Pulmonary venous flow in cardiac tamponade: influence of left ventricular dysfunction and the relation to pulsus paradoxus.

The pattern of left atrial filling was studied in nine closed-chest dogs during cardiac tamponade before and after production of microembolic left ventricular dysfunction produced by intracoronary injection of 54 +/- 4 microns (SD) microspheres. With cardiac tamponade, a significant increase in the ratio of systolic/diastolic pulmonary venous flow velocity integral both before (1.65 +/- 0.24 versus 2.77 +/- 0.43 [SE], p less than 0.05) and after production of left ventricular dysfunction (0.57 +/- 0.12 versus 1.77 +/- 0.44, p less than 0.05) was seen. Compared with baseline, cardiac tamponade caused a significant inspiratory decrease in systolic pulmonary venous velocity both before (7.3 +/- 2.0 versus 1.2 +/- 1.4 cm/sec) and after left ventricular dysfunction (3.4 +/- 0.4 versus 1.0 +/- 0.9 cm/sec, both p less than 0.05). The magnitude of respiratory variation (expiration-inspiration) of the pulmonary venous flow velocity integral with tamponade was significantly greater before than after left ventricular dysfunction (1.6 +/- 0.2 cm versus 0.8 +/- 0.2 cm, p less than 0.05). A significant correlation was found between the inspiratory fall in aortic systolic pressure and the flow velocity integral of pulmonary venous flow before left ventricular dysfunction (r = 0.58, p less than 0.05). After coronary embolization, neither pulsus paradoxus nor significant respiratory variation (expiration-inspiration) of the pulmonary venous flow integral was observed with cardiac tamponade. In this model of cardiac tamponade and left ventricular dysfunction, left atrial filling occurs predominantly during ventricular systole. These changes may be helpful in recognizing hemodynamically significant pericardial effusion and have implications for the pathophysiology of cardiac tamponade.