The prevalence of penicillin allergy is estimated to be approximately 10% of the United States population,1 but it is also estimated that the large majority of these patients are not actually sensitive to the drug.2 A 2017 review supports this misrepresentation, noting that negative penicillin skin tests in multiple inpatient cohorts showed anywhere between 79% to 100% negative testing.3 Whereas the selection of antibiotics is ever expanding, so too are the number of resistant bacteria, and apt antibiotic stewardship applauds the ability to narrow coverage to penicillin when appropriate. Indeed, it is imperative to consider the implications of adding an allergy to a patient's medical record, particularly when it comes to antimicrobials. There are certain cutaneous drug eruptions labeled as allergies, when, in fact, rechallenge of the offending medication may be reasonable. Of course, there are other cutaneous adverse drug reactions (ADRs) that absolutely prohibit readministration of the culprit drug. This is, in part, why demystifying drug rashes is so important.
Drug rashes in the general population have been reported to occur to almost all prescription medications, at rates of more than 10 cases per 1,000 new users.4 There are at least 35 reported cutaneous drug-reaction patterns in the general population.5 In the pediatric population, antibiotics, antiepileptics, and nonsteroidal anti-inflammatory drugs (NSAIDs) are the most frequent culprits of ADRs, but vaccines and other medication classes are also implicated.6 As of 2001, it was estimated that the incidence of ADRs in the outpatient pediatric population was about 1.5% (and more than 6 times this number in the pediatric inpatient setting).7 It is thought that ADRs in the pediatric population are increasing due to growing rates of medication administration (prescription and over-the-counter) as well as polypharmacy in the pediatric population. A 2001 Canadian study found that 70% of prescription medications were prescribed to only 25% of the pediatric population.8 In general, drug reactions in the pediatric population can manifest differently depending on the age of the child, variations in immune system, and metabolic processing of drugs, as well as variability in the types of drugs being used.9
Pediatric Drug Reactions
There are multiple ways to categorize pediatric cutaneous ADRs. In this article, we focus first on classification by primary clinical morphology, then subclassify on hypersensitivity type, severity, and time to onset.
The clinical appearance of cutaneous drug eruption can vary widely. There are many instances in which cutaneous drug eruptions can overlap, and some are on a spectrum. Moreover, the clinical differential for drug rashes invariably includes nondrug etiologies.
Hypersensitivity classification of drug rashes is most commonly based on the classic Coombs and Gell type I-IV hypersensitivities. Type I is immunoglobulin E (IgE)-dependent and is an immediate hypersensitivity. This type of reaction is less common in neonates and young infants due to a less effective antibody (IgE)-mediated response.9 Type II is cytotoxic, type III is immune-complex dependent, and type IV is delayed-type hypersensitivity. As with clinical morphology, drug reactions often do not fit perfectly into one hypersensitivity type, but understanding the basic pathophysiology can be helpful in establishing a timeline and in predicting overall clinical outcomes.
Severity of the drug eruption is determined by the morbidity and mortality that can result. The World Health Organization defines a severe cutaneous adverse reaction as one that results in death or is life-threatening, requires inpatient care (or prolongation of inpatient care), or results in significant disability/incapacitation.10 Cutaneous manifestations can be harbingers of other systemic end organ damage such as in drug rash with eosinophilia and systemic symptoms (DRESS), but also may be the cause of morbidity and mortality such as in the case of Stevens-Johnson syndrome/toxic epidermolytic necrolysis (SJS/TEN). Understanding the potential severity of cutaneous drug eruptions can help guide decisions on whether to discontinue the causative medication or if rechallenge is acceptable.
As noted above, a basic understanding of the expected time to onset of cutaneous manifestations of ADRs is valuable in identifying the suspected culprit drug and is another important tool in the overall classification of these cutaneous reactions.
Wheals, Annular, Targetoid Lesions
At most, approximately 10% of episodes of urticaria are drug related.11 Urticaria is a transient rash consisting of edematous, erythematous wheals (Figure 1). Wheals can appear in multiple morphologies including serpiginous and annular, and can affect the trunk, face, and extremities. Urticaria tends to be pruritic.
Urticaria. Erythematous annular to polycyclic transient wheals on the back.
A clinical tip to elicit the transient nature of the rash is to circle a wheal and ask the patient if the wheal within the circle is still present 24 hours later. As mentioned previously, urticaria is less common in neonates and young infants due to a less dominant antibody (IgE) response.
Hypersensitivity classification is usually type I, although nonimmunologic mast cell destabilizers can cause urticaria as well. The condition is severe only if associated with anaphylaxis, otherwise it is typically mild. Average time to onset is minutes to hours. Common offending drugs include penicillin, sulfonamides, aspirin/NSAIDs, and contrast media.
Serum Sickness-Like Reaction/Urticaria Multiforme
The authors of this article consider the above entities to be on a spectrum, although some sources will differentiate based on whether drug (usually an antibiotic) administration preceded onset of rash, and whether symptoms of joint pain and fever are present. Cutaneous manifestations begin with urticarial-appearing wheals, often annular, that progress to form a violaceous to gray- colored center.12 Although these lesions can appear similar to true urticaria, the lesions in this condition are not transient, and typically can display a more dusky central color (Figure 2). Time of onset is also different from true urticaria, usually 1 to 3 weeks after drug initiation for serum sickness-like reaction/urticaria multiforme. The patient may be otherwise clinically unaffected; however, in some cases, the patient is febrile and may have accompanying joint pain.
Serum sickness-like reaction. Annular to polycyclic erythematous fixed wheals with dusky centers.
Acute hemorrhagic edema of infancy (AHEI) can present with a similar, dramatic morphology. In AHEI, however, there is true vasculitis noted on skin biopsy. Hypersensitivity classification is multifactorial; it is not immune-complex mediated, whereas classic serum sickness is a type III reaction.
Severity is mild to moderate discomfort and it is reasonable to “treat through” this rash if necessary. Average time to onset is 1 to 3 weeks. Common offending drugs include cephalosporins (cefaclor most commonly associated historically), penicillins, minocycline, sulfonamides, and buproprion.
Erythema multiforme (EM) presents with targetoid papules and plaques (should be palpable) (Figure 3), classically with three separate concentric rings of color, but atypical lesions with only two concentric rings may occur. Distribution tends to be fairly symmetric and commonly includes the extremities, particularly hands, elbows, and wrists. The face and trunk can be affected as well.10 EM is most often seen secondary to infection, rather than drug. The most common infectious causes include herpes simplex virus and Mycoplasma pneumoniae, so it is important to consider preceding infectious etiologies in these patients rather than the subsequent antimicrobial therapy. Mucosal involvement (Figure 4) can be notable, particularly in EM major and M. pneumoniae-induced rash and mucositis. Systemic symptoms of fever and arthralgias are seen with EM major. The presence of fever, mucosal symptoms, and EM-like lesions should also raise concern for Kawasaki disease, a separate, non–drug-induced entity. In the pediatric population, it is important to note that it is rare to see EM in infants, particularly EM major.13 A recent review suggested that when EM does occur in infants, it is most commonly associated with vaccination9 or infection.7 There have not been reported cases of EM in infants attributed to a drug. The lesions of EM can persist for weeks and tend to heal without scarring.10 Hypersensitivity classification is type IV. When ocular involvement is pronounced, patients EM major require prompt ophthalmology consultation to prevent permanent sequelae. Otherwise, EM minor tends to run a mild course. Average time to onset is generally 1 to 7 days. Common offending drugs include NSAIDs, sulfonamides, anticonvulsants, and allopurinol.
Erythema multiforme with targetoid lesions on the lower extremity. This 7-year-old also had oral mucosal erosions. Used with permission from Dr. Anthony J. Mancini, Northwestern University Feinberg School of Medicine.
Erythema multiforme major with mucosal involvement.
Papules, Macules, and Purpura
Morbilliform/Exanthematous Drug Rash
Named after the genus Morbillivirus (which includes rubeola virus, the cause of measles), this rash features erythematous macules and papules, leading to the commonly used moniker “maculopapular,” which often coalesce into patches and plaques (Figure 5). This eruption tends to start on the trunk then affects proximal and sometimes distal extremities. Of note, the rash can progress to superficial desquamation but does not typically form blisters or vesicles, does not slough, and may be itchy but is not generally tender or painful. Hypersensitivity classification is delayed type IV. Severity is mild; however, pruritus may be bothersome to the patient. Average time to onset is approximately 7 to 14 days but may be earlier if previously sensitized. Common offending drugs include penicillins, sulfonamides, and anticonvulsants.
Morbilliform drug eruption. Erythematous macules and papules, some coalescing into patches and plaques.
Drug Reaction with Eosinophilia and Systemic Symptoms
In a recently published small case series of pediatric cases of DRESS,14 all 10 patients invariably presented with fever >101.3°F and nine presented with a morbilliform rash morphology. Seven of 10 patients had accompanying facial edema. Other rash morphologies included purpuric and bullous lesions. Importantly, the cutaneous manifestations of DRESS are variable, and it is important to consider the entire clinical picture and assess for facial edema, lymphadenopathy, and hepatosplenomegaly. The timeline for DRESS helps to distinguish it from other morbilliform rashes, as it classically appears 2 to 6 weeks after initiation of drug. DRESS patients tend to have associated systemic symptoms of fever and general malaise, as well as adenopathy. The rash in DRESS tends to spread caudally,11 and as noted above is usually accompanied by facial (especially periorbital) edema and associated pruritus. The cutaneous manifestations of DRESS may progress to erythroderma.
DRESS may feature distinctive laboratory abnormalities including eosinophilia and atypical (reactive) lymphocytes, as well as other manifestations of internal organ damage, which may range from mild to severe. The liver is the most commonly affected internal organ, but the heart, lungs, kidneys, brain, and others may be affected as well.15 It is important to note that thyroid abnormalities due to DRESS may be delayed for months after initial rash onset, so these patients need close follow-up. DRESS is a severe ADR and rechallenge with the culprit drug should not be attempted. Hypersensitivity classification is delayed type IV. The severity is severe, and the average time to onset is 2 to 6 weeks. Common offending drugs include carbamazepine, phenytoin, phenobarbital, cephalosporins, dapsone, lamotrigine, penicillins, anti-retrovirals, sulfasalazine, and vancomycin.
Stevens-Johnson Syndrome/Toxic Epidermal Necrolysis
SJS/TEN is covered in the Bullae section below; however, due to the severity of this cutaneous ADR, it is mentioned in this section as well, as the initial morphology may appear morbilliform with dusky to erythematous macules coalescing into patches (Figures 6–8). Patients will likely appear ill and may complain of skin tenderness.
Toxic epidermal necrolysis. Dusky macular rash progressing to bullae.
Toxic epidermal necrolysis. Bullae with epidermal detachment and marked mucosal involvement.
Toxic epidermal necrolysis. Resolved eruption with residual hyper- and hypopigmentation.
Hypersensitivity Vasculitis (Leukocytoclastic Vasculitis)
Drug-induced leukocytoclastic vasculitis (LCV) is much more common in adults than in the pediatric population. LCV presents as palpable purpura, often on the lower extremities (Figure 9). A recent case series of 56 pediatric patients with LCV found that only 2% of cases may have been attributable to an ADR (and it is difficult to know if these were due to a drug or a preceding infection, as a culprit drug was never definitively identified).16 Therefore, it is important that drug-induced cutaneous small vessel vasculitis be a diagnosis of exclusion, and other causes such as infectious etiologies, or more rarely cryoglobulinemia, be ruled out. Additionally, in a patient with gastrointestinal symptoms and LCV, consideration should be given to IgA-mediated vasculitis (regardless of etiology), and renal function and blood pressure should be monitored. Hypersensitivity classification is type II or III. Severity is potentially severe if internal organ involvement, and the average time to onset is 1 to 3 weeks. Common offending drugs include penicillins, NSAIDs, cephalosporins, and sulfonamides.
Leukocytoclastic vasculitis. Palpable purpura on the lower extremity.
Acute Generalized Exanthematous Pustulosis
This rash often begins in intertriginous areas with erythema and will then progress to feature small, nonfollicular pustules on a background of erythema (Figure 10). The pustules in acute generalized exanthematous pustulosis (AGEP) are subcorneal, hence easily denuded, and may require an attentive examination to appreciate. Patients with AGEP may be febrile. The eruption usually resolves over weeks with superficial desquamation. AGEP is unusual in pediatric patients and may occur in the absence of drug administration. There are pediatric cases reported in response to viral infection, mercury exposure, and in response to systemic loxoscelism from brown recluse bite.17 Hypersensitivity classification is type IV, the severity is severe, and the average time to onset is 2 to 4 days. Common offending drugs include beta lactams, macrolides, and clindamycin.
Acute generalized exanthematous pustulosis. (A and B) Nonfollicular-based superficial homogenous pustules.
Symmetric Drug-Related Intertriginous and Flexural Erythema
Symmetric drug-related intertriginous and flexural erythema begins as an erythematous, pruritic eruption, localized to flexural skin surfaces, and often beginning in the gluteal and/or inguinal areas but ultimately involving any skin fold. The rash tends to be well demarcated and can be associated with eczematous scale and/or desquamation of the affected areas.18 Hypersensitivity classification is type IV, severity is mild to moderate, and the average time to onset is 1 to 7 days. Common offending drugs include penicillins and cephalosporins.
Stevens-Johnson Syndrome/Toxic Epidermolytic Necrolysis
As noted above, SJS/TEN may initially present with a morbilliform eruption. In these cases, the patient will present with notable symptoms of systemic illness, oftentimes the skin will be tender, and the examination will reveal mucosal involvement. Time to onset and rapid progression of the rash can help to differentiate SJS/TEN from DRESS, as DRESS may feature mucosal involvement as well. The morphology of SJS/TEN progresses rapidly from dusky macules and patches to the formation of bullae and complete denudation of the epidermis as a result of epidermal necrosis.
SJS and TEN are differentiated depending on the percentage of body surface area (BSA) sloughing. SJS is defined as <10%, overlap as 10% to 30%, and TEN as >30% of the BSA.19 SJS/TEN is rare in the pediatric population. A large cross-sectional study looking at data from 2009 to 2012 noted an estimated frequency of 4.3 to 5.8 per million hospitalized children for SJS and substantially less for TEN (0-0.7 per million).19 Hypersensitivity classification is type IV, severity is severe, and average time to onset is 7 to 21 days. Common offending drugs include sulfonamides, NSAIDs, and anticonvulsants.
Other Cutaneous ADRS
In addition to the above, more important cutaneous ADRs, others warrant at least a mention and include drug-induced lupus, fixed drug eruptions (Figure 11), lichenoid drug eruption, acneiform drug eruptions, and bullous drug eruptions. As mentioned earlier, there are dozens of other reported cutaneous drug eruption morphologies as well.
Fixed drug eruption. Dusky patch on the palm, which was recurrent upon drug ingestion.
Diagnosis and Testing
Diagnosis of cutaneous drug rashes is overwhelmingly clinical. Particularly in pediatric patients, where quantity of daily medications is lower, identifying a recent drug as a culprit can potentially be straightforward. It is helpful to establish a timeline of start date for each drug and plot against the start date of the cutaneous eruption. The morphology of each rash, as noted above, can have some clinical overlap, so deciding on a general category of morphology can be helpful. After establishing the category and timeline, it is important to consider all other medications the patient is taking as possible culprits (including over-the-counter agents and herbal supplements). It is important to keep in mind that even if the patient has presently discontinued the drug, the rash may still be apparent, as some rashes resolve very slowly, despite discontinuation of the offending drug.
A thorough review of systems is useful to determine if the patient is experiencing systemic symptoms such as fever, joint pain, and/or malaise. As noted above, endorsing skin pain is a classic finding in SJS/TEN.
Laboratory analysis can be useful in some drug rashes, particularly in situations where it is unclear whether the rash is drug driven. Blood testing may reveal peripheral eosinophilia, which can help tip the scale toward cutaneous ADR. With respect to differentiating between some of the above eruptions, DRESS can feature elevated hepatic transaminases, as well as acute kidney injury, atypical lymphocytes, and eosinophilia. AGEP may show similar laboratory abnormalities. SJS/TEN may be accompanied by end organ damage due to dehydration and/or sepsis.
Skin biopsy can be quite helpful in clarifying specific causes and in differentiating more severe reactions, such as those characterized by cell-mediated cytotoxicity, from others.
Management and Prevention
Management of cutaneous drug eruption depends on the severity. In mild reactions, such as morbilliform drug eruption, as well as serum sickness-like reaction and even some mild presentations of AGEP, it is important to weigh the risk and benefit of continuing the culprit drug. In some cases, it may be necessary to treat through with symptomatic control (antihistamines, topical corticosteroids, wet wraps), while observing the patient to assess for potential progression to a more concerning clinical picture.
In patients with severe cutaneous drug eruptions, the culprit drug should be discontinued immediately and added to the patient allergy list, and alternative therapy should be pursued. Treatment of cutaneous manifestations beyond drug discontinuation depends upon the type of drug reaction and is further detailed below.
For AGEP, topical steroids are often indicated. The skin manifestations improve over the course of 1 to 2 weeks, followed by superficial desquamation. In rare, very severe or symptomatic cases, a taper of systemic steroids may be warranted.
For DRESS, due to the potential systemic manifestations of this ADR, systemic corticosteroids are often warranted and may be initiated at the time the diagnosis is suspected. Clinicians should anticipate a prolonged steroid taper due to the risk of rebound. The length of steroid taper is typically a minimum of 6 weeks, with close follow-up. If there is concern for rebound, the steroid treatment may need to be extended. In addition, as noted above, patients with DRESS will need monitoring of thyroid function for months after the rash has improved, as delayed-onset hypothyroidism has been reported with this drug reaction. There are pediatric cases of type III polyglandular autoimmune syndrome following DRESS, as well.20 Supportive care as needed for renal and hepatic function is most appropriate, depending on the extent of end organ damage.
Similarly, SJS/TEN requires intensive supportive care and a multidisciplinary approach including ophthalmology, nephrology, dermatology, surgery (ideally a specialized burn care unit if available), infectious diseases, and other subspecialties. Beyond supportive care there are a few systemic agents that may be considered on a case-by-case basis, including intravenous immunoglobulin, some of the newer biologic agents, and cyclosporine.
Drug reactions in the pediatric population are an important entity to consider on the list of differential diagnoses for cutaneous eruptions. If a drug reaction is suspected, it is important to determine the most likely type of reaction, which can help elucidate the predicted severity of the rash. Timeline of rash onset plotted against timeline of the drug start date can help in identifying the culprit agent. Potential or actual severity of the eruption then helps guide the next best steps in management.
- Albin S, Agarwal S. Prevalence and characteristics of reported penicillin allergy in an urban outpatient adult population. Allergy Asthma Proc. 2014;35(6):489–494. doi:10.2500/aap.2014.35.3791 [CrossRef] PMID:25584917
- Joint Task Force on Practice Parameters American Academy of Allergy, Asthma and ImmunologyAmerican College of Allergy, Asthma and ImmunologyJoint Council of Allergy, Asthma and Immunology. Drug allergy: an updated practice parameter. Ann Allergy Asthma Immunol. 2010;105(4):259–273. doi:10.1016/j.anai.2010.08.002 [CrossRef] PMID:20934625
- Sacco KA, Bates A, Brigham TJ, Imam JS, Burton MC. Clinical outcomes following inpatient penicillin allergy testing: a asystematic review and meta-analysis. Allergy. 2017;72(9):1288–1296. doi:10.1111/all.13168 [CrossRef] PMID:28370003
- Stern RS. Clinical practice. Exanthematous drug eruptions. N Engl J Med. 2012;366(26):2492–2501. doi:10.1056/NEJMcp1104080 [CrossRef] PMID:22738099
- Litt J. Litt's Drug Eruptions & Reactions Manual. New York, New York: CRC Press; 2013:469–474. doi:10.1201/b15347-27 [CrossRef]
- Smyth RM, Gargon E, Kirkham J, et al. Adverse drug reactions in children—a systematic review. PLoS One. 2012;7(3):e24061. doi:10.1371/journal.pone.0024061 [CrossRef] PMID:22403604
- Impicciatore P, Choonara I, Clarkson A, Provasi D, Pandolfini C, Bonati M. Incidence of adverse drug reactions in paediatric in/out-patients: a systematic review and meta-analysis of prospective studies. Br J Clin Pharmacol. 2001;52(1):77–83. doi:10.1046/j.0306-5251.2001.01407.x [CrossRef] PMID:11453893
- Khaled LA, Ahmad F, Brogan T, et al. Prescription medicine use by one million Canadian children. Paediatr Child Health. 2001;8(suppl A):6A–56A. doi:10.1093/pch/8.suppl_A.6A [CrossRef]
- Schachner LA, Hansen RC. Pediatric Dermatology. New York, New York/USA:Mosby;2010:1698.
- Bolognia J, Schaffer JV, Lorenzo C. Dermatology. New York, NY: Elsevier; 2018:332–373.
- Paller AS, Mancini AJ. Hurwitz Clinical Pediatric Dermatology. New York, NY: Elsevier; 2016:467.
- Sempau L, Martín-Sáez E, Gutiérrez-Rodríguez C, Gutiérrez-Ortega MC. Urticaria multiforme: a report of 5 cases and a review of the literature. Actas Dermosifiliogr. 2016;107(1):e1–e5. doi:10.1016/j.adengl.2015.11.010 [CrossRef] PMID:25843790
- Zoghaib S, Kechichian E, Souaid K, Soutou B, Helou J, Tomb R. Triggers, clinical manifestations, and management of pediatric erythema multiforme: a systematic review. J Am Acad Dermatol. 2019;81(3):813–822. doi:10.1016/j.jaad.2019.02.057 [CrossRef] PMID:31331726
- Oberlin KE, Rahnama-Moghadam S, Alomari AK, Haggstrom AN. Drug reaction with eosinophilia and systemic symptoms: pediatric case series and literature review. Pediatr Dermatol. 2019;36(6):887–892. doi:10.1111/pde.13949 [CrossRef] PMID:31423638
- Sola D, Rossi L, Sainaghi PP, Pirisi M. DRESS syndrome with cerebral vasculitis. Intern Med. 2013;52(12):1403–1405. doi:10.2169/internalmedicine.52.8762 [CrossRef] PMID:23774556
- Johnson EF, Wetter DA, Lehman JS, Hand JL, Davis DM, Tollefson MM. Leukocytoclastic vasculitis in children: clinical characteristics, subtypes, causes and direct immunofluorescence findings of 56 biopsy-confirmed cases. J Eur Acad Dermatol Venereol. 2017;31(3):544–549. doi:10.1111/jdv.13952 [CrossRef] PMID:27580419
- Meadows KP, Egan CA, Vanderhooft S. Acute generalized exanthematous pustulosis (AGEP), an uncommon condition in children: case report and review of the literature. Pediatr Dermatol. 2000;17(5):399–402. doi:10.1046/j.1525-1470.2000.017005399.x [CrossRef] PMID:11085672
- Dogru M, Ozmen S, Ginis T, Duman H, Bostanci I. Symmetrical drug-related intertriginous and flexural exanthema (baboon syndrome) induced by amoxicillin-clavulanate. Pediatr Dermatol. 2012;29(6):770–771. doi:10.1111/j.1525-1470.2011.01577.x [CrossRef] PMID:22044408
- Hsu DY, Brieva J, Silverberg NB, Paller AS, Silverberg JI. Pediatric Stevens-Johnson syndrome and toxic epidermal necrolysis in the United States. J Am Acad Dermatol. 2017;76(5):811–817.e4. doi:10.1016/j.jaad.2016.12.024 [CrossRef] PMID:28285784
- Morita C, Yanase T, Shiohara T, Aoyama Y. Aggressive treatment in paediatric or young patients with drug-induced hypersensitivity syndrome (DiHS)/drug reaction with eosinophilia and systemic symptoms (DRESS) is associated with future development of type III polyglandular autoimmune syndrome. BMJ Case Rep. 2018;2018:bcr-2018225528. doi:10.1136/bcr-2018-225528 [CrossRef] PMID:30368475