Evaluation of drug patch tests in children

Abstract

Background:

Patch tests are used to diagnose nonimmediate T-cell–mediated drug hypersensitivity reactions. The aim of this study was to evaluate the results of patch tests performed with suspect drugs in children.

Methods:

Patients < 18 years of age who had a drug patch test at the pediatric allergy outpatient clinic of our hospital between January 2014 and January 2020 were included in the study. Age, sex, culprit drug(s), reaction characteristics, and patch test results were recorded from the patients' files.

Results:

A total of 105 drug patch tests were performed on 71 patients during the study period. The patients' median age was 7 years (interquartile range, 4–11 years), and 57.7% (n = 41) were boys. Twenty-three patients (32.3%) had severe cutaneous adverse reaction (Stevens-Johnson syndrome in 11, drug reaction with eosinophilia and systemic symptoms in 9, and acute generalized exanthematous pustulosis in 3 patients), 45 (63.3%) had maculopapular rashes, and 3 (4.2%) had fixed drug eruption. A total of 20 patch test results (28%) were positive: 18 of 44 patch tests (40.9%) with antiepileptic drugs and 2 of 48 patch tests (4.1%) with antibiotics. Positive results were obtained in 23% of the patch tests (6/26) in 20 patients with severe cutaneous adverse reactions and in 17.7% of the patch tests (14/79) in 51 patients with mild cutaneous reactions. No adverse reactions occurred during or after the patch tests.

Conclusion:

In our study, patch test positivity was more common with antiepileptic drugs and in patients with severe cutaneous drug reaction.

Drug hypersensitivity reactions can be difficult to deal with for both patients and physicians. There are still no standardized diagnostic tests for delayed and severe cutaneous adverse reactions (SCAR), especially in children.¹ Patch tests (PT) are the preferred and first ordered test in patients with delayed mild cutaneous adverse drug reactions (CADR) and SCARs due to the lower probability of reactivation.² Also, PTs have high specificity in drug hypersensitivity reactions.³ Clinical conditions with the highest PT positivity are maculopapular exanthema (MPE), contact dermatitis, acute generalized exanthematous pustulosis (AGEP), fixed drug eruption (FDE), and drug reaction with eosinophilia and systemic symptoms (DRESS), whereas lower rates of positivity are seen in bullous reactions such as Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN).^4,5

Apart from the clinical presentation, the culprit drug also affects PT positivity. When a single drug is suspected, there may be 40–100% positivity,^6,7 whereas the positivity rate may decrease to 0.4–40% as the number of suspect drugs increases.^5,8 Furthermore, PT positivity is more common with β-lactam antibiotics, carbamazepine (CMZ), other antiepileptic drugs, and heparin derivatives,⁹ and can also be influenced by the method of PT application, such as the drug form, location, concentration, and the properties of the vehicle used.⁹ Although there are several published guidelines with regard to drug PT protocols and the concentrations to be applied^2,4,10,11; standardized data for children are still not available. Although we have some insight about drug PTs in adults, there are insufficient data about the results of PTs among children with suspected drug hypersensitivity. Therefore, in the present study, we evaluated PT results for delayed drug reactions in children.

METHODS

Study Population

Patients ages < 18 years who underwent a drug PT for delayed CADR or SCAR in the pediatric allergy outpatient clinic of our hospital in the 6-year period between January 2014 and January 2020 were included in the study. The study was approved by the ethics committee of the Ankara City Hospital Children's Hospital (Decision E1/410/2020).

Study Procedures

Patients' data were analyzed retrospectively. Questions in the standard European Academy of Allergy and Clinical Immunology/European Network of Drug Allergy history questionnaire were reviewed from patient files.¹² Demographic characteristics of patients (age, sex, diseases, atopic status, presence of concomitant allergic disease), drug(s) responsible for the reaction, reaction features (times, types, and characteristics), PT features (drugs used, concentrations, and results), other tests for drug allergy (if any), and skin test results for atopy (if any) were recorded.

The clinical histories of the patients who underwent PT were examined from their records. CADRs were classified as MPE and FDE; SCARs were classified as DRESS, AGEP, and SJS/TEN. Patients with mucocutaneous rash, fever, lymphadenopathy, internal organ involvement, and hematologic abnormalities (e.g., eosinophilia and/or atypical lymphocytes [RegiSCAR score > 3]) were considered DRESS.¹³ Criteria for SJS were acute onset of mucous membrane involvement (at least two mucosal surfaces) and skin symptoms (maculae, target-like, bullae or erosion, and positive Nikolsky sign, together with epidermal detachment of <10% of the total body surface area). Patients with similar clinical signs but > 30% epidermal detachment were classified as TEN.¹⁴ AGEP was defined as nonfollicular, pustular lesions of <5 mm on erythematous skin that occurs within a few days of the first dose of the drug (EuroSCAR score > 4).¹⁵ Drug exposure, followed by acute onset of one or more oval erythematous plaque lesions that heal with hyperpigmentation was considered FDE.¹⁰

Drug PT

PT with the suspect drugs were performed at least 6 weeks after the drug reaction. In patients who present with DRESS and SJS/TEN, the tests were performed at least 6 months later. Informed consent was obtained from the patients' family for testing. During PT application, we did't perform PT if patient had symptoms such as infection, fever and eczema. PTs were prepared and performed by our allergy clinic according to the general European Academy of Allergy and Clinical Immunology/European Network of Drug Allergy guidelines.⁸ Powder vial forms of the suspect drugs were preferred, but tablet forms were used if the powder vial form was not available. For each drug, the recommended concentration was determined from the guideline based on the characteristics of the reaction.^8,16

Powdered drugs were weighed with sensitive scales and were mixed with white petroleum jelly to obtain 5, 10, and 30% concentrations. The prepared mixes were placed in plastic or aluminum chambers (IQ Chambers city in Sweden or Finn Chambers city in Finland respectively) by using a spatula. A negative control, which consisted of only white petroleum jelly, was also included. The patches were applied to the patient's upper back, in the upper middle of the scapulae (a clean and lesion-free area was preferred). A PT was also placed on residual hyperpigmented lesion scars in patients with FDE. Labels that indicated the drug and concentration in each chamber were affixed to the patch. The families were advised not to open, wet, or move the patch for 48 hours.

The patches were opened after 48 hours. The first reading was made after waiting for 15–20 minutes. Standard scoring systems of the international and European contact dermatitis research groups (ICDRG and ECDRG) were used to evaluate the results.⁹ The presence of erythema, indurations, and tiny vesicles was considered a positive test result. If there wasn't any reaction, it was considered as a negative result. Patients with negative or inconclusive results were reevaluated at 72 hours. If the results were still negative at 72 hours, then the families were informed and advised to come back to the clinic in case of rash or any lesions within 1 week.

Drug Provocation Test

Drug provocation tests were performed only in patients with negative PT results and mild cutaneous lesions. Oral challenge was not conducted for SCAR or FDE. The suspect drug was administered in divided doses (3–5 doses) every 30 minutes until a cumulative dose close to the age- and weight-adjusted daily dose of the drug was achieved. If any reaction occurred, then the test was discontinued. The challenge test was open and carried out by experienced physicians of our clinic.

Statistical Analysis

SPSS 17 (IBM, Armonk, NY) was used for statistical analyses. Discrete variables are expressed in numbers and percentages. Continuous variables are summarized by using means ± standard deviations for normally distributed data and as median and interquartile range (IQR) for non-normally distributed data. The χ2 test was used to examine nonparametric statistics, and continuous variables were compared by using the Mann–Whitney test for non-normally distributed data and the Student's t-test for normally distributed continuous data. A value of p < 0.05 was considered significant.

RESULTS

In this study, we analyzed the results of 105 drug PTs performed on 71 children. The patients' median age was 7 years (IQR, 4–11 years) and 57.7% of the patients (41/71) were boys. Only five patients had known atopy before the drug reaction: two had egg allergy and three had pollen sensitization. A total of nine patients (12.6%) had concomitant other allergic disease and seven (77%) of them had asthma. Fifty-four patients (76%) of them had other chronic diseases; of these, 81% (44/54) had epilepsy, 9% (5/54) had kidney disease, and 9% (5/54) had another disease, such as malignancy, metabolic disease, or rheumatoid disease. All 71 patients had dermatologic symptoms, which consisted of mild maculopapular reactions in 63.4% of the patients (45/71), severe cutaneous reactions in 32% (23/71 [DRESS in 9, AGEP in 3, SJS in 11]), and FDE in 4% (3/71). The demographic summaries of the study are shown in Table 1.

Table 1.

Descriptive summaries of study

IQR = Interquartile range; DRESS = drug rash with eosinophilia and systemic symptoms; AGEP = acute generalized exanthematous pustulosis; SJS = Steven-Johnson syndrome.

PTs Result

The 105 PTs were conducted with 30 different drugs at different concentrations. Antibiotics were the culprit drugs in 45.7% of the tests (48/105), and β-lactams accounted for 75% of the antibiotics (36/48). The second most common drugs were antiepileptics in 41.9% of the tests (44/105). All patch tested drugs were applied at 5% and 10% concentrations, except for montelukast, which was applied at concentrations of 5, 10, and 30%. The median time between PT and the last reaction was 3 months (IQR, 2–6 months; minimum-maximum, 1–54 months). In patients with DRESS and SJS, the time interval was at least 6 months (median, 6 months; minimum-maximum, 6–54 months). A total of 20 patients (28%) had positive PT results. Two patients had positive results at 48 hours, and 18 patients showed positive PT results at 72 hours. Of the positive PT results, 90% (18/20) were performed with antiepileptics and only 10% (2/20) were performed with antibiotics (β-lactams). The details of the positive PT results are shown in Table 2.

Table 2.

Details of positive drug patch tests

CADR = Cutaneous adverse drug reaction; AGEP = acute generalized exanthematous pustulosis; CAM = amoxicillin-clavulanic acid; DRESS = drug rash with eosinophilia and systemic symptoms; CMZ = carbamazepine; VPA = valproic acid; SJS = Steven Johnson syndrome; MPE = maculopapular exanthema; PHE = phenobarbital; FDE = fix drug eruption.

Antiepileptic PTs.

Forty-four PTs were performed with antiepileptics, 40% (18/44) of which used CMZ. Forty percent (18/44) of the antiepileptic PTs were positive, with CMZ showing the highest positivity rate at 66% (12/18) (Fig. 1). The details of antiepileptic PTs are shown in Table 3.

Figure 1.

Photographs of positive patch test results (A) with 10% carbamazepine, (B) with 10% carbamazepine, and (C) with 10% valproic acid.

Table 3.

Patch test with antiepileptics

n = numbers of patch test applied; SJS = Steven Johnson syndrome; DRESS = drug rash with eosinophilia and systemic symptoms.

Antibiotic PTs.

Forty-five PTs were performed with antibiotics. Of these, 84.4% (38/45) used β-lactams, and amoxicillin clavulanate accounted for 42% (16/38). Four percent (2/45) of the antibiotic PT results were positive, and both were amoxicillin clavulanate. The details of antibiotic PTs are shown in Table 4.

Table 4.

Patch test with antibiotic

n = number of patch test; SJS = Steven Johnson syndrome; AGEP = acute generalized exanthematous pustulosis; FDE = fix drug eruption; DRESS = drug rash with eosinophilia and systemic symptoms; + = positive patch test; – = negative patch test; PEN-V = penicillin V; PEN-G = penicillin G; CAM = amoxicillin-clavulanic acid; SAM = sulbactam-ampicillin; PIP-TAZ = piperecilin-tazaobactam; TMX = trimetoprim-sulfamethaxazole.

PTs according to CADR Phenotype

Thirty-two percent of the reactions (23/71) were SCAR. The PT details for the patients with CADR are shown in Table 5.

Table 5.

Drugs according to CADR phenotype

CADR = Cutaneous adverse drug reaction; MPE = maculopapular exanthema; FDE = fix drug eruption; DRESS = drug rash with eosinophilia and systemic symptoms; SJS = Steven Johnson syndrome; TEN = toxic epidermal necrolysis; AGEP = acute generalized exanthematous pustulosis; IQR = interquartile range; TMX = trimetoprim-sulfamethaxazol; VPA = valproic acid; Fe2 = iron; CAM = amoxicillin-clavulanic acid; CMZ = carbamazepine; PIP-TAZ = piperecilin-tazaobactam.

Provocation Test Results

Five patients who presented with MPE and had a negative PT result underwent an oral provocation test, which was also negative. Of these, four patients were challenged with β-lactam antibiotics and one was challenged with the ferrous form of iron. Alternative drugs were recommended to all the patients who could not be provoked, even if their PT result was negative.

DISCUSSION

In this study, we evaluated 71 children who underwent PT for suspected delayed drug hypersensitivity reaction. Of these, 32% (23/71) had SCAR, 28% (20/71) had positive PT results, and seven of the patients with positive PT test results had SCAR. The drug class with the highest rate of positive PT results was the antiepileptics (66% [12/18]). The reported frequency of PT positivity ranges widely in the literature, especially among children (range, 0,9–89%).^17,18 In a β-lactam hypersensitivity study with 213 children, Vila et al.¹⁷ reported only 0.9% PT positivity, whereas Lezmi et al.¹⁹ observed 19% PT positivity in a study of nonimmediate β-lactam reactions in 550 children. Atanasković-Marković et al.¹⁸ showed that 89.39% of PT results (59/66) were positive in children with a history of confirmed hypersensitivity reaction to antiepileptic drugs. This rate was higher than those in previous studies because it was associated with confirmed hypersensitivity reactions.¹⁸ However, the high positivity rate may also be related to their investigating only PTs of antiepileptic drugs, especially CMZ when considering that previous studies showed that CMZ has high PT positivity.^6,9,20,26 Consistent with these studies, our PT positivity rate was 28% overall and was higher among antiepileptic PTs (40%) and highest with CMZ (66%).

Previous adult studies reported that PT positivity varies by SCAR type.^6,21 Although studies that evaluated PT positivity in children are limited, they also demonstrate variability in PT positivity with CADR features, as in adults. In a study of SCARs, Liccioli et al.²² reported 22.6% (n = 7) positivity in 31 PTs; of those seven patients, five had DRESS, one had AGEP, and one had linear immunoglobulin A bullous disease, whereas all patients with SJS had negative PT results (n = 12). In another pediatric study, Misirlioğlu et al.²³ reported positive PT results in 3 of 3 patients with DRESS. Atanasković-Marković et al.¹⁸ observed that 9.1% of PT positivity was in patients with SCAR (7.6% DRESS and 1.5% SJS) and 90% were in those with cutaneous self-limiting and benign reactions. Among the children who underwent PT due to delayed CADR in our study, 63.3% (n = 45) had a clinical history of mild MPE, 32.3% (n = 23) of SCAR, and 4% (n = 3) of FDE. Among those with SCAR, the distribution was 47% with SJS/TEN (n = 11), 39% with DRESS (n = 9), and 13% with AGEP (n = 3).

PT sensitivity also varies, depending on the culprit drug class.^6,24,25 In previous studies, a positive drug PT result was detected most frequently with antibiotics (especially in the β-lactam group)^6,24,25 or antiepileptics (most frequently with CMZ).^6,26 Lammintausta et al.²⁷ investigated the usefulness of skin tests in CADR reactions and showed that PT was most useful with antibiotics (especially β-lactam and without macrolides), cardiovascular drugs, and antiepileptic drugs (especially CMZ and phenytoin). Santiago et al.²⁸ evaluated drug PT results in adults with DRESS and found 32.1% PT positivity, with 94.4% of positive PTs (17/18) performed with antiepileptics and 76% of those who used CMZ (13/17). In contrast, they also found that none of the 19 allopurinol PTs resulted in a positive reaction. In a study of children, Atanasković-Marković et al.¹⁸ reported the highest positivity rate among antiepileptics with CMZ (40% positive PT result [25/59]), followed by 32% positivity with lamotrigine (19/59), 16.9% with valproate (10/59), and 10% with phenobarbital (6/59). Similarly, antiepileptics accounted for 90% of the PT positivity (18/20) in our patient series, with CMZ being the most frequently positive antiepileptic. We also observed positivity with phenobarbital (2/6), lamotrigine (1/1), and valproate (3/11). PT data of previous studies^9,27 were mostly about adults and positivity was found most frequently with CMZ; our pediatric population's PT positivity was also mostly related with CMZ.

The second most common culprit drug with PT positivity was the β-lactam antibiotic amoxicillin (10% [2/20]). Our positivity with antibiotics (β-lactams) was lower than in some studies in the literature.^6,24,25 Pinho et al.29 analyzed the PT results of 260 patients ages 12–93 years who presented due to nonimmediate CADR to antibiotics. They obtained positive PT reactions in 56 patients (21.6%), and PT reactivity was higher for penicillins (25.4%). However, in their β-lactam hypersensitivity study in children, Vila et al.¹⁷ performed 203 PTs (152 patients with delayed reaction and 51 with undetermined reaction type) but reported that only two were positive, consistent with our study.

In addition to the drug administered, the concentration applied is an important factor in PT sensitivity. Low concentrations may cause false negativity, whereas high concentrations may, in rare cases, lead to reactivation of the reaction or to a more severe clinical reaction.¹⁶ Because SCARs such as DRESS and SJS can relapse at high concentrations, it is recommended to start with low concentrations.⁵ In this respect, we selected appropriate concentrations according to the clinical presentation of the patient and the nature of the culprit drug. We used most drugs at concentrations of 5% or 10%, in accordance with PT guideline.⁸ In patients with SCAR, the highest concentration we used was 10%; therefore, there may have been false-negative results in our study. However, in our study, there were 23 patients with severe drug reactions, and none of them had a recurrence or any reaction to the test. Although PT is safer than intracutaneous and provocation tests, there have been rare reports of reactions induced by PT. In a case report, Salman et al.³⁰ observed AGEP recurrence with ceftriaxone, whereas Mashiah and Brenner³¹ observed a recurrence of AGEP with acetaminophen. Similarly, some drugs may trigger a relapse when applied in PT, as in case reports of CMZ, acyclovir, and pseudoephedrine PTs that caused reactivations in the form of mild CADR or SCAR.¹ We observed no reactions from the 18 CMZ PTs and one acyclovir PT in our study.

CONCLUSION

Drug PTs are a preferable method in children due to their safety in the assessment of delayed drug reactions. PTs can assist in the diagnosis and identification of drugs that can be used in the future, especially in patients with SCAR. The sensitivity of PT was higher with antiepileptic drugs, especially CMZ, compared with other drugs.