Developments in the field of allergy in 2009 through the eyes of Clinical and Experimental Allergy

Summary

In 2009 the journal published in the region of 200 papers including reviews, editorials, opinion pieces and original papers that ran the full gamut of allergic disease. It is instructive to take stock of this output to determine patterns of interest and where the cutting edge lies. We have surveyed the field of allergic disease as seen through the pages of Clinical and Experimental Allergy (CEA) highlighting trends, emphasizing notable observations and placing discoveries in the context of other key papers published during the year. The review is divided into similar sections as the journal. In the field of Asthma and Rhinitis CEA has contributed significantly to the debate about asthma phenotypes and expressed opinions about the cause of intrinsic asthma. It has also added its halfpennyworth to the hunt for meaningful biomarkers. In Mechanisms the considerable interest in T cell subsets including Th17 and T regulatory cells continues apace and the discipline of Epidemiology continues to invoke a steady stream of papers on risk factors for asthma with investigators still trying to explain the post-second world war epidemic of allergic disease. Experimental Models continue to make important contributions to our understanding of pathogenesis of allergic disease and in the Clinical Allergy section various angles on immunotherapy are explored. New allergens continue to be described in the allergens section to make those allergen chips even more complicated. A rich and vibrant year helpfully summarized by some of our associate editors.

Asthma and rhinitis

Definition and phenotypes of asthma

Although the latest GINA definition of asthma is widely accepted, it remains confused because it fails to identify a primary characteristic which separates it from other diseases [1]. These disadvantages can be overcome by objective measurements of the components that make this disease heterogeneous [2, 3]. One of the phenotypes of asthma is intrinsic or non-allergic asthma, which is characterized by progressive loss of lung function beginning later in life in the absence of atopy. Holt and Sly [3] place this entity between the two extremes of chronic airway inflammatory diseases represented by atopic asthma and COPD. Among the mechanisms underlying intrinsic asthma, it has been proposed a role of chronic pulmonary infections in amplifying inflammation and/or in promoting expansion of T cells and IgE production driven by bacterial superantigens [4, 5]. An alternative view sustained by Black and Roth [6] is that an abnormality in the airways smooth muscle cells (SMC) is the primary or instigating event. A large proportion of patients with adult-onset non-allergic asthma are female with a peak in the prevalence around the age of 50 years. In these patients, Van der Berge et al. [7] proposes a role for female sex hormones such as oestrogens and progesterone. However, there is controversy in this area. Interestingly, oestrogens have been shown to protect against insulin resistance and menopause is accompanied by an increased insulin resistance due to a decline of oestrogen levels [8]. This is in line with the observation of Thuesen et al. [9], that insulin resistance is a stronger predictor for development of adult-onset asthma than the presence of obesity. It is unclear whether insulin resistance itself contributes to the onset of the disease or whether both insulin resistance and asthma is caused by the low-grade systemic inflammation present in obesity.

Environmental risk factors

There are a number of new findings on the influence of food and diet on the development of asthma [10]. In Finland, where low vitamin D intake, a short exposure to sunlight and high prevalence of asthma are seen, higher maternal vitamin D intake from foods during pregnancy is protective from childhood asthma and allergic rhinitis (AR) in the offspring [11]. In Eastern Europe, like Belarus, where the prevalence of allergic disease is lower, one of the factors most strongly associated to symptoms of asthma and AR was the use of probiotics for abdominal pain, diarrhoea and in association with antibiotics [12]. The finding is intriguing because it contrasts with the hygiene hypothesis, unless probiotic use is merely a marker for life-style or behavioural characteristics that increase the risk of allergic disease, rather than an independent cause. In contrast, studies conducted in Australia [13] and United States [14] confirmed the hygiene hypothesis. Specifically, exposure to siblings and infections protected against early onset AR (i.e. before age 7), whereas increased health care and medication use favoured sensitization to environmental allergens.

There is a debate whether early nutritional interventions may modify the development of atopic disease in infant and children. Food allergy has been identified by Schroeder et al. [15] as a risk factor of asthma in children independent from aeroallergen sensitization and family history of asthma. Subsequent dietary behaviour may modify the beneficial effect of breastfeeding. In fact, Mai et al. [16] showed that fast food consumption was associated with asthma in 8–10-year-old children and potentially counteracts the protective effect of prolonged breastfeeding on asthma.

Genetic risk factors

Over 100 genes have been associated with asthma or atopy-related phenotypes. Common polymorphisms have been identified in genes suspected to play a role in asthma and allergy according to their biologic mechanisms underlying immune responses [17]. More recently, data from positional cloning studies have suggested that allergic diseases may result from an impairment of the epithelial barrier in affected tissues [18]. Less than 50% of the polymorphisms have been replicated in two or more independent samples. Genuneit et al. [19] investigated the association of 55 relevant polymorphisms with wheeze and allergy in a case-control sample from Phase 2 ISAAC. Despite the biological plausibility of IgE-related mechanism in asthma very few of the tested candidates were associated with both wheeze and increased IgE. The strongest association with wheeze were detected in genes for IL-4R, MS4A2, Toll-like receptor 4 (TLR 4) and TLR 9. Vergara et al. [20] were able to replicate the association of G-protein-coupled receptor 154 with asthma and total IgE in a population of the Caribbean coast of Colombia. Like other complex diseases, asthma is influenced by both genetic and environmental factors. The findings of Rogers et al. [21] support an important gene-by-environment interaction between two common factors. In fact, increased risk of asthma conferred by GSTM1-null homozygosity in children was restricted to those with a history of intrauterine exposure to tobacco smoke.

Mechanisms

It is well recognized that the immune response in allergic asthma is polarized towards type 2 T cells (Th2). The increase of Th2-associated allergic diseases and Th1-related autoimmune diseases has hence raised the hypothesis that these conditions might result from a functional or a quantitative deficiency of regulatory T cells (Treg), that provide both Th2 and Th1 immune responses [22]. There’s a variety of Treg subsets that either develop as a normal part of the immune system or in response to particular antigenic exposure. The best known are CD4⁺T cells that constitutively express high levels of surface CD25 (CD25^+high) and IL-10-producing T cells. Considering this, Borrego et al. [23] tried to characterize and define the differences in immunological status of 50 young children with recurrent wheeze (with and without clinical risk factors for developing asthma) with 30 age-matched healthy controls by assessing the role of CD4⁺CD25⁺T cell population in regulating the immune response. Children with recurrent wheeze appeared to have lower absolute numbers of Treg population (CD4⁺CD25⁺ and CD4⁺CD25^+high) and lower expression of IFN-γ than healthy controls. There are many possible explanations for the finding of a reduced suppressive activity of Treg in allergy and asthma. One hypothesis is that atopic and/asthmatic individuals have an inherent defect that reduces their suppressive capacity. Another is the hygiene hypothesis that suggests that the lack of exposure to infectious agents and environmental microbes in Western population in childhood may cause a failure in the development of protective Treg responses to allergens. In this contest, the role of innate immunity appears relevant. Dendritic cells (DCs) represent the bridge between innate and acquired immunity. For instance, endogenous danger signals trigger DCs to stimulate T cell responses to inhaled allergens [24]. These signals, such as adenosine triphosphate (ATP) or uric acid may play a role at starting and maintaining allergic disease. Particularly extracellular ATP seems to be involved in DC maturation and can induce a rapid and valuable death in CD4⁺CD25⁺Treg.

A cellular component, including T cell infiltration, is observed mainly in the late-phase of allergic response. Moed et al. [25] investigated skin responses after intracutaneous injection of house dust mite (HDM) in children with AR. The magnitude of late-phase skin response correlated with the ability of T cells, taken from peripheral blood before skin test, to proliferate in vitro, suggesting a causal role for antigen-specific T cells in late-phase response to allergens.

Bronchial SMC are crucial in airway remodelling characteristic of asthma. Solarewicz-Madejek et al. [26] showed that SMCs are one of the targets of T cells and eosinophil activity. Indeed, infiltrating T cells and eosinophils are able to trigger bronchial SMC death by apoptosis and necrosis. Apoptotic SMCs were detected particularly in central airways of fatal asthma. The interpretation of these results is intriguing because one of the most notable features detected in segmental and subsegmental airways from severe asthma was an increased thickness of smooth muscle area compared with non-severe asthma [27]. In the attempt to understand the mechanism of airway smooth muscle (ASM) hyperplasia, Saunders et al. [28] evaluated CCR3, CCR7, CXCR1, 3, and 4 expression and function in asthma. No differences in chemokine receptor expression between ASM cells derived from asthmatics and non-asthmatics were detected, suggesting that their importance in asthma is uncertain.

While infiltrating inflammatory cells in bronchial wall were comparable in severe and non-severe disease in the study of Macedo, Shiang et al. [29] demonstrated that cell density of mast cells, eosinophils and neutrophils in adventitial layer of pulmonary arteries was increased in fatal asthma. The latter observation contributes to the understanding of distal lung alteration in severe asthma, because much of previous knowledge on vascular changes in asthma were based on the analysis of the submucosal plexus of vessels of the bronchial circulation. Changes in airway microcirculation are generally regarded as factors of severity and chronicity in asthma. Recent observations suggest the contrary. Kristan et al. [30] demonstrated that angiogenesis-related factors, like angiogenin and vascular endothelial growth factor (VEGF), are significantly increased in the induced sputum supernatant of patients with rhinitis without asthma as well as in asthmatic patients. Khor et al. [31] found that airway vascular leakage is increased in parallel with early clinical deterioration of asthma control and precedes airway inflammatory cellular changes and angiogenesis.

Occupational asthma

Occupational asthma induced by diisocyanates and other low molecular weight chemicals is an interesting model of adult-onset asthma because the causing agent is known, atopy is not a confounder, the exposure (onset, levels, duration, cessation) to offending agent can be assessed. The mechanism of sensitization is controversial because specific IgE antibodies against isocyanates are found in a minority of subjects with asthma. Isocyanates may directly damage airway epithelium through a GSH-dependent mechanism [32] and increase the permeability of bronchial epithelium through a VEGF-mediated pathway in vitro [33]. Occupational exposure to isocyanates enhances spontaneous IL-10 production by peripheral blood mononuclear cells (PBMC) [34]. In addition, spontaneous production of soluble human leucocyte antigen-G (HLA-G), a non-classical human leucocyte antigen class I molecule with tissue-protective properties, was significantly higher in PBMC from subjects with isocyanate asthma compared with non-occupational allergic asthmatics, whose PBMC produced HLA-G only after LPS stimulation. The results suggest that there is a different regulation of HLA-G in asthma induced by isocyanates and in non-occupational allergic asthma.

At variance of asthma caused by low molecular weight chemicals, occupational asthma caused by vegetal or animal proteins shares the same mechanisms of common allergic asthma and the demonstration of a specific sensitization relays on conventional in vivo (prick test) or in vivo (specific IgE) tests. However, the multi-centre study of van Kampen et al. [35] found that, even for the relatively common baker’s asthma, there is a wide variability in the antigen preparations from different producers. The recombinant lipid transfer protein (LTP) Tri a 14, a major allergen associated with wheat allergy, characterized by Palacin et al. [36] represents a new potential tool for baker’s asthma diagnosis, based on its physicochemical and immunologic similarity with its natural counterpart.

Treatment

Allergen-specific immunotherapy (IT) is one of the treatment options for AR and asthma. A concern of subcutaneous IT is the risk of adverse events. A prospective Italian survey reported that 95 systemic reactions and no fatality occurred after a total of 60785 injections for IT [37]. Only three severe systemic reactions were recorded, i.e. two severe asthma and one anaphylaxis. Ragweed and grass extracts caused significantly more side-effects than other allergens. The conclusion of this study is that injection IT has an acceptable risk/benefit ratio when prescribed and carried out according to recommendations. An alternative route of administration for IT is sublingual. In a multinational study the efficacy and safety of five-grass pollen sublingual IT was tested in 628 adults with allergic rhinoconjunctivitis [38]. No serious systemic events or anaphylaxis were observed. The authors concluded that risk-benefit ratio was favourable in all of these patients subgroups, regardless the severity profile, sensitization status, and presence of asthma.

The other treatment directed against a mechanism of allergic asthma is omalizumab, a monoclonal anti-IgE antibody. The combination of omalizumab and specific IT was tested in patients with seasonal allergic rhinoconjunctivitis and asthma [39]. The patients who received combination therapy exhibited significantly less symptoms than those on IT only. Unfortunately, the study cannot tell what would be the efficacy of omalizumab without IT, i.e. what IT adds to omalizumab treatment. There is evidence that a successful specific IT functions through the induction of various Treg subsets [40]. It has been shown that the administration of a single dose of corticosteroid induces a burst of IL-10 synthesis, which is an essential requirement for the induction of IL-10 secreting Treg (Treg1) [41]. There is also evidence that vitamin D leads to an increase of IL-10 production by human T cells [42]. However, the attempt to enhance the efficacy of specific IT in asthmatic children allergic to HDM by administration of prednisone and vitamin D on the day of antigen extract injection was unsuccessful [43]. Actually, intervention with prednisone impaired the induction of Treg in this in vivo study.

Topical corticosteroid is considered the medication of choice for treatment of asthma and AR. However, the response to treatment is not always satisfactory. The reason(s) for poor corticosteroid responsiveness is controversial. In a cross-sectional study in daily practice, Quaedvlieg et al. [44] showed that asthma uncontrolled by treatment is associated with higher degree of sputum eosinophilia, but the design of the study cannot tell which factors are predictive of a worse outcome. Prospective studies are more informative. Indeed, Valera et al. [45] examined biopsies of nasal polyps before and after treatment with topical budesonide and found that higher expression of IL-1β, ICAM-1 and NF-κB before treatment predicted a poorer response to corticosteroid.

Mechanisms in allergy

Mucosal tolerance to allergen sensitization

The mucosal surface (e.g. airway and skin) provides an active barrier to prevent the host from damage. Der p 2 from HDM and Ph l 1 from Timothy grass have been shown to activate or injure airway epithelium, leading to epithelial barrier breach and subsequent development of allergic sensitization and adaptive immune response [46, 47]. In contrast, normal flora or commensal bacteria maintain mucosal tolerance via increasing CD4⁺CD25⁺ Tregs cells. DC act as a ‘sensor’ for the development of tolerance to allergens. Mucosal signals, including pattern-associated molecular patterns from microbes and damage-associated molecular patterns generated in the host, keep resident DCs in an immature or non-inflammatory state to promote Tregs development. Animal asthma model studies suggest that deletion of Treg cells was essential to break the tolerance towards ovalbumin (OVA) [48]. Likewise, during the grass allergy season, reduced allergen-specific Tregs were observed in PBMC of allergic rhinoconjuctivitis patients [49]^. Staphylococcus endotoxin B (STB) stimulation to nasal polyps and turbinate tissues decreased Tregs, but increased the pro-inflammatory cytokines IL-6, IFN-γ and IL-4. Neutralization of IL-6 in nasal tissues increased IL-10 and Foxp3 mRNA as well as Tregs following STB stimulation [50]. These studies suggest that breaching mucosal tolerance is the first step in the inception of allergy. Therefore, more research efforts aimed at maintaining mucosal tolerance to various allergens are needed to prevent or reverse the allergic diseases.

Modulation of allergen sensitization by microbes

Endotoxin and microbial exposure in early childhood may play a critical role in shaping immune system and hence modulating the development of atopy [51, 52]. The hygiene hypothesis has shifted our research focus towards understanding the role of natural microbiota and microbial infection in the regulation of allergen sensitization. Previous studies have indicated the immunomodulatory role of gut microbiota that colonizes shortly after birth and microbial infection during early age of life in immune system development. Sjogren et al. [53, 54] demonstrated that increased colonization of Bifidobacterium species during early childhood was associated with reduced risk of developing allergic diseases via increasing humoral immunity whereas less colonization of certain lactobacilli group I bacteria Bifidobactreium adolescentis and Clostridium difficile can increase the risk of allergy.

Pathogenic microbial infections can alter the natural microbiota that might induce allergen sensitization or exacerbate allergic diseases by damaging epithelial barrier integrity and inducing airway inflammation [55]. Microbiome in the lung is relatively less known compared with that in the gut. Recent publications suggest that lung microbial diversity is lost in asthmatics especially during an exacerbation. Analysis of microbiome of asthma patients showed predominance of probacteria like Haemophilus spp whereas bacteroids like Prevotella spp were more common in control subjects [56]. Early childhood hypopharyngeal colonization with Streptococcus pneumoniae, Moraxella catarrhalis, Haemophilus influenzae, or a combination of these microbes significantly increased wheezing and hospitalization [57]. Elucidation of microbial diversity and its role in host immune response will promote more research focusing on the application of probiotics and TLR agonists in the prevention and treatment of allergic diseases.

Allergic inflammation

Breach of mucosal tolerance generates Th cytokine imbalance in allergic diseases including asthma. Although tremendous efforts have been made to understand the cytokine network, targeting Th2 cytokines (such as IL-5) and/or inflammatory cells has not been very effective in controlling allergic diseases [58]. Studies of new subtypes of T helper cells including Th17 and Th9 cells have generated a greater interest in better understanding allergic diseases [59]. Owing to its critical role in driving neutrophil inflammation in severe asthma, IL-17 has been suggested to be an attractive therapy for asthma [60]. Caution should be taken in utilizing IL-17 neutralizing therapy as many studies have indicated a pivotal role of IL-17 in maintaining mucosal immunity. Recurrent or persistent bacterial and viral infection is commonly found in severe asthma especially during exacerbation. Bacterial or viral infection can drive a Th17 response and uncontrolled airway neutrophilic inflammation in an allergic host [61]. As there is no effective therapy in place, persistent airway inflammation has been associated with higher morbidity of severe asthma [23, 62]. Current treatments and prevention strategies are focused on eosinophilic inflammation and allergen avoidance, respectively. Future therapies may be shifted to improve host mucosal innate immunity and reduce neutrophilic inflammation in allergic diseases such as severe asthma.

Genetic studies in allergic diseases

Single nucleotide polymorphisms (SNP) in genes especially those involved in immune recognition or inflammatory response can render a host susceptible or resistant towards developing allergic diseases. SNPs in IL-4 and IL-4RA are significantly associated with increased susceptibility to allergic diseases such as asthma [63]. IL-12 is known to promote the generation of Th1 cells. Recently, Shimokawa et al. [64] discovered an SNP (CTCAA/GC allele) in IL-12B gene promoter region, which was associated with increased IL-12 promoter reporter activity. β2-adrenergic receptor (ADRB2) is one of the most studied asthma candidate genes considering its importance in asthma therapy. Unfortunately, it appears that ADRB2 SNPs are not associated with asthma aetiology [65]. A clinical study demonstrated reduced percentage of CD14⁺ monocytes and membrane-bound CD14 in PBMC of infants with a high risk developing allergic diseases [66]. Numerous studies suggest an important role of genetic factors in the inception of allergy, but due to contradictory or inconclusive results from various studies, contribution of SNPs or genetics to the development of allergic diseases remains poorly understood. Careful and well-designed genetic studies are warranted to meet the challenges in personalized medicine in allergic diseases.

Potential new therapies to control allergic diseases

Better understanding of molecular mechanisms of allergic diseases will not only help us identify new biomarkers for diagnosis [67, 68], but also drive the discovery of novel therapies or improve the use of existing drugs for more effective control of allergic diseases. Immunotherapies for various seasonal allergens have shown promising results in the prevention of allergy. Sublingual grass pollen IT was associated with increased Foxp3⁺ cells (Tregs) and systemic humoral changes such as increased allergen-specific IgG4 and IgA. Regulating innate immune responses to allergen and/or microbial exposure might also provide an effective treatment controlling the inflammation in allergic diseases. For example theophylline when administered at a therapeutic concentration reduced (60–70% reduction) myeloid DC differentiation from monocytes [69] and a combination of budesonide and formeterol additively or synergistically inhibits RSV-induced secretion of inflammatory mediators such as CCL5, CXCL8, CXCL10, as well as VEGF secretion from human airway epithelial cells [70].

Future research directions in allergic diseases

Future research work could be more centred on intervening the allergen sensitization. Preserving the microbial diversity at mucosal surface has been recognized as an essential factor in maintaining mucosal tolerance. Therefore, probiotics or TLR agonists may be utilized to modify mucosal microbiota and tolerance to allergens. Advancement of our knowledge in human gut and lung microbiome will help us predict host susceptibility to allergic diseases, disease severity, or therapeutic efficacy.

As the cytokine network is still considered to be critical in the development or progression of allergic diseases, more innovative research approaches are likely to be attained with a better understanding of the natural history of allergic diseases. For example, by understanding how allergic inflammation is stabilized or resolved without various therapies (e.g. steroids), we could develop more effective and safer therapies.

Epidemiology of allergic disease

Prevalence and genetics of allergic disease

The prevalence of allergic manifestations (e.g. asthma, eczema, hayfever, and food allergy) may have reached its plateau in some industrialized countries [71]; however, it is still increasing in many parts of the world [72–74] although Kramer et al. [12] found a relatively low incidence of allergic disease in eastern Europe. Overall, the prevalence of allergic manifestations varies between 25% and 35% in children and around 10% in adults [75]; while the prevalence of allergic senitization, for instance in the US population, has reached 54% in adults [76]. The concept of susceptible individuals developing new allergic diseases from infancy through childhood, the so-called ‘allergic march’, is firmly embedded in the literature and much time and energy has been devoted to how this might be prevented. However the way in which the allergic march develops at a population level has not been well established. In a novel epidemiological approach Punekar and Sheikh [77] have retrospectively constructed a large birth cohort from a primary care practice database, which has allowed them to determine the nature of the allergic march. Consultations for allergic disease were very common with 50% of the cohort diagnosed with at least one allergic condition. There were 15 different patterns of disease progression so that it seems to be not so much a march as a meander. Most strikingly they have found that easily the most common pattern was eczema followed by asthma which fits very well with the emerging picture that defects in the skin barrier lead to Th2 sensitization to aeroallergens through the skin, which in turn leads to asthma as a result of allergen-mediated effects on the airways. Modifiable environmental risks factors are of particular interests because they provide critical hints about what public health can do to prevent disease and which new treatment may be promising to be developed. Two papers addressed the question of how and when sensitization to aeroallergens develops. Oryszczyn et al. [78] studied cat allergy in adults and found that those people exposed during childhood appeared to develop a state of tolerance. On the same theme Matricardi et al. [79] investigating the development of sensitization in a birth cohort found that IgE responses to airborne allergens underwent dynamic changes throughout childhood with a high frequency of new sensitizations and remission with more persistent sensitization being associated with early onset and high levels of specific IgE. The importance of sensitization to airborne allergens in causing severe asthma was emphasized by Wang et al. [14] who found a strong correlation with increased health care and medication use in a group of inner city children.

Genetic variation has been documented to play a role in allergic disease. There are a large number of studies associating polymorphisms in candidate genes and allergic outcomes although many are underpowered or not replicated [17, 19]. A detailed and robust example of this type of study was undertaken by Vergara et al. [20] who found a positive association between asthma and total IgE and SNPs in the candidate gene G-protein-coupled receptor 154 in a Columbian population. In addition, the respective genetic polymorphisms are often detected only in small subgroups of the population, and do not explain a large proportion of allergic diseases. To follow Mendelian rules, genetic models are often restricted. This does not facilitate the application of novel explanatory models used in epidemiologic research. Gene–environment interactions are at the heart of the cause of disease and these are now beginning to be studied. Rogers et al. [21] investigated the interactions between the glutathione-S-transeferase MI null mutation and exposure to tobacco smoke. They found that the risk of asthma in children and the homozygous null mutation was restricted to those with a history of intrauterine smoke exposure. Furthermore, recent studies reveal that a change in the genetic pool is insufficient to account for the temporal and spatial correlates of the increase of allergic manifestations and suggest that modifications of the epigenome (e.g. DNA methylation) and gene expression may be more important. Such modifications are thought to result from gene–environment interactions occurring during the developmental periods before and after birth [80], and in other transitional periods (puberty, menopause). Vercelli suggested that one of the major challenges facing geneticists is to understand how environmental and developmental factors interact with genetic determinants to increase disease susceptibility [17]. Epidemiologic studies investigating gene–environment interaction or studies analysing epigenetic effects seem to identify stronger risks [81]. Gene–environment and epigenetic effects are particularly topical. Compared with Mendelian genetic explanations, the strength of gene–environment interactions, in particular those including the gestational environment, has been known [82, 83]. However, epigenetic explanations have provided new concepts on how the memory of past exposures can be stored and thus how gene–environment interactions can maintain their effects even after caesura of the environmental exposure. The environment seems to play a key role in activating or silencing genes by altering DNA methylation, histone modification, and chromatin structure. Hence, we are observing a remarkable shift in epidemiologic allergy research. After 100 years of predominance of Mendelian models, the Lamarckian concept assuming that nurture can change nature (i.e. the epigenome), long thought of no relevance to modern research, is enjoying a quiet resurgence with epigenetic theories of inheritance [84]. Perera et al. [85] undertook one of the first studies to investigate the effect of gene–environment interaction related changes in DNA methylation, and asthma in children. During the third trimester of pregnancy, mothers completed personal prenatal air monitoring for polycyclic aromatic hydrocarbons (PAH) resulting from inner-city diesel exposure in New York, United States. Methylation sensitive restriction fingerprinting was used to analyse umbilical cord white blood cell DNA of 20 cohort children. Over 30 DNA sequences were identified whose methylation status was dependent on the level of maternal PAH exposure. Six sequences were found to be homologous to known genes having one or more 5′-CpG island(s) (5′-CGI). Of these, acyl-CoA synthetase long-chain family member 3 (ACSL3) exhibited the highest concordance between the extent of methylation of its 5′-CGI in cord white blood cells and the level of gene expression in matched foetal placental tissues in the initial 20 cohort children. ACSL3 was therefore chosen for further investigation in a larger sample of 56 cohort children. Parents were asked whether their child had a doctor’s diagnosis of asthma before age 5. Methylation of the ACSL3 5′-CGI was increased when maternal airborne PAH exposure was higher (OR 13.8 [95% confidence interval (CI) 3.8–50.2]), indicating that epigenetic marks of this gene are altered by exposure. Regarding the disease impact of the gene, methylation was found to be associated with parental reports of asthma symptoms in children. The sample size of this study is small and the specific genetic link with the disease was not replicated. However, this pioneering study provides a proof of concept that environmental exposure may change epigenetic marks such as DNA methylation, which in turn may be associated with allergic diseases. A future surge of publications on epigenetic associations is anticipated.

Risk Factors for allergic disease

Epidemiological studies are particularly suited to defining potential risk factors of allergic disease and 2009 saw steady flow of papers in both Clinical and Experimental Allergy and other allergy journals investigating this aspect of allergic disease. Another remarkable shift in allergy research can be observed. Whereas in the 1990s, iatrogenic effects of one specific asthma treatment were debated [86], in the first decade of the 20th century there is an emerging interest of iatrogenic effects resulting from treatments of other conditions, but affecting allergies. These include unwanted effects of folate application during pregnancy. One of the success stories in public health is the folic acid supplementation during pregnancy, which has repeatedly shown to reduce the risk of neural tube defects and other congenital malformations. Related public health measures include folic acid supplementation in pregnant women. Several countries, including the United States, fortify flour with folic acid to help ensure adequate blood levels. Because Norway does not, women take supplements of folic acid, which eases the assessment of folic acid by questionnaire in Norway compared with countries with folate fortification of food. Mechanistically, folate serves as methyl donor required for epigenetic changes. As part of the Norwegian Mother and Child Cohort Study, pregnant women were asked to record in 4-week period the supplements they used according to their supplement container. Wheeze (months 6-18) and lower respiratory tract infections (LRTIs, months 0–18) in offspring were used as outcome variables. The sample included 32077 children [87]. Folate supplements were used by 79.3%; 22.3% in the first trimester only, 13.8% after the first trimester, and 42.6% in both periods. About 40% of the children had wheezing between 6 and 18 months; 16.8% of the children suffered from LRTI. Comparing folic acid supplementation before and after 12 weeks of gestation, the study demonstrates a slightly increased risk of wheeze in children of mothers who used folate in the first trimester (relative risk [RR] 1.06; 95% CI 1.03–1.10) and of LRTI (RR 1.09; 95% CI 1.02–1.15). This study provides critical information from an unbiased large sample. One limitation is that respiratory symptoms in 0–18-month-old children are likely to be transient. However a more recent Australian study reported that folate acid intake – in late pregnancy – is also related to an increased risk of childhood asthma at age 3.5 years (RR 1.26; 95% CI 1.08–1.43) and 5.5 years (RR 1.17; 95% CI 0.96–1.42) [88]. It is obvious that the potential link between folate supplementation during pregnancy and childhood asthma needs further attention. Public health measures of folate supplementation, in the best intention, may have done harm, because unwanted effects of high folate supplementation were not assessed before it was introduced. We may have to balance risk and benefits of intended and non-intended effects and may have to focus supplementation on specific groups. Methodologically, the work demonstrates that large cohort studies such as the Norwegian Mother and Child Cohort Study are critical to address new emerging risk factors in an unbiased way.

Prenatal use of acid-suppressive drugs use is also attracting interest [89]. In western societies, about 25% of the population report having heartburn at least once a month and 12% at least once per week. In order to decrease heartburn, the major symptom of gastroesophageal reflux, often related to eating styles, three types of acid-suppressive drugs are frequently used: (a) Antacids perform a neutralization reaction and reduce the acidity in the stomach; (b) H2 receptor antagonists decrease the production of acid by epithelium cells in the stomach via blocking effects of histamine on these cells; and (c) Proton pump inhibitors block enzymes producing gastric acid. For instance, in 1999, approximately 10% of Norway’s population had used antacids in the last 2 weeks [90]. These drugs are among the most widely sold drugs in the world. Nevertheless, they have unwanted effects regarding allergy. Acid levels comparable to human stomach pH levels (1–3) are required to destroy the allergen potential of some foods. Murine and human models have demonstrated that reduction of gastric pH levels by antiacids, H2 antagonists, and proton pump inhibitors facilitate the production of food-specific antibodies.

The study population of mother and children born between 1995 and 2004 was extracted from the Swedish Medical Birth Register and this register was used to identify mothers who took acid-suppressive medication during pregnancy. Data from the Swedish Hospital Discharge were used – following International Classification of Diseases Codes – to define asthma and allergy cases (1995–2006). The Swedish Prescribed Drug Register was used to identify children born after 1995, who received prescriptions for asthma medications, antihistamines, steroid ointments, or epinephrine autoinjectors during the years 2005 and 2006. After exclusion of children with reported perinatal risk factors for allergy such as Caesarean section and preterm birth, the study population included 585 716 offspring that were stratified into three groups. Only 1% of the pregnant women received a prescription and purchased acid-suppressive drugs. The RR of any allergic disease in offspring was 7.17% in mothers who had acid-suppressive drugs during pregnancy and 5.01% in those who did not. This difference was statistically significant. The association was not depending on the type of acid-suppressive drugs (antacids, H2 receptor antagonists, or proton pump inhibitors). Childhood asthma comprised most of the allergic disease cases in offspring. No association with acid-suppressive drugs was found for other allergic manifestations. The large unselected sample and the lack of systematic misclassification using administrative data nicely demonstrate a moderate association between acid-suppressive drugs in pregnant mothers and asthma in their offspring. A limitation of this approach is the potential of non-differential misclassification (prescription and dispense) of this class of drugs, which is likely to underestimate the RR. Another limitation is the lack of data that could help to elaborate the specific risk. In an editorial discussing this paper, Kemp describes two pathogenetic explanations [91]. One possibility is that pregnant women using acid-suppressive drugs may have higher concentration of allergens, which then are transferred to the foetus. Another explanation is that women continue the use of acid-suppressive drugs after pregnancy leading to a transfer of these drugs to the offspring via breastmilk. Additionally, one may consider that gestational exposure to acid-suppressive drugs may alter epigenetic regulations in the developing foetus, which in turn may facilitate the development of asthma later in life.

Prenatal paracetamol use has been proposed as a risk factor for the development of asthma [92]. Paracetamol or acetaminophen is one of the most frequently used drugs among pregnant women and is considered the first-choice analgesic and antipyretic in most cases. For instance, data from the Danish National Birth Cohort (N~100 000 women) shows that 9% used acetaminophen during the first trimester, 21% during the second, 28% during the third up to time of delivery. The result of six studies suggest a moderately increased risk of eczema, asthma, hayfever in offspring of mothers with prenatal paracetamol exposure although one study did not support these associations [92, 93]. The latter study had the strength of prospective ascertainment of exposure during pregnancy in a large sample (1505 pregnant mothers), but the limitation, that two thirds of the pregnant women were recruited because of having asthma. Nevertheless, no stratification by maternal asthma status was provided. To separate whether maternal asthma modifies this association, information was collected from mothers of 1741 children in the Spanish province of Murcia (mean child’s age 4.1 years) [94]. In a self-administered questionnaire, data were gathered on childhood wheezing, the use of paracetamol during pregnancy, on maternal asthma, and alternative risk factors such as breastfeeding, pet exposure, and siblings. The proportion of participation was 61%. In non-asthmatic mothers use of paracetamol at least once a month during pregnancy was related to an increased odds ratio (OR) of wheezing in offspring (OR 1.94 [95% CI 1.15–2.61]). It was disappointing that only 3.6% of the participating mothers had asthma (n = 65). In this group, no association between paracetamol and offspring wheezing was detected.

This Spanish study provides evidence that paracetamol use may constitute a risk for offspring asthma if the mother does not have asthma. The sample of asthmatic mother size is too small to accurately assess the risk of their offspring. Because evidence against paracetamol is accumulating, future regulations may be required. It is remarkable that prenatal paracetamol but not aspirin use is associated with an increased risk for the offspring. Because exposure is avoidable, difference between paracetamol and aspirin, the role of maternal asthma, and risk variations with gender of the child urgently need further evaluation.

Caesarean sections have been implicated as a risk factor for allergic disease [95–97]. Rates of Caesarean deliveries have risen in most of industrialized countries from approximately 5% in the 1970s to over 30% in 2000 in some regions of the world. Compared with children born by vaginal delivery infants born by Caesarean section have a delayed microbial exposure; indeed, it has been shown that the intestinal flora in section deliveries differs from those with vaginal deliveries [98]. This setting is considered to setback the maturation of the neonatal immune system. Previous reports do not provide consistent evidence for an association between Caesarean section and allergy or asthma in offspring. Nevertheless, a meta-analysis found that, on average, Caesarean section increased risk of asthma in children by 22% [95]. In the PIAMA study [99], mothers were recruited during pregnancy in three regions of the Netherlands. A total of 4146 mothers who delivered their child between 1996 and 1997 participated in the study (proportion of participation 53%). Every year, parents were asked to report information regarding their child’s asthma. The 8-year questionnaire on asthma was returned by 3238 parents; after exclusion of missing data asthma information was available for 2917 and serum IgE measurement for 1454 children. Perinatal data were obtained from two self-administered questionnaires completed in the third trimester and 3 and 12 months after birth. A total 8.5% of 2917 children were born by Caesarean section; in 39.6% one parent had a history of allergy, and both in 9.1%. The prevalence of asthma at age 8 was 12.4%. An allergic sensitization was identified in 41% of the children (n = 1454). With increasing status of parental allergy the OR of Caesarean section on asthma increased, whereas the OR for having allergic sensitization decreased. Strengths of this study include the prospective collection of information, which reduces recall bias, and the large sample size. A weakness is that the opposed trends of the combined effects of parental allergy and Caesarean section for asthma and allergic sensitization were not deciphered. It would be of interest to understand whether the operation is associated with atopic or non-atopic asthma, and hence, whether Caesarean sections have an effect on lung development or the development of allergy. Future studies should distinguish whether the operation was elective or an emergency decision. In addition, in increasing number of sections are conducted in obese mothers, hence, it is possible that maternal obesity plays a role?

The effects of oral contraceptive (OCP) and hormone replacement therapies have been suggested as risk factors for allergic disease and sex differences in airway behaviour and airway disease have been documented [7]. Among women, early age at menarche is associated with adult asthma; irregular menstruation is related to lower lung function; and menopausal transitions affect asthma. Women carrying a female foetus had significant lower lung function during pregnancy than those carrying a male foetus. Hence, sex steroid hormone may influence airways and allergies, which motivates research on extraneous sex hormones in women, namely OCP use and hormone replacement therapy. In a follow-up study of a cross-sectional survey conducted in seven Northern European centres (1990–1994), 5791 women age 25–45 were selected to investigate associations between OCP and asthma, asthma symptoms, and hayfever [100]. Women using OCP had a higher prevalence of self-reported asthma, allergic asthma, asthma symptoms, and hayfever (ORs between 1.3 and 1.5). Because fat tissue produces sex steroid hormone and a higher body mass indicates changes in metabolism, the interaction of body mass index (kg/m², BMI) with OCP was analysed. The OR of having asthma related to OCP is low in underweight women (OR 0.41, and increases with increasing BMI to an OR of 1.91 (95% CI 1.2–3.02). Against that, the OR for the association of hayfever and OCP does not vary over the three BMI groups. Strengths are the unselected sample and large study population of women 20–45 years of age.

Limitations include the cross-sectional analyses and relying upon individual reports. However, the authors could rule out a detection bias, because the relation of women with asthma symptoms that were diagnosed with asthma did not depend on taking OCP or related doctors’ visits. The different trends of the asthma-OCP and the hayfever-OCP relations over the three BMI groups need further clarification. Are these associations related to allergic sensitization or the reactivity of bronchial tissue? Are these associations driven by to endocrine or metabolic changes? Surprisingly, in another study, use of prenatal OCP was found to be associated with asthma, wheezing, and coughing in offspring 11–12 years of age [101]. This may suggest that the risk is restricted to women but may be transmitted from mothers to offspring. Hence, given this large Northern European cohort, it would be of foremost importance to investigate whether the combined OCP-BMI effect can also be detected in offspring of the female participants. Gender differences in allergic disease are not restricted to asthma. Soost et al. [102] using double-blind placebo-controlled challenges found that food allergy in Germany was twice as common in women as men despite similar levels of sensitization.

In a modification of the original observations of David Strachan which led him to posit the hygiene hypothesis Matheson et al. [13] found that increased numbers of siblings was much more protective against early than late onset rhinitis.

In the course of delivery, after rupture of the membranes, amniotic fluid becomes colonized with microbes. Amniotic fluid microbial invasion is common in women who underwent operations after the onset of labour and rupture of the membranes (70%) and less frequent in those undergoing operations with intact membranes (20%) [103]. To investigate the effect of intrauterine microbial invasion for the development of asthma and allergic sensitization, Kerki-Nisula et al. [104] followed offspring of mothers who delivered their infants by Caesarean section between 1990 and 1992 in Finland. During the operation, amniotic fluid was aspirated or intrauterine swab samples were taken for bacterial cultures. Postal questionnaires were mailed to the mother, when the children were 15–17 years of age. The study sample with complete data consisted of 460 children. A sub-sample was invited to participate in a skin prick test (SPT); 323 children participated. Microorganisms were identified in 31.5% of the amniotic or intrauterine samples. In children whose maternal samples were colonized with Streptococcus species or by potential pathogenic anaerobic bacteria the ORs for reporting doctor-diagnosed asthma were 2.53 times (95% CI 1.19–5.38) and 4.51 times (95% CI 1.56–13.0) higher, respectively. The risk was further elevated if the neonate received antibiotics. No statistically significant associations were found between maternal infection during delivery and allergic sensitization of the children. Limitations of this pioneering investigation include a loss of follow-up of children (460 or 323 of 749 eligible offspring). Another limitation is that all children were delivered by Caesarean section. Comparisons with bacterial colonization in vaginal deliveries would improve the evaluation of risks. The study suggests that microbial exposure during delivery poses a risk for offspring asthma. The authors also discuss an alternate explanation, namely that mothers or neonates with a susceptible immune system may be more prone to develop infections during delivery and respiratory infections, which may explain the association of the perinatal bacterial exposure and asthma in offspring. Future studies need to explore the role maternal and offspring infections and/or investigate genetic polymorphisms or epigenetic marks that are related to an increased susceptibility to infections.

The role of maternal fat consumption during pregnancy for the development of allergy, eczema, and asthma is in dispute [105]. The Osaka Maternal and Child Health Study is a prospective cohort study [106]. Women were recruited during pregnancy. Of the 1002 women, 867 mother-child pairs participated in the second (2–9 months post-partum), and 763 in the third survey (16–24 months post-partum). During pregnancy, information was gathered on diet using a self-administered, semi-quantitative diet history questionnaire. Estimates of daily intake were calculated and compared with a 3-day dietary diary. Another questionnaire asked about potential confounding variables including maternal intake of vitamins D and E and paternal and maternal history of asthma, atopic eczema, and hayfever. In the second and third survey, information on wheezing, asthma, and eczema were collected. Adjusting for confounders, the OR of diet variables for the development of wheezing and eczema were estimated by comparing the fourth quintile of the diet variable distribution with the first quintile. Of the 763 children in the third survey, 22% had wheeze and 18.6% had eczema. The results suggest that maternal intake of α-linolenic acid and docosahexaenoic acid during pregnancy may be preventive against childhood wheeze (OR 0.52 [95% CI 0.28–0.97]; OR 0.37 [0.15–0.91], respectively). The intake of n-6 polyunsaturated fatty acids seems to increase the risk of eczema in offspring (OR 2.25 [95% CI 1.13–4.54]). The strength of the study is the prospective design of the study, which reduces recall biases of food items. Regarding selection biases the moderate proportion of follow-up is balanced because central maternal variable did not differ in continuing participants and drop-outs. One limitation is the high number of statistical tests without adjustment for false discovery rate. Another limitation is the approximation of food consumption using questionnaires. Because we do not understand the effect of specific food items and their interaction with pre-existing conditions, future studies need to take pregnancy and delivery conditions into account and take repeated measurements of critical markers in maternal serum and breastmilk.

Pesticide use has been linked with respiratory disease in several studies and farmers are more likely to be diagnosed with asthma than other occupational groups. The Agricultural Health Study is a large prospective investigation of Iowa and North Carolina pesticide applicators, mostly farmers, and their wives [107]. All male applicators age 20 year and older with complete information on asthma history and potential confounders, including smoking, height and weight, and respiratory diseases were included. Individuals diagnosed with asthma before age 20 were excluded. The applicators provided detailed information on lifetime pesticide use for 50 pesticides and high pesticide exposure events. The information was used to develop lifetime exposure matrices. Of the 19 704 farmers included in the analysis, 2.2% reported adult-onset asthma with 0.6% allergic and 1.6% non-allergic asthma. A history of high pesticide exposure events was associated with a statistically significant twofold increased odds of having asthma. For allergic asthma, three herbicides (2,4,5-TCP, EPTC, and paraquat), three organochlorine insecticides (chlordane, heptachlor, lindane), three organophosphate insecticides (diazinon, parathion, and coumaphos), one fungicide (captan), and two fumigants (ethylene dibromide/carbon tetrachlorine and carbon disulphide) were statistically significantly associated with allergic asthma. Non-allergic asthma was statistically related with the use of one herbicide (petroleum oil), the organochlorine insecticide DDT, and two organophosphate insecticides (phorate and malathion). A limitation is that the study relied on self-reported exposure and asthma data. Another limitation is the lack of lifetime farming exposure to animals, hay, grain, which also may have contributed to asthma. Surprisingly, although allergic asthma is less common in adults than in children and far less common in farmers than in other occupational groups, the associations with pesticides were much stronger for allergic than non-allergic asthma.

Conclusion

In recent years, various publications showed an evolving interest in adverse effects of medication and treatment during pregnancy and delivery on the risk of allergy and asthma. Folate supplements, use of paracetamol, and acid-suppressive drugs during gestation and Caesarean sections constitute frequent events. If further evidence supports the presented findings, then a large proportion of asthma may be attributed to iatrogenic effect established during the perinatal period. There is no doubt that this proportion belongs to the group of avoidable diseases. Hence, the challenge for allergy researchers is to initiate discussions with obstetricians. Public health officials and obstetricians need to be made aware of a possible increased risk of asthma and allergy in offspring that originate during gestation and delivery.

It is peculiar that the studies presented on acid-suppressive drugs during pregnancy and on Caesarean section reported an increased risk of offspring asthma, but not of other allergic manifestation. Research needs to be done to understand the pathogenesis of the different perinatal exposures. Another peculiarity is the role of maternal or paternal preconditions. The association of Caesarean section with offspring asthma seems to be moderated by parental allergy; and the adverse effect of paracetamol use during gestation may be restricted to non-asthmatic mothers. Considering that peculiarity indicate that we do not yet sufficiently understand the role to pre-existing parental conditions during pregnancy, then conclusions about perinatal risk factors (acid-suppressive drugs, paracetamol, and Caesarean section) may be premature.

Another evolving research area is the role of the intestinal microbial flora in infants, whose role is to support the development of the immune system. It is possible, that the common risk factor of Caesarean section and continuing acid-suppressive drug use during lactation is an effect on the microbial gut flora of the infant. Again, we have to connect all dots before we can provide substantial evidence of iatrogenic effects.

To improve our understanding of perinatal and microbial conditions, it is necessary to improve and expand routine sample collection in neonates and infants. Stool samples are easily accessible and storable samples for future studies. Collection of cord blood, placental, and breastmilk samples will provide precious information for future studies, in particular to understand if and how epigenetic changes and immune maturation go hand in hand. This knowledge may be able to help us to prevent childhood or adult development of immune diseases by improving gestational conditions.

To advance the epidemiology of allergy, two types of investigations are needed. Large investigations such as the Agricultural Health Study or the Scandinavian register studies can identify environmental and iatrogenic exposure. However, large projects need to be complemented by smaller studies collecting details of exposure, genetic and epigenetic conditions, such as the pioneering work of the inner-city study on polycyclic aromatic hydrocarbons in New York. In an ideal design, both aspects would be combined and matched by more detailed disease information, which in turn would facilitate a better understanding of the pathogenesis of allergy.

Experimental models of allergic disease

Model systems

Experimental models of allergic disease continue to be a valuable resource for investigating the mechanisms underlying allergic pathology. As such these models need to reflect the human clinical phenotype as closely as possible. The classical OVA driven model is widely used and continues to be refined in order to improve the applicability of the model for investigation. One drawback to this model has been the use of the adjuvant alum, because this promotes mast cell independent inflammation. A comparison of an adjuvant vs. adjuvant free protocol has revealed that subcutaneous OVA sensitization in the absence of alum generates a phenotype that is comparable to the commonly used alum protocols [108]. This system now lends itself to the study of mast cell dependent pathways, an important facet of the human disease.

The particular contribution of IgE to allergic pathology has been investigated in a number of different models, including those induced by OVA/alum and chemical haptens. IgE antibodies were found to enhance pulmonary inflammation caused by inhalation of the contact sensitizing hapten trinitrobenzene sulphonic acid (TNBS). In this study mice deficient in IgE had reduced airway hyperresponsiveness (AHR) and airway inflammation after inhalation of TNBS [109]. In contrast in an OVA model of experimental asthma IgE was found to play a central role in mediating the allergic immune response, in particular the immunoglobulin classic antigen-binding site was shown to play a crucial role in creating an interface between the allergen and IgE, suggesting for the first time that specific IgE can regulate these immune responses [110]. Analysis of serum and monoclonal IgE antibodies in the NC/Nga dermatitis prone mice revealed multiple clones that showed specificity for the autoantigen histone H3. Moreover, immunologically detectable histone H3 antigens could be detected in skin sections from dermatitis sites, suggesting that anti-histoneH3 auto-antibodies may contribute to the development of dermatitis [111].

An increasing trend in the literature is to develop models, which use common environmental allergens to investigate mechanisms. Inhaled HDM has been used to induce a allergic inflammation and AHR and has the advantage of generating pathology in naive mice without the need for prior peripheral sensitization with adjuvant. Prolonged exposure to inhaled HDM extract induces a robust eosinophilic inflammation in conjunction with AHR and a polarized pulmonary Th2 response [112]. Cockroach allergen was found to have similar effects [113]. These models highlight the contribution of the epithelium to development of allergic pathology because the pulmonary epithelium represents the first point of contact with allergen. Molecules secreted at this site in response to allergen exposure exert a critical role in driving the innate and adaptive immune response to inhaled allergen [114].

Genetic and environmental influences

Evidence from clinical studies suggests that environmental factors influence the course of allergic reactions and so experimental models need to take into account age, obesity, sex differences. A number of studies in the Journal have used a range of experimental models to assess genetic and environmental influences on the development of some of the cardinal features of the allergic response. Genome wide association screens have identified a number of genes thought to influence susceptibility to asthma. Similarly contrasting mouse strains have revealed differential responses in IL-13 and TGFβ following allergen exposure. The asthma ‘susceptible’ strain BALB/c and asthma ‘resistant’ strain C57Bl/6 both developed Th2 inflammation after OVA challenge, although only BALB/c mice showed evidence of airway remodelling. Rather than differences in IL-13 itself there were shown to be differences in the activation of downstream signalling pathways, including STAT6 and SMAD2, suggesting that these may be more relevant than inflammation alone as a measure of disease progression [115]. Diet is increasingly thought to influence disease development and obesity has been associated with reduced airway function and asthma. De Vries et al. [116] used an experimental model to show that a high-fat diet redirects local immune responses to allergen in the lungs and leads to attenuated pulmonary eosinophil recruitment. Although asthma is a common childhood disease, the vast majority of experimental models are conducted with adult animals. Several studies have sought to redress this using neonatal or juvenile mice to determine mechanisms. Early life administration of cockroach allergen was found to induce airway inflammation and remodelling which was augmented with the co-administration of the TLR 4 ligand endotoxin [113]. Similarly early life sensitization to allergen was found to be enhance by co administration of TLR 7 ligands such as viral ssRNA [117]. Endotoxin is a common component of bacterial cell walls and rhinovirus and RSV are ssRNA viruses common in childhood. Similar studies in adult mice have shown that infection with low dose Mycoplasma pneumonia infection enhances allergic inflammation; due to a reduction in prostaglandin E₂ which was found to be an endogenous inhibitor of airway inflammation in this model [118]. In contrast gastrointestinal helminth infections were associated with a significant inhibition of airway inflammation due to an increase in regulatory pathways such as IL-10 [119]. Taken together these studies underscore the importance of considering age as an influencing factor in development of allergic responses, but also show that allergen sensitization is affected by the presence of infections agents which activate pattern recognition receptors on the epithelial surface. Indeed, these effects may occur even prenatally because maternal nasal exposure to bacteria during pregnancy results in protection of the offspring from OVA-induced airway inflammation but required intact TLR signalling in the mother [120].

Tolerance to harmless inhaled particles is achieved in the lung via a complex array of regulatory cells and molecules. It is thought that breakdown in any of these pathways leads to sensitization to otherwise innocuous particulate such as pollen or house dust. Mouse models have been used to understand how these pathways regulate the response inhaled allergen. Depletion of CD4-CD25 regulatory cells impairs induction of tolerance in an OVA model of asthma [48], while suppressor of cytokine signalling 1 was found to be a physiological regulator of the allergic response in vivo [121]. The potential to manipulate these pathways for patient benefit is an attractive prospect, but a careful understanding of the underlying mechanisms is vital.

Clinical allergy

Four topics dominated the Clinical Allergy section of the journal in 2009, IT, allergy investigations, new treatments and food allergy.

Immunotherapy

In patients with a combination of seasonal asthma and rhinoconjunctivitis Kopp et al. [39] in a randomized double-blind and placebo-controlled study reported that a combination of omalizumab and IT was more efficacious than IT alone in terms of symptoms, asthma control and quality of life (QoL). One of the issues facing an allergist considering IT is whether polysensitization will impact on efficacy. In an important contribution Malling et al. [38] demonstrated that there was no difference in the efficacy of treatment with a Stallergenes grass pollen preparation between monosensitized and polysensitized individuals. In the same issue Horak and colleagues confirmed the efficacy of the 300IR dose of the Stallergenes grass pollen sublingual IT preparation and Stelmach and colleagues demonstrated that a high dose ultra rush regime with the Stallergenes preparation was effective and safe in children [122, 123]. European allergy continues to be very innovative in trying to maximize the benefit of IT [124, 125]. This approach was typified by Senti et al. [126] who modified the A-type CpG oligodeoxynucleotide which is a stronger induced of IFN-γ but less stable than the B type. However by packaging it with virus-like particles they have been able to stabilize the adjuvant. They reported an early-phase clinical trial showing that this approach appeared efficacious and safe. In a similar vein while sub-cutaneous IT with allergen extracts is an effective form of treatment for severe rhinitis, it carries a risk of serious adverse events. Modified allergens which are thought to tolerise T cells without cross-linking mast cell/basophil bound specific IgE, are considered to have a better side-effect profile, but uncertainty about their effectiveness persists. Majak and colleagues tried a different approach to using adjuvants to improve the efficacy of HDM IT in children with asthma by giving it in combination with steroids and vitamin D based on work showing this combination induces Treg cells. However they found the steroids if anything impaired effectiveness [43, 127]. Ceuppens et al. [128] have undertook a blinded and randomized trial of a gluteraldehyde modified birch pollen vaccine. After a rapid up-dosing schedule and 18 months of treatment there was a significant improvement in a global symptom and treatment score with a very good side-effect profile in an admittedly small study. More options to consider when choosing which IT regime is right for your patient. The efficacy of IT for venom-induced anaphylaxis is well established although there is often a delay in referral for a specialist opinion [129]. Patients who have experienced reactions after being stung often understandably become quite phobic about exposure to wasps and bees and therefore it would be expected that venom IT might also impact on QoL and point discussed by Heddle and Brown [130]. The group of Dr Dubois have previously shown this to be the case in patients who have suffered major systemic symptoms. Patients who have only had urticaria and angioedema are not always offered IT. The group therefore compared QoL outcomes in patients undergoing IT or just given adrenaline auto-injectors and found that patients who underwent IT had a much better outcome in terms of QoL [131]. It would be interesting to know if that was also the case for people who had just large local reactions where IT is not usually recommended. The safety of subcutaneous immunotherapy (SIT) continues to be an important subject of debate especially with the increasing popularity of sub-lingual treatment which appears to have a much lower risk of systemic side-effects. Further information was provided by a large prospective audit of systemic reactions to SIT for respiratory allergy from Schiapolli and colleagues which reported the experience of 11 Italian clinics. Just over 3% of patients had anaphylaxis with no fatalities. Half occurred during the maintenance phase. Once again asthma was shown to be an important a risk factor. The authors felt that this was an acceptable level of side-effects. Whether that is the case or not this paper provides hard data on which to allow patients to make an informed decision on whether to have this treatment or opt for a sublingual approach [37].

Investigations

There is increasing interest in the link between venom-induced anaphylaxis and mastocytosis [132]. Potier et al. [133] undertook a retrospective analysis of the relationship between venom anaphylaxis and serum tryptase and demonstrated a distinct clinical pattern in those with a raised tryptase (several of whom had previously undiagnosed mastocytosis), with more severe reactions with marked flushing but less urticaria. The basophil activation test (BAT) is increasingly used in a clinical setting to complement skin testing particularly where measurement of specific IgE is often unhelpful such as drug allergy. In an extension of the body of work on this test Gomez et al. [134] compared BAT with SPT in patients with definite immediate type allergy to pyrazalones a chemical structure that is found in a wide range of medicines. They found as is common with BAT that it was good, but not quite good enough with 33% of patients who were SPT negative (where such a test would have most clinical usefulness) being positive on BAT. Along the same lines hypersensitivity to non-selective COX inhibitors such as aspirin remains a fascinating, important, but largely unexplained phenomenon. The immediacy of the reactions strongly points towards either a mast cell or basophil mediated process, but direct evidence is lacking. A number of studies have suggested that the BAT is useful in helping to diagnose this type of hypersensitivity. However a careful study by Celik et al. [135] demonstrated that the test is not sufficiently specific to draw any meaningful conclusions from the results, especially at the dose of aspirin commonly used in the investigation. Similarly Seitz and colleagues found that the BAT was not helpful in diagnosing fluoroquinolone hypersensitivity although skin tests were confirmed to very helpful in diagnosing penicillin allergy [136, 137]. In contrast Kerosec and colleagues investigated the utility of the BAT in patients with a good history of venom anaphylaxis but negative specific IgE and SPT. They found that in these difficult cases which made up about 4% of their practice that the BAT was often positive even when the intradermal test was negative and had good specificity [138, 139]. There continues to be considerable interest in using SPTs and specific IgE levels to predict the likelihood of a positive food challenge. Diequez et al. [140] have added to this debate by demonstrating that in egg allergic children the presence of a 7 mm weal of a sIgE of >1.3 KU/L to egg white means that it is highly likely that the person will be symptomatic on exposure to egg so there is no need for a challenge. Very useful information for a busy clinician. Skin testing is a cheap and effective way to determine the presence of specific IgE and together with a consistent history enable a confident diagnosis of type I allergic disease. However the results are only as good as the skin test solution. There is an assumption that a commercially available solution will have proven reliability with little to choose between manufacturers, but the less commonly used allergens may not have undergone rigorous field testing. This is clearly demonstrated by van Kampen and colleagues who in a very thorough and detailed study have investigated four sources of skin test solutions for wheat and rye flour in bakers. They have found considerable variability in their reliability compared with in vitro testing or challenge, with the sensitivity, perhaps predictably, largely depending on the protein content of the solutions. This emphasises the importance of allergen standardization. In the meantime don’t take your skin test solution for granted [35]. An in vivo investigation for AR is the nasal provocation test. The gold standard is allergen challenge but this requires a considerable degree of expertise. An adenosine mono-phosphate challenge is easier and is thought to be mediated by mast cell degranulation but does it give the same information particularly regarding the efficacy of antihistamines. The paper from Brian Lipworth’s group suggests that in terms of the response to antihistamines AMP and allergen challenge give very similar responses which is useful information for those developing new treatments for AR [141]. Similarly Torkildsen et al. [142] validated conjunctival allergen challenge as a way to determine response to treatment with antihistamines. Anaphylaxis represents the purest expression of type 1 immune responses and follows a classical early and late response pattern seen after allergen challenge. Late responses can be dangerous if a patient has been discharged and not under medical observation. Mehr and colleagues asked whether there are any features that could predict the likelihood of a late-phase reaction in children presenting to the emergency room with anaphylaxis. Eleven per cent of children went on to have a late response. If there was no need for a second dose of adrenaline or fluid resuscitation there was a 99% chance of not having a late response whereas all anaphylactic late responses were associated with at least one of these interventions [143]. A useful guide to management in the emergency room where a family may be keen to get home after a successfully treated anaphylactic event.

Treatments

There continues to be considerable interest in the potential value of probiotics in both preventing and treating allergic disease particularly eczema, an area reviewed by Johannsen and Prescott and the subject of a Cochrane meta-analysis by Boyle and colleagues [144, 145]. The potential role of homoeostatic gut organisms in protecting against allergic disease continues to be a topic of considerable interest to allergists. However the overall message remains confusing both in terms of whether the endogenous gut flora is important in determining whether you become allergic and whether supplementing diet with probiotics reduces the risk of developing allergic disease. There is considerable complexity involved in this area. In terms of the use of exogenous probiotics there are issues of dose, type of organism, timing of treatment and outcome measures all of which have several permutations. In terms of the gut flora many studies techniques which can accurately reflect the complexity of the gut microbiota are only just becoming widely used. In addition although there are plausible hypotheses to provide an immunological explanation for an effect of gut flora and probiotics on allergy risk this is not yet a coherent and consistent story. Sjögren and colleagues examined gut flora using quantitative RT-PCR to measure amounts of indicator species. They found that allergy was less common in children with a more diverse gut microbiota. However this was related to other known protective factors such as endotoxin exposure and family size so it remains to be seen whether it is causal or simply an association [54]. A study by Soh et al. [146] examined the effect of probiotic supplementation to the diet of children during the first 6 months of life on eczema. In this carefully designed and executed study they found no benefit compared with placebo which would suggest that there is no value, at least in terms of preventing eczema, in adding probiotics to infant diet. Shibata et al. [147] investigated the benefits on ketose a fructo-oligosaccharide which stimulated the activity of bifidiobacteria on atopic dermatitis in infants. They found a significant improvement in the ketose group compared with placebo in the SCORAD score although this was not obviously linked to an effect on the count of bifidobacteria. A very novel approach to treatment of allergic disease is the use of hookworm infection which in field studies is associated with lower rates of atopy and allergic disease [reviewed in [148]]. In an initial safety study Feary et al. [149] found that that larval passage through the lungs had no effect on AHR on lung function in people with AR and AHR but without clinical asthma. This paved the way for a subsequent clinical trial of hookworm infection reported this year [150].

Food allergy

Urticaria is a common and distressing condition and many patients feel that food is playing a role. A number of food additives have been implicated in particular salicylates but also food dyes, preservatives and aromatic volatile compounds which have been termed pseudoallergens [151]. However the evidence that these compounds play a role is limited and controversial [152]. Bunselmeyer et al. [153] addressed this difficult subject by firstly trying to induce remission by putting patients on a strict pseudo-allergen free diet and then re-introducing foods in a structured manner. They found that 17 of 100 of patients gained complete remission the great majority of whom developed urticaria on re-introduction of the offending food additive. Good evidence that a careful attention to additive avoidance really is worthwhile. Food allergy has a major effect on QoL yet until recently there have been relatively few food allergy questionnaires. A number have now been developed including a self-administered one for children reported by Flokstra-de Blok et al. [154] which should prove useful in research into this increasingly common problem. The authors also discuss the methodological problems associated with the development of QoL instruments in a letter written in response to an article by Leung and colleagues reporting on the effects of food allergy on QoL in China [155, 156]. The idea that peanuts (Arachis hypogaea), are nuts is very ingrained in the general psyche and it is important to be constantly reminded that it is a member of the legume family (Fabaceae) and that cross reactivity with other members of the legume family are more likely than with tree nuts. Lupin (Lupine in North America), seeds and flour are being increasingly used in food preparation as an alternative to soy products. Lupin is also in the legume family although in a different genus (Lupinus) to peanuts (Arachis). Nonetheless allergy to lupin flour is increasingly reported. Fiocchi et al. [157] investigated the important issue of how commonly patients with peanut allergy are likely to be allergic to lupin flour. They found that two out of 12 children with peanut allergy were allergic to lupin containing pasta in a DBPCFC. A real problem therefore that emphasises the importance of including lupin flour as an ingredient on food labelling. Sensitization does not always mean disease and in food allergy the gold standard for deciding if someone is allergic is a food challenge which is time consuming and carries a potential risk. Less invasive markers of the likelihood of symptoms in the presence of sensitization would therefore be very welcome. Most interest has focussed on the level of specific IgE or in some cases the SPT response but both are far from perfect. In this deceptively simple study Tripodi and colleagues found that a threshold dilution of 1 : 256 gave a remarkable sensitivity and specificity for discriminating between a positive and a negative challenge of 95% and 100%, respectively. This offers a simple and convenient alternative for the practising allergists that avoids the stress and resources required for a food challenge and provides data on which to base advice to parents who are unwilling to proceed to a challenge [158].

Allergens

Three major topics have been covered in this section during the last year: (1) the identification, characterization and determination of biological activity of allergens from plant foods, pollens, mites and ant sting [36, 159–166]; (2) the use of molecular markers and the biochemical modification of pollen and HDM allergens to enhance diagnosis and IT [167–172], and (3) the effect of proteolytic digestion and heat treatment on the allergenic properties of plant foods and cow’s milk proteins [173–176].

Allergen characterization

Non-specific LTPs (nsLTPs), particularly peach Pru p 3, are the main allergens causing food allergy in the adult Mediterranean population. Lauer et al. [160] and Sirvent et al. [161] identified novel nsLTPs from A. hypogaea (Ara h 9) and Sinapis alba (Sin a 3) as major allergens in peanut- and yellow mustard-allergic patients from Spain, respectively. Both nsLTPs showed strong cross-reactivity to peach Pru p 3. Profilin Sin a 4 was also described as yellow mustard allergen. Palacin et al. [36] focused on recombinant nsLTP Tri a 14, an inhalant allergen from wheat flour, as a molecular tool for the diagnosis of baker’s asthma, based on its equivalent IgE binding capacity, and heat and proteolytic resistance, to those of its natural counterpart.

Seed storage proteins belonging to the 2S albumin and 11S globulin family have also been studied in the context of plant food allergy. Ahn et al. [162], described two major Pistachia vera allergens (Pis v 1 and Pis v 2) from these proteins families in pistachio-allergic patients, the isolation of their corresponding cDNAs, and cross-reactivity with their homologous allergens from cashew nut. Porterfield et al. [159] and Blanc et al. [163], evaluated the effector activity of peanut allergens by measuring degranulation of humanized rat basophilic leukaemia cells sensitized by serum IgE from peanut-allergic subjects. Ara h 2 and Ara h 6, both 2S albumins (conglutins), accounted for the majority of effector activity found in peanut extracts. Regarding pollinosis, Gastaminza et al. [164], detected allergenicity of pine (Pinus radiata) pollen in the North of Spain, as well as an unidentified major 42 kDa allergen and, as expected, a high level of cross-reactivity with pollen from other Pinus species.

Allergens of animal origin are represented by isoforms of Blo t 12 from the tropical mite Blomia tropicalis investigated by Zakzuk et al. [165]. Two isoallergens isolated from mites collected in different areas (Colombia and Singapore), showed differences in IgE-binding capacity, as well as in the amino acid sequence of a putative major IgE epitope. Thus, potential targets for site-directed mutagenesis to produce hypoallergenic molecules were located. Additionally, Lee et al. [166], identified and characterized a 23 kDa protein, homologous to members of the antigen 5 family of proteins from Hymenoptera, as the major allergen of Pachycondyla chinensis in ant sting anaphylaxis patients.

Immunotherapy

Several reports have been devoted to pollen allergy diagnosis and IT. Barber et al. [167], undertook a large (1329 pollen-allergic patients), geographically diverse survey in Spain to compare the results from an in vitro method testing a panel of isolated allergens with those from SPTs using crude extracts and purified pan-allergens (profilin, polcalcin, and LTP). They found that SPT with conventional extracts failed to diagnose patients with sensitization to pan-allergens, and that routine SPT to them is a way to improve diagnostic efficacy. Regarding IT, Hejl et al. [168], reported on an extensive T and B cell cross-reactivity towards the allergens of the Pooideae grasses (i.e. Phleum pratense, Lolium perenne, Dactilys glomerata, etc), and that IgG4 induced by IT with P. pratense would recognize allergens from most of these species. Therefore, treatment of grass pollen-allergic subjects with pollen extract of only one species (i.e. P. pratense) would protect against allergic responses induced by most of other temperate grass species. Chemical modification of allergens to reduce IgE-binding, but retaining their immunogenicity, has been suggested to improve safety of vaccines. This strategy was illustrated by acetylation of the major mugwort pollen allergen Art v 1 carried out by Perovic et al. [169], and explored by Carnes et al. [170], in depigmented allergoids of Betula alba. Relevant inhalant animal allergens, such as HDM (Dermatophagoides pteronyssinus) Der p 1 and Der p 2, have been also modified to better fulfil requirements for future safer specific IT. Thus, Asturias et al. [171], engineered hybrid molecules derived from Der p 1 and Der p 2, which had lower IgE reactivity and potency to induce cutaneous reactions, but provoked higher T cell proliferation responses, than the natural mite allergens. Immunization of mice showed the immunogenicity of the hybrid molecules by their capacity to induce IgG responses against both natural allergens. Burtin et al. [172], have explored an alternative way to produce hypoallergenic molecules in the near future. Different modified forms of Der p 1 bearing mutations within either the N-glycosylation sites and/or the cysteine protease-active site were successfully expressed in a tobacco plant system.

Food allergens

The potential significance of food allergen digestion by simulated gastrointestinal fluids are still controversial. In this context, Bøgh et al. [173], showed that peanut Ara h 1 was broken down in an in vitro model of the human gastrointestinal digestion process, but the resulting peptide fragments retained the capacity to sensitize Brown Norway rats, being capable of inducing specific IgG and IgE antibodies. A differential effect of heat treatment on the allergenicity and susceptibility to enzymatic digestion of cow’s milk allergens β-lactoglobulin and α-casein has been reported by Morisawa et al. [174]. Heat treatment impaired B-cell epitopes of the former by directly causing conformational changes and by increasing its susceptibility to digestion. In contrast, α-casein was resistant to heat treatment, but highly liable to suffer enzymatic digestion by pepsin even before heating. Adachi et al. [175], also found that food processing affects pepsin resistance of soybean β-conglycinin, with this allergen completely hydrolysed in soy milk whereas it remained almost intact in tofu. Finally, Worm et al. [176], showed that heat processing (roasting) of native hazelnut reduced its allergenicity, as indicated by different ex vivo and in vivo tests in hazelnut-allergic patients. On the other had, challenge with encapsulated hazelnut flour to circumvent the oral mucosa elicited oral allergy syndrome in the majority of challenged patients, suggesting this reaction as part of a systemic response.