Esmolol in Sepsis: Microcirculatory Savior or Autoregulation Annihilator?
If venous pressure were always the downstream pressure of an artery, then when flow is zero, the arterial pressure should equal the venous pressure. For example, if one were to measure the coronary artery pressure and coronary sinus pressure at zero flow, their pressures should be equal. Yet, this has been done; when flow is zero in the heart’s vascular tree, an instantaneous measurement of arterial and venous pressure reveals a large gradient. How can this be so? The reason is likely pre-capillary arterial tone, which prevents flow below a certain threshold pressure in the artery. This threshold pressure is called the ‘critical closing pressure’ [Pcc] and creates ‘vascular waterfall’ [VW]. VW simply means that the downstream [venous] pressure no longer regulates flow, just as the height of a waterfall does not affect the volume of water per unit time rolling over its crest.
Maintaining a Pcc in low flow states may be beneficial to maintain a pressure gradient across capillary beds, for example, during CPR. With this background and previous data demonstrating mortality benefit with this beta-blocker, a very recent article examined the effects of esmolol in a canine model of sepsis.
What They Did
In a canine model, endotoxemia was introduced using an E. coli infusion. This resulted in septic shock and all 16 dogs were given norepinephrine and fluids to maintain hemodynamics. To measure the arterial critical closing pressure [Pcc], a series of inspiratory holds were performed at progressively higher pressure. Following 8 seconds of an inspiratory hold, renal artery pressure and flow were plotted. This line was then extrapolated to zero flow to estimate the Pcc. They performed a similar analysis for the renal vein to extrapolate its zero flow intercept and compared this with the Pcc. If the Pcc was greater than the renal vein zero flow pressure, then VW was in effect.
What They Found
As suspected, the introduction of septic shock greatly reduced the Pcc, in effect abolishing VW. The Pcc fell from ~ 65 mmHg to – 14 mmHg [yes, a negative Pcc]. Both fluids and norepinephrine increased the Pcc, but fluids to a greater extent. Then, with the infusion of esmolol, Pcc was increased to 44 mmHg as compared to control which remained at – 14 mmHg [yes, negative]. Thus, they argue that esmolol increases the Pcc, restores vascular waterfall, which is beneficial for the microcirculation.
I found the discussion of this paper to be somewhat chaotic and there is no clear discussion of mechanism; further, critical limitations, I think, were not considered. The most notable limitations from this study arise from 1. Their assumptions about pressure-flow relationship and 2. The tissue beds that they studied.
Firstly, they assumed that arterial pressure-flow are functionally independent and they made no comment on the importance of auto-regulation. Myogenic auto-regulatory responses occur very rapidly, in less than 3-4 seconds and their measurements occurred at the end of 8 second inspiratory holds. Thus, as the inspiratory hold progressed and renal pressure fell, there would have been a strong impetus for arteriolar dilation within the kidney. Their protocol, therefore, may be less of a test of Pcc and more a test of auto-regulation reserve. For example, if flow is maintained by vascular dilation as pressure falls, then the slope of the pressure-flow line will become less steep and the intercept [i.e. the Pcc] will fall. Ideally, the pressure-flow relationships would have been done very rapidly, which they weren’t [see figure 1 A & B].
Secondly, they chose the kidney for their tissue of interest. The kidney has very complicated hemodynamics. For example, the renal medulla is less capable at auto-regulation than the cortex; that is, the medulla pressure-flow relationship is more linear. Considering the medulla and cortex together would tend to steepen the slope of the renal pressure flow relationship and raise Pcc. Importantly, during sepsis, flow moves away from the medulla, to the cortex, and this physiology is amplified by norepinephrine. With a greater fraction of blood moving through vascular beds that have more auto-regulatory reserve [i.e. the cortex], the slope of the pressure-flow diagram is expected to fall and lower the Pcc – especially if measurements are taken beyond 4 seconds – as delayed measurement allows auto-regulation to take hold.
Figure 1A: The renal auto-regulation curve is presented for reference. Flow is maintained by renal vascular diameter change as indicated. Figure 1B: The pressure flow extrapolations. Imagine two experiments where pressure is reduced from P1 to P2 [on x-axis]. In one experiment, pressure and flow are measured instantly [light blue] with extrapolation to the x-intercept [Pcc in blue]. In a second experiment, pressure is dropped from P1 to P2, but a long measurement time is used such that auto-regulation takes effect and flow is higher at P2 because of arteriolar dilation. The slope of this line [red-dashed] is less steep with a lower [negative] Pcc. Finally, and hypothetically, if esmolol caused a drop in pressure during the inspiratory holds, the pressure-flow relationship may have fallen down the leftmost arm of the auto-regulation curve. This would result in a steeper slope [purple dashed] and a higher Pcc.With respect to the mechanism by which esmolol raises the Pcc, there is little discussion and not enough data is presented to determine if auto-regulation is the actual underlying culprit. We are given baseline heart rate, filling pressures and cardiac output, but not the instantaneous values during the inspiratory holds. If, for example, during inspiratory hold whilst on esmolol the renal perfusion pressure fell below the lower threshold of auto-regulation, then the pressure-flow relationship becomes steep and linear again. Additionally, if esmolol antagonized some of the effect of norepinephrine and diverted blood from the renal cortex to medulla, the pressure-flow relationship would become more linear secondary to the inherent characteristics of the medulla. Both effects would raise the extrapolated Pcc as compared to pressure-flow points made within the range of normal renal auto-regulation [see figure 1B, purple line].
After 1 hour, there was no statistically significant change in the venous-arterial carbon dioxide gradient between control animals and those on esmolol which points away from improved tissue perfusion. However, at hour 2, there was a gradient observed. It is not indicated from where the blood gas samples were drawn. If venous carbon dioxide was drawn from the right atrium, the reduction in heart rate from esmolol and, therefore, prolonged diastolic coronary perfusion may have resulted in a lower right atrial PvCO2 and a diminished A-V CO2 gradient. The use of lactate as a marker for tissue perfusion should probably be abandoned, especially whilst on a beta-blocker. Further, urine output remained the same.
Thus, the experimental paradigm of Du et. al. cannot definitively inform us of vascular waterfall in the septic kidney. This is because their pressure-flow measurements were prolonged; unfortunately, their results may have simply been a test of renal cortical auto-regulation.