[Follow-up study of short course chemotherapy of pulmonary tuberculosis complicated with diabetes mellitus].

A retrospective analysis was made for 644 patients with pulmonary tuberculosis newly diagnosed during the time of 1977 to 1985 to evaluate the influence of diabetes mellitus (DM) on the drug response rate and the long-term relapse rate in the treatment of tuberculosis. These patients were divided into four groups: (1) 123 patients with DM on 9- to 12-month short course regimens; (2) 79 with DM on 13- to 36-month long-term regimens; (3) 379 nondiabetic patients on short-term regimens; and (4) 63 nondiabetics on long-term regimens. Bacteriological relapse after chemotherapy was defined as positive cultures growing at least 20 or more colonies. 1) Bacteriological negative conversion rates were similar in both diabetic and nondiabetic patients who had received combination regimens including INH and RFP. The degree of control of DM did not affect the conversion rate. 2) Of 297 patients who had received short-course chemotherapy and the information for analysis were available as of August 1989, 8 (10.3%) had relapses in 78 diabetic, and patients 23 (10.5%) in 219 non-diabetic patients; the difference was not statistically significant. There were also no discernible differences in the relapse rates between patients on the short-course regimens and those on the long-term regimens. 3) Most of the relapses occurred around 6 months and 30 months after completing the short-course chemotherapy. Similar pattern of relapses was observed also in the long-term therapy group. 4) Pretreatment radiographic findings and quantity of the acid-fast bacilli in the sputum, and the presence of cavitary lesions at the completion of therapy bore no significant relation to the development of relapse. 5) In the diabetic patients the degree of control of DM contributed little to the development of recurrence. 6) Although most of the patients without DM relapsed with sensitive strains and achieved a good response to retreatment, diabetic patients frequently relapsed with resistant strains and had a grave prognosis. 7) Of 41 patients who died after having been on the short-course regimens, two were attributed to tuberculosis. Only one was attributed to tuberculosis of 25 patients who died after receiving the long-term regimens. 8) The data obtained here confirmed that the 12-month regimen for diabetic patients could achieve favorable results in the response rate and the long-term relapse rate, as that for nondiabetic patients. However the cases of tuberculosis complicated with DM frequently showed a poor prognosis once the relapse took place. These results suggest that pyrazinamide-containing 4-drug combination regimens in an initial intensive phase is the preferred treatment for the patients with DM.