To the Editor: Studies comparing the frequency of adverse cutaneous reactions (ACRs) secondary to hydroxychloroquine (HCQ) therapy inpatients with dermatomyositis (DM) and cutaneous lupus erythematosus (CLE) have shown mixed results. Pelle et al and Wolstencroft et al both found increased ACRs in patients with DM.1,2 In contrast, no differences were found by Gonzalez et al.3This study had 2 objectives: first, to evaluate if increased ACRs exist in patients with DM and, second, to identify specific factors (eg, history of atopy, other drug allergies, baseline eosinophilia) in patients who developed ACRs secondary to HCQ therapy that might enable us to predict the reaction, and thereby avoid it. For this latter purpose, we sought to increase our dataset by including patients with lichen planopilaris (LPP), as dermatologists frequently prescribe HCQ to these patients.

Based on ICD-10 codes, 1163 patients with DM seen at Mass General Brigham outpatient dermatology clinics between January 2000 and December 2019 were identified. Patients for inclusion had to have an M-33 code that was recorded at least twice. Individual medical chart reviews were performed to confirm diagnoses and to ascertain the use of HCQ.These inclusion criteria produced a dataset of Table I. Patient characteristics by disease type patients with DM who were prescribed 200-800 mg HCQ daily. Age and gender-matched CLE (n = 271) and LPP (n = 74) controls were then extracted in an identical manner for comparison in our dataset. After reviewing the dataset, these criteria yielded 150 patients with CLE and 36 patients with LPP. Only cases of ACRs diagnosed by a dermatologist and supported either clinically or histopathologically were included.The overall incidence of ACRs due to HCQ therapy was 4.2% (13/306) and the incidence was elevated inpatients with DM (P< .001, Table I). This finding of increased incidence of ACRs in patients with DM is consistent with the findings of Pelle et al and Wolstencroft et al. However, only 9.2% of patients with DM who developed ACRs underwent HCQ therapy, which was lower than that reported by Pelle et al (31%) and Wolstencroft et al (20.7%).1,2 There were no statistically significant differences in terms of daily HCQ dose based on actual covert hepatic encephalopathy or ideal body weight between patients with DM with HCQinduced ACRs and those of controls (P SC79 research buy = .58).

In contrast to chronic virus infection our hypothesis that patients with DM may also develop more non-HCQ drug allergies, non-HCQ drug allergies were more common in patients with CLE (P = .009). Specifically, hypersensitivity to sulfa was more common in patients with CLE (P = .004). Baseline eosinophilia was most frequently found in patients with LPP (P = .002). Of those who presented with ACRs in response to HCQ therapy (Table II), those with a higher mean age were more likely to develop HCQ-associated ACRs (56.1 versus 46.6, P = .048), but this significance was lost when multiple comparison adjustments were performed.

In patients with DM, those on immunosuppressive therapy were less likely to develop ACRs in response to HCQ therapy than those who were not (odds ratio, 0.18; P = .02). This is potentially clinically useful information because if immunosuppressant use is a protective factor, the addition of HCQ only after immunosuppression may help reduce the risk of HCQ allergy. In conclusion, we did not identify statistically significant risk factors in patients with DM that would allow prediction about which patients develop HCQ-induced ACRs. While larger than previous studies, this study is limited by the fact that it is a retrospective study and represents only a limited sample from a single institution.