Objective To explore the application and efficacy of glucocorticoids in severe non-antineoplastic drug-induced liver injury (DILI).

Methods A retrospective study was conducted on patients with severe non-antineoplastic DILI who were hospitalized at Zhongshan Hospital, Fudan University from November 2016 to October 2025. Clinical data, laboratory results, suspected causative drugs, glucocorticoid use, and prognosis were collected. The time for peak total bilirubin decreased by 50% was used as an efficacy indicator. Patients were divided into a glucocorticoid group and a non-glucocorticoid group based on glucocorticoid use.

Results A total of 102 patients with severe DILI were included, and 58 (56.9%) received glucocorticoid therapy. The most common suspected causative drugs were traditional Chinese medicine (65.7%), and the primary type of liver injury was hepatocellular (72.5%). The proportion of female patients (P＝0.030), baseline peak total bilirubin (P＜0.001), and international normalized ratio (P＝0.034) were significantly higher in the glucocorticoid group than in the non-glucocorticoid group. The median time for peak total bilirubin decreased by 50% was 9.5 days in the glucocorticoid group and 11 days in the non-glucocorticoid group, with no statistically significant difference between the two groups. No significant benefit of glucocorticoid therapy was observed in subgroups with positive autoantibodies, elevated immunoglobulin G (IgG), or elevated γ-globulin ratio. Low-dose and gradual tapering were the main methods of glucocorticoid administration. Adverse reactions occurred in 8 patients (13.8%) in the glucocorticoid group, which resolved after drug withdrawal or symptomatic treatment.

Conclusions The usage rate of glucocorticoids is relatively high in patients with severe non-antineoplastic DILI. Glucocorticoid treatment may accelerate the improvement of cholestasis in these patients, with mild adverse reactions and good safety.