OIT effectiveness is normally assessed using two possible outcomes: desensitization or SU (32, 33). As mentioned above, desensitization is the patient’s ability to increase the minimal amount of allergen required to elicit an allergic reaction; to be maintained, it requires daily allergen exposure (34). An allergen threshold increase provides a certain degree of protection on the accidental ingestion of the causative allergen (32). Moreover, some desensitized patients can regularly ingest a full serving dose without reactions. However, the ultimate OIT goal is the achievement of SU, which is, in previously desensitized patients, the ability to safely consume any amount of the offending food, even after a prolonged period of allergen avoidance (35). Desensitization does not preclude SU development. Indeed, desensitization is necessary to reach, afterwards, a state of SU.

It has long been established that 50% of cow’s milk allergic children and up to 80% of hen’s egg-allergic children develop tolerance by the age of 4-6 years (36–38). Moreover, recent studies showed that an increasing number of children tends to outgrow their cow’s milk and egg allergies after the preschool age (37, 39, 40). However, although FA often spontaneously resolves by preschool age, at least in patients allergic to cow’s milk and egg (41–43), the rate of patients outgrowing FA in adolescence is lower, with individuals suffering from persistent FA (37, 39, 40). Therefore, in these patients, immunotherapy represents a possible way to modify the natural course of their persistent FA.

A recent meta-analysis confirmed that OIT induces desensitization in most patients allergic to peanut, milk, and egg: 76.9% of OIT patients reached desensitization vs 8.1% of patients tolerating the foods following allergen avoidance or placebo (26). Although the capacity of OIT to induce SU has not been clearly demonstrated, this meta-analysis showed that SU developed in 31.8% of patients after OIT vs 11.1% after allergen avoidance or placebo (26). Of note, the likelihood of achieving SU appeared to be related to the duration of OIT. Indeed, 28% of egg-allergic patients reached SU after 22 months of egg OIT (35), compared to 50% after 4 years of treatment (44). Moreover, the rate of SU was higher in peanut-allergic patients when OIT was initiated at a younger age (9-36 months) (45). Although the ability of OIT to induce SU seems limited (26), in accordance to experts’ opinion, families may consider desensitization as an acceptable outcome, as it protects children from potentially severe allergic reactions due to accidental allergen exposure (46).

OIT efficacy/effectiveness is presumably dependent on its effects on the allergen-specific immune response (47, 48). The immunological basis underlying desensitization and SU during OIT are still poorly understood. This is in part due to the scarcity in studies on immunological mechanisms in OIT trials and to other possible limitations inherent to conducting research in FA pediatric patients (49). Notwithstanding, it is well established that OIT tends to reduce allergen-specific Ig (sIg)E (after an initial increase), and this is followed by a boost in sIgG₄. The latter compete with sIgE for allergen binding, thus decreasing effector cell activation (i.e., basophils, mast cells), that release the mediators responsible for acute allergic reactions, including anaphylaxis (50). IgG₄ has been the most studied sIg in OIT, but other subclasses may contribute to the overall blocking and inhibitory sIgG response in FA (50). Considering cellular immunity, the spotlight has mainly been on Treg cells, which typically increase during OIT and exert a beneficial, but transient, immunosuppressive function (48). Therefore, while there is evidence that OIT induces protective immunological mechanisms (47, 48, 50), our understanding of these circuits is still fairly limited, and therapeutic approaches to make them endure after OIT interruption remain to be clarified (49).

Recently, it has been shown that tolerance should not be considered as an “all-or-nothing phenomenon”. For example, some milk- and egg-allergic children tolerate baked forms of these foods (51, 52). This is due to the impact of heat-processing on the structure of immunodominant allergens of egg and milk (53, 54), which reduces the number of sIgE-binding sites (55, 56). Along this line, Esamaelizadeh et al. conducted a RCT in 84 children presenting with cow’s milk allergy but tolerant to baked milk. Patients were divided into a case (baked milk consumers) and a control group (baked milk avoiders) for one year: 88.1% of patients in the case group and 66.7% of those in the control group developed tolerance to unheated milk at the end of the study (57). Therefore, the use of cow’s milk as part of baked food in OIT may increase safety and may favor the resolution of allergies towards the native allergens (raw or unprocessed) (58–60). Nevertheless, further studies are needed to confirm this line of evidence and to ascertain involved mechanisms (57, 61, 62).

In addition to the use of baked or hypoallergenic foods, other approaches have been explored to improve OIT efficacy. One of them included the use of anti-IgE monoclonal antibodies as adjuvant treatment for OIT patients (63). Omalizumab was first used in combination with OIT in milk-allergic children (64). Subsequent studies showed that omalizumab facilitated a faster achievement of higher OIT maintenance doses as compared to regular OIT (32, 65–69). Importantly, while omalizumab used as an adjuvant to OIT improved its safety and tolerability, it did not lead to higher efficacy (67), although larger studies are required on this topic (70). In this context, biologics intended to block other pathways of the Th2 immune response, such as dupilumab have begun to be explored in FA patients (71). The idea of interfering with IL-4Rα, thus blocking IL-4/IL-13 signaling, would prevent IgE re-generation from any memory B cell reservoir that requires IgE class-switching and its commitment to a plasma cell lineage (49, 72, 73); in addition, de novo Th2 polarization would be hampered. Consequently, the concomitant interference with IL-4 and IL-13 signaling in FA, may not only impair the machinery re-generating IgE but also potentiate regulatory pathways leading to desensitization, SU or oral tolerance (72, 74, 75).

Even though OIT has been evaluated for different allergens through several trials, in most cases this approach still lacks standardized protocols and current evidence has been generated only in a selected proportion of pediatric FA patients and only for certain foods (20). As abovementioned, OIT efficacy in clinical trials should clearly distinguish between desensitization and SU. Furthermore, patient-reported outcomes measures should be included in the clinical trials as well, because they represent powerful and irreplaceable tools to quantify the patient’s perception of the disease status and of its improvement (32).

OIT trials have shown effective desensitization in many patients and SU in some, but it is still unclear if the most relevant and important outcomes were measured. At present, there is no consensus on the core outcome domains, and validated instruments to assess these domains are lacking as well (29). The heterogeneity in OIT products, protocols, outcomes, age and clinical features of the enrolled patients does not allow to adequately assess the treatment effectiveness (27). Some preliminary estimations may be drawn for certain FA (e.g., cow’s milk, egg, and peanut). However, these only apply to children, because OIT in adult patients appears to not lead to successful desensitization (26). Importantly, children have a considerable likelihood of spontaneously acquiring tolerance, particularly to cow’s milk and egg, which questions the utility of OIT in them (76). Therefore, in some cases, it may be more appropriate to wait for the natural development of tolerance before proposing such a challenging and time-consuming intervention. On the other hand, those children that do not spontaneously outgrow their FA and become adults, may have lost a sensible window to modify their hypersensitive immunological status.

Another line of criticism towards OIT considers that the evidence in support of OIT efficacy in FA is weak and relying on limited data for cow’s milk, egg, and peanut. At present, OIT is not recommended for many foods, neither in adults nor in children (20). Most OIT clinical trials assessing effectiveness do not consider adherence problems although such issues remain critical in real-life settings. Indeed, OIT is a very demanding therapeutic option, and its efficacy strictly depends on patients’ adherence to treatment (20). In addition, consistent clinical and laboratory data on SU are scarce (77) and OIT is currently not widely used in the adult population worldwide. SU may be clinically confirmed by performing a food challenge after OIT has been discontinued for a certain period of time. In most studies, SU has not been assessed and there is not enough information on the possible efficacy of OIT in acquiring it after treatment discontinuation (26). Moreover, there is no consensus on when SU should be assessed, especially on how much time after OIT cessation. The majority of studies addressing SU evaluate it up to 8 weeks after treatment discontinuation (34), which may be insufficient to reach a firm conclusion. Furthermore, even once SU is confirmed, it may be lost over time. For example, it has been demonstrated that after a 2-weeks proven SU, patients treated with OIT for cow’s milk, egg and wheat may still experience allergic clinical manifestations after longer periods of avoidance (78). The lack of evidence for SU implies that OIT patients should pursue a life-long “maintenance phase” to prevent potentially dangerous adverse reactions (ARs) after the consumption of the involved food. To date, protocols do not include accurate information on possible quantity and frequency of food intake after the end of the maintenance phase. Therefore, in most cases there is a lack of standardization and of recommendations on this aspect of patients’ after-treatment management.

Several immunological changes have been reported during OIT, some of which appear to be consistent across different allergens and OIT protocols (50). As indicated earlier, OIT studies usually show an increase in sIgG₄ levels, and its persistence upon therapy cessation has been associated with SU (79). Other immunological changes reported during OIT include expansion of Treg cells, and reduction in total and sIgE levels (47, 50). However, reliable biological markers to assess the evolution of OIT patients or desensitization/SU persistence are still unavailable. The suppression of the immunologic response during OIT seems to be transient (80, 81) and unable to control persistent populations of pathogenic Th2-cells, which have been detected in patients with peanut allergy after 12 to 24 months of OIT (80). Hence it is unclear if immunological markers can serve as a practical and reliable outcome of OIT (48). A detailed characterization of the immune cells affected by OIT is required to uncover immunological changes indicative of durable SU or effective desensitization, which can be tracked after OIT discontinuation (48, 49, 77). In this context, reaching a better understanding on the immunological basis of persistent FA may lead to the identification of novel biomarkers that may help in better defining OIT outcomes (49, 72–74).

The use of omalizumab as adjuvant therapy for OIT showed some benefits in facilitating a faster achievement of higher maintenance doses, as discussed above. However, omalizumab failed to improve SU acquisition. This may be potentially connected to the fact that the benefit obtained is lost after the discontinuation of the treatment with this monoclonal antibody (69).

At present, OIT may be considered for recommendation only in cow’s milk, egg or peanut allergic children and desensitization is assured only if the adherence to the treatment is very high. In most cases, appropriate products and defined protocols for OIT are still lacking. OIT to baked foods may be particularly challenging in centres lacking expertise. To date, effectiveness in inducing SU has not been demonstrated neither clinically nor in terms of persistent immunological modifications. Therefore, post-desensitization strategy remains unclear and further studies are needed. New approaches such as the utilization of omalizumab to improve OIT effectiveness failed to achieve long-term SU. Of note, lack of agreed core outcomes in FA slows down the process of high-quality evidence collection and OIT effectiveness assessment.

FA patients and their relatives should be always aware of the potential risks associated with OIT. Various studies assessed OIT safety comparing intervention data with patients on elimination diet. Almost all OIT patients experienced mild or moderate ARs (46), as perioral rash, local urticaria, rhinitis, or minor gastrointestinal clinical manifestations (26), and most ARs resolved without treatment or simply after administration of oral antihistamines (82).

The risk of systemic ARs during OIT is relatively uncommon, and no OIT-related deaths were reported in the literature. Still, all FA patients should be trained to use, and carry, emergency drugs (epinephrine auto injector), and should be accurately monitored, especially in the OIT up-dosing phase (83). A recent baked-milk OIT study reported ARs in 21 out of 63 patients, with only one of them developing anaphylaxis (84). Other studies confirmed a low incidence of ARs requiring epinephrine injection during peanut (30, 85) and egg (86) OIT. In contrast, Kauppila et al. reported that 14% of patients had anaphylaxis during raw milk OIT, but a cofactor was often considered as an element involved for inducing the AR, or the ARs appeared after a period of allergen elimination diet (87).

Cofactors may play a role in triggering an AR, which mainly occurs during the OIT build-up phase. Importantly, patients already tolerating a specific allergen dose, may sometimes experience an AR during the OIT maintenance phase. In those cases, a cofactor altering immune homeostasis (e.g., viral infections, fever, exercise, non-steroidal anti-inflammatory drugs intake, or hormonal changes) may be involved (46). Indeed, it is important to recommend avoidance of physical activity at least one hour before and three hours after the intake of a food allergen. Likewise, children suffering from fever or infectious diseases should suspend, or at least halve, the OIT maintenance dose for a few days (82). Other conditions contributing to ARs during OIT are poorly controlled asthma, seasonal pollen allergy, and circumstances such as consumption of the OIT on an empty stomach (82).

Once long-term secondary effects are concerned, it is still debated whether OIT might play a role in the onset of eosinophilic esophagitis (EoE) (82, 88, 89). A meta-analysis reported a not negligible prevalence of EoE in OIT-treated patients (90). However, EoE is a condition that can resolve following specific therapy or OIT discontinuation (89). Along this line, the results of a recent trial on 15 adults undergoing peanut OIT showed that possible OIT-induced EoE and gastrointestinal eosinophilia are usually transient and not always associated with gastrointestinal signs and symptoms (91). Recurrence of ARs may also be a cause of OIT withdrawal (92, 93). A recent meta-analysis reported an overall rate of OIT discontinuation of about 14%; only 4.7% was due to clinical manifestations possibly related to EoE (92). In addition, Blumchen et al. showed that, using a peanut OIT protocol with a low maintenance dose, the proportion of dropouts due to ARs was 6.7%, without need of epinephrine use and with no EoE development (31). Nevertheless, it is still unclear if OIT can induce EoE as a direct side effect of the treatment.

The decision to initiate OIT should be tailored on the patient’s allergic profile, and on personal choice of patients and their families. Awareness of the possible risks should be raised, and details of the heterogeneity of reported ARs should be clearly outlined and put into context of the specific treatment proposed (88, 94).

Patient safety is a critical issue for OIT. ARs, including life-threatening events, appear to be more frequent during OIT, in comparison with food allergen avoidance (28, 88, 89, 94). Studies report that 10–35% of FA children withdrew OIT trials because of anaphylaxis, acute or repeated ARs and especially chronic abdominal pain (95, 96). As mentioned above, many identifiable but often unavoidable factors may cause ARs (including fever or infections, exercise, temperature changes, dosing on an empty stomach, menstruation, seasonal allergies, asthma, and non-compliance), and some identifiable cofactors are often unknown (28, 82, 97, 98). For example, near-fatal reactions have been observed in asthmatic teenagers showing poor compliance (99). Therefore, the presence of cofactors requires frequent allergen dose adjustments to obtain a safe dosing profile (100). In this regard, the European Academy of Allergy and Clinical Immunology (EAACI) OIT guidelines suggest continuous surveillance of OIT patients for ARs and clinical manifestations of new-onset EoE, especially during the up-dosing phase (20).

EoE is a possible secondary effect of OIT, but the relationship between EoE and OIT remains controversial, being unclear whether OIT causes EoE or reveals a pre-existing condition (89, 101). The overall prevalence of EoE following OIT was reported as 2.7% (82, 90). However, a prospective food OIT study reported that EoE or eosinophilic gastroenteritis developed in 7 (7.2%) patients out of 97 children included in a milk OIT group and in 2 (6.4%) out of 31 patients in a egg OIT group (102). During OIT, sIgG and sIgA may enhance eosinophil activation and contribute to EoE onset (103). To evaluate the real EoE prevalence during and following OIT, it is necessary to consider that an esophagogastroduodenoscopy is not performed in all patients with dose-limiting gastrointestinal clinical manifestations. Therefore, the real EoE rate may be higher than reported (104). Furthermore, longitudinal data of EoE in food OIT are insufficient to study long-term safety (93). A recent comprehensive review analyzed data from 110 studies (92) and found that EoE-confirming biopsies were performed in 18 studies only, and EoE was diagnosed in 5.3% of OIT patients.

Many studies show that epinephrine use is variable and related to the OIT protocol (105). Wasserman et al. retrospectively reviewed charts from 352 patients evaluated in 5 allergy centres, and they found that epinephrine administration was necessary in 36 patients (10.2%) (106). A Cochrane systematic review on milk OIT reported ARs in 97 out of 106 patients (91%), while epinephrine was required in 9% of patients receiving milk OIT (107). A more recent prospective study on milk OIT registered 1,548 ARs, most of which occurring during the escalation phase (89.6%). Anaphylaxis and chronic late-onset gastrointestinal clinical manifestations accounted for 15.8% of ARs and represented the primary reason for protocol withdrawal. Interestingly, a higher rate of sIgE for α-lactalbumin and casein at baseline was associated with an increased risk of anaphylaxis during milk OIT, while patients with higher sIgE for β-lactoglobulin had a lower risk (108). Moreover, egg OIT was associated with serious ARs in all 10 RCTs included in a Cochrane systematic review. Epinephrine was required in 21 out of 249 (8.4%) of children in the egg OIT group but never in the control group (105). A recent systematic review of >1,000 peanut-allergic patients evidenced that peanut OIT increases the risk of ARs, anaphylaxis development and epinephrine use, either during the build-up or the maintenance phase. The authors estimated the risk of anaphylaxis in patients undergoing OIT about 22% in comparison with a baseline risk of 7% (28).

Some centres are using omalizumab during OIT because it may reduce the risk of ARs, especially in children with severe FA. However, the duration of the therapy with omalizumab is still debated, as well as the long-term outcomes after omalizumab discontinuation (66). Some experts consider the use of omalizumab as a helpful measure for rapid up-dosing but, when decreasing omalizumab doses, ARs become more frequent. Therefore, omalizumab appears to confer short-term protection from OIT-related ARs (109). Furthermore, the use of omalizumab may be associated with side reactions including skin inflammation and anaphylaxis in 0.1-0.2% of patients, likely due to its engagement with Fcγ receptors (110).

Further investigations on OIT safety should be carried on prior to routine OIT use in clinical practice (105, 111). A careful and complete explanation of the ARs risks vs OIT clinical benefits should be given before starting the therapy to both children and their families (83).

Both FA and the consequent elimination diet have a negative impact on QoL due to food-related anxiety, fear of accidental exposures and the ever-present burden of social and dietary limitations (11, 95, 112–114). Allergen elimination diet remains the gold standard for FA management. This approach, however, requires constant responsibility from both patient and caregiver. FA children are vulnerable to unintentional allergen ingestion and possible anaphylaxis, which has a negative influence on the QoL (35).

OIT aims to increase the allergen reactivity threshold to reduce the risk of serious allergic reactions after inadvertent allergen exposure. This approach should provide a safer social life for FA children, with less fear of being accidentally exposed to the allergen, and eventually result in QoL improvement. A complete resolution of the allergy following OIT treatment remains the ultimate OIT goal, which would clearly improve patients’ QoL. An increasing number of studies evaluated the impact of OIT on children’s QoL, mainly for egg (115, 116), peanut (11, 31, 96, 117–119), and cow’s milk allergy (120), but also on patients suffering from multiple FA (121–124). The results of these studies are encouraging, even though most of them report major limitations mainly related to the small sample included in the study and the limited number of RCTs.

Most of the studies confirmed that desensitization correlates with an improvement in children’s QoL, as perceived by their caregivers (96, 115, 121, 122, 124, 125). Three factors were found to be associated with a more substantial improvement in QoL: having an allergy to a single food, presenting with a history of anaphylactic reactions prior to OIT initiation, and having a very low QoL before the beginning of OIT (124). Blumchen et al. found that achieving desensitization during OIT improved the QoL as perceived by caregivers and by patients in a double-blind RCT (31). After peanut OIT, there was a significant improvement in QoL in the domain of “risk of accidental exposure” and “emotional impact” in children, when compared with the placebo group. Other studies that evaluated the QoL during and/or at the end of OIT showed a more substantial improvement in children (116, 117) or adolescents (117) when compared to the caregiver-reported QoL. In contrast, Reier-Nielsen et al. found statistically significant improvements in QoL reported by caregivers, while no significant change was recorded in children (119).

Several studies evaluated QoL improvement by OIT, as perceived by caregivers, in patients achieving SU (11, 120, 126). Specifically, a double-blind placebo-controlled RCT (127) evaluated the impact of Probiotic and Peanut Oral Immunotherapy (PPOIT) on health-related QoL. The authors concluded that PPOIT had a sustained beneficial effect on the psychosocial impact of FA at 3 months and 12 months after completion of the treatment. The improved QoL was specifically associated with the acquisition of SU. Indeed, for this study, a post-hoc analysis revealed that no improvement in QoL was seen for either PPOIT-treated or placebo-treated patients who failed to achieve SU (128). A study assessed the impact of QoL evaluated by children and adolescents after achieving peanut SU (126); in this study, there was a general trend towards an improvement in QoL at the end of peanut OIT, but it did not reach statistical significance, probably because of the small sample size.

Studies assessing QoL as one of the outcomes in patients undergoing OIT are heterogeneous, use different methods, and are often limited to a small sample size. According to some experiences, compared to food allergen avoidance, OIT may be associated with an improvement in QoL for both patients and caregivers, especially at the end of the treatment and in the absence of ARs (122, 129). However, these data should be viewed with caution as additional evidence is needed.

FA patients present with a decreased QoL, with repercussions on their general health and lifestyle (20, 130). Although OIT is a promising therapeutic approach, it is demanding, especially in terms of protocol duration, adherence-related issues, and safety concerns (20). It is therefore evident that improvements in patients’ QoL throughout and after OIT must be carefully assessed (26, 131). The heterogeneity of factors possibly influencing QoL makes it difficult to achieve a uniform approach. Some important ones such as the perception of treatment burden, are rarely considered in OIT trials (31). Health-related QoL has only recently started to appear as a potential outcome in OIT clinical trials, hence data on the matter are limited (131) and a meta-analysis by Chu et al. found no evidence that OIT improves the QoL (28).

Improved QoL after OIT has been reported only in a few trials for egg (115, 116) and cow’s milk allergy (120), while a larger number of studies considering QoL is available regarding peanut (31, 96, 118, 119, 126, 127) and multiple food desensitization (121, 122, 124, 125). Most of these studies are characterized by substantial limitations in the reported results, being based only on parent-reported QoL (96, 121, 123, 128) or lacking a control group (117). As aforementioned, discrepancies between children self-assessment and parental reports have emerged, showing significantly better QoL scores reported by parents when compared to their children (119). These data suggest that parents may overestimate the impact of OIT on child’s QoL, calling into question the appropriateness of parental proxy reports use as a valid outcome of OIT effectiveness. In addition, OIT may also result in QoL worsening, for example in the build-up phase, as demonstrated by Ebstein-Rigbi et al., probably because of the occurrence of ARs in this treatment phase (122). The heterogeneity of the studies and of the methodology of QoL assessment, along with the very limited number of data do not allow for definitive conclusions but the most up to date meta-analysis does not add optimism (28).