To the Editor
Allergic disease diagnosis begins with a clinical history and physical
examination to identify allergic symptoms associated with a relevant
allergen.1 Immunoglobulin E (IgE) antibody
sensitization is then confirmed with in vivo skin tests or in vitro
blood test. If there is a mismatch between history and these primary
sensitization diagnostic tests, then a provocation test may be required.
During oral food challenges (OFC), test administrators must be specially
trained in acute severe allergic reaction management and OFCs should
only be performed with immediate access to intensive care
units.2 Therefore, evaluating specific IgE (sIgE)
sensitization are often used to help predict the outcome of the OFC
before referring to a well-equipped tertiary hospital.
Skin prick tests are not the preferred diagnostic tool if physicians are
not well-trained allergists and in many countries, and typically, the
number of antigens that can be evaluated at one time using serologic
tests is limited. These limitations lead physicians to prefer to perform
the multiple-antigen simultaneous tests for food allergy (FA) diagnosis,
the results of which are semi-quantitative and often limited in
predicting OFC outcome.3 The purpose of this study was
to analyze the possibility of sIgE antibody detection using troops from
self-reported food allergic symptoms.
From May 2011 to December 2013, medical records of 377 patients (3 years
old or younger) who visited the Department of Pediatrics at Ajou
University Hospital were collected for egg white-, cow’s milk-, walnut-,
and soybean-sIgE sensitization (sIgE ≥ 0.35 kU/L, UniCAP, ThermoFisher
Scientific, USA). Five components of patients’ clinical history were
collected: 1) way of exposure: direct (ingestion) or indirect (skin,
inhalation); 2) type of exposure: isolated or mixed; 3) onset time: ≤2
hours or >2 hours; 4) symptom characteristics: anaphylaxis,
urticaria, itchiness, vomiting, or diarrhea; and 5) consistency:
negative past history or positive past history with consistency or
inconsistency. Each clinical history was classified into class 1:
direct-isolated intake resulting in anaphylaxis or hives without
inconsistency; class 2a: class 1 with inconsistency, class 2b:
indirect-mixed intake resulting in anaphylaxis or hives regardless of
consistency, or class 2c: direct/indirect-isolated/mixed intake
resulting in itch without hives, vomiting, or diarrhea without
inconsistency; or class 3: class 2c with inconsistency or asymptomatic
to direct, isolated exposure. All class 1 cases that were not of
isolated ingestion were considered vague and were reclassified as class
2. An exception was made for anaphylaxis due to skin contact, which was
still regarded as class 1. If the symptom onset time was recorded as
“next day,” these classes were also reclassified by adding 1 (Table
1). Receiver operating characteristic curves were analyzed using SPSS
version 22.0 (SPSS Inc., Chicago IL). The study was approved by the
Institutional Review Board of Ajou University Medical Center
(MED-KSP-12-381).
In class 1 cases, the sensitization rate (i.e., sIgE positive) was the
highest in walnuts (9 of 10; 90.0%), followed by hen’s egg white (49 of
55; 89.1 %) and cow’s milk (71 of 81; 87.7%). However, soybean-sIgE
sensitization was only found in 5 of 8 cases (Table 2). Meanwhile, in
class 2 cases, sIgE positivity was found in only 2 of 8 cases for
soybean (25%) and 10 of 33 cases for cow’s milk (30.3%). However,
within class 2 cases, sIgE-positivity was also high for egg white (17 of
22 cases; 77.3%) and walnut (2 of 3 cases; 66.7%). Egg white-sIgE
demonstrated an area under the curve (AUC) of 0.717 and a positive
predictive value of 89.1% in class 1 cases. When class 2 cases were
included in this analysis in addition to class 1 cases, the AUC of egg
white-sIgE positivity increased to 0.750, the negative predictive value
increased to 68.6% compared with 47.5% for class 1 cases only, and the
accuracy increased from 67.2% for class 1 cases only to 77.6% for
class 1 and 2 cases. However, for cow’s milk- and walnut-sIgE
sensitization rate, class 1 cases were the most predictable (AUC of
0.790 and 0.755, respectively), with an accuracy rate of 78.0% and
76.5%, respectively. Soybean-sIgE sensitization rate had a lower AUC of
0.662 in class 1 cases than other allergens (Table 3).
FAs can be highly anticipated if anaphylaxis or objective symptoms are
repeated more than once within a few hours after ingestion of a specific
antigen.4 Accordingly, class 1 cases were limited to
patients who experienced an immediate onset of objective symptoms
(anaphylaxis or hives) after isolated ingestion. This study included
patients aged < 3 years who often refused to perform an OFC
using unfamiliar foods; hence, this study is limited as an OFC was not
often performed. Therefore, the medical history technique described in
this study cannot be deemed a definitive method for confirming FA. In
fact, in class 3 cases, including asymptomatic or subjective
manifestations to allergens, approximately 20–40% were sIgE positive
and highly likely to tolerate the specific allergen and therefore would
not be deemed to have an FA (sIgE positive: egg white, 38.5%; milk,
29.4%; walnut, 25.0%; and soybean, 33.3%). While it appears that many
patients must confirm FA through OFC, even if there is sIgE
sensitization, this study provides information for primary physicians
who may be limited in implementing OFC. The rate of sIgE sensitization
in class 1 cases was high (87.7–90.0% depending on antigen),
supporting the detailed collection of important elements of the history
taking pertaining to allergens. This study suggests that a clinical
history of allergen exposure and characteristics of reactions can help
determine whether sIgE sensitization and further OFC testing are
required.
Egg white, cow’s milk, and tree nuts are known as highly likely to be
OFC positive, especially if patients have a history of adverse reactions
within 5 minutes of direct exposure.5 In infants and
young children who are allergic to foods, it is rare that they
experience respiratory or gastrointestinal symptoms alone and this study
continues to support this outcome. When collecting medical history from
patients for a suggested FA, physicians must record all symptoms that
occur sequentially, including the type of intake, symptom development
and characteristics, and previous allergic symptoms after ingestion.
This study demonstrated differences in sIgE sensitization according to
the type of food antigen type, and it would be beneficial to add
additional history-taking techniques based on these results. For
instance, the accuracy of the egg white-sIgE sensitization testing
increased when interpreting both class 1 and class 2 cases based on
patient history, as it included objective cases of symptoms caused by
both direct and indirect exposures. However, the accuracy decreased in
the milk-sIgE sensitization testing of class 2 cases. It is thought that
this may be due to young patients’ subjective symptoms, such as
gastrointestinal symptoms alone. As walnuts were frequently consumed in
a mixed form rather than isolated, there was a possibility of an
increase in accuracy for the walnut-sIgE testing of both class 1 and
class 2 cases.
Prior to FA diagnosis, a detailed medical history could screen for
potential allergens to be checked for IgE sensitization; however, it
does depend on the type of allergen. The sIgE sensitization of egg
white, cow’s milk, and walnuts are easy to predict by history alone but
methods for increasing the predictability of each allergen are slightly
different.