Vasodilation and Cardiopulmonary Bypass

The Role of Bradykinin and the Pulmonary Vascular Endothelium

Low systemic vascular resistance (SVR) commonly occurs during and early after cardiopulmonary bypass (CPB), and is easy to accept as a bothersome side effect of bypass since it is usually transient and easy to treat. Occasionally, patients have a more severe and persistent fall in SVR, referred to as postoperative vasodilatory shock.¹ Treatment is frequently required to maintain adequate perfusion pressure during CPB and to establish satisfactory hemodynamics when ready to separate the patient from bypass. This has usually entailed counteracting the effect of the vasodilatory mediators by administering adrenergic vasoconstrictor drugs such as phenylephrine or norepinephrine. Although usually effective and safe, these drugs can redistribute blood flow in such a way as to compromise the splanchnic and renal circulations.² Another etiology of low SVR that has been described¹³ is abnormally low vasopressin levels that respond quite well to low-dose vasopressin infusion even when the usual vasoconstrictor drugs are not effective in restoring vascular tone. There are a number of known responsible mediators, including bradykinin, produced during CPB, many of which are related to the systemic inflammatory response elicited by blood contactwith the extracorporeal circuit and the physiologic changes in the circulation with CPB.⁴

In this issue of CHEST (see page 1776), Cugno et al have documented an increase in bradykinin levels during CPB and have correlated these levels with a decrease in SVR. Moreover, they have shown that bradykinin levels drop when circulation is reestablished to the pulmonary vascular bed, but still do not return to baseline levels at the end of operation. Bradykinin is one of the vasoactive substances known to be released in response to CPB, but the authors point out that documentation of levels has been difficult because of the laboratory methods needed to accurately determine levels.

Other mediators involved in the inflammatory response to CPB were also measured, but a direct correlation of these with bradykinin levels was not found despite substantially elevated levels. The levels of tumor necrosis factor that they obtained were considerably higher than those reported in similar patients by other investigators,⁵ and raises the question as to whether their patients experienced greater systemic inflammatory response to CPB than usually documented. Their findings also support activation of the fibrinolytic system, and although they found no direct correlation between simultaneously measured levels of bradykinin and variables associated with fibrinolysis, both increased substantially over the period of CPB.

Bradykinin is a potent vasoactive polypeptide that causes vasodilatation and increased capillary membrane permeability⁶⁷ and also stimulates release of tissue plasminogen activator from peripheral vascular endothelium.⁸ The consequences of increased circulating levels are loss of SVR, sequestration of fluid in tissues, and increased fibrinolysis and bleeding,² all problems associated with CPB. The primary site of metabolism and inactivation of bradykinin is the pulmonary vascular endothelium.⁹ Under normal circumstances, bradykinin is rapidly metabolized, with > 80% being deactivated with first pass through the lungs.⁹ The enzyme responsible for its deactivation is angiotensin-converting enzyme (ACE), which also converts angiotensin I to angiotensin II, also important in modulating SVR.⁹

Preoperative administration of ACE inhibitors as well as reduced left ventricular systolic function are independent risk factors for low SVR and vasodilatory shock after operations requiring CPB.¹¹⁰ Indeed, one of the reported mechanisms for effectiveness of ACE inhibitors for hypertension control is an increased level of bradykinin that may be especially important in patients with low renin hypertension.¹¹ None of the patients in the study by Cugno et al were receiving ACE inhibitors before operation. However, had some patients been receiving one of these drugs, one would expect them to have had even higher bradykinin levels, further delay in metabolism after reestablishing full pulmonary circulation, and, consequently, lower SVR from bradykinin effect and inhibition of angiotensin I to angiotensin II conversion.

The potential for significant clinical benefit by reducing perioperative bradykinin release was not discussed by Cugno et al, but should not be ignored. The use of aprotinin in patients undergoing cardiac surgery requiring CPB has been directed primarily to decreasing bleeding during re-operative open cardiac surgery. Aprotinin is also a potent inhibitor of kallikrein and the kinin system, which are activated by CPB and result in the production of bradykinin.⁷¹² Because of evidence that aprotinin also decreases the systemic inflammatory response to CPB,¹³ in part because of kallikrein and kininogen inhibition, use of aprotinin has been liberalized in many institutions. There is some evidence from an earlier study¹² of aprotinin in CPB that it may prevent the fall in SVR by this mechanism, making this an intervention worth further study, particularly in patients receiving ACE inhibitors preoperatively.

This study serves also to rather neatly highlight and explain a number of observations regarding the physiology of CPB and its effect on SVR and suggests management modifications that might be beneficial. During standard full CPB, the lungs are excluded from the circulation except for the small amount of bronchial blood flow. Since the pulmonary vascular endothelium has important metabolic functions, including deactivation of bradykinin, serotonin, leukotrienes, and norepinephrine as well as production of angiotensin II, prostacyclin,⁹ and endothelin 1,¹⁴ isolation of the lungs can have important consequences. The authors noted that bradykinin levels decreased in patients in whom partial CPB was established for a period of time before full separation from CPB. Furthermore, they measured levels across the lung in a selected smaller group of patients and found that 60% of the bradykinin was removed in one pass through the lungs. This reduced transpulmonary metabolism together with continued elevation of bradykinin at the end of the operation, after many passes through the pulmonary vascular bed, suggests that some degree of pulmonary endothelial dysfunction was also present.

It is common practice to keep the patient on full CPB, bypassing the lungs, until shortly before the patient is ready for separation from bypass. This delays the potentially beneficial effect of blood contact with the pulmonary vascular endothelium and may result in more prolonged need for pharmacologic support of SVR both during the rewarming period as well as early after separation from CPB. However, it has been our practice to resume circulation to the lungs by establishing partial bypass and cardiac ejection, as soon as cardiac recovery is satisfactory. Normal “full flow” CPB is maintained, but controlled retard on the venous return line maintains low normal filling pressures and permits cardiac ejection, augmenting systemic and pulmonary blood flow. This provides opportunity for blood contact with the pulmonary vascular bed to metabolize high levels of bradykinin and to resume the normal conversion of angiotensin I to angiotensin II, thus mitigating two of the potential causes of intraoperative and early postoperative decreased SVR. The delay in full metabolism of bradykinin and the lower-than-expected transpulmonary gradient of bradykinin described by Cugno et al are evidence of pulmonary vascular endothelial dysfunction that may also be partially ameliorated by earlier resumption of pulmonary blood flow.¹⁵ This technique has the added benefit of providing pulsatile flow and perhaps better organ perfusion during recovery and rewarming.

There is certainly more to the pathophysiology of CPB and particularly to the abnormal vasodilatation that results than transient elevation of bradykinin levels.⁴ Contact activation in the extracorporeal circuit may be decreased with albumin pretreatment¹⁶ or with heparin-coated circuits,⁴ and newer direct inhibitors of other inflammatory mediators in addition to aprotinin may all be of help. Meanwhile, awareness of the important functions of the pulmonary vascular endothelium may better guide the management of CPB to help restore satisfactory levels of SVR earlier after cardiac repair.