香京julia种子在线播放

By characterizing molecular components of a given allergen that are recognized by patients' IgE, molecular allergology (MA) can be used to confirm suspected IgE-mediated food allergy (12). One of the main strengths of MA is that it allows for differentiation between co-sensitization and cross-reactivity, improving accuracy (13). The typical example is sensitization to plant foods secondary to pollen allergy—detecting IgE to the cross-reactive component (e.g., Ara h 8, Cor a 1) helps differentiate from primary systemic food allergy, associated with sensitisation to primary food allergens (for example, seed storage proteins Ara h 1/2/3/6 and Cor a 9/14). Further, MA has applicability to risk-stratification, helping to guide clinical management. Despite this, MA has demonstrated comparable specificity, but lower sensitivity, than OFC so far, limiting its value for primary diagnosis, with sensitivity of 67% and specificity 93% for casein, 74% and 91% for hen's egg ovomucoid, 82% and 92% for peanut Ara h 2 respectively (14). Exact values of sIgE titres are of limited clinical value, as severe reactions may still occur with sIgE of low levels and high affinity while conversely sIgE can also be present in the serum of resolved patients (15). Cost of MA is a limitation compared to SPT, though integration of novel multiplex platforms may help to minimize economic barriers, in the cases that require testing to many allergen components.

The BAT assesses degranulation of basophils in vitro following stimulation with the allergen. Basophils drive acute allergic responses via IgE recognition of the allergen and activation via the high-affinity IgE receptor. Following stimulation by allergens, flow cytometry is used to look for upregulation of the key markers CD63 and CD203c as a proxy for basophil degranulation. BAT has shown excellent diagnostic performance in clinical trials, having the potential to reduce the need for OFC (16). BAT has improved specificity compared to SPT and sIgE, while retaining sensitivity- limiting the number of false positives- with optimal cutoffs showing 98% sensitivity and 96% specificity for peanut (17). Further, BAT has potential utility for predicting disease trajectory, with some studies using repeated BAT to monitor acquisition of desensitization and remission of the allergic response (18). The ability of BAT to categorize allergic disease phenotypes through threshold dose reactivities offers opportunities for increased granularity of diagnostic approach, facilitating personalised care (19). Some of the barriers to further integration of BAT into clinical practice include the ca. 10% of individuals with basophils selectively unresponsive to FcεRI-mediated signalling. As an ex-vivo test on fresh (whole) blood, BAT features logistical difficulties in terms of procuring the sample from patients, requiring time-constrained transport to the lab and processing.

Various approaches have been suggested attempting to improve feasibility and deliverability of OFC. In patients with minimal risk of reacting to the allergen, cumulative OFC has been adopted to reduce the time involved and the risk of desensitization during OFC, which has been associated with false-negative OFC. Cumulative OFC consists in a single dose OFC which resembles the age-appropriate portion size. Another approach often adopted for baked milk and baked egg is the low-dose OFC, which offers a way to establish a dose below which patients can continue eating the food at home. A lower dose has been used in a standard OFC, with one study defining the dose as 1%–10% of usual OFC dose (20, 21). The danger of a low dose OFC is the fact that the threshold of reactivity can change over time and allergen-content of foods is not always declared, thus care needs to be taken to avoid the allergen in circumstances known to be associated with a reduction in allergen threshold (e.g., infections, sleep deprivation, intense exercise, menstruation, etc), and to ingest foods with known allergen content that is within the recommended range for that patient.

Similarly to the BAT, the mast cell activation test (MAT) uses flow cytometry to observe the in vitro activation of mast cells to ascertain the degree of allergic response to particular allergens. Diagnostic performance of the MAT has been shown to exceed the SPT with an AUC of 0.99 compared to 0.73 for SPT (24). Mast cell lines or primary mast cells derived from peripheral blood hematopoietic stem cells are passively sensitized with the patient's serum or plasma, then incubated with allergen to assess response. The requirement for serum/plasma rather than fresh blood is a key distinguishing factor between BAT and MAT, potentially improving the feasibility of MAT and allowing for testing samples collected prospectively over time in the same experiment. However, growing mast cells can be challenging and primary cells are difficult to isolate, require tissue or large volumes of blood and are highly variable, with passive sensitization limiting sensitivity significantly (25). Requiring transfer of IgE to cells that are not autologous, MAT has lower sensitivity than the BAT with one study showing 75% to peanut compared to over 83% for BAT, so could be more useful in a sequential process following BAT, especially in cases of basophil-non-responders (26, 27).

The bead-based epitope assay (BBEA) represents a scalable, high through-put multiplex assay for IgE and IgG4 to several sequential epitopes, requiring microliter samples only (28). This technology utilises peptides mechanically coupled to beads, which are incubated with serum or plasma samples, followed by quantification of fluorescence by fluorophore-labelled antibodies. Sensitivity and specificity of epitope detection, at 92% and 94% respectively, has been defined for peanut allergy. BBEA can also be used for phenotypic stratification and severity prediction (29, 30). A particular strength of BBEA is its reproducibility, which is key for effective widespread use. Errors pertaining to technical details (well position, reading order), as well as batch effects, are possible limitations.

BBEA requires precise prior knowledge of epitope amino acid sequences corresponding to each allergenic food of interest (29). Further implementation of BBEA is therefore predicated on continued advances in basic research concerning protein sequencing, as well as the sharing and processing of large epitope datasets developed from robustly characterised patient cohorts. Consequently, in the near future, BBEA likely has most applicability to diagnosis of the well-researched peanut allergy. An important limitation is that BBEA is limited to sequential epitopes, whereas in reality conformational epitopes also play a key role in food allergy. BBEA analysis also does not capture the totality of the IgE response, as it does not reflect components such as affinity and cellular response which are better characterized by the cellular assays mentioned earlier, as well as only demonstrating allergen sensitisation rather than overt clinical allergy (31). Utility of BBEA results may therefore be improved with parallel investigation of cellular assays, and correlation with clinical history.

Glycosylation of natural allergens can lead to cross-reactivity mediated by glycan-specific antibodies (14, 32). For instance, sensitization to cross-reactive carbohydrate determinants (CCD) can cause detectable IgE to peanut whilst resulting in negative SPT or BAT to peanut, reflecting the absence of clinical relevance in this CCD sensitisation. In Ghanaian children, CCD was identified as the underlying reason for IgE cross-reactivity between IgE to peanut and IgE to parasites, such as Schistosoma Mansoni (33). While IgE to glycans is often regarded as a barrier to specificity of in vitro diagnostics, it could represent another component of the allergic response to fully investigate in addition to conventional IgE serology, defining distinct patient sensitization profiles (34). For instance, IgE to α-Gal has previously been associated with meat-allergy delayed anaphylaxis (35).

A different prism is that of IgE glycosylation, which has been implicated in the pathogenicity of IgE in its ability to induce mast cell and basophil degranulation in vitro, and anaphylaxis in a mouse model, when compared with non-allergic individuals (36). However, the functional consequences of differences in IgE glycosylation between allergic and tolerant individuals has yet to be assessed between allergen-specific IgE molecules.

Some studies have shown potential significance in testing for specific IgG4 against foods (37, 38). IgG4 is an IgG subclass that is produced under the influence of the same cytokines as IgE and has the unique ability for monovalent bispecific binding via Fab-arm exchange, theorized to be associated with IgE class-switching. However, it has been suggested that IgG4 merely represents a bystander of immunological tolerance rather than a factor driving such immune response (39). In the LEAP study, peanut-specific IgG4 increased over time in the study arm who consumed peanut from the first year of life, who also had less peanut allergy (40). In a recent cross-sectional egg allergy study, sIgG4 to egg, egg white and egg components, ovomucoid and ovalbumin, was higher in allergic compared with tolerant children and in consumers compared with avoiders (41). More research is needed to understand the role of IgG4 in food allergy and, certainly, to understand its utility in food allergy diagnosis.

sIgA is also presumed to play a role in the pathophysiology of food allergy, as sIgA deficiency has been linked to higher rates of allergies including food allergy, and salivary IgA has correlated with oral immunotherapy response for peanut allergy (42, 43). One cohort study indeed revealed higher plasma sIgA counts in children with peanut allergy, though no evidence was found for protection of food-specific IgA against development of food allergy or its potential utility for food allergy diagnosis (44).

A novel food allergy biomarker involves gauging the frequency of pathogenic T cell subsets and their activation in response to allergen stimulation. T cells are key immune cells in orchestrating food allergic and tolerant responses. In particular, Th2A cells, a subset of terminally differentiated memory CD4 + helper T cells, with characteristic expression of CD molecules and distinct functionality from conventional Th2 cells, have been associated with a specific stable allergic disease-related phenotype (45). One study examining pathophysiological trajectories of peanut allergy in a cohort of patients receiving oral immunotherapy showed clinical utility of subgroup division according to peanut-specific T-cell mediated immune profile (46). Discrete peanut-reactive T cell immunotypes based on relative proportions of Th2A cells were demonstrated, which were significantly related to distinct responses to allergen-specific immunotherapy. Additionally, Tfh13 cells, a subset of T follicular helper cells, may also be implicated in inducing anaphylaxis through induction of high-affinity specific IgE via sequential switching of B cells from IgG1 to IgE (47). In a study elucidating the position of T regulatory cells in spontaneous cow's milk allergy recovery, artificial neural networks were used to identify the markers FoxP3, Nfat-C2, IL-16, and GATA-3 as associated with the persistence of cow's milk allergy, with higher FoxP3 intensity in cow's milk allergic children, linking Th2 and Treg expression profiles with food allergy disease activity (48). An enhanced understanding of further Th subtypes could permit personalised care via assessment of individual T cell phenotypic signatures, designing and delivery of novel immunomodulatory therapies, as well as prediction of treatment response.

Playing an integral role in the production of allergen-specific antibodies, characterization of B cells is another possible biomarker for food allergy, determining specific attributes of the B cell receptor. IgE + B cells are directly related to the pathogenesis of food allergy, but are low in frequency and therefore difficult to identify. Investigative studies into IgE + B cells have identified the majority of IgE + B cells as immature plasmablasts with a common clonal family (49). IgE + memory B cells have been detected in the circulation, but they can also be present in the tissue, namely in the stomach and duodenum (50, 51). IgE + clonal signatures from gut samples could therefore represent one potential avenue to identify food allergy, though limited by access to relevant tissue sites in the gut. As with T cells, specific B cell subsets may also allow for further insights into food allergy, such as type 2-polarised memory B cells which are implicated in generating sIgE in early sublingual immunotherapy for birch pollen/house dust mite allergy (52). Similarly, a unique population of CD23 + IgG1 + memory B cells has also been described in the long-term persistence of peanut allergy in children, with frequency correlating to serum IgE levels (53, 54).

In a similar vein to characterization of T cell responses, studying microbiome dynamics could permit discrete categorisation of food allergy trajectories. Disparities in the prevalence of certain bacterial species between peanut-allergic and non-allergic patients were demonstrated in a multidimensional multi-omic study of the oral microbiome, accompanied by significant differences in short-chain fatty acid and IL-4 titres (55). In particular, abundances of Prevotella species were protective against a Th2-mediated response, whereas milieus tending towards Neisseria categorized Th2-mediated allergic response. This corroborates with literature suggesting a role for Neisseria-related mucosal inflammation. In a similar study using faecal samples, a metabolome network analysis from infancy to mid-childhood was conducted, revealing significant associations between reduced microbiome diversity, temporal bacterial variation, and histidine, butyrate, isovalerate pathway metabolites with development of peanut allergy (56).

Iterating on the evidence provided by associative trials, oral and gut microbiomes have been clinically implicated in influencing disease reaction thresholds. A prospective multiscale study revealed characteristic oral and gut microbiome signatures affiliated with peanut allergy, observing significant associations between prevalence of Veillonella and Bacteroides species with high-threshold peanut-allergic patients (57). Consistency of ordinal trends was also noted between these discriminative bacteria and non-allergic participants. These findings suggest integrated analysis of combined oral and gut environments could be worthwhile in exploring crosstalk and categorical summation of host-microbe-immune dynamics. Advances in microbiome analysis may be laterally transferable to other allergenic foods, as pro-allergenic micro-environments are likely to be shared between food allergies without requiring precise knowledge of peptide sequences, as in BBEA. Despite modest predictive performance in preliminary research, the complex and interdependent mechanisms of the gut and oral microbiome create fundamental limitations for the specificity of any proposed diagnostic test involving the microbiome, as confounding effects (diet, ethnicity, comorbidities) may create statistical noise, making microbiome analysis unlikely to be significantly useful outside the context of a broader picture of host metabolites and IgE levels (58, 59).

As a natural extension from effector immune cell responses, characterization of the cytokine milieu could further define patient-specific immunopathogenetic signatures. Plasma cytokine profiles have been studied as predictors of food allergy diagnosis and resolution among sensitized infants, with specific reference to higher levels of allergy-defining IL-4, IL-13, IL-12p70, and lowered IL-10 (60, 61). Heterogeneity of cytokine profile has also been described between peanut and egg-allergic patients, which has been related to disparities in resolution rate (62). One prospective study explored the clinical utility of these markers, observing significantly raised median TNF- α and IL-6 levels before a therapeutic elimination diet in food-allergic participants, followed by significant decreases post-diet. The existing literature suggests a role for plasma cytokines in quantifying success of immunotherapeutic interventions, with decreases in IL-5 and IL-13 observed after food allergen immunotherapy, though it is worth noting the transience of this immunologic suppression (63, 64). These results are positive toward integration of plasma cytokine levels in initially confirming food allergy and tracking disease trajectory, but must be interpreted with particular consideration to time-course due to their fluctuant nature. More utility may therefore be derived from multiple readings longitudinally and single-cell cytokine studies. The efficacy of single-cell cytokine analysis is highlighted in the work of Chiang et al., who used single-cell profiling to interrogate heterogenous Th2 effector subsets in peanut-allergic patients, revealing a delayed, IL-2-dependent CD154 upregulation on Treg cell marker-expressing cells (65). Similarly, Neeland et al. showed significantly greater Il-6 and TNFα production, and lower IFN-γ production, with single-cell analysis of CD3/CD28 stimulated naïve CD4+ T cells in food allergic adolescents (66).

AI methods hold a revolutionary potential for precisely identifying food allergy and synthesizing results of various tests for enhanced diagnostic accuracy. In comparison to conventional computerised protocols, artificial intelligence is able to fluidly adapt to provided datasets; learning from prior experience and applying this to novel situations. Specifically, machine learning (ML) frameworks, defined as a subset of AI involving teaching machines to emulate human learning while avoiding explicit programming, have utilized biomarkers including omics, total IgE, sIgE, sIgG4, and SPT data to successfully predict food allergy in infant populations (67). ML has also been efficacious as part of a medical algorithm identifying food allergy from data on methylation of highly discriminating cytosine-phosphate-guanine dinucleotides, reaching 100% accuracy through an ensemble voting scheme (68). Furthermore, prediction of tolerance to oral food challenge has been successfully demonstrated with Random Forest model sensitivity and specificity of 98% and 95% respectively for heated egg (69). ML processes have also been integrated within the novel bead-based epitope profiling mechanism described earlier, with the ability to accurately examine baseline variables influencing response to oral immunotherapy (70).

One of the major benefits of artificial intelligence approaches is the possibility for exploratory, unbiased data-driven analysis such as K-Means Clustering, which is an unsupervised algorithm enabling the grouping of unlabelled datasets into discontinuous clusters. Effective application of this to the MAT technique has been demonstrated, creating distinct clusters of phenotypes within a dataset of patients with similar responses to MAT (24, 71). AI models have also been successfully applied to MA multiplex arrays (72). As described in previous sections, the process of food allergy diagnosis lends itself to categorization of specific phenotypes in continuous unlabelled datasets, indicating high potential to accelerate our understanding of allergic patient presentations across a wide array of allergens and laboratory techniques. As well as this lateral applicability, there is also immense forward capability for analysis of methods that have yet to be discovered or refined, permitting swifter validation and optimization.