Drug-induced erythroderma, also known as exfoliative dermatitis drug eruption, is a type of severe drug eruption and manifested as diffuse erythema and desquamation on the body surface area (BSA). This is a serious and potentially life-threatening systemic inflammatory skin disease, which progresses rapidly and has a long treatment cycle with mortality rate of 10%-20%.^[1-3] This study aims to retrospectively analyze hospitalized patients with acquired immunodeficiency syndrome (AIDS) suffering from drug-induced erythroderma from January 2013 to December 2020 in the First Affiliated Hospital of Zhejiang University School of Medicine.

The clinical data of AIDS patients with drug-induced erythroderma were retrospectively analyzed. Patients met the criteria for drug-induced erythroderma.^[1] Medical records were reviewed by at least two experts. The patients were excluded if erythroderma was caused by drugs, tumors, or psoriasis.

The clinical data including age, gender, medications, allergies, drug eruption, modified severity-of-illness score for toxic epidermal necrolysis (SCORTEN) scores, treatments, time to improvement (no new rash, normal body temperature), and length of hospital stay were collected.

The modified SCOTERN scores^[4,5] were recorded within 24 hours of admission based on seven clinical indicators: age >40 years, malignant tumors, heart rate >120 beats/minute, blood sugar >14 mmol/L, bicarbonate <20 mmol/L, blood urea nitrogen >10 mmol/L, and affected BSA >40%; 1 point for each item, 0 point for without any clinical indicators.

Statistical analysis was performed with SPSS (SPSS Inc., USA). Continuous data were presented as mean±standard deviation (SD) for normal distribution and as median (range) for non-normal distribution. Categorical data were recorded as frequency and percentage.

Twelve AIDS patients with drug-induced erythroderma were included in the study. The average age was 34 years (range 23-83 years): seven (58.3%) patients were ≤40 years, and five (41.7%) patients >40 years. There were 11 males and one female.

The drugs that caused drug-induced erythroderma were: antiviral drugs (n=4), antituberculosis drugs (n=2), compound sulfamethoxazole (n=1), carbamazepine (n=1), Chinese medicine (n=1), antibiotics (amoxicillin) (n=1), immune checkpoint inhibitor (PD-1 monoclonal antibody) (n=1), and thalidomide (n=1).

Eight patients had infections before onset of AIDS: pityrosporum infection (n=3), tuberculosis infection (n=2), cyanobacteria marneffei infection (n=1), and combined more than two infections (n=2). Three patients had tumor: lung cancer (n=1), pancreatic cancer (n=1), and hepatocellular carcinoma (n=1).

The seven clinical indicators within 24 hours of admission were: age >40 years (n=5), malignant tumors (n=3), heart rate >120 beats/minute (n=5), blood sugar >14 mmol/L (n=5), bicarbonate <20 mmol/L (n=5), blood urea nitrogen >10 mmol/L (n=4), and affected BSA >40% in all patients. The modified SCOTERN score for all patients averaged 3.01±0.99.

All patients were treated with glucocorticoids, with simultaneous intravenous immunoglobulin (IVIG) pulse therapy in nine patients. Three patients were not treated with IVIG due to renal dysfunction or economic reasons. During IVIG treatment, no obvious adverse reactions were observed. The average time to effectiveness was 7.08±2.23 days, and the average length of hospital stay was 17.92±8.46 days. Among the 12 patients, 11 were cured, and one died of secondary multiple infections, whose modified SCORTEN score was 5 points. The mortality rate was 8.3%.

Erythroderma is a severe drug eruption. The clinical manifestations are diffuse erythema, swelling, and massive desquamation on the BSA, often accompanied with high fever, enlarged lymph nodes, liver and kidney injuries, and hypoproteinemia. The average age of onset is (42.1±20.6) years,^[6,7,8,9] and the average age of the patients in this group was 34 years (range 23-83 years). Compared with previous data, the age of patients in this group was younger, and this may be related with the fact that human immunodeficiency virus (HIV) patients were complicated with erythroderma. AIDS patients are usually young. Compared with the general population, HIV patients have impaired immune function, and are therefore susceptible to Epstein-Barr virus (EBV) and other viral infections. They are routinely treated with antiviral drugs.

It has been reported that the drug eruption rate in HIV-positive patients is 100 times that of HIV-negative patients.^[10,11,12] However, AIDS patients with drug-induced erythroderma is rare, and the mechanism is still unclear.^[13] Common agents that cause drug-induced erythroderma in HIV-negative persons include carbamazepine antiepileptic drugs, sulfa antibiotics, penicillins, and cephalosporins.^[1] The most used drugs that caused erythroderma in this study were antiviral drugs (33.4%), which may be related to the current need for long-term highly active antiretroviral treatment for AIDS. In antiviral therapy, non-nucleoside reverse transcriptase inhibitors nevirapine and efavirenz can cause severe skin eruption.^[10,11] Anti-tuberculosis drugs are the second most common cause of drug-induced erythroderma in our study. It is suggested that AIDS patients should be tested for tuberculosis infection early, and skin changes should be closely observed during the course of anti-tuberculosis treatment.

Drug-induced erythroderma is dangerous and has a high mortality rate.^[1] Therefore, proper evaluation of the patients is important for guiding treatment and prognosis. The modified SCORTEN scoring system has been used in recent years to evaluate severe bullous drug eruption.^[13-15] Among the seven indicators, age, malignant tumors, blood sugar, bicarbonate, blood urea nitrogen, affected BSA, and heart rate are related to the severity of the underlying diseases, which suggests that patients’ condition and underlying diseases are important. The disease severity can be assessed at an early stage.^[4,5] In lack of objective evaluation indicators for drug-induced erythroderma,^[16,17,18] we used this scoring system to assess the severity of drug-induced erythroderma. In this study, the modified SCORTEN score of the fatal case was 5 points, which was higher than the average SCORTEN score (3.01±0.99 points). This indicates that the modified SCORTEN may be a feasible tool to predict the severity and prognosis of drug-induced erythroderma.

Although biologic agents have been safe and effective in the treatment of severe drug eruptions in recent years, there is no evidence-based case report of biologic agents for the treatment of AIDS patients with severe drug-induced erythroderma. It is currently believed that glucocorticoids are still the first-line treatment for AIDS with drug-induced erythroderma. Immunocompromised HIV patients have a high incidence of co-infection, and glucocorticoids may enhance virus reactivation and even increase the risk of spreading infection.^[8] In our group of patients, glucocorticoid therapy was used, and good curative effects were achieved, which may be related to the co-administration of IVIG therapy at the early stage. Its mechanism of action is to inhibit antibody production, accelerate antibody metabolism, neutralize autoantibodies and complements, interfere with antibody-dependent cell-mediated cytotoxicity, affect T-cell activation, restore Th1/Th2 cell balance, inhibit cell adhesion, regulate cell proliferation and apoptosis, and affect glucocorticoid receptor sensitivity.^[19,20,21] We observed no adverse reactions caused by IVIG. We think that glucocorticoids and IVIG may have a synergistic effect in the treatment of this disease.

AIDS patients with drug-induced erythroderma may have high modified SCORTEN scores, severe illness, and complications, including multiple and severe infections. In AIDS patients with drug-induced erythroderma, glucocorticoid combined with IVIG should be used. However, due to the limited number of included patients, this method needs to be investigated further in studies with a larger sample size to observe the efficacy of IVIG in this disease.

We are grateful to Shu-juan Song for her contribution to part of the study.

Funding: The study was supported by National Natural Science Foundation of China (81972931).

Ethical approval: This study was approved by the Ethics Committee of the First Affiliated Hospital of Zhejiang University School of Medicine (2019-0022).

Conflicts of interests: No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

Contributors: WFZ, DRF, and HF contributed to the data collection, statistical analysis, and interpretation of the results. WFZ wrote the first draft. All authors have approved the final version of the manuscript.