The blood vessels that run on the surface of the heart and through its muscle are compliant tubes that can be affected by the pressures external to them in at least two ways. If the pressure outside these vessels is higher than the pressure at their downstream ends, the vessels may collapse and become Starling resistors or vascular waterfalls. If this happens, the flow through these vessels depends on their resistance and the pressure drop from their inflow to the pressure around them and is independent of the actual downstream pressure. In the first part of this review, the physics of collapsible tubes is described, and the possible occurrences of vascular waterfalls in the body is evaluated. There is good evidence that waterfall behavior is seen in collateral coronary arteries and in extramural coronary veins, but the evidence that intramural coronary vessels act like vascular waterfalls is inconclusive. There is no doubt that in systole there are high tissue pressures around the intramyocardial vessels, particularly in the subendocardial muscle of the left ventricle. The exact nature and values of the forces that act at the surface of the small intramural vessels, however, are still not known. We are not certain whether radial (compressive) or circumferential and longitudinal (tensile) stresses are the major causes of vascular compression; the role of colagen struts in modifying the reaction of vessel walls to external pressures is unknown but possibly important; direct examination of small subepicardial vessels has failed to show vascular collapse. One of the arguments in favor of intramyocardial vascular waterfalls has been that during a long diastole the flow in the left coronary artery decreases and reaches zero when coronary arterial pressure is still high: it can be as much as 50 mmHg in the autoregulating left coronary arterial bed and ~ 15-20 mmHg even when the vessels have been maximally dilated. These high zero flow pressures, especially during maximal vasodilatation, have been regarded as indicating a high back pressure to flow that is due to waterfall behavior of vessels that are exposed to tissue pressures. Because direct evidence for waterfall behavior is lacking, alternative explanations for the high zero flow pressures have been sought. One major theory notes that the large intramyocardial blood volume with its long time constants is in series with the larger coronary arteries with their short time constants. When the coronary arteries have discharged their blood so that flow is zero at the origin of the larger coronary arteries, there is still flow downstream in the microvessels; zero flow in the arteries occurs at a pressure that is the input pressure to the microvessels and therefore much higher than downstream venous or right atrial pressures. The compliance and blood volume of the intramyocardial vessels also play an important role in phasic myocardial blood flow. When the left ventricle contract the high intramyocardial pressures around the subendocardial vessels squeeze blood forward from the capillaries and intramyocardial venules into the extramural coronary veins and also causes retrograde flow into the extramural coronary arteries. This retrograde flow is usually concealed by the compliance of the extramural coronary arteries. The compliance of these extramural arteries is high enough to store most of the blood that enters these arteries in systole from the aorta so that it is not possible to assess phasic myocardial perfusion by inspecting the phasic flow pattern in the proximal large coronary arteries. Although coronary blood flow depends on these physical factors, its quantity is usually regulated by nonphysical factors, mainly biochemical and neurohumoral. If coronary perfusion pressure is altered while myocardial oxygen consumption remains constant, coronary flow initially changes in the direction of the pressure change but then tends to return to near its initial value; the phenomenon is termed autoregulation. It is likely that the mechanisms responsible for this feedback are related directly or indirectly to oxygen supply to the heart muscle, although the precise chain of events has not yet been elucidated. The responsible mechanisms may well be the same as those that are involved in adjusting the level of coronary flow to the metabolic needs of the heart. In fact, coronary flow can be related very closely to coronary perfusing pressure and myocardial oxygen consumption.
