Secondary Logo

Journal Logo

Endothelium-Dependent Vasodilation in the Skin Microcirculation of Patients with Septic Shock

Kubli, Sandrine*; Boëgli, Yann*; Ave, Anne Dalle*; Liaudet, Lucas; Revelly, Jean-Pierre; Golay, Sandrine*; Broccard, Alain; Waeber, Bernard*; Schaller, Marie-Denise; Feihl, François*

Basic Science Aspects
Free

The evidence for endothelial dysfunction in sepsis is mostly restricted to animal models. We investigated endothelial function in the skin microcirculation of eight patients hospitalized for septic shock in an intensive care unit (ICU). All patients required adrenergic support. Twelve hemodynamically stable ICU patients without sepsis who did not receive any vasoactive medication were used as controls. The two groups were of similar age and sex ratio. For additional reference, 16 healthy, nonsmoking subjects matched for age and sex to the first two groups were also studied. The evaluation of endothelial function was based on the comparison of skin blood flow responses to iontophoretically applied acetylcholine (Ach, an endothelium-dependent vasodilator) and sodium nitroprusside (SNP, an endothelium-independent vasodilator). Skin blood flow was measured on the volar face of the forearm using laser Doppler imaging. Before application of Ach or SNP, the mean baseline skin blood flow was below 100 perfusion units (PU) in all subjects and did not differ between groups. The maximal increase in blood flow elicited by both agents was significantly depressed in the patients with sepsis (Ach: 167 ± 63 PU; SNP: 138 ± 34 PU, mean ± SD) compared with the ICU control patients (Ach: 291 ± 135 PU, P < 0.05; SNP: 261 ± 121 PU, P < 0.01) and the healthy, nonsmoking groups (Ach: 336 ± 98 PU, P < 0.01; SNP: 304 ± 81 PU, P < 0.01). The ratio of responses to Ach and SNP did not significantly differ between groups (septic: 1.22 ± 0.40; ICU control 1.18 ± 0.46, healthy, nonsmoking 1.12 ± 0.24, P = 0.86). Thus, sepsis was not associated with a selective depression of the endothelium-dependent response. These results suggest that the capacity of the endothelium to produce signals for vasorelaxation remains intact in the skin microcirculation of patients with septic shock.

*Division of Clinical Pathophysiology, Medical Intensive Care Unit, and Surgical Intensive Care Unit, University Hospital, CH-1011, Lausanne, Switzerland

Received 22 May 2002;

first review completed 20 Jun 2002; accepted in final form 11 Sep 2002

Address reprint requests to F. Feihl, MD, Division of Clinical Pathophysiology, BH19-317, CHUV, CH-1011 Lausanne, Switzerland.

This work was supported by the Tossiza Foundation and by the Swiss National Foundation for Scientific Research (grant no. 32-55946.98).

Back to Top | Article Outline

INTRODUCTION

It is currently accepted that the vascular endothelium plays a crucial role in the complex pathogenesis of sepsis-induced organ damage (1,2). In this condition, the endothelial cells (ECs) take on an inflammatory phenotype characterized in particular by the enhanced surface expression of procoagulant activity and leukocyte adhesion molecules. According to theory, these modifications are instrumental in the formation of microvascular dysfunction, as well as in the recruitment and activation of blood leukocytes. In the process, ECs are injured by various mechanisms, possibly including a direct action of bacterial products (3) and neutrophil-mediated oxidant damage (4). As a result, the controlled production of endothelium-derived vasoactive factors such as nitric oxide (NO), prostanoids, and endothelins is disrupted so that ECs loose the ability to regulate vessel tone, possibly amplifying disturbances in the distribution of blood flow and thus regional cellular hypoxia.

In animal models of sepsis, it has been repeatedly observed that the endothelium-dependent vasorelaxation elicited by acetylcholine (Ach) or other mediators either in vivo or in isolated vessels, was attenuated or lost entirely (5–9). Of note, the opposite finding has also been reported (10). In contrast to the abundance of experimental data, there is no direct evidence for an alteration of endothelium-mediated regulation of vascular tone in human sepsis. Ex vivo studies of isolated human vessels present obvious ethical difficulties. Recently, omental arteries excised from subjects with sepsis at the time of surgery for peritonitis showed no abnormality of endothelium-dependent relaxation in vitro (11). In vivo, the reactive hyperemia that follows the release of a transient interruption of regional blood supply has been found depressed in the skin or skeletal muscle of patients with severe sepsis (12–16). Reactive hyperemia is partly endothelium-dependent, but its many other determinants preclude the drawing of any specific conclusion (17). Other, more specific methods for the in vivo study of endothelium-dependent relaxation in humans, i.e., the response of regional blood flow to the intra-arterial infusion of Ach or the ultrasonic assessment of flow-mediated vasodilation in limb arteries are hardly feasible in a critical care setting.

In recent years, it has been proposed to observe endothelium-dependent vasodilation in the dermal microcirculation, the relevant stimulus being the noninvasive local application of Ach with iontophoresis, and skin blood flow being measured with laser Doppler flowmetry (18,19). As an endothelium-independent control, the microvascular response to the iontophoresis of sodium nitroprusside (SNP) can be similarly recorded. Due to its practical simplicity and noninvasive nature, this method is technically feasible in the ICU. We recently demonstrated its reproducibility (20). In the present study, we have used it to compare endothelium-dependent and endothelium-independent vasodilation in the dermal microcirculation of ICU patients who were nonseptic and as well as severely septic. We hypothesized that septic conditions would be associated with a blunting of the dermal blood flow response to iontophoretic application of Ach relative to that of SNP.

Back to Top | Article Outline

MATERIALS AND METHODS

Subjects

Three groups of subjects were studied, following approval by the Institutional Review Committee. All subjects gave written informed consent to the investigation. In the case of patients in septic shock on mechanical ventilation and heavy sedation, consent was sought, also in written form, from a relative.

Back to Top | Article Outline

Group with sepsis

This group comprised patients hospitalized in our medical intensive care unit who met criteria for severe sepsis. The latter required the fulfillment of both the screening criteria for possible sepsis and the additional confirmatory criteria for severe sepsis as defined in a recent epidemiological study (21).

The screening criteria were the presence of all of the following features: temperature > 38.3°C or < 35.6°C rectally, b) respiratory rate > 20/min or mechanical ventilation, c) heart rate > 90/min, and d) clinical evidence of infection; or one or more blood cultures positive for pathogen after 48 h of incubation.

The confirmatory criteria were: Presence of any of the following seven criteria, without alternative explanation: PaO2/FIO2 < 280 (intubated) or 40% face mask in use (nonintubated), arterial pH < 7.30; urine output < 30 mL/h; systolic blood pressure < 90 mmHg of fall in systolic blood pressure > 40 mmHg sustained for 2 h despite fluid challenge; systemic vascular resistance < 800 dynes/sec/cm−5; prothrombin time or partial thromboplastin time greater than normal or platelets <100 × 109/L or platelets decreased to <50% of most recent measurement before current day; and deterioration in mental status within 24 h.

Back to Top | Article Outline

ICU control group

These patients were hospitalized in the same ICU, but at the time of examination, they fulfilled none of the various items in the screening or confirmatory criteria cited above. Furthermore, they received no vasoactive medication. They matched the patients with sepsis regarding both age and sex.

Back to Top | Article Outline

Healthy group

These were healthy, nonsmoking normocholesterolemic subjects who also matched the patients with sepsis regarding both age and sex.

Back to Top | Article Outline

Endothelial function in the skin microcirculation

The details of the methods used in the present study have been previously published (20,22) and will only be given here in summarized form.

Back to Top | Article Outline

Measurement of dermal blood flow

We used a recently developed laser Doppler Imager (LDI; Moor Instruments, Axminster, UK). In this device, a moving mirror directs a beam of laser light (633 nm) onto the skin. Computer-controlled rotations of the mirror allow the scanning of a square region, the size of which may be adjusted by the operator. From the analysis of the backscattered Doppler-shifted light, microvascular blood flow in each of up to 256 x 256 adjacent spots (“pixels”) is calculated. The final result is a computer-generated, color-coded image of the spatial distribution of microvascular blood flow. Total flow, expressed in perfusion units (PU) according to the principle of laser Doppler flowmetry, can be calculated later by summing the pixel values in an arbitrarily shaped region of interest within the scanned area. In comparison with usual fiber optic laser Doppler probes, the LDI allows the measurement of microvascular blood flow of a much larger area with no skin contact.

Back to Top | Article Outline

Iontophoresis of vasoactive agents

Iontophoresis is a noninvasive method for transdermal transfer of charged molecules locally on the skin by means of an externally applied electrical current (23).

To follow with LDI the response of skin blood flow to iontophoretically applied vasoactive agents, we used custom-made ring-shaped chambers fitted with a copper electrode that was connected to a iontophoresis controller (MIC1-e; Moor Instruments). The chambers were affixed with double-sided tape to the skin on the volar face of the forearm, filled to the rim with a solution of either 1% Ach (endothelium-dependent stimulus) or 0.1% SNP (endothelium-independent stimulus), and covered with a transparent glass slip. The internal diameter of the chamber was 10 mm, so that exposed skin area was 0.78 mm2 An indifferent ECG electrode was placed on the wrist. Polarity was adapted to the electric charge of the vasoactive molecule (chamber positive for Ach and negative for SNP). The pulsed iontophoretic protocols were as described (20). With these protocols, we found in healthy nonsmokers that the delivered doses of Ach and SNP were largely supramaximal, i.e., they exceeded those required for a maximal effect by a factor of at least 5 (B. Waeber and F. Feihl, unpublished observations).

Electrical current alone may induce a vasodilatation due to the stimulation of local sensory nerves, a response that may be inhibited by local anesthesia (23,24). In the present study, Ach and SNP were therefore applied on skin pretreated for 1 h with a local anesthetic cream (5% EMLA cream; Astra Pharmaceutica, Dietikon, Switzerland) under an occlusive dressing (Tegaderm; 3M Health Care, Loughborough, UK). The effective inhibition of current-induced vasodilatation was systematically controlled by applying the same amount of current to a adjacent control chamber filled with isotonic saline.

Back to Top | Article Outline

Protocol

The responses of skin blood flow to the iontophoresis of Ach and SNP were determined in close succession on two adjoining but distinct sites on the volar face of the forearm. The recording of both responses took <1 h. The iontophoresis chambers were placed on sites of apparently normal skin. In six patients with sepsis, the responses to Ach and SNP were determined twice 24–48 h apart (days 1 and 2), taking care to use the precisely the same sites on both occasions.

Back to Top | Article Outline

Data analysis

The responses of skin blood flow to Ach and SNP were quantified as the maximal increase above the baseline value recorded on the same skin site immediately before iontophoresis. The measurements obtained in the three groups were compared with one-way simple analysis of variance (ANOVA). When the F value was significant, pairwise comparisons were carried out using modified t tests (25). The alpha level of all tests was set at 0.05. In the following text, results are summarized as the mean ± SD.

Back to Top | Article Outline

RESULTS

A total of 12 septic subjects were studied at least once in the ICU. Among them, eight met the standard criteria for septic shock (26) and required adrenergic support to maintain a mean blood pressure >70 mmHg. Neither vasopressor nor inotropic therapy was needed in the remaining four patients. We felt that this latter number was too small to justify subgroup analysis, and we dealt with the ensuing heterogeneity by restricting the presented data to the eight patients in septic shock (Table 1). Twelve control ICU patients (Table 2) and 16 healthy subjects were included.

Table 1

Table 1

Table 2

Table 2

The source of infection was known in all subjects with septic shock, and the responsible microorganisms were identified in all but one case (Patient 2). The time lapse from the appearance of the screening criteria to the onset of septic shock was 2 days or less in all but one patient (Patient 2, 10 days). The interval from the diagnosis of septic shock to day 1 was 3 days or less. All patients with sepsis were on mechanical ventilation, had a pulmonary catheter in place for clinical reasons, and had been receiving adrenergic support, which always included norepinephrine (Table 1), for 12–72 h before the first laser Doppler examination. In accordance with the rules for inclusion, no patient in the ICU control group had any vasoactive medication at the time of examination. The Sepsis-Related Organ Failure Assessment (SOFA) score shown in Table 1 was calculated as described (27,28), except that the neurologic Glasgow coma score was omitted due to the difficulty of obtaining reliable data on this item in the patients with sepsis, all but two of whom were receiving heavy sedation (continuous i.v. infusion of either propofol or a midazolam/opioid combination). With this modification, the SOFA score had a maximal possible value of 20. It averaged 13.8 ± 3.6 in the group with sepsis and 2.8 ± 2.3 in the ICU control group (P < 0.001). The hemodynamic, treatment, and SOFA score data in Table 1 were recorded on the day of assessment of microvascular reactivity (day 1 in the patients with sepsis who were examined on two occasions).

The three study groups were matched for sex ratio (septic: 2 female, 6 male; ICU control: 3 female, 9 male; healthy: 4 female, 12 male) and did not differ statistically for age (septic: 58.0 ± 12.9 years; ICU control: 59.8 ± 12.4 years; healthy: 58.7 ± 12.4 years, P = 0.96). The mean blood pressure in the healthy group (89 ± 7 mmHg) and in the ICU control group (87 ± 14 mmHg) was significantly higher than in the group with sepsis (70 ± 9 mmHg, P < 0.01).

The control measurements made in absence of vasoactive agents in the iontophoretic chambers indicated that the iontophoretic current per se had no effect on skin blood flow (data not shown).

Figure 1 shows the average skin blood flow responses to the iontophoresis of Ach and SNP in the three groups. To construct the average curve of the group with sepsis, data from the six patients who underwent the assessment twice were those collected on day 1. The flows recorded in baseline conditions, i.e., in the minute preceding iontophoresis, were of comparable magnitude on the sites used to administer Ach and SNP, and did not differ between groups (Table 2). The maximal increase in flow caused by Ach and SNP was significantly smaller in patients with sepsis than in ICU control and healthy subjects (Fig. 2 and Table 2). The ratio of the maximal changes induced by Ach and SNP was the same in the three groups (Table 2). Furthermore, when individual responses to Ach were plotted against the corresponding responses to SNP, the data points for all subjects appeared to cluster around a common line (Fig. 2). These observations indicate that the effects of Ach and SNP on skin blood flow were blunted to a similar degree in the patients with sepsis when compared either with the ICU control or with the healthy subjects.

Fig. 1

Fig. 1

Fig. 2

Fig. 2

In the group with sepsis, there was no statistically significant correlation of the responses of Ach and SNP with any of the following variables: mean systemic blood pressure, cardiac index, systemic vascular resistance, blood lactate, SOFA score, dose of norepinephrine, or duration of the septic shock.

In six patients with sepsis, the skin blood flow responses could be measured twice, 24–48 h apart.(days 1 and 2). In four of these subjects, the severity of the septic state decreased from days 1 to 2, as shown by reduced requirements for adrenergic support in the face of a maintained or higher blood pressure, and this was associated with an increase in the skin blood flow responses to both Ach and, except in one patient, SNP (Fig. 3, closed symbols and plain lines). In the other two patients, the opposite evolution occurred (Fig. 3, open symbols and dotted lines). The changes observed from days 1 to 2 in responses to Ach and SNP were tightly correlated (P < 0.001), with a linear relationship of slope close to 1 (Fig. 4).

Fig. 3

Fig. 3

Fig. 4

Fig. 4

Back to Top | Article Outline

DISCUSSION

In the present study, we examined, for the first time in patients with sepsis, the reaction of microvascular skin blood flow to an endothelium-dependent (Ach) and to an endothelium-independent vasodilator (SNP). In these subjects, all of whom were in septic shock when first examined on day 1 (Table 1), the vasodilation induced by both agents was blunted in comparison with two different control groups matched for age and sex (Fig. 1). This effect of septic shock was not unexpected, as it is consistent with the depression of reactive hyperemia in the skin microcirculation reported by others in similar conditions (13,14,29). The new finding of our study consists in the complete lack of a differential effect of sepsis on endothelium-dependent and endothelium-independent vasodilation, as shown by the similar ratio of the responses to Ach and to SNP in the three groups (Table 2 and Figure 2), as also shown by the constancy of this ratio over time in spite of marked changes in the severity of the septic state (Figs. 3 and 4). These results suggest that the capacity of the endothelium to produce signals for vasorelaxation remains intact in the skin microcirculation of patients with septic shock.

Why are the responses of skin blood flow to vasodilatory stimuli impaired in septic shock? Potentially, sepsis and shock could reduce the efficiency of transdermal iontophoretic drug delivery. However, this is unlikely because the doses of Ach and SNP used in the present study were largely supramaximal (see “Materials and Methods”). Neither can the results shown in Figure 1 be exclusively explained by a background vasodilated state because the baseline blood flows (i.e., as measured before iontophoresis) did not differ between the three groups of subjects (Fig. 1 and Table 2). Unfortunately, our data contain no information regarding the actual mechanisms of impaired microvascular responses in the patients with sepsis. We can only speculate on several nonexclusive possibilities, which include a low capillary hematocrit related to rheologic abnormalities (because laser Doppler flowmetry specifically captures the movement of red blood cells), microvascular obstruction, vascular smooth muscle deficit, and an excess of vasoconstrictor influences, either endogenous or related to treatment with norepinephrine. Finally, the response of skin blood flow to locally applied vasodilators depends not only on events in the local dermal microcirculation, but also on perfusion pressure, which can be decreased either by upstream regional arterial/arteriolar constriction or by systemic hypotension.

Our main conclusion, namely that endothelial function seems preserved in the skin microcirculation of patients with septic shock, highly rests on assumptions concerning the differential role of the endothelium in the vasodilations induced in this vascular bed by either Ach or SNP. The first point concerns the mode of local application of these agents using iontophoresis. The iontophoretic current may antidromically stimulate afferent nerve endings, leading to the local release of vasoactive mediators (24). Our data are not confounded by this factor, as we pretreated the tested skin with lidocaine in the form of an EMLA patch, and so completely prevented the potential response to current alone, as previously observed by ourselves (20) and others (23,24). The septic conditions might have differentially affected the penetration of Ach and SNP through the skin. Although this possibility cannot be rejected, it is unlikely to have biased our results, for the reasons already explained (i.e., the use of largely supramaximal doses of each agent). As a second point, few would dispute that SNP, a donor of nitric oxide, exerts an endothelium-independent effect. However, the endothelium dependence of the microcirculatory response to the local application of Ach in the skin has not been formally demonstrated. Indeed, such a demonstration entails the experimental destruction of the endothelium while preserving the underlying vascular smooth muscle, and is utterly impossible in an intact microvascular bed. Nevertheless, several arguments make it reasonable to assume that the response to Ach recorded with the methodology used in the present study is mediated by the endothelium. First, a mass of experimental data involving various vascular beds and vessel size in various species, including humans, consistently indicate that an intact endothelium is a necessary condition for Ach to relax vascular smooth muscle (30). Second, pharmacological interventions aimed at inhibiting the production of known endothelium-derived relaxing factors such as prostacyclin (19,31) and nitric oxide (32) have been able to attenuate the response of skin blood flow to the local application of Ach in humans, although findings have not been completely uniform in that respect (19). Third, insulin-mediated vasodilation in the whole leg, a known endothelium-dependent response, was found correlated with the skin blood flow response to Ach but not SNP (33). Finally, using the exact methodology described here, we recently characterized the skin blood flow responses of middle-aged healthy male smokers (20 packs/year), a population at risk for incipient endothelial dysfunction. Using nonsmokers matched for age and sex as a reference, we found that the response of the smokers to Ach was markedly more attenuated than their response to SNP (34).

Because we only studied patients with sepsis requiring adrenergic support, our data are open to the criticism that a confounding influence of this treatment potentially obscured any differential effect that sepsis might have on the endothelium-dependent and -independent vasodilatory responses. Although we cannot entirely discount this possibility, we have limited data that argue against it. First, there was no statistically significant correlation of the responses to Ach and SNP on one hand, and the dose of norepinephrine received by each patient on the other. Second, in all patients in whom measurements were carried out twice, the responses to Ach and SNP covaried form days 1 to 2 (Fig. 4), including in the three cases where adrenergic support was no longer present on day 2 (Fig. 3). Finally, the aforementioned four additional subjects with sepsis who did not require vasopressor therapy (see beginning of the “Results” section) did not display a selective depression of the Ach relative to the SNP response (mean maximal change in skin blood flow ± SD: Ach 275 ± 75 PU, SNP 255 ± 75 PU; mean Ach/SNP ratio: 1.10 ± 0.14).

The apparent preservation of endothelium-dependent vasorelaxation in the skin microcirculation of patients with severe sepsis must be qualified in several respects. As a first point, this observation should not be extrapolated to other organs, in particular to those more essential to patient survival than the skin. Unfortunately, the skin is the only window presently available for the study of endothelial function in such critical conditions. Second, it must be repeated that we probed the microcirculation, whereas much of the experimental work that led to the concept of sepsis-induced loss of endothelium-dependent vasodilation was based on the observation of larger vessels. In rat endotoxic shock, interestingly, the vasodilation induced by Ach was abolished in femoral arteries, reduced in proximal (A1), but normal in more distal mesenteric arterioles (A2 and A3) (35). In a morphologic study by the same group in the same species, endotoxemia was associated with severe damage to the endothelium of the femoral artery, but not of the mesenteric arterioles (36). These observations point to the possibility that the endothelial damage inflicted by sepsis could be less marked in the micro-than in the macrocirculation. We must acknowledge that this issue is not entirely clear because in a different model of sepsis induced by cecal ligation and puncture in the rat, distal arterioles in the cremaster muscle had a blunted dilation response to the local administration of Ach (37).

Apart from the present work, we know of only one other study that informs on the status of endothelium-dependent vasodilation in human sepsis. In eight patients with severe sepsis who required adrenergic support and were surgically treated for acute peritonitis, Stoclet and coworkers (11) were able to isolate omental arteries and found an unimpaired relaxation response to the endothelium-dependent vasodilator bradykinin. Our data are consonant with this report. It appears that in human severe sepsis or septic shock, the ability of the endothelium to generate signals for vasorelaxation is conserved in at least some vascular beds. The functional significance of this fact remains to be determined. We might speculate that it is of some importance in mitigating the local disturbances of oxygen supply.

Back to Top | Article Outline

ACKNOWLEDGMENTS

The authors thank the nursing staff for their help at the bedside, and Camille Anglada for outstanding technical assistance. The purchase of the laser-Doppler imager was made possible by a grant from the Tossiza Foundation.

Back to Top | Article Outline

REFERENCES

1. Glauser MP, Heumann D, Baumgartner JD, Cohen J: Pathogenesis and potential strategies for prevention and treatment of septic shock: an update. Clin Infect Dis 18(Suppl 2):S205–S216, 1994.
2. Maier RV, Bulger EM: Endothelial changes after shock and injury. New Horizons 4:211–233, 1996.
3. Brigham KL, Meyrick B: Endotoxin and lung injury. Am Rev Respir Dis 133:913–927, 1986.
4. Minamiya Y, Abo S, Kitamura M, Izumi K, Kimura Y, Tozawa K, Saito S: Endotoxin-induced hydrogen peroxide production in intact pulmonary circulation of rat. Am J Respir Crit Care Med 152:348–354, 1995.
5. Wylam ME, Samsel RW, Umans JG, Mitchell RW, Leff AR, Schumacker PT: Endotoxin in vivo impairs endothelium-dependent relaxation of canine arteries in vitro. Am Rev Respir Dis 142:1263–1267, 1990.
6. Kuo L, Chilian WM, Davis MJ, Laughlin MH: Endotoxin impairs flow-induced vasodilation of porcine coronary arterioles. Am J Physiol 262:H1838–H1845, 1992.
7. Fullerton DA, McIntyre RC, Hahn AH, Agrafojo J, Koike K, Meng X, Banerjee A, Harken AH: Dysfunction of cGMP-mediated pulmonary vasorelaxation in endotoxin-induced acute lung injury. Am J Physiol 268:L1029–L1035, 1995.
8. Wang P, Ba ZF, Chaudry IH: Endothelium-dependent relaxation is depressed at the macro- and microcirculatory levels during sepsis. Am J Physiol 269:R988–R994, 1995.
9. Wang P, Ba ZF, Cioffi WG, Bland KI, Chaudry IH: Salutary effects of ATP-MgCl2 on the depressed endothelium-dependent relaxation during hyperdynamic sepsis. Crit Care Med 27:959–964, 1999.
10. Chen SJ, Wu CC, Yen MH: Alterations of ex vivo vascular reactivity in intraperitoneal sepsis. J Cardiovasc Pharmacol 24:786–793, 1994.
11. Stoclet JC, Martinez MC, Ohlmann P, Chasserot S, Schott C, Kleschyov AL, Schneider F, Andriantsitohaina R: Induction of nitric oxide synthase and dual effects of nitric oxide and cyclooxygenase products in regulation of arterial contraction in human septic shock. Circulation 100:107–112, 1999.
12. Hartl WH, Gunther B, Inthorn D, Heberer G: Reactive hyperemia in patients with septic conditions. Surgery 103:440–444, 1988.
13. Haisjackl M, Hasibeder W, Klaunzer S, Altenberger H, Koller W: Diminished reactive hyperemia in the skin of critically ill patients. Crit Care Med 18:813–818, 1990.
14. Young JD, Cameron EM: Dynamics of skin blood flow in human sepsis. Intensive Care Med 21:669–674, 1995.
15. Astiz ME, DeGent GE, Lin RY, Rackow EC: Microvascular function and rheologic changes in hyperdynamic sepsis. Crit Care Med 23:265–271, 1995.
16. Neviere R, Mathieu D, Chagnon JL, Lebleu N, Millien JP, Wattel F: Skeletal muscle microvascular blood flow and oxygen transport in patients with severe sepsis. Am J Respir Crit Care Med 153:191–195, 1996.
17. Koller A, Kaley G: Role of endothelium in reactive dilation of skeletal muscle arterioles. Am J Physiol 259:H1313–H1316, 1990.
18. Morris SJ, Shore AC, Tooke JE: Responses of the skin microcirculation to acetylcholine and sodium nitroprusside in patients with NIDDM. Diabetologia 38:1337–1344, 1995.
19. Noon JP, Walker BR, Hand MF, Webb DJ: Studies with iontophoretic administration of drugs to human dermal vessels in vivo: cholinergic vasodilatation is mediated by dilator prostanoids rather than nitric oxide. Br J Clin Pharmacol 45:545–550, 1998.
20. Kubli S, Waeber B, Dalle-Ave A, Feihl F: Reproducibility of laser-Doppler imaging of skin blood flow as a tool to assess endothelial function. J Cardiovasc Pharmacol 36:640–648, 2000.
21. Sands KE, Bates DW, Lanken PN, Graman PS, Hibberd PL, Kahn KL, Parsonnet J, Panzer R, Orav EJ, Snydman DR: Epidemiology of sepsis syndrome in eight academic medical centers. Academic Medical Center Consortium Sepsis Project Working Group. J Am Med Assoc 278:234–240, 1997.
22. Kubli S, Feihl F, Waeber B: β-Blockade with nebivolol enhances the acetylcholine-induced cutaneous vasodilation. Clin Pharmacol Ther 69:238–244, 2001.
23. Morris SJ, Shore AC: Skin blood flow responses to the iontophoresis of acetylcholine and sodium nitroprusside in man: possible mechanisms. J Physiol (Lond) 496:531–542, 1996.
24. Wardell K, Naver HK, Nilsson GE, Wallin BG: The cutaneous vascular axon reflex in humans characterized by laser Doppler perfusion imaging. J Physiol (Lond) 460:185–199, 1993.
25. Winer BJ:Statistical Principles in Experimental Design. New York: McGraw-Hill, 1971, pp 172–175.
26. Abraham E, Matthay MA, Dinarello CA, Vincent JL, Cohen J, Opal SM, Glauser M, Parsons P, Fisher Jr, CJ Repine JE: Consensus conference definitions for sepsis, septic shock, acute lung injury, and acute respiratory distress syndrome: time for a reevaluation. Crit Care Med 28:232–235, 2000.
27. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, Reinhart CK, Suter PM, Thijs LG: The SOFA (Sepsis-Related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 22:707–710, 1996.
28. Ferreira FL, Bota DP, Bross A, Melot C, Vincent JL: Serial evaluation of the SOFA score to predict outcome in critically ill patients. J Am Med Assoc 286:1754–1758, 2001.
29. Pšschl JM, Weiss T, Fallahi F, Linderkamp O: Reactive hyperemia of skin microcirculation in septic neonates. Acta Paediatr 83:808–811, 1994.
30. Vanhoutte PM: How to assess endothelial function in human blood vessels. J Hypertens 17:1047–1058, 1999.
31. Khan F, Davidson NC, Littleford RC, Litchfield SJ, Struthers AD, Belch JJ: Cutaneous vascular responses to acetylcholine are mediated by a prostanoid-dependent mechanism in man. Vasc Med 2:82–86, 1997.
32. Warren JB: Nitric oxide and human skin blood flow responses to acetylcholine and ultraviolet light. FASEB J 8:247–251, 1994.
33. Serne EH, Gans ROB, ter Maaten JC, ter Wee PM, Donker AM, Stehouwer CDA: Capillary recruitment is impaired in essential hypertension and relates to insulin's metabolic and vascular actions. Cardiovasc Res 49:161–168, 2001.
34. Pellaton S, Kubli S, Feihl F, Waeber B: Blunted vasodilatory responses in the cutaneous microcirculation of cigarette smokers. Am Heart J 144:269–274, 2002.
35. Baker CH, Sutton ET: Arteriolar endothelium-dependent vasodilation occurs during endotoxin shock. Am J Physiol 264:H1118–H1123, 1993.
36. Sutton ET, Zhou Z, Baker CH, Price JM, Chen Y: Differences in arterial and arteriolar endothelial structure during endotoxin shock. Circ Shock 41:71–76, 1993.
37. Tyml K, Yu J, McCormack DG: Capillary and arteriolar responses to local vasodilators are impaired in a rat model of sepsis. J Appl Physiol 84:837–844, 1998.
Keywords:

MeSH; sepsis; endothelium; vascular; microcirculation; skin; laser Doppler flowmetry; acetylcholine

©2003The Shock Society