A controlled clinical trial of daily and intermittent regimens of rifampicin plus ethambutol in the retreatment of patients with pulmonary tuberculosis in Hong Kong A Hong Kong tuberculosis treatment services/Brompton hospital/British medical research council investigation

Abstract

In a controlled clinical trial, 4 regimens of rifampicin plus ethambutol are being compared with the standard reserve regimen of ethionamide plus pyrazinamide plus cycloserine in the management of patients in Hong Kong with pulmonary tuberculosis whose first-line chemotherapy has failed. The results at 1 year are presented. A total of 575 Chinese adults with sputum-positive pulmonary tuberculosis, with strains resistant to isoniazid, who had previously been treated with first-line drugs but not with the drugs under study, were allocated at random to: 1. a. rifampicin plus ethambutol daily (ER7); 2. b. rifampicin plus ethambutol twice a week (ER2); 3. c. rifampicin plus ethambutol once a week (ER1); 4. d. rifampicin plus ethambutol daily for 2 months and then once a week (ER7ER1); 5. e. ethionamide plus pyrazinamide plus cycloserine daily for 6 months and then ethionamide plus pyrazinamide daily (EtZC). There were 384 patients (107 ER7, 89 ER2, 57 ER1, 69 ER7ER1, 62 EtZC) in the main comparison. The radiographic improvement and monthly sputum culture results were broadly similar in all 5 regimens. At 12 months 89 per cent of 107 ER7, 80 per cent of 89 ER2, 82 per cent of 57 ER1, 91 per cent of 69 ER7ER1 and 95 per cent of 62 EtZC patients had a favourable status. Bacteriological failure on the rifampicin regimens was associated with the early emergence of rifampicin resistance, there being evidence that the Hong Kong strains were less sensitive pretreatment to ethambutol than British strains were. Of 396 patients (91 ER7, 80 ER2, 78 ER1, 85 ER7ER1, 62 EtZC) in the analysis of adverse reactions, 2 of 91 ER7 patients, but no other patient on the other rifampicin regimens, had reactions to ethambutol. Nine (10 per cent) of the ER7 patients had reactions to rifampicin; 111 (46 per cent) of the 243 patients on intermittent rifampicin had reactions to rifampicin, and 34 (55 per cent) of the EtZC patients had adverse reactions. Reactions to daily rifampicin were either cutaneous or transient increases in serum transsaminase levels without jaundice. Six main types of adverse reaction to intermittent rifampicin occurred; cutaneous, abdominal, ‘flu’ syndrome, respiratory, hepatic and purpura. The ‘flu’ syndrome was the commonest, but was less common during twice-weekly treatment (22 per cent of 80 ER2 patients) than during once-weekly treatment (54 per cent of 85 ER1 patients) with the same dose size of rifampicin (P=00001), and less common during once-weekly rifampicin (35 per cent of 85 ER7ER1 patients) when this was preceded by a 2-month daily phase (P = 0.03). Most adverse reactions to intermittent rifampicin were relatively easy to manage clinically, and only 11 patients (5 per cent of 243 on intermittent rifampicin) had to have the drug terminated. Adverse reactions to the EtZC regimen (the commonest of which were cutaneous, hepatic, gastro-intestinal, neuropsychiatric reactions and arthralgia) led to the termination of one drug in 11 and all 3 drugs in 4 of 34 patients with reactions. Circulating rifampicin-dependent antibodies, detected by the indirect antiglobulin method, were found in 1 (2 per cent) of 47 ER7 patients, in 30 per cent of 40 ER2, 41 per cent of 44 ER1, and 32 per cent of 44 ER7ER1 patients and in none of 36 EtZC patients. They were significantly associated with the ‘flu’ syndrome (P=0.001). In the light of this study, twice-weekly rifampicin plus ethambutol, with or without an initial daily phase, is now in use as a reserve regimen in Hong Kong.