Recent advances in the management of nut allergy

Co-existent allergy peanut and TN have been described for many years. Initially, in the 90s, Sicherer et al reported that 34% of patients allergic to peanut or 1 TN may present with multiple nut allergy. However, further studies reported a large variation in the proportion of patients reacting to multiple nuts, ranging from 12% to 96.7%.^{,,,  ,  ,  ,  ,  ,  ,}  These data have been summarized in Table 1.

The influence of pollen allergy, the population studied as well as its ethnicity are all confounding factors that might influence the results. In addition, these differences can be explained by differences in the methodology of the studies.

Thus, studies that reported on specific IgE found that the co-sensitization rate among TN and peanut ranged between 60.6% and 96.7%.^{,  ,  ,,  ,} 

When a positive clinical history is required to diagnose nut allergy, but without a confirmatory OFC, the proportion of patients with multiple nut allergy is lower than expected, ranging between 23% and 68%^{,,  ,  ,  ,}  (Table 1).

Studies including OFCs to prove co-existent peanut/TN allergy, considered as the gold standard, report a rate co-existent allergy of 12%–38.8% and confirmed initial data by Sicherer^,,, (Table 1). However, a recent prospective multicenter study in Europe (Pronuts study) based on 122 patients that underwent sequential OFC to determine allergy versus tolerance, showed a higher rate of co-existent peanut/TN or sesame seed allergy at 60.7%. These results could be explained by the fact that the Pronuts study was prospective as opposed to retrospective in the previous studies,^, assessed all 9 TNs (compared to other studies testing less TNs),^, and included sesame seed which belongs to the Pedaliaceae (seeds) family. The NutCracker study which was also a prospective study including OFCs, reported a lower prevalence of multiple nut allergy below 30%, however, this study (based on a cohort of 83 children with TN allergy in Israel) included only OFCs to a subset of TN (walnut, pecan, cashew, pistachio, hazelnut, and almond), which could potentially underestimate the rate of co-existent allergy. Couch et al in a recent retrospective study, found similar results; 67 patients with a history of TN allergy underwent an OFC to another TN to which they were sensitized, but not exposed to before. Interestingly, only 14% of the included patients had a positive OFC to another TN. However, this study was retrospective and patients in this study had relatively low levels of specific IgE (90% of these patients had a sIgE <2 kU/L and at least half had a level <0.35 kU/L); thus, these patients were probably selected for OFC based on lower IgE tests to confirm non-allergy by OFC and this would have contributed to an underestimation of co-existent peanut/TN allergy.

The age of participants may also play a role in the rate of co-existent peanut/TN allergy. Indeed, although most TN allergies become apparent when a patient is young, many studies have shown that the rate of co-existent allergy between peanut and TN increases with age. The HealthNuts study showed that children who had peanut allergy at 1 year old had a 27% chance of having an OFC-confirmed TN allergy at 6 years of age. Brough et al in a retrospective study reported similar results. This increase can be explained particularly by the fact that nuts are introduced later than other foods. Indeed, Clark and Ewan showed that the number of nut consumption increased with age (23% eating more than one nut at 2 years, versus 73% by 10 years); they postulated that this could lead to higher rates of multisensitization (19% at 2 years, 86% at 5–14 years) and multiallergy (2% at 2 years to 47% at 14 years). Conversely, Elizur et al proposed the opposite hypothesis, that elimination of TN in multiple-food-allergic patients could promote the development of sensitization and allergy to TN years later. Peanut/TN co-sensitizations are common and distinguishing asymptomatic sensitization from clinical food allergy is currently based on OFCs, whicht may lead to life-threatening reactions. The clinical relevance of serological cross-reactivity between peanut/TN therefore needs to be better defined.

Peanut/TN allergy has 2 main dimensions. One is the cross-reactivity for the components, and another is severity. Recent advances in the field of component resolved diagnostics (CRD) provides the clinician with more information as to whether the patient has secondary nut allergy due to pollen food syndrome (also known as oral allergy syndrome) or primary nut allergy, more likely to lead to systemic symptoms.

Indeed, there are different sensitization profiles in peanut/TN allergy. Patients can be, therefore, sensitized to different families of proteins within the nut. The physico-chemical properties of the proteins to which peanut/TN allergic patients may be sensitized are responsible for allergic reactions of varying severity. The most well-known protein family is the seed storage protein family (eg, Ara h 2, Cor a 9, Cor a 14) responsible for severe anaphylactic reactions, explained in part by their thermostability and digestive resistance. Other families of proteins that are also responsible for severe reactions are the oleosins, defensin and LTP family.

Other sensitization patterns can lead to less severe symptoms in the majority of cases, such as the reprensented sensitization to PR-10 and profilins family. This is due to the fact that these protein families are degraded by heat and digestion.

Components, protein families and cross-reactivity between components are referenced in Table 2.

Allergies to certain well-defined combinations of nuts may be due to the presence of similar or closely related epitopes. Such closely related epitopes are more common in phylogenetically closely related nuts such as cashew and pistachio, walnut and pecan,^,,, peanut and soybean.^{,,  ,  ,} 

Thus, the Pronuts and NutCracker studies found that 97%–100% of pistachio and pecan allergic children were allergic to cashew nut and walnut, respectively.^,  Moreover, 64.2%–83.3% of patient allergic to cashew or walnut were respectively co-allergic to pistachio and pecan.^, In a retrospective study, Andorf and al reported similar results. Pistachio and cashew nuts belong to the same Anarcadiaceae family (homology 79% between rPis v 3 and rAna o 1, and homology 66% between Pis v 1 and r Ana o 3).^, High homology between pecan and walnut protein sequences, which belong to the same botanical family (the Juglandaceae family), have also been described. Indeed, 2S albumin allergens in walnut (Jug r 1) and pecan (Car i 1) have 88% sequence identity and legumin allergens in walnut (Jug r 4) and pecan (Car i 4) have 95% sequence identity.

Other studies have reported lower prevalence of co-allergy between cashew nut and pistachio. Indeed, Van der Valk et al and the HealthNut studies found that only 31%–36% of the cashew-allergic patients reacted to pistachio.^, There is an uni-directionality of the co-existent allergies, as a lower proportion of patients allergic to walnut and cashew are allergic to pecan and pistachio, respectively. This suggests that some allergenic proteins are shared while others are unique to cashew and walnut and therefore result in mono-allergy.

For peanut and soybean, studies have demonstrated similarities between both legumes allergens, such as Ara h 1, Ara h 3, and Ara h 8 with Gly m 5, Gly m 6, and Gly m 4, respectively, between 38,4%–70%^{,,  ,  ,}  (Table 2). Despite this homology, studies show a low rate of cross sensitization and cross reaction. Indeed, in a study from several years ago, 31% of peanut-allergic children had cosensitization with soy, and only 3% had clinical reactivity to soy. In other studies, the cross-reactivity rate has been estimated to be between 6.5% and 15%.^, Another study by Savage et al reported that 98% of patients with a soy allergy also had a peanut allergy. As with nut allergies, these data suggest that some proteins are common to peanut and soybean and some are specific to soybean and peanut.

Different families of proteins such as the seed storage protein families (vicillins, 2 S albumins and legumes), the family of lipid transfer proteins (LTP) family, and pathogenesis-related protein type 10 (PR-10) family also help explain the cross-reactivity among peanuts and other legumes (eg, peanut-lupine). In addition, they also help explain, in part, why unrelated nuts such as TN and peanuts may exhibit serological and clinical cross-reactivity.^,,

Peanut and lupine have a high degree of cross-reactivity; therefore, risk associated with cross reaction is also high as compared to other legumes. Studies show that 14.5%–89% of peanut allergic patients were sensitized to lupine however this cross sensitivity is clinically significant in only 4%–35% of cases.^{,  ,  ,  ,  ,}  (Table 2). Cross-reactivity has been reported to be mediated by Lup a 1 (vicilin-like protein)^, (Table 2). In 2017, the lupine profilin Lup a 5 was registered, which is highly cross-reactive to other profilins (eg, Ara h 5) and which is recognized by the sera of both lupine and peanut-allergic patients (www.allergome.org).

Due to structural homology, lipd transfer proteings (LTPs) from different allergen sources are generally IgE cross-reactive; however, sensitization profiles are extremely heterogeneous, and individual cross-reactivity patterns may range from a single LTP to many different LTPs (from food or pollens).^,

Some studies report a significant number of peanut/TN allergies associated with LTP sensitization, which may be responsible for severe systemic reaction.^, The peach LTP Pru p 3 has been shown to be the primary sensitizing allergen for cross-reactivity with other LTP, including peanut (Ara h 9), hazelnut (Cor a 8), walnut (Jug r 3), and almond (Pru du 3) (Table 2). It has been shown that sensitization to LTP leads to a large variety of clinical manifestations; although oral allergy syndrome (OAS) is probably the most frequent clinical expression, LTPs can be also responsible for severe systemic reactions.^, Thus, it is the most frequent cause of primary food allergy in the Mediterranean area.^,

LTP sensitization can occur via the gastrointestinal tract, but the predominant presence of the LTP syndrome only in the Mediterranean region suggests that environmental factors play a major role. Indeed, Vereda et al showed that in peanut allergic patients, LTP sensitization rate varied by country: in Spain, 60% of patients are sensitized to peanut LTP (Ara h 9) while these proportions were 7.7% and 14.3% in the United States and in Sweden, respectively. The reasons for these geographical distributions are still poorly defined. Studies hypothesize that these distributions are in part due to variations in environmental homologous pollen allergens exposures in LTP-endemic areas such as Art v 3 from mugwort, or Pla a 3 from plane tree.^, In agreement with Pastorello, Scala et al reported that, in LTP allergic patients, co-sensitization with PR-10 proteins, is associated with milder symptoms.. In addition, the higher the levels of birch pollen in a certain area, the lower the prevalence of LTP hypersensitivity.

TN and peanut allergy may display serological as well as clinical cross-reactivity with pollens. The majority of these patients suffer from OAS. Patients initially allergic to birch pollens through sensitization to a PR-10 protein, may develop a secondary allergy (pollen-food syndrome) to peanuts or TN (OAS);^, they develop mainly mild symptoms limited to the oropharynx, with pruritus, tingling, erythema, and mild edema of the mouth upon ingestion of peanut or TNs(67). Pollen food syndrome (PFS) is triggered by a cross-reaction between allergens in pollen and allergens in peanuts/TN.^, Homologous proteins have been identified between hazelnut, walnut, peanut, and soybean and have been shown to cross-react with Bet v 1.^{,  ,} 

The prevalence of PFS ranges from 4.7% to greater than 20% in children sensitized to pollens.^, The PR-10 family also plays a significant role in PFS. Bet v 1 from birch pollen is well known of these proteins and is one of the major pan-allergens in PFS. Uotila et al in a retrospective study found that among subjects with birch sensitization, 84% were cosensitized to hazelnut, 71% to almond, and 60% to peanut; amongst these nut-sensitized patients, 40% of patients sensitized to hazelnut, 34% of those sensitized to almond, and 36% of those sensitized to peanut reported typical symptoms of PFS. A retrospective review from Northern France, where there is a high level of birch pollen exposure, reported a 43.2% co-existent TN allergy rate amongst patients with peanut allergy (43,2%), with hazelnut being the most common TN allergy observed.

Symptoms associated with PR-10 sensitivity are mainly mild.^{,,  ,}  However, the thermostability of the proteins in this family are variable. Heat processing such as roasting significantly reduces the rosacea fruit protein allergenicity in patients with birch-pollen allergy, but some sensitized individuals can still experience positive reactivity toward roasted peanut, soy, and TNs.

Although the DBPCFC is the gold standard for diagnosis of food allergies, this is costly, resource and time-consuming, and carries the risk of potentially life-threatening reactions. Some patients or their parents refuse to perform an OFC due to the fear of triggering a severe reaction. In the Pronuts study, Brough et al reported that 8.2% of children did not perform an OFC due to fear of reaction or history of previous severe reaction on exposure to the incriminated nut. In this clinical setting, not performing an OFC can potentially lead to unnecessary and prolonged peanut/TN avoidance, which may have the unintended risk of increasing peanut/TN allergy risk. However, it is necessary to find new, less invasive diagnostic tools for the diagnosis of peanuts/TN allergies. Studies have shown that combination of SPT, sIgE, and basophil activation test (BAT) improved the ability to identify allergic and tolerant patients. In the case of peanut allergic patients, this approach could potentially lead to a reduction of OFC of 76.6%–97%.^,

As for other allergies, it is of major importance to interpret peanut/TN SPT and sIgE in the context of the clinical history. The diagnostic value of SPT and specific IgE to whole allergen extracts has been found to vary significantly among studies. Indeed, those different results might be explained by differences in the population studied, prevalence of pollen allergy, and the methodology used in the study.^,,, These data have been summarized in Table 3. As an example, while a SPT <3 mm has a good negative predictive value;^{,  ,  ,}  SPT<3 mm still requires further investigations in the context of a convincing clinical history of nut allergy. In contrast, a SPT ≥3 mm to a specific nut, without an appropriate clinical context has a poor predictive value and is associated with high rate of false positives.^,, Clark et al showed that amongst patients with a history of reaction to peanut or TN, a SPT ≥8 mm had a predicted clinical reactivity greater than 95% accuracy. Ho et al confirmed this threshold value for cashew, hazelnut, and walnut.