Derivation of pulmonary capillary pressure from arterial occlusion in intact conditions

To investigate the reliability of the pulmonary capillary pressure measurement with the arterial occlusion technique.

Prospective, randomized, controlled study on anesthetized animals.

A cardiopulmonary research laboratory.

Seven healthy, mongrel dogs.

The animals were anesthetized, and left thoracotomy was performed. A 7-Fr pulmonary artery flotation catheter was inserted to monitor the pulmonary arterial pressure. Arterial flow to the left lower lobe was monitored with a cuff-type flow probe. A laser Doppler flow probe was placed on the surface of the left lower lobe to monitor flow in the microcirculation.

Arterial occlusions were performed by inflating the flotation balloon located in the left pulmonary artery. A monoexponential curve was fitted to a stretch of data between 0.2 to 2 secs post-occlusion and extrapolated back toward time zero. Time zero was defined as the instant when a change in the arterial pressure was first observed. When the balloon was inflated, pulmonary arterial flow and pressure decreased simultaneously; flow reached zero after 72 +/- 5 msecs, while pressure decreased rapidly and thereafter continued to decline more slowly. A change of flow in the main artery was followed by a change in microvascular flow with an 80 +/- 20 msec lag. Thus, if flow in the large arteries was at a peak or a nadir, a peak or a nadir flow in the microcirculation would occur 80 msecs later. Therefore, as a first approximation, we estimated that flow in the arterioles stopped 80 msecs after it had reached zero in the main artery. At this instant of time when flow in the arterioles stopped, the pressure across the arterial tree would have equilibrated. We calculated the arterial occlusion pressure at time zero (when pressure or flow began to change), at the time when pressure and flow had fully equilibrated across the arterial tree, and two other selected instants in between. The extrapolated pressure at these four instants were all < 1.1 mm Hg apart.

Back extrapolation of the postarterial occlusion data to 80 msecs after flow in the main artery reached zero, provided a physiologically correct estimate of capillary pressure. This approach would be equivalent to extrapolating to 152 msecs after the initial change in pressure was noted. Thus, precapillary pressure can be accurately estimated by identifying time zero as described above, fitting the data between 0.2 to approximately 2.0 secs to a -single exponential, and calculating the pressure on the curve at 152 msecs. However, under clinical conditions, only time zero is identifiable from the pressure tracings. Our results show that back extrapolation to any point between zero and 152 msecs is acceptable. The breakpoint on the arterial pressure tracing (if discernible) is perhaps most practical because it falls between zero and 152 msecs. In humans, wave transmission time generally would be in the same range, and thus, the same criteria may be applied.