Immediate Reactions Following Iodinated Contrast Media Injection: a Study of 38 Cases

Home , Iobitridol, Iodinated contrast, Iodixanol, Iohexol, Iomeprol, Iopamidol

European Journal of Radiology 77 (2011) 495–501

Contents lists available at ScienceDirect

European Journal of Radiology

journal homepage: www.elsevier.com/locate/ejrad

Immediate reactions following iodinated contrast media injection: A study of 38 cases

Pascale Dewachter a,∗, Dominique Laroche b,1, Claudie Mouton-Faivre c,2, Evelyne Bloch-Morot d,3, Jean-Pierre Cercueil e,4, Liliane Metge f,5, Marie-France Carette g,6, Marie-Claude Vergnaud h,7, Olivier Clément i,8 a Service d’Anesthésie-Réanimation Chirurgicale & SAMU de Paris, Hôpital Necker-Enfants Malades, AP-HP, Université Paris-Descartes, 149 Rue de Sèvres, 75015 Paris, France b Service de Biophysique, Centre Hospitalier Universitaire, Avenue de la Côte de Nacre, Caen, France c Pôle d’Anesthésie-Réanimation Chirurgicale, Centre Hospitalier Universitaire-Hôpital Central, Avenue du Maréchal de Lattre de Tassigny, Nancy, France d Service de Médecine Interne, Hôpital Européen Georges Pompidou, 20 Rue Leblanc, Paris, France e Département d’Imagerie Medicale, Centre Hospitalier Universitaire, Hôpital du Bocage, 2, Boulevard du Maréchal de Lattre de Tassigny, Dijon, France f Département d’Imagerie Médicale, Centre Hospitalier Universitaire Caremeau, Place du Pr Robert Debré, Nîmes, France g Service de Radiologie, Hôpital Tenon, 4 Rue de la Chine, Paris, France h Service de Médecine Polyvalente, Centre Hospitalier Universitaire, Avenue de la Côte de Nacre, Caen, France i Service de Radiologie, Hôpital Européen Georges Pompidou, AP-HP, Université Paris Descartes, 20 Rue Leblanc, Paris, France article info abstract

Article history: Objectives: To investigate the pathomechanisms involved in cases of immediate hypersensitivity reactions Received 2 August 2009 occurring after the administration of iodinated contrast media. Accepted 17 September 2009 Materials and methods: Patients having presented clinical signs of immediate hypersensitivity suggesting allergy after iodinated contrast medium were investigated. Histamine and tryptase concentrations were Keywords: measured, and/or skin tests were performed. Patients with positive skin tests to the culprit contrast agent Anaphylaxis were classified as IgE-mediated allergic hypersensitivity (Group I) and patients with negative skin tests Contrast media as non-allergic hypersensitivity (Group II). Hypersensitivity Immediate Results: 38 patients were included. Most reactions appeared after non-ionic (n = 32). Reactions were more Skin tests frequently severe following ionic than non-ionic (p = 0.014). Skin testing was not performed in 11 patients. Histamine Skin tests with the culprit contrast agent were negative in 26% of the patients (Group II, n = 7) whereas Tryptase they were found positive with the contrast agent in 73% of the patients (Group I, n = 19). Latex-induced reaction was diagnosed in one patient, and was consequently excluded from the cohort. In Group I, the frequency of cross-reactivity with the other commercialized iodinated contrast media was low (7%). Cardiovascular signs were present in Group I (52.6%, n = 10), and absent in Group II (p = 0.023). Histamine and tryptase concentrations were higher in patients who had cardiovascular signs (p < 0.02). Conclusion: Immediate reactions with clinical signs suggesting allergy along with positive skin tests with the administered contrast agent confirm immediate allergic hypersensitivity (anaphylaxis) to this agent. Consequently, the culprit contrast agent should be definitely avoided as well as cross-reactive ICM in order to prevent further recurrences. © 2009 Elsevier Ireland Ltd. All rights reserved.

∗ Corresponding author. Tel.: +33 1 44 49 54 07; fax: +33 1 44 49 54 10. E-mail addresses: [email protected] (P. Dewachter), [email protected] (D. Laroche), [email protected] (C. Mouton-Faivre), [email protected] (E. Bloch-Morot), [email protected] (J.-P. Cercueil), [email protected] (L. Metge), [email protected] (M.-F. Carette), [email protected] (M.-C. Vergnaud), [email protected] (O. Clément). 1 Tel.: +33 2 31 06 47 03; fax: +33 2 31 06 51 60. 2 Tel.: +33 3 83 85 14 03; fax: +33 3 83 85 85 59. 3 Tel.: +33 1 56 09 27 92; fax: +33 1 56 09 38 45. 4 Tel.: +33 3 80 29 36 86; fax: +33 3 80 29 32 43. 5 Tel.: +33 4 66 68 33 10; fax: +33 4 66 68 38 12. 6 Tel.: +33 1 56 01 62 91; fax: +33 1 56 01 69 32. 7 Tel.: +33 2 31 06 46 79; fax: +33 2 31 06 51 92. 8 Tel.: +33 1 56 09 38 51, Fax: +33 1 56 09 38 50.

1. Introduction 2.4. Allergological assessment

Anaphylaxis remains one of the rare but significant events After informed consent was obtained, skin tests (prick-tests fol- specifically related to iodinated contrast media (ICM) that can lead lowed by intradermal tests) were performed with ICM at least 4 to morbidity and mortality, even in previously healthy patients. weeks after the occurrence of the immediate reaction. Prior to skin However, patients experiencing immediate reactions in the radi- testing, cutaneous reactivity was assessed with a positive (codeine ological area are usually not explored. Thus, in the absence of phosphate 50 ␮gmL−1) and a negative (saline solution) controls. systematic follow-up and investigation of these patients along with Prick-tests (PTs) were performed on the forearm and were con- underreporting of these reactions, the incidence of anaphylaxis sidered positive if, within 15 min of injecting the pure drug on following ICM remains currently unknown. However, immediate the forearm, a wheal equal to at least half the positive control reactions to ICM are a major concern for radiologists, even if their and larger than the negative control appeared. Intradermal tests overall frequency has probably diminished with the increasing (IDTs) were performed after PTs, on the patient’s back, by inject- use of non-ionic ICM [1]. Thus, a multidisciplinary network of ing 0.02–0.03 mL of the ICM solution at a diluted concentration radiologists, allergologists, biochemists and anesthesiologists was ranging from 1:1000 to the pure solution, raising a wheal of about created in order to collect clinical cases of documented immedi- 3 mm. The IDT was considered positive if an erythematous and pru- ate reactions following ICM. The aim of the present study was to riginous wheal, at least doubling the size of the injection wheal, identify the pathomechanisms involved during immediate reac- and surrounded by a flare, appeared within 15–20 min after the tions following ICM on the basis of definite clinical signs and injection. IDTs were performed with the culprit ICM and with allergological assessment. This evaluation included skin testing the other commercially available ICM. PTs were also performed and biochemical assessment in order to prove the diagnosis of with two standardized commercial natural rubber latex extracts anaphylaxis, identify the culprit agent and therefore to prevent (Stallergènes, Antony, France and Allerbio, Varennes-en-Argonne, recurrences. France) as recommended elsewhere [4].

2. Materials and methods 2.5. Statistical analysis

2.1. Patients Results were expressed as mean ± SD. The Chi-square test and Fisher’s exact test were used to compare qualitative data between Patients who experienced an immediate reaction occurring groups. The Mann–Whitney rank test was used to compare quan- less than one hour following injection of ICM had biochemical titative data between groups (Statview 5.0, SAS Cary NC, USA). measurements and/or skin tests. Patients experiencing immedi- Signiﬁcance was assumed when p < 0.05. ate reactions with clinical signs suggestive of allergy and positive skin tests with the culprit ICM entered Group I (IgE-mediated 3. Results allergic hypersensitivity, i.e. immunological mechanism strongly suspected) whereas those with negative skin tests entered Group II (non-allergic hypersensitivity, i.e. immunological mechanism 3.1. Total cohort characteristics excluded) [2]. Thirty-eight patients (23 female, 15 male) who had an immediate reaction following ICM injection were studied between January 2.2. Data collection 2001 and December 2003 (Tables 1–3). Mean age was 47.7 ± 17.6 years (range: 13–78 years); 44.0 ± 14.7 years for women and The following data were collected: age at the time of the reac- 53.5 ± 20.5 years for men (p > 0.05). The clinical severity of the reaction, gender, history of asthma, atopy, drug, food or latex allergy, tion was Grade I in 15 patients, Grade II in 13 patients, Grade premedication before the injection, name of the contrast agent III in 8 patients and Grade IV in 2 patients. Cutaneous-mucous injected, onset delay between the ICM injection and the occurrence signs were present in 35 patients (urticaria = 21; angioedema = 17; of the clinical reaction, previous exposures to ICM and previous erythema = 15), cardiovascular signs in 13 (hypotension = 3; car- reaction(s) to ICM. Clinical signs were classiﬁed according to the diovascular collapse = 8; cardiac arrest = 2), respiratory signs in Ring and Messmer four-step grading severity scale [3], adapted as 12 (cough = 7, dyspnea = 8, bronchospasm = 1) and gastrointestinal follows: Grade I involves cutaneous-mucous signs, Grade II corre- signs in 4 (nausea = 3; diarrhea = 1). The precise interval between sponds to moderate clinical signs associating cutaneous-mucous, ICM injection and onset of the reaction was recorded in 27 cases, cardiovascular or respiratory signs. The cardinal sign of Grade III and was 5 min or less in 19 cases (70.4%) and between 6 and reaction is cardiovascular collapse which may be associated with 20 min in 8 cases (29.6%). The reactions occurred after non-ionic cutaneous-mucous symptoms or bronchospasm and Grade IV is ICM injection in 32 cases (84.2%) and after ionic ICM injection in 6 cardiac arrest. Dyscomfort and subjective signs, such as a sensation cases (15.8%). The reactions were life-threatening in 6 patients who of heat, pain or dizziness were not considered. reacted after non-ionic ICM (18.7%) and in 4 patients who reacted after ionic ICM (66.7%) (p = 0.014). Eight patients were pretreated 2.3. Biochemical assessment and 19 were not (data missing for 11 patients). The frequency of life-threatening reactions was 37.5% (n = 3) in patients who had Blood samples were collected in EDTA and plain tubes as soon received pretreatment and 26.3% (n = 5) in patients who had not as the patient was clinically stable. Plasma or serum was obtained (p > 0.05). Nine patients had a positive previous allergological his- ◦ by centrifugation at 10–15 C, followed by gentle aspiration as far tory and 25 had not (data missing for 4 patients), the respective ◦ as possible from the white cell layer, and was frozen at −20 C frequencies of life-threatening reactions were 22.2% (n = 2) and 24% until use. Biochemical tests included duplicate measurements (n =6) (p > 0.05). The frequency of life-threatening reactions was of plasma histamine (Immunotech Radioimmunoassay, Beckman 31.8% (7 cases out of 22) in patients who previously received an Coulter, Luminy, France) and serum total tryptase concentrations ICM and 16.7% (2 cases out of 12) in patients who did not (p > 0.05) (UniCAP Fluoroimmunoassay, Pharmacia & Upjohn, Uppsala, Swe- (data missing for 4 patients). Among the 22 patients who previ- den). ously received ICM, 5 had a clinical history of previous reaction and P. Dewachter et al. / European Journal of Radiology 77 (2011) 495–501 497 , 1 1 − − gl ) ) 1 1 ␮ − − gl gl ␮ ␮ Tryptase ( Tryptase ( and <13 1 ) ) − 1 1 − − (nmol l 6 5.2 Histamine (nmol l 116ND18.13.9 9.7 ND 8.5 3.4 Respiratory signs Respiratory signs Histamine Cardiovascular signs Gastrointestinal signs Gastrointestinal signs normal values of plasma histamine and serum tryptase are <6 nmol l vity mechanism (normal values of plasma histamine and serum tryptase are <6 nmol l signs mucous signs GEAUU CVCGE, CVC A Hypotension Cough 4 ND 14.2 Cough, dyspnea 34.2 12.9 ND 5 5 Premedication Cutaneous-mucous , corticoids, U, corticoids A GE CVC Hypotension Dyspnea ND ND ND ND 1 1 1 1 1 1 1 Premedication Cutaneous- Previous reaction (grade) Previous reaction (grade) injection injection Culprit ICM Previous ICM Culprit ICM Previous ICM Time to onset (min) Time to onset (min) grade grade , respectively). 1 − gl ␮ 123 524 73 M5 54 M IV 32 F III 76 F III M MD III 2 III 1 1 1 Ioxaglate Yes Iohexol Iomeprol Iopamidol Yes Ioxaglate Yes Yes Yes Yes (IV) No No No No Yes (I) No MD MD Anti-H U, A U, A A Nausea Cardiac CVC arrest CVC CVC ND Bronchospasm ND 1305 ND ND 76 ND ND 89 73 32 M M II II 20 1 Iodixanol Iohexol No Yes No Anti-H No A Cough ND ND 67 56 49 F F III III MD 5 Ioxaglate Yes Ioxaglate Yes No Yes (I) Anti-H Anti-H 20 37 F II MD Iobitridol Yes No MD U Cough, dyspnea ND ND 10 51 F II MD Iohexol Yes Yes (II) MD GE Dyspnea ND ND Patient Age Sex Severity Patient Age Sex Severity 21222324 6325 6826 M 78 M 73 M 29 II F 45 II F I F I 2 I MD I MD MD 10 2 Iohexol Iohexol Iobitridol Iobitridol Yes Yes Yes Iohexol Yes Iobitridol No No No Yes (I) No No No No No MD U No A No U U GE, U, A GE, U, A Nausea dyspnea Cough 3.5 11.4 1.1 2.3 5.7 ND 2.7 1 5 ND 16.4 6.9 111213 31 43 F 32 F II F II II 5 5 5 Iomeprol Iopentol MD Iobitridol Yes No No MD MD Anti-H A GE Hypotension Dyspnea ND ND ND ND 141516 2517 53 F18 18 F19 32 II F 68 I F 66 I M I 1 F I 15 I 10 5 20 1 Iopamidol Iobitridol No Iobitridol Yes Iobitridol No Iobitridol No Yes Iopentol Yes No No No No Anti-H No No No Anti-H U, A GE, U, A GE, U GE, A Table 1 Clinical and biochemical characteristics of the patients of Group I (positive skin tests to the culprit ICM) suggesting allergic hypersensitivity ( GE: generalized erythema; U: urticaria; A: angioedema; CVC: cardiovascular collapse. MD: missing data, ND: not done. Table 2 Clinical and biochemical characteristics of the patientsand of <13 Group II (negative skin tests to the culprit ICM) suggesting non-allergic hypersensiti GE: generalized erythema; U: urticaria; A: angioedema; MD: missing data; ND: not done. respectively). 498 P. Dewachter et al. / European Journal of Radiology 77 (2011) 495–501

the current reaction was life-threatening in 3 of them. Among the ) 1

− 17 patients who had no previous history of reaction, 4 had a life- gl

␮ threatening reaction. The frequencies of life-threatening reactions Tryptase ( between previous reactors and others were not different (p > 0.05). )

1 Skin tests were performed in 27 patients, together with bio- − chemical tests in 14 (Tables 1 and 2), whereas 11 patients only had

72.71.5 12.2 1.66.1 4 2.83 2.5 5.2 2.9 6.3 2.8 4.3 5.3 biochemical investigations (Table 3). Histamine (nmol l

3.2. Biochemical tests

Blood samples were drawn after the reaction in 25 patients out dyspnea dyspnea Respiratory signs of 38. Plasma histamine concentrations were increased in 9 patients (range: 6–1305 nmol l−1, normal values < 6 nmol l−1), and tryptase concentrations in 4 patients (range: 16.4–76 ␮gl−1, normal values<13␮gl−1). Concentrations of histamine (358 ± 632 versus

, respectively). ± −1 ± 1 10.6 25.9 nmol l ; p = 0.019) and tryptase (33.9 27.3 versus −

gl ±

Cardiovascular signs 6.5 3.6; p = 0.004) were higher in patients with life-threatening ␮ reactions (Grades III and IV) than in patients with moderate reactions (Grades I and II). Similarly, histamine (290 ± 569 nmol l−1 −1 −1 and <13 versus 10 ± 27 nmol l ; p = 0.009) and tryptase (29.1 ± 27.1 ␮gl 1 − versus 6.5 ± 3.7 ␮gl−1; p = 0.016) concentrations were higher in

mol l patients who had presented cardiovascular signs than in patients Gastrointestinal signs who had not. No signiﬁcant difference was seen in patients with and without respiratory signs for histamine or tryptase concentrations.

3.3. Allergological assessment

One patient (Grade III reaction: cardiovascular collapse; data not

signs shown) had negative IDTs with ICM but positive PTs with latex. He was excluded from the cohort because the diagnosis was supported by a latex-induced allergic reaction. Latex allergy was ruled out by negative PTs with latex in the other 26 patients. These 26 patients were classiﬁed into two groups, according to the results of skin tests with the administered ICM. Group I com- Premedication Cutaneous-mucous prised 19 patients (73.1%) with positive skin tests. Skin tests with the injected ICM were positive in 56% of patients with Grade I reactions, in 70% with Grade II reactions and 100% with Grade III or IV reactions. The culprit ICM were: iobitridol (n = 5), ioxaglate (n = 4), iohexol (n = 3), iomeprol (n = 2), iopamidol (n = 2), iopentol (n =2) and iodixanol (n = 1). The 19 patients had positive intradermal tests Previous reaction (grade) to the culprit ICM, while only one patient (6%) had a positive PT. The positive IDTs were obtained with the ICM diluted at 1:100 in 3 cases (15.8%), 1:10 in 9 cases (47.4%) and with the pure ICM solution in 7 cases (36.8%) (Table 4). Skin cross-reactivity was assessed

injection in the 19 patients with one through 10 other ICM (Table 5). Nine out of the 129 IDTs (7%) performed appeared positive. Skin cross- reactivity to one or two ICM different from the culprit one was present in 6 patients whereas the 13 other patients had positive IDT with the culprit ICM alone. Group II consisted of 7 patients (26.9%) with negative PTs and Suspected ICM Previous ICM IDTs to the culprit ICM (Table 2). IDTs were negative with the pure solution in all of them. No cross-reactivity with the other ICM was found (59 IDTs performed with non-culprit ICM).

Time to onset (min) 3.4. Comparison of Group I and Group II (Tables 1 and 2)

3.4.1. Characteristics of the patients Group I (IgE-mediated allergic hypersensitivity) comprised 13

grade female and 6 male patients, and Group II (non-allergic hypersensitivity) comprised 4 female and 3 male patients. The gender distribution and the patients’ age (48.2 ± 17.8 years in Group I; 56.1 ± 19.1 years in Group II) were not different between the two groups. Non-ionic ICM were more frequently involved than ionic in both groups (78.9% in Group I and 100% in Group II; p > 0.05). The reactions occurred less than 5 min after the ICM injection in 30 13 M II 1 Ioxitalamate No MD GE Cough, Patient Age Sex27 Severity 2829 19 64 M 59 F IV F III II MD 5 MD Iomeprol Iobitridol Iodixanol No No Yes No No No No U GE Nausea Diarrhea CVC Cardiac arrest 99 26 3.6 6.8 39.6 6.7 3132 283334 F 4535 4536 II F 5237 M 35 I M 55 I F 43 I 5 F I M I ? I 5 7 10 Ioxitalamate MD 5 Yes Iobitridol Iohexol Iohexol Iobitridol MD Iobitridol No Yes Iohexol Yes No MD No No No MD MD Corticoid GE No MD No U U No U GE, U, A GE, U, A U, A Cough,

Table 3 Clinical and biochemical characteristics of the patients who did not undergo skin tests (normal values of plasma histamine and serum tryptase are <6 n GE: generalized erythema; U: urticaria; A: angioedema; CVC: cardiovascular collapse. MD: missing data. 75% and in 66.6% of cases in Group I and II, respectively (p > 0.05). P. Dewachter et al. / European Journal of Radiology 77 (2011) 495–501 499

Table 4 Skin tests results in Group I (Allergic hypersensitivity): prick-tests (PT) and intradermal tests (IDT) expressed as the diameters (mm) of the injection papule (IP) and of the ﬁnal papule (FP) (IP/FP) obtained with the culprit ICM at different dilutions, and cross-reactivity to other ICM (see Table 5).

Patient Culprit ICM PT IDT 10−3 IDT 10−2 IDT 10−1 IDT pure Cross-reactivity

1 Ioxaglate Positive Negative Negative 3/8 Negative 2 Iohexol Negative Negative Negative 4/11 Negative 3 Iomeprol Negative Negative Negative Negative 3/8 Negative 4 Iopamidol Negative Negative Negative Negative 3/8 Negative 5 Ioxaglate Negative Negative Negative 3/8 Ioxitalamate 6 Ioxaglate Negative Negative Negative 3/9 Ioxitalamate, iobitridol 7 Ioxaglate Negative Negative Negative 3/8 Negative 8 Iodixanol Not done Negative Negative 3/7 Negative 9 Iohexol Negative Negative Negative 3/9 Ioversol, iomeprol 10 Iohexol Negative Negative Negative Negative 3/8 Iopromide 11 Iomeprol Not done Negative Negative 3/8 Negative 12 Iopentol Negative Negative 3/13 Negative 13 Iobitridol Negative Negative Negative Negative 4/11 Negative 14 Iopamidol Negative Negative Negative Negative 3/12 Negative 15 Iobitridol Negative Negative 3/8 Ioxitalamate, iohexol 16 Iobitridol Negative Negative Negative Negative 2/9 Negative 17 Iobitridol Negative Negative Negative 4/12 Negative 18 Iobitridol Negative Negative Negative Negative 3/6 Iopromide 19 Iopentol Negative Negative 3/12 Negative

Life-threatening reactions were frequent in Group I (n = 7, 36.8%) ing for 1 patient) had previously received an ICM, 4 had a previous but absent in Group II (p > 0.05). Cutaneous or cutaneous-mucous immediate reaction (itching, urticaria, erythema with dyspnea, car- signs were present in 18 patients (94.7%) of Group I and in all the diovascular collapse in one case each). In Group II, 5 patients had patients of Group II (p = NS). Cardiovascular signs were present in previously received an ICM, in one case it was followed by a cuta- 10 patients (52.6%) of Group I and were not observed in Group II neous reaction (urticaria). (p = 0.023). Respiratory signs were observed in 36.8% of patients of Group I and 42.8% of patients of Group II (p > 0.05). Gastrointestinal 3.4.3. Pretreatment signs were seen in one patient of each group. In Group I, 7 patients were pretreated (H1-receptor antagonists Fourteen of the skin-tested patients had biochemical measure- in 7 cases associated to corticosteroids in 2 cases) (data missing ments. Plasma histamine concentrations were higher in Group I for 5 patients). Three life-threatening reactions and 4 moderate − − (210 ± 485 nmol l 1; n = 7) than in Group II (2.1 ± 1.1 nmol l 1; n =5) reactions were observed (Table 1). None of the patients in Group II (p = 0.0045). Serum tryptase concentrations did not differ between had been pretreated (data missing for 2 patients), only moderate − − Group I (17.8 ± 23.7 ␮gl 1; n = 9) and Group II (9.1 ± 4.8 ␮gl 1; reactions were seen (Table 2). n =5)(p > 0.05). 4. Discussion 3.4.2. Previous allergological and radiological histories In Group I, 6 patients (33.3%) had a previous positive allergolog- Immediate reactions following ICM can occur through multi- ical history (asthma, conjunctivitis, drug allergy in 2 cases each) ple mechanisms. However, the most severe reactions are probably (data missing for 1 patient). In Group II, 3 patients (42.8%) had directly related to allergic hypersensitivity, i.e. anaphylaxis [5–11]. a previous positive allergological history (rhinitis, asthma, drug In the present study, immediate reactions to ICM were classiﬁed allergy in 1 case each). Thirteen patients in Group I (data miss- on the basis of clinical ﬁndings and skin tests results. The clinical

Table 5 Cross-reactivity with the different ICM, diluted and pure, in the patients of Group I (Allergic hypersensitivity). Intradermal tests (IDT) are expressed as the diameters (mm) of the injection papule (IP) and of the ﬁnal papule (FP) (IP/FP) (see Table 4 for culprit ICM).

Patient Amidotrizoate Ioxaglate Ioxitalamate Iobitridol Iodixanol Iohexol Iomeprol Iopamidol Iopentol Iopromide Ioversol

1 ND Culprit Negative ND Negative Negative ND ND ND ND ND 2 Negative Negative Negative Negative Negative Culprit Negative Negative Negative Negative Negative 3 Negative Negative Negative Negative Negative Negative Culprit Negative Negative Negative Negative 4 ND Negative ND Negative ND Negative Negative Culprit Negative ND ND 5 ND Culprit 10−1 3/6 ND ND ND ND Negativea Negativea ND Negativea 6 ND Culprit Pure 2/5 Pure 2/5 Negative Negative ND Negative Negative ND Negative 7 Negative Culprit ND ND ND ND ND Negative ND Negative Negative 8 ND Negative Negative Negative Culprit Negative Negative Negative Negative Negative Negative 9 ND Negative Negative Negative Negative Culprit 10−1 3/15 Negative Negative Negative 10−1 3/15 10 ND Negative ND Negative Negative Culprit ND Negative Negative Pure 3/8 Negative 11 ND ND ND ND ND ND Culprit Negative ND ND ND 12 ND Negative Negative Negative Negative Negative ND Negative Culprit Negative Negative 13 Negative Negative Negative Culprit Negative Negative Negative Negative Negative Negative Negative 14 ND Negative Negative Negative Negative Negative Negative Culprit Negative Negative Negative 15 ND ND 10−2 5/12 Culprit ND 10−2 5/10 ND ND ND ND ND 16 Negative Negative Negative Culprit Negative Negative Negative Negative Negative Negative Negative 17 Negativea Negativea Negativea Culprit Negativea Negativea ND Negativea ND ND ND 18 Negative Negative Negative Culprit Negative Negative Negative Negative Negative Pure 3/6 Negative 19 ND Negative ND ND Negative Negative Negative ND Culprit ND ND

ND: not done. a Tests performed only up to 10−1 dilution. 500 P. Dewachter et al. / European Journal of Radiology 77 (2011) 495–501 eligibility criteria for this study were: (i) the occurrence of clinical trations were higher in patients with cardiovascular signs, but not signs belonging to the Ring and Messmer adapted scale [3,4], along in patients with respiratory signs. A raised tryptase concentration with (ii) an onset delay of less than one hour between the ICM indicates mast cell activation and is consistent with anaphylaxis injection and the occurrence of the clinical reaction. In most cases, [17–19]. In the present series, only 3 patients out of 9 with posi- the onset delay was less than 5 min, a time-frame consistent with tive skin tests had increased or borderline tryptase concentrations. immediate allergic hypersensitivity reactions. In the present study, These 3 patients had Grade III reactions, while the 6 others expe- non-ionic ICM were more frequently involved than ionic ICM, this rienced Grade I or II reactions. The mild severity of the reactions trend probably reflects the current market share. Skin testing with might explain the lower levels of tryptase [6,20]. The comparison immediate reading was used to classify the patients between aller- with the baseline level has been suggested to interpret such data gic and non-allergic hypersensitivity [12]. Intradermal tests were [17], but was not available in this study. Alternatively, inappro- performed with diluted ICM up to the pure solution. It is unusual priate delays between reaction and sampling could lead to false to use pure drug solutions for IDTs, and we anticipated that false- negative values, as samples obtained during the first 15 min may positive results could occur due to the high osmolality or toxicity of not contain tryptase yet, and as tryptase may have disappeared in the ICM commercial solution. However very few tests (9/188 = 4.8%) samples obtained hours after the reaction [17]. appeared positive when using pure non-culprit ICM, demonstrating Finally, our results confirm other reports [6,10,14,15] demon- that this hypothesis was not true. This finding confirms that skin strating that allergic hypersensitivity is a frequently involved mast cells do not release histamine in vivo through non-specific pathomechanism elicitating immediate reactions to ICM. Allergic mechanisms, as it was previously shown in vitro [13]. The positive hypersensitivity reactions usually recur with possibly increased IDTs with the pure ICM solution were thus considered as specific, severity on re-challenge, and despite pretreatment. Thus, in case of indicating an allergic mechanism. Thus, an ICM-induced allergic an ICM documented anaphylaxis, the culprit ICM and the skin cross- hypersensitivity reaction was considered proven in 73% of the reactive ICM should be avoided. Negative skin-tested ICM may be patients. On the contrary, prick-tests with ICM showed insufficient proposed for subsequent radiological procedures [21]. diagnostic sensitivity. For these patients, 93% of IDTs performed In conclusion, following an immediate reaction to ICM, the com- with the other commercialized ICM were negative, indicating a low bination of clinical, biochemical and skin test evidence will identify rate of cross-reactivity among the different commercialized ICM the culprit agent and will allow the practitioner to avoid recur- despite closely related molecular structures. Thus, we suggest that rences. cross-reactivity with other ICM should be assessed by using IDTs, in order to propose a safe alternative (i.e. non-reactive ICM by skin Conflict of interest testing) for further procedures. Two recent reports studied skin reactivity of ICM reactors by The authors disclose any actual or potential conflict of inter- using IDTs with 1:10 diluted solutions. The former found positive est including any financial, personal or other relationships with IDTs in 28% of the 32 patients [14]. The latter, a European mul- other people or organizations within 3 years of beginning the work ticentric study, found positive IDTs in 24.6% of the 122 patients submitted that could inappropriately influence their work. [15]. The percentages of positivity were lower than in the present study, even when considering the same dilutions, as we found pos- References itive IDTs with 1:100 or 1:10 dilution in at least 46% of the patients (n = 12/26). In the present study, the higher frequency of skin tests [1] Katayama H, Yamaguchi K, Kozucka T, et al. Adverse reactions to ionic and positivity may be explained by the precise knowledge of the name nonionic contrast media. A report from the Japanese Committee on the Safety of Contrast Media. Radiology 1990;175(3):621–8. of the culprit ICM for each patient, whereas in the other studies, it [2] Johansson SGO, Hourihane J, Bousquet J, et al. A revised nomenclature for was unknown for 60% and 48% of patients, respectively [14,15]. This allergy. An EAACI position statement from the EAACI nomenclature task force. emphasizes that a detailed description of all administered drugs Allergy 2001;56(9):813–24. and the chronology of the clinical event are fundamental in order [3] Ring J, Messmer K. Incidence and severity of anaphylactoid reactions to colloid volume substitutes. Lancet 1977;1(8009):466–9. to perform the appropriate skin tests. Radiologists should therefore [4] French Society of Anesthesiology and Intensive Care Medicine. Reducing the conform to the recommendation of indicating the precise name of risk of anaphylaxis during anaesthesia. Abbreviated text. Ann Fr Anesth Reanim the injected ICM on every radiological report. 2002;21(Suppl. 1):7–23. [5] Kanny G, Maria Y, Mentre B, et al. Case report: recurrent anaphylactic shock to Taken altogether, these previous studies and our results indicate radiographic contrast media. Evidence supporting an exceptional IgE-mediated that immediate allergic hypersensitivity to ICM must be con- reaction. Allergol Immunol 1993;25(10):425–30. sidered. In our series, allergic hypersensitivity to ICM appeared [6] Laroche D, Aimone-Gastin I, Dubois F, et al. Mechanisms of severe immediate reactions to iodinated contrast material. Radiology 1998;209(1):183–90. three times more frequent than non-allergic hypersensitivity. Res- [7] Mita H, Tadokoro K, Akiyama K. Detection of IgE antibody to a radiocontrast piratory and/or cutaneous signs did not discriminate between medium. Allergy 1998;53(12):1133–40. allergic and non-allergic mechanisms. Conversely, cardiovascu- [8] Brockow K, Vieluf D, Püschel K. Increased post-mortem serum mast cell tryptase in a fatal anaphylactoid reaction to nonionic radiocontrast medium. J lar signs and/or life-threatening reactions were only seen during Allergy Clin Immunol 1999;104(1):237–8. allergic hypersensitivity reactions, confirming previous reports [9] Alvarez-Fernandez JA, Valero AM, Pulido Z, et al. Hypersensitivity reaction to [5,6,8–11,16]. Surprisingly, about 30% of patients who had an aller- ioversol. Allergy 2000;55(6):581–2. [10] Dewachter P, Mouton-Faivre C, Felden F. Allergy and contrast media. Allergy gic reaction had not been exposed to ICM previously, suggesting 2001;56(3):250–1. that they were sensitized by other means and therefore indicating [11] Valfrey J, Newinger G, Arbogast R, et al. Choc à l’ioxaglate (Hexabrix 320) that anaphylaxis may follow a first-time ICM administration. lors de coronarographies. A propos de deux cas. Rev Fr Allergol Immunol Clin Several anaphylactic reactions occurred despite pretreatment, 2002;42(2):157–62. [12] Mirakian R, Ewan PW, Durham SR, et al. BSACI guidelines for the management as reported for other cases [5,10,11] and their severity was inde- of drug allergy. Clin Exp Allergy 2009;39(1):43–61. pendent of pretreatment. [13] Stellato C, De Crescenzo G, Patella V, et al. Human basophil/mast cell releasabil- Histamine and tryptase concentrations were higher during ity. Heterogeneity of the effects of contrast media on mediator release. J Allergy Clin Immunol 1996;97(3):838–50. life-threatening reactions, i.e. anaphylaxis than during moderate [14] Kvedariene V, Martins P, Rouanet L, et al. Diagnosis of iodinated con- reactions therefore confirming a previous study [6] and case- trast media hypersensitivity: results of a 6-year period. Clin Exp Allergy reports [10]. Histamine concentrations were higher in patients with 2006;36(8):1072–7. [15] Brockow K, Romano A, Aberer W, et al. Skin testing in patients with hypersen- positive skin tests, suggesting that histamine is more likely released sitivity reactions to iodinated contrast media: A European multicenter study. during allergic than during non-allergic reactions. Tryptase concen- Allergy 2009;64(2):234–41. P. Dewachter et al. / European Journal of Radiology 77 (2011) 495–501 501

[16] Dewachter P, Tréchot P, Mouton-Faivre C. Allergie à l’iode: le point sur la [19] Harper NJ, Dixon T, Dugue P, et al. Suspected anaphylactic reactions associated question. Revue de la littérature. Ann Fr Anesth Reanim 2005;24(5):40– with anaesthesia. Anaesthesia 2009;64(2):199–211. 52. [20] Laroche D. Immediate reactions to contrast media: mediator release and value [17] Ebo DG, Fisher MM, Hagendorens MM, et al. Anaphylaxis during anaesthesia: of diagnostic testing. Toxicology 2005;209(2):193–4. diagnostic approach. Allergy 2007;62(5):471–87. [21] Dewachter P, Mouton-Faivre C, Laroche D, et al. Allergie immédiate aux [18] Dewachter P, Mouton-Faivre C. What investigation after an anaphylactic reac- produits de contraste iodés et prévention des réactions. Rev Med Interne tion during anaesthesia? Curr Opin Anaesthesiol 2008;21(3):363–8. 2009;30:872–81.