Abstract: Objective  To evaluate the efficacy and safety of new chemotherapeutic regimens for retreatment pulmonary tuberculosis. Methods  A total of 345 cases were enrolled，according to the results of drug susceptibility testing, 161 cases with drug-sensitive results were assigned to a test group（85 cases） and a control group (76 cases),184 with drug-resistant results assigned to the test group（124 cases） and the control group (60 cases). The chemotherapeutic regimens were as follows：(1)For the drug-sensitive group, new chemotherapeutic regimen with 2HRZES/6-10HRE was used in the test group and primary regimen with 2H₃R₃Z₃E₃S₃/6H₃R₃E₃  was used in the control, respectively(8 months for PTB but no diabete cases,12 months for PTB with diabete cases). (2)For the drug-resistant group, 3R（H）ZES±Lfx/6-9R(H)ZE±Lfx was used in the test group. R or H could be replaced each other for R or H resistant patients. Am could be used in the case of S resistance. Am and Lfx could be used in HS or RS resistant cases. 2-3HRZES/6HRE regimen was used in the control(We selected the control subjects in the same place, same situation but historical controls). Results  (1)In drug sensitive group, the sputum conversion rate was 77.6%(66/85) and 88.2%（67/76）, radiographic improvement rate was 63.5%(54/85) and 55.3%（42/76）, effective rate was 96.5%(82/85) and 90.8%（69/76）and cavity closing rate was 68.3%(28/41) and 56.1%（23/41）,there was no statistically significant difference between the test group and the control group（χ²＝2.397 24，χ²＝0.821 42，χ²＝1.354 777，χ²＝0.829 58, P＞0.05）. Also in drug sensitive group complicated with diabetic, the sputum conversion rate was 66.7%（16/24） and 73.3%（11/15）, radiographic improvement rate was 58.3%（14/24） and 46.7%（7/15）,there was no statistically significant difference between the test group and control group complicated with diabetic（χ²＝0.191，χ²＝0.511，P＞0.05）. Meanwhile, the sputum conversion rate was 82.0%（50/61） and 91.8%（56/61）, radiographic improvement rate was 65.6%（40/61） and 57.4%（35/61）in the test group but no diabetes and control group but no diabetes. There was no statistically significant difference between the two groups（χ²＝2.589 62，χ²＝0.861，P＞0.05）.（2）In the drug resistant group,  at the end of treatment, the radiographic improvement rate was 79.8%（99/124）in the test group and 66.7%(40/60) in control. There were statistically significant difference between the test group and control （χ²＝3.81, P＜0.05）;Radiographic improvement rate was 62.9%(78/124) in test group and 45.0%(27/60) in the control, there were statistically significant difference between the test group  and the control（χ²＝4.584 29, P＜0.05）; The Incidence of adverse reactions were 1.1%（1/92） and 1.2%(1/82) in the test and control groups in drug sensitive group, respectively, and the Incidence of adverse reaction in the test group of drug-resistant cases was 2.8％（4/142）. Conclusion  For retreatment drug-sensitive cases, whether or not complicated with diabetes, can receive the primary chemotherapeutic regimen. New chemotherapeutic regimens had got good effect in the retreatment drug-resistant cases and the Incidence of adverse reactions were lower and no sever adverse reactions developed.

Key words: Tuberculosis,pulmonary/drug therapy, Recurrence, Clinical protocols, Controlled clinical trial, Diabetes mellitus