive medications as per Appendix III. Moreover,
further management will be left at the discretion of the attending physician,
including choice of additional therapy, if indicated.
The trial control group is aimed at avoiding hypertension. Should clinically
important hypotension occur physicians are encouraged to manage this, but
choice of corrective therapy will be left at the discretion of treating
physician.