Response to ‘Flow versus pressure?’

We are tremendously grateful for the positive comments shared by Dr Furst (Furst, 2020a), and regret that we were not aware of his book (Furst, 2020b) sooner, in order to give his ideas due consideration in our original article (Joyce and Wang, 2020).

We echo Dr Furst's argument that the regulation of blood flow takes precedence over blood pressure. This was indeed eloquently expressed almost a century ago, when the Austrian physiologist Adolf Jarisch Jr stated ‘for the development of the doctrine of the circulation, it was undoubtedly fatal that the measurement of blood flow was comparatively laborious, but that blood pressure could be determined so easily. That is why the sphygmomanometer gained such a fascinating influence, although most organs do not need pressure, but flow’ (Jarisch, 1928). It is unfortunate that this has not become more widely appreciated.

Central venous pressure (CVP) represents a case in point, and has, as explained by Furst (2020a), been challenged as a reliable indicator for haemodynamic status in the clinical context (Marik et al., 2008). In a classic Starling curve, cardiac output (CO) is expected to increase when CVP rises. However, this only holds true under defined conditions (Berlin and Bakker, 2015) and critically depends on what is determining the change in central venous pressure. For example, CVP will decrease in conditions where increased cardiac contractility is increased (Joyce and Wang, 2020).

In situ perfused slider turtle (Trachemys scripta) hearts provide a curious example. Farrell et al. (1994) clearly demonstrated the archetypal Starling response when filling pressure was increased by increasing the height of the column filling the heart (Fig. 1A). However, turtle atria contain smooth muscle (an apparently unique trait amongst vertebrates; Joyce et al., 2020) that, when constricted, reduces CO. Using a similar perfused heart preparation, we could stimulate this atrial smooth muscle to constrict by adding histamine, whilst the heart filled from a constant pressure head (Joyce et al., 2019). Under this condition, CVP rose whilst CO fell. Conversely, inhibiting smooth muscle contraction with wortmannin caused CO to increase and cardiac filling pressure to decrease (Fig. 1B). This shows that increased CVP does not necessarily augment CO, as a simplified Starling curve would imply. CVP is only a suitable indicator for cardiac preload under certain circumstances when it acts as a surrogate for end-diastolic volume (Berlin and Bakker, 2015). This is, of course, because the force of cardiac contraction is not determined by the filling pressure per se, but rather by the stretch of the myofilaments (i.e. the ability to form cross-bridges between actin and myosin), which is determined by the volume of blood in the ventricle, not its pressure.

CVP is not only determined by venous return and cardiac function but also affected by the extra-vascular pressure adjacent to the heart, i.e. the ‘juxta-cardiac pressure’ (Berlin and Bakker, 2015). In experimental preparations, i.e. with open thorax and pericardium, this is insignificant, but in the intact animal it is liable to change. For example, in turtles, periods of ventilation are accompanied by a large fall in CVP (Joyce et al., 2018). This decrease in CVP coincides with increased CO, and can be attributed to decreases in visceral and intrapericardial pressures (as a result of the actions of the ventilatory muscles). Here, a decrease in CVP does not represent decreased cardiac filling; rather, it may promote it.

We hasten to add that we do not negate the utility of measuring CVP to understand cardiovascular physiology. Rather, the opposite; we believe it should earn greater prominence, but only when considered in its proper context, and when measured in parallel with other haemodynamic measurements.