Published ahead of print on December 12, 2002, doi:10.1164/rccm.200208-840OC
© 2003 American Thoracic Society
Left Atrial Pressure Can Be Accurately Transmitted to the Pulmonary Artery despite Zone 1 ConditionsDenver Health Medical Center and Department of Medicine, University of Colorado Health Sciences Center, Denver, Colorado; and Department of Medicine, University of Washington Medical Center, Seattle, Washington Correspondence and requests for reprints should be addressed to Richard K. Albert, M.D., Denver Health Medical Center, 777 Bannock, MC 4000, Denver, CO 80204-4507. E-mail: Ralbert{at}dhha.org
Pulmonary arterial occlusion pressure is not thought to reflect left atrial pressure (Pla) when alveolar pressure (PA) exceeds pulmonary venous pressure because alveolar capillaries collapse and the required continuous fluid column between the pulmonary artery and left atrium is interrupted. However, arterial-to-venous flow can occur when PA exceeds both the pulmonary arterial pressure (Ppa) and pulmonary venous pressure (i.e., in Zone 1 conditions), indicating the existence of a continuous patent vascular channel. Accordingly, Ppa should reflect Pla under these conditions. To investigate this connection cannulas were placed in the pulmonary arteries and left atria of eight excised rabbit lungs. Ppa and Pla were set 5 cm H2O above PA, which ranged from 0 to 25 cm H2O. Pla was then reduced in 2 to 4 cm H2O decrements while recording Ppa when arterial-to-venous flow ceased. At all PAs greater than 0 cm H2O, Pla was accurately reflected by the Ppa when both were exceeded by PA. The greater the PA, the lower the Ppa could track Pla below PA. Pla can be accurately measured by a pulmonary arterial catheter under Zone 1 conditions.
Key Words: pulmonary circulation • ARDS • pulmonary arterial wedge pressure The pulmonary arterial occlusion pressure (Ppao) cannot accurately reflect left atrial pressure (Pla) unless a continuous column of fluid connects the tip of the pulmonary arterial catheter with a downstream pulmonary vein that, in turn, empties into the left atrium. The zonal perfusion model states that arterial-to-venous flow ceases when PA exceeds pulmonary arterial (Ppa) and pulmonary venous pressure (Ppv) because the alveolar capillaries collapse (i.e., Zone 1 conditions) (1). A similar situation exists in Zone 2 (Ppa > PA > Ppv) when the balloon of a pulmonary artery catheter is inflated, as any effect of Ppa on pulmonary capillary pressure is removed, and the capillaries will again collapse when PA exceeds Ppv. Accordingly, chapters in numerous pulmonary and critical care medicine textbooks (2–7) and reviews of hemodynamic monitoring (8–12) make the point that Ppao cannot reflect Pla if the tip of the pulmonary arterial catheter is located in a region of lung in which Zone 1 or 2 conditions exist. We (13–16) and others (17–22) have previously noted that pulmonary arterial-to-venous flow can occur when the lung is completely in Zone 1. Accordingly, a continuous fluid channel connecting the tip of a pulmonary arterial catheter to a downstream pulmonary vein must exist. The purpose of this study was to determine if such a functional pathway allowed measurement of Pla when PA exceeded both Ppa and Ppv.
Isolated Lung Preparation The experimental protocol was approved by the University of Washington Animal Care Committee. New Zealand white rabbits of either sex (n = 8, weighing 2.7–3.0 kg) were anesthetized with pentobarbital sodium (30–50 mg/kg) and given papaverine (3 mg/kg) and heparin (1,000 U) through an ear vein. A tracheotomy was performed and an endotracheal tube was inserted. Animals were mechanically ventilated with a constant-volume piston pump (Harvard Apparatus Co., South Natick, MA), with a VT approximately 10 ml/kg and a rate of 30 breaths per minute. A median sternotomy was performed, the animals were exsanguinated, and cannulas were placed in the pulmonary artery and left atrium. The heart–lung block was placed on the dorsal lung surface, in a warmed, humidified chamber. The lungs were perfused with 30 to 60 ml of TRIS-buffered Tyrode's solution containing 2% albumin, 4% dextran (both from Sigma Chemical, St. Louis, MO) and 15 mg/L papaverine from the arterial reservoir until the venous outflow appeared to be free of blood. The venous outflow was then directed through a narrow horizontal tube that allowed observation of individual drops of fluid at low flow. Ppa and Ppv were measured at the tips of their respective cannulas. The zero pressure reference was the dorsal lung surface. Ventilation was stopped and static PA was adjusted per the experimental protocol. All pressures were measured with a Graphtec Mark 12 Data Managed System, DMS 1000 (Irvine, CA).
Experimental Protocol In three lungs the above protocol was performed at PA = 5, 15, and 25 cm H2O (set on deflation) before and after generating pulmonary edema. Edema was produced by lowering PA to 10 cm H2O and raising Ppa and Ppv to 30 cm H2O for 5 minutes. At the end of the experiment, all lungs were weighed and dried in a vacuum oven until the weight was stable and a wet/dry ratio was determined. The data were plotted with Ppv on the x axis and Ppa on the y axis, and second order polynomial equations of the raw data were used to calculate Y intercepts on inflation and deflation and before and after creating pulmonary edema. These were compared using Students paired t test. A p value of less than 0.05 was considered to be statistically significant.
Pla could be accurately measured in the upstream pulmonary artery when Pla was less than PA at all PAs greater than 0 cm H2O (Figure 1) . An effect of the lung volume history was also apparent. At PAs greater than 5 and less than 20 this effect was statistically significant (Figure 1). Subtracting the mean Y intercept on inflation and deflation from the PAs at which they were measured demonstrates that the greater the PA, the more Pla could be reduced and still be accurately reflected by Ppa (Figure 2) . The mean ± SD wet/dry ratio of the five lungs studied without causing pulmonary edema was 6.1 ± 0.7. All of the polynomial equations for the individual studies at all PAs had R values greater than or equal to 0.99.
Pla was accurately reflected in the upstream pulmonary artery when Pla was less than PA at PA = 5, 15, and 25 cm H2O despite the presence of pulmonary edema (Figure 3) . The Y intercepts of the three lungs increased after creating edema, and the change reached statistical significance at PA = 15 and 25 cm H2O. Again, the greater the PA the more Pla could be reduced and still be accurately reflected by Ppa, but edema had the effect of reducing the distance into Zone 1 that Ppv could be lowered and still be reflected by Ppa (Figure 4) . The mean ± SD wet/dry ratio of the three lungs studied before and after causing pulmonary edema was 16.8 ± 2.6 kg. All the polynomial equations for the individual studies done at PA = 15 and 25 cm H2O had R values greater than or equal to 0.98. At PA = 5 cm H2O, five of the R values were greater than or equal to 0.97 and one was 0.77.
The important findings of this study are that (1) Pla can be accurately measured in the upstream pulmonary artery when both Ppa and Pla are below PA (i.e., when the entire lung is entirely in Zone 1 conditions), (2) the distance below PA that Pla can still be measured in the pulmonary artery increases with increasing PA, (3) the transmission of Pla to the pulmonary artery is reduced but can still occur despite considerable lung edema, and (4) the patency of the vascular channel connecting the left atrium to the pulmonary artery under Zone 1 conditions is, at least in part, dependent on alveolar surface tension.
Methodology We held PA constant to eliminate potential cyclic changes in zonal conditions and to avoid questions pertaining to time constants of changing zonal conditions. Several studies have demonstrated that pulmonary venules and veins can constrict in response to hypoxia and a variety of vasoactive mediators (23–25). Accordingly, pressures measured in the pulmonary artery during flow could be considerably higher than Pla in the setting of pulmonary venoconstriction. We therefore measured Ppa under no-flow conditions and also administered papaverine to eliminate any potential effects of changes in arterial or venous resistance. Although the no-flow condition also mimics the situation that would exist when the balloon of a pulmonary arterial catheter is inflated, the duration of balloon occlusion in patients is generally far shorter than the time we allowed for pressure equilibration (frequently > 5 minutes). Using a cell-free perfusate reduced any potential problems resulting from settling, hemoconcentration, or clogging of vessels occurring during the time they were exposed to a constant Ppa, Pla, and PA under no-flow conditions. We have directly observed Zone 1 flow in animals with normal hematocrits in vivo (15), although the rate of flow is reduced when the perfusate contains blood (17).
Ability of Upstream Pressures to Reflect Downstream Pressures in Zone 1 To answer this question we reviewed each of these studies in detail. In some, it was not possible to discern the specific zonal conditions because the method used to establish the zero pressure reference was not specified (26, 28, 32) or the measurements were not made under Zone 1 conditions (31). In other studies, only summary data were presented such that the vascular and alveolar pressures in the individual experiments could not be determined (28, 30). Three studies presented individual pressure measurements made under well-described zonal conditions. Roy and colleagues (29) found that when the pulmonary artery catheter was an average of 2.5 cm above the left atrium, Ppao did not reflect Pla when Pla was less than PA (i.e., mean Ppao = 9.9 and 15.2 cm H2O at PAs of 10 and 15 cm H2O, respectively, whereas mean Pla was 6.9 and 8.9 cm H2O, respectively). In the 21 measurements made when PA = 10 cm H2O and Pla was less than PA, however, Ppao was identical to or less than or equal to 0.7 cm H2O above Pla in five, and was actually 2 cm H2O below Pla in one. When PA = 15 cm H2O, Ppao ranged as high as 8.8 cm H2O above PA and exceeded PA by more than 2 cm H2O in 10/24 studies. The duration of balloon occlusion in these experiments was 30 seconds. Similar disparities between Ppao and positive end-expiratory pressure (PEEP) were made for catheters with tips located above the left atrium by Berryhill and Benumof (35), who found that Ppao averaged 4.3 cm H2O above PA. The duration of balloon occlusion was not specified. Although Ppao might reflect PA under Zone 1 conditions (i.e., critical closure occurring at the alveolar level), explanations for how Ppao could exceed PA, particularly by as much as 8.8 cm H2O, are not as clear. We can propose five possibilities. First, there might be problems with the zero pressure reference. Second, the position of the pulmonary arterial catheter or the diameter of the vessel in which it is placed might change as PEEP is increased such that during balloon occlusion the catheters might become over-wedged leading to false high readings as has been reported previously (4, 6). Third, because these studies were done with positive pressure ventilation continuing as PEEP was adjusted, the higher PA produced during tidal inflation could have exceeded the closing pressure for what would otherwise have been a patent Zone 1 pathway at the lower level of PEEP. Fourth, pulmonary venous resistance could increase at higher levels of PEEP (24), and this could retard the runoff of blood from the tip of the pulmonary artery catheter to the left atrium. If this occurred, and insufficient time were given for arterial-to-venous flow to cease, pressures would not equilibrate. Others studies, however, have demonstrated that pulmonary venous resistance actually decreases with increasing levels of PEEP (13, 14). Finally, disparities between Ppao and Pla will increase in direct proportion to the resistance in the circuit downstream of the occlusion and the time allowed for equilibration to occur. The studies noted previously interrupted arterial-to-venous flow for 1 minute or less. We allowed flow to continue until less than 1 drop of perfusate drained from the circulation in 10 seconds. This frequently required up to 5 minutes. Interestingly, Roy and colleagues (29) also found that when the pulmonary artery catheter was approximately 5 cm below the left atrium, Ppao tracked Pla rather nicely, even when the lung was in Zone 1 (i.e., Ppao = 4.9 and 7.8 cm H2O at PAs of 10 and 15 cm H2O, whereas Pla was 4.2 and 5.9 cm H2O, respectively). This finding suggests that catheter location, rather than the specific zonal conditions, might determine whether downstream pressures can be measured by catheters in the upstream segment. Hasan and colleagues (33) correlated Ppao and Pla in vivo in animals with normal lungs and in others with unilateral or bilateral lung injury. Ppao rather accurately reflected Pla (i.e., a difference of < 2 mm Hg when Pla was < 9 mm Hg) in animals with normal lungs and those with bilateral injury, until PEEP reached or exceeded 15 mm Hg. At a PEEP of 22 mm Hg, however, Ppao was not able to track Pla in normal lungs when Pla was 13 mm Hg below PA or when Pla was 13.8 mm Hg below PA in lungs with bilateral injury. Although we did not distend our lungs to 22 mm Hg, we found that, at a PA of 25 cm H2O (i.e., 18.3 mm Hg), we could track Pla in the pulmonary artery when Pla was as much as 10 to 15 cm H2O (i.e., 7.3–11.0 mm Hg) below PA in normal lungs and as much as 10 to 12.5 cm H2O (i.e., 7.3–11 mm Hg) below PA in edematous lungs. Most importantly, in animals with unilateral lung injury, Hasan and colleagues (33) found that Pla was rather accurately reflected by Ppao when Pla was as much as 20 cm H2O into Zone 1 (see their Figure 4). In the normal lungs of animals with unilateral injury, however, the ability of Ppao to track Pla was lost at PEEP levels of greater than approximately 5 cm H2O. The considerable disparity between the Ppao-Pla difference measured when both lungs were normal versus that measured when the pulmonary arterial catheter was positioned in the normal lung of animals with unilateral lung injury was not addressed (33).
Effects of Edema The patency of the pulmonary arterial-to-venous pathway in Zone 1 lungs depends on the presence of surface tension (13, 14, 17–22). When surface tension is either very low as a result of a low PA (Figure 1, PA = 0 cm H2O) or when it is eliminated by completely filling the lung with liquid (14), the pathway is collapsed. Accordingly, some of the differing results in the literature may be explained by differences in lung volume and/or degrees of edema or injury.
Clinical Implications Received in original form August 9, 2002; accepted in final form December 11, 2002
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
