Phenylephrine increases cardiac output by raising cardiac preload in patients with anesthesia induced hypotension

Phenylephrine is conventionally thought to negatively affect CO, or at best to have no influence if the cardiac contractility is able to overcome the increased afterload without loss of SV [10]. Our hypothesis, however, was on the contrary, that if a relative hypovolemia is present due to anesthesia-induced excessive vasodilation of the capacitance vessels, this could be corrected by phenylephrine, inducing an improved centralization of the available blood, eventually resulting in an increase in CO.

The main finding of this prospective study was that in patients with anesthesia-induced hypotension and preload dependency—defined as PPV > 12%, phenylephrine increases CO by virtue of an increase in return function. This is reflected in multiple distinct indices, all indicating an increase in CO, owing to a rightward shift in the position of the heart on the Frank–Starling relationship: the dissociation of MAP and SVR at T₁, CVP, PPV, CO, and EtCO₂.

In patients undergoing sigmoidectomy, a relative hypovolemia is common owing to the combination of several factors: the patients have been fasting from the night before, had bowel preparation, and received an epidural loading dose, combined with general anesthesia, all causing vasodilation.

This pharmacologically induced vasoplegia, together with gravitational venous pooling in the lower limbs due to the leg-down patient positioning—to prevent conflict with the surgeon’s arms—and an increase in abdominal pressure often results in a markedly decreased venous return [11]. While an increase in cardiac preload enhances the CO, recent evidence revealed that a zero-balance fluid approach is recommended in the elective perioperative setting to avoid adverse effects of unnecessary, excessive fluid administration [12, 13]. In bowel surgery, the additional concern for return of peristalsis, and oedema of the gut tissue emphasizes the importance of moderating fluid administration. Fluid restriction is therefore considered a part of the care package for enhanced recovery after colorectal surgery [14].

Importantly, this relative hypovolemic state is just of temporary occurrence. When surgery ends, and the patient awakens, the venous tonus and consequently CO will increase again spontaneously. As such, an alternative to enhance venous return, other than the irrevocable administration of fluids would be desirable. In those cases, the option of recruiting internal blood volume by increasing venous return through reversible pharmacological means such as vasopressors acting predominantly on capacitance vessels might be a potential alternative.

In an animal study, the average effect of phenylephrine on CO was related to the preload dependency of the heart: when the heart was preload dependent, phenylephrine induced an increase in CO [7]. Similarly, in a study with human patients, CO and SV decreased in preload-independent patients through an increase in cardiac afterload, but by virtue of increased venous return were unchanged in those that were preload-dependent [6].

In contrast to most studies analysing the combined effects of phenylephrine [3], its beneficial effect on the return function is much more pronounced in our patients where the combination of general anesthesia, leg-down position and epidural analgesia—which reduces alpha tone—induced distinctly different physiological conditions [15]. The actual blood flow in the body is determined by the intersection of the cardiac function and the return function, the latter defined by the stressed blood volume, of which the bulk is in the small venules and veins [16]. As such, the CO response to phenylephrine is very dependent on the starting condition of the return function: if the patient is volume replete, with good reserves in unstressed volume and minimal initial tone in the veins draining the compliant region, phenylephrine can recruit unstressed volume into stressed volume by contractions of the smooth muscles in the walls of the vessels of the compliant part of the circulation, increasing the venous elastic recoil pressure. When this effect is greater than the increase in venous resistance, this will result in increased venous return and CO [16].

Figure 3 and Table 1 show that 90 s after an initial parallel increase in MAP and SVR following phenylephrine administration, the SVR curve started to decrease steeply, while the MAP curve drops more slowly. This dissociation indicates a second phenomenon increasing the blood pressure independently of the vascular resistance system. Because MAP ~ CO × SVR, an increase in venous return offers an additional contribution to the effect of phenylephrine on the MAP. The onset of this effect—about 90 s following the initial increase in MAP—corresponds to the expected time to reach a significant concentration of phenylephrine in the venous capacitance vessels. Remarkably, at T₁₀, the SVR is not significantly higher compared to T₋₁, while MAP and CO are still 24 and 20% higher, respectively (Table 1).

Next, the evolution of CVP—derived from the central venous catheter—also demonstrates a centralisation of venous blood, resulting in increased right ventricular preload, reflecting the postulated effects of phenylephrine on capacitance vessels. Even more, despite an increase in (left ventricular) CO—which depletes blood from the venous side—there is a persistent increase in CVP, implying that a higher cardiac preload is the primary driving force of the increased CO. This is also reflected in the steep increase in SV from 55 to 70 mL between T₋₁ and T₅, despite the increase in systemic afterload, with no significant change in HR (Table 1). As surrogate measures of cardiac preload dependency, PPV—determined from the systolic and diastolic blood pressure measurements through the ventilation cycle—and SVV also distinctly drop. While PPV is not strictly a measure of preload, its decrease indicates a right-shift of the heart on the Frank–Starling relationship and thus a transition from preload dependence to fluid unresponsiveness without fluid administration [6]. Since all patients were preload dependent, no comparison in hemodynamic effects of phenylephrine between preload-dependent and preload-independent states was possible.

As a separate independent measure, the evolution of the EtCO₂—measured by absorption spectrometry—indicates an increase in CO following phenylephrine [17, 18]. The significant average(SD) increase in EtCO₂ from 38(4) to 42(5) mmHg during stable ventilatory settings and invariant HR also indicates an increase in CO owing to increased cardiac preload.

An important pharmacological consideration is the rather long biological half-life of phenylephrine of 2–3 h. While clinical experience based on the evolution of the MAP following phenylephrine administration gives the impression of a very short half-life, the fast decline of the MAP after the initial increase merely reflects the redistribution of the phenylephrine, but not its elimination. The long biological half-life of phenylephrine permits to reach a clinically significant plasma concentration in the capacitance vessels. While norepinephrine also has a potent α-mimetic effect, it has a biological half-life of only 2–6 min [19, 20]. Because of the fast enzymatic degradation of norepinephrine [20], the concentration in the capacitance vessels remains much lower—remind that the splanchnic capacitance vessels only receive a relatively small fraction of the CO, while harbouring 25% of the total blood volume [21]. Given these pharmacokinetic differences, the ratio of the effects on CO and MAP will arguably be more balanced following phenylephrine compared to norepinephrine administration. This should, however, be substantiated in further research.

This emphasizes the physiological complexity determining the ultimate effect of the used vasopressor on the CO: (1) the balance of α- and β-adrenoceptor effects, (2) the pharmacokinetic properties of the vasoactive molecule, and (3) the preload dependency of the patient.

Our study has several limitations: firstly, no echocardiographic measurements of diastolic right and left volumes were performed to assess the preload effects of phenylephrine. The used ProAQT/Pulsioflex device is yet to be formally validated as sufficiently accurate to measure the absolute values of the variables of interest. Although calibrated devices offer more accurate absolute values, the relative changes to baseline values can be acceptably described by pulse wave contour analysis technology for assessing trend changes within a moderate range of acceptable physiological values. Additionally, the accuracy of the calculated SV based on wave-contour analysis may have been affected by the change in arterial elastance following phenylephrine administration [22]. The evolution of the CO, however, is only one of several variables indicating an increase in cardiac filling due to phenylephrine. The particular advantage of this device based on fast-reacting algorithms is its high temporal resolution, averaging 4 “sliding” intervals of 7.5 s, which results in complete recalculation within 30 s [23]. Secondly, the patients were rather volume dependent due to the combination of vasoplegia by epidural analgesia and hypovolemia by bowel preparation, which may explain the distinct effects observed in this study, compared to other reports. Thirdly, atropine was administered at the start of anesthesia to attenuate the negative effects on the MAP and CO after induction with TIVA [24]. This doesn’t influence the effects of phenylephrine on the venous return, but it probably blunts the reflex bradycardia induced by phenylephrine. While this is beneficial to more clearly demonstrate the effects on the venous return and global hemodynamics, it may narrow the external validity of its beneficial effects on the CO. Fourthly, Pmsa and Pvr values were derived mathematically and therefore, coupled with CO. Ideally, these values should have been assessed independently of CO. In our study, this was not done because the inspiratory hold manoeuvres to measure Pmsa would have unavoidably disturbed the accuracy of the primary research variables. The algorithm to calculate Pmsa and Pvr was validated previously [25]. Finally, no radiographic confirmation of the tip of the central venous catheter was performed at the time of CVP measurement. While this would arguably have minor influence on the calculated changes to baseline values of the studied variables, it may have affected the accuracy of the CVP measurements.

With respect to the described limitations, the current results must be interpreted within the constraints of potential shortcomings. Neither preload, contractility nor afterload were directly measured by the pulse contour technology. While the observed results suggest an improvement of venous return by virtue of phenylephrine administration, further research relying on direct measurements, like echography or thermodilution will be needed to fully describe the changes in cardiac preload, contractility and afterload in these clinical conditions.

In this study, the evolution of the hemodynamic variables was investigated after injection of a single dose of phenylephrine. This strategy was selected in order to most reliably depict the chronology of the hemodynamic changes. Additional research investigating the hemodynamic effects of phenylephrine in different baseline preload states may further elucidate the preload-dependency of these compound effects. The effects during continuous infusion on the investigated variables will have to be precisely determined, as well as comparison with the effects of alternative vasopressors. Ideally, an independent measure of CO, such as ultrasound or thermodilution should have been used to confirm the observed evolution of waveform-derived variables.