Clinical Characteristics, Disease Course, and Outcomes of Patients With Acute Generalized Exanthematous Pustulosis in the US

Key Points

Question

What are the clinical characteristics, disease course, and outcomes of patients with acute generalized exanthematous pustulosis (AGEP) in the US?

Findings

In this case series evaluation of 340 patients with AGEP, common physical examination and laboratory findings were fever, neutrophilia, eosinophilia, and hypocalcemia. A small percentage of patients experienced acute elevations in creatinine or liver enzyme levels; newly started treatment with antimicrobials was the most common suspected cause, and most patients had experienced resolution or improvement as of the most recent available data.

Meaning

The findings of this case series suggest that AGEP, commonly triggered by antimicrobial agents, may be associated with systemic complications in a minority of patients and typically resolves with symptomatic treatment.

Abstract

Importance

Acute generalized exanthematous pustulosis (AGEP) is a rare, severe cutaneous adverse reaction associated with systemic complications. Currently available data are largely limited to small retrospective case series.

Objective

To describe the clinical characteristics, disease course, and outcomes of a heterogeneous group of patients with AGEP across the US.

Design, Setting, and Participants

A retrospective review of a case series of patients was conducted from January 1, 2000, through July 31, 2020. All 340 included cases throughout 10 academic health systems in the US were scored retrospectively using the EuroSCAR scoring system, and patients with a score corresponding to probable or definite AGEP and aged 18 years or older were included.

Main Outcomes and Measures

Patient demographic characteristics, clinical course, suspected causative agent, treatment, and short- and long-term outcomes.

Results

Most of the 340 included patients were women (214 [62.9%]), White (206 [60.6%]), and non-Hispanic (239 [70.3%]); mean (SD) age was 57.8 (17.4) years. A total of 154 of 310 patients (49.7%) had a temperature greater than or equal to 38.0 °C that lasted for a median of 2 (IQR, 1-4) days. Of 309 patients, 263 (85.1%) developed absolute neutrophilia and 161 patients (52.1%) developed either absolute or relative eosinophilia. Suspected causes of AGEP were medications (291 [85.6%]), intravenous contrast agents (7 [2.1%]), infection (3 [0.9%]), or unknown (39 [11.5%]). In 151 cases in which a single medication was identified, 63 (41.7%) were β-lactam antimicrobials, 51 (33.8%) were non–β-lactam antimicrobials, 9 (6.0%) were anticonvulsants, and 5 (3.3%) were calcium channel blockers. The median time from medication initiation to AGEP start date was 3 (IQR, 1-9) days. Twenty-five of 298 patients (8.4%) had an acute elevation of aspartate aminotransferase and alanine aminotransferase levels, with a peak at 6 (IQR, 3-9) days. Twenty-five of 319 patients (7.8%) experienced acute kidney insufficiency, with the median time to peak creatinine level being 4 (IQR, 2-5) days after the AGEP start date. Treatments included topical corticosteroids (277 [81.5%], either alone or in combination), systemic corticosteroids (109 [32.1%]), cyclosporine (10 [2.9%]), or supportive care only (36 [10.6%]). All-cause mortality within 30 days was 3.5% (n = 12), none of which was suspected to be due to AGEP.

Conclusions and Relevance

This retrospective case series evaluation of 340 patients, the largest known study cohort to date, suggests that AGEP onset is acute, is usually triggered by recent exposure to an antimicrobial, may be associated with liver or kidney complications in a minority of patients, and that discontinuation of the triggering treatment may lead to improvement or resolution.

This case series examines the development of acute generalized exanthematous pustulosis.

Introduction

Acute generalized exanthematous pustulosis (AGEP) is a rare, severe cutaneous adverse reaction with an estimated incidence of 1 to 5 cases per million per year.¹ The reaction is defined by acute development of generalized erythema with dozens to hundreds of overlying sterile nonfollicular pustules that begin to resolve after the offending agent, usually a new medication, is discontinued.² Although generally associated with less morbidity than other severe cutaneous adverse reactions, systemic complications of AGEP, such as acute kidney insufficiency and hepatitis, and a mortality rate of less than 5% have been reported.^2,3 Because of its rarity, most AGEP studies are retrospective case series with fewer than 50 cases.^4,5,6,7,8,9 In this study, we sought to perform a large-scale retrospective review of AGEP cases across 10 different academic health systems in the US.

Methods

Setting

This was a retrospective case series using medical records conducted at 10 academic dermatology departments: Mass General Brigham/Harvard Medical School, University of Pennsylvania, Yale University, New York University, University of California San Francisco, Vanderbilt University Medical Center, Medical College of Wisconsin, Northwell Health, Oregon Health Sciences University, and the University of Rochester. The study was approved by the institutional review boards at each site. Patient consent was waived by all institutional review boards because the data were deidentified.

Identifying Potential Cases

Study sites used any or all of the following methods to identify cases of potential AGEP at their institution from January 1, 2000, through July 31, 2020: free-text search of emergency department, inpatient, or ambulatory patient notes; free-text search of pathology records; International Statistical Classification of Diseases, 10th Revision codes L27.0 (generalized skin eruption due to drugs and medicaments taken internally) and L53.8 (other specified erythematous conditions); and inpatient dermatology consult logs; search terms are available in eTable 1 in the Supplement. Cases by institution are listed in eTable 2 in the Supplement. Demographic data were obtained from the electronic health record. Data on race were included in too numerous ways for our study population to be compared with those of similar studies surrounding AGEP.

Screening for Study Inclusion

Potential cases were reviewed to confirm that AGEP had been considered as a differential diagnosis of a consulting or primary dermatologist in a patient aged 18 years or older at the time of diagnosis. All potential cases were scored using the EuroSCAR scoring system,¹ and those with a score of at least 5 (probable AGEP) were included. If a patient had multiple episodes of AGEP reported in their record, the earliest episode within our study period was used.

Time Course and Causative Agent

The AGEP time course was defined as the time between the AGEP start and end dates (definitions in eTable 1 in the Supplement). To identify the causative agent, each dermatology team’s initial consult and progress notes were reviewed to determine the suspected agent(s) for each case and the dates of exposure. For inclusion in the further analysis of specific medications, the dermatology team had to have listed a single medication as the suspected cause, and the start date of the medication had to be known and whether it was initiated before or on the AGEP start date. If infection was considered to be the cause of AGEP, relevant serologic and culture results were collected when available.

Systemic Involvement

Acute kidney insufficiency was defined as a creatinine level during the AGEP episode 1.5 times greater than the creatinine level measured in the 7 days before development of AGEP. Acute transaminitis was defined as alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels greater than twice the upper limit of normal during the AGEP course. If the patient’s most recent AST and ALT values in the 6 months preceding development of AGEP were also twice normal, transaminitis was considered chronic and these patients were excluded from the acute transaminitis analysis.

Treatment and Follow-up

Initial consult and progress notes were screened to collect data on treatments initiated for AGEP in each patient. For systemic medications, the maximum dose and treatment dates were collected when available. Medical records were reviewed to determine whether each patient had follow-up with dermatology to confirm that AGEP remained the most likely diagnosis. Follow-up data were also screened to determine whether patch testing had been performed to confirm allergy.

Statistical Analysis

Descriptive statistics were used to summarize the patient and disease characteristics, treatments, and outcomes. Categorical variables were summarized using frequencies and proportions. Continuous variables were summarized using means with SDs or medians (IQR) and range (minimum and maximum), as appropriate. Data analysis was conducting using Microsoft Excel for Mac, version 16.16.26 (Microsoft Corp) and Stata, version 16 (StataCorp LLC).

Results

Study Population

A total of 354 patients across all 10 sites met our study inclusion criteria and received a detailed medical records review. Fourteen of these patients (4.0%) had follow-up visits with a dermatologist that indicated a new diagnosis other than AGEP was more likely (pustular psoriasis [n = 5], drug reaction with eosinophilia and systemic symptoms [n = 5], morbilliform drug eruption [n = 2], Sneddon-Wilkinson [n = 1], and immunoglobulin A pemphigus [n = 1]), and were excluded from our analysis.

The final study population of 340 patients had a mean (SD) age of 58 (17.4) years; 214 were women (62.9%), 126 were men (37.1%), 206 were White (60.6%), and 239 were non-Hispanic (70.3%) (Table 1). Further characteristics included a history of diabetes (97 [28.5%]), chronic kidney disease (79 [23.2%]), and a history of psoriasis (24 [7.1%]); 64 patients (18.8%) had a previous drug hypersensitivity reaction (urticaria, anaphylaxis, or angioedema), 27 had a morbilliform drug reaction (7.9%), 6 (1.8%) had a history of a drug reaction with eosinophilia and systemic symptoms, and 10 (2.9%) had a history Stevens-Johnson syndrome/toxic epidermal necrolysis.

Table 1. Study Population.

Characteristic	No. (%)
Patients, No.	340
EuroSCAR score, median (IQR) [range]	8 (7-9) [5-12]
Probable AGEP (score 5-7)	128
Definite AGEP (score 8-12)	212
Age, mean (SD) [range], y	57.8 (17.4) [18-90]
Sex
Male	126 (37.1)
Female	214 (62.9)
Ethnicity
Hispanic or Latino	20 (5.9)
Not Hispanic or Latino	239 (70.3)
Not specified	81 (23.8)
Racea
American Indian/Alaska Native	1 (0.3)
Asian	15 (4.4)
Black or African American	58 (17.1)
Native Hawaiian or Pacific Islander	1 (0.3)
White	206 (60.6)
Other	20 (5.9)
Not specified	39 (11.5)
Associated comorbidities
Diabetes	97 (28.5)
Chronic kidney disease	79 (23.2)
Psoriasisb	24 (7.1)
Plaque	8 (33.3)
Pustular	7 (29.2)
Erythrodermic	4 (16.7)
Inverse	1 (4.2)
Guttate	0
Type unknown	9 (37.5)
Psoriatic arthritis	3 (0.9)
History of drug rash
Hypersensitivity (urticaria, anaphylaxis, or angioedema)	64 (18.8)
Morbilliform	27 (7.9)
Lichenoid	2 (0.6)
Fixed drug eruption	3 (0.9)
History of SCAR
DRESS	6 (1.8)
SJS/TEN	10 (2.9)

Abbreviations: AGEP, acute generalized exanthematous pustulosis; DRESS, drug reaction with eosinophilia and systemic symptoms; SCAR, severe cutaneous adverse reaction; SJS/TEN, Stevens-Johnson syndrome/toxic epidermal necrolysis.

^a Classifications were made based on how patient demographic data were available within the electronic medical records at home institution of Mass General Brigham. These were the choices available to participating study sites in the standardized data entry form. Collaborators then chose from these available choices based on what information was available in the demographic data of their home institution. Other race was chosen when a patient’s racial demographic information was not available or was listed as something other than one of the available choices in the standardized data entry form.

^b All subtypes of psoriasis reported for each patient were included.

Clinical Course and Diagnostic Criteria

A total of 178 patients (52.4%) developed AGEP while hospitalized (Table 2), although 100% of the study cohort was ultimately admitted to the hospital. The median EuroSCAR score of the included population was 8 (IQR, 7-9), with 128 patients (37.6%) receiving a score indicating probable and 212 patients (62.4%) receiving a score indicating definite AGEP. Twenty patients (5.9%) had mucosal involvement of the pustular eruption; the oral mucosa was involved in 18 patients. Two hundred twenty patients (64.7%) experienced desquamation as part of their clinical course. A skin biopsy was performed in 245 patients (72.1%), with 221 patients (90.2%) having some histologic features consistent with AGEP and 50 (20.4%) demonstrating both papillary edema and spongiform subcorneal and/or intraepidermal pustules (Table 2).

Table 2. Clinical Course.

Variable	No. (%)
Site of initial presentation
Inpatient (developed while already admitted)	178 (52.4)
Outpatient
Dermatology	28 (8.2)
Other	28 (8.2)
Emergency department	94 (27.6)
Other/not specified	12 (3.5)
Suspected cause of AGEP
Medication	291 (85.6)
Single potential agent	180 (61.9)
Multiple potential agents	111 (32.6)
Infection	3 (0.9)
Intravenous contrast agent	7 (2.1)
Other/unknown	39 (11.5)
Mucosal involvementa	20 (5.9)
Ocular	0
Oral	18 (90.0)
Nasal	1 (5.0)
Genital	3 (15.0)
Biopsy performed	245 (72.1)
Both papillary edema and spongiform subcorneal and/or intraepidermal pustules	50 (20.4)
Either papillary edema with any kind of pustule other than spongiform or spongiform subcorneal and/or intraepidermal pustules	167 (68.2)
Only exocytosis of polymorphonuclear neutrophils	4 (1.6)
Other	24 (9.8)
Fever (temperature ≥38.0 °C)	154/310 (49.7)
Fever duration, median (IQR) [range], d	2 (1 to 4) [0 to 21]
Start of fever relative to AGEP start date, median (IQR) [range], d	1 (0 to 3) [−8 to 11]
Fever temperature, median (IQR) [range], °C	38.8 (38.3 to 39.4) [38.0 to 40.6]
Elevated PMNs	263/309 (85.1)
Eosinophilia	161/309 (52.1)
Elevated WBC count	270/289 (93.4)
Hypocalcemia	207/315 (65.7)
Acute transaminitis (uncorrected)	60/298 (20.1)
Baseline AST/ALT values available	34/60 (56.7)
Acute transaminitis after considering baseline AST/ALT values	25/34 (73.5)
Corrected frequency of acute transaminitis	25/298 (8.4)
Days to peak AST/ALT values relative to AGEP start date, median (IQR)	6 (3-9)
No. with creatinine measured during AGEP course	319 (93.8)
No. with baseline creatinine in 7 d prior to AGEP	86/319 (27.0)
No. with AKI	25/86 (29.1)
Prevalence of AKI	25/319 (7.8)
AKI relative to AGEP start date, median (IQR)	4 (2-5)

Abbreviations: AGEP, acute generalized exanthematous pustulosis; AKI, acute kidney insufficiency; ALT, alanine aminotransferase; AST, aspartate aminotransferase; PMN, polymorphonuclear neutrophils; WBC, white blood cell.

^a All types of mucosal involvement for each patient were included.

A total of 154 of 310 patients (49.7%) with data available had a confirmed fever (temperature ≥38.0 °C) while in the health care setting, and these fevers had a median duration of 2 (IQR, 1-4) days, started a median of 1 (IQR, 0-3) day after erythema or pustules were first noticed, and reached a median maximum temperature of 38.8 °C (IQR, 38.3-39.4 °C). A total of 263 of 309 patients (85.1%) developed absolute neutrophilia, and 161 of 309 (52.1%) developed either absolute or relative eosinophilia. Hypocalcemia (uncorrected) was found in 207 of 315 patients (65.7%).

Suspected Causative Agent

According to the consulting or primary dermatology teams, a medication was the suspected causative agent in 291 patients (85.6%), with 180 (61.9%) cases narrowed to a single suspected mediation. An intravenous contrast agent was the suspected cause in 7 patients (2.1%) and infection was the suspected cause in 3 patients (0.9%, 2 with incomplete negative or negative serologic testing and 1 of whom had both blood- and wound-culture–positive results for Staphylococcus aureus). In 39 patients (11.5%), no suspected causative agent could be determined.

There were 151 patients (44.4%) in whom a single medication was suspected and was known to be started before or on the day that erythema or pustules were first noted (Table 3); 114 (75.4%) were antimicrobials. Sixty-three cases (41.7%) were suspected to be caused by a β-lactam antimicrobial: 27 (17.9%) were a cephalosporin and 17 (11.3%) were piperacillin with tazobactam. Non–β-lactam antimicrobial was the second most common medication category at 51 cases (33.8%), followed by anticonvulsants (9 [6.0%]), analgesics (5 [3.3%]), calcium channel blockers (5 [3.3%]), and hydroxychloroquine (4 [2.6%]). The median time from medication initiation to the AGEP start date in this group of 151 cases was 3 (IQR, 1-9; range, 0-79) days.

Table 3. Medication Frequencies and Initiation Date Relative to AGEP Start Date for Cases in Which a Single Agent Was Suspected and Medication Start Date Was Before or on AGEP Start Date^a .

Medication type	No. (%)
β-lactam antimicrobials	63 (41.7)
Amoxicillin	7 (4.6)
Amoxicillin with clavulanic acid	5 (3.3)
Ampicillin with sulbactam	1 (0.7)
Cefalexin	2 (1.3)
Cefazolin	6 (4.0)
Cefepime	6 (4.0)
Cefpodoxime	2 (1.3)
Ceftazidime	4 (2.6)
Ceftriaxone	6 (4.0)
Cefuroxime	1 (0.7)
Ertapenem	3 (2.0)
Meropenem	1 (0.7)
Oxacillin	2 (1.3)
Piperacillin with tazobactam	17 (11.3)
Non–β-lactam antimicrobials	51 (33.8)
Azithromycin	2 (1.3)
Ciprofloxacin	3 (2.0)
Clindamycin	9 (6.0)
Dapsone	1 (0.7)
Daptomycin	1 (0.7)
Doxycycline	2 (1.3)
Levofloxacin	3 (2.0)
Metronidazole	1 (0.7)
Minocycline	1 (0.7)
Moxifloxacin	1 (0.7)
Trimethoprim with sulfamethoxazole	11 (7.3)
Vancomycin	16 (10.6)
Antifungals	2 (1.3)
Nystatin	1 (0.7)
Terbinafine	1 (0.7)
Analgesics	5 (3.3)
Celecoxib	1 (0.7)
Diclofenac sodium with misoprostol	1 (0.7)
Ibuprofen	1 (0.7)
Oxycodone	1 (0.7)
Oxycodone with paracetamol	1 (0.7)
Anticonvulsants	9 (6.0)
Carbamazepine	1 (0.7)
Lamotrigine	2 (1.3)
Levetiracetam	1 (0.7)
Phenytoin	2 (1.3)
Sodium valproate	1 (0.7)
Calcium channel blockers	5 (3.3)
Diltiazem	3 (2.0)
Nicardipine	1 (0.7)
Nifedipine	1 (0.7)
Others	13 (8.6)
Bupropion	3 (2.0)
Brentuximab	1 (0.7)
Hydroxychloroquine	4 (2.6)
Imatinib	1 (0.7)
Metoprolol	1 (0.7)
Montelukast	1 (0.7)
Venlafaxine	1 (0.7)

Abbreviation: AGEP, acute generalized exanthematous pustulosis.

^a Total of 151 patients.

Systemic Involvement During AGEP Course

A total of 298 patients had AST and ALT values checked at least once during the AGEP course; 26 patients had liver function tests recorded during their AGEP course but did not have prior baseline liver function tests to compare to determine whether their transaminitis was potentially chronic. Levels in 60 patients (20.1%) who had data on their liver function testing during AGEP as well as at baseline were twice the upper limit of normal in at least 1 of the 2 values (Table 2). Thirty-four of those 60 patients (56.7%) had additional AST and ALT values determined from within 6 months before development of AGEP, 9 of which indicated potentially chronically elevated AST and ALT levels. The 25 remaining cases composed 8.4% of the 298 patients whose AST and ALT levels were checked at least once in the AGEP course. Among this subgroup, the median time to peak AST and ALT values was 6 (IQR, 3-9) days after the AGEP start date.

A total of 319 patients had their creatinine level measured during the AGEP course, with 86 (27.0%) of those individuals also having a baseline creatinine level within 7 days before the start of the AGEP course. Of those 86 patients, 25 (29.1%) had an increase of at least 1.5 times the baseline level. Among this subgroup, the median time to the peak creatinine level was 4 (IQR, 2-5) days after the AGEP start date. Those 25 patients were 7.8% of all 319 patients with available creatinine level data.

Treatment

In addition to stopping treatment with the suspected medication, the most common treatment was topical corticosteroids, with 277 patients (81.5%) receiving these agents either alone or in combination with other treatments (Table 4). The use of systemic treatments included corticosteroids (109 [32.1%]), cyclosporine (10 [2.9%]), and/or acitretin (2 [0.6%]), infliximab (2 [0.6%]), intravenous immunoglobulin (2 [0.6%]), or dapsone (1 [0.3%]). Supportive care only was used for 36 patients (10.6%).

Table 4. AGEP Treatment in 340 Patients.

Treatmenta	No. (%)
Supportive care only	36 (10.6)
Topical corticosteroids	277 (81.5)
Systemic corticosteroids	109 (32.1)
Other systemic treatment	17 (5.0)
Acitretin	2 (0.6)
Cyclosporine	10 (2.9)
Dapsone	1 (0.3)
Infliximab	2 (0.6)
Intravenous immunoglobulin	2 (0.6)

Abbreviation: AGEP, acute generalized exanthematous pustulosis.

^a All treatments used for each patient were included.

Outcomes and Follow-up

As of the last known record for this patient cohort, AGEP had completely resolved (190 [55.9%]), improved (98 [28.8%]), or progressed (1 [0.3%]) (Table 5). All-cause mortality within 30 days was 3.5% (n = 12), none of which was suspected to be due to AGEP. Data were available on 102 patients to identify the date of complete pustular resolution as determined by a dermatologist. Among this subgroup, the median length from initial presentation to complete pustular resolution (ie, the AGEP course) was 8 (IQR, 5-12) days.

Table 5. Outcomes in 340 Patients.

Variable	Median (IQR) [range]
Postpustular desquamation, No. (%)	220 (64.7)
Last available outcome, No. (%)
Improvement of rash	98 (28.8)
Resolution of rash	190 (55.9)
Progression of rash	1 (0.3)
Deatha	12 (3.4)
Other/unknown	39 (11.5)
Length of hospitalization, d	9 (6-13) [0-144]
Length of AGEP course, d
End date determined by complete pustular resolution according to a dermatologist, d	8 (5-12) [2-144]
No.	102
End date determined by complete pustular resolution according to a nondermatologist, d	8 (6-10) [3-62]
No.	35
End date (hospital discharge date), d	8 (6-10) [2-28]
No.	104

Abbreviation: AGEP, acute generalized exanthematous pustulosis.

^a Sepsis/septic shock (n = 4), cardiopulmonary arrest (n = 2), pulseless electrical activity (n = 1), multiorgan failure (n = 1), motor vehicle crash (n = 1), surgical complications (n = 1), cardiopulmonary arrest (n = 1), and septic shock (n = 1).

Eighty-nine patients (26.2%) had a recorded follow-up visit with a dermatologist after the AGEP course. Of the 291 cases in which medication was the suspected cause, 4 patients (1.4%) had records of subsequent patch testing.

Discussion

In this retrospective case series evaluation, we identified and characterized the experience of 340 patients with AGEP across 10 academic health systems in the US. The study population was predominantly female, consistent with previous studies.^3,5,9,10,11 A total of 52.4% of the patients developed AGEP while already admitted to the hospital; the rest developed AGEP in the community and then presented to an outpatient clinician and/or hospital. Ninety-seven patients (28.5%) had a history of diabetes, which is higher than the nationally reported prevalence among all adults (13%) but comparable to the prevalence among adults aged 65 years or older (26.8%), which likely better reflects our study population whose mean age was 58 years.¹² Similarly, our observed prevalence of chronic kidney disease (23.2%), although higher than the nationally reported prevalence of 7%, is comparable to the prevalence among adults aged 65 to 79 years (21.7%).¹³

The observed prevalence of personal history of psoriasis (7.3%) in this study was the same as that previously reported among patients with AGEP in the US and Europe.^9,10 Although AGEP and generalized pustular psoriasis are considered to be separate entities, despite their similarities in morphologic and histologic characteristics, associations with interleukin 36 receptor antagonist gene variations have been found for both conditions, a variation that could make patients susceptible to pustular eruptions.² Although this issue was beyond the scope of our study, a personal history of psoriasis was absent in most of this population.

Prior Drug Reaction History

Although previous studies have reported the prevalence of any and all previous unspecified drug reactions among AGEP cohorts to be between 8.3% and 86%,^7,9,14 we determined the prevalence of previous drug reactions specifically for hypersensitivity (urticaria, anaphylaxis, and angioedema) reaction (64 [18.8%]), morbilliform drug reaction (27 [7.9%]), Stevens-Johnson syndrome/toxic epidermal necrolysis (10 [2.9%]), drug reaction with eosinophilia and systemic symptoms (6 [1.8%]), fixed drug eruptions (0.9%), and lichenoid (0.6%) drug eruptions. The prevalence of previous hypersensitivity reactions among this population may be falsely increased, given that there is a low threshold for entering unverified drug allergies into the medical record based on patient-reported history.¹⁵

Hematologic and Physical Examination Findings

The prevalence of neutrophilia (85.1%) was consistent with previous studies (range, 88%-100%).^6,7,8,9,14 Mucosal involvement, mostly oral, was rare (5.9%), even more so than the previously reported prevalence (range 15.4%-25%).^4,6,8 Both of these findings support established AGEP diagnostic scoring criteria.¹

Eosinophilia, which is not part of established diagnostic scoring, was present in 52.1% of the study population and has been variably present in previous studies (range, 8%-83%).^6,7,9,14 Fever measured and recorded in the health care setting was prevalent in 49.7% of the study population, which is lower than previously published case series that reported nearly universal fever,^6,8,14 but greater than the prevalence reported by others.^4,9 This discrepancy could be due to variable inclusion of patient-reported fevers without documented temperatures, but overall, these findings suggest that development of fever during AGEP is variable, and its inclusion in diagnostic scoring systems should be re-evaluated with future prospective studies.

Kidney and Liver Involvement

Of 319 patients with at least 1 creatinine level recorded during their AGEP course, only 86 had a baseline creatinine level available from within 7 days before the AGEP start date, and 25 of those patients (29.1%) experienced a creatinine level increase of at least 1.5 times the baseline level. Therefore, of 319 patients with a recorded creatinine level, 25 (7.8%) experienced acute kidney insufficiency, which is slightly fewer than previously reported (10.3%).³

Regarding hepatic involvement, although 60 of 298 patients (20.1%) with available data had AST or ALT values twice the upper limit of normal, after excluding patients with missing baseline values or baseline values that indicated chronic transaminitis, our corrected frequency of acute transaminitis during the AGEP course was 25 of 298 patients (8.4%), which is consistent with previous results on a smaller scale.³ Our methods do not allow for determination of a causative association between AGEP and end-organ damage, as these abnormalities could be secondary to other comorbidities or iatrogenic causes, such as initial infection and subsequent antimicrobial therapy. These frequencies of acute kidney insufficiency and acute transaminitis are likely less than the true frequencies given our strict exclusion of patients without comparison baseline data, but even these conservative values demonstrate the need to monitor these patients for systemic complications.

Causative Agent

In line with previous studies, most AGEP cases were thought to be caused by a medication (85.6%). Data from cases in which the dermatology team was able to narrow the suspected agents to 1 medication administered before AGEP developed revealed most medications were started within a few days before the eruption (median, 3 [IQR, 1-9] days) and were largely antimicrobials (114 [75.4%]), with both of these results consistent with the findings of previous smaller scale retrospective and prospective studies.^{1,4,5,6,8,9,10,14}

The question of whether infections themselves can be a cause of AGEP has long been debated.² In our series of 340 patients, only 3 patients (0.9%) had an infection listed as the suspected cause by the dermatology team (2 with incomplete negative or negative results of serologic testing, and 1 with both blood- and wound-culture–positive results for Staphylococcus aureus). Although it is theoretically possible for a foreign pathogen to cause an immune response similar to that of a foreign substance, such as a medication or intravenous contrast agent^7,9 (suspected cause in 2.1% of our cases), it appears to be rare, and future studies will need to better elucidate this association, if one exists.

Treatment

Consistent with previous studies, AGEP had either improved or fully resolved in 95.4% of patients with available outcome data as of the last available note with the dermatology team, with the majority of patients being treated with topical corticosteroids (81.5%). Systemic treatment was used in 37.1% of cases, with corticosteroids (n = 109) and cyclosporine (n = 10) being the 2 most common medications. A recent study that likely included a portion of the patients from this cohort found similar outcomes between patients treated with either of these 2 systemic agents.¹⁶ There is no definitive end point of AGEP, and patients often have prolonged recovery after pustule formation ceases that usually includes desquamation (64.7% of patients in this study). Among the 105 patients for whom the dermatology team monitored and noted detailed physical examinations in the patient record, the median time for complete pustular resolution was 8 (IQR, 5-12) days, supporting the less than or equal to 15-day cutoff for pustular resolution included in the EuroSCAR criteria proposed by Sidoroff et al.¹

Mortality

Although the mortality rate of 5% for AGEP is commonly cited,² the origin of this data point is unclear.^1,17 In our cohort of 340 patients, mortality within 30 days was low (3.5%), and all cases of mortality were attributable to causes other than AGEP.

Limitations

This study has limitations. This was a retrospective study conducted across 10 different academic sites. The use of a standardized data entry form with standardized clinical definitions and inclusion criteria helps to prevent intersite variability, yet differences in EMR data availability may affect the information available for each patient. Initial case identification varied depending on the capability of each site, and there is the possibility for selection bias depending on how potential cases were initially identified. There are no standard diagnostic criteria for AGEP, and although the EuroSCAR scoring system has been applied in many previous retrospective studies, its implementation and interpretation are user dependent; thus, there is the possibility that cases were misclassified as either being or not being AGEP. The EuroSCAR criteria also prohibited the inclusion of acute localized exanthematous pustulosis, which some consider an emerging variant of AGEP.¹⁸ Lack of a control group limits the ability to identify an association between AGEP and any of the clinical and systemic findings observed during these patients’ AGEP course, as well as limits the ability to assess the risk of developing AGEP from exposure to suspected causative agents.

Conclusions

The results of this retrospective case series analysis of 340 patients with AGEP support many previous hypotheses surrounding the cutaneous adverse reaction. The onset is acute and usually triggered by a new antimicrobial, the discontinuation of which leads to improvement or complete resolution in most patients. The majority of patients experienced neutrophilia and eosinophilia, and about half developed a fever (temperature ≥38.0 °C). Kidney and liver complications may develop secondary to AGEP, yet these appear to be rare. In addition, no cases of patient mortality directly attributable to AGEP were encountered in the study population.

Supplement.

eTable 1. Study Definitions

eTable 2. Cases by Institution