Background: During tricuspid valve replacement in a patient with previous mitral valve surgery, we made an incidental observation that the right atrium can be opened without caval snaring and without air entering the venous reservoir. We tested this hypothesis on an animal model.
Methods: Two patients underwent right atrial surgery using percutaneous cannulation, and no air was entrained without caval snaring. This principle was tested in an animal model using 2 pigs weighing 80 kg each. Percutaneous cannulae were placed under epicardial echo guidance with their tips 4 cm from the right atrium. A “collapsible bag with air drainage system” was introduced into the venous return system to quantify air return from the superior vena cava (SVC) and inferior vena cava (IVC). Two types of percutaneous cannulae with (Cardiovations Quick Draw) and without (Biomedicus) proximal side holes were tested.
Results: In the animal model using Biomedicus cannulae, upon opening the right atrium, air was entrained from the SVC cannula at 60 mL/minute with no air in the IVC. There was no difference in the amount of air between the two cannulae. Pressures measured were 5 cm of water in the IVC and -20 cm water in the SVC. Epicardial ultrasound demonstrated complete collapse of both vena cavae. Partial clamping of the SVC cannula reduced the amount of air to 60 cc/min, and placing a small straight clamp at the SVC atrial junction eliminated the air. No air was noted in IVC cannula.
Conclusions: Inferior vena caval drainage by percutaneous cannula does not entrain air with either type of cannula and without snaring (both in clinical cases and animal model). This might be explained by the presence of a competent Eustachian valve. However, the SVC is not immune to air. Minimal air (approximately 60 mL/minute) could be managed by partial clamping or completely be avoided by placing a small straight clamp without snaring.
*Pennsylvania Hospital, University of Pennsylvania, Philadelphia, PA; and †Drexel University College of Medicine, Hahnemann Hospital, Philadelphia, PA.
Presented at the ISMICS 8th Annual Meeting, New York, NY, June 1–4, 2005.
Address correspondence and reprint requests to Dr. Dandolu Reddy, 230 West Washington Square, Farm Journal Building, Philadelphia, PA 19107; e-mail: Reddy.firstname.lastname@example.org.
Snaring to gain control of the inferior and superior vena cavae can be tedious and dangerous when performed through small incisions. This procedure is often performed through a 10-cm right minithoracotomy or by a totally endoscopic procedure (robotic surgery). During mobilization of the inferior vena cava (IVC), the hepatic veins, the right and left atria, or the pulmonary veins can be injured. Similarly the right pulmonary artery or atria can be injured during mobilization of the superior vena cavae. The risks are greater when port access surgery is done for redo cases.
During tricuspid valve replacement in a patient with previous mitral valve surgery, we made an incidental observation. Port access surgery was performed with a right 10-cm minithoracotomy and femoral vein (21 Fr) and right jugular vein (17 Fr) cannulation using long cannulae. As a result of adhesions and limited exposure and after several attempts, snaring of the vena cavae was abandoned. The end holes of both venous cannulae were positioned approximately 5 cm from the cava atrial junction using transesophageal echocardiographic guidance. Upon opening the right atrium, no air was noted in the IVC drainage line. A small amount of air was noted in the superior vena cava (SVC) line, which was successfully managed by the perfusion team by using a partial clamp. The procedure was successfully completed without caval snaring.
This technique was repeated during a case of atrial septal defect closure using port access technique and femoral and jugular cannulation without snaring the vena cavae. No air was noted in the IVC return cannula and only intermittent minimal air in the SVC cannula.
This study was performed to evaluate this technique in an animal model and to determine the quantity of air entrained. The results suggest potential use in a broader patient population.
Two pigs, each weighing 85 kg, were used. Institutional animal care committee approval was obtained. The subjects were premedicated using Terazol (tiletamine/zolazepam 4 mg/kg) and midazolam (0.01 mg/kg). The pigs were intubated and anesthesia maintained using isoflurane gas and intermittent boluses of Fentanyl 5 mcg/kg.
Midline sternotomy was performed, and the ascending aorta was cannulated after heparinization. Both groins were dissected and the femoral vessels exposed. The right internal jugular vein was exposed with a cut down. Two different types of venous cannulae with proximal side holes (Cardiovations Quick draw, Somerville, NJ) and without were used (Biomedicus, Medtronic, Minneapolis, MN) (Fig. 1).
We designed a system to quantitatively measure air return in each venous cannula. A collapsible air bag, each with an approximate 500 mL capacity, was inserted in each of the venous return lines with a 1-way valve. The bag was hung in a nondependent position, enabling air entering the line to rise. A 50-mL syringe was attached to the other end of the bag via a 3-way stopcock. Air was measured and removed with the syringe. An adjustable metallic clamp with a screw system was placed to create a differential vacuum in each venous return line (Fig. 1).
A standard roller pump with oxygenator and crystalloid prime of 1000 mL was used.
CONDUCT OF STUDY
A 21-Fr venous cannula with side holes (Cardiovations) was inserted via the right femoral vein. The tip was positioned 4 cm from the right atrium, guided by an epicardial ultrasound probe (Accuson) (Fig. 2). Similarly, a 21-Fr venous cannula with side holes was inserted into the right jugular vein with its tip positioned 4 cm from the SVC right atrial junction (Fig. 2). Cardiopulmonary bypass was started with an average flow rate of 4 L/minute, and normothermia was maintained. Vacuum assist of –20 cm was used which could be adjusted using adjustable clamps on venous return lines. The amount of air entrained via each cannula was measured after making a 5 cm opening in the lateral wall of the right atrium. At the same time blood in the right atrium entered from the coronary sinus and the vena cavae. Bypass was continued for 5 minutes and air return values noted. The aorta was cross-clamped for 5 minutes, thereby eliminating coronary venous return. The right atrium was completely cleared of blood, simulating conditions of an atriotomy. Air return from the venous cannulae was measured with an empty right atrium. The right atriotomy was then temporarily closed using a side biting clamp, and the pig was weaned off bypass and the venous cannulae removed. During various times, pressures were measured in the SVC and IVC.
Recannulation using 21-Fr venous cannulae without side holes was done via the femoral and jugular veins. Similarly, normothermic bypass was started and continued for 5 minutes. During this time, the amount of air entrained with the open right atrium was measured. Air-return values were again measured with the aorta cross-clamped and an empty right atrium. Pressures in the SVC and IVC were recorded during various times (Table 1).
At the end of this experiment, the animal was weaned off bypass and euthanized as per animal protocol.
During the animal 1 study (Table 1), there was no air entrained in either venous cannula with the right atrium closed. Upon initial cannulation using long cannulae with proximal side holes (Cardiovations) and open right atrium, air was noted only in SVC venous return line at approximately 60 mL/minute. This small amount of air in the SVC drainage line was managed by partially clamping it to reduce the degree of vacuum. Bypass was continued without air locking of the SVC drainage line. No air was noted in the IVC drainage line. However, upon cross clamping the aorta and emptying the right atrium, massive amounts of air were noted in the SVC drainage line with no air noted in the IVC line. Massive air return in the SVC drainage line could not be controlled by partial clamping. At this time the pressures measured –20 cm in SVC and 5 cm H2O in the IVC. Images from the epicardial ultrasound (Accuson) showed complete collapse of the vena cavae. A small clamp was easily placed at the SVC right atrial junction without the need for any dissection. This clamping maneuver completely eliminated air in the SVC drainage line.
Animal 1 was weaned off bypass; venous cannulae removed and recannulated using 21-Fr long venous cannula without proximal side holes (Biomedicus). The experiment was repeated with measurement of air return in the venous drainage lines with right atrium closed, open and emptied after cross clamping. There was no difference in the amount of air return with either type of cannula (Table 1).
Another pig weighing approximately 80 kg (animal 2) was used to repeat the experiment on a different day. Results from these animal studies are shown in Table 1.
Snaring of the vena cavae can be difficult and tedious when performed as part of minimally invasive cardiac procedures. This includes performing the procedure by a 10-cm right minithoracotomy as done for port access surgery or by a totally endoscopic procedure as in robotic surgery. Injury to pulmonary veins, atria, or vena cavae can occur and may result in catastrophic hemorrhage. The risk of injury is especially high in patients with previous cardiac surgery.1
In the animals used for this study, it was noted that if the venous cannula introduced via the femoral vein was positioned with its end hole at least 4 cm from the IVC–right atrial junction, no air was entrained. This was true using an end hole only type of cannula or using one with additional proximal side holes. No air was noted, even when the aorta was cross-clamped and the right atrium emptied. The pressures in the IVC during these periods were in the positive range of 5 cm of H2O, indicating closed-system drainage. This could be explained by a competent Eustachian valve. This valve opens only toward the right atrium, thereby preventing air entry into the IVC despite negative pressure exerted by the vacuum drainage system.
Air return was noted in the SVC drainage line when the right atrium was opened. In the clinical cases, the amount of air was minimal and was handled by partial clamping the line, which decreased the amount of negative pressure in the SVC. In our animal experience, large amounts of air entered the SVC drainage line when the atrium was completely emptied after aortic cross-clamping. The pressure in the SVC was noted to be in the range of –20 cm H2O. This could be explained by lack of a valve at its junction with the right atrium. Probably the drainage of the SVC into the right atrium is also assisted by gravity, thereby explaining no need for a valve. This air return was completely abolished by placing a small clamp at the SVC right atrial junction. This is easily accomplished without the need for dissection, as the SVC is completely collapsed by vacuum drainage.
The Eustachian valve is a semilunar valve attached by its convex margin to the anterior margin of the IVC orifice. The concave margin is free. The left side of the concave margin becomes continuous with the anterior edge of the limbus of the fossa ovalis while the right side is lost in the lateral wall of the right atrium. The valve is formed by a duplication of endocardium of the right atrium, enclosing few muscle fibers. During gestation, this valve is large and serves to direct flow from the IVC into the left atrium via the foramen ovale. The valve varies considerably in size and is sometimes cribriform, filamentous; occasionally it is absent.2–4 In an autopsy series, Yates found the Eustachian valve in 86% of 120 hearts with an average height of 3.6 mm.5 During transesophageal echocardiography study, the Eustachian valve is found at the level of coronary sinus and typically moves backward and forward twice for each cardiac cycle. The primary driving force for its motion appears to be the central venous pressure.6 Transesophageal echocardiography may be useful for planning cannulation in minimally invasive right atrial procedures.
The following observations can be made from our clinical and laboratory experience. IVC drainage without snaring using a cannula placed through femoral vein can be accomplished without entraining air. This seems to be true even when the right atrium is completely evacuated after aortic cross-clamping. SVC drainage using trans jugular cannulae without snaring results in entrainment of at least a moderate amount of air in the venous line. This could be potentially managed by using partial clamping of the drainage line or completely avoided by placing a small clamp at SVC right atrial junction. In a clinical scenario of surgery on the right heart, the atrium could be opened without snaring either cava, and perhaps the orifice of the SVC occluded with a balloon catheter. A new SVC drainage cannula with a balloon at its tip and proximal side holes for drainage might be useful.
1. Tevaearai HT, Mueller XM, Jegger D, et al. Optimization of the pump driven venous return for minimally invasive open heart surgery. Int J Artif Organs.
2. Stix MS, Bakola A, Lappas DG. Interpretation of Eustachian valve motion. J Cardiothorac Vasc Anesth.
3. Saric M, Rosenzweig BP, Kronzon I. Unusual Eustachian valve function. J Am Soc Echocardiogr.
4. Gray's Anatomy. 36th Edition
. Ed. Warren H. Lewis. New York: Williams & Warwick. 643.
5. Yates WM. Variations and anomalies of the venous valves of the right atrium of the human heart. Arch Pathol.
6. Limacher MC, Gutcesell HP, Vick GW, et al. Echocardiographic anatomy of the Eustachian valve. Am J Cardiol.