Systemic inflammation is seen in all surgery and trauma; however, it is more profound in cardiac surgery that uses cardiopulmonary bypass (CPB). The contact of blood components with non-biological surfaces such as the pump tubing, oxygenator membranes and altered propulsion by the roller pumps is a significant factor in amplifying the inflammatory effects of CPB. While CPB is currently unavoidable in valve replacement surgery, there are choices between conventional surgery using CPB and off-pump (OP) techniques in coronary surgery, especially if reducing the extent of the systemic inflammatory response syndrome (SIRS) is an objective.
As coronary surgery accounts for over 70% of cardiac surgery, the potential for minimising the deleterious effects of the SIRS has extensive implications as it would benefit large numbers of patients. However, surgical and pharmacological strategies to prevent or minimise SIRS may have untoward consequences with respect to the completeness of revascularisation, graft patency and the degree and duration of invasive monitoring.
Evidence of Systemic Inflammation
Massive release of inflammatory markers into the circulation occurs. These include activated complement components C3a and C5a, interleukins (IL-6, IL-8 and IL-1β) and tumour necrosis factor (TNF)-α – all cytokines that promote neutrophil and monocyte mobilisation in response and directed to the sites of inflammation and perceived vascular injury. In addition, many other vasoactive mediators are released, including endothelin, histamine, prostacyclin and bradykinin.1,2 Almost invariably the concentrations of these cytokines increase by between 10- and 100-fold over baseline, with peak-level increases during CPB occurring four to 12 hours after completion of CPB before falling back to normal by between 48 and 72 hours.2
In parallel, there is organ dysfunction of varying degrees, particularly in the lungs, kidneys and central nervous and gastrointestinal systems. These parallel changes are in part attributed to the intense affects of the newly released vasoactive molecules, but may also be due to the non-pulsatile flow associated with CPB, lower mean pressures and microemboli.1–3 Inflammatory lung injury appears to be due to the effects of activated leukocytes on the pulmonary capillaries resulting in increased vascular permeability, extravascular water content and shunting. In the brain, CPB is associated with an increase in extracellular fluid, as noted on magnetic resonance imaging (MRI).4
Post-operative, Physiological and Haemodynamic Effects
In previous decades, patients on CPB were usually cooled to 26–28ºC and were often hypothermic and hypertensive and had high systemic vascular resistance post-operatively.5 These physiological reactions to cold often masked or counteracted the effects of complement activation and circulating cytokines. Contemporary perfusion at normothermia or mild hypothermia has allowed greater expression of the vasoactive mediator effects in terms of measurable physiological and haemodynamic effects, particularly with respect to temperature, systolic and mean blood pressure, systemic vascular resistance and cardiac output in the post-operative period.6–8
Causes of Systemic Inflammatory Response Syndromes and Their Relative Contribution
Operative trauma is a potent stimulus, particularly when vascularised tissue is breached. In coronary surgery there is extensive trauma to cutaneous, subcutaneous, muscular and periosteal tissues and marrow and vascular injury (to varying extents) via sternotomy, mediastinal, internal thoracic artery and additional conduit harvesting.9 With specific reference to OP coronary surgery, similar patterns of substantial cytokine and vasoactive mediator release are seen, although of a lesser intensity.2,10,11 The anti-inflammatory cytokine IL- 10 is released to similar levels in both forms of coronary surgery,12 implying that the sternotomy, conduit harvesting, mediastinal dissection and cardiac handling are significant and possibly the most important factors. CPB, with exposure of blood to artificial surfaces and areas of endovascular trauma (aorta, right atrium), mechanical disruption of blood elements and scavenged shed blood suction, is an additional contributor. The length of CPB and surgery may also influence the degree of SIRS.1–4,13
Clinically Relevant Responses to Systemic Inflammation
The patient’s temperature usually rises from baseline upon return to the intensive care unit (ICU) by 1.5ºC. It usually peaks at 12 hours post-operatively and falls towards baseline thereafter. Typical temperature readings are 36ºC upon return to the ICU, a peak of 37.5ºC 12 hours post-operatively, returning to 36.5ºC–37ºC by 24 hours postoperatively, presuming there is no other pathology nor intervention.6,8
Similarly, cardiac index (CI) tends to increase over the first 12 hours post-operatively by approximately 20%, typically from a CI of 2.5l/min/m2 baseline post-operatively to a peak of 3.0–3.1l/min/m2 and tending back towards the baseline of 2.5l/min/m2 by 24 hours.6,8 Conversely, systemic vascular resistance index (SVRI) falls (reflecting significant systemic vasodilatation post-operatively) from a baseline of 2,200–2,500dyne/cm5/m2 to 1,800–2,000dyne/cm5/m2 at 12 hours post-operatively – approximately a 20% fall – and then tends towards a baseline of 2,200–2,500dyne/cm5/m2 by 24 hours.6,8
These patterns of temperature, CI and SVRI are all consistent and parallel and complement each other. In addition, they exactly follow the patterns of cytokine and vasoactive mediator released into the circulation and the clinical vascular effects of these chemicals.1,2,4,14 These responses abate as the mediators are cleared from the circulation. In a recent study of post-operative haemodynamic function following coronary bypass surgery performed at normothermia or mild hypothermia (33–35ºC), we found that up to 40% of patients had extremely low SVRI (less than 1,500dyne/cm5/m2) and up to a further 56% of patients had a normal SVRI (1,500–2,900dyne/cm5/m2) in the immediate 24 hours post-coronary surgery, i.e. up to 96% of patients had low or normal SVRI. These findings were in parallel with high cardiac outputs and CIs, with up to 98% of patients having a CI consistently lower than 2.2l/min/m2 in the first 24 hours,6 supporting similar findings from a less extensive study by Velissaris and colleagues.8
Does Off-pump Coronary Surgery Reduce the Systemic Inflammatory Response?
At a biochemical level, the activation of white cells, complement system, cytokines and vasoactive mediators are all significantly reduced in OP coronary surgery compared with conventional coronary artery bypass graft (CABG) using CPB. Although many studies indicating this are not randomised (and generally the OP surgeries are shorter and have fewer grafts), in all randomised comparisons there is convincing and significant evidence to support a less profound systemic response at a cellular and biochemical level. In general, the patterns of neutrophil activity and cytokine and mediator release are similar, peaking at one to six hours post-operatively for both OPCAB and conventional CABG. However, concentrations of elastase, C3a, C5a, IL-6, IL-8 and TNF-α were 25–50% lower in OPCAB than in conventional CABG.2,4,7,14
At practical and clinical levels, the post-operative physiological and haemodynamic indicators of temperature, blood pressure, SVR and CI all support the notion of a significant haemodynamic response to systemic inflammation following coronary surgery. However, a randomised study with 50 patients in each group in which continuous measurement of T, BP, CVP, PAW, PA pressures and hourly calculations of cardiac index and SVRI were performed showed no difference between OPCAB and conventional CABG in these haemodynamic parameters.6 We found that the SVRI fell immediately post-operatively and continued to fall for the first 18 hours. The mean CI rose and the highest mean CI was achieved at 12–18 hours post-operatively. There was no difference between OPCAB and on-pump. As noted above, up to 96% of patients had a normal to low SVRI in the first 24 hours and there was no difference between groups. This was also true for normal to high CIs and the infusion of the vasopressor noradrenaline (up to 10% of patients required NA in the first 24 hours; results were similar in both groups). In the few published studies on the post-operative haemodynamics and patterns of SVR, no difference has been found between OPCAB and conventional CABG.6,8
How does one account for this? Some have postulated that the trauma of surgery (incisions, conduit harvesting, mediastinal dissection and cardiac manipulation), which is similar in both OPCAB and ONCAB, is the dominant driver of SIRS, rather than CPB.9 Also, once a certain level of white cell, complement activation, cytokine and vasoactive mediator release has been achieved, i.e. a certain threshold for a significant effect, further release, or higher circulating levels of these mediators, does not have any further vasoactive effects.
To a degree, vasodilation and low SVR in the immediate post-operative phase is beneficial as it reduces pre-load and afterload, thus optimising the haemodynamic milieu for the revascularised heart, provided that systemic blood pressure changes are not too profound, i.e. systolic BP is maintained at >90mmHg and mean BP remains at >70mmHg, so that coronary and general organ perfusion pressures are maintained.6
Clinical outcomes, such as pulmonary and neurocognitive functions that are affected by SIRS and CPB, have not shown a benefit from OPCAB.1,2,4,7 Observational, matched and randomised trials have not been able to show any difference in the degree of pulmonary dysfunction.2,7,9
Similarly, multiple studies that have specifically attempted to address neurocognitive dysfunction post-coronary surgery have all failed to show any benefit for OP cardiac surgery in the short and intermediate term, and also at five years post-operatively.15 These findings suggest that inflammatory insults over and above CPB dominate the pathological changes in the lungs and brain.
The incidence of renal dysfunction is similar in off- and on-pump CABG for the majority of patients. However, randomised and case-matched studies have shown a clear benefit for OPCAB in those patients with pre-operative renal dysfunction (which often includes older patients and those with more extensive coronary artery disease and generalised atherosclerosis).2,13,16,17 However, the reason for this may have more to do with the maintenance of pulsatile flow and higher systolic and mean pressures during OPCAB, rather than the effects of the inflammatory response. Blood loss, transfusion requirements and discharge haemoglobin are all improved by OPCAB.2,13,16–18 Thus far it has not been possible to separate the contributions of the different anticoagulation regimens.
Other Prophylactic Measures
The use of corticosteroids has a sound theoretical basis and has been shown to reduce release of IL-6, IL-8 and TNF-α, and to reduce complement activation. However, no clinical benefit has been demonstrated, and furthermore their use may potentiate hyperglycaemia and deep sternal and other infections.19
Aprotinin reduces the release of cytokines and vasoactive mediators. Although its primary use in cardiac surgery is a protease inhibitor antifibrinolytic agent to reduce blood loss, it appears to have powerful effects in modulating and at least partly blocking the inflammatory response through proteolytic enzyme inhibition, diminished neutrophil and complement activation and reduced cytokine release. Clinically, aprotinin is neuroprotective in coronary surgery, but its use in routine coronary surgery, whether on- or off-pump, is restrained by its cost and reports of possible reduced graft patencies and an increased incidence of renal dysfunction.20 Antioxidants, particularly vitamins C and E, have been shown to be of benefit in the resuscitation of patients with major trauma, but these benefits have not been translated to coronary artery surgery.21
Management of Systemic Inflammatory Response Syndromes, Low Systemic Vascular Resistance and High Cardiac Output States
As noted, low SVR may be beneficial in reducing afterload and optimising pre-load conditions for the left ventricle. However, blood pressures <90mmHg systolic and <70mmHg mean may compromise perfusion. Hence, if such blood pressure parameters are reached and the patient is clearly in a high cardiac output state (CI >2.5l/min/m2) in conjunction with a documented low SVRI (<1,500dyne/cm5/m2), this should be treated. Our initial strategy is to infuse 1l of colloid followed by 1l of crystalloid (depending on the patient size). If the BP parameters are not achieved, i.e. systolic >90mmHg and mean >70mmHg, we commence a noradrenaline infusion, the usual required dose being 1–5μg/min. As the SIRS haemodynamic effects on SVR and CI last for at least 18 hours, the noradrenaline infusion will usually be required for that period of time.6
Low systemic vascular resistance and high cardiac output states secondary to vasodilatation from systemic inflammatory response are common (almost universal) after coronary surgery conducted at normothermia or mild hypothermia. It occurs with equal frequency in both on- and off-pump surgery. Although the intensity of the inflammatory response appears to be much less in OP, according to biological markers there is no difference in clinical outcomes except for degrees of renal dysfunction in high-risk patients and transfusion requirements – both of which may have other explanations. OP coronary surgery does not protect against vasodilation and low SVR states. However, the clinical results from either form of coronary surgery are excellent, and low SVR states do not necessarily present a significant clinical problem.