Key words CT - contrast agents - drugs/reactions - safety
Introduction
Iodinated contrast media (ICM) are the most frequently utilized contrast agents worldwide.
More than 70 million CT scans were performed in the United States alone in 2007 [1 ] with a clear trend toward increased future utilization [2 ]. Based on this, it is estimated that the administration of ICM is performed at least
100 million times each year worldwide. Although low-osmolality ICM agents have an
excellent safety/risk profile, adverse drug reactions (ADRs) in general or hypersensitivity
reactions (HSRs) in particular occur in approximately 3 – 30 % and 1 – 3 %, respectively
[3 ]
[4 ]. HSRs are per definitionem so-called type B reactions [5 ] and therefore are not predictable. Despite this fact, several authors have identified
risks predisposing patients to HSRs [6 ], in particular a history of previous CM-related hypersensitivities [6 ]. In other words, adequate screening of patients prior to contrast administration
remains an important prerequisite for the safe performance of contrast-enhanced CT
(CECT).
Safe ICM administration is dependent on several preconditions, one of which is the
use of precise nomenclature. Imprecise or incorrect terminology potentially places
the patient at risk. One of the most important nonspecific terms within the context
of ICM-related ADRs is the concept of an “iodine allergy”. Indeed a true “iodine allergy”,
meaning an allergy against elemental iodine, would be lethal [7 ]. We hypothesize that this nomenclature is more problematic than helpful and poses
a potential safety risk. A thorough search of the available literature did not identify
a formal evaluation of patients with an “iodine allergy” undergoing contrast-enhanced
radiologic procedures. To verify the aforementioned hypothesis, patients with an alleged
“iodine allergy” listed in the medical record were analyzed. The goal of the present
paper is threefold: 1) to identify the possible clinical meaning of individual “iodine
allergies”, 2) to analyze the resulting clinical consequences/prophylactic actions
arising from alleged “iodine allergies” including possible ADRs that occurred after
prophylaxis, and finally 3) to recommend a practical method to ensure the safe administration
of ICM in future contrast-enhanced procedures for patients diagnosed with an “iodine
allergy”.
Patients and Methods
Investigated population and study design
The radiologic information system (RIS database) at a radiology department within
an academic medical center was searched to identify patients with a history of an
“iodine allergy” (group 1) for whom imaging had been subsequently ordered. Either
a full-text search of all relevant dates (1997 – 2015) or a screening by individual
timeframes (01/2014 – 03/2014 and 01/2015 – 12/2015) was performed. Two age-, sex-,
and procedure-matched control groups were identified within the RIS database consisting
of patients with ADRs related to ICM who subsequently underwent imaging. These were
divided into patients for whom a nonspecific “iodine contrast medium (ICM) allergy”
was documented but the exact culprit ICM was unknown (group 2) and those for whom
a specific culprit agent for ICM allergy was known (group 3 or substance allergy group).
The institutional review board (IRB) approved this study. Informed signed consent
was waived because of the retrospective nature of the study.
The inclusion criteria were adult patients (> 18 years old) with a history of a so-called
“iodine allergy” for whom contrast-enhanced CT imaging was ordered between January
1997 and December 2015. All documented explanations or descriptions of the term “iodine
allergy” were explored in the medical record. Epidemiologic baseline data were collected
including age, sex, medical history, history of allergic disorders, any available
details about previous reactions to ICM, the administered contrast materials, the
type of radiological procedure, prophylactic actions (e. g. premedication or change
of procedure), and finally the outcome (i. e., whether a reaction ultimately occurred
after CM administration). The severity of both historical reactions in the patients
with an “iodine allergy” and of any subsequent reactions to the administration contrast
material was classified according to the previously published grading system (see
[Table 1 ]) [8 ]. We also screened the data for AEs or HRSs that occurred in spite of different prophylactic
actions. Evidence of a CM-related AE was analyzed using the Naranjo adverse drug reaction
probability scale [9 ].
Table 1
Clinical features and severity grading of CM-induced reactions from the literature
[8 ].Tab. 1 Klinische Symptome und Schweregrade der KM-induzierter Reaktionen nach Literatur
[8 ].
severity grade
features
actions
1 – mild
heat/cold feeling, nausea, vomiting, flushing, mild urticaria or other anatomically
limited skin disease (e. g. fixed drug eruption), itching
no treatment necessary
2 – moderate
angioedema, generalized urticaria or other skin disease (e. g. maculopapular exanthema),
bronchospasm, dyspnea, stridor, wheeze, moderate hypotension, tachycardia, chest or
throat tightness, dizziness
anti-allergy drug treatment (e. g. with corticosteroids and H1-antihistamines) can
be performed in the radiology department
patient can be discharged after remission of symptoms
3 – severe
prolonged hypotension, ventricular fibrillation, angina, myocardial infarction, cardiac
arrest, severe bronchospasm, pulmonary edema, respiratory arrest, cardiovascular collapse,
confusion, loss of consciousness, convulsion, Stevens–Johnson syndrome, toxic epidermal
necrolysis, contrast-induced nephropathy, nephrogenic systemic fibrosis
hospitalization necessary
Statistics
Continuous variables were expressed as mean ± standard deviation (SD). Categorical
variables were expressed by frequencies and percentages. The Chi-square test (Pearson)
and Fisher’s exact test were used to compare qualitative (categorical) variables and
the Student’s t-test or the Mann-Whitney U-test to compare quantitative (continuous)
variables.
Multiple logistic regression analysis was used to identify the risk factors related
to AEs in patients in groups 1 and 2. The magnitude of risk was calculated with the
odds ratio (OR). 95 % two-sided Cornfield`s confidence intervals (CI) were constructed
according to Clopper-Pearson.
P-values (two-tailed tests) < 0.05 were considered statistically significant. Statistical
analysis was performed with Stata version 12 software for Windows (StataCorp LP, Lakeway
Drive, College Station, Texas, USA).
Results
Patient profiles
Group 1 consisted of 300 patients (men n = 108/36.0 %) with a documented history of
the alleged “iodine allergy”. The mean age of the study population was 63.3 ± 13.3
years. The mean age of men and women was identical (63.3 ± 12.5 years vs. 63.3 ± 13.7
years; p = 0.95).
Group 2 consisted of age-, sex-, and procedure-matched patients (n = 230) diagnosed
with an “ICM allergy” to an unknown culprit agent. The proportion of men was 39.1 %
and the mean age was 61.5 ± 13.6 years, does not statistically significantly different
from group 1 (p = 0.94).
Group 3 consisted of age-, sex-, and procedure-matched patients (n = 70) with a clear
and specifically documented ICM allergy (e. g., “iopromide allergy”). 15.7 % of the
patients were men and the mean age in the group was 59.3 ± 14.0 years, again this
was not statistically significantly different from the other groups (p = 0.96).
Diagnostic explanation of the term “iodine allergy”
275 out of the 300 patients of group 1 were clearly diagnosed as having an “iodine
allergy” and 25 (8.3 %) were documented as having a “suspected iodine allergy”. For
the analysis, these two subgroups were grouped together and defined as “iodine allergy”
patients.
In 253 (84.3 %) of the group 1 patients the term “iodine allergy” was not further
explained. In 47 (15.7 %) patients the diagnosis was further specified in reports
as: 24 (8.0 %) cases with CM allergy, 20 (6.7 %) antiseptic allergies, and 4 (1.3 %)
seafood allergies. In one case, both CM allergy and antiseptic allergy were documented.
No instances of an amiodarone drug allergy, sometimes referred to as an “iodine allergy”
as well [10 ], were found in our sample.
Disease spectrum, prophylaxis, and subsequent adverse reactions
Generally, the more specific the diagnosis, the more information was available. Only
11.7 % of patients in group 1 had documented symptoms pertaining to the previous hypersensitivity
reaction, while in 88.3 % both the clinical features and the severity of the so-called
“iodine allergy” remained unknown ([Table 2 ], [3 ]). However, in patients either with a nonspecific ICM allergy (group 2) or a specific
ICM allergy (group 3), previous clinical symptoms and severity were documented in
65.2 % and 90 % of cases, respectively (see [Fig. 1a, ]
[Table 2 ], [3 ] for details).
Table 2
Severity of symptoms and their frequency of occurrence within the three analyzed groups
for past ADRs described in the medical record. Of note, the majority of patients in
group 1 did not have sufficient documented information to enable the classification
of symptoms.Tab. 2 Schwere der Symptome und ihre Häufigkeit in der letzten dokumentierten UAWs in den
drei analysierten Gruppen. Es ist bemerkenswert, dass bei den meisten Patienten der
Gruppe 1 keine ausreichende Information vorhanden war.
severity grade
group 1
“Iodine allergy“
group 2
“ICM allergy“
group 3
specific allergy
1 – mild
37.1 %
61.3 %
76.2 %
2 – moderate
22.9 %
20.7 %
14.3 %
3 – severe
40.0 %
18.0 %
9.5 %
absolute numbers
35
150
63
Table 3
Frequency (absolute numbers and percentages) of past symptoms in groups 1 – 3 (n.s. = not
significant).Tab. 3 Häufigkeit (absolute Anzahl und prozentual) der klinischen Symptome der letzten KM-UAW
der Gruppen 1 – 3 (n.s. = nicht signifikant).
clinical features and symptoms
group 1
“Iodine allergy“
group 2
“ICM allergy“
group 3
Specific allergy
p-value
group 1 vs. 2
group 1 vs. 3
group 2 vs. 3
unknown
88.3 %
34.8 %
10.0 %
< 0.0001
< 0.0001
< 0.0003
known
11.7 %
65.2 %
90.0 %
< 0.0001
< 0.0001
< 0.0003
cutaneous symptoms
4.0 %
34.8 %
48.6 %
< 0.0001
< 0.0001
< 0.04
shivering/fatigue
0.0 %
1.3 %
0.0 %
n.s.
n.s.
n.s.
delayed reaction
0.3 %
3.5 %
4.3 %
< 0.03
n.s.
n.s.
cardiopulmonary symptoms
3.0 %
10.9 %
11.4 %
< 0.0007
< 0.005
n.s.
rhinitis/sneezing
0.0 %
0.4 %
7.1 %
n.s.
n.s.
n.s.
gastrointestinal symptoms
0.3 %
12.2 %
24.3 %
< 0.0004
< 0.0001
< 0.02
neurological/ophthalmological
0.0 %
4.8 %
1.4 %
n.s.
n.s.
n.s.
anaphylaxis
4.7 %
7.8 %
7.1 %
n.s.
n.s.
n.s.
absolute numbers
300
230
70
Fig. 1 a Correlation between the prevalence of ADRs (black bars – left y-axis) and the prevalence
of nonspecific clinical features (grey line – right y-axis) (top). b Correlation between the frequencies of prophylactic management/frequencies of CT
examinations (bars – left y-axis) in the three groups and the frequency of cancelled
scans (grey line – right y-axis) (bottom).
A spectrum of different prophylactic actions was performed in the three groups prior
to subsequent diagnostic imaging. In the majority of procedures, some prophylaxis
was documented, but up to 10 % of the examinations were performed without any prophylactic
measures ([Table 4 ]). The prophylactic actions varied throughout the studied groups. For instance, non-contrast
CT imaging was performed more often in group 1 than in group 2 (p < 0.02) and group
3 (p < 0.003). Moreover, in the control groups (2 and 3), the culprit ICM was often
omitted and another ICM was injected, while in group 1 the chemical class of contrast
agents was changed to barium or gadolinium (see [Fig. 1b, ]
[Table 4 ] for details). In general, the more accurate the diagnosis, the higher the likelihood
that the originally ordered examination and modality was performed ([Fig. 1b ]).
Table 4
Comparison of prophylaxis/clinical consequences of previous ADRs in groups 1 – 3 (n.s. = not
significant).Tab. 4 Vergleich der Prophylaxe / klinischen Konsequenzen bei vorausgegangenen UAWs der
Gruppen 1 – 3 (n.s. = nicht signifikant).
clinical consequences
group 1
“Iodine allergy“
group 2
“ICM allergy“
group 3
Specific allergy
p-value
group 1 vs. 2
group 1 vs. 3
group 2 vs. 3
no consequence
10.3 %
5.2 %
8.6 %
< 0.04
n.s.
n.s.
altered examination:
36.7 %
29.1 %
18.6 %
n.s.
< 0.005
n.s.
23.7 %
44.8 %
30.0 %
< 0.0001
n.s.
< 0.03
17.7 %
17.0 %
17.1 %
n.s
n.s.
n.s.
5.7 %
2.6 %
0.0 %
n.s.
–
n.s.
2.3 %
0.0 %
0.0 %
–
–
–
1.7 %
0.0 %
0.0 %
–
–
–
1.7 %
1.7 %
25.7 %
n.s.
< 0.0001
< 0.0001
0.0 %
0.4 %
0.0 %
–
–
–
0.7 %
0.0 %
0.0 %
–
–
–
absolute numbers
300
230
70
The less specific the terminology utilized to describe previous ADRs, the greater
the prevalence of AEs/ADRs in the three patient groups (r = 0.981; p < 0.00 001).
In fact, no group 3 patients experienced an AE/ADR, whereas 9 (3.0 %) patients in
group 1 experienced a reaction. A reaction occurred in one group 2 patient (0.4 %)
(see [Fig. 1a ], [Table 5 ] for details).
Table 5
Overview of individual patients with ADRs despite prophylactic action.Tab. 5 Überblick einzelne Patienten mit UAWs trotz prophylaktischer Maßnahmen.
severity grade
classification
patient age, sex
prophylactic management
adverse event
1 – mild
group 1
41, female
none
vomiting
group 2
60, female
none (premedication refused)
cutaneous symptoms (neck)
group 1
68, female
none
local reaction by PVP iodine
group 1
61, male
premedication[1 ]
conjunctivitis
group 1
47, female
change of modality (MRI)
cutaneous symptoms
2 – moderate
group 1
62, female
premedication1
larynx edema, cutaneous symptoms
group 1
76, male
premedication1
breathing difficulties, nausea
group 1
49, female
change of modality (MRI)
tachycardia
group 1
64, female
change of modality (MRI)
erythema, palpitations, arrhythmias, hypertension
3 – severe
group 1
75, male
change of CM type (barium)
aspiration
1 Intravenous injection of 2 mg clemastine and 125 mg methylprednisolone.
All reactions in group 1 were ADRs: eight reactions could be related to the CM administration,
and one reaction occurred due to the application of polyvidone (PVP) iodine. In group
2, only one woman who had refused premedication had a reaction. No patients in group
3 experienced an ADR following the prophylactic action. An OR of 9.24 for an ADR (95 %
CI 1.1642 to 73.4541; p < 0.04) was calculated for group 1 as compared to the two
other groups.
Discussion
Previous papers dealing with “iodine allergy” and ICM application either focused on
implying that this kind of allergy does not exist or on questionnaire results [6 ]
[11 ]
[12 ]
[13 ]
[14 ]
[15 ]
[16 ]. This study examines for the first time the incidence of subsequent ADRs in patients
referred to as having a history of “iodine allergy” versus the ADR rate in cases where
a specific or nonspecific history of ICM allergy is documented. 9 of 10 ADRs occurred
in the “iodine allergy” group. A clear correlation between the precision of diagnosis
and propensity to have a subsequent reaction was found.
The diagnosis “iodine allergy” is inexact and associated with insufficient documentation
The first issue to be addressed in this study was the analysis of the term “iodine
allergy”, which remained unexplained in the majority (84.3 %) of the patient cohort.
Only in a remaining 15.7 % patients were further annotations such as CM allergy, antiseptic
allergy, and fish/seafood allergy documented. As expected, the definition of an “iodine
allergy” was heterogeneous. According to the literature, the term “iodine allergy”
describes four different conditions: ICM allergy, disinfectant allergy, amiodarone
allergy and/or seafood (shellfish or fish) allergy [10 ]
[12 ]
[16 ]
[17 ].
Because of its vagueness, we recommend that the term “iodine allergy” should be omitted
in the future and replaced with the offending agent when possible and the associated
reaction. Patients with an allergy to iodinated antiseptics, amiodarone, or seafood,
for example, can safely receive all ICMs without premedication (except in the case
of a concurrent ICM allergy).
Antiseptics based on PVP iodine may induce local irritation or in a minority of cases
an allergic reaction such as a so-called type IV allergy or contact dermatitis [18 ]. Immediate type (type I allergy) reactions with allergen-specific IgE have been
rarely described [19 ]. All reactions should be diagnosed by adequate test procedures (patch tests for
delayed reactions and prick/intradermal tests for immediate reactions). One should
be aware that polyvidone iodine is broadly used as an antiseptic compound not only
in disinfectants but also in different drugs and sometimes even in ICM solutions [20 ]. Unfortunately, current package labeling for available ICM solutions does not mention
polyvidone iodine content. Patients with PVP hypersensitivity should receive an alternative
antiseptic material (e. g. chlorhexidine, alcohol). In this study, one patient in
group 1 was treated with PVP iodine for an interventional radiological procedure and
subsequently developed an allergic reaction.
Seafood allergies were also initially assumed to be an “iodine allergy” [12 ]
[13 ]
[16 ]
[21 ]. Now it is clear that tropomyosin, a muscle protein of crustaceans, or parvalbumin,
a muscle protein of fish, contribute to immediate type allergies in patients with
seafood allergy [13 ]. Therefore, patients with hypersensitivity (allergic or non-allergic) to seafood
should omit these in their diet, but can receive all iodinated CM.
Prophylactic actions
Patients truly at an increased risk for iodinated CM ADR (i. e. history of a CM-induced
hypersensitivity reaction) should undergo a special prophylactic management program.
In this study, clinical management was altered in all patient groups; only in a minority
of patients was no special prophylaxis task performed ([Table 4 ]). However, the nonspecific term “iodine allergy” and the often associated lack of
clinical documentation about the ADR led to uncertainty in the management of these
cases.
In approximately 20 % of patients in group 1, an anti-allergic premedication was applied
before an ICM was subsequently given ([Table 4 ]). Premedication as a prophylactic procedure has both advantages and disadvantages.
If the patient has a true allergy to ICM, it may be helpful. However, premedication
protocols lack standardization, and even drugs utilized for premedication can induce
ADRs. Finally, breakthrough reactions can occur even with premedication [22 ]. In special cases, premedication may be indicated; however, it is far from a universal
solution.
In most patients with an “iodine allergy”, the modality was changed to an MRI examination.
This is in line with several papers that recommend MRI, GBCA application for CT, or,
to a lesser extent, ultrasound in patients with an ICM allergy or “iodine allergy”
[23 ]
[24 ]
[25 ]
[26 ]. However, GBCA can induce both AEs and hypersensitivity reactions. Moreover, even
initial contact with a GBCA can lead to an allergic reaction as reported previously
[27 ] and also found in the unpublished observations of the authors. In fact, the change
to MR examinations in our study group was accompanied by AEs as well ([Table 5 ]). In 4 out of 51 examinations when the modality was changed, ADRs were observed.
A change of the CM class was performed rarely: twice barium was used instead of ICM
for CT, once GBCA, and once CO2 . Unfortunately, one patient received per os barium in lieu of ICM resulting in the
aspiration of barium. As this case demonstrates, changing of the CM class is not necessarily
a safe alternative. A change of the contrast medium within the group of iodinated
CMs was performed in groups 2 and 3 ([Table 4 ]). This prophylactic task was not accompanied by adverse reactions, and can be recommended
as described previously [28 ].
Some patients received gastrointestinal CM application as prophylaxis. However, it
should be noted that there is still some risk related to this application route [29 ].
In 16 cases the examinations were cancelled because the patients had a positive history
of an “iodine allergy” ([Table 4 ], [Fig. 1a ]). The cancellation rate of radiological examinations declined with the quality of
information concerning the allergy from 5.3 % in the reference group to 2.2 % in group
2. In group 3, no examinations were cancelled.
Adverse events
Despite a prophylactic management program, 9 patients in group 1 ultimately experienced
ADRs ([Table 5 ]). One ADR occurred in group 2, and none in group 3. Therefore, the uncertainty of
the diagnosis associated with term “iodine allergy” is associated with an increased
risk for AEs/ADRs. For safety, it is thus recommended to eliminate the term “iodine
allergy” from the medical lexicon.
Limitations of this study
One limitation of our study is its retrospective nature. Very likely, the detected
number of patients with “iodine allergy” is incomplete, because the diagnosis “iodine
allergy” is mentioned in variable sections of the RIS database. Moreover, due to different
spelling styles (e. g. “iodine-allergy”, “iodine allergy”, “iodineallergy”, “iodine”),
it is difficult or impossible to identify all patients with this diagnosis. Therefore,
the current analysis was limited to 300 cases.
Conclusion
Taken together our data show, not unexpectedly, that the term “iodine allergy” covers
different medical entities. We agree with Dewachter et al. [13 ] who stated that “asking a patient if he/she is ‘allergic to iodine’ is a question
that should be avoided because its significance is null.” To this, it can now be added
that the diagnosis of “iodine allergy” is associated both with missing information
about the clinical symptoms of previous adverse reactions and an increased risk for
ADRs even after special prophylactic management. Therefore, to better understand a
patient’s risk and to start an effective prophylactic program, an ICM-allergy work-out,
the exact name of the culprit compound, and the clinical symptoms of the last ADR
must be determined and possibly supplemented with the results of in vivo and in vitro
allergy tests.
