Incidence and clinical features of liver injury related to Kampo (Japanese herbal) medicine in 2,496 cases between 1979 and 1999: problems of the lymphocyte transformation test as a diagnostic method.

We retrospectively examined the summaries of all admission records of patients from 1979 to 1999 in our department, and selected for further study all liver injuries suspected of being related to Kampo medicines. Among 2,496 summaries, 30 summaries described liver disorders suspected of being related to Kampo medicines. Whether there was a causal relationship between the use of Kampo medication and the occurrence of liver injury was assessed according to the criteria described by Haller and Benowitz (2000), independently of the results of the lymphocyte transformation test (LTT). Among 30 events, we concluded that 9 were definitely unrelated, and 6 were probably unrelated to the use of Kampo medicines. Nine events (0.36% of 2,496 patient admissions and 0.06% of 14,616 outpatients) were considered possibly related, and only 6 events (0.24% of 2,496 patient admissions and 0.04% of 14,616 outpatients) were judged to be definitely or probably related to Kampo medicines. Low-grade eosinophilia was observed in a few patients of these "related" groups, and no fever or rash was observed in these "related" groups. Other clinical features, including type of liver injury, duration of Kampo medicine-use, recovery period and laboratory data, were not different from liver injuries associated with western drugs. Most patients in the definitely "unrelated" group were positive in the LTT for the suspect Kampo medicine, suggesting that the LTT may be unreliable for the diagnosis of Kampo-medicine-induced liver injury. From 1979 to 1999, our use of Kampo medicines to treat patients resulted in a low rate of liver injury and no fatalities.