Hypovolemia, resulting from blood or water loss, vasoplegia, or a capillary leak, is one of the most common reasons for shock states in the critically ill. Consequently, fluid replacement therapy is one of the cornerstones of the treatment in patients with trauma and/or septic shock. In addition, due to decreased peripheral vascular resistances in combination with altered microvascular blood flow, most frequently the ICU physician is also obliged to use vasoactive drugs in order to control hemodynamics and tissue perfusion. In this context, the question of the adequate perfusion pressure may assume crucial importance. In fact, according to the “two pores” theory for transcapillary fluid exchange, the rate of fluid escape from the vascular into the interstitial space depends on both the hydrostatic capillary pressure and the microvascular permeability [1].

In this issue of Intensive Care Medicine, Dubniks et al. [2] present an interesting article on the effects of increasing mean arterial pressure with i.v. noradrenaline on plasma volume in a rodent model of increased permeability, induced by injection of dextran 70 in order to obtain an anaphylactic reaction. Plasma volume was measured before and after albumin infusion, and mean arterial pressure was maintained throughout the experiment by noradrenalin infusion to a level above baseline (on average 100 mmHg). The main result was that the increase in blood pressure induced a loss of plasma volume, which was more pronounced in the animals with increased vascular permeability. In previous experiments both in hemorrhaged cats [3] and in rodents with anaphylaxia-related increased permeability similar to the present study [4], the authors' group had already reported that the relatively poor plasma expansion capacities of various colloid solutions were caused by increased transcapillary leakage due to increased microvascular permeability. In their current study, the authors now demonstrate that the flux of fluids from intravascular to extravascular spaces is, at least in part, driven by the intravascular hydrostatic pressure, and, moreover, that increasing the colloid osmotic pressure fails to compensate this effect.

This very interesting result immediately raises the following points:

  1. 1.

    What would happen with another synthetic colloid? In this context we may focus on a previous study from this group. In the same model of dextran 70-induced hyperpermeability the authors reported the effects of several synthetic colloids. Albumin was a more effective plasma volume expander than hydroxyethyl starch or gelatin. However, all colloids exhibited lower plasma expanding properties than observed before in animals with normal permeability [5]. Nevertheless, even when using the best plasma expander in this model (albumin), increasing the intravascular pressure fostered a larger volume loss into the extravascular space.

  2. 2.

    What would happen with another model of capillary leakage? In Dubnik et al.'s study [2], hyperpermeability was induced in rats by an anaphylactic reaction to dextran 70 infusion. In contrast to anaphylaxia, sepsis is a much more frequent situation in intensive care practice, and in a different experimental model of capillary permeability, results may thus have been different. In fact, in a porcine model of septic shock-associated capillary leakage, Marx et al. reported the effects on plasma volume using either hydroxyethyl starch or modified fluid gelatin [6]. The artificial colloids maintained plasma volume and colloid osmotic pressure, but increased albumin escape rate. Interestingly, mean arterial pressure was lower than baseline in all treated animals during the fluid resuscitation period, and the authors did not investigate the impact of increasing perfusion pressure with vasoactive drugs. In the light of the results reported by Dubnik et al., one could speculate that this “hypotensive” resuscitation may have limited the loss of plasma volume.

  3. 3.

    The transfer of the present results into the clinical scenario needs to be tempered. Indeed, the results of this study might suggest at first glance that restoration of perfusion pressure is detrimental in term of plasma volume loss. According to the experimental protocol noradrenalin infusion was targeted to achieve a mean arterial pressure just above the baseline value. This target pressure may be considered too high when taking into account the protocol of early goal-guided therapy strategy [7] or the very recent recommendations of the International Consensus Conference on hemodynamic monitoring in shock and implications for management [8]. In addition, Ledoux et al. reported in 10 patients with septic shock the effects that increasing mean arterial pressure with norepinephrine from 65 mmHg to 85 mmHg did not significantly affect systemic oxygen metabolism, skin microcirculatory blood flow, urine output, or splanchnic perfusion [9]. Similarly, Bourgoin et al. reported that in patients with septic shock, increasing mean arterial pressure from 65 mmHg to 85 mmHg by incremental noradrenalin neither affected metabolic variables nor improved renal function [10]. Finally, reducing the needs for plasma expanders per se in patients with increased permeability is debatable. On the one hand, the therapeutic strategy of large fluid challenge increased mortality in ARDS patients: in fact, in a prospective clinical study in 113 patients Simmons et al. reported that a large fluid intake induced a higher mortality when associated with a weight gain [11]. Furthermore, using pulmonary artery catheterization and thermal-green dye double-indicator dilution measurements of extravascular lung water, Shuller et al. reported that survivors had no significant fluid gain or change in extravascular lung water but decreased pulmonary artery occlusion pressure and body weight [12]. Finally, a randomized clinical trial confirmed the benefit of restrictive strategy of fluid management in patients with pulmonary edema (reduced extravascular lung water, ventilator days and ICU days) [13]. On the other hand, in an experimental study in endotoxic rats, we used terlipressin (a potent V1a agonist) in combination with fluid challenge in order to increase mean arterial pressure from 70 mmHg to 90 mmHg. In contrast to terlipressin alone, cardiac output and ileal microcirculation as well as survival rate improved only when terlipressin was combined with fluid challenge [14]. Moreover, Rivers et al. nicely demonstrated the important fact that early and large amounts of fluid improved survival in septic patients [7].

We thank the Grände team for their systematic experimental approach regarding the effect of fluid losses in various models of normal or increased vascular permeability, and we would like to encourage a further assessment of the authors' concept in more clinically relevant—but probably more difficult—long-term, large-animal models, such as endotoxic or polymicrobial shock, investigating various amounts of different types of fluids and vasoconstrictive drugs together with metabolic markers of perfusion or oxygen exchange and/or mortality.