[Six-months chemotherapy (2HRZS or E/4HRE) of new cases of pulmonary tuberculosis--six year experiences on its effectiveness, toxicity, and acceptability].

The incidence of tuberculosis in Japan, 33.7 per 100,000 in 1997, is very high compared with USA or Western European countries. The decrease in the incidence has slowed down from the early 1980s, and the average annual rate of decrease has been 3.8% in the last 5 years. About 9 percent of tuberculosis patients defaulted from the nine-month regimen (6HRS or E/3HR) in urban areas. Regimens shorter than nine-month are needed to achieve high effectiveness of tuberculous chemotherapy. Out of 1128 new pulmonary tuberculosis patients, six-hundred twenty started treatment with six-month (2HRZS or E/4HRE) in Fukujuji Hospital, JATA, in Tokyo from January 1991 to December 1996. Out of 620, four-hundred twenty eight were both smear and culture positive, 136 were smear negative and culture positive and 56 were bacilli negative. Out of 564 bacilli positive cases, 530 were susceptible to INH and RFP. Out of 530 drug susceptible cases three hundred ninety-three patients completed the regimen. Ninety-three percent of these patients had converted to negative at two months of chemotherapy and all of them at five months. Out of 450, two-hundred ninety five completed 6-month regimen, one-hundred fifty-five were changed their regimen or prolonged duration of chemotherapy. Out of 295, nine patients (3.1%) relapsed after the completion of 6-month chemotherapy. Mean follow-up period was 17.2 months and the median was 15.5 months. The relapse rate was 2.2 per 100 person-years. Six of the relapsed cases were complicated with Diabetes Mellitus. Relapse rate was higher in patients with Diabetes Mellitus than in patients without (6/54, 7.9 per 100 person-years vs 3/237, 0.8 per 100 person-years) (p < 0.001). Drug-induced hepatotoxicity was defined as elevated serum transaminase level with clinical symptoms of hepatitis or elevated serum transaminase level more than 5 times of upper limit of normal range with or without symptoms. Drug-induced hepatotoxicity developed in 43 (8.0%) of 535 with initial normal liver function test results, this rate was similar to that in patients treated with nine-month regimen (34/420, 8.1%). But the frequency of hepatotoxicity of more than 400 IU/ml of serum transaminase level was higher in patients treated with PZA-containing regimen than with nine-month regimen (16/536, 3.0% vs 4/420, 1.0%), but this deference was not statistically significant. Hepatotoxicity developed in 13/85 (15.3%) of patients treated with PZA-containing regimen with abnormal liver function tests at the beginning of chemotherapy, and this frequency was similar to 7/65 (10.8%) in patients with nine-month regimen. The relapse rate in patients with Diabetes Mellitus was statistically higher than in without Diabetes Mellitus (7.9 vs 0.8 per 100 person-years). We concluded that the six-month regimen was highly effective, but the frequency of severe hepatotoxicity was relatively higher than in nine-month regimen and the duration of chemotherapy was not enough for patients complicated with Diabetes Mellitus. Further study is needed for sufficient chemotherapy in patients with Diabetes Mellitus.