Side-effects of insect venom immunotherapy:results from an EAACI multicenter study

Authors

  • H. Mosbech,

    1. Allergy and Chest Unit, Section for Internal Medicine B, Frederiksberg University Hospital, DK-2000 Copenhagen F, Denmark
    Search for more papers by this author
  • U. Müller,

    1. Allergy and Chest Unit, Section for Internal Medicine B, Frederiksberg University Hospital, DK-2000 Copenhagen F, Denmark
    Search for more papers by this author
  • On Behalf Of The Study Group*

    Search for more papers by this author
    • * Participants: W. Aberer, A. Geusau, Universitäts-Hautklinik Wien, Austria; J. Birnbaum, Departement des Maladies Respiratoires, Hôpital Sainte-Marguerite, Marseille, France; P. Blaauw, P. van Valenberg, W. Pieters, L. O. M. J. Smithuis, Elkerliek Ziekenhuis, Helmond, The Netherlands; F. Bonifazi, Servizio di Allergologia Respiratoria, Ospedale Reg. Umberto I, Ancona, Italy; J. Bousquet, Service des Maladies Respiratoires, Hôpital Arnaud de Villeneuve, Montpellier, France; M. T. Gallesio, Servizio di Allergologia, Ospedale Mauriziano, Torino, Italy; L. Díez-Gómez, Servicio de Alergia, Hospital Ramón y Cajal, Madrid, Spain; G. Klein, Dermatologische Abteilung, Krankenhaus der Elisabethinen, Linz, Austria; M. Kosnik, University Clinic of Respiratory and Allergic Diseases, Golnik, Slovenia; G. Kunkel, H. Ratti, Institut für Klinische Immunologie und Asthmapoliklinik, Virchow-Klinikum, Berlin, Germany; F. de Maat-Bleeker, Poli Allergie, Academisch Ziekenhuis, Utrecht, The Netherlands; H. Mosbech, Allergiklinikken, Rigshospitalet, København, Denmark; U. Müller, Medizinische Klinik, Zieglerspital, Bern, Schwitzerland; E. Pastorello, Instituto di Clinica Medica Generale e Terapia Medica I, Milan, Italy; E. Pedro, A. G. Palma Carlos, Servico de Medicina III – Allergologia, Hospital de Santa Maria, Lisbon, Portugal; B. Przybilla, F. Ruëff, Klinik und Poliklinik für Dermatologie und Allergologie, Ludwig-Maximilians-Universität, München, Germany; B. Wüthrich, J. Fäh, Dermatologische Klinik, Universitätsspital, Zürich, Schwitzerland; R. Zerboni, M. G. Severino, P. Campi, Allergologia e Immunologia Clinica, Nuovo Osp. San Giovanni di Dio, Firenze, Italy; C. van der Zwan†, Ziekenhuis Eemland, Amersfoort, The Netherlands.


Dr H. Mosbech
Allergy and Chest Unit
Section for Internal Medicine B
Frederiksberg University Hospital
DK-2000 Copenhagen F
Denmark

Abstract

Background: The effect of venom immunotherapy (VIT) is well documented, but fear of systemic side-effects (SE) may prevent its use. The study aimed to analyze the character and frequency of SE and risk factors.

Methods: In a prospective study, 19 European centers included patients starting on VIT for systemic reactions to insect stings. Various dose regimens were applied.

Results: Data from 840 patients with a total of 26601 injections were obtained. Seventy-one percent were treated with Vespula-venom extract and 27% with honeybee-venom extract. Twenty percent of patients had SE corresponding to 1.9% of injections during dose increase and 0.5% during the maintenance phase. The vast majority of the 280 reactions were mild: only one-third required medical treatment. Injected or inhaled adrenaline was applied in six patients, of whom only one had a drop in blood pressure and collapse. Female sex, bee-venom extract, and rapid dose increase, but not severity of insect sting reactions, increased the risk of SE. The severity of SE was less in males but was not related to age, treatment phase, species of insect, or severity of insect sting reactions.

Conclusions: The frequency of SE was low, and the majority of thesecould be managed without treatment. Risk was increased in females, in bee-venom-treated patients, and in those with rapid dose increase.

Side-effects have been crucial when discussing the pros and cons of allergen-specific immunotherapy. In some countries, the fear of side-effects has led to strict regulations preventing a more widespread use of this treatment ( 1, 2).

One could expect the patient's type of allergy and severity of symptoms to influence the frequency, severity, and kind of side-effects with subsequent immunotherapy. Since insect-venom-allergic patients after being stung present with a variety of severe reactions in several organs, a fear in patients and physicians of more severe and frequent side-effects with venom immunotherapy (VIT) than with, for instance, pollen immunotherapy, is not uncommon. However, in an early study, this fear turned out to be groundless ( 3).

In the USA, a large-scale study of insect allergy was performed just after the introduction of venom extracts in the 1970s ( 3). In Europe, at present, reliable and comparable large-scale information on the prevalence, character, and severity of side-effects to VIT is scarce and not available from either pharmaceutic companies or national authorities. The aim of the present study was to collect data prospect-ively in order to be able to assess the types and frequencies of side-effects with VIT with different extracts, dose regimens, and patient groups.

Material and methods

Design

All allergy centers represented by the Subcommittee on Insect Allergy of the European Academy of Allergology and Clinical Immunology (EAACI) were invited to join the study. In participating centers, patients with systemic reactions starting on VIT were included prospectively. For each patient, a form was filled in describing symptoms, atopy, and results of diagnostic testing. A reporting form had to be filled in if a side-effect occurred, and at regular intervals, lists of the number of injections given to each patient were updated. Insect sting reactions and side-effects during VIT were reported and subsequently graded according to a scale ( Table 1) ( 4). The inclusion period was 1 year except in a few centers with a large intake of patients, where the number of patients had to be limited to avoid dominance. The subsequent study period was 2 years.

Table 1.  Classification of abnormal sting reactions and side-effects. Modified from Mueller ( 4)
Type of reactionSymptoms
0) Large localSwelling >10 cm for >2 days
1) MinorItching, urticaria, edema, malaise, anxiety
2) GeneralChest tightness, palpitations, dizziness, nausea, abdominal pain
3) SevereSomnolence, respiratory difficulties, vomiting, diarrhea, incontinence
4) AnaphylacticConfusion, drop in blood pressure, feeling of impending doom, unconsciousness, cyanosis, death

In several centers, patients were referred to their own general practitioners for further VIT after a certain number of injections. To ensure standardized reporting of data, only injections given in the participating centers before referral were included in the study.

Centers

Twenty-three centers intended to participate in the study. Four did not include patients; subsequently, data from 19 centers in 10 European countries could be evaluated.

Patients

A total of 840 patients (457 males and 383 females) were included. The mean age was 41 years (range: 5–77 years) in the total group and in the major subgroup treated with extract from one insect species only. All patients had experienced at least one systemic reaction to an insect sting, and none had received VIT previously. Basic clinical data are listed in Table 2. All patients had either a positive skin test or specific IgE to the relevant allergen.

Table 2.  Clinical data at inclusion
 One-allergen groupTotal group
Sex
 F47%46%
 M53%54%
Age – years (mean, range)41 (5–77)41 (5–77)
Unrelated allergy
 Asthma 5% 5%
 Rhinitis13%12%
 Urticaria 5% 5%
Interval (years) after sting (mean, range)2 (0–26)2 (0–26)
Severity of sting reaction
 Mueller class 112%12%
 Mueller class 224%23%
 Mueller class 341%41%
 Mueller class 423%24%
Positive skin test
 Wasp83%84%
 Honeybee46%47%
Specific IgE for
 Wasp82%83%
 Honeybee52%53%

VIT

No effort was made to harmonize the treatment regimen in the different centers, and, although the number of venom extracts was limited, most centers used their own treatment protocols. Updosing was performed either by weekly injections, by rush regimens with several injections on a few consecutive days, or by clustered regimens with accumulated injections on separate days, often with weekly intervals. No attempt was made to compare data from patients receiving different rush regimens.

The majority, 417 males and 365 females, were treated with one venom extract. Fifty-eight patients had two venom-extract treatments concomitantly ( Table 3). Commercial extracts with venom from wasp (Vespula species), honeybee, and paper wasp (Polistes species) were applied. The number of patients treated in various centers differed considerably ( Table 4). Extracts were in an aqueous solution either with or without aluminum hydroxide added to give a depot effect. Some patients switched from one type of extract to another, the majority switching to the depot version when the maintenance dose was reached, and a few switching due to repeated side-effects with one type of extract.

Table 3.  Allergen extracts and dose-increase regimen for 840 patients included and 782 treated with venom extract from only one insect species. Some patients changed from depot to nondepot, or vice versa
 No. of treatment series
 One-allergen groupTotal group
Extract
 Depot122174
 Nondepot656720
 Not specified  4  4
Venom
 Vespula species557615
 Honeybee212256
 Polistes species 13 27
Updosing regimen
 Weekly113179
 Clustered 63 97
 Rush/ultrarush508524
 Mixed 98 98
Table 4.  Number of patients treated with various allergens in participating centers (patients treated with venom extract from only one insect species)
 IT allergen
 HoneybeeVespulaPolistesTotal
Centre
1841 49
21316 29
3235 37
41944467
523 5
639113
72529
844 8
91518 33
10 61 61
11419 23
12433 37
133110 41
141061273
1512 3
1650122 172
172937 66
181316433
19221 23
Total21255713782

Statistics

Cross-tabulations of categoric variables were analyzed by the chi-square test. Group comparison of continuous variables was done by the Mann-Whitney U-test (or the Kruskal-Wallis test if more than two groups), and correlation was expressed by the nonparametric Pearson correlation coefficient. Stepwise logistic regression was applied to determine the influence of the various parameters on the occurrence of side-effects in the individual patients. A level of probability above 5% was regarded as significant. Statistical analyses were done by SPSS 7.0 for Windows®.

Results

A total of 26601 injections were given, 23602 to patients receiving treatment with only one extract. In this subgroup, 12122 injections were given during the dose-increase phase (mean: 15 per patient, SD: 5) and 11480 during the maintenance phase (mean: 15 per patient, SD: 16).

Ninety-two percent of the patients received a maintenance dose of 100 μg, 3% left the study before reaching a maintenance level, 2% had to stay on a lower level (20–90 μg), and 3% received 110–200 μg at each injection during the maintenance phase.

A total of 299 systemic side-effects were reported; of these, 280 occurred in patients treated with one venom only. In this subgroup (and in the total group), 20% of the patients had at least one systemic reaction and 1.2% of injections elicited reactions. Side-effects occurred during dose increase at a median dose of 30 μg (range: 0.01–150 μg) and during the maintenance-phase at a median dose of 100 μg (range: 30–150 μg). The detailed analyses below were made in the subgroup treated with one venom only.

The majority of systemic symptoms were very mild ( Fig. 1), and only one-third required treatment. Oral antihistamine was the drug most frequently applied ( Table 5). A drop in blood pressure was reported in nine cases, but only one patient received adrenaline. This patient and one other patient suffered fainting/collapse. In the latter case, no adrenaline was administered.

Figure 1.

Side-effects during dose increase and maintenance phases classified according to severity.

Table 5.  Treatment of side-effects to VIT in patients treated with venom extract from only one insect species
Type of treatment% of reactions
No treatment66
Oral treatment (total)18
 Antihistamine17
 Corticosteroid 6
Injection (total)19
 Adrenaline 3
 Antihistamine15
 Corticosteroid11
Inhalation (total) 6
 Adrenaline spray 3
 Inhaled β-agonist 2

The frequency of reactions was higher during the dose-increase phase than the maintenance phase (mean: 1.9%vs 0.5% of all injections). When analyzed separately, female sex, rapid dose-increase regimens, and treatment with bee-venom extract seemed to increase the risk of side-effects ( Table 6). If analyzed in a model taking the correlation between the different parameters into account, only female sex and rapid dose-increase (rush) regimens turned out to be risk factors for systemic side-effects. In addition, patients with pre-existing allergic rhinitis more often had side-effects (29%vs 19%, P<0.05). The following parameters did not influence the risk of systemic side-effects in either separate analyses or logistic regression: age, pre-existing asthma or urticaria, severity of original insect sting symptoms, time interval between sting and symptoms, number of systemic sting reactions, progression in sting reactions, type of extract (with or without aluminum hydroxide), and number of venom extracts used for treatment (one or two).

Table 6.  Systemic side-effects to VIT (patients treated with venom extract from only one insect species). Calculations are based on frequency of side-effects in individual patients
 Occurrence of systemic side-effects
(mean and range)
Type of
patient/treatment
Percentage of
injections during
updosing
Percentage of
injections during
maintenance
Percentage
of
patients
  1. P<0.05.

Sex
 F2.0 (0.0–52)* 0.9 (0.0–100)  25*
 M1.2 (0.0–63)* 1.2 (0.0–100)  15*
Extract
 Depot1.2 (0.0–21)  0.2 (0.0–14)   17 
 Nondepot1.6 (0.0–63)  1.3 (0.0–100)  20 
Venom
 Vespula species1.5 (0.0–63)  0.6 (0.0–100)* 19 
 Honeybee1.8 (0.0–52)  2.4 (0.0–100)* 24 
 Polistes species   0 (0.0–0.0)   0 (0.0–0.0)   0 
Updosing regimen
 Weekly0.8 (0.0–21)* 0.2 (0.0–14)   12*
 Clustered1.3 (0.0–9.1)*0.1 (0.0–3.8)  22*
 Rush/ultrarush2.1 (0.0–63)* 1.5 (0.0–100)  24*
 Mixed0.5 (0.0–9.1)*1.0 (0.0–67)    8*
Total group1.6 (0.0–63)  1.1 (0.0–100)  20 

When side-effects occurred, they were less severe in males ( Fig. 2). Severity was not related to pre-existing asthma, rhinitis, or urticaria; severity of original insect sting reaction ( Fig. 3); whether or not the patient had had a positive skin test or allergen-specific IgE for the relevant venom pretreatment; type of extract; or the insect venom used for treatment.

Figure 2.

Side-effects in women and men classified according to severity.

Figure 3.

Severity of last reaction to relevant insect for different severity of side-effects.

Side-effects occurring during the dose-increase phase did not differ from side-effects during the maintenance phase with respect to severity ( Fig. 1), interval after injection, or duration, but reactions during the maintenance phase were treated more often (52%) than reactions during updosing (31%). Reactions occurred in patients not differing with respect to age or severity of the original insect sting reaction.

Rapid onset of side-effects correlated positively to short duration of these reactions (R=0.47, P<0.001). Severity was not correlated to age of patient, time interval from original insect sting to symptoms, or time interval from injection to occurrence of side-effects.

Discussion

The present multicenter study has made it possible to collect a very large set of data covering a variety of treatment regimens administered to a high number of patients with different symptoms. This has enabled analyses not possible in local settings.

We intended to make the reporting forms as self-explanatory as possible and collected data on a running basis in order to correct misunderstandings and make the data as uniform and complete as possible. The basic information describing patients at inclusion was obtained in a standardized form. The reporting of side-effects was less standardized, as some centers registered every slight symptom/sensation in a detailed fashion, whereas other using the same extract and dose regimen reported fewer but more substantial side-effects. Subsequently, reporting of Mueller class I side-effects was probably somewhat less reliable than the reporting of more severe reactions. However, the main conclusion of the study that severe side-effects to VIT are very infrequent was not invalidated by the methodological problems mentioned. An analysis including only objective symptoms would have reduced the frequency considerably, but could not be done retrospectively.

The fact that only one-third of the systemic reactions were treated illustrates the low severity of the reactions. Adrenaline was applied in only one of the nine cases in which a drop in blood pressure was reported. Blood pressure levels were not registered in this study, but the limited need for this effective and appropriate treatment must reflect limited changes in blood pressure, since all centers could be regarded as experienced in VIT and the treatment of side-effects.

The finding that rush updosing seems to imply a higher risk of side-effects is not unexpected ( 5). A slower and less “aggressive” regimen would have a less drastic impact on the immune system and thereby also be less likely to induce side-effects. However, in one comparative study with VIT, the same frequency per injection (and a lower frequency per patient) was observed ( 6). A very low frequency of systemic side-effects (7% of patients) was seen in an extremely rapid dose-increase schedule (ultrarush) ( 7). In fact, such regimens were applied in some patients in a few of the participating centers, but since no attempt was made to standardize or obtain very detailed information on the many dose regimens applied in this study, only broad conclusions could be drawn. The number of injections needed to reach the maintenance dose did not differ between the dose-regimen groups. This indicates that only a minority of patients were offered the ultrarush regimen. Additional investigations are needed if, for instance, subtypes of rush regimens are to be compared.

The present study confirms the finding that there seem to be more side-effects with bee-venom than wasp-venom treatment ( 7–11). The reason is unknown. No differences were found in side-effects in patients starting treatment with aqueous and with aluminum hydroxide-precipitated venom extracts. A few patients switched during updosing due to side-effects, and others switched after reaching the maintenance dose to be able to increase the dosing interval. Since the background for the switch was not registered systematically, a detailed analysis comparing the two types of extract was not possible.

As in a recent British study ( 10), female patients had more side-effects than male and these tended to be more severe than the reactions registered in the male group. This finding might be explained by a lower threshold among women for reporting symptoms.

In other studies, the frequency of side-effects with VIT has varied between 0% and 100% in different patient groups exposed to different extracts and different dose regimens ( 3, 5, 10, 12). In the present study, about one-fifth of the patients had one or more systemic side-effects. This figure may seem high, but several aspects have to be taken into account. Firstly, the little need for medical treatment of the side-effects confirms that these were generally very mild. Secondly, a reaction rate of below 2% during dose increase and 0.5% at the maintenance dose is generally lower than side-effects reported for other allergens administered for less serious diseases ( 13). In addition, adrenaline self-injection, the alternative to VIT, seems inconvenient to many patients, and the efficacy of adrenaline treatment is not 100% ( 14); therefore, VIT is to be preferred in spite of the side-effects.

In several centers, maintenance treatment was continued by general practitioners. The lower risk of side-effects after updosing confirms that this is a sensible way to organize treatment if specialist resources are limited or are far away from the patient. This lower risk during maintenance treatment has been a general finding ( 3, 10). However, since reduction accounts only for the frequency, and not the severity, of side-effects during this phase, general practitioners, as well as specialists, should be ready to treat anaphylactic reactions. In the present study, we intended to collect data from general practice if patients continued treatment there. Unfor-tunately, the information was not sufficiently extensive to allow analysis. However, there is no reason to expect a higher frequency of side-effects in general practice, since many of the patients continuing treatment in hospital were in fact “problem patients” selected because of previous side-effects.

Selection of candidates for VIT by sting challenge, as was done in one of the participating centers, can exclude more than half of the patients with otherwise positive diagnostic tests ( 15). One would expect the remaining patients to have an increased risk of side-effects. However, this could not be demonstrated in the present study, in which only a very small percentage of the patients had been challenged in hospital before treatment. Similarly, one would have expected previous severe insect sting reactions to indicate an increased sensitivity to the venom extract and, similarly, an increased risk and severity of side-effects. As in another large-scale study ( 8), we could not confirm this hypothesis. One explanation could be that the Mueller classification cannot categorize intermediate and severe reactions properly. Recently, a new side-effect classification has been proposed (primarily for patients with inhalant allergy) ( 2), but this might imply a better prediction.

In conclusion, it has been possible to carry out a large and long-term multicenter study under the auspices of the Insect Venom Allergy Subcommittee of the EAACI. The results obtained should encourage others to carry out collaborative studies on different aspects of allergy. Our results showed that the majority of systemic side-effects of VIT were very mild even though they occurred in one-fifth of the patients. Few predictive factors were identified. Rapid dose regimens gave rise to more side-effects than slow. However, ultrarush treatment was not investigated separately. The higher risk of side-effects in women remains unexplained. It seems safe to let a well-prepared general practitioner perform the treatment, at least during the maintenance phase.

Footnotes

  1. * Participants: W. Aberer, A. Geusau, Universitäts-Hautklinik Wien, Austria; J. Birnbaum, Departement des Maladies Respiratoires, Hôpital Sainte-Marguerite, Marseille, France; P. Blaauw, P. van Valenberg, W. Pieters, L. O. M. J. Smithuis, Elkerliek Ziekenhuis, Helmond, The Netherlands; F. Bonifazi, Servizio di Allergologia Respiratoria, Ospedale Reg. Umberto I, Ancona, Italy; J. Bousquet, Service des Maladies Respiratoires, Hôpital Arnaud de Villeneuve, Montpellier, France; M. T. Gallesio, Servizio di Allergologia, Ospedale Mauriziano, Torino, Italy; L. Díez-Gómez, Servicio de Alergia, Hospital Ramón y Cajal, Madrid, Spain; G. Klein, Dermatologische Abteilung, Krankenhaus der Elisabethinen, Linz, Austria; M. Kosnik, University Clinic of Respiratory and Allergic Diseases, Golnik, Slovenia; G. Kunkel, H. Ratti, Institut für Klinische Immunologie und Asthmapoliklinik, Virchow-Klinikum, Berlin, Germany; F. de Maat-Bleeker, Poli Allergie, Academisch Ziekenhuis, Utrecht, The Netherlands; H. Mosbech, Allergiklinikken, Rigshospitalet, København, Denmark; U. Müller, Medizinische Klinik, Zieglerspital, Bern, Schwitzerland; E. Pastorello, Instituto di Clinica Medica Generale e Terapia Medica I, Milan, Italy; E. Pedro, A. G. Palma Carlos, Servico de Medicina III – Allergologia, Hospital de Santa Maria, Lisbon, Portugal; B. Przybilla, F. Ruëff, Klinik und Poliklinik für Dermatologie und Allergologie, Ludwig-Maximilians-Universität, München, Germany; B. Wüthrich, J. Fäh, Dermatologische Klinik, Universitätsspital, Zürich, Schwitzerland; R. Zerboni, M. G. Severino, P. Campi, Allergologia e Immunologia Clinica, Nuovo Osp. San Giovanni di Dio, Firenze, Italy; C. van der Zwan†, Ziekenhuis Eemland, Amersfoort, The Netherlands.

Ancillary