Characteristics and risk factors for drug-induced lung injury in hospitalized patients: A matched case-control study

Abstract

Abstract Objective: Many drugs can cause drug-induced lung injury (DLI), but its clinical features and pathogenic factors are not fully clear. The present study aims to explore the clinical characteristics and related risk factors of DLI and to provide a reference for the safe application of the drug. Materials and Methods: The clinical features of patients diagnosed with DLI in our hospital between 2014 and 2018 were retrospectively analyzed. DLI was diagnosed by using the Japanese Respiratory Society criteria. A retrospective case-control study was conducted according to suspected drugs, race, and significant diagnosis-matched controls, and logistic regression was used to identify risk factors. Results: 38 patients were diagnosed with DLI, and the median time from initiation of medication to the diagnosis of DLI was 74 days. The main clinical symptoms were shortness of breath (81.6%), cough (71.1%), expectoration (55.3%), and fever (42.1%); chest CT showed mainly ground glass opacity (GGO) (81.6%), nodular pattern (36.8%), increased of lung markings (34.2%) and interlobular septal thickening (23.7%). In laboratory tests, the levels of WBC, ALT, LDH, CRP, and KL-6 were significantly increased, while creatinine and albumin levels were significantly decreased. A total of 32 patients were treated with steroids. Of all the 38 patients, six were utterly cured, 27 showed improvement in their symptoms, two were poorly treated, and three died; In the matched case-control study, 38 DLI patients were successfully matched with 152 non-DLI controls. After multifactorial analysis, no factors were significantly associated with an increased incidence of DLI. Conclusions: The prognosis of DLI is poor, so it is necessary to be highly vigilant when using drugs that are easy to induce DLI. LDH and KL-6 may be useful biomarkers related to DLI. No significant risk factors for DLI were found in the present study, further high-quality original studies are needed to draw definitive conclusions.