Prevalence and Clinical Features of Drug Reactions With Eosinophilia and Systemic Symptoms Syndrome Caused by Antituberculosis Drugs: A Retrospective Cohort Study.

PURPOSE: Although there have been reported cases of drug reactions with eosinophilia and systemic symptoms (DRESS) syndrome caused by antituberculosis drugs, there has been no research to examine its prevalence. This study assessed the prevalence and clinical characteristics of DRESS syndrome caused by antituberculosis drugs.

METHODS: The electronic medical records of a cohort consisting of adult patients diagnosed with tuberculosis between July 2006 and June 2010 were reviewed and retrospectively inspected. We searched the surveillance system for adverse drug reactions and the electronic medical records to identify patients who reported severe cutaneous adverse reactions to antituberculosis drugs. These patients were then re-assessed using a European Registry of Severe Cutaneous Adverse Reactions to Drugs and Collection of Biological Samples (RegiSCAR) scoring system. Clinical characteristics, including the symptoms and latency of DRESS syndrome, the therapeutic dosage and period of steroids, and the final duration of tuberculosis therapy, were examined.

RESULTS: Of the 1,253 adult patients with tuberculosis receiving antituberculosis drugs, 15 were identified as potential cases of DRESS syndrome (prevalence of 1.2%). Ethambutol was the most frequently used drug (53.5%), followed by rifampicin (26.7%), pyrazinamide (20.0%), streptomycin (13.3%), and isoniazid (6.7%). The median latency after day 1 of antituberculosis medication was 42 days. The median daily dose of steroids, expressed in prednisone-equivalent units, was 33-mg/day, and the median dosing period was 14 days. The duration of tuberculosis treatment was 76 days longer than the standard treatment period of 180 days. There was a significant difference in the peak eosinophil counts of DRESS syndrome patients according to RegiSCAR scores. Moreover, there was a significant quantitative correlation between the RegiSCAR score and peak eosinophil count. A negative correlation was also found between the RegiSCAR score and latency.

CONCLUSIONS: This study confirmed the prevalence of DRESS syndrome in a cohort of adult patients with tuberculosis.