Description of studies

1. Diagnostic characteristics of cutaneous tests for immediate HSR

The diagnostic characteristics of cutaneous tests for acute (immediate) hypersensitivity reactions (HSR) to contrast media (CM) were evaluated in 4 studies (Caimmi, 2010; Kim, 2013; Salas, 2013; Sesé, 2016).

Caimmi (2010) studied 159 patients. Patients were tested with the culprit iodine-based contrast medium (ICM) and a set of other ICM if they were positive for the culprit ICM or if its name was unknown. To know which ICM was involved, either patients already knew which drug had supposedly caused the reaction, or the authors contacted the hospital in which the reaction had occurred. The ICM used were: amidotrizoate, ioxithalamate, iopamidol, iohexol, ioversol, iopromide, iomeprol, iobitridol, iodixanol and ioxaglate. Skin tests were performed firstly as prick tests with the undiluted commercially available solution and then, if negative, by intradermal tests (IDT) at a 1: 10 dilution. Prick tests were considered positive if, after 15 min, the size of the weal was at least 3 mm in diameter. For IDT, positivity was considered when the size of the initial weal increased by at least 3 mm in diameter after 15 to 20 min, considering as non-irritant a maximum dilution of 1/10. The negative predictive value was defined as the proportion of patients with negative skin test results to at least one ICM at first testing who had a further injection with that ICM without reacting. One hundred participated (75.5% participation rate). Seventy-one of them (59.2%) were females of a median age of 56 (45–65) years. Most of the reactions were immediate (101 out of 120, 84.2%), and in two cases, it was not possible to assess whether the reaction was immediate or nonimmediate. For immediate reactions, 42 (41.6%) were of grade 1, 34

(33.7%) of grade 2, 20 (19.8%) of grades 3 and five (4.9%) of grade 4. Only one (5.9%) of the 17 nonimmediate reactions was moderate, all the others were mild (16 to 94.1%).

Kim (2013) retrospectively included 1048 patients. The mean (SD) age was 55.1 (14.5) years; 501 (47.8%) were male. Intradermal test with the RCM that was to be used in the pending nonionic CM-enhanced CT was performed just before the CT examinations. The nonionic CM used was iopromide, iomeprol, iohexol, and iodixanol. Intradermal tests were conducted on the volar surface of the forearm with a negative control, saline. A 1:10 solution of contrast medium (0.03 to 0.05 mL), which has been accepted as a non-irritating concentration, was gently injected into the skin to produce a small superficial bleb of 2 to 4 mm. Skin test positivity was determined when the diameter of the wheal increased by at least 3 mm, and surrounding erythema was observed after 15 to 20 minutes. If a patient had a negative response to skin tests, CT was performed as scheduled (provocation). Of the 376 patients previously exposed to CM, 61 (16.2%) had a history of at least 1 mild CM-associated reaction: 56 (91.8%) had immediate and 5 (8.2%) nonimmediate reactions.

Salas (2013) included 90 patients with a history of immediate HSR after contrast media (CM). Immediate HSR was classified according to the Ring and Messmer scale. Skin tests (ST) were carried out using the following CM: iobitridol, iomeprol, iodixanol, iohexol, ioversol, iopromide and ioxaglate. Prick tests were performed using undiluted CM and IDT using 10- fold dilutions. In those with a negative ST, a single-blind placebo-controlled provocation test was performed with the CM involved, as described. In patients with a positive ST and/or provocation test, a basophil activation test (BAT) was performed with iohexol (3; 0.3 mg/ml), iodixanol (3; 0.3 mg/ml), iomeprol (3.5; 0.35 mg/ml) and ioxaglate (5.8; 0.58 mg/ml) (based on dose–response curves and cytotoxicity studies). The median age of the subjects evaluated was 54.50 ± 27 years; 63 (60%) were women. The CM involved in the reaction was iomeprol in 26 cases (28.89%), iodixanol in 19 (21.11%), iohexol in 11 (12.22%), iopromide in 9 (10.00%) and unknown in 25 (27.78%). According to the clinical history, most cases developed reactions with skin involvement (65.65% urticaria/ angioedema and 30% generalized erythema), and only 4.44% had airway or cardiovascular involvement. Regarding symptom severity, 69 cases (76.71%) had grade I reactions, 18 (20%) grade II and 3 (3.33%) grade III. No patients had grade IV reactions.

Sesé (2016) included 37 patients with a definite history of immediate HSR due to iodine- based contrast media (ICM). Immediate HSR was classified according to the Ring and Messmer scale. Skin tests were performed at least 6 weeks after the HSR on the volar forearm with the suspected ICM and with four other ICM. Skin prick tests (SPTs) involved freshly prepared undiluted ICM commercial solutions, and intradermal tests (IDTs) were performed successively with 100-fold and then 10-fold solution diluted in 0.9% sterile saline. Saline and chlorhydrate histamine were negative and positive controls, respectively. In total, 37 patients (24 women, mean age 49.3 years at the time of the reaction) completed the tests. The clinical severity of the reaction was grade I for 26 (70%), grade II for 4 (11%), and grade III for 7 (19%); 35 (95%) reported skin or mucosal symptoms, including pruritus (n = 11), facial erythema (n = 6), generalized erythema (n = 20), urticaria (n = 7), and angioedema (n = 5).

2. Diagnostic characteristics of cutaneous tests for non-immediate HSR

The diagnostic characteristics of cutaneous tests for delayed (nonimmediate) hypersensitivity reactions (HSR) to iodine-based contrast media (ICM) was evaluated in one study (Torres, 2012).

Torres (2012) included a total of 161 subjects with a history of a nonimmediate reaction imputable to at least one CM was evaluated. One patient who developed Stevens–Johnson syndrome was not included. The median age was 58.5 years (IR: 48.85 to 66.5) with 82 men (50.9%). According to the information obtained from the clinical history, the CM involved in the reaction were iomeprol in 53 (32.9%), iodixanol in 46 (28.6%), iohexol in 27 (16.8%),

iobitridol in 4 (2.5%), ioversol in 3 (1.9%), iopromide in 3 (1.9%), ioxaglate in 2 (1.2%) and

unknown in 23 (14.3%). According to the clinical history, 108 cases (67.1%) developed symptoms compatible with exanthema and 53 (32.9%) with delayed urticaria. Regarding symptom severity, 16 cases (9.9%) had mild reactions, 143 (88.8%) moderate reactions, and 2 severe reactions (1.2%) consisting of desquamative exanthema. Concerning the number of episodes, 132 cases (82%) had one episode and 29 cases (18%) two episodes.

3. Other tests

Three studies analysed different tests to determine hypersensitivity to contrast media (Kim, 2019; Meucci, 2020; Schrijvers, 2018).

Kim (2019) in a prospective cohort studied 36 patients with a history of immediate adverse drug reactions to radiocontrast media (RCM), presenting at the Allergy and Asthma Clinic of Severance hospital in South Korea from 2017 to 2018. Mean age was 57.3 ± 13.9 years and 69.4% (n=25) was female. The index test was intradermal testing (IDT) with diluted (1:10) RCM: iobitridol, iohexol, iopamidol, iopromide, and iodixanol. The IDT was considered positive when the diameter of the initial wheal had increased ≥3mm and was surrounded by erythema, confirmed at 20 minutes and at 3 days after IDT. The comparator test was similar to the index test, only performed with undiluted RCMs. No reference test was performed.

Meucci (2020) studied retrospectively 98 patients with previous reactions to iodinated contrast media (ICM) presented at the Allergology Unit in a hospital in Italy, from 2015 to 2018. Median (range) age was 65.6 (23-90) years and 54.2% (n=53) was female. The index test was the (less sensitive) skin prick test with undiluted ICMs: iohexol, iopromide, iodixanol, iopamidol, and ioversol. The skin test was considered positive when the diameter of the initial wheal had increased ≥3mm and was surrounded by erythema after 15 minutes. Furthermore, a distinguishment was made between immediate hypersensitivity reactions (IHR) (<1 hour after ICM administration) and delayed hypersensitivity reactions (DHR) (>1 hour after ICM administration). The comparison test was an IDT with diluted (1:10) ICM: iohexol, iopromide, iodixanol, iopamidol, and ioversol. The IDT was considered positive

when the diameter of the initial wheal had increased ≥3mm and was surrounded by erythema after 20 minutes. The reference test was a DPT, where the choice of ICM was based on the following: in case of a mild, recent (<12 month) reaction with negative skin tests for the culprit ICM, the DPT was performed with the culprit ICM. In case participants refused administering of culprit ICM, or if culprit ICM was unknown, another ICM was chosen. A subgroup of patients was re-exposed to ICM as part of their regular medical care; this re-exposition was used as a reference test to analyse their entire diagnostic protocol (skin tests + DPT).

Schrijvers (2018) in a retrospective cohort studied 597 patients with a history of ICM- mediated drug hypersensitivity reaction, presenting at the Allergy Department of the University Hospital, France, February 2001 to September 2014. Median (range) age was 60 (13-92) years and 68.0% (n=406) was female. The index test was a skin prick test with undiluted ICM: amidotrizoate, ioxitalamate, iopamidol, iohexol, ioversol, iopromide,

iomeprol, iobitridol, iodixanol, and ioxaglate. The skin test was considered positive when the diameter of the initial wheal had increased ≥3mm and was surrounded by erythema after 15 minutes. When the skin test was negative, and intradermal test (IDT) was performed as well. The IDT was considered positive when the diameter of the initial wheal had increased ≥3mm and was surrounded by erythema after 20 minutes. No reference test was performed, but

re-exposure to a skin test negative ICM occurred in 233 (39%) patients as part of their regular medical care.

4. Hypersensitivity reactions to gadolinium-based contrast agents (GBCA)

For GBCAs there was even less literature available, as hypersensitivity reactions to these agents are infrequent with an estimated prevalence of 0.004%-0.7% (Ahn, 2022). Skin tests are performed only in case reports or small case series and outcome measures as NPVs can therefore not be calculated (Gallardo-Higueras 2021, Grüber 2021). As pathogenetic mechanisms for GBCA-mediated hypersensitivity reactions are considered similar to those elicited by ICM and skin tests are performed according to comparable protocols, the recommendations for GBCA are extrapolated from those for ICM.

Results

Due to the heterogeneity in study designs, reported outcomes and follow up times, pooling of data could not be performed.

1. Diagnostic characteristics of cutaneous tests for immediate HSR

Caimmi (2010) revealed that ICM skin tests were positive in 21 patients (17.5%). Seventeen of them (80.9%) had a history of immediate reaction (four with grade 1, eight grade 2, four grade 3 and one grade 4). Prick tests were all negative. IDT were positive at 20 min for 15 patients with an immediate history and for the patient with unknown chronology. Caimmi (2010) found one single false negative; the negative predictive value of ICM skin tests was 96.6% (95% CI: 89.9 to 103.2).

Kim (2013) showed that among the 1046 patients who had negative responses on skin tests, 52 (5.0%) showed immediate-type adverse reactions after CT using radio contrast media.

However, most reactions were mild and cutaneous, such as pruritus, urticaria, and mild angioedema. Only 1 patient (0.1%) had a grade II moderate immediate reaction accompanied by breathing difficulty and mild laryngeal oedema, which were relieved with an antihistamine. The negative predictive value of the pre-screening skin test for immediate hypersensitivity reactions before contrast media administration was 95.0%. The negative predictive value of the skin test for immediate hypersensitivity reactions in patients with a history of contrast media hypersensitivity reactions was 80.3% (n= 49/61) and that in patients without a history was 95.9% (n= 945/985).

Salas (2013) showed that five subjects (5.56%) had a positive skin test: three by prick test (one to iodixanol, one to iomeprol and one to iohexol) and five by intradermal testing (four to iohexol, three iodixanol and two to iomeprol). In cases with a negative skin test to all CM tested (N = 74), provocation test was carried out with the culprit CM if known, being positive in three cases: one to iodixanol, one to iomeprol and one to iodixanol, iohexol plus iomeprol. In total, 11 patients with a negative ST refused to undergo a provocation test, resulting in a negative predictive value to immediate hypersensitivity reactions of 95.26%. Eight (8.9%) cases were confirmed as having IHR, 5 (62.5%) by ST and 3 (37.5%) by provocation test. Five from those confirmed as IHR (62.5%) had a positive BAT.

A systematic review of the literature was performed to answer the following question: What is the diagnostic value of skin testing for hypersensitivity reactions to contrast media?

P (Patient category) Patients with hypersensitivity reactions after radiological examinations with contrast media.

I (Intervention) Cutaneous tests: skin test, patch test (PT), intradermal test (IDT), skin prick test (SPT) or scratch test.

C (comparison) Clinical diagnosis of hypersensitivity reaction after contrast administration.

R (Reference) Drug provocation test.

O (outcome) Correctly confirmed diagnosis of hypersensitivity reaction to contrast media (sensitivity, specificity, area under the curve, positive predictive value, negative predictive value).

Relevant outcome measures

The working group considered sensitivity and specificity critical outcome measures for the decision-making process; and considered the area under the curve and the positive and negative predictive values important outcome measures.

Search and select (Methods)

On April 22^nd, 2021, a systematic search was conducted in the databases Embase (embase.com) and Medline (OVID) from 2017 onwards, using relevant key words for systematic reviews, RCT’s, observational studies and other study designs about

hypersensitivity reactions after contrast media. Specifically, the value of serum and/or urine tests, either skin tests or prophylactic measures were sought. The literature search yielded 400 unique references.