A Persistent Left Superior Vena Cava: Erroneous Measurement of Cardiac Output

Abstract

To the Editor: Placement of a pulmonary artery catheter (PAC) through a persistent left superior vena cava (PLSVC) may lead to erroneous cardiac output (CO) measurements. We report a 42-yr-old male who was scheduled for elective aortic valve reconstruction. After induction of anesthesia, a PAC was placed without difficulty and obviously in a correct distal position as confirmed by pressure recording (at 60 cm). However, pulmonary artery thermodilution measurement revealed a double-hump suggesting a shunt phenomenon (Fig. 1A).Figure 1.: (A) Thermodilution curves with a double-hump as obtained with the injection site of the pulmonary artery catheter in the coronary sinus. (B) Thermodilution curves without a double-hump as obtained after repositioning of the pulmonary artery catheter with a correct injection site in the right atrium.The PAC was retracted prior to cardiopulmonary bypass and re-advanced (at 65 cm) after successful weaning. The thermodilution curve then did not show a double-hump shape and CO was more than doubled (8.6 L min−1) (Fig. 1B). A postoperative chest radiograph (Fig. 2A) and computed tomography scan (Fig. 2B) revealed the PAC in a PLSVC.Figure 2.: (A) Chest radiograph showing a persistent left superior vena cava (PLSVC). Arrow head = tip of the pulmonary artery catheter, arrow = central venous catheter. (B) Chest computed tomography scan showing the position of the pulmonary artery catheter through the coronary sinus (indicated by the arrow). Arrow head = persistent left superior vena cava (PLSVC).In general, a PLSVC, being the most common thoracic venous anomaly (0.5%), is of no pathological consequence (1). In up to 92% of cases, the PLSVC drains into the right atrium through the coronary sinus (2) and patients with a PLSVC have an increased risk for atrial fibrillation due to bilateral pacemaker areas in the myocardium (3). Measurement of CO by the thermodilution technique is based on the Stewart–Hamilton principle: CO = m0/AUC Δt dt (m0 = mass of the indicator injected, AUC = area under the temperature curve). Whenever the AUC is increased (here by “re-circulation”) while keeping the amount of indicator injected unchanged, CO will be under-estimated. In our case, cold saline was injected into the coronary sinus yielding a pulmonary artery thermodilution curve with a double-hump which is similar to one observable in left-to-right shunt during which the pulmonary artery thermodilution curve is typically prolonged. Since a part of the indicator that has already passed the tip of the PAC re-circulates, a second, but less pronounced, decrease in temperature appears (4). In our patient, an intracardiac shunt could be excluded by echocardiography. Therefore, “re-circulation” from the coronary sinus into the right atrium was the cause of the described double-hump phenomenon. Consequently, the first AUC was artificially elevated, explaining the underestimation of CO. Re-positioning of the PAC to an increased insertion depth of about 65 cm resulted in a correct right atrial injection site, and a higher CO was measured without the “double-hump” phenomenon. In summary, a PAC advanced through a PLSVC may cause erroneous measurements of CO and particular attention has to be paid to the shape of the thermodilution curve in this setting. Wolfram Schummer, MD Claudia Schummer, MD Samir G.Sakka, MD, PhD Department of Anesthesiology and Intensive Care Medicine Friedrich-Schiller-University of Jena D-07740 Jena, Germany [email protected]