INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In anaphylactic shock, vascular collapse results when mediators with vasoactive properties are released from basophils and mast cells during allergen challenge (1). Of the mediators released, the autacoids, which include histamine, prostaglandins, and leukotrienes, have assumed important etiological roles in contributing to cardiovascular collapse in anaphylactic shock (1, 4). Histamine may act through H1, H2, and H3 receptors to promote shock during allergen challenge (5, 7). H1 receptors mediate coronary vasoconstriction and cardiac depression (5, 11), whereas H2 receptor agonists produce coronary and systemic vasodilation as well as increases in heart rate (HR) and ventricular contractility (11). On the other hand, histamine H3 receptors have recently been identified on presynaptic terminals of sympathetic effector nerves that innervate the heart and systemic vasculature (7, 9). These receptors have been found to inhibit endogenous norepinephrine release from the sympathetic nerves (7). H3 receptor activation would therefore be expected to accentuate the degree of shock observed during antigen challenge since compensatory neural adrenergic stimulation would be blocked.

The eicosanoids that encompass the prostaglandins and leukotrienes may also contribute to the cardiovascular collapse observed during allergen challenge (4, 12). Cyclooxygenase pathway products, such as prostaglandin (PG)I₂, PGE₁, and PGA₁ cause systemic vasodilation and positive inotropy (14), whereas other products such as thromboxane A₂ (TXA₂) and PGD₂ are vasoconstrictors and cause cardiac systolic dysfunction either by direct or indirect effects (14). Alternatively, the lipoxygenase pathway yields mainly systemic vasoconstrictors of which leukotriene (LT)D₄, LTC₄, and LTE₄ have been purported to cause cardiac depression in many experimental preparations (4).

In an initial study, we showed that left ventricular (LV) contractility was decreased in a canine model of anaphylaxis and that this decrease could be ascribed to the release of mediators of anaphylaxis (15). In a subsequent study, we showed that pretreatment with the histamine H1 blocker chlorpheniramine maleate or the prostaglandin inhibitor indomethacin attenuated the cardiovascular collapse observed in this model (16). A constant dose of intravenous antigen challenge was given with each treatment, and both of these treatments appeared to block mast cell release of mediators as compared with a nontreatment protocol, although there appeared to be an additional receptor blocking effect on the cardiovascular system with histamine H1 blockade (8). Histamine H1 activation has been shown to cause cardiac depression in many experimental models (5, 8), and in our subsequent experiment, cardiac output () was higher under H1 receptor blockade as compared with the control study, for similar mean pulmonary wedge pressures () between studies (16).

However, from that study, it was not clear whether H1 blockade with chlorpheniramine maleate or cyclooxygenase inhibition with indomethacin would still be beneficial, if the antigen dose were increased, such that it was now sufficient to induce mast cell release of mediators. Moreover, in the latter study (16), pretreatment with other drugs, which included a histamine H2 receptor blocker, a histamine H3 blocker, and a leukotriene pathway inhibitor, did not improve hemodynamics in this model. Nevertheless, measurements of cardiac mechanics were not obtained, which makes the changes in LV performance observed during anaphylaxis difficult to interpret. Any beneficial effect of pretreatment with these agents on LV contractility may be difficult to analyze because of the concomitant decreases in LV preload and afterload that occur during allergen challenge (15, 16).

In the present study, we examined LV mechanics in this ragweed model and determined the effects of pretreatment with histamine H1, H2, H3 receptor blockers, and cyclooxygenase and leukotriene pathway inhibitors. A chronically instrumented preparation was used in which LV volumes were determined by sonomicrometry. The dose of allergen was varied to produce similar degrees of shock between treatments. The primary objective was to determine whether any of these treatments was useful in preventing the depression in LV contractility known to occur in this model.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Anaphylaxis Model

This experiment was approved by the University Animal Care Committee, and the investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, 1996). Our canine model of ragweed anaphylaxis has previously been described (15). Newborn dogs receive their first dose of antigen (0.5 mg ragweed pollen extract) mixed with 30 mg Al(OH)₃ intraperitoneally within 24 h after birth. Injections consisting of the same dose are repeated at weekly intervals for 8 wk, at biweekly intervals for approximately 30 wk, and then monthly to maintain hypersensitivity. This regimen results in mean immunoglobulin (Ig)E anti-ragweed antibody titers of > 256 dilutions when measured approximately 2 wk prior to study by passive cutaneous anaphylaxis. Nonsensitized littermate control animals show no IgE titers by this test. Sensitized animals show bronchoconstriction in response to inhalation challenge and cardiovascular collapse in response to intravenous challenge. In the present study, the animals were examined at approximately 1 yr of age when they were large enough for implantation of the ultrasonic crystal transducers.

Animal Preparation

This preparation has previously been described and will only be briefly detailed here (15). Approximately 2 mo before anaphylactic measurements were obtained, sensitized dogs were anesthetized with pentobarbital sodium (30/kg) and were ventilated with a constant-volume ventilator (tidal volume, 12 ml/kg) in which the inspired oxygen concentration was 50%. Under sterile conditions, a left thoracotomy was performed and the pericardium was opened. Three pairs of subendocardial ultrasonic crystal transducers (piezoelectric crystals; Channel Industries, Santa Barbara, CA) were placed along the orthogonal axes of the LV. Subendocardial crystal transducers were implanted along the anterior-posterior (AP) and septal-lateral (SL) minor axes and apex-base (AB) major axis. The crystal transducers were later attached to a sonomicrometer (Triton Technology, San Diego, CA) to record LV dimensions.

Upon closure, the pericardium was loosely reapproximated and the chest closed without drainage. Thus, this preparation is essentially an open pericardial preparation. All wires were tunneled subcutaneously and brought out the back of the neck. A collar was worn by the animals to protect the wires until the time of study. Preoperative and postoperative antibiotics (cloxacillin and gentamicin) were given. At the time of study, all animals were fully recovered and healthy.

Measurements of LV mechanics were obtained approximately 1 mo after crystal implantation. Baseline measurements (see PROTOCOL) were obtained approximately 1 h after pentobarbital anesthesia (30 mg/kg) when pentobarbital concentrations were approxmately 20 to 25 mg/L (20). At this concentration, Unruh and coworkers (20) showed that there was little effect of this anesthesia on LV mechanics. The animal was ventilated (tidal volume, 12 ml/kg) with the rate adjusted to maintain pH at approximately 7.3 to 7.4. Supplemental oxygen was given to maintain arterial PO₂ > 100 mm Hg throughout the study.

A high-fidelity transducer tipped catheter (Millar Instruments, Houston, TX) was advanced into the LV through a carotid artery incision and was used for measuring the LV end-diastolic (LVEDP) and LV systolic pressures. These pressures were correlated with simultaneous LV dimension tracings to define end-diastolic (LVEDD) and end-ejection dimensions (LVEED), respectively. LV end-ejection dimension was defined by maximal negative pressure decline (dP/dt max) (21). LVEDP was defined as the point in the cardiac cycle where the rate of change in LV pressure increased by 150 mm Hg/s, with increase sustained for  50 ms (21). LVEDP was related to the simultaneous LV dimension tracing to define end-diastolic dimension (21).

Through a femoral vein, a Fogarty catheter was inserted into the inferior vena cava. When the balloon on the catheter was inflated, venous return would be reduced to obtain multiple LV stroke work versus LV end-diastolic volume (LVEDV) coordinates (see DATA ANALYSIS). Mean arterial pressure () was measured with a polyethylene catheter inserted into the femoral artery. Through the right jugular vein, a thermodilution Swan-Ganz catheter was advanced into the pulmonary artery to measure mean pulmonary arterial pressure (), , mean right atrial pressure (), and . Catheters were connected to respective transducers (Cobe Industries, CO) and were referenced relative to the left atrium. All signals were displayed on an eight-channel recorder (Astra-Med Recorder, Warwick, RI). Measurements were obtained at end-expiration, so that respiratory variation would not affect the results. Stroke volume (SV) was calculated from /HR, while systemic vascular resistance (SVR) was calculated from (  )/.

Plasma mediators of anaphylaxis were also measured during pre-shock and shock conditions immediately after measurements of LV mechanics were obtained. Plasma mediators of anaphylaxis included determinations of histamine, leukotrienes (LTE₄), as well as the stable breakdown products of prostacyclin (PGI₂) and TXA₂ (i.e., 6-keto- PGF₁ and thromboxane B₂ [TXB₂], respectively). Concentrations of mediators were measured by radioimmunoassay techniques as previously described (18). For all mediators, samples were obtained in duplicate and stored at 70° C until analyzed.

Protocol

Six separate studies (control, n = 9; and five drug studies) were performed approximately 3 wk apart in randomized design. Previous experiments have indicated that the anaphylactic response is stable when experiments are performed over this 3-wk interval (18). The drug studies included an H1 blocker study (n = 6) in which 10 mg/kg of chlorpheniramine maleate was intravenously infused prior to challenge (22), an H2 blocker study (n = 6) in which raniditine HCl (20 mg/kg) was administered (17), an H3 blocker study (n = 7) in which the specific H3 blocker thioperamide maleate (1 mg/kg) was given (9), a cyclooxygenase inhibition study (7) in which pathway inhibition was produced by pretreatment with indomethacin (2 mg/kg) (23), and a lipoxygenase pathway inhibition study (n = 5) in which pathway inhibition was produced by the 5-lipoxygenase activating protein (FLAP) antagonist MK-0591 (24).

Except for MK-0591, the drugs were administered intravenously over a 20-min period and the doses denoted were based on previous reported findings. MK-0591 was administered initially as a 2 mg/kg intravenous bolus and then as 8 µg/kg/min constant intravenous infusion for the remainder of the experiment. MK-0591 was kindly supplied by Dr. A. W. Ford-Hutchinson of Merck Frosst Canada, Inc.

In the nontreatment study and the five drug studies, the animals were studied at baseline, 1 h after drug treatment, and during allergen challenge. Approximately 45 min after the preparation was completed during which the animal was stable, baseline (i.e., preshock) measurements were determined. If needed, the animal was given intravenous volume (6% hetastarch in normal saline solution) to bring LVEDP to approximately 10 to 15 mm Hg. Then, measurements were repeated 1 h postbaseline: for the first 20 min of this 1-h interval, normal saline (250 ml) was administered in the nontreatment study; in the drug studies, the treatment mixed in normal saline solution was given (except for MK-0591, which was administered as described previously). Then, intravenous ragweed was administered through a central intravenous line (shock condition). The antigen dose was progressively doubled (starting from 100 µg) until shock was ascertained. Shock was defined as a decrease in of approximately 50% from that measured preshock. During shock, measurements were obtained at the point of lowest when a plateau in had occurred. As determined in previous studies, this plateau lasts for approximately 10 to 15 min (15, 18). After completion of each of the six studies, vascular accesses were surgically closed, and the animal was returned to its cage.

In the protocol, although the animals were randomized between treatment studies, not all dogs could be examined with each treatment. This occurred because of equipment failure in some animals, and also because some dogs participated in another study. However, a nontreatment study was performed in all animals.

Moreover, to examine the effect of time on our hemodynamic measurements, we performed a separate sham shock protocol (n = 7) in which hemodynamic variables were obtained under conditions of the different blocking agent when only the diluent for the ragweed allergen (normal saline) was given after the respective drugs were administered. Measurements were obtained at baseline, drug administration, and sham-anaphylaxis.

Data Analysis

To evaluate LV mechanics during anaphylaxis, we performed the analysis in terms of preload recruitable stroke work (SW) (25). In a given study, we matched SW at comparable LVEDV between conditions to determine whether LV systolic function was depressed during shock. SW was calculated from [(mean LV ejection pressure  LVEDP) × SV] as described by Glower and coworkers (25). SV was derived from (LVEDV  LVEEV), where LV end-diastolic volume (LVEDV) and LV end-ejection volume (LVEEV) were calculated as described by Sodums and coworkers (26) who used a similar preparation. Volume was calculated with the assumption that LV was a general ellipsoid and that the three dimensions formed the principal axes of this structure. The following equation was used in which the dimension (D) of interest (i.e., end-ejection dimension [EED], or end-diastolic dimension [EDD]) was inserted into the equation: volume = /6 × DAP × DSL × DAB. We previously showed that SV measured by thermodilution techniques correlated well with that calculated by sonomicrometric techniques (27).

The measurement protocol was as follows. In each of the three conditions (i.e., baseline, placebo/treatment, and shock), hemodynamic and LV mechanic measurements were initially obtained without preload reduction by occlusion with the Fogarty catheter (i.e., static measurements). Hemodynamic measurements included , , SV (by thermodilution), HR, , SVR, , and , whereas the LV mechanic measurements included LVEDV, LVEDP, LVEEV, and SW (calculated by LV mechanics). Arterial blood samples were also obtained for mediator analysis. In each of preshock conditions, static measurements were followed by LV mechanic measurements in which the balloon on the Fogarty catheter was inflated. In this way, multiple SW versus LVEDV coordinates could be obtained by a reduction in LV preload. During shock, however, a full range of SW versus LVEDV coordinates could not be determined because LV systolic pressures dropped to such a low value. It was therefore possible to obtain only one LVEDV versus SW coordinate during shock. The objective of this protocol was to match the LVEDV obtained during anaphylaxis with comparable LVEDV obtained during the preshock conditions. In this way, SW could be compared at similar LVEDV between conditions in a given study.

Statistical analysis included one-way analysis for repeated measures (ANOVA1R; NWA Statpak, Portland, OR) and Student-Newman-Keuls (SNK) multiple comparison test when variables were compared in a single study. When treatments were compared between studies, the change between placebo/treatment and shock was compared by ANOVA1R and SNK. Results are reported as mean ± 1 SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the control study, the dose of antigen required to produce shock ranged from 0.1 to 80 mg and averaged 24 ± 9 mg. In Study_{H1 blocker}, the dose of antigen required to produce shock approximately doubled (53 ± 20 mg, p < 0.05) as compared with the control study. In the other studies, the doses were not significantly changed as compared with the control study: in Study_{H2 blocker}, the mean dose was 33 ± 7.5 mg; in Study_{H3 blocker}, the dose was 24 ± 9 mg; in Study_cyclo, the dose was 32 ± 7.5 mg; and in Study_lipox the dose was 30 ± 5.9 mg.

In Figure 1, there were no differences in baseline, placebo/ treatment, or shock values of , , and measured between treatments. These parameters decreased approximately 50% during shock as compared with preshock conditions. Findings in SVR, , and were also similar between treatments (Table 1). On the other hand, in Study_{H3 blocker}, HR significantly increased during shock, whereas there was no change found with the other treatments (see Figure 1, lower right hand panel ).

View larger version (31K): [in this window] [in a new window]   Figure 1.   Pa, , Ppaw, and HR are shown in the different studies at baseline, placebo/treatment, and shock. These measurements were obtained without occlusion by the Fogarty catheter. Pa, , and Ppaw decreased during shock compared with the preshock values (p < 0.05 by one-way ANOVA and SNK). HR in the H3 blocker study increased during shock compared with the preshock values (*p < 0.05), whereas there were no changes in HR in the other studies. The findings observed in HR between treatment and shock in StudyH3 blocker were statistically different from those measured in the other studies by ANOVA and SNK (+p < 0.05 versus other studies). In the different treatments, n = 9 in the control study; n = 6 in the H1 blocker study; n = 7 in the cyclooxygyenase inhibition study; n = 6 in the H2 blocker study; n = 5 in the lipoxygenase study; and n = 7 in the H3 blocker study.

In all studies, LVEDV measured during shock decreased approximately 50% as compared with the preshock conditions. This decrease in LVEDV was accompanied by corresponding reductions in SV (by thermodilution), SW (by LV mechanics), and LVEDP (Figure 2). Of note, in Study_{H3 blocker}, the reduction in LVEDV tended to be greater, whereas the reduction in SW tended to be less than corresponding values found in the other studies. When SW was normalized for a constant LVEDV measured between conditions, the findings in SW in Study_{H3 blocker} were different from those found in the other treatments. As can be observed from Figure 3, SW did not decrease in Study_{H3 blocker}, whereas it decreased with the other treatments.

View larger version (31K): [in this window] [in a new window]   Figure 2.   LVEDP, LVEDV, SV (by thermodilution), and SW (by LV mechanics) are shown in the different studies at baseline, placebo/treatment, and shock. Numbers of experiments are as described in Figure 1. These measurements were obtained without occlusion by the Fogarty catheter. By one-way repeated measures ANOVA and SNK, all variables decreased during shock (p < 0.05) compared with preshock values.

View larger version (21K): [in this window] [in a new window]   Figure 3.   In the treatment studies, SW was normalized for matched LVEDV in the three conditions. Number of experiments are as described in Figure 1. In all studies except for StudyH3 blocker, SW decreased during shock compared with the preshock value. The findings in StudyH3 blocker were statistically different from those in the other studies by ANOVA and SNK (+p < 0.05 versus other studies).

In the sham-shock protocol, there was no change in when placebo was administered with any of the treatments. In the control study, measured 152 ± 13 mm Hg at baseline, 148 ± 14 mm Hg during placebo treatment, and 150 ± 20 mm Hg during placebo shock.

The mediators of anaphylaxis found in the different treatment studies are shown in Figure 4. Baseline concentrations of mediators were unchanged between studies. In the H1 blocker study, the increase in histamine during challenge was higher than that found in the other studies but did not reach statistical significance between treatments. In Study_cyclo, TXB₂ and 6-keto-PGF₁ did not increase during shock, whereas the increase in LTE₄ was higher as compared with the other studies. In Study_lipox, the increase in LTE₄ observed during challenge was eliminated.

View larger version (27K): [in this window] [in a new window]   Figure 4.   Mediators are shown in the different studies at baseline, placebo/treatment, and shock. Numbers of experiments are as described in Figure 1. In the histamine measurements, values determined during shock were different from preshock values (p < 0.05) in all studies by ANOVA and SNK. For the other mediators, significance between shock and preshock is given by *p < 0.05, shock versus baseline is given by +p < 0.05. Significant findings observed between placebo/treatment and shock in the different studies are given by #p < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The results obtained in the present study extend our understanding of the changes in LV mechanics observed during anaphylaxis in this model. In an initial study, we showed that LV contractility is depressed during ragweed challenge, and that this effect could be ascribed to the release of mediators of anaphylaxis (15). In a second study, we administered a constant dose of antigen under conditions of the five pretreatments used in the present study (16). The results showed that pretreatment with the H1 receptor blocking agent chlorpheniramine maleate and the cyclooxygenase inhibitor indomethacin attenuated the cardiovascular collapse found in this model. The present study shows that when the dose of allergen was increased to the extent that hypotension developed during challenge, there were no effects in Study_{H1 blocker} and Study_cyclo of pretreatment on hemodynamics or on left ventricular mechanics as compared with the control study.

On the other hand, histamine H3 blockade appeared to have a beneficial effect on cardiovascular collapse in our model. In Study_{H3 blocker}, HR increased during shock as compared with the other treatments. Moreover, as determined by SW, the depression in LV contractility observed during challenge was eliminated in Study_{H3 blocker}. When adjusted for LVEDV, SW was unchanged between pre- and postshock conditions during H3 receptor blockade. This finding could not be appreciated in the static (i.e., non-Fogarty occluded) measurements because of the decreases in preload and afterload that are found during allergen challenge in this model. As shown in Figure 2, of all the treatments, LVEDV in Study_{H3 blocker} appeared to decrease the most during challenge. This effect would have obscured any beneficial finding of the drug on our measurement of LV contractility.

In the cardiovascular system, evidence has accumulated that histamine H3 activation may be important in the modulation of sympathetic neurogenic activity in diseased conditions (7, 9, 28). H3 receptors function as inhibitory presynaptic receptors (heteroceptors) that modulate norepinephrine release from adrenergic nerve endings. Imamura and coworkers (28) reported that, in the heart, H3 receptors caused a decrease in norepinephrine release in pathophysiological conditions associated with enhanced adrenergic activity, such as myocardial ischemia. Under normal adrenergic activity, histamine release from the heart and other organs appears to be too low to cause H3 activation. However, in pathological conditions such as myocardial ischemia and sepsis, the results have indicated that histamine H3 activation may be observed (28, 30).

Only very small concentrations of histamine are required to cause histamine H3 activation in the heart. We previously showed that in an isolated ventricular trabecular preparation, this concentration ranged between 10⁷ to 10⁹ M (30). On the other hand, the concentration of histamine required to cause histamine H1 and H2 activation appears to be larger and is in the range of 10⁵ to 10³ M (30). In the present study, during allergen challenge, plasma concentrations of histamine in the control study increased from near negligible concentrations to approximately 50 nM. This concentration would be sufficient to cause histamine H3 activation in our model.

Thus, the lack of the HR response observed during anaphylaxis in the control study was at least in part related to histamine H3 activation, because H3 receptor blockade in Study_{H3 blocker} resulted in an increase in HR during anaphylactic shock. Moreover, it appears that an increase in adrenergic neural stimulation during shock was sufficient to maintain LV contractility during challenge. As shown in Figure 3, a depression in SW otherwise found during allergen challenge was eliminated by H3 receptor blockade in Study_{H3 blocker}.

Nevertheless, the mechanism by which LV depression initially occurs in our model is not totally resolved, because the neural adrenergic activation that was improved by H3 blockade must have been a compensatory mechanism to restore LV function toward the preshock value. In our previous study (16), we found that when pretreatment was undertaken with chlorpheniramine maleate, the cardiovascular collapse observed in this model was attenuated. In that study (16), there was some release of histamine during H1 receptor blockade, although the other mediators were not released. Because, for a given , under H1 blockade was higher than that measured in the control study (16), we hypothesized that despite the release of histamine under H1 blockade, there was peripheral effect of the H1 blocker on the heart that prevented the cardiac depression ascribed to H1 activation in previous studies (5, 8).

Yet, in the present study, chlorpheniramine maleate did not attenuate hemodynamic collapse as compared with the control study. In Study_{H1 blocker}, the dose of antigen was progressively increased until the mast cell release of mediators and consequently shock occurred. The dose of antigen required to produce shock in Study_{H1 blocker} was significantly higher than that found in the other studies. Ting and coworkers (31) found that when the antihistamine hydroxyzine was given orally for 3 d, there was an inhibitory effect on antigen-induced mast cell release in ragweed-sensitized individuals who were challenged by intradermal skin reaction. They suggested that hydroxyzine exerted a direct protective action against mast cell activation. In the present study, it was necessary to double the amount of antigen given in Study_{H1 blocker} as compared with the control study to produce shock. If anything, however, when shock was eventually produced in Study_{H1 blocker}, the quantity of histamine release appeared greater than that found in the other studies. Besides an increase in release, moreover, an additional mechanism for the elevation in blood histamine concentration found during shock in Study_{H1 blocker} may reflect the effect of the anti-H1 receptor blocker on interfering with histamine metabolism in this model. Our results do not distinguish between the relative effects of these two possibilities.

In Study_{H1 blocker}, it appears that enough histamine was released during allergen challenge to overcome any peripheral blockade of the H1 receptor, so that any peripheral blocking effect of chlorpheniramine maleate on the H1 receptor was not apparent. A similar finding was observed by Silverman and coworkers (22) in an Ascaris suum model of anaphylaxis. Thus, our results indicate that under conditions of H1 receptor blockade, if sufficient allergen is administered, then H1 receptor blockade may be overcome. However, under conditions of H3 receptor blockade, the amount of histamine released during shock was not sufficient to overcome receptor H3 blockade. This may be due to tighter binding between the H3 blocker and H3 receptor as compared with the H1 blocker and H1 receptor (32) or because somewhat less histamine was released during challenge in Study_{H3 blocker} than in Study_{H1 blocker}.

In both our previous (16) and present anaphylaxis study, we found that under conditions of H3 receptor blockade, SVR was unchanged as compared with the control study. In contrast, in a guinea pig preparation, McLeod and coworkers (9) found that activation of peripheral H3 receptors may cause a decrease in tone in arterial resistance vessels. In Study_{H3 blocker}, we expected that SVR would increase as compared with the control study, because there would be more neural sympathetic activity in Study_{H3 blocker}. In a canine model of sepsis (30), we also found that under conditions of H3 receptor blockade, LV contractility increased, whereas SVR remained unchanged compared with a control study. Thus, it appears that histamine H3 receptors do not play an important role in modulating arteriolar resistance in canine models of anaphylaxis or sepsis.

In Study_cyclo, there was no change in cardiovascular variables as compared with the control study, even though the release of thromboxane and prostacyclin was eliminated during shock. Thus, in contrast to our previous study (16), the present study shows that when enough allergen was administered to produce shock, there was no effect on indomethacin pretreatment on the anaphylactic response in our model. Moreover, in Study_lipox, there was also no change in cardiovascular variables as compared with the control study, even though the release of LTE₄ was eliminated during shock. Although the leukotrienes, prostaglandins, and thromboxanes have been shown to be important contributors to cardiovascular dysfunction in other models of anaphylaxis (4, 13, 14) there were little effects of these mediators on the modulation of hemodynamics in this canine model.

In the present study, hemodynamics were examined while the animals were anesthetized. It would not be possible to study these animals in the conscious condition. Of the anesthetic agents available, all agents would affect some aspect of the anaphylactic shock in our model, and because pentobarbital anesthesia is often used in investigating hemodynamics in allergic models (22, 33), this anesthesia was administered in the present study. A previous study showed that plasma concentrations of pentobarbital would remain within a relatively constant range over the course of the experimental interval (20). In the present experiment, because all control and treatment studies were performed on different occasions with identical anesthesia, there would be constancy of anesthesia effect between studies. Nevertheless, pentobarbital anesthesia is known to increase HR and blood pressure, which may be related to the vagolytic effect of the agent, although baroceptor reflex effects cannot be excluded (20, 34). Because it is not possible to exclude an effect of anesthesia on our results, it is recognized that the conclusions of this study must be cautiously applied to the human condition.

Among the many mediators that are released in anaphylaxis, histamine, leukotrienes, and prostaglandins have assumed an important role in causing the cardiovascular collapse observed. We recognize that many other mediators are released during challenge that may also contribute to this collapse. Based on the findings in the literature, we concentrated on those mediators that were likely to be important in mediating cardiovascular collapse in anaphylaxis (12). For instance, platelet-activating factor is grouped with the autacoids, but appears to cause damage by modulation of other classes of mediators, and not by a primary process (12). Furthermore, it is important to note that the present study looks at pretreatment with various drugs in anaphylaxis, and that the results might not be applicable to treatment of anaphylaxis once shock has occurred.

Our results indicate that during shock, under conditions of H3 receptor blockade, LV systolic function and HR were higher as compared with a nontreatment study. It is recognized that the application of animal models to the human condition must be interpreted cautiously. The present study shows that histamine H3 activation by inhibiting adrenergic neural norepinephrine release contributes to cardiovascular dysfunction in anaphylactic shock.