These vasoactive mediators have direct cardiac effects [9, 15-18] or, more frequently, effects on peripheral blood vessels [4, 8, 9, 18, 19]. Although the pattern of peripheral effects may differ markedly in individual patients, common phenomena in anaphylaxis are a decrease in peripheral resistance and an increase in vascular permeability. These phenomena may result in a maldistribution of the intravascular fluid and a plasma loss up to 50% within 10 minutes [19, 20], leading to anaphylactic shock.

A compensatory release of the vasopressors epinephrine, norepinephrine, and angiotensin II has been observed within several minutes after the induction of anaphylactic shock in animal models and in organ models [7, 9, 18, 21]. Angiotensin II may be derived from the conversion of plasma angiotensin I, as well as from activated endothelial cells [22, 23], where it contributes to the local regulation of vascular tone in heart, lungs, and kidneys. A release of catecholamines and angiotensins has also been observed in clinical studies on medication-induced anaphylaxis [9, 20, 24]. Except for one study on the release of catecholamines during the first 2 minutes after insect-sting challenge in children (Hauk and colleagues. Catecholamine release on insect-sting challenge. European Congress on Allergy and Clinical Immunology; Glasgow, May 1990), there are no published prospective studies in humans on the systemic release of cardiovascular mediators during insect-sting anaphylaxis.

The recurrence rate of insect-sting anaphylactic reactions has been reported to range from 20% to 80%. This rate is influenced to a great extent by study design (prospective versus retrospective, references [3, 4, 6, 25]). It has been shown before [3, 6, 25] that most patients with a history of insect-sting anaphylaxis are treated unnecessarily and that in some patients treatment is withheld incorrectly, when the classical selection criteria for venom immunotherapy (high insect-specific immunoglobulin E [IgE] in plasma or positive skin test) are applied. Therefore, in The Netherlands and some other European countries, patients may be selected for venom immunotherapy by sting challenge under intensive-care conditions. This sting-challenge approach offered us the opportunity to study prospectively the recurrence rate of insect-sting anaphylactic reactions as well as the sequential release of catecholamines and angiotensins in relation to the development of hypotension during anaphylaxis.

Methods

Top Methods Results Discussion Author & Article Info References

Patients

One hundred thirty-eight patients with a history of an anaphylactic reaction were subjected to a provocation test with the insect concerned. The insect was identified by the patient's recognition of the doctor's descriptions of the two predominant Hymenoptera in The Netherlands, supplemented by skin tests and specific IgE. Eight healthy volunteers were also challenged, including five persons who had been stung in the past by an insect of the species in question without experiencing anaphylactic symptoms. All patients were in good health; none was pregnant or had a cardiac disorder. None of the patients used a ß-blocker, antihistamine, mast-cell stabilizing agent, or any other drug. Electrocardiogram, chest radiograph, and results of routine blood examination were normal in all participants. The patients were first seen in the outpatient clinic, where detailed extensive oral and written information was given about the study protocol. After giving informed consent, the patients were entered in the study protocol, which had been approved by the medical ethical review board of the Eemland Hospital.

Study Protocol

The provocation procedure with living insects was performed as described previously [3, 6, 14]. In short, all participants were placed on a continuous heart monitoring device in the intensive care unit before the challenge, and an intravenous catheter was inserted in each arm. A yellow jacket of the species Vespula germanica or a honey bee of the species Apis mellifera was induced to sting the lower left arm for a period of 30 seconds. After that the insect was killed. Blood pressure was recorded automatically at 1- to 5-minute intervals in unstable conditions and at 5- to 15-minute intervals in stable conditions. Mean arterial pressure was calculated as two thirds diastolic pressure plus one third systolic pressure.

The severity of the anaphylactic reaction was graded according to Mueller [1]. Briefly, symptoms are classified as described below, in which more severe reactions may be accompanied by symptoms of less severe reactions: 0 = no systemic reaction; I = skin symptoms (generalized urticaria, itching, erythema) or anxiety; II = gastrointestinal symptoms (stomach or chest pain, nausea or vomiting), generalized edema; III = respiratory symptoms (shortness of breath, difficulty in swallowing, hoarseness, or stridor); IV = hypotension defined as a decrease of 15 mm Hg or more in mean arterial pressure from initial value, requiring immediate intervention, with or without other cardiovascular symptoms such as cyanosis, collapse, arrhythmias, or angina pectoris. In this article, we refer to Mueller 0 as no reaction, Mueller I as mild, Mueller II and III as intermediate, and Mueller IV as severe or anaphylactic shock reactions. Only patients with intermediate or severe reactions after challenge were started on venom immunotherapy [26].

Blood Sampling

The interval between the sting and the onset of clinical symptoms may vary considerably, as reported previously [1-4, 6] and as confirmed by our study. Blood was collected before sting challenge ("pre") and at 1, 5, 15, and 60 minutes after the moment at which the patient indicated that a reaction had started. Similar blood sampling was done in participants who did not show an anaphylactic reaction, arbitrarily starting 15 minutes after the sting.

The blood sampling procedure has been described elsewhere [3]. Serial samples were collected from only 69 participants: from the 8 volunteers, the 39 patients with an anaphylactic reaction after provocation, and the first 22 of the 99 patients with no reaction. Because the data of the other 77 nonreacting patients were not expected to provide additional information, postchallenge samples were not collected from these patients. Immediately after collection in lithium heparin tubes, the blood samples were put on ice and centrifuged. Metabisulfite was added, and the plasma samples were deep-frozen at –70°C, all within 15 minutes after collection. All five samples of each individual patient were tested with the same assay procedure.

Assays

Epinephrine, norepinephrine, and dopamine levels were measured with high-performance liquid chromatography system. Quantification was performed after calibration with epinephrine, norepinephrine, and dopamine standards (Sigma, Amsterdam, The Netherlands). Angiotensin I and II levels were measured with a radioimmunoassay (Immuno Technology Service Production B.V., Wychen, The Netherlands), as described previously [28].

Statistical Analysis

Except for the mean arterial pressure, none of the variables appeared to be normally distributed. In such cases, the data were log-transformed before analysis. Paired Student t-tests were used to compare data of the same patients at different time intervals. Unpaired Student t-tests were used to compare data between different patient groups. Correlations were calculated by linear regression analysis and P values with the Pearson product moment correlation coefficients on the combined data of the 39 patients with anaphylactic reactions. A P value < 0.05 was considered to represent a significant difference or correlation. For statistical reasons, the four patients with an intermediate reaction after challenge were excluded from group-wise mediator analysis but not from correlation analysis.

Results

Top Methods Results Discussion Author & Article Info References

Clinical Course

The patient groups were identical in distribution of insect species concerned, grade of previous reaction, age, sex, insect-specific IgE and IgG4, and skin thresholds to venom, as described previously [3]. In Table 1, the severity of the previous anaphylactic reaction was related to the severity of the reaction after the provocation test in 138 patients. No increase in individual severity grades of reaction was observed, as compared with the previous anaphylactic reaction after the field sting, although an increase in severity within the same grade did occasionally occur. In the 39 patients with an anaphylactic reaction after insect-sting challenge, the symptoms started 1 to 40 minutes (median, 10 minutes) after the sting. No significant relation was observed between this time interval and the clinical severity of the reaction.

Two patients developed an acute hypotensive reaction after the sting challenge that was probably due to causes other than acute vasodilation. One of these two patients developed angina pectoris with documented electrocardiographic changes characteristic of ischemia 25 minutes after the sting challenge. Her cardiac complaints were accompanied by hypotension, presumably from acute heart failure. The other patient suffered from a protracted form of anaphylaxis, developing hypotension 35 minutes after the onset of symptoms (itching). This was accompanied by bradycardia and sweating, in all likelihood representing a vasovagal collapse.

The maximum mean differences between mean arterial pressure values before and after the insect-sting challenge were –7.4 mm Hg, –4.7 mm Hg, and –2.4 mm Hg for the volunteers, the nonreacting, and the mildly reacting patients, respectively. In patients with anaphylactic shock, however, the mean arterial pressure decreased from 97 ± 11 mm Hg to 65 ± 19 mm Hg (mean ±SD; 63 ± 21% of initial values; P < 0.001; Figure 1 5 minutes after the first symptoms of anaphylaxis. Individual mean arterial pressure values at 5 minutes for the four groups of subjects are illustrated in Figure 2, and for a representative patient with anaphylactic shock in Figure 3. The use of plasma expanders and epinephrine prevented interpretation of blood pressure values at 15 and 60 minutes.

View larger version (24K): [in this window] [in a new window]   Figure 1. Changes in mean arterial pressure, catecholamines, and angiotensins in relation to time of insect-sting anaphylactic shock. A. Mean arterial pressure values (mm Hg). B. Plasma levels of epinephrine (nmol/L). C. Norepinephrine (nmol/L). D. Angiotensin I (pmol/L). E. Angiotensin II (pmol/L). The values are shown (for mediators on a log-transformed scale) for individual patients with anaphylactic shock following in-hospital challenge with a live yellow-jacket or honey-bee sting. Values are shown before (pre) and at varying intervals after the patient indicated the onset of clinical symptoms. Bars indicate median values. Three of the patients with an anaphylactic shock were excluded from the norepinephrine and epinephrine bars from the time of epinephrine administration onward (one patient at 5 minutes and two at 15 minutes).

View larger version (23K): [in this window] [in a new window]   Figure 2. Individual values of mean arterial pressure, catecholamines and angiotensins in insect-sting anaphylaxis 5 minutes after the onset of clinical symptoms, classified according to the severity of reaction. A. Mean arterial pressure value (mm Hg). B. Plasma levels of epinephrine (nmol/L). C. Norepinephrine (nmol/L). D. Angiotensin I (pmol/L). E. Angiotensin II (pmol/L). For the nonreacting participants, levels are shown at 19 minutes after the sting. Bars indicate median levels of the groups. One patient in the anaphylactic shock (severe reaction) group was excluded from this graph because epinephrine had been administered immediately before this time point.

View larger version (13K): [in this window] [in a new window]   Figure 3. Mean arterial pressure and epinephrine in a representative patient with insect-sting anaphylactic shock. Mean arterial pressure (mm Hg) and levels of epinephrine (nmol/L) are shown before and during anaphylaxis. The x axis indicates the time in minutes during the sting challenge, with 0 as the moment that the patient indicated the onset of the reaction. The arrows indicate the times at which an antihistaminic drug was administered. The bar below the arrows indicates the period that one-half liter of plasma expander was infused.

Heart rate varied within seconds to minutes with marked individual differences. In 10 of the 17 patients with anaphylactic shock, tachycardia (> 25% increase from initial value) was observed immediately before the onset of hypotension, whereas in 5 of these 17 patients, a bradycardia (> 25% decrease from initial value) preceded the onset of hypotension. After this time, heart rates fluctuated due to emotional state and individual therapy.

An antihistaminic drug (Tavegil; Sandoz, Basel, Switzerland; 1 mg/mL intravenously, maximum of 6 mL) was given to all patients with intermediate or severe reactions and to 1 of the 18 patients with mild reactions because of severe itching. Severe reactors also received fluid replacement (Haemaccel; Behringwerke AG, Wardorf, Germany; intravenously, maximum of 2.5 L). Epinephrine (Farmachemie; Haarlem, The Netherlands; 0.1 mg/mL intravenous drip, maximum of 5 mL) was given to 3 of the 17 patients with a persistent severe reaction. Because of our in-hospital setting with constant monitoring from the first anaphylactic symptom, the other 14 patients with anaphylactic shock could be managed with antihistamines and plasma expanders only. All but two patients with anaphylactic reactions recovered fully within 4 hours after the sting challenge. These two patients suffered from sustained erythema and malaise only but did not require additional therapy. All 17 patients with anaphylactic shock were kept overnight in the intensive care unit, but no late reaction occurred. All 38 patients left the hospital in good clinical condition.

Plasma Levels of Catecholamines

Median prechallenge levels of epinephrine and norepinephrine for the four groups of participants combined were 0.3 nmol/L (range, 0.2 to 2.3 nmol/L) and 1.5 nmol/L (range, 0.5 to 6.7 nmol/L), respectively. These levels did not differ among the four groups. Mean epinephrine and norepinephrine levels did not change significantly in the volunteers or in the patients with mild or no reactions. The epinephrine and norepinephrine levels in the three patients in shock who received epinephrine therapy were excluded from the start of this therapy (one patient at 5 minutes, two patients at 15 minutes). Epinephrine and norepinephrine levels in the group of patients with anaphylactic shock rose to 2.5 nmol/L (range, 0.2 to 35.7 nmol/L; P < 0.05; see Figure 1) and 5.9 nmol/L (range, 1.6 to 30.9 nmol/L; P < 0.01; see Figure 1), respectively, within 5 minutes after the onset of clinical symptoms. Peak values of epinephrine and norepinephrine were significantly higher in the patients with a shock reaction than in the patients with a mild reaction (P < 0.001 and P < 0.05, respectively). Individual levels of epinephrine and norepinephrine for the four groups of participants at 5 minutes are illustrated in Figure 2. The course of epinephrine levels in a representative patient is shown in Figure 3. Dopamine levels fluctuated near the detection level of the assay (0.9 nmol/L) in all patients (data not shown).

Plasma Levels of Angiotensins

Median prechallenge levels of angiotensin I and angiotensin II for the combined four groups were 82 pmol/L (range, 16 to 315 pmol/L) and 61 pmol/L (range, 7 to 217 pmol/L), respectively. These levels did not differ significantly among these four groups. Angiotensin I and II levels did not change after the sting challenge in the volunteers or in the patients with mild or no reactions. Angiotensin I levels decreased nonsignificantly to 61 pmol/L (range, 8 to 177; see Figure 1 60 minutes after the onset of symptoms in the patients with anaphylactic shock. Angiotensin II levels in patients with anaphylactic shock rose to 105 pmol/L [range, 11 to 286 pmol/L; P < 0.01; see Figure 1 5 minutes after the onset of anaphylaxis. Individual angiotensin I and angiotensin II levels for the four groups at 5 minutes are shown in Figure 2.

Relation between Mean Arterial Pressure, Catecholamines, and Angiotensins

The change in mean arterial pressure of the combined 39 patients with anaphylactic symptoms after insect-sting challenge inversely correlated with the change in plasma levels of epinephrine, norepinephrine, and angiotensin II (r = 0.60, 0.63, and 0.53, respectively; P < 0.001) 5 minutes after the onset of clinical symptoms. Epinephrine and norepinephrine levels of the patient with anaphylactic shock who had already received epinephrine therapy at 5 minutes were excluded from analysis (Figure 4). Epinephrine and norepinephrine (but not the angiotensin II) levels at 1 minute correlated significantly with the decrease in mean arterial pressure 5 minutes after the onset of clinical symptoms (r = 0.53, P < 0.001; r = 0.76, P< 0.001; and r = 0.24, P > 0.05, respectively). No significant relation was observed between the change in mean arterial pressure and angiotensin I values (r = 0.13).

View larger version (21K): [in this window] [in a new window]   Figure 4. Relation between changes in mean arterial pressure and plasma levels of catecholamines in 39 patients with anaphylactic symptoms after insect-sting challenge. A. Epinephrine (nmol/L) and B. norepinephrine (nmol/L) levels 5 minutes after the onset of anaphylactic symptoms were corrected for individual prechallenge values and related to the change in mean arterial pressure. One patient with anaphylactic shock was excluded from this graph, because epinephrine had been administered immediately before this time point.

Discussion

Top Methods Results Discussion Author & Article Info References

The goal of our study was to establish the occurrence of a recurrent anaphylactic reaction after sting challenge in persons with previous insect-sting anaphylactic reactions, as well as its relation to the release of catecholamines and angiotensins. We found that only 39 of 138 (28%) patients developed a second anaphylactic reaction after sting challenge. A significant release of catecholamines and angiotensins was only observed in patients with anaphylactic shock, and this release was related to the degree of hypotension. The pulmonary symptoms in the three patients with a grade III reaction (see Methods) were not life threatening and disappeared without epinephrine or inhaler therapy. Upper airway obstruction was not seen. Catecholamine levels in these three patients did not change after sting challenge (data not shown). Patients with anaphylactic shock tended to develop symptoms sooner after the sting than the other anaphylactic patients, but this difference was not significant (median, 7; range, 1 to 37 minutes versus median, 12; range, 2 to 40 minutes).

One of the key signs of a life-threatening anaphylactic reaction is acute hypotension. Anaphylactic shock is mediated by the effects of mast-cell or basophil products, such as histamine, on the heart itself and on peripheral blood vessels [3, 4, 7-9]. Acute cardiac shock may be mediated by the direct effect of histamine on the myocardium itself [9, 17, 18], as well as by histamine-induced coronary spasm [9, 15, 16, 28, 30]. Peripheral effects of histamine include endothelial damage, which may induce dilation of arterioles and postcapillary venules, opening of intercellular gaps, and increase in vascular permeability [4, 8, 9, 18, 19]. The ensuing decrease in total peripheral resistance and plasma loss up to 50% within 10 minutes induce a distributive type of shock [9, 19, 20]. In a recent study [3], we showed that the release of mast-cell histamine and tryptase, but not of prostaglandin D2, correlated with the drop in blood pressure in insect-sting anaphylaxis. This supports the role of mast-cell mediators in the pathogenesis of anaphylactic shock. Other mast-cell mediators (platelet activating factor, leukotrienes, other prostaglandins [10, 11]) and possibly also endothelial factors [12, 13], or factors from the complement [14], the kinin [9], and the fibrinolytic systems (van der Linden and colleagues. Controlled insect-sting challenge in 55 patients: relation between activation of plasminogen and the development of anaphylactic shock; in preparation) may amplify the hypotensive effects of histamine.

Most of the data on the release of catecholamines and angiotensins during anaphylaxis have been obtained in animal studies and clinical case histories. In our prospective study on 138 patients and 8 volunteers, we observed a rise in epinephrine, norepinephrine, and angiotensin II levels that was related to a sudden drop in blood pressure after sting challenge (P < 0.001; see Figure 4. Actually, mediator levels may have been higher because we did not correct for vascular leakage or plasma dilution caused by the use of plasma expanders. The changes in mean arterial pressure in our patients may be influenced by the fact that therapy was instituted when the mean arterial pressure dropped more than 15 mm Hg. Schellenberg and colleagues [31] showed a rise in norepinephrine but not in epinephrine when histamine was infused in eight volunteers until mean arterial pressure values dropped to less than 15 mm Hg from initial values, comparable to our mildly reacting patients (see Methods). Possibly, mild hypotension induces norepinephrine release only, whereas severe hypotension triggers the release of both norepinephrine and epinephrine. Hauk and colleagues did not observe a correlation between epinephrine or dopamine levels and the severity of the reaction at 1 or 2 minutes after an insect-sting challenge in children (Catecholamine release on insect-sting challenge. European Congress on Allergy and Clinical Immunology; Glasgow; May 1990). These authors did not sample blood after the first 2 minutes after a sting, whereas we observed the highest levels of norepinephrine and epinephrine 5 minutes after the onset of anaphylactic symptoms (median of 15 minutes after the sting). Therefore, it is possible that Hauk and colleagues have missed the changes in catecholamine levels in their patients.

We failed to observe a common course of heart-rate changes during anaphylactic shock, possibly due to the intricate balance between the patient's emotions, the many vasoactive mediators and target organs involved, the different compensatory responses in blood pressure regulation in individual patients, and the administered therapy. Immediately before the development of hypotension, 5 of the 17 patients developed bradycardia, probably reflecting a vasovagal effect (fear) or cardiac depression. In 10 of these 17 patients, tachycardia preceded the onset of hypotension. This may have been caused by the combined effects of anxiety and momentarily released histamine [3, 8, 9], C3a [14], and catecholamines [21, 31, 32]. No differences in catecholamine levels were observed between the initially bradycardiac or tachycardiac patients. Thus, we observed a rise in plasma levels of endogenous epinephrine and norepinephrine that correlated with a drop in blood pressure in patients with anaphylactic symptoms after Hymenoptera sting challenge.

In addition to endogenous norepinephrine and epinephrine, blood pressure maintenance in insect-sting anaphylactic shock may be regulated by a dopamine system [33] and by the renin-angiotensin system [7, 9, 34, 35]. Dopamine is a potent vascular agent and a precursor of norepinephrine and epinephrine [33], but the role of circulating dopamine in blood pressure regulation is not known. Dopamine levels in this study varied along the detection level of 0.9 nmol/L (data not shown), although the assay detected significant levels of dopamine after dopamine infusions in other patients (van der Linden PWG, unpublished results). Therefore, we favor the hypothesis that circulating endogenous dopamine is not involved in maintenance of blood pressure.

Angiotensin I levels in the patients with shock reactions decreased nonsignificantly to 74% of initial values 60 minutes after the first symptoms (see Figure 1), probably reflecting leakage through endothelial gaps, dilution by plasma expanders, and conversion by angiotensin converting enzyme. Angiotensin II levels in the patients with a cardiovascular reaction rose in relation to the drop in blood pressure [r = 0.53, P < 0.001]. This rise in angiotensin II may be due to direct release of endothelial-derived angiotensin II [22] or of angiotensin converting enzyme [23]. In our study, prechallenge levels of angiotensin I and II did not predict the outcome of insect-sting anaphylaxis, as suggested by Hermann and colleagues in a retrospective study [35]. Thus, the norepinephrine and epinephrine system as well as and the angiotensin system may contribute to maintenance of blood pressure in insect-sting anaphylactic patients. These phenomena may explain why patients on ß-blockers [36] and perhaps angiotensin converting enzyme inhibitors as well [37] are prone to develop cardiovascular anaphylactic reactions after an insect sting that are often difficult to treat.

In conclusion, we have shown that only 28% of subjects with a previous anaphylactic reaction to an in-sect sting developed a recurrent anaphylactic reaction after a sting challenge. During this recurrent reaction, plasma levels of endogenous epinephrine, norepinephrine, and angiotensin II rose in relation to hypotension.