Characterizing Granuloma Annulare in 73 Pediatric Patients

Background Granuloma annulare (GA) is a common, benign, idiopathic inflammatory dermatosis. Aside from case reports and small studies, there are limited data about the characteristics of GA in children. Objective This study aimed to better characterize the epidemiologic and clinical features, triggering factors, disease associations, and outcomes of GA in the pediatric population. Methods We conducted a retrospective study of 73 pediatric patients diagnosed with GA at the University of Rochester Medical Center over a 7-year period. Results The most common subtype was localized GA (71.2%, n = 52), followed by subcutaneous (also known as “deep GA”; 16.4%, n = 12) and generalized (12.3%, n = 9) subtypes. Over 90% of patients had idiopathic GA, with the remaining patients reporting viral infection or trauma as triggers. Half of the patients studied had comorbid conditions, most frequently atopic dermatitis (17.8%, n = 13), obesity (9.59%, n = 7), asthma (6.85%, n = 5), and allergic rhinitis (6.85%, n = 5). The median duration of the disease was 11.00 months (interquartile range (IQR) 15.75 months); generalized GA had the shortest duration (median 10.00 months, IQR 15.50 months), while subcutaneous GA had the longest duration (median 12.00 months and IQR 29.00 months). Although recurrence rates for subcutaneous and generalized GA were high at 45.5% and 33.3%, respectively, most patients achieved clearance or improvement with treatment. Conclusion Most cases of GA in our study were idiopathic, with no clear differences between GA subtypes and associated comorbidities. Topical steroids were the most prescribed treatment with mixed efficacy.


Introduction
Granuloma annulare (GA) is a benign infammatory dermatosis of unknown etiology.It is common in children and adults, with an estimated incidence of 0.04% in the United States [1].Several subtypes of GA exist, including the localized, generalized, subcutaneous, perforating, and patch variants, each with distinct clinical features [2].In the pediatric population, the localized and subcutaneous subtypes are the most common [3].
Several aspects of GA remain poorly understood.Its association with other diseases, for example, remains heavily debated.Te association with diabetes mellitus has been particularly controversial, with studies yielding conficting results [4][5][6].Very little data exist regarding associated diseases in pediatric patients or between various GA subtypes.Tere are also limited data on efcacious treatments for GA.Generalized GA, which has been historically difcult to manage, has had several proposed treatments with mixed results largely based on case reports and small retrospective studies [7].
Given the paucity of studies characterizing GA in the pediatric population, we are presenting this study to review the epidemiologic and clinical features, triggering factors, disease associations, and outcomes of GA in pediatric patients.

Study Design.
We conducted a retrospective singlecenter study at the Pediatric Dermatology Clinic in the University of Rochester Medical Center (URMC) Department of Dermatology in Rochester, NY.Tis study was approved by the Institutional Review Board (IRB) at the University of Rochester Medical Center.Inclusion criteria included age <18 years at the time of GA diagnosis, an ICD-10 diagnosis of "granuloma annulare" (L92.0) in the EPIC electronic medical record, and presentation to URMC between January 1, 2014, and December 20, 2020.

Data Collection and Analysis.
Once inclusion criteria were met, a retrospective electronic medical record chart review was conducted to obtain information on patient demographics (gender, ethnicity/race, age of GA onset, and GA evaluation), medical history, family history, known triggers, clinical and histological fndings, treatments used, and disease course.All patient data were deidentifed and recorded in a secure spreadsheet.Statistical analyses were performed using the GraphPad Prism Software Version 9.3.1.Descriptive statistics were used for demographic variables including sex, age, and race.Quantitative variables were expressed as mean ± standard deviation (SD), whereas qualitative variables were expressed in frequencies.Te duration was expressed as median ± interquartile range (IQR).

Results
Table 1 describes the demographics of our study cohort and clinical features of patients' disease.Our cohort consisted of 73 patients, most of whom were female (63.0%, n � 46) and Caucasian (53.4%, n � 39).Te mean age of GA onset was 6.21 years (standard deviation (SD) 4.34 years).
Localized GA commonly presented on the arms, legs, hands, or feet with multiple asymptomatic papules in an annular or circinate pattern mainly located on the extensor surfaces of the lower and upper extremities, with a minority of patients experiencing symptoms of pruritus (23.1%, n � 12) or pain (3.85%, n � 2) (Figure 1).Generalized GA had a similar morphology, but typically involved the trunk in addition to the extremities, and had associated pruritus in less than half of cases (44.4%, n � 4) (Figure 2).Finally, the subcutaneous subtype was commonly a solitary, frm, nonulcerated, nontender, and sometimes pinkish nodule.Te most common location was on the lower extremities, especially the pretibial area, followed by the hands (Figure 3(a)).Buttocks, scalp, and forehead were less commonly afected.Te nodules were symptomatic in 41.7% of the cases (n � 5), notable for symptoms of pruritus (33.3%, n � 4) or pain (8.33%, n � 1).
Te median duration of the disease was 11.00 months (interquartile range (IQR) 15.75 months).Generalized GA had the shortest duration (median 10.00 months and IQR 15.50 months), while subcutaneous GA had the longest duration (median 12.00 months and IQR 29.00 months).Topical steroids were the most common treatments for localized and generalized GA, used in 86.5% and 100.00% of cases, respectively.Topical immunomodulators (e.g., tacrolimus) were used in some cases of localized GA (9.62%, n � 5) and rarely in generalized and subcutaneous subtypes (11.1%, n � 1 and 8.33%, n � 1, respectively).Intralesional steroids were given to one of the nine patients in our cohort with generalized GA (11.1%), as well as to one patient with localized (1.92%) and one patient with subcutaneous (8.33%) GA subtypes.
Excision was utilized for 50% of cases of subcutaneous GA.Te decision to pursue excision was made either due to an unclear diagnosis and/or the lesion causing the patient discomfort.Subcutaneous and generalized GA had the highest recurrence rates (45.5%, range: 3 months to 5 years post-excision and 33.3%, range: 1 month to 17 months postinitial assessment, respectively).Approximately 9.59% (n = 7) of patients across subtypes had persistent disease despite treatment, and 17.8% (n = 13) of patients were lost to follow-up.

Discussion
Te study we present sheds further light on the growing body of knowledge regarding GA.Although GA is not uncommon, its etiology, presentation, and treatment are not entirely understood, particularly with regard to its rarer subtypes.Localized GA, the most common subtype, classically presents as fesh-colored or red papules that can coalesce into annular plaques [5].Generalized GA constitutes about 15% of cases and presents with lesions with similar morphology to the localized variant, but it is usually defned as >10 lesions, although others have defned it as the involvement of the extremities and trunk [2,8].Subcutaneous GA is characterized by an asymptomatic, expanding nodule most commonly seen on the lower extremities [9].Te patch and perforating variants are less common and were not observed in our cohort but are characterized by erythematous patches on the trunk and proximal extremities or 3-4 millimeter papules in annular confguration often with central umbilication, respectively [3,10].Histologically, conventional GA exhibits palisading 2 Dermatology Research and Practice histiocytes around degenerated collagen, with increased dermal mucin.Subcutaneous GA has histologic features indistinguishable from rheumatoid nodules, with histiocytes arranged around necrobiotic collagen in the deep dermis or subcutaneous fat (Figures 3(b) and 3(c)) [11].Te pathogenesis of GA remains unclear.One popular hypothesis is that of primary necrobiosis secondary to trauma [5,12].Tis may be mediated by the release of nonspecifc esterase, acid phosphatase, and lysosome by histiocytes.An alternative hypothesis is that of a delayedtype hypersensitivity reaction.It is believed that a T H 1 reaction stimulates macrophages to produce matrix metalloproteinases, leading to connective tissue degradation, which is consistent with histopathological fndings of focal   Dermatology Research and Practice collagen degradation [13].Previous studies have found cytokines produced by T-helper cells, including TGF-β and IL-2, in lesions of patients with GA which further supports this hypothesis; however, studies on immunocompromised patients with GA have shown that the pathogenesis of this disease may be more complex [5].It is unclear what proportion of GA patients have a preceding trigger.Prior reports have noted diverse triggers, including vaccines (e.g., hepatitis B, tetanus, and Bacillus Calmette-Guérin), viral infections (e.g., varicella, herpes zoster, and COVID-19), medications (e.g., allopurinol, amlodipine, and anti-TNF agents), trauma, arthropod bites, and malignancy [14].Although most of our cases were idiopathic, approximately 7% of patients endorsed triggers.Tese included viral infections (including Epstein-Barr virus) and trauma (including arthropod bites), consistent with data in existing literature [14].Te variety of triggers may refect GA's uncertain etiology, which could involve a type IV delayed-type hypersensitivity reaction to an unknown antigen [15].A highly debated topic is whether GA is associated with other conditions, such as metabolic, autoimmune, and neoplastic disorders.One of the largest recent case-control studies (n � 5137 patients with GA, 51169 controls) used ICD-10 codes to demonstrate an association between GA and hyperlipidemia (adjusted odds ratio (aOR) 1.15, 95% confdence interval (CI) 1.08-1.23),hypothyroidism (aOR 1.59, 95% CI 1.14-1.22),systemic lupus erythematosus (SLE) (aOR 3.06, 95% CI 1.86-5.01),rheumatoid arthritis (aOR 2.05, 95% CI 1.34-3.13),and diabetes (aOR 1.67, 95% CI 1.55-1.80);however, this study was largely based on the adult population [4].Although few large-scale pediatric GA studies exist, the cooccurrence of pediatric GA with type 1 diabetes mellitus, hematologic malignancy, and atopy has been documented in existing literature [3].A recent case series further highlighted diabetes mellitus as well as thyroid disease and hyperlipidemia as comorbidities of pediatric GA, which was observed in our cohort in small numbers [16].Te prevalence of diabetes, specifcally, in our cohort (2.74%, n � 2) was higher than that reported in youth in the general population (2.22 per 1,000 individuals) [17].However, the association between pediatric GA and diabetes mellitus remains heavily debated [4,6,7,18].We had several patients with manifestations of atopy, notably atopic dermatitis (n � 13), allergic rhinitis (n � 5), and asthma (n � 5), which is in line with a recent pediatric GA retrospective study in which 23 of 47 patients included had evidence of atopy [18].However, none of these atopic comorbidities were clearly associated with a particular GA subtype in our cohort.Furthermore, the relatively low sample size of our study and lack of a control group limit our ability to infer a clear association between GA and the comorbidities discussed.
Finally, the management of GA remains challenging.As seen in our study, steroids are often the most frequently prescribed treatment [19].Yet, their efcacy is mixed, with some studies demonstrating rapid clearance and others fnding improvement in only about half of GA patients [8,19].Treatment efcacy may be limited in part by the GA subtype.Whereas localized GA tends to be self-limited, other subtypes, particularly generalized GA, can have high recurrence rates.Many treatments have been investigated for generalized GA with inconsistent results and varied side efects, including topical and intralesional steroids, topical immunomodulators (tacrolimus), antimicrobials (doxycycline, dapsone, and hydroxychloroquine), phototherapy, TNF-alpha inhibitors, and oral vitamin E [19][20][21].Of the patients in our cohort who pursued these therapies, namely, intralesional steroids or topical immunomodulators, there was no clear association with a particular outcome.
Although there is a growing body of evidence regarding GA, much remains to be elucidated.Further large-scale, prospective studies should be conducted to further characterize the various GA subtypes, their management, and associated comorbid conditions.

Figure 1 :Figure 2 :Figure 3 :
Figure 1: Localized granuloma annulare presenting as erythematous to violaceous annular plaques on the foot (a), knee (b), and dorsal hand (c) on Fitzpatrick skin types IV through V.

Table 1 :
Patient and disease characteristics in pediatric GA (N � 73).