Postoperative Pericardial Flush, to evaluate the effects of pericardial flush with a crystalloid on blood loss after CABG.

This innovative prospective research project has to deliver the world*s first
proof of concept for the use of postoperative pericardial flush (PPF).
Continuous postoperative flushing of the pericardial cavity has never been
described and is essentially innovative. PPF has the potential to develop into
a versatile technique to control the pericardial environment; making cardiac
surgery procedures cleaner en safer. A reduction of common major complications
such as postoperative bleeding, atrial fibrillation and infection will lead to
a decrease of RBC transfusions and surgical re-exploration which are
independent risk factors for increased sternal wound infections,
transfusion-related complications and higher overall societal costs. When
proven (cost-)effective, PPF needs to be integrated into the standard
postoperative protocol for all cardiac surgical procedures, allowing it to be
used on a very large (global) scale in standard care. The primary clinical goal
is to reduce postoperative bleeding and decrease transfusion requirements.
Secondly, by reducing postoperative atrial fibrillation and preserving right
ventricular function improved patient outcomes may be expected.

Population
Cardiovascular disease (CVD) is a major cause of morbidity and mortality in the
world; in the Netherlands accounting for 39.735 annual in 2010, of which 18.581
(46.8%) men and 21.154 (53.2%) female. The proportion of CVD in the overall
mortality in 2010 between the ages of 0-54 was 14% (622 deaths) in women vs.
19% (1151 deaths) in men. The proportion of CVD in the total mortality above
the age of 85 is 37% (11,303 deaths) in women and 34% (5,163 deaths) in men.
[C] Within the group of CVD the majority of mortality is caused by ischemic
heart disease and stroke: 47% in women and 51% amongst men. Besides ischemic
heart disease and stroke (including cerebral infarction and cerebral
hemorrhage) the category 'other heart disease' (which includes heart failure
and atrial fibrillation) is a major contributor to mortality in CVD. In 2010
atrial fibrillation accounted for 3.2% of total CVD mortality with a surprising
1:2 ratio between women and men [C]. The intended study results will benefit an
increasing population. Over the last 15 years the number of open heart
surgeries performed yearly in The Netherlands has increased from 13.327 (1995)
to 16.128 (2009) (1107 and 1245 per million respectively). It is expected that
the number of open-heart surgery in 2020 will increase further to 18.420; which
might be an underestimation of the true increase. There are two important
factors that this assessment has not taken into account: an increase in the
number of valve and coronary bypass surgeries (CABG) on the basis of increasing
life expectancy and changes in indication for operation. [A,B] The expected
estimation of the number of open-heart surgeries in 2020 is therefore an
underestimation of the true number.

Blood loss and (pericardial) blood
Excessive postoperative blood loss (>2L/24 hours or >200 mL/hour) is one of the
most common complications of cardiac surgery. Predictive factors for hemorrhage
and reoperation/re-exploration following cardiac surgery include: age, obesity,
renal insufficiency, cardiopulmonary bypass (CPB) time and intracardiac repair.
Reoperation/re-exploration for bleeding is a strong independent risk factor for
adverse outcome following cardiac surgery. Specifically, operative mortality,
prolonged mechanical ventilation, acute respiratory distress syndrome, sepsis,
and atrial arrhythmias are increased in these patients. In addition,
postoperative bleeding requiring multiple transfusions and surgical
re-exploration is associated increased sternal wound infection,
transfusion-related infection and higher costs [1, 48]. A surgical cause of
bleeding is only found in half of patients undergoing
reoperation/re-exploration for bleeding. In the remainder of patients the cause
is multifactorial and probably an acquired/surgical related hemostatic defect
is responsible for diffuse mediastinal hemorrhage [1].
It is clear that during the first hours after cardiac surgery, clotting is
suboptimal while a large internal wound remains. Although strongly fixed
together, the two sternal halves are considered an important focus of
postoperative bleeding. The normal amount of blood loss after cardiac surgery
varies between 300 and 1500 mL during the first 12 hours. To evacuate blood
from the pericardial-, and if necessary pleural cavities, chest tubes are left
postoperative. However, when blood loss is excessive or when cloths start to
develop more rapidly, the drains often fail in their function to evacuate all
accumulated blood. Stasis of blood and cloths in the pericardial (and/or
pleural) cavity leads to high fibrinolytic activity, maintenance of blood loss
and in some cases to cardiac tamponade [2-6].
After removal of the chest tubes, often one day postoperatively, some blood and
cloths remain in the pericardial cavity. During the next days, under the
influence of fibrinolysis and hemolysis, a highly osmotic liquid solution with
high concentration of large molecular proteins (haemoglobin) is formed. This
osmotic active solution is known to contribute to the increase of pericardial
effusions and late cardiac tamponade in the first weeks postoperatively. The
blood and cloths that remain in the pericardial cavity can also induce an
inflammatory reaction that may play a role in the decline of right ventricular
function, the occurrence of atrial fibrillation and the formation of adhesions
[9-11, 46]. Besides directly causing patient discomfort and leading to
hemodynamic compromise, postoperative atrial fibrillation is associated with
worse clinical outcome regarding mortality, morbidity and costs [2-18].
During redo surgery for postoperative bleeding or tamponade, removal of blood
and cloths and flushing the pericardial cavity with warm saline is enough to
stop the bleeding immediately in a vast majority of patients. Reoperation for
bleeding is a strong independent risk factor for adverse outcome following
cardiac surgery. Specifically, operative mortality, prolonged mechanical
ventilation, acute respiratory distress syndrome, sepsis, and atrial
arrhythmias are increased in these patients. Controlling and stopping the
bleeding early postoperatively also means that anticoagulation therapy can be
started earlier; reducing the risk of thromboembolic events on the ICU.

Transfusion requirements
In addition, postoperative bleeding requiring multiple transfusions and
surgical re-exploration is associated with increased sternal wound infection
and transfusion-associated lung injury [7]. The incidence of blood transfusions
in cardiac surgery have been reported to vary from 27% to 90% [8]. It has been
pursued with the assumption that transfusing an anemic patient will improve the
outcome. Blood transfusion has a clearly defined role in the management of
hemorrhagic shock and is presumably beneficial in situations where a critically
low hematocrit is contributing to a state of oxygen-supply dependency. A number
of studies have demonstrated that low hemoglobin (Hb) concentrations and
decreased oxygen delivery increase mortality [19,20]. Defoe et al. [21] found
that patients who had a lower hematocrit during CABG surgery were associated
with a higher risk of in-hospital mortality [21]. The potential benefits are,
however, countered by many transfusion-associated complications: the risk of
transfusion-associated lung injury [7], transfusion associated immunomodulation
[22], transfusion-related circulatory overload [23], and cellular hypoxia [24].
Blood transfusions have also been linked to postoperative renal dysfunction
[25], pneumonia [26], wound infections [27] and severe sepsis [28].
There have been several recent well-designed randomized control trials in
critically ill or CABG patients showing a significant association of
transfusion with increased short- and long-term postoperative mortality of 66%
[29]; morbidity [30,31] and healthcare costs [32]. Different studies showed
that lower Hb and Ht levels increase mortality as well [19-21]. The short-term
adverse effects of blood transfusion in cardiac surgical patients are well
documented but there are few studies conducted assessing long-term survival
[29,33-35]. Others [36] stated that bleeding, increased chest tube production,
is associated with higher mortality through mechanisms not related to blood
transfusion. Despite all the available evidence, the transfusion practices vary
substantially [8,38]. Efforts made to decrease transfusion rates in cardiac
surgery show persistent effects for several years [39]. The use of blood and
other blood product transfusion in cardiac surgical patients still remains very
high. Adult cardiac surgery utilizes a significant proportion of all packed red
blood cell (PRBC) transfusions all over the world. The incidence of blood
transfusion in patients undergoing cardiac surgery has been reported to vary
from 27% to 90% [8]. Transfusion of a single RBC unit may increase the direct
medical costs due to hospitalization with 10% [40].

Atrial fibrillation
Besides directly causing patient discomfort and leading to hemodynamic
compromise, several studies have demonstrated that post-operative atrial
fibrillation (AF) is associated with: an increased risk of postoperative
stroke, represented by an annual stroke rate of 5% in patients with AF [12-14];
a higher rate of in-hospital mortality (5.8% vs. 2.2%, P=0.003) [12]; longer
intensive care unit and hospital stays [12,13,15,16,41] and higher costs of
treatment [17,18]. Both paroxysmal and chronic AF have shown a significant
increase of the risk of stroke, especially in older patients [42].

Right ventricular function
In CABG, right ventricular function is impaired directly postoperatively and
recovers to preoperative state levels within six months [49]. Various
hypotheses regarding the pathogenesis of the selective decline in RVF after
cardiac surgery have been put forward. However, no clear cause has been found.
It is possible that the thin-walled right ventricle may be more susceptible to
dysfunction secondary to inflammation or effusions postoperatively [50]. These
effusions may result from local tissue damage or from a systemic inflammatory
response. Another theory suggests that postoperative pericardial adhesions may
impair right ventricle filling. Prospective studies are needed to elucidate
this phenomenon.

Rationale
We believe that the occurrence of all previous outlined postoperative
complications is influenced by a common denominator that we consider to be a
pericardial cavity contaminated with blood. We hypothesize that a new
therapeutic technique, continuous postoperative flushing of the pericardial
cavity with saline (Postoperative Pericardial Flush) can contribute to a
cleaner pericardial space; hereby reducing postoperative blood loss,
transfusion requirements, atrial fibrillation, inflammatory response and
infection, and postoperative right ventricular function impairment. This will
improve clinical outcome with respect to mortality and morbidity, increase
HRQoL post cardiac surgery and decrease societal health-care costs.

Objectives of the randomized controlled trial
Primary objective:
To determine whether postoperative pericardial flush with a crystalloid can
reduce postoperative blood loss by effectively removing contaminated
pericardial blood after CABG. The primary study endpoints include: blood loss
12h postoperative and delta haemoglobin (between sternal closure and 12h
postoperative).

Secondary objectives:
To determine the effect of postoperative pericardial flush on:
- RBC transfusion (After randomization);
- In-hospital complications (Operative mortality, mortality, prolonged
mechanical ventilation, infection, atrial arrhythmias, surgical re-
exploration, sternal wound infection);
- Haemoglobin levels at discharge;
- ICU stay and total hospitalization duration;
- Adverse effects within the follow-up period of 6 months (Late cardiac
tamponade, infection, 30-day mortality; mortality within the follow-up period);
- Right ventricular function (6 months follow-up);
- Health related quality of life (HRQoL using EQ-5D+ at 6 months follow-up);
- Cost-effectiveness (Compared by assessing cost per quality adjusted life
year (QALY), calculated from the health utility gain scores obtained with the
EQ-5D+ questionnaire. Costs are divided in direct medical costs, consisting of
hospital days, ICU-days, medication, transfusion requirements, costs related
with in-hospital complications and indirect costs).


Objectives of the substudy
Primary objective:
To determine whether postoperative pericardial flush with a crystalloid can
reduce postoperative intrapericardial and systemic inflammatory response. The
primary study endpoints of the substudy include: difference in inflammatory
markers (IL-1, IL-6, IL-8, TNF-a, MCP-1, CRP, leukocytes, monocytes,
fibrinogen, D-Dimeer, albumin, Na, Cl, TF, and tPA) between sternal closure and
12h postoperative.

This study is designed as a prospective single-blind randomized trial with
evaluation of patient outcomes after a clinical follow-up of six months.
Patients are included consecutively within the AMC.

This study is supported by ZonMw within the *Doelmatigheidsonderzoek 2013-2015*
program for which approval was obtained on 15-10-2013. After re-approval of the
MERC this study will start a preparation period of 1 month (December 2013)
before the first patient is included. Our aim is to include 3 CABG patients a
week; patients are operated on Mondays, Tuesdays and Fridays. Inclusion of all
170 patients will take 12 months. Data analysis will be performed as soon as
the last patient has had his 6-month follow-up; data analysis will take 4-6
months.

Postoperative Pericardial Flush (PPF) system
The PPF system is aimed to extract the contaminated pericardial blood, dilute
thrombi, decrease postoperative blood loss, and prevent the formation of
postoperative pericardial adhesions via dilution and rotation of the fluid
stream. The perfusion fluid, NaCl 0,9%, has minor or no effects on the
pericardial tissue, haemostatic and fluid balance of the patient, but dilutes
possible thrombi. The pericardial inflow Redon drain will be inserted through a
small extra incision that is similar to the incision for the standard chest
tubes. This extra incision is needed because previous studies (safety and
feasibility study and a currently ongoing randomized trial in CHD patients)
have shown that leakage of the flushing fluid might occur when the inflow drain
is inserted through the same incision as one of the standard incisions for the
chest tubes. If necessary, a separate drainage tube will be inserted into the
pleural cavity. Every opened (pleural/pericardial) cavity is drained
separately. The outflow tract is a closed low vacuum drain collection system,
draining the pericardial cavity under low / negative pressure of 15cmH2O and
with a total volume of 2000mL. If required, fluid replacement or blood
transfusion will be given.

PPF will be performed continuously after operation, starting from the moment
the sternum is closed (T-1) until the total flushing volume of 7000ml has been
completely infused. After the Pericardial Flush system has stopped
approximately 14 hours post T-1, the system will be disconnected from the
inflow Redon drain. In the case of persistent blood loss after disconnection of
the PPF system, standard ICU thoracic drain protocol will be respected.

Implementation of the PPF system is standardized on a previous safety and
feasibility trial that was performed at our institution and on a similar
randomized clinical trial (in patients with congenital heart disease) which is
currently beeing performed at our institution (NL42595.018.12). PPF flow rate
(inflow) and thoracic drain production (outflow) are double monitored. Primary,
data is entered into a specially designed and integrated input field in the
ICU*s PDMS, which is done hourly by ICU staff. PDMS automatically calculates
the net thoracic drain production. Secondary, the VTBI is checked on the
infusion pump by a research nurse after 1h, 3h, 6h, 12h and 24h; the same is
done for thoracic drain production. Flow rate adjustments are made in the event
of fluid accumulation and or total drain obstruction. In the event of >200ml
fluid accumulation the PPF flow rate will be lowered to 100mL/hour and if not
evacuated during the next two hours, the PPF system will be stopped. In the
event of total thoracic drain / outflow tract obstruction the PPF will be
stopped until the obstructed chest tube(s) are functional again. During the
whole course of the experiment the patients* hemodynamics are monitored
extensively by continuous arterial and central venous line measurements
according to standard intensive care unit protocols. Routinely much attention
is given to signs of fluid accumulation in pericardial and/or pleural cavities.
When indicated chest X-ray or echocardiography will be performed to evaluate
fluid retention in pleural space and/or pericardial cavity. If required, fluid
replacement or blood transfusion will be given.

Inflow tract:
- NaCl 0,9% (7000ml via standard infusion line);
- Volumetric pump (flow rate of: 500mL/hour);
- Enflow® Blood and Fluid warming system;
- Perforated silicon drain.

Outflow tract:
- Perforated silicon drain (MEDICA Europe) 30Ch (CE0344);
- Standard low vacuum drain collection system.

Time points 6 months follow-up
T-D = One day preoperative;
T-4 = Baseline before induction of anesthesia;
T-3 = Before institution of CPB;
T-2 = After discontinuation of CPB;
T-1 = Closure sternum / start Pericardial Flush system (infusion of 7000ml at a
constant flow rate of 500ml/h);
T0 = ICU arrival;
T6 = ICU arrival + 6 hours;
T12 = ICU arrival + 12 hours;
(stop of the Pericardial Flush System after 14 hours);
T24 = ICU arrival + 24 hours.
TW = Arrival on thoracic surgery ward
T+D = Discharge from hospital
T+6M = 6 months follow-up

Additional blood samples
Additional blood samples on top of those needed for regular treatment will be
collected at six timepoints, as shown in Figure 3 of the study protocol. All
bloos samples from the systemic circulation will be drawn from the central
artery line that is part of the regular treatment. Sampling by itself will
therefore not be an extra burden for the patient. Extra samples include: blood
samples from the central artery line (intra- en postoperative), samples of
pericardial blood (intraoperative), and samples of mediastinal chest tube
drainage fluid (postoperative). Extra systemic blood samples will be collected
at the same moment and following blood samples that are needed for standard
treatment. Timepoints: before induction of anesthesia (1xEDTA and 1xCitrate
from the systemic circulation), before institution of CPB (1xEDTA and 1xCitrate
from the systemic circulation), after discontinuation of CPB (1xEDTA and
1xCitrate from the systemic circulation and 2 samples of pericardial blood),
before sternal closure (1xEDTA and 1xCitrate from the systemic circulation and
2 samples of pericardial blood), Arrival on ICU (1xEDTA and 1xCitrate from the
systemic circulation and 2 samples of mediastinal chest tube drainage), 12
hours after ICU arrival (1xEDTA and 1xCitrate from the systemic circulation and
2 samples of mediastinal chest tube drainage). Total samples per patient
needed: blood samples from central venous line (12), samples of pericardial
blood (4), samples of mediastinal chest tube drainage fluid (4). The following
markers will be determined per sample IL-1, IL-6, IL-8, TNF-a, MCP-1, CRP,
leukocytes, monocytes, fibrinogen, D-dimeer, albumin, Na, Cl, TF, and tPA.
Blood samples will be put into centrifuge directly after collection. Plasma
will then be distributes over 5 smaller samples. All samples will be labelled
and stored in a freezer at -80°C.

The development of expected adverse events, infection and fluid retention (in
pleural and pericardial cavities), will be closely monitored via several
imaging techniques and laboratory measurements. The blood samples necessary for
the study are not expected to negatively influence the result of treatment. A
possible benefit for a patient receiving the pericardial perfusion might be
active (central) warming when a patient suffers from hypothermia.

Adverse Events (AE)
An adverse event (AE) is defined as any undesirable experience occurring to a
subject during this trial, whether or not considered related to postoperative
pericardial flush. The difference between adverse events and serious adverse
events is that AE does not result in events as described in the section serious
adverse events. All adverse events observed by the investigator or his staff
will be recorded in the eCRF by the investigator.

Serious Adverse Events (SAEs)
This study population has a high risk of serious complications, which are
inherent to their vulnerable condition and unrelated to postoperative
pericardial flush, which is under evaluation in this trial. Complications are
included in the primary and secondary outcomes of this study and are recorded
in the Case Report Form (eCRF). All SAE*s are reported to the DSMB every 6
months during the 2 year study period. SAE*s that result in death or are life
threatening should be reported expedited. The expedited reporting will occur
not later than 7 days after the responsible investigator has first knowledge of
the adverse reaction. This is for a preliminary report with another 8 days for
completion of the report.