Troubleshooting critical events during cardiopulmonary bypass (CPB) requires vigilance, expertise, and communication with both the surgeon and perfusionist. Obstruction of the CPB circuit or oxygenator results in circulatory arrest, which can lead to catastrophic sequelae for the patient. We describe the diagnosis, management, and a probable etiology of acute cessation of CPB.
Informed consent was obtained from the patient’s family for publication of this report.
An 80-year-old woman with a history of critical aortic stenosis, atrial fibrillation, left ventricular systolic and diastolic dysfunction, and obstructive coronary artery disease of the left main (50% stenosis) and posterior descending (70% stenosis) coronary arteries was admitted for aortic valve replacement and cardiac artery bypass graft.
After induction of anesthesia, 5 g aminocaproic acid was administered as a bolus, which was followed by a 1 g/h infusion (5 g was also used to prime the CPB machine). A median sternotomy was then performed while the left internal mammary artery was dissected concurrently with saphenous vein harvesting. Fifteen thousand units of unfractionated heparin was administered by the intracardiac route into the right atrium 4 minutes before initiating CPB resulting in an increase of activated clotting time (ACT) from a baseline of 122 to 705 seconds. CPB was initiated after cannulation of the ascending aorta and the right atrium. ACT levels were checked every 30 minutes and ranged between 501 and 804.
The aortic valve was replaced with a 23-mm Magna Ease bioprosthetic valve (Edwards Lifesciences, Irvine, CA). After closure of the aortotomy, the surgeon proceeded to anastomose the posterior descending artery to the saphenous vein graft. During this time, the right atrium and ventricle became suddenly distended, while drainage from the venous cannula abruptly decreased causing the patient’s mean arterial blood pressure to decrease to 30 mm Hg. We routinely use conventional gravity siphon venous drainage in our institution to achieve satisfactory venous drainage. As venous return decreased, vacuum suction was applied to the venous cannula; however, this paradoxically worsened the blood return. The venous cannula was repositioned multiple times without improvement.
Transesophageal echocardiography (TEE) revealed a distended right atrium containing a large echogenic mass that was a mobile clot occluding the venous cannula. The surgeon proceeded to bicaval cannulation of the superior and inferior cavae which allowed CPB to be resumed at full flow. The single-stage venous cannula was removed along with an attached clot. The right atrium was then opened, and an organized 2.1 × 2.0 × 0.9 cm clot was evacuated. The decision was made to abort bypass of the left main coronary artery as rewarming was nearly complete and there was concern for the possibility of further clotting.
Separation from the CPB machine was otherwise uneventful, and no clots were seen on the post-CPB TEE. She wa s, however, found to have a soleal vein thrombus in the midcalf during her postoperative workup.
We describe an acute circulatory failure during CPB from mechanical occlusion of the venous cannula by a blood clot not observed using TEE before CPB. While other conditions such as malposition of the venous cannula, kinks or obstruction of the venous lines by a smaller cannula, or airlock could cause acute disruption in venous return, these are not consistent with the echocardiographic findings. It is unlikely that physiologic states such as venodilation or hypovolemia would cause such an acute disruption in venous return. The most likely source of the organized clot was from dislodgement from a lower extremity or pelvic vein and embolization to the right side of the heart during CPB. Postoperatively, the patient was indeed found to have a soleal thrombus in the left lower extremity.
This event was captured in real time by TEE and probably is indicative of the natural process of developing a pulmonary embolism from a dislodged thrombus. Midesophageal TEE images (Figs. 1 and 2) revealed a mobile, dense echogenic mass in the distended right atrium, which was not present before initiation of CPB (Fig. 3). The mass occluded the venous inflow cannula in a ball-and-socket mechanism, leading to a near cessation of venous inflow. This provides an explanation of decreased inflow with the application of suction to the venous cannula as the thrombus was pulled toward the cannula causing further occlusion (Fig. 2). Although unlikely, there remains the possibility of the clot being generated from contact of blood with the extracorporeal circuit during CPB despite appropriate target levels of ACT. A full TEE examination was performed both before and after CPB without identification of thrombus. Furthermore, the organized pathologic appearance of the thrombus makes the acute process of clot generation even more unlikely.
In 2012, Kim et al.1 reported generation of a left atrial thrombus during CPB in a patient undergoing mitral valve repair in which the ACT was maintained at >500 seconds. They cite extensive thrombin generation during CPB initiated through contact of blood with the extracorporeal circuit, which was not fully inhibited by heparin as a possible etiology.1 A second report describes the spontaneous formation of blood clots in both arterial and venous circulation despite adequate anticoagulation; however, the patient was found to have heparin-induced thrombocytopenia and thrombosis.2 Testing for heparin-induced thrombocytopenia was preformed postoperatively in our patient and was negative. King and Kane3 report 2 instances of clots forming in the oxygenator during CPB, but both occurred in instances where CPB had to be reestablished after protamine administration with ACTs of 100 and 135 seconds. Our hypothesis is that manipulation of the left lower extremity during saphenous vein harvesting in a sedentary patient with a history of deep vein thrombosis may have been instrumental in having a clot dislodged in the operating room. This would make sense given the organized nature of the clot, which would be different from a fresh clot formed de novo. While genetic hypercoagulable states such as factor V Leiden, protein C or S deficiency, prothrombin G20210A mutation, and antiphospholipid antibodies predispose patients to the development of deep vein thrombosis, they have not been reported to have caused clotting during CPB. From epidemiological data, we know that a pulmonary embolism event is not uncommon in the perioperative setting of cardiac surgery, with a series of 1033 patients demonstrating an incidence of 3.7%.4 While rare, the possibility of thrombus migration, especially in sedentary patients undergoing vein harvesting should be considered if venous inflow abruptly decreases during CPB. This as well as other etiologies such as malposition or migration of the venous cannula can be easily visualized using TEE.
1. Kim SY, Song JW, Jang YS, Kwak YL. Formation of intracardiac thrombus during cardiopulmonary bypass despite full heparinization and adequate activated clotting time -A case report-. Korean J Anesthesiol. 2012;62:571–4
2. Legare JF, Arora R, Wood JW. Massive intracavitary clot formation during cardiopulmonary bypass. Can J Cardiol. 2004;20:825–6
3. King DO, Kane PB. Blood clot formation during multiple cardiopulmonary bypass procedures necessitated by massive hemorrhage. Anesth Analg. 1978;57:273–6
4. Josa M, Siouffi SY, Silverman AB, Barsamian EM, Khuri SF, Sharma GV. Pulmonary embolism after cardiac surgery. J Am Coll Cardiol. 1993;21:990–6