Abstract

Septic shock carries a high mortality. Characteristic hemodynamic changes include hypotension due to arterial and venous dilatation and impaired cardiac contractility, but there are also profound changes in metabolic, respiratory, hematologic, and host-defense functions.¹²³ Decreased peripheral responsiveness to vasoactive agents contributes to the progressive decline of the systemic blood pressure, which ultimately leads to tissue hypoperfusion and circulatory failure.¹²³⁴

Bacterial wall lipopolysaccharide (endotoxin) is considered an important etiologic agent in the pathophysiology of septic shock⁵ and has been administered to animals to produce a model of the human condition.⁶⁷⁸ An understanding of the mechanisms by which endotoxin produces changes in vascular behavior is likely to lead to novel therapies. Two major approaches to the exploration of the vascular effects of endotoxin have been adopted: studies in vitro have focused on functional, biochemical, and molecular changes induced by direct exposure of cells or blood vessels to endotoxin,^9101112 whereas studies in vivo have assessed vascular changes after initiation of a generalized inflammatory response induced by systemic administration of endotoxin.¹³¹⁴ Two phases in the response to endotoxin have been revealed: an acute phase that occurs over 5 to 90 minutes followed by a delayed phase that begins 3 to 4 hours after administration of endotoxin and lasts for up to 24 hours.¹⁴¹⁵ With repeated administration, tolerance of the effects of endotoxin develops.¹⁰¹⁶¹⁷

The proposed mechanisms that underlie the acute vascular effects of endotoxin include enhanced synthesis of bradykinin,¹¹¹³ NO,⁴⁸ prostanoids,¹² platelet-aggregating factor,¹⁸ cytokines¹⁹²⁰ or leukotrienes,²¹ and direct effects of endotoxin on the vascular endothelium.²² Some studies have shown increased production of the vasodilator mediator NO in the vessel wall due to expression of an inducible isoform of NO synthase,²³ and enhanced production of vasodilator prostanoid may occur after induction of COX-II.²⁴ Most studies have used arterial vessels²⁵ ; however, venodilatation is an important component of the pathophysiology of septic shock and could contribute to the changes in cardiac filling pressure and cardiac output. It is unclear whether the mechanisms and action of endotoxin are similar in veins, resistance vessels, and conduit arteries.

Systemic administration of endotoxin to healthy human volunteers produces cardiovascular changes similar to those seen in animal models; arterial¹⁵²⁵ and venous²⁶²⁷ dilatation and hypotension are seen within 60 minutes and persist for up to 8 hours,¹⁵²⁷ but it is difficult to dissect the mechanisms in whole-body studies. Human vessels in vitro show inconsistent responses to endotoxin,²⁸ and it is possible that the vascular responses seen in vivo are dependent on involvement of other cell types or tissues, increased concentrations of cytokines circulating in blood,²⁰ infiltration of inflammatory cells into the vessel wall,²³ or neurohumoral effects of systemically administered endotoxin.²⁹ One approach to gaining insight is to investigate the local vascular effect of endotoxin in vivo. This is what we have attempted in the experiments reported here. We used superficial hand veins to explore the mechanisms of the changes seen and examined the phenomenon of tolerance to endotoxin.

Methods

Studies were approved by the local research ethics committee and were performed on male (14) and female (18) subjects aged 19 to 38 years. Subjects who were included gave their informed, written consent and stated that they were healthy and were not taking any medication. Throughout the study, subjects lay supine in a temperature-controlled laboratory (28°C to 30°C) with one hand placed on an angled support above the level of the heart.

Assessment of Venous Response to Norepinephrine

To assess the responses to norepinephrine in a single vein on the back of the hand, a congesting cuff was placed around the upper arm and inflated to 40 mm Hg. Drugs or physiological saline (0.9%) were infused continuously (0.25 mL/min) through a 23-gauge needle inserted into the study vein. A lightweight probe was placed 5 to 10 mm downstream from the tip of the needle on the skin overlying the summit of the vessel. We assessed internal venous diameter every 5 minutes by recording the linear displacement of the probe when the pressure in the congesting cuff was lowered from 40 to 0 mm Hg and then inflated to 40 mm Hg again.³⁰ The diameter of a vessel under constant distension pressure is determined by the state of contraction of the smooth muscle (Laplace's relationship), and contraction leads to a reduction in diameter.

Baseline recordings of vein size were made for 10 minutes during infusion of saline. The continuous infusion of saline (0.25 mL/min) during periods of cuff inflation does not by itself alter venous diameter,³¹³² because the venous outflow to the limb is not occluded but simply maintained at constant pressure. Cumulative dose-response curves to norepinephrine (5 to 1280 pmol/min, with each dose increment representing a doubling of the previous dose) were constructed. Each dose was infused for 5 minutes, and doses were increased until a maximal response was achieved (no further constriction despite a doubling in dose). The response to norepinephrine varies between subjects but is consistent and reproducible within a single study in an individual.³³ To limit the duration of the study, four doses of norepinephrine were selected for repeat dose-response curves in a single study; these doses produced 0%, 20% to 40%, 40% to 70%, and 70% to 100% constriction on the first occasion (the doses varied between individuals and for clarity were designated doses A, B, C, and D). Maximal constriction refers to the response to dose D unless otherwise stated. Vein size was recorded in the same place as on the first occasion. Dose-response curves to norepinephrine were constructed before and after local administration of endotoxin.

Single Deep Breath as a Venoconstrictor Stimulus

We induced sympathetic venoconstriction by asking subjects to take a single deep inspiration over a period of 5 seconds, to hold this for a comfortable period (approximately 10 seconds), and then to breathe out slowly before they resumed normal breathing. The transient constriction evoked by this maneuver is blocked by drugs that inhibit the sympathetic nervous system.³⁴ Two adjacent dorsal hand veins were studied simultaneously. One vein was isolated and received endotoxin while the other was left unoccluded. Deep-breath maneuvers were performed before and 1 hour after endotoxin was instilled.

Instillation of Endotoxin

To instill endotoxin or control solution (saline), a length of the vein under study was isolated from the circulation by means of two wedges placed 2 to 3 cm apart on the skin overlying the vessel.³⁵ The wedges were weighted to occlude the inflow and outflow to the isolated segment. Endotoxin (100 EU in 1 mL saline) or saline (1 mL) was injected into the isolated segment, and although we did not assess it formally, there appeared to be no significant leakage (the vessel stayed distended despite deflation of the upper arm cuff). This dose of endotoxin gives a calculated local concentration of about 20 ng/mL, similar to that reported in the blood of patients with severe sepsis.³⁶ One hour later, the contents of the segment were aspirated and the wedges removed so that the circulation of the blood through the vessel was reestablished. At this stage, dose-response curves to norepinephrine were repeated.

Effects of Endotoxin: Time Course (Study 1)

In 12 subjects, the time course of the response to endotoxin was explored. Dose-response curves were constructed to norepinephrine before and at 1 and 2 hours after exposure to endotoxin or saline in 4 subjects, at 3 hours (after endotoxin only) in 4 subjects, and at 4 hours (after endotoxin only) in 4 subjects. In 3 additional subjects, deep breath–induced venoconstriction was studied before and 1 hour after endotoxin. For deep-breath studies, two adjacent veins were compared. One was occluded and received endotoxin while the other was left unoccluded. Deep-breath studies were performed in both veins simultaneously, before and 1 hour after endotoxin. At the end of the study, phentolamine hydrochloride (25 nmol/min for 20 minutes) was infused into the control vein and the deep-breath response repeated.

In all subsequent studies, the response to endotoxin was assessed at 1 hour after endotoxin, and to study the response to drugs that might inhibit or reverse the effects of endotoxin, we selected subjects who in preliminary studies demonstrated a large response to endotoxin (arbitrarily defined as >40% suppression of contraction to norepinephrine).

Effects of Endotoxin: Local or Systemic Effect? (Study 2)

To determine whether endotoxin was producing a local rather than a systemic effect, dose-response curves to norepinephrine were constructed simultaneously in two adjacent veins on the same hand (n=3). One vein was isolated and received endotoxin as before, while the other was left unoccluded. Dose-response curves to norepinephrine were constructed in both veins simultaneously, before and 1 hour after one vein was exposed to endotoxin.

Effects of Repeated Administration of Endotoxin (Study 3)

In five subjects, the effects of daily instillation of endotoxin into a single vein for 3 days were explored. On each day, dose-response curves to norepinephrine were constructed before and 1 hour after endotoxin. To determine whether the effects of repeated exposure to endotoxin initiated a local or systemic response, an adjacent vein on the same hand (ie, a vein that had not been exposed to endotoxin previously) was exposed to endotoxin on day 3, and a dose-response curve to norepinephrine was constructed before and 1 hour after endotoxin.

Effects of Inhibition of Cyclooxygenase and/or NO Synthase on the Response to Endotoxin (Study 4)

Five subjects were given soluble aspirin (1 g, a dose that inhibits the dilation to arachidonic acid³⁷ ), and the response to endotoxin was determined 2 hours later. Dose-response curves were established to norepinephrine before and 1 hour after exposure to endotoxin. In a separate set of studies in five subjects, a dose-response curve to norepinephrine was constructed before and 1 hour after exposure to endotoxin, and at the end of the second dose-response curve, L-NMMA (100 nmol/min for 10 minutes) was coinfused with the same dose of norepinephrine that produced the maximal constriction before endotoxin. The combined effects of aspirin and L-NMMA were also studied. Five subjects were given oral aspirin (1 g) 2 hours before the study, and the response to norepinephrine was determined before and at 1 hour after exposure to endotoxin. In addition, immediately after the second dose-response curve to norepinephrine had been constructed, L-NMMA (100 nmol/min) was coinfused with norepinephrine and the dose-response curve repeated for a third time.

Effects of Hydrocortisone on Response to Endotoxin and to Development of Endotoxin Tolerance (Study 5 and Study 6)

Five subjects received hydrocortisone (100 mg) 2 hours before the study. Dose-response curves to norepinephrine were constructed before and 1 hour after exposure to endotoxin (study 5).

To determine the effects of the glucocorticoid (study 6) on repeated daily dosing of endotoxin (for 3 days, as above), five subjects were given hydrocortisone (100 mg) 2 hours before the study on day 1 and day 2. On day 3, no steroid was given. Dose-response curves to norepinephrine were constructed before and 1 hour after endotoxin.

Drugs

Endotoxin (EC-5, 10 000 EU/vial) was obtained from USP. Vials were stored at −4°C. The endotoxin was rehydrated with 20 mL sterile saline (0.9%) to give a solution of 1000 EU/mL. The vial was then shaken for a minimum of 20 minutes. The solution was divided into aliquots and stored at −20°C for up to 10 weeks. To determine activity of the endotoxin, dose-response curves to endotoxin were constructed at four weekly intervals in J774 murine macrophage cell lines, and nitrite production was determined as described previously.³² All endotoxin was filtered through a 0.8-μm bacterial filter (Acrodisc PF, Gelman Sciences). Dispersible aspirin was obtained from Aspar Pharmaceuticals Ltd, L-NMMA from Welcome Foundation Limited, hydrocortisone (20 mg) from MSD, norepinephrine (2 mg/vial) from Sanofi Winthrop, ascorbic acid (100 mg/mL) from Evans Medical Ltd, heparin (100 U/mL) from CP Pharmaceuticals Ltd, and phentolamine (10 mg/mL) from Ciba. Ascorbic acid was added to norepinephrine stock solutions to prevent auto-oxidation. Heparin (100 U) was added to the endotoxin solution before administration to prevent thrombus formation.

Calculations and Statistics

Vein size was measured in arbitrary units and converted to millimeters after calibration of the transducer at the end of the experiment. The response of the resting vein to drugs is expressed as a percentage reduction in diameter from that measured during infusion of saline alone. Results were compared by use of Student's t test for paired data or ANOVA of the means as appropriate; a value of P<.05 was considered statistically significant.

Results

Local instillation of endotoxin into the isolated vein had no effect on resting vein size and caused no local adverse or systemic effects. The average basal vein size (ID at constant pressure) for each study is shown in the Table. In each subject, norepinephrine produced a dose-dependent venoconstriction (Fig 1) but as is recognized,³³ there was variability in the response between individuals, with the dose that produced maximal constriction ranging from 40 to 1280 pmol/min.

Effects of Endotoxin: Response and Time Course (Study 1)

Local instillation of endotoxin caused a rightward shift in the dose-response curve to norepinephrine (for example, the dose of norepinephrine that produced a 40±3% constriction before endotoxin produced a 6±3% constriction 1 hour after endotoxin) and suppressed the maximal constriction achieved (Fig 2). In seven subjects, the dose of norepinephrine given after endotoxin was increased in an attempt to produce full constriction. In these subjects, maximal constriction (89±11%) was produced in response to 623±64 pmol/min before endotoxin, but 1 hour after endotoxin, the constriction reached a plateau at 52±14%, and even doses up to 2560 pmol/min had no additional effect. Attenuation of the norepinephrine response was seen at 1, 2, and 3 hours after endotoxin, but by 4 hours, the venoconstrictor potency of norepinephrine was fully restored (Fig 3). The maximal constriction to norepinephrine (dose D) before endotoxin was 87±4% and 1 hour later was 52±8% (P<.05). In contrast, the maximal constriction to norepinephrine 4 hours after endotoxin was back to 99±0.2% of the control value (P=NS; Fig 3). The effect was specific for endotoxin, because there was no change in the dose-response curve to norepinephrine at 1 or 2 hours after instillation of saline (maximal constriction: before saline, 77±4%; 1 hour later, 88±5%; 2 hours later, 82±2%; Fig 2).

The potential physiological significance of attenuation of the constrictor response to norepinephrine is illustrated in the three subjects who took part in the study of deep-breath responses. A deep breath produced simultaneous transient venoconstriction in both veins (13±4% constriction; Fig 4). After endotoxin, the constrictor deep-breath response was abolished in the treated (0% constriction) but not in the control vein (11±3% constriction). Infusion of phentolamine (25 nmol/min for 20 minutes) into the control vein abolished the deep-breath response in this vessel (Fig 4), which confirmed that the constriction produced by this maneuver was due to sympathetic nervous system activation and norepinephrine release.

Effects of Endotoxin: Local or Systemic Effect? (Study 2)

In experiments that compared two adjacent veins on the same hand, there was a rightward shift of the norepinephrine dose-response curve in the vein that received endotoxin, and there was suppression of maximal constriction at 1 hour (maximal constriction: before endotoxin, 81±7%; after endotoxin, 26±6%; P<.05). In contrast, in the control vein, there was no change in the response to norepinephrine at the same time points (maximal constriction: time 0, 86±13%; 1 hour after instillation of endotoxin into the adjacent vein, 90±10%; Fig 5).

Effects of Repeated Administration of Endotoxin (Study 3)

Instillation of endotoxin into the same vein on 3 consecutive days resulted in the development of tolerance to the effects of endotoxin, so that by day 3, there was no shift in the dose-response curve to norepinephrine after endotoxin (maximal constriction before endotoxin was 91±6%, 98±2%, and 94±4% on days 1, 2, and 3, respectively; 1 hour after instillation of endotoxin, maximal constriction was 32±11% [P<.001], 67±6% [P<.05], and 85±7% [P=NS], respectively; Fig 6). In contrast, on day 3, the adjacent vein that had not been exposed to norepinephrine previously still demonstrated sensitivity to the effects of endotoxin (maximal constriction: before endotoxin, 94±4%; 1 hour after endotoxin, 39±6%; Fig 6).

Effects of Cyclooxygenase Inhibitors and/or NO Synthase Inhibitors on the Response to Endotoxin (Study 4)

Neither the cyclooxygenase inhibitor aspirin nor the NO synthase inhibitor L-NMMA modified the shift in the norepinephrine response induced by endotoxin (Fig 7).

Coinfusion of L-NMMA for 10 minutes with a dose of norepinephrine (124±66 pmol/min) that produced maximal constriction before endotoxin did not alter vein size (maximal constriction: before endotoxin, 88±11%; after endotoxin, 30±10%; after endotoxin and in the presence of L-NMMA for 10 minutes, 31±10%; P<.05; Fig 7).

Similarly, the combination of aspirin (1 g) taken 2 hours before the study plus coinfusion of L-NMMA with norepinephrine did not affect the endotoxin-induced dose-response shift to norepinephrine (maximal constriction: before endotoxin, 90±5%; 1 hour after endotoxin, 33±5%; 1 hour after endotoxin and in the presence of aspirin and L-NMMA, 27±5%; n=5; P<.05; Fig 7).

Effects of Hydrocortisone on Response to Endotoxin and Development of Tolerance (Study 5 and Study 6)

Oral hydrocortisone (100 mg) taken 2 hours before the study abolished the endotoxin-induced shift in the response to norepinephrine at 1 hour (maximal constriction: before endotoxin, 77±4%; 1 hour after endotoxin, 70±5%; Fig 8).

Hydrocortisone taken 2 hours before each study inhibited the effects of endotoxin on the response to norepinephrine on day 1 and day 2, but on day 3 (in the absence of steroid), endotoxin caused a rightward shift in the dose-response curve to norepinephrine (maximal constriction before endotoxin was 89±5%, 99±1%, and 100±0% on days 1, 2, and 3, respectively; 1 hour after endotoxin, maximal constriction was 99±1%, 97±3%, and 40±6%, respectively; P<.05; Fig 9).

The in vitro endotoxin assay (nitrite production by J774 cells³⁸ ) demonstrated that the endotoxin solution retained full activity throughout the period of the study.

Discussion

Injection of endotoxin into animals or humans causes systemic arterial¹⁵ and venous²⁷ dilation and a fall in blood pressure. These effects are similar to the changes seen in patients with septic shock. The results of the present study provide direct evidence for a local vascular action of endotoxin in human veins in vivo and demonstrate that this effect is suppressed by a glucocorticoid. Using a system in which drugs are given into a single superficial vein in very low doses sufficient only to produce changes in the study vessel, we found that endotoxin attenuates the constrictor response to norepinephrine, causing a shift in the dose-response curve and suppressing the maximal constriction achieved. This effect was greatest 1 hour after exposure to the endotoxin and had waned by 4 hours, by which time the norepinephrine response had fully returned. Pretreatment with oral hydrocortisone abolished the endotoxin-induced hyporesponsiveness to norepinephrine; however, the local vasoactive mediators NO and prostaglandins appeared not to contribute to the changes seen. Repeated exposure to endotoxin induced tolerance.

Under the experimental conditions used, provided subjects are comfortably warm and relaxed, superficial hand veins have no intrinsic tone and cannot be dilated further.³² To see a dilatation, it is necessary to preconstrict the vein, and we did this with norepinephrine (a situation similar to a vessel in an organ bath in vitro). Norepinephrine released from sympathetic nerves is the major determinant of venous tone,³⁹ and our study suggests that systemic administration of endotoxin inhibits contractions induced by neuronally derived norepinephrine and thereby causes venodilation. This suggestion is directly supported by the observation that the venoconstriction produced by activation of the nervous system after a deep breath was taken was abolished by treatment with endotoxin.

The hypotensive and vasodilator effects of endotoxin in healthy volunteers have been reported previously,¹⁵ but it has not been determined whether these effects are due to a systemic reaction to endotoxin or direct effects on the vessels themselves. Indeed, after intravenous injection of endotoxin, a series of systemic effects occurs that includes fever, altered white cell count, increased concentration of cytokines circulating in blood, and activation of coagulation factors.^10151920 In the present study, a segment of the vein was isolated from the circulation, and the endotoxin was instilled locally for 1 hour and then withdrawn, at which time the vessel was opened again to the circulation. The hyporesponsiveness to norepinephrine produced by this maneuver was due to the endotoxin, because instillation of sterile saline had no effect. Several observations support the conclusion that the effect of endotoxin was local rather than systemic: the total dose of endotoxin instilled and then removed was less than the minimum pyrogenic dose⁴⁰ ; subjects experienced no systemic symptoms; and, most conclusively, reactivity of an adjacent, untreated vein on the same hand was unaltered. However, it remains possible that the blood cells trapped within the isolated segment contributed to the response to endotoxin.

Hyporesponsiveness to norepinephrine was evident 1 hour after endotoxin and persisted for at least 3 hours. This rapid effect is consistent with the acute effects of endotoxin seen in certain animal models²⁵ and with the time course of the hemodynamic changes that occur in healthy volunteers after systemic administration of endotoxin.¹⁵ The effects of endotoxin persisted for more than 2 hours after the endotoxin had been removed, at a time when the vessel was no longer exposed to this stimulus. This suggests that exposure to endotoxin triggered longer-lasting pharmacological and biochemical changes in the vessel wall.

The effects of endotoxin could be abolished by prior administration of oral hydrocortisone, which indicates that whatever mediates the response to endotoxin is suppressed by an anti-inflammatory dose of glucocorticoid. Studies in animals and in vitro have demonstrated that exposure to endotoxin leads to the delayed expression of iNOS and COX-II, a process that is inhibited by glucocorticoids.⁴¹⁴² Furthermore, the blocking of the activity of iNOS or COX⁴³ reverses many of the vascular effects of endotoxin, which indicates that the products of these enzymes, NO and prostanoids, contribute to the changes seen. However, in the hand veins, although glucocorticoids were effective, the NO synthase inhibitor L-NMMA (which should inhibit NO synthesis within minutes) and the cyclooxygenase inhibitor aspirin had no effect on the hyporesponsiveness to norepinephrine. Thus, NO or prostanoids are unlikely to contribute to the response to endotoxin that we detected in these veins. This finding is consistent with certain previous studies of the acute response to endotoxin in animals⁴² and a study of the delayed effects of endotoxin on human saphenous veins in vitro.²⁸ In addition, a biochemical study⁴⁴ on human saphenous vein vascular smooth muscle cells suggested that interleukin-1 might elevate the concentration of cGMP independent of NO generation. It is unlikely in the present study that insufficient doses of the blocking agents were used, because we have demonstrated previously that the dose of L-NMMA used in the present study attenuates the response to endothelium-dependent dilators⁴⁵ and that the dose of aspirin used inhibits dilation to arachidonic acid.³⁷ Thus, the mechanism of the change we have observed in the veins remains uncertain, but the possibilities include direct stimulation of guanylate cyclase,⁴⁴ local generation of platelet-aggregating factor, leukotrienes, cytokines, isoprostanes, or other glucocorticoid-suppressible inflammatory mediators.^18192046 Further studies with specific antagonists of these mediators are now needed. It is also important to develop methods to study the direct local actions of endotoxin on the arterial circulation in humans (resistance vessels and conduit vessels).

By 4 hours, the effects of endotoxin had disappeared. This contrasts with the effects seen when endotoxin is administered systemically, in which case the hypotension persists for 8 hours or longer.¹⁵ The reasons for this are not known but might include persistence of endotoxemia after systemic administration, differences between arteries and veins or between vascular beds, or the contribution of systemic rather than direct local effects of endotoxin to the delayed and more long-lasting response to endotoxin. The chronic vascular response to endotoxin seems to be due in large part to expression of iNOS⁴²³⁴³ in the vessel wall, a process that takes longer than 2 to 3 hours and might be due to a systemic response, with infiltration into the vessel wall by activated circulating macrophages.²³ Clearly, such a systemic response would have been absent in the present study, and it now should be possible to dissect which mechanisms are generated locally and which are systemic in origin.

Repeated daily exposure to endotoxin for 1 hour for 3 consecutive days induced tolerance to the effects of endotoxin by the third day. Again, this effect appeared to be a local phenomenon, because at a time when the treated vein was tolerant to endotoxin, an adjacent vein on the same hand was not. Interestingly, tolerance to endotoxin on day 3 did not occur when the biological response (hyporesponsiveness to norepinephrine) on days 1 and 2 had been blocked by prior administration of hydrocortisone. These results suggest that tolerance occurs as a result of a downregulation or depletion of pathways that mediate the response to endotoxin rather than downregulation of the endotoxin receptors. This is consistent with studies performed in vitro on macrophage and monocyte cell lines that demonstrated that tolerance occurred as a result of decreased, endotoxin-induced expression of mRNA for tumor necrosis factor and decreased G protein function,⁴⁷ with no alteration of endotoxin receptors.⁴⁸ Our results do not support the suggestion that systemic elevation of endogenous glucocorticoids is the mechanism of vascular tolerance to endotoxin.⁴⁹ There is interest in the possibility that inducing endotoxin tolerance could be used therapeutically in high-risk patients.⁵⁰ Our results suggest that tolerance to the vascular effects of endotoxin is achievable but indicate that it may not be possible to induce tolerance without first eliciting the biological response.

We have developed a safe and reproducible model that allows the study of the acute local effects of endotoxin on the venous wall in situ. To the best of our knowledge, similar experiments have not been undertaken previously in animals or humans. The model permits exploration of the role of inflammatory mediators and their effects on constrictor or dilator responses and might also be extended to explore the effects of endotoxin on endothelial function, blood cell/vessel wall interactions, or changes in permeability. The finding that endotoxin produces an acute venous hyporesponsiveness to the constrictor effects of norepinephrine or activation of the sympathetic nervous system might account for the changes in cardiac filling pressure that occur in septic shock.¹² This condition is associated with profound venodilatation despite high circulating levels of catecholamines and activation of the sympathetic nervous system (the major determinant of venous tone).⁵¹ The observation that the hyporesponsiveness to norepinephrine is inhibited acutely by glucocorticoids but appears not to involve prostanoids or NO suggests that other unidentified mediators contribute to the overall vascular response to endotoxin. This suggestion is supported by the finding that animals that lack the gene that encodes for iNOS still show a degree of hypotension in response to endotoxin.⁵² Considerable species variability in the response to and effects of endotoxin has been reported,²⁵ and it should now be possible to explore the local mechanisms of action of endotoxin in detail in humans.

Acknowledgment

Dr Bhagat is supported by the British Heart Foundation.

Selected Abbreviations and Acronyms

• Download figure|
• Download PowerPoint

Figure 1. Dose-response curve to norepinephrine; n=7.

• Download figure|
• Download PowerPoint

Figure 2. In four subjects, dose-response curves to norepinephrine (NA in figure) were constructed before and at 1 and 2 hours after the vein was exposed to endotoxin (n=4) or saline (n=4). For the repeat dose-response curves to norepinephrine at 1 and 2 hours after saline or endotoxin, 4 doses of norepinephrine were selected that produced 0%, 20% to 40%, 40% to 70%, and 70% to 100% constriction on the first occasion (designated A, B, C, and D). Left: A, 23±19 pmol/min; B, 55±40 pmol/min; C, 208±148 pmol/min; and D, 425±292 pmol/min. Right: A, 6±2 pmol/min; B, 41±22 pmol/min; C, 93±41 pmol/min; and D, 220±60 pmol/min. ○, before saline or endotoxin; •, 1 hour after saline or endotoxin; ▪, 2 hours after saline or endotoxin.

• Download figure|
• Download PowerPoint

Figure 3. In 12 subjects, the time course of the response to endotoxin was explored. Dose-response curves were constructed to norepinephrine (NA in figure) before and at 1, 2, 3, and 4 hours after exposure to endotoxin (n=4 at each time point). Results are expressed as maximal percentage constriction to norepinephrine relative to the initial control value for that study.

• Download figure|
• Download PowerPoint

Figure 4. In three subjects, sympathetically mediated venoconstriction responses were assessed simultaneously in two adjacent veins. The control vein (top) was left unoccluded while the other vein was isolated and received endotoxin (bottom). Deep-breath (↑) venoconstrictor responses were performed simultaneously in both veins before and 1 hour after instillation of endotoxin. A deep breath produced simultaneous transient venoconstriction in both veins. After endotoxin, the constrictor deep-breath response was abolished (bottom). Finally, in the control (untreated) vein, phentolamine (25 nmol/min for 20 minutes) was infused, and the deep-breath response was assessed again (top).

• Download figure|
• Download PowerPoint

Figure 5. In three subjects, dose-response curves were constructed simultaneously in two adjacent veins. The control vein was left unoccluded while the other vein was isolated and received endotoxin. Dose-response curves to norepinephrine (NA in figure) were constructed in both veins before and 1 hour after one vein was exposed to endotoxin. For the repeat dose-response curves to norepinephrine at 1 and 2 hours after saline or endotoxin, four doses of norepinephrine were selected that produced 0%, 20% to 40%, 40% to 70%, and 70% to 100% constriction on the first occasion (designated A, B, C, and D). Left: A, 4±1 pmol/min; B, 10±0 pmol/min; C, 40±0 pmol/min; and D, 107±27 pmol/min. Right: A, 7±2 pmol/min; B, 30±10 pmol/min; C, 67±14 pmol/min; and D, 160±0 pmol/min. ○, before endotoxin; •, 1 hour after endotoxin.

• Download figure|
• Download PowerPoint

Figure 6. In five subjects, the effects of repeated instillation of endotoxin for 3 days into a single vein were explored. On each day, dose-response curves to norepinephrine (NA in figure) were constructed before and 1 hour after endotoxin. To determine whether the effects of repeated exposure to endotoxin initiated a local or systemic response, on the third day, an adjacent vein on the same hand was exposed to endotoxin and a dose response to norepinephrine was constructed before and 1 hour after endotoxin.

• Download figure|
• Download PowerPoint

Figure 7. The effects of inhibition of cyclooxygenase and/or NO synthase inhibitors on the response to endotoxin were explored. A, Five subjects were given soluble aspirin (1 g), and the response to endotoxin was determined 2 hours later. Dose-response curves to norepinephrine (NA in figure) were established before and 1 hour after exposure to endotoxin. (A, 7±2 pmol/min; B, 40±11 pmol/min; C, 224±113 pmol/min; and D, 464±220 pmol/min.) B, In five subjects, a dose-response curve to norepinephrine was constructed before and 1 hour after endotoxin. At the end of the second dose-response curve, L-NMMA (100 nmol/min for 10 minutes, indicated by long bar) was coinfused with the same dose of norepinephrine that produced the maximal constriction before endotoxin. (A, 7±2 pmol/min; B, 23±7 pmol/min; C, 41±9 pmol/min; and D, 124±66 pmol/min.) C, The combined effects of L-NMMA and aspirin were studied in five subjects. All subjects received oral aspirin (1 g) 2 hours before the study, and the response to norepinephrine was determined before and 1 hour after exposure to endotoxin. In addition, immediately after the second dose-response curve to norepinephrine had been constructed, L-NMMA (100 nmol/min) was coinfused with norepinephrine and the dose-response curve repeated for a third time. (A, 8±1 pmol/min; B, 32±5 pmol/min; C, 80±18 pmol/min; and D, 173±32 pmol/min.) ○, before endotoxin; •, 1 hour after endotoxin; ▪, 1 hour after endotoxin with L-NMMA coinfused.

• Download figure|
• Download PowerPoint

Figure 8. In five subjects, oral hydrocortisone (100 mg) was taken 2 hours before the study and the response to endotoxin determined 2 hours later. Dose-response curves to norepinephrine (NA in figure) were established before and 1 hour after exposure to endotoxin. A, 6±1 pmol/min; B, 28±11 pmol/min; C, 57±22 pmol/min; and D, 133±39 pmol/min. ○, before endotoxin; •, 1 hour after endotoxin.

• Download figure|
• Download PowerPoint

Figure 9. On 2 consecutive days, five subjects were given oral hydrocortisone (100 mg) 2 hours before the study. Dose-response curves to norepinephrine (NA in figure) were constructed before and 1 hour after endotoxin. On day 3, no steroid was given, and the dose-response curve to norepinephrine was established before and 1 hour after endotoxin. Day 1: A=7±1 pmol/min; B=18±2 pmol/min; C=36±4 pmol/min; and D=88±20 pmol/min. Day 2: A=9±1 pmol/min; B=36±11 pmol/min; C=72±23 pmol/min; and D=144±46 pmol/min. Day 3: A=8±1 pmol/min; B=28±5 pmol/min; C=56±10 pmol/min; and D=112±20 pmol/min. ○, before endotoxin; •, 1 hour after endotoxin.

Footnotes