Vascular Waterfall Research Articles

Effects of adenosine on pressure-flow relationships in an in vitro model of compartment syndrome.

Blood flow through skeletal muscle is best modeled with a vascular waterfall at the arteriolar level. Under these conditions, flow is determined by the difference between perfusion pressure (Pper) and the waterfall pressure (Pcrit), divided by the arterial resistance (Ra). By pump perfusing an isolated canine gastrocnemius muscle (n = 6) after it was placed within an airtight box, with and without adenosine infusion, we observed an interaction between the pressure surrounding a muscle (as occurs in compartment syndrome) and baseline vascular tone. We titrated adenosine concentration to double baseline flow. We measured Pcrit and Ra at box pressures (Pbox), which resulted in 100 (Pbox = 0), 90, 75, and 50% flow without adenosine; and 200, 180, 150, 100, and 50% flow with adenosine. Without adenosine, each 10% decline in flow was associated with a 5.7 mmHg increase in Pcrit (P < 0.01). With adenosine, the same decrease in flow was associated with a 2.6-mmHg increase in Pcrit (P < 0.01). Values of Pcrit at 50% of flow were almost identical. Each 10% decrease in flow was also associated with 2.2% increase in Ra with or without adenosine (P < 0.001). Ra decreased with adenosine infusion (P < 0.05), and there was no interaction between adenosine and flow (P > 0.9). We conclude that increases in pressure surrounding a muscle limit flow primarily through changes in Pcrit with and without adenosine-induced vasodilation. The interaction between Pbox and adenosine with respect to Pcrit but not Ra suggests that Pbox affects the tone of the vessels responsible for Pcrit but not Ra.

The Pharyngeal Critical Pressure: The Whys and Hows of Using Nasal Continuous Positive Airway Pressure Diagnostically

The Pharyngeal Critical Pressure: The Whys and Hows of Using Nasal Continuous Positive Airway Pressure Diagnostically

A nonlinear model for myogenic regulation of blood flow to bone: equilibrium states and stability characteristics.

A simple compartmental model for myogenic regulation of interstitial pressure in bone is developed, and the interaction between changes in interstitial pressure and changes in arterial and venous resistance is studied. The arterial resistance is modeled by a myogenic model that depends on transmural pressure, and the venous resistance is modeled by using a vascular waterfall. Two series capacitances model blood storage in the vascular system and interstitial fluid storage in the extravascular space. The static results mimic the observed effect that vasodilators work less well in bone than do vasoconstrictors. The static results also show that the model gives constant flow rates over a limited range of arterial pressure. The dynamic model shows unstable behavior at small values of bony capacitance and at high enough myogenic gain. At low myogenic gain, only a single equilibrium state is present, but a high enough myogenic gain, two new equilibrium states appear. At additional increases in gain, one of the two new states merges with and then separates from the original state, and the original state becomes a saddle point. The appearance of the new states and the transition of the original state to a saddle point do not depend on the bony capacitance, and these results are relevant to general fluid compartments. Numerical integration of the rate equations confirms the stability calculations and shows limit cycling behavior in several situations. The relevance of this model to circulation in bone and to other compartments is discussed.

Maximal vasodilation does not eliminate the vascular waterfall in the canine hindlimb.

Previous studies have shown that blood flow through skeletal muscle is regulated by changes in an arteriolar vascular waterfall [critical pressure (Pcrit)] and a proximal (arterial) resistance (Ra) element. To determine whether Pcrit still exists during maximal vasodilation, we pump perfused vascularly isolated canine hindlimbs. We set outflow pressure to zero and measured Pcrit, perfusion pressure (Pper), and regional elastic recoil pressure (Pcl; by a stop-flow technique) and calculated both Ra and venous resistance before and after maximal vasodilation with adenosine and nitroprusside. Pcrit was 56.4 +/- 5.1 mmHg before vasodilation and decreased to 11.0 +/- 0.6 mmHg after vasodilation, which was less than the downstream pressure in the venous compliant region (Pel). Therefore, Pcrit should not have affected flow at normal Pper levels under vasodilated conditions. However, we could still measure Pcrit because our technique allowed Pel to decline and Pcrit becomes apparent once Pel < Pcrit. With vasodilation, Ra decreased to < 8.1 +/- 2.6% and Rv decreased to 41 +/- 6% of control values. In contrast to the nonvasodilated vasculature, increases in venous pressure during maximal vasodilation caused immediate increases in Pper. This also suggests that the vascular waterfall is inactive under conditions of maximal vasodilation. We conclude that a small arteriolar Pcrit is still present in the maximally vasodilated hindlimb but is less than the downstream pressure and does not affect flow under these conditions.

Distinct behavior of portal venous and arterial vascular waterfalls in porcine liver

Purpose: Hepatic dysfunction is associated with morbidity and mortality in critically ill patients. Understanding liver hemodynamics in pathological states requires characterization of the normal portal venous and hepatic arterial circulations. Using pressure flow analysis, we tested the hypothesis that vascular waterfalls determine blood flows in the normal liver. Methods: In 14 vascularly isolated porcine livers, steady-state pressure-flow relationships, which defined a slope (incremental resistance) and a zero flow pressure intercept (Po), were generated for each vessel over a range of hepatic venous pressures (Phv). Results: Critical closing pressures occurred in the portal venous circulation (Po = 3.8 ± 0.4 mm Hg) with classical waterfall physiology observed as Phv was raised. The hepatic arterial critical closing pressure (Po = 8.3 ± 1 mm Hg) showed a constant positive pressure difference of 5 mm Hg versus Phv as the latter was increased from 0 to 28 mm Hg ( P < .05). Portal venous resistance decreased when Phv was greater than Po ( P < .05), but no effect on hepatic arterial resistance was seen as Phv was increased. Conclusion: Both critical closing pressures and incremental resistances showed markedly different responses to increased outflow pressures in the portal venous and hepatic arterial circulations. The results provide the physiological basis to analyze hemodynamic changes in the liver under normal and pathological conditions.

Regional control of venous return: liver blood flow.

The aim of the study was to determine whether closing pressures or vascular distensibility can be used to describe liver venous hemodynamics when right atrial pressure is raised. The study was performed using a vascularly isolated pig liver preparation that allowed the independent control of portal vein and hepatic artery inflows and of outflow pressure (Pout). Pressure-flow (P-Q) relationships of both liver vessels were generated at multiple levels of Pout. At Pout of 0 mm Hg, the portal vein P-Q relationship was linear, with a convexity toward the pressure axis at low flows (5 to 10 ml/min/kg). The zero flow pressure was 1.5 +/- 0.2 mm Hg, greater than Pout (p < 0.05). On raising Pout from 0 to 15 mm Hg, the shape of the portal vein P-Q relationships became progressively more linear, with a decrease in slope; no difference between zero flow pressure and Pout was observed. At Pout of 0 mm Hg, the hepatic artery presented a zero flow pressure > Pout. Raising Pout from 0 to 15 and 30 mm Hg resulted in a zero flow pressure always > Pout (p < 0.05). The behavior of the liver vein system is characterized by a zero flow pressure mimicking a classic vascular waterfall and by distensibility, once the waterfall is exceeded. Both factors act to minimize the reduction in venous return with an increased central venous pressure. Flow through the hepatic artery is affected by an increase in backpressure occurring upstream from the sinusoids, reducing arterial inflow for a constant perfusion pressure.

Pressure-flow relationships in in vitro model of compartment syndrome.

Compartment syndrome is a condition in which an increase in intramuscular pressure decreases blood flow to skeletal muscle. According to the Starling resistor (i.e., vascular waterfall) model of blood flow, the decrease in flow could occur through an increase in arterial resistance (Rart) or an increase in the critical closing pressure (Pcrit). To determine which explains the decrease in flow, we pump perfused a canine gastrocnemius muscle placed within an airtight box, controlled box pressures (Pbox) so that flow ranged from 100 to 50%, and measured Pcrit, Rart, arterial compliance, small venular pressure (measured by the double-occlusion technique), and venous pressure. An increase in Pbox limited flow mainly through an increase in Pcrit (75-85%), with only small changes in Rart (15-25%) and no change in arterial compliance. Increases in Pbox also produced a vascular waterfall in the venous circulation, but small venular transmural pressure always remained less than control levels. We conclude that increases in Pbox mostly limit blood flow through increases in Pcrit and that Rart plays a minor role. Transmural pressure across the small venules decreases with increases in intramuscular pressure, which contradicts the currently held belief that compartment syndrome is due to a cycle of swelling-ischemia-swelling.

Effects of nifedipine on vascular waterfall and arterial resistance in canine hindlimb.

Pressure-flow relations in the canine hindlimb can be well explained by a vascular waterfall at the arteriolar level. Under these conditions, P(art) = Pcrit + Q.Rart, where P(art) is the arterial pressure, Pcrit is the waterfall pressure, Q is regional flow, and Rart is the arterial resistance of the vessels upstream from the waterfall. To determine whether calcium channels in vascular smooth muscle affect Pcrit in addition to Rart, we pump perfused canine hindlimbs and measured both variables over a range of perfusion pressures (Pper) before and during the infusion of the calcium channel blocker nifedipine. Nifedipine significantly decreased Pcrit and Rart at each Pper. Increasing Pper under control conditions from 50 to 150 mmHg significantly increased Pcrit from 24.2 +/- 1.5 to 42.5 +/- 2.2 mmHg. During nifedipine infusion, increasing Pper from 25 to 100 mmHg also increased Pcrit from 14.5 +/- 1.5 to 20.2 +/- 1.9 mmHg, but the rate of increase was less. In contrast to the rise in Pcrit with increasing Pper, Rart significantly decreased from 10.7 +/- 1.1 to 8.1 +/- 1.2 mmHg.min.100 g.ml-1 before nifedipine infusion, and from 5.7 +/- 0.4 to 2.2 +/- 0.1 mmHg.min.100 g.ml-1 during nifedipine infusion. Venous resistance (Rven) significantly decreased with increases in Pper and during nifedipine infusion. The regional elastic recoil pressure (Pel, a measure of small venular pressure) increased with both an increase in Pper and nifedipine. These results suggest that nifedipine decreases Pcrit, Rart, and Rven and that at constant Pper nifedipine increases Pel.

Effects of increased and decreased tissue pressure on haemodynamic and capillary events in cat skeletal muscle.

1. The controversial problem concerning the unusual haemodynamics of the deranged circulation during increased hydrostatic tissue pressure (PT) was elucidated by detailed studies of arterial, capillary and venous functions in cat skeletal muscle exposed to graded experimental changes of PT over a wide range. 2. The results indicated that the impaired circulatory state in skeletal muscle during raised tissue pressure is characterized by the following train of events: (a) a primary partial passive compression of the most distal part of the venous system due to negative vascular transmural pressure selectively at this site, in turn leading to the prompt development of a distinct 'venous outflow orifice resistance' graded in relation to the PT rise; (b) a consequent reduction of blood flow graded in relation to this resistance increase; (c) a rise in intramuscular venous pressure proximal of the 'venous outflow orifice' by the same extent as the PT increase; (d) transmission of the raised venous pressure to more proximal vessels in relation to the prevailing segmental resistance ratios; (e) a consequent maintenance of clearly positive transmural pressures in all vascular sections proximal to the 'venous outflow orifice', preventing collapse of these vessels; (f) maintenance of a largely normal capillary filtration coefficient and functional capillary surface area; and (g) an increase in capillary pressure by approximately 85% of the PT rise which reduces the rate of net transcapillary fluid absorption to about one-seventh of that expected from the PT rise per se. 3. Previous concepts of a 'vascular waterfall phenomenon', a capillary collapse, or an arteriolar 'critical closure phenomenon' did not seem to be valid for the skeletal muscle circulation during increased PT. 4. The rate of net transcapillary fluid flux per unit PT change was much smaller during positive than negative PT, since capillary pressure rose considerably when PT was increased above control, but was largely unchanged when PT was decreased below control. 5. Possible ways to improve the circulatory state in conditions with an oedema-induced tissue pressure rise are discussed.

Effects of arterial and venous pressure alterations on transcapillary fluid exchange during raised tissue pressure.

To assess local haemodynamic effects of raised tissue pressure per se and of arterial and venous pressure variations during raised tissue pressure, and to evaluate the vascular waterfall phenomenon. An isolated pump-perfused sympathectomised cat skeletal muscle enclosed in a plethysmograph. Hydrostatic capillary pressure (Pc), tissue volume alterations and blood flow were recorded at various arterial (PA) or venous (PV) pressure levels at a raised tissue pressure (Ptissue). Total vascular resistance (Rtot) and its three consecutive sections, arterial resistance (Rart), venular resistance (Rvenule), and venous outflow orifice resistance (Rorifice), were recorded. Increase in Ptissue increased Pc due to a marked increase in Rorifice and unchanged Rart and, when Ptissue > PV, by about 90% of the Ptissue increase. During the raised Ptissue, a PA increase (95 to 115 mmHg) increased Pc (by 3.1 +/- 1.1 mmHg) causing fluid filtration, and a PA decrease (95 to 75 mmHg) decreased Pc (by 2.4 +/- 0.5 mmHg) causing fluid absorption. Rart was unchanged, indicating impaired autoregulation. Increased PV had no haemodynamic effects when PV < Ptissue due to a gradual decrease in Rorifice towards zero when PV reached Ptissue. At PV > Ptissue blood flow and Pc increased gradually, causing fluid filtration. Tissue volume is increased by raised and decreased by lowered PA, the latter may be of use to decrease ICP in the injured brain. The results indicate that PEEP or head elevation will not influence ICP from the venous side if CVP < ICP. Finally, the "vascular waterfall phenomenon" was rejected as Rorifice is a normal variable fluid resistance.

Effects of positive end-expiratory pressure and body position on pressure in the thoracic great veins.

Positive end-expiratory pressure (PEEP) commonly decreases cardiac output. The major cause of this is believed to be decreased venous return due to increased right atrial pressure. We hypothesized that when the lungs were hyperinflated they could also restrict venous return by directly compressing the thoracic vena cavae. We measured the longitudinal distribution of pressure in the thoracic vena cavae of 10 dogs on and off 10 mm Hg PEEP, in the supine (S), prone (P), right lateral (RL), and left lateral decubitus (LL) positions. In the superior vena cava (SVC) both on and off PEEP, and in the inferior vena cava (IVC) off PEEP, pressure fell uniformly from the thoracic inlet to the right atrium. However, in the IVC on PEEP, intravascular pressure fell abruptly by up to 5 mm Hg. This pressure drop occurred in a discrete (1 to 2-cm) segment of the IVC, suggesting a localized increased in extravascular surface pressure. When this pressure inflection was present, changes in right atrial pressure had no effect on pressure in the IVC upstream of the inflection, consistent with a "vascular waterfall." These observations were most prominent in the LL, least common in the RL, and variably present in the P and S positions. Occlusion of the right bronchus intermedius prior to PEEP (preventing right lower, middle, and accessory lobe inflation) prevented the appearance of the pressure inflection during PEEP in the LL but not in the S or P positions. We conclude that PEEP impedes venous return partly by direct compression of the IVC, predominantly in positions in which the IVC is non-dependent.(ABSTRACT TRUNCATED AT 250 WORDS)

A model of the instantaneous pressure-velocity relationships of the neonatal cerebral circulation

The instantaneous relationship between arterial blood pressure (BP) and cerebral blood flow velocity (CBFV), measured with Doppler ultrasound in the anterior cerebral artery, is represented by a vascular waterfall model comprising vascular resistance, compliance, and critical closing pressure. One min recordings obtained from 61 low birth weight newborns were fitted to the model using a least-squares procedure with correction for the time delay between the BP and CBFV signals. A sensitivity analysis was performed to study the effects of low-pass filtering (LPF), cutoff frequency, and noise on the estimated parameters of the model. Results indicate excellent fitting of the model (F-test, p < 0.0001) when the BP and CBFV signals are LPF at 7.5 Hz. Reconstructed CBFV waveforms using the BP signal and the model parameters have a mean correlation coefficient of 0.94 with the measured flow velocity tracing (N = 232 epochs). The model developed can be useful for interpreting clinical findings and as a framework for research into cerebral autoregulation.

A model of phasic arterial and venous blood flow in the coronary circulation

A numerical model of the coronary circulation was developed which simultaneously addresses both the arterial and venous sides of the circulation. The model is based upon the concept of an intramyocardial pump rather than the vascular waterfall mechanism and utilizes only phasic pressures as input. The left ventricular myocardium was modelled as three equal and parallel compartments, each consisting of fixed arterial and venous resistances on either side of a lumped arterial and venous intramyocardial capacitance. A time-varying intramyocardial pressure, P i ( t), different for each compartment, was applied through the capacitance. Input pressures were obtained from direct measurements of phasic aortic, right atrial and intramyocardial pressures in six open-chest anesthetized dogs. The model accurately predicted the phasic waveforms of coronary arterial inflow and venous outflow and the nearly 180° phase shift between coronary inflow and outflow. The predicted magnitudes of mean systolic coronary arterial inflow (4.7 ± 1.0 ml/min), mean diastolic coronary inflow (38.4 ± 0.5 ml/min) and mean total coronary arterial inflow (27.7 ± 1.0 ml/min) in the left anterior descending coronary artery were not significantly different from those measured in dogs. The measured values were 5.2 ± 1.0, 28.1 ± 4.2, and 22.2 ± 3.0 ml/min, respectively. The model also predicted a transmural gradient of coronary arterial blood flow which favors the subepicardial layers during systole and the subendocardial layers during diastole. The latter is consistent with experimental observations in the beating canine left ventricular myocardium.

Superior and inferior vena caval flows during respiration: pathogenesis of Kussmaul's sign.

Respiratory induced changes in superior (QSVC) and inferior (QIVC) vena caval flows and abdominal pressures were evaluated in anesthetized closed-chest dogs. QSVC and QIVC were measured with ultrasound transit time flow probes (n = 5), and general (Pab) and regional subdiaphragmatic (Pd) abdominal pressures were measured by air-filled balloons (n = 5), during two respiratory maneuvers produced by phrenic nerve stimulation, i.e., simulated spontaneous inspiration (SSI), and Mueller maneuver (MM), with the airway occluded to minimize diaphragmatic descent. With hypervolemia: during SSI, QSVC decreased, QIVC increased, and right atrial pressure increased (P less than 0.05) despite a decrease in esophageal pressure (Pes), i.e., Kussmaul's sign; during MM, both QSVC and QIVC increased (P less than 0.05) without Kussmaul's sign. The ratios delta Pab/delta Pes and delta Pd/Pes were larger during SSI than MM (P less than 0.01). With hypovolemia: during SSI and MM, QSVC increased and QIVC decreased with a venous pressure gradient across the diaphragm (P less than 0.05), consistent with development of a vascular waterfall. These results suggest that 1) changes in abdominal pressure may affect the patterns of QSVC and QIVC during respiration, depending on blood volume status; 2) a prolonged inspiration with hypovolemia may decrease QIVC because of the development of a vascular waterfall; 3) QSVC and QIVC may be interdependent during respiration; and 4) the essential mechanism of Kussmaul's sign is a substantially larger inspiratory increase in abdominal pressures produced by diaphragmatic descent compared with the decrease in intrathoracic pressure under hypervolemic conditions, rather than the presence of pericardial pathology or right heart dysfunction.

Hemodynamics of vascular ‘waterfall’: is the analogy justified?

The concept of ‘vascular waterfall’ has been used for collapsible vessels in different hemodynamic states which have little similarity to each other from a dynamic standpoint. Examples include (a) flow through large systemic veins entering the thorax, (b) flow through microvessels, such as pulmonary, cardiac, hepatic, cerebral, and (c) flow through the jugular vein of the giraffe. The dynamics of freely falling liquids (waterfall) as compared with the flow through collapsible blood vessels (in vivo and in vitro0 and in collapsible tubes are dissimilar in too many respects to justify analogy. The flow through collapsible tubes and blood vessels can be explained satisfactorily on the basis of elementary principles of fluid mechanics (Bernoulli-Poiseuille). Hence, the term waterfall as a metaphor is misleading and unjustified. We suggest that the use of the term be discontinued for describing vascular dynamics.

Effects of inspiratory diaphragmatic descent on inferior vena caval venous return.

To explain the contradictory results in the literature regarding the effects of inspiratory diaphragmatic descent on inferior vena caval (IVC) venous return, we evaluated changes in total IVC flow as well as regional splanchnic and nonsplanchnic IVC flows by use of ultrasound flow probes placed around the thoracic and subhepatic abdominal IVC during phrenic nerve stimulation (PNS) in anesthetized open-chest dogs. With the abdomen closed (n = 6), PNS under hypervolemic conditions increased the total IVC flow by enhancing the splanchnic IVC flow, with a transient decrease in the nonsplanchnic IVC flow (P less than 0.05). Under hypovolemic conditions, PNS initially increased the total IVC flow but later decreased the total IVC flow by reducing the nonsplanchnic IVC flow, associated with a venous pressure gradient in the IVC across the diaphragm (P less than 0.05), consistent with development of a vascular waterfall. With the abdomen widely open, the mobile abdominal contents eviscerated, and the subhepatic IVC occluded (n = 5), PNS increased the splanchnic IVC flow associated only with an increase in focal contact pressure over the liver without any increase in general abdominal pressure (Pab) (P less than 0.05). These results suggest that our previously proposed concept of abdominal vascular zone conditions (J. Appl. Physiol. 69: 1961-1972, 1990) is useful as a global approximation to understand the effects of respiratory-induced changes in Pab's on the total and regional IVC venous return. Nonhomogeneous distribution of Pab's during diaphragmatic descent may need to be considered to explain all aspects of the behavior of the intact IVC system.

Coronary Sinus Pressure: A Condition per Se of Return Volume via the Coronary Sinus in the Healthy Animal? Surgical Considerations in Atriopulmonary Anastomosis Procedures

In order to test the hypothesis of the "vascular waterfall" mechanism in coronary venous circulation, the role of coronary sinus pressure in the regulation of coronary return volume (CRV) via the coronary sinus (CS) was studied in healthy animals. An experimental model of pressure regulation in the CS was prepared and aortic and left atrial pressures, EKG, and the cardiac output (measured by thermodilution) were recorded. The return volume via the CS was measured at CS pressures of 10 or less, 15, 20, and 25 mm Hg or more, for a total of 36 determinations. Overall, increased CS pressure did not produce significant changes in aortic and left atrial pressures, heart rate, cardiac index, or CRV via the CS. When CS pressure was 25 mm Hg or more, however, there was a significant decline in the average CRV via the CS. Nevertheless, stepwise variant regression showed that the CS per se does not condition the volume of coronary return via the CS. The authors' results suggest that in healthy animals, the hypothesis of the "vascular waterfall" mechanism in coronary venous circulation is not valid. Thus, in correcting congenital cardiac malformations by use of atriopulmonary anastomosis procedures, there is no justification for employing techniques that ensure CS drainage into the left atrium in order to avoid the hemodynamic repercussions attributable to the hypothesis of the "vascular waterfall."

Endotoxin and the mechanical properties of the canine peripheral circulation

To determine if alterations in venous vascular mechanics by endotoxin contribute to the hemodynamic disorder of septic shock, we studied eight dogs before and after the injection of 5 to 10 mg/kg of Escherichia coii endotoxin. We isolated the venous drainage of the peripheral (ie, superior plus inferior vena cava below the renal veins) and splanchnic vasculatures, and measured venous pressure (Pv) and flow (electromagnetic flow meter) in each region. Blood drained through vascular waterfalls that controlled the Pv of each region and then into a reservoir. A constant flow pump returned blood to the right atrium. By temporarily stopping all flows, we measured the elastic recoil pressures of each region. We changed regional volumes by changing Pv and integrating the transient changes of outflow. This allowed the calculation of compliance (Cv in mL/mm Hg/kg), venous resistance (Rv), venous time constant (τv in seconds), and fractional flow of each region. Endotoxin produced a large loss of reservoir volume but did not change the fractional flows. Peripheral Cv increased and Rv did not change, so that peripheral τv was 3.8 ± 1.0 before and 5.7 ± 2.7 (mean ± SD; P < .024) after endotoxin. The splanchnic compliance tended to decrease (0.776 ± 0.801 before and 0.487 ± 0.418 after; P = NS) and Rv did not change, so that splanchnic τv decreased from 14.6 ± 5.6 before to 10.8 ± 5.1 after endotoxin ( P < .024). In the intact animal, without a reservoir, the decrease in stressed volume would decrease cardiac output and the decrease in the splanchnic τv would increase cardiac output.

Venous waterfalls in coronary circulation

Several studies of flow through collapsible tubing deformed by external pressures have led to a concept known as the "vascular waterfall". One hallmark of this state is a positive zero-flow pressure intercept (Pe) in flow-pressure curves. This intercept is commonly observed in the coronary circulation, but in blood-perfused beating hearts a vascular waterfall is not the only putative cause. To restrict the possibilities, we have measured flow-pressure curves in excised non-beating rabbit hearts in which the coronary arteries were perfused in a non-pulsatile way with a newtonian fluid (Ringers solution) containing potent vasodilator drugs. Under these circumstances, vascular waterfalls are believed to be the only tenable explanation for Pe. In physical terms the waterfall is a region where the vessel is in a state of partial collapse with a stabilized intraluminal fluid pressure (Pw). It is argued that the most probable site of this collapse was the intramural veins just before they reached the epicardial surface. In accord with the waterfall hypothesis, Pe increased as the heart became more edematous, but flow-pressure curves also became flatter, implying multiple waterfalls with differing Pws, leading to complete collapse of some of the venous channels. The principal compressive force is believed to have been the interstitial fluid pressure as registered through a needle (Pn) implanted in the left ventricular wall, but a small additional force (Ps) was probably due to swelling of interstitial gels. A method is presented for estimating Ps and Pw. Unlike rubber tubing, blood vessels are both collapsible and porous. Apparently because of increased capillary filtration, Pn was found to increase linearly with the perfusion pressure. Thus, Pw was not the same at all points on the flow-pressure curve. This finding has interesting implications with respect to the concept of coronary resistance.

Tone-dependent waterfall behavior during venous pressure elevation in isolated canine hearts.

We examined the "vascular waterfall" hypothesis, which proposes that coronary flow is unaffected by elevations in outflow pressure until the latter reaches a critical threshold level, in 29 isolated canine hearts. In fibrillating hearts vasodilated with adenosine or carbocromen, coronary flow and the coronary pressure-flow relation were not affected by changes in great cardiac vein pressure (PGCV) below a threshold value of 11 +/- 0.9 (mean +/- SEM) mm Hg. Further elevations of PGCV reduced flow and shifted the pressure-flow relation to the right, increasing its pressure-axis intercept (Pf=0). When vasomotor tone was augmented with vasopressin, threshold PGCV increased to 25 +/- 2.7 mm Hg (p less than 0.001). Once again, the pressure-flow relation was unaffected by changes in PGCV below the threshold value and shifted to the right when this value was exceeded. The amount by which spontaneous values of Pf=0 exceeded threshold values of PGCV was greater when vasomotor tone was augmented than during vasodilation. Pf=0 continued to exceed PGCV when the latter was raised above the threshold level. Both Pf=0 and threshold values of PGCV were less during a long diastole than during ventricular fibrillation. We reached the following conclusions. 1) During changes in PGCV below a threshold value, the coronary circulation exhibits traditional waterfall behavior. 2) The threshold pressure for altering waterfall behavior is affected by vascular tone and mechanical activity. 3) Pf=0 remains above PGCV when the latter is increased above the threshold value needed to alter flow.