High Risk Infants who are Low Dose Tolerant after Peanut Oral Food Challenges

Introduction

The landmark Learning Early About Peanut Allergy (LEAP¹) trial, follow up study (LEAP-On²), and other recent data suggest that early peanut (PN) introduction can help to prevent PN allergy in a significant number of at risk infants. Expert panel guidelines have since been published on PN introduction and oral food challenges (OFCs) that allocate risk level and stratify challenges based on who should be challenged at home, in a supervised visit or with an office OFC.^3,4 Despite these developments and changes in practice, our understanding of the most effective ways to implement early PN introduction is incomplete.

The LEAP trial identified children at high risk for PN allergy in a narrowly defined population (infants with moderate to severe eczema and/or egg allergy) and of those, excluded infants who had skin prick tests over 4 mm. The NIAID consensus guidelines recognize some uncertainty in management of children with larger skin tests, recommending PN introduction or OFC in those who have skin prick tests ≤ 7mm. But these recommendations do not specifically address what the allergy specialist should do with infants who have larger SPTs (or high PN-IgE levels) ^3,4. Although these recommendations and previous data suggest that those with SPT ≥8 mm are ‘probably allergic’, effectiveness of early PN introduction in young infants with larger SPTs is not known and was not addressed in the LEAP trial. Conceivably many children in this category would benefit from PN-OFC.

To date, published data on implementation of the new PN guidelines in the clinical setting is limited. In a recent retrospective study, the feasibility of implementing these guidelines was addressed in a large referral population of infants under 12 months who were evaluated for PN allergy following the consensus recommendations. Observed feeding or graded food challenges to PN were only offered if SPT <8 mm, while those with larger test results were advised strict avoidance. Reaction rate was low and most were mild in severity⁵.

Both the NIAID guidelines and LEAP trial utilize a goal consumption dose of 3.9 g PN protein for OFC for those with positive testing (LEAP gave single 2 gram PN protein dose to those with negative testing). This corresponds to approximately 1 tablespoon of smooth PN butter or 40 pieces of Bamba, which in our experience, is not feasible for a young infant to reliably consume in one sitting. There are currently no published data on the use of lower-dose PN protein for infant OFCs to assess for low-dose tolerance, though this could potentially allow more successful PN introduction in young infants who are concurrently developing feeding skills and preferences. Taking advantage of low-dose tolerance has been discussed in a clinical communication with older PN allergic children which helped support our clinical reasoning.⁶ Given the major shift in the paradigm for PN introduction after the LEAP trial but prior to the release of the guidelines, we sought to widely offer infant food challenges in a safe setting using low-dose PN-OFC. The “high-risk” infants as described in the LEAP trial excludes many who we deemed, along with the respective families, to be worth challenging given that SPT and PN-IgE levels are not standardized in this age group. Most reference values are based on infants > 12 months^7,8 although the cutoff of >4mm is referenced for children < 2 years of age.⁹ Because OFC is the gold standard for diagnosis, we considered offering OFCs to the “high-risk” and highest-risk infants an ethically sound service; the potential benefits of tolerating OFC offered in a hospital-based setting with ample emergency resources are great. Based on clinical observations, our approach evolved quickly to using lower PN protein doses (1 gram) for OFCs since this is more age-appropriate in lieu of the almost 4 grams reported in LEAP and standard OFCs. This dose was then slowly increased at home by a set of recommendations to families thought to be more akin to real-world food introduction and escalation. With extensive counseling of families before and after the OFCs, this was considered to be a reasonable off-label clinical service not requiring a formal research protocol.

Methods

We defined highest-risk infants (HRIs) as those with PN-SPT >7 mm although we offered the challenges to all infants coming to the clinic if home introduction was not desired. Three families at the beginning were offered 2.5 or 5 gram challenges, but most of the HRIs were offered 1 gram as we tailored our approach over time. We have standing IRB approval for OFCs and maintaining a database of these outcomes.

The goal PN protein dose was 1 gram in six gradually escalating doses. Families were referred often for positive IgE testing performed by an outside provider, moderate to severe eczema, egg allergy, possible reaction at first PN introduction, or a sibling with food allergy (FA). Having a food allergic sibling has in some studies been shown to increase risk of FA;^10-12 thus, we included this in our highest-risk criteria as many of the families were apprehensive about home introduction and delayed introduction was likely to occur.

Informed consent was obtained prior to OFC, in keeping with our standard practices considering OFCs as procedures. The main OFC outcome was the assessment of tolerance to 1 gram of PN protein. PRACTALL guidelines were used to determine if patients did or did not tolerate OFC¹³.

A retrospective chart review was conducted of PN-OFCs performed on 70 infants ages 6-13 months between December 2015 to December 2017 at our pediatric allergy referral center. We compared PN-SPT size, PN-IgE and Ara h 2 IgE between those who were low-dose tolerant and those who had an IgE-mediated reaction during the challenge. There were 3 infants who had delayed reactions including persistent emesis 1.5-3 hours after exposure. These infants were not included in this analysis but represent a previously rare population with PN-FPIES who we have described elsewhere¹⁴.

If the infant was low-dose PN tolerant, we advised that families introduce 1 gram PN protein (1 teaspoon PN butter, PN powder/flour or 1/4 bag of Bamba) at home 3 times per week and increase by 25% per serving (about 1/4 teaspoon) every 2-4 weeks as tolerated up to a goal of 2 grams of PN 3 times per week up to age 5 years. We discussed with families of children who had the highest testing (SPT > 7mm) that their child is not considered to have tolerance to PN and that they must continue to carry an epinephrine autoinjector at all times until further follow up. Epinephrine autoinjector prescriptions were kept up to date for these families. We also advised that infants be observed closely after home doses for 1-2 hours and that families consider skipping PN on days with fever or gastrointestinal symptoms. We did not consider this to be oral immunotherapy per se since starting doses were much higher and home dosing was not highly regimented.

Statistical analysis was performed in Stata/SE 13.1 using Student’s t test, rank sum, Fisher’s exact analyses and receiver operating characteristic (ROC) area under the curve.

Results

Over a two-year period, most infants ages 6-13 months at our referral center who were considered high- or highest-risk were low-dose tolerant of a 1 gram PN-OFC (54/70, 77%). 11 patients had successful home introduction and were not challenged (SPT 0-2mm or IgE 0-0.39 kU/L). 6 infants were tested and parents preferred to avoid PN (SPT median 9.5 mm (range 0-12), and median PN-IgE 24.2 kU/L (range 13.8 to >100 kU/L). There were about 8 infants deemed tolerant who had mild, transient, self-resolving hives during the food challenges.

There was no statistical difference in age, gender or race between the groups regardless of SPT size (see Table 1). We had a relatively large percentage of African-Americans compared to LEAP and showed that these infants tolerated PN as often if not more than Caucasians (not significant). This population had predominantly mild atopic dermatitis (AD; 74%) with only about 26% of patients having moderate-to-severe AD determined by clinical judgment by various providers in our group (N.B. we did not have baseline eczema as a reference prior to referral, so eczema might have been more severe). There was no difference in AD, moderate-to-severe AD, or egg sensitization between the groups.

Biomarkers

We examined SPT size, total PN-IgE, and Ara h 2 IgE because these are laboratory values widely used in clinical practice, although Ara h 2 IgE is missing for some patients due to problems with availability of this test at the beginning of our study. The median SPT size for those who were low-dose PN tolerant was 4 mm compared to 9.5 mm for those who were non-tolerant (Table 1). Of the infants with PN-SPT ≤4 mm nearly all (29/30, 97%) tolerated PN, indicating these infants are likely safe to try PN at home or in an office setting, which falls in line with current guidelines. Those infants with PN-SPT >4 mm mostly tolerated PN as well (25/40, 63%). With a worst-case assumption that those who refused challenges were intolerant then the tolerance rate would be 54% (25/46).

There was considerable overlap in the PN-SPT size and PN-IgE levels between those who were tolerant and not tolerant, however Ara h 2 IgE appeared to help distinguish if an infant would react, especially in those who already had a large SPT >4 mm (median Ara h 2 IgE was 6.49 kU/L in those who reacted vs. 0.62 kU/L in those who were tolerant). Of those infants with SPT >7mm (n=21), tolerant infants (11/21, 52%) had a median SPT 9mm (8-11mm), PN-IgE 1.1 kU/L (range 0.3-10.7 kU/L) and Ara h 2 IgE 1.6 kU/L (range 0-9.57 kU/L). Those who reacted (10/21) had median SPT 10 mm (8-16mm), PN-IgE 4.6 kU/L (1.18-31.2 kU/L), Ara h 2 IgE 6.8 kU/L (0.63-19.3 kU/L). Ara h 2 IgE was the only marker that was significantly different between the two groups (p<0.03). See Figure 1 for a comparison of Ara h 2 IgE by SPT size. ROC curves using data from all patients showed that Ara h2 of 1.05 kU/L or above was associated with a 75% sensitivity and 72% specificity for PN reaction and a level of 6.05 kU/L or greater was associated with 58% sensitivity and 92% specificity (Area under the curve 0.8). If we assume that those who refused OFC were intolerant, the tolerance rate would be 37% (10/27). Age, sex, race, atopic dermatitis and egg sensitization did not affect tolerance regardless of SPT size.

• Download figure
• Open in new tab

Figure 1.

Ara h 2 IgE by SPT size and OFC outcome. Overall, Ara h 2 IgE 1.05 kU/L or above was associated with a 75% sensitivity and 72% specificity for PN reaction and a level of 6.05 kU/L or greater was associated with 58% sensitivity and 92% specificity (Area under the curve 0.8).

There were no biomarkers that could help predict an FPIES reaction in 3 patients. Two of these three patients had moderate to severe eczema, which is higher than the overall population of infants studied here, although clinical and statistical significance was not determined due to the small numbers.

Discussion

This study demonstrates successful low-dose PN tolerance in infants who are at high risk for PN allergy including those with the largest PN-SPT (>7 mm). Those who are ≤4 mm might be able to try PN at home or in a more general office setting, given that 97% were tolerant. The highest risk infants in our study would normally be advised to avoid PN based on current guidelines.^3,15-17 Our PN-OFC results indicate that this would omit many infants who could introduce low-dose PN and possibly avoid a lifelong PN allergy by taking advantage of the theoretical window of opportunity for development of tolerance.^{1,2,12,17,18,19}

Potential prevention of PN allergy is arguably the leading advance in food allergy in the last few decades given that the natural history of PN allergy is to persist, especially for children with PN-SPT>13mm or PN-IgE >5 kU/L²⁰. Early consumption appears to be critical in many smaller studies^19,21,22 and was confirmed with a moderately high level of certainty in a meta-analysis based on the LEAP, LEAP-On, and EAT studies²³. We also know that PN-OIT likely has better outcomes for sustained unresponsiveness in children less than 5 years who already have a known PN allergy¹⁹. Together, these findings compelled us to try to prevent PN allergy in as many children in our own population as possible.

Given the high impact of prevention and variability of cutoff values based on study site, testing materials, and study population, we offered OFC to our patients as the best way to determine true tolerance. Given the high level of parental anxiety about feeding PN in those who have siblings with FA (n=4) we thought that it was important to include these infants to ensure early introduction even if home introduction was feasible without testing^10-12. These infants were often referred anyway due to elevated PN-IgE previously obtained by the pediatrician.

We chose to introduce a low dose of 1 gram of PN protein during the PN-OFC instead of 3.9 grams as in LEAP and 5 grams as in standard OFCs. This dose was chosen based on our experience with infants and with PN-OFCs; we have seen that many infants are not easily able to consume more than 1 gram, especially when they are around 6 months of age and have only recently started solid foods. Furthermore the EAT trial demonstrated that it was difficult for families to be adherent given the variation in developmental readiness to eat solid food¹⁸. For this reason, we chose to challenge infants to a developmentally appropriate dose and increase more gradually at home as a more real world food introduction approach. We strongly recommended that parents give the 1 gram of PN two to three times per week without a long break in consumption based on the LEAP methodology. We also encouraged parents of the highest-risk infants to maintain access to epinephrine autoinjectors until further follow-up.

Based on our data, Ara h 2 IgE seems to be the most valuable test to lab predict reactions during PN-OFCs especially in those who already have a large SPT. PN-SPT and PN-IgE do not appear to be sufficient to exclude an infant from a low-dose OFC. Further studies are required to determine if PN-SPT and Ara h2 in tandem can provide enough assurance that an OFC is not indicated. However, given our current clinical setting within an academic institution with access to emergency and critical care, we feel comfortable offering OFC and managing reactions if the final outcome is to prevent a substantial number of children from becoming PN allergic.