Bibliografisk note

Funding Information:
This task force has been approved and received grants from EAACI for its organization: EAACI budget code 40703. Annick Barbaud and Lene Heise Garvey (LHG) chaired the EAACI Drug Provocation Test Working Group. Annick Barbaud and Lene Heise Garvey led the discussions and drafted the general considerations section. Each author reviewed the literature, contributed to the general table A (supplement) and wrote his or her own section. Jose Julio Laguna, Anca Mirela Chiriac, Annick Barbaud, Lene Heise Garvey for beta-lactams, Maria Torres and Jean Christoph Caubet for NSAIDs, Knut Brockow for contrast agents, Kathrin Scherer Hofmeier and Josefina Cernadas for anti-infectives, Alessandra Arcolaci and Patrizia Bonadonna for PPIs and antihistamines 2, Lene Heise Garvey for perioperative anaphylaxis, Annick Barbaud for corticosteroids and anticonvulsants. All contributors took part in discussions and corrections during face-to-face and online meetings. All contributors approved the final version. Rapid drug desensitization (RDD) is presently the main therapeutic approach to HSRs to both monoclonal antibodies (mAbs) and chemotherapeutics. However, emerging data suggests DPT might prevent a significant number of patients from undergoing unnecessary RDD.101–103 For chemotherapeutics, studies report negative DPT in 30%–56% of patients who could receive treatment without RDD.101,104 Patients with paclitaxel HSR tolerated rechallenge more often than those with platin HSRs.105 When multiple drugs are administered simultaneously, they should all be considered and DPT might prevent misdiagnosis. Leucovorin was the culprit in up to 11% of “oxaliplatin-reactive” patients.106 Limited data on DPT with mAbs have been published. In patients with unequivocal history of a DHR to mAbs, 30% showed negative DPT and could avoid RDD.101 DPT protocols exist for rituximab, anti-TNF agents, trastuzumab and omalizumab. A diagnostic algorithm for DHRs to biologicals has been proposed.103 For both chemotherapeutics and biologicals, risk management strategies and contraindications follow general recommendations (Table 2). A risk stratification approach to rechallenge patients has been developed by Picard et al.104 (Table S13). DPT should only be performed if there is therapeutic indication for the drug. It is considered a high-risk procedure.101 The patient's next scheduled treatment should be used for DPT, avoiding delays or overdose. The standard approach involves administering the infusion under normal conditions. Intensified premedication is not recommended.102,103 The TF recommends following previous guidelines proposed by EAACI.102,103 For chemotherapy, the TF recommends performing DPT in moderate-low severity IR according to Picard et al.104 (Table S13) with negative skin tests or negative BAT and normal serum tryptase. A RDD is recommended in high-risk patients with severe reactions, positive skin tests, or cardiovascular or respiratory comorbidities. For biologicals, the TF suggests DPT in expert drug hypersensitivity centers, if the biological cannot be substituted with other skin test- and/or BAT-negative agents. In practice, the heterogeneity of DPT protocols between centers renders data comparison difficult and a standardized procedure has not yet been agreed. The approach to allergy to beta-lactam antibiotics has evolved compared to the more restrictive approach recommended previously.7 Many examples of risk stratification and algorithms have been published recently.7,33–42 Whilst focus remains on patient safety, individual risk stratification means, that in patients considered to be at low risk of allergy a diagnosis can be achieved with fewer resources. However, in the literature three main areas still lack consensus in the adult population: What defines a low-risk patient? This has been addressed in this position paper in Figure 1, Table 3 and Table S2. The place of DPT without preceding ST in adults. The need for investigations before DPT in patients with NIRs and a low-risk profile is the subject of much debate (Table 5). Regarding children, there is strong evidence based on multiple large studies to provide a firm recommendation for direct oral DPT without skin testing first in low-risk patients.17,37,43–45 Geographical variations in epidemiology of reported drug allergy Inclusion criteria and recruitment Definition of “low-risk” is clinical and does not always foresee a positive allergy work-up (see “general considerations” paragraph). A proportion of patients with positive skin test result have unknown/poorly defined reactions. According to an individual risk–benefit analysis, ST could be favored in patients with severe comorbidities even if their history is that of a low-risk reaction Note: See Tables S1 and S2 and general considerations paragraph. Different inclusion criteria and recruitment bias the reporting and interpretation of prevalence of drug allergies across the world. In studies from specialized drug allergy centers, prevalence and severity of reactions is generally higher. However, when a delabeling approach is favored, with recruitment in general departments, prevalence and severity are lower. In adults, there is increasing evidence that such an approach could be safe in targeted populations.15,16,46–48 An overview of recent studies on direct DPT without previous skin testing in non-severe NIR in adults is summarized in Table S3. From these studies, among 5948 adult patients who had penicillin DPT without ST, 289 reactions (4.9%) occurred with only 2 cases of anaphylaxis (0.03%). In adults with NIR where a drug allergy expert has performed a risk evaluation and considers the risk of a reaction on DPT to be low- as defined in Table 3 and Figure 1- the TF suggests that there is enough evidence to consider DPT, without skin testing first. The value of prolonged DPT Prolonged DPT exceeding 1 day for NIR diagnosis is also the subject of much debate.39,46,49–52 There are many different protocols for performing prolonged DPT and there is no consensus on a preferred procedure. Details of all analyzed studies and their references are reported in Table S4. From 23 articles, including 6484 patients with prolonged DPT, reactions occurred in 495/6484 (7.6%); during the initial DPT in 147/6484 (2.3%) and during the prolonged DPT in 347/ 6337 (5.5%). As there is only one study with a washout period50 presently, it is difficult to conclude if prolonged DPT increases sensitivity. Some studies suggest that the patient would be more likely to take the drug in the future after a prolonged DPT,16 but others find a higher rate of refusal of further intake in the prolonged DPT group.53 Arguments for one-day versus prolonged drug provocation test are summarized in Table S5. Based on the literature, at the present time the TF cannot make a recommendation for or against performing prolonged DPT. The TF suggests that if prolonged provocation is performed in NIR: at least the maximum single therapeutic/unit dose should be reached; a washout period should be introduced between doses, to allow the initial dose to potentially elicit a reaction. There is some evidence that the initial dose can elicit reactions up to 48 h–7 days later,42,52 and ideally a minimum 48 h wash out period should be respected. However, to ensure patient compliance and/or assess risk of NIR, some groups perform prolonged DPT for 2–4 days16,43 with a wash out period between initial dose and the following morning, that is, 16–18 h; DPT approaching the duration of a full treatment (7–10 days) is not recommended. In the case of antimicrobial allergy, the risk of developing antimicrobial resistance with prolonged DPT must be weighed against the benefits of delabeling inaccurate drug allergy labels. In patients with severe IR to BL more than 6 months before investigation, who display negative results on IgE, skin tests and DPT, a risk of resensitization should be considered.7 To identify resensitization in remote IRs, few highly specialized centers perform prolonged DPTs on this indication, on an experimental basis, and advice re-testing with ST or IgE tests 4–6 weeks later.54 HRs to macrolides are rare. Skin tests are often irritant and badly standardized. In most cases DPT is needed to confirm or rule out hypersensitivity and to investigate cross-reactivity.55,56 Although older data state that skin tests with fluoroquinolones (FQ) can be irritative10 and that up to 50% of case may show cross-reactivity,57,58 a new method for IDT reading has been recently suggested.59 The authors studied 163 allergic sequential patients with an history of HS to FQ. Intradermal tests were performed according to ENDA protocol12 using histamine as a positive control at a concentration of 0.1 mg/mL. For IDT to ciprofloxacin, levofloxacin, and moxifloxacin concentrations of 0.025 and 0.005 mg/mL were used. Authors considered negative tests those with non-specific wheal without flare which frequently occurs to multiple skin tested patients to FQ. 159 patients were negative using this criteria and 82/82 of them underwent single dose DPT to the index FQ or other FQ (levofloxacin 250 mg, ciprofloxacin 250 mg, moxifloxacin 200 mg).59 Negative oral DPT resulted in the removal of FQ allergy or revision to confirm tolerance of an alternative FQ, as it is known that FQ may cross-react, Protocols vary in terms of doses and steps used, but most achieve full therapeutic dose in 1 or 2 days.58 Krantz et al.60 challenged 204 patients with low-risk IR and NIR, without allergy work-up, in a one- or two-step single dose procedure with tolerance of the drug in 94%. Literature is based on small case series. Skin tests show low sensitivity and DPT is needed to confirm the diagnosis.61 Infrequent use of tetracyclines results in very low rates of hypersensitivity primarily with FDE and photosensitivity, but also cases of DRESS.62 Very few graded DPT have been reported.62 IR are rare, and NIR occur from <1% to 10.5%, including SCARs. There are only few reports with graded oral DPT.63 Vancomycin and teicoplanin can induce anaphylaxis, MPE, vasculitis and SCARs. Non-allergic reactions to vancomycin (red man syndrome, Vancomycin flushing syndrome) occurs in 5%–14% of patients usually due to rapid infusion. Symptoms are itching and rash typically of face, neck and upper body caused by direct release of histamine from mast cells and basophils. It can typically be overcome by slowing infusion rate and premedication with antihistamine., Cross-reactivity between vancomycin and teicoplanin may occur.64 DPT is rarely needed.65 One study reported graded intravenous DPT with gentamicin in skin test negative patients with IR.66 Data on HSs to anti-TB drugs are limited, but one study reported reactions in 3%–4% of patients.67 HSs are usually benign MPEs, occurring few weeks after starting therapy. Some patients develop HS to multiple anti-TB drugs. A pragmatic approach to standardize the management is needed. When a HS occurs, all anti-TB drugs should be withdrawn, and skin testing should be performed.68 In NIR, the TF suggest that when skin tests are negative, all necessary anti-TB drugs may be sequentially re-introduced (rechallenge or graded escalation procedure) at an interval of 3–4 days, at target therapeutic doses, allowing time to detect a severe NIR with minimal risk of drug resistance. Details on skin tests and procedures with anti-TB drugs are given in Table S6. NSAID hypersensitivity can be allergic with selective responders (SR) or non-allergic with cross-reactive patients (CR). DPT is used to confirm sensitization, study cross-reactivities or find alternatives. Different protocols have been suggested.23,69–76 In reviewed original studies (17 exclusively in children), 2374 adults, 3363 children and 100 not specified cases were included (Tables S7–S9). Most subjects (76.9%) were CR. In all patients, 19.7% of DPTs were positive, with a higher percentage in subjects confirmed as CR (23.3%) than in those confirmed as SR (8.6%). Skin tests were rarely performed and were only carried out with culprit NSAID before DPT in 93 cases. In most studies, DPT was performed with an alternative NSAID (95%). Only 55.4% were tested with the culprit, and 50% were tested with culprit and alternative. The culprit drug was tested more frequently in SR (96.3%) than in CR (43.2%) and more often in children. The DPT mainly tested positive for pyrazolones (in 65.7% subjects) and diclofenac (in 42.1%). The results of DPT for other NSAIDs are shown in Table S7. In CR cases, skin tests are not helpful, and DPT with aspirin (ASA) or other NSAIDs are recommended to determine tolerance. In CR cases, cofactors such as preexisting CSU, asthma, mastocytosis can negatively influence the intensity of adverse effects and must be considered before DPT. In preexisting CSU, DPT with COX-2 inhibitors or preferential COX-2 inhibitors (nimesulide, meloxicam) can be done under concurrent antihistamine treatment. DPT with COX-1 inhibitors might be done after complete remission of CSU. In NSAID-exacerbated respiratory disease there is a risk of severe bronchospasm and graded DPT must be conducted cautiously. For CR patients, the sequence of NSAIDs given for DPT vary greatly in the literature. The test plan should consider the probability of inducing a reaction, which is lowest for selective COX-2 inhibitors (“coxibs”), and lower for partial COX-1 inhibitors (e.g., paracetamol)77 than for full COX-1 inhibitors (e.g., ibuprofen, ASA). The DPT dose increments are shown in Table S9. In 104 cases, the NSAID was administered as a single dose. The drug was administered at 90 min-interval in most cases (1117; 53.2%), followed by 60 min-intervals (546; 26%) and 120–180 min-interval (375; 17.8%). Although they may reduce the intensity of adverse effects of a DPT with NSAIDs/paracetamol, in well-controlled CSU patients, the TF suggests maintaining treatment with antihistamines or anti-IgE, as well as maintaining treatment with leukotriene agonists in patients with well-controlled asthma. A decision algorithm for NSAID-induced urticaria/angioedema is provided in Figure 2. DHR to proton pump inhibitors (PPIs) and Histamine-2-receptor antagonist (H2RA) have increased in the last decade. Most reactions to PPIs are IR78 and skin test specificity is high, but sensitivity is low (22.6%–58.8%) thus DPT is necessary79 when skin tests are negative. Single-blind, placebo-controlled DPT has been performed in skin test-negative patients, with the culprit in low and intermediate risk cases and with an alternative after severe reactions.78–81 Titration and dosing were similar in all studies reaching full single dose in 1 day (Table S10).24,78–80 Time interval between the incremental doses ranged from 30 to 60 min.24,79–80 Cross-reactivity is reported among different PPI belonging to the same subgroup (Table S10), but patterns are variable and inconsistent.24 Rabeprazole is in the “lansoprazole” group as esomeprazole and pantoprazole are in the “omeprazole group”. Most PPIs are slow release and dose titration is challenging due to the formulation and potential loss of drug function. Due to the slow release, HRs may be delayed. Consequently, the TF suggests prolonged observation of ≥3 h after the last step of an oral DPT with PPI. IV DPT might be an option, but there is very limited experience with this approach. PPI also induce NIR ranging from mild MPE to SCARS, but data on IDT with delayed reading and patch testing are limited. Cross-reactions to all PPI have been reported in DRESS.81,82 For H2RA, DPT has been performed with 2–6 steps with 30-min intervals until the single therapeutic dose was reached in 1 day (Table S11).83 Cross-reactivity between ranitidine and nizatidine has been found.84 Data about non-immediate reactions are scarce. The task force recommends to perform DPT with the culprit PPI in patients with IR and negative skin test to the culprit; perform DPT with alternative skin test-negative PPIs in patients with positive skin test to the culprit and in patients with a high-risk of severe anaphylaxis; extend the observation time for oral DPT due to unpredictable absorption with the delay of the initial reaction. A negative skin test to iodinated (radio-)contrast media (ICM) cannot exclude hypersensitivity in all patients with ICM reaction. However, negative skin tests are associated with less severe non-allergic reactions and positive skin tests with more severe (and dangerous) allergic reactions.85–99 DPT have been performed by allergists particularly after more severe reactions, which are often associated with positive skin tests, whereas ICM re-exposure conducted by a radiologist at the next needed x-ray/CT scan has been the traditional option for non-severe reactions (e.g., isolated urticaria or non-severe maculopapular exanthema) to confirm tolerability.85 During the last decade DPT protocols have also been established85–99 (Table S1). Protocols for DPT with ICM are neither standardized nor validated; there is no consensus regarding dose and dose increments.85 Series including up to 161 adult patients reported various modalities and doses for DPT. Maximum single doses ranged from 10 to 120 mL, dose increments from 1 to 7 steps, duration from 1 to 2 days. The second day in NIR was separated by 1–2 weeks in two centers89,95 (Table S12). Intervals between doses were 30–120 min for IR and 60–1440 min for NIR. All studies performed DPT with skin test negative ICM and most used an alternative ICM, while some challenged the culprit ICM when negative in skin tests. One study, most likely using re-exposure in combination with premedication, re-exposed skin test positive non-culprit ICM and confirmed a high positive predictive value.99 There is no evidence to recommend optimal doses for ICM DPT. The TF recommends that both ICM re-exposure with an alternative skin test negative ICM by a radiologist, or DPT in an allergy department with a skin test negative alternative or culprit ICM, are both valid options depending on the risk involved for patients needing further ICM examinations. Patients with severe anaphylaxis should receive DPT in an allergy department, whereas for those with non-severe reactions, such as urticaria or MPE and negative skin testing an alternative ICM can be re-exposed in the radiology department according to the recommendations issued by the allergist. ICM can cause acute kidney damage. The TF recommends excluding contraindications before DPT. Higher risk is seen in patients with reduced renal function; therefore, low-osmolality/iso-osmolar ICMs should be used, and oral serum bicarbonate, oral N-acetylcysteine, and/or intravenous saline solution (0.9%) may be given as prophylaxis90 also before DPT.97 For gadolinium-based contrast media, there is only one low-dose one-step study performing DPT with 1 mL skin test-negative gadoteric acid in 14 patients.100 The risk of inducing other adverse effects like nephrogenic fibrosis must be balanced against the need for DPT. Evidence is insufficient for recommending a specific protocol. Rapid drug desensitization (RDD) is presently the main therapeutic approach to HSRs to both monoclonal antibodies (mAbs) and chemotherapeutics. However, emerging data suggests DPT might prevent a significant number of patients from undergoing unnecessary RDD.101–103 For chemotherapeutics, studies report negative DPT in 30%–56% of patients who could receive treatment without RDD.101,104 Patients with paclitaxel HSR tolerated rechallenge more often than those with platin HSRs.105 When multiple drugs are administered simultaneously, they should all be considered and DPT might prevent misdiagnosis. Leucovorin was the culprit in up to 11% of “oxaliplatin-reactive” patients.106 Limited data on DPT with mAbs have been published. In patients with unequivocal history of a DHR to mAbs, 30% showed negative DPT and could avoid RDD.101 DPT protocols exist for rituximab, anti-TNF agents, trastuzumab and omalizumab. A diagnostic algorithm for DHRs to biologicals has been proposed.103 For both chemotherapeutics and biologicals, risk management strategies and contraindications follow general recommendations (Table 2). A risk stratification approach to rechallenge patients has been developed by Picard et al.104 (Table S13). DPT should only be performed if there is therapeutic indication for the drug. It is considered a high-risk procedure.101 The patient's next scheduled treatment should be used for DPT, avoiding delays or overdose. The standard approach involves administering the infusion under normal conditions. Intensified premedication is not recommended.102,103 The TF recommends following previous guidelines proposed by EAACI.102,103 For chemotherapy, the TF recommends performing DPT in moderate-low severity IR according to Picard et al.104 (Table S13) with negative skin tests or negative BAT and normal serum tryptase. A RDD is recommended in high-risk patients with severe reactions, positive skin tests, or cardiovascular or respiratory comorbidities. For biologicals, the TF suggests DPT in expert drug hypersensitivity centers, if the biological cannot be substituted with other skin test- and/or BAT-negative agents. In practice, the heterogeneity of DPT protocols between centers renders data comparison difficult and a standardized procedure has not yet been agreed. DPT with potent anesthetic drugs, for example, hypnotics and neuromuscular blocking agents should only be performed in highly specialized centers in close collaboration between anesthetists and allergists.20 Readers with a specific interest in DPT with drugs used exclusively in the perioperative setting are referred to literature specifically addressing this.22,107 Opioids rarely cause IgE-mediated allergic reactions but may induce direct mast cell histamine release via the opioid receptor or the MRGPRX2 receptor.108,109 This is common with natural opioids, for example, morphine and codeine, and rare with synthetic opioids, for example, tramadol or fentanyl.108 Opioids elicit gastrointestinal adverse effects, itch or rash often misinterpreted as allergy. A risk stratification algorithm for investigation of suspected opioid allergy has been suggested.109 Skin testing using non-irritant concentrations20 is useful for synthetic opioids, but shows limited usefulness for natural opioids.110 Titrated DPT with opioids has been shown to be safe in experienced hands.108 True allergic IR to local anesthetics (LA) are extremely rare. Vasovagal reactions, hyperventilation/anxiety attacks or toxicity symptoms may mimic hypersensitivity, but skin symptoms are typically missing.111,112 Skin testing is recommended before DPT when hypersensitivity is likely, or when patients need reassurance. Due to a high level of anxiety a placebo-controlled DPT should always be considered.112 Cross-reactivity has been reported in both IR and NIR to LA and a skin test and DPT-negative alternative LA should always be identified,111 ideally for both dental treatment and regional anesthesia. LA provocation with vasoconstrictor to reproduce symptoms such as tachycardia, may reassure patients that such symptoms are benign.112 Most common dyes are patent blue and methylene blue used for sentinel node procedures in cancer surgery. Skin testing is gold standard and there is currently no provocation model. Provocation tests with corticosteroids (CS) should be performed at least 1 week after skin tests, as patch tests and IDT can show delayed positivity.113,114 In case of IR to depot corticosteroid preparations, reactions to excipients, most commonly polyethylene glycol or carboxymethylcellulose, must be ruled out.115 Anaphylaxis to methylprednisolone may be restricted to this corticosteroid, as cross-reactivity is unpredictable, a skin test- and DPT-negative alternative CS must be found in both IR and NIR.114 The route of administration depends on patient need, and DPT with injectable steroids is usually performed intramuscularly or intravenously. Graded administration to reach the full dose or up to 1 mg/kg/day has been suggested116 (Table S1). Anticonvulsants are classified as aromatic and non-aromatic (Table S14).117 They mainly induce NIR, with a high frequency of SCARs. In NIR, cross-reactivity between anticonvulsants is frequently reported between aromatics but cross-reactivity between aromatic anticonvulsants and lamotrigine, remains controversial.118,119 Most studies were based on cases of DRESS, known for its high risk of multiple sensitizations without relation to chemical structures. Finally, some severe reactions could be explained by accumulation of toxic metabolites. Non-aromatic anticonvulsants such as benzodiazepines, gabapentin, levetiracetam and sodium valproate117 are frequently proposed as alternatives in case of NIR to aromatic anticonvulsants/lamotrigine. The TF suggests following DPT protocol with slowly increasing doses, as evaluated in the highest number of cases.120 This protocol determines the dose per body weight, with incremental dosages week after week (Table S1). DPT can never exhibit 100% sensitivity or specificity; however, this is not required for making recommendations. Cofactors such as virus infection/reactivation, physical and psychological stress and simultaneous intake of other drugs are not reproduced during a DPT. It is difficult to standardize the method of DPT for drugs with different mechanisms and patterns of hypersensitivity. The dose used during provocation may also be insufficient to elicit a response in some cases leading to a risk of false negative testing.

Publisher Copyright:
© 2023 The Authors. Allergy published by European Academy of Allergy and Clinical Immunology and John Wiley & Sons Ltd.