Discontinuation of Cardiopulmonary Bypass


When discontinuation of CPB is imminent, the anesthesia provider should prepare for a safe and efficient transition to the native circulation so that full attention can be devoted to the patient during this critical period. One should attempt to anticipate potential adverse events, however to a certain extent this is not possible, hence the importance of adequate preparation (so that minimal time is spent doing rote tasks, ex. making up infusions). Necessary tasks include re-dosing paralytics and anesthetic agents (necessary as the patient rewarms), preparation of all infusions, loading of milrinone (if necessary), and calculating protamine doses. Several authors have attempted to predict ionotropic needs following CPB, the results of which are displayed below:

Predictors of Ionotrope Needs After CPB

Predictors of Ionotropic Use Following CPB (McKinlay. JCVA 18: 4, 2004, n=1009) Cross clamp time (OR 1.013 per minute) RSWMI (OR 4.2) Reoperation (OR 2.4) Combined CABG/MVR (OR 3.6) Moderate/Severe MR (OR 2.3) Preoperative EF < 35% (OR 2.4)
Other predictors from smaller studies include Low EF, advanced age, and history of CHF [JFB4. A&A 66: 461, 1998, n=149], low EF, old age, cardiac enlargement, female gender, and elevated LVEDP [Royster A&A 72: 729, 1991, n=128], and CI < 2.5, LVEDP >= 20, EF < 40%, and GFR < 60 mL/min/1.7BSA [Ahmed. J CTS 4:24, 2009, n=97]

The above data suggest that prolonged cross clamp time (especially during combined procedures), reoperation, mitral regurgitation, low preoperative EF/CI, advanced age, and elevated LVEDP are particularly worrisome for increased vasopressor requirements following CPB

Metabolic/Hematologic Optimization

The minimum acceptable [Hgb] for coming off bypass is not known but is classically taught to be around 7 g/dL. In patients with impaired ventricals or following non-optimal revascularization, increased oxygen carrying capacity may be indicated. Note, however, that a retrospective examination of over 8000 CABGs suggested that while a lower intraoperative HCT is associated with a higher risk of low-output cardiac failure, transfusion of RBCs does not ameliorate this risk and in fact may increase it, even when intraoperative [Hgb] < 7 g/dL [Surgenor SD et al. Circulation 114 1S: I43, 2006; FREE Full-text at Circulation]

Kaplan recommends administering bicarbonate to correct pH to near-normal levels prior to discontinuing cardiopulmonary bypass – while pH > 7.2 generally results in minimal cardiac depression, the post-CPB patient may be at increased risk [Kaplan JA, ed. Essentials of Cardiac Anesthesia. Saunders, 2008 p 535 (ISBN 978-1-4160-3786-6)]. Note, however, that the use of bicarbonate to correct pH is controversial, especially in light of the strong ion acid-base theory [Kellum JA. Critical Care 4: 6, 2000], which suggests that the hydroxide ions contributed by NaHCO3 will bind to H+ dissociated from water, which over a period of minutes negates the effect of adding OH-. A small study comparing NaHCO3 to NaCl for the treatment of metabolic acidosis (starting pH 7.22) showed a brief (30 mins) increase in pH with the bicarbonate group but no improvement in hemodynamics [Cooper DM et al. AIM 112: 492, 1990]. These results were essentially replicated in a similar study [Mathieu D et al. Critical Care Medicine 19: 1352, 1991]

Post-CPB patients are generally hypocalcemic. Because of the perceived risk of potentiating reperfusion injury, Kaplan recommends reserving Ca++ for patients who are both hypocalcemic and relatively unresponsive to vasopressors (i.e., 1) start your drips first 2) only add Ca++ if necessary and 3) be careful not to overshoot) [Kaplan JA, ed. Essentials of Cardiac Anesthesia. Saunders, 2008 p 535 (ISBN 978-1-4160-3786-6)]


SVR and MAP have a biphasic response to temperature – during true hypothermic CPB, SVR and MAP will tend to rise as temperatures are increased, however once ~ 32C is reached MAP and SVR will begin to fall again. The anesthesia provider should monitor temperature vigilantly – rapid rewarming (> 0.5 C/min in animal models) and/or excessive rewarming temperatures (> 37C in humans) are associated with poor neurologic outcomes [Grigore AM et al. Anesth Analg. 109: 1741, 2009; FREE Full-text at Anesthesia & Analgesia (beginning Dec 2010)]

In order to accelerate warming, increased pump flows and/or arterial vasodilators may be helpful – skeletal muscle and subcutaneous fat are particularly slow to rewarm

Reperfusion and Restoration of Native Circulation


Prior to removing the cross-clamp, the surgeon will often ask the perfusionist to partially clamp the venous line, allowing blood to pass into the right heart, lungs, ultimately filling the left ventricle so that intracardiac air can be identified and removed. The patient is often placed in head-down position to direct air (after cross-clamp removal) away from the carotids.

Cross-Clamp Removal

Just prior to cross-clamp removal, the surgeon will then ask the perfusionist to “go down on flows” i.e. temporarily lower the pressure in the aorta. When the cross-clamp is removed, the heart will be perfused by systemic blood (and not intermittent boluses of cardioplegia). In the case of a CABG, if the proximal anastomoses have not been completed, MAP may then be increased to 70 or 80.

V-fib treatment and prophylaxis

Defibrillation is commonly indicated following cross-clamp removal. Start with 5-10 J. Defibrillation is more effective if the heart has been adequately rewarmed. After four unsuccessful attempts at defibrillation, consider making attempts to further warm the heart, correct electrolyte abnormalities, and begin antiarrhythmic therapy. Note that 100 mg IV lidocaine just prior to cross clamp removal is superior to both amiodarone and placebo and significantly reduces the incidence of ventricular fibrillation after cross clamp removal [Ayoub CM et al. Eur J Anaesthesiol 26: 1056, 2009].

Metabolite Washout

As venous drainage is reduced (letting the patient “work”) in order to transition back to the native circulation, metabolic byproducts of the previously cold, ischemic heart will return directly to the circulation (rather than being picked up by the pump first) – some of the decrease in SVR that occurs near the termination of CPB is due to metabolite washout during reperfusion of the heart (as occurs with other organs), especially adenosine, which is a potent vasodilator.

Avoidance of Calcium

There are compelling reasons to believe that administration of calcium following myocardial ischemia (such as occurs during cardiopulmonary bypass) may be harmful [Chen RH. Ann Thorac Surg 62: 910, 1996].

Calcium Considerations according to Chen RH. Ann Thorac Surg 62: 910, 1996 1. After hypoxia/ischemia, intracellular Ca2+ is increased primarily by activities of the Na/Ca exchanger 2. K+ based cardioplegia, which raises the myocardial resting membrane potential, does NOT inhibit L-type Ca channels (which are activated at -20 to -30 mV) 3. Data from single-cell myocyte models suggests that myocardial cells that are reoxygenated prior to receiving Ca2+ survive, whereas those who receive Ca2+ prior to reoxygenation hypercontract and may not fully recover 4. Stunned myocardium actually has increased intracellular [Ca2+], and thus increased energy demand, despite low output 5. Transient intracellular hypercalcemia mimics myocardial stunning

In his excellent review of the subject, Chen states “…there is no justification for the use of calcium chloride before and in the early phase of reperfusion in cardiac surgical procedures. This practice can only increase the already high [Ca2+], and worsen the reperfusion injury. Conversely, hypocalcemic reperfusion in conjunction with hyperkalemia can divert the newly available ATP for the repair and recovery of cardiomyocytes” [Chen RH. Ann Thorac Surg 62: 910, 1996].