Embed Size (px)
Citation preview
8213
Saturday 24 January 1981
CONTROLLED TRIAL OF FOUR THRICE-WEEKLY REGIMENS AND A DAILY REGIMENALL GIVEN FOR 6 MONTHS FOR PULMONARY
TUBERCULOSIS
HONG KONG CHEST SERVICE/BRITISH MEDICALRESEARCH COUNCIL
Summary Five 6-month antituberculosis regimens,allocated at random to patients with acid-fast
bacilli in their sputum on microscopy, were studied. Four,given three times a week from the start, contained isoniazidand rifampicin together with (1) streptomycin, pyrazinamide,and ethambutol, (2) streptomycin and pyrazinamide, (3)streptomycin and ethambutol, or (4) pyrazinamide andethambutol. The fifth was daily isoniazid, rifampicin, pyra-zinamide, and ethambutol. All 833 patients with drug-sensi-tive strains of bacilli before treatment had a favourable bac-
teriological response during chemotherapy, and the bacterio-logical relapse rates during 12 months after stopping chemo-therapy were 2% or less for all regimens except thrice-weeklyisoniazid, rifampicin, streptomycin, and ethambutol (theonly regimen without pyrazinamide), which had a relapserate of 8%. The results were equally good for the 138 patientswith bacilli resistant to isoniazid, streptomycin, or both drugsinitially. The incidence of potentially serious toxicity waslow. The daily regimen is relevant to programmes in whichpatients self-administer their drugs, and the 3 pyrazinamide-containing intermittent regimens are relevant to fully super-vised outpatient programmes.
Introduction
THE recommended treatment for pulmonary tuberculosisin Britain is daily isoniazid and rifampicin for 9 months, withethambutol for the first 2 months.’ However, this is an ex-pensive regimen, and there is no certainty that patients taketheir prescribed drugs. A study in Singapore has shown thatthe total duration could be reduced to 6 months when pyra-zinamide and streptomycin were given instead of etham-butol.2 The policy of the tuberculosis services in urban areasof many countries, including Hong Kong, is to give everydose of chemotherapy under full supervision and predomi-nantly on an outpatient basis. The present study was there-fore designed to explore completely intermittent 6-monthregimens, which require fewer attendances at treatment
centres than daily supervised regimens, and which are alsoless costly and less likely to cause toxicity.In a previous study in Hong Kong,3 a fully intermittent
6-month regimen of streptomycin, isoniazid, rifampicin, andpyrazinamide three times a week for 4 months, followed bystreptomycin, isoniazid and pyrazinamide twice a week, had a
bacteriological relapse rate of 6% in patients with drug-sensi-tive strains of tubercle bacilli before treatment, but a higherrate in patients with strains resistant to both isoniazid andstreptomycin. In planning the present study, it was arguedthat maintaining the frequency of drug administration atthree times a week and continuing rifampicin administrationfor the whole 6 months would substantially strengthen theregimen. Further, in order to study the relative importance ofpyrazinamide, streptomycin, and ethambutol, four intermit-tent regimens were compared, each containing all three oronly two of these drugs in addition to isoniazid and rif-ampicin. A daily regimen of the four orally administereddrugs was included as a control. The evidence from earlierstudies was that streptomycin would make at most a smallcontribution 4,5 and that ethambutol was unlikely to have anymeasurable effect, except for patients with strains initially re-sistant to both isoniazid and streptomycin. 3
Materials and Methods
Patients and Treatment RegimensChinese patients aged 15 years or more, with pulmonary tubercu-
losis, positive for acid-fast bacilli on sputum microscopy, and ahistory of no previous antituberculosis chemotherapy or no morethan 2 weeks during the last year, and no more than 4 months in all,were eligible for admission to the study.They were allocated at random to the following 6-month treat-
ment regimens:1. HRSZE3 isoniazid, rifampicin, streptomycin, pyrazinamide, andethambutol three times a week for 6 months.2. HRSZ3 the same as regimen 1, but without ethambutol.3. HRSE3 the same as regimen 1, but without pyrazinamide.4. HRZE3 the same as regimen 1, but without streptomycin.5. HRZE7 the same as regimen 4, but given daily instead of 3 times aweek.The dosages of drugs given intermittently were: isoniazid 15
mg/kg, rifampicin 600 mg, streptomycin 1 g, pyrazinamide 2 - 0 or2 - 5 g, and ethambutol 30 mg/kg. The daily dosages were: isoniazid300 mg, rifampicin 450 or 600 mg, pyrazinamide 1’ 5 or 2 - 0 g, andethambutol 25 mg/kg for 2 months and then 15 mg/kg. The higherdosages of rifampicin and pyrazinamide were given to patientsweighing 50 kg or more.
In all 5 regimens, the drugs were given together. Patients receivedevery dose under the direct supervision of outpatient clinic orhospital staff. Of the 1027 patients who started treatment 368 (36%)were initially admitted to hospital, 121 of them for 4 weeks or less.
Assessment of ProgressThree sputum specimens were examined bacteriologically before
treatment; 1 specimen was examined at monthly intervals up to 18months, an additional specimen being examined at 4, 5, and 6months. All were examined by smear and culture; positive cultureswere identified and tested for drug sensitivity. 6A posteroanterior chest radiograph was obtained before treat-
ment. The serum alanine aminotransferase concentration was
172
measured before treatment, and at 1, 2, 3, 6, and 7 months. Inthe HRSE3 and HRZE3 series, serum was tested for rifampi-cin-dependent antibodies at 4 and 6 months. Urine tests for isoni-azid8 were done without warning on a random sample of patientsduring chemotherapy, and on all patients at 7 and 9 months.
Definitions of Bacterial Drug Resistance
Drug resistance was defined as growth (20 colonies or more) in atleast 1 culture on the following or higher drug concentrations:isoniazid 0 - 2 mg/l, rifampicin 32 mg/1, and pyrazinamide 50 mg/l (atthe selected pH9); streptomycin resistance was defined as a
geometric mean minimum inhibitory concentration of 32 mg/l ormore.
Population StudiedA total of 1207 patients (244 HRSZE3, 243 HRSZ3, 239 HRSE3,
242 HRZE3, 239 HRZE7) were admitted, of whom 89 wereexcluded from the bacteriological analyses for reasons presentbefore treatment-for 66 all the pretreatment cultures were negativeand from 4 only mycobacteria other than Mycobacterium tuberculosiswere isolated; 18 had received excess previous chemotherapy, and 1had extrapulmonary tuberculosis. Another 143 were excluded forreasons arising during chemotherapy; 62 patients missed at least 6weeks of the allocated regimen because of drug toxicity, and 57 (11HRSZE3, 15 HRSZ3, 11 HRSE3,14 HRZE3, 6 HRZE7) because ofdefault; 9 had treatment changed because of pregnancy; 8 died fromcauses unrelated to their tuberculosis, and 1 (HRZE7) fromtuberculosis, but on the 17th day of treatment; the remaining 6 wereexcluded for miscellaneous reasons.975 patients remained available for the bacteriological
analyses-833 (160 HRSZE3, 161 HRSZ3, 170 HRSE3, 174
HRZE3, 168 HRZE7) with drug-sensitive strains ofM. tuberculosisbefore treatment, and 142 (34 HRSZE3, 22 HRSZ3, 29 HRSE3, 25HRZE3, 32 HRZE7) with strains resistant to 1 or more of the drugsisoniazid, streptomycin, and rifampicin. Of the 975, only 64 withsensitive strains and 13 with resistant strains reported havingreceived any previous chemotherapy.
Condition on Admission
All patients in the bacteriological assessments had positivesputum smears and cultures before treatment. For patients withsensitive strains, pretreatment distributions (not tabulated here) ofthe following factors were similar for all 5 regimens: sex, 72% male(range 69-75%); age, 32% more than 45 years (range 29-36%) and35% less than 25 years (range 32-39%); and weight, mean 48’ 0 kg(SD==7-2 2 kg). On a single posteroanterior chest radiograph, theextent ofdisease 10 was slight for 41 %, limited for 32%, moderate for17%, extensive for 9%, and gross for 1%; and the extent of obviouscavitation I was nil for 73%, slight for 16%, moderate for 10%, andextensive for 1%.
Results
Patients with Drugsensitive Strains before TreatmentBacteriological response during chemotherapy.-The pro-
portions of patients with negative cultures at months 1 and 2are shown in table I. At 2 months, the difference between theregimens was highly significant (p«0 00 1, 4df), this beingdue to the difference between the HRSE3 regimen and the 4pyrazinamide-containing regimens. By 3 months, 97% ormore of the patients on all 5 regimens had negative cultures.There were no bacteriological failures during chemotherapy.
Bacteriological relapse after chemotherapy.-Bacteriologicalrelapse after the end of chemotherapy was defined as a culturegrowing 10 or more colonies ofM. tuberculosis at 2 out of any3 consecutive monthly assessments. The bacteriologicalrelapse rates (table II) were 1% or 2% for the regimens con-taining pyrazinamide, but 8% for the HRSE3 regimen(p«0’001, 4 df). Of the total of 20 bacteriological relapses,17 started during the first 6 months after the end of chemo-
TABLE I-CULTURE RESULTS AT 1 AND 2 MONTHS FOR PATIENTS
WITH DRUG-SENSITIVE STRAINS BEFORE TREATMENT
*H = isoniazid; R = rifampicin; S = streptomycin; Z = pyrazinamide;E=ethambutol; 3=three times a week; 7=daily.
therapy and all 20 were with strains of tubercle bacilli whichwere still drug-sensitive. One other HRSZ3 patient, with sili-cosis, who did not relapse bacteriologically, had retreatmentstarted in the 9th month after the end of chemotherapybecause of radiographic deterioration.
Patients with Drug resistant Strains before Treatment
All of the 142 patients assessed who had drug-resistantstrains before treatment (table III) responded satisfactorily tochemotherapy except 1 HRZE3 patient, with a streptomycin-resistant strain, who died during the 5th month, having hadpositive sputum cultures at 1 to 4 months, in which no addi-tional drug resistance emerged. During the 12 months afterthe end of chemotherapy, of 104 patients treated with a pyra-zinamide-containing regimen, 3 (3%) relapsed bacteriologi-cally; 2 had strains resistant to isoniazid, streptomycin, andrifampicin before treatment, and the 3rd to isoniazid alone, Incontrast, of the 27 HRSE3 patients, 7 (26%) relapsed bac-teriologically ; only 2 had strains resistant to both isoniazidand streptomycin before treatment, the remainder havingstrains resistant to a single drug. Only 1 patient (HRSE3)1with strains resistant to isoniazid and streptomycin, acquiredadditional resistance, namely to rifampicin. The differencebetween the pyrazinamide-containing regimens and the
HRSE3 regimen for patients with strains resistant to
isoniazid, streptomycin, or both these drugs (that is excludingthe 4 patients with rifampicin-resistant strains) was highlysignificant (p <0 001, Cochran’s test).
Urine Tests for Isoniazid
Of the 321 surprise urine tests for isoniazid during chemo-therapy on 283 patients, 297 (92 5%) were positive. A furtherspecimen was tested in 19 of the 23 patients with a nega-tive result, 17 of these were positive. Of the 1886 tests done on
TABLE II-BACTERIOLOGICAL RELAPSES DURING 12 MONTHS AFTER
THE END OF CHEMOTHERAPY IN PATIENTS WITH DRUG-SENSITIVE
STRAINS BEFORE TREATMENT
*See footnote to table I
173
TABLE III-BACTERIOLOGICAL RESULTS FOR PATIENTS WITH
DRUG-RESISTANT STRAINS BEFORE TREATMENT
* See footnote to table I
’jAIl had a favourable response except 1 HRZE3 patient with pretreatmentresistance to streptomycin who died in the 5th month.
973 patients after the end of chemotherapy, 1880 (99 - 7%)were negative.
Adverse Reactions
Among the 1205 patients who started on their allocatedchemotherapy (table IV), adverse reactions were reportedmost frequently by those on the two regimens which con-tained both streptomycin and pyrazinamide, and least
frequently by those on the HRSE3 regimen, the only onewhich did not include pyrazinamide. However, only a smallproportion ofpatients had 1 or more of their drugs withdrawnbecause of reactions. The drugs withdrawn were isoniazid for9 patients, rifampicin for 8, streptomycin for 19, pyrazina-mide for 20, and ethambutol for 9. Of the 50 patients who hadI or more drugs terminated, 41 had 1 drug terminated, 6 had2 drugs, and 3 had 4 drugs. A high proportion of thesereactions were mild gastrointestinal episodes, many of whichwere not due to toxicity, but consisted of nausea occasionedby swallowing a large number and bulk of tablets. Cutaneousand vestibular reactions (most of them mild) accounted formost of the other reported reactions.
TABLE IV-POSSIBLE ADVERSE REACTIONS AND THEIR MANAGEMENT
* See footnote to table I
Hepatic reactions, defined as abnormalities in liver func-tion, with or without symptoms, leading to a modification tothe regimen, were commonest in patients on the dailyregimen, and did not occur with the HRSE3 regimen. Of thetotal of 21 hepatic reactions, 16 were syrhtomlessabnormalities in liver function, and only 3 patients (1HRSZE3, 2 HRZE7) had jaundice, in months 1, 1, and 2,respectively; 1 (HRZE7) of these, who was also being treatedwith ampicillin for an upper respiratory tract infection,developed exfoliative dermatitis with hepatitis at the end ofthe first month, and died in hepatic coma 43 days afteradmission to the study, the antituberculosis drugs havingbeen withdrawn 11 days before death. 15 of the 21 patientshad their chemotherapy resumed uneventfully after an
interruption, once their liver function tests had returned tonormal.2 patients (HRSZE3, HRZE7) had thrombocytopenic
purpura, this being associated with hepatitis in the patient onthe daily regimen. Rifampicin was withdrawn in both cases.Only 4 patients (2 HRSZ3, 2 HRZE3) had episodes of the"flu" syndrome,12 all mild. Of the 199 HRSE3 and 201HRZE7 patients tested at 4 and 6 months, only 1 and 2,respectively, had rifampicin-dependent , antibodies, allwithout symptoms. 3 patients (2 HRSZE3, 1 HRSE3) hadethambutol withdrawn because of transient visual changes(reduced acuity in 2, and altered colour perception in 1).
Discussion
A study in Singapore2 produced a very promising 6-monthdaily regimen for smear-positive pulmonary tuberculosis ofisoniazid and rifampicin, with streptomycin and pyrazina-mide for the first 2 months; the effectiveness of this regimenhas recently been confirmed by the British Thoracic Associ-ation.13 The present study has produced four more 6-monthregimens each containing isoniazid, rifampicin, and pyra-zinamide, and either one or both of the drugs streptomycinand ethambutol, which are highly effective not only forpatients with drug-sensitive tubercle bacilli before treatment,but also for those with strains initially resistant to isoniazid,streptomycin, or both these drugs. One of the regimens is adaily regimen of orally administered drugs, relevant to pro-grammes in which patients self-administer their drugs. Theother three are given three times a week from the start andtherefore are particularly relevant to programmes in-whichthe policy is to give every dose of chemotherapy fully super-vised on an entirely outpatient basis. Indeed, a fullyintermittent three-times-weekly regimen is now underassessment on a service basis in the urban areas of HongKong. Such assessments are essential to find out whethershort-course regimens which are found to be highly effectivein controlled clinical trials are also effective under
programme conditions. 14In the present study, the high relapse rate which followed
the intermittent regimen of rifampicin, isoniazid, strepto-mycin, and ethambutol, without pyrazinamide, confirms theimportance of pyrazinamide in short-course chemotherapyand the very small contribution of ethambutol and strepto-mycin.3-5,14 The reason for including ethambutol was that itmight have prevented bacteriological failure associated withthe emergence of additional drug resistance in patients withpretreatment strains resistant to both izoniazid and strepto-mycin, an important problem in Hong Kong. In the event, itappears to have been unnecessary. The limited role ofstrepto-mycin, even when given daily for 6 months as a supplement to
174
isoniazid and rifampicin,4 and the present findings, suggestthat it too contributes little or nothing as a sterilising drug(though it may contribute to the prevention of drug resist-ance) when added to the combination of isoniazid, rifampicinand pyrazinamide, an important possibility which is underfurther investigation in Singapore.The four highly effective regimens vary in drug cost. At
1980 prices in the United Kingdom, the daily regimen is themost expensive, costing 286 for a patient weighing 50 kg.The three-times-weekly regimen of isoniazid, rifampicin,streptomycin, and pyrazinamide costs only £ 108, this risingto D44 when the streptomycin is replaced by ethambutol tomake a fully oral regimen, and the 5-drug regimen costs £ 65.In comparison, the 9-month daily regimen of isoniazid,rifampicin, and ethambutol recommended by the BritishThoracic Association’ costs £ 290.Adverse reactions were not a major problem even with the
intermittent regimens, which include higher dosages ofisoniazid and pyrazinamide. The frequency with which oneor more of the drugs was terminated varied from 6% to 3%,the regimens which included both streptomycin and pyra-zinamide having the highest rates. Hepatic reactions, definedas abnormalities in liver function leading to a modification tothe regimen, were commonest in patients on the daily regi-men (p«0 - 001, 4df), and did not occur in those on the onlyregimen which did not include pyrazinamide. Nevertheless,of 21 hepatic reactions in 967 patients on pyrazinamide, 16were no more than transient and symptomless abnormalitiesin liver function, and in only 3 did jaundice occur. Reactionsto rifampicin were rare; thrombocytopenic purpura occurredin 2 patients, and the "flu" syndrome in 4. Rifampicin-dependent antibodies were found in only 3 of 400 patients (1 1on intermittent and 2 on daily rifampicin), none of whom hadsymptoms, which confirms previous experience that rifam-picin 3 times a week is no more likely to sensitise patients thandaily rifampicin. 6In conclusion, four 6-month regimens (3 fully intermittent
and 1 daily) in this study are highly effective as assessed up to12 months after the end of chemotherapy, even for patientswith strains initially resistant to isoniazid and streptomycin,and previous experience suggests that these excellent resultsare likely to be maintained.l5
It is important to find out whether equally good results canbe achieved if pyrazinamide is given for a shorter period, ifstreptomycin is dispensed with or given only for an initialperiod, and if ethambutol is dispensed with, measures toreduce the cost and the toxicity risk of the treatment, as well asthe number of tablets and injections. Two of these possi-bilities are being investigated in Hong Kong.
In Hong Kong, the medical and laboratory staffs, the medical social workers,health visitors, and health auxillaries, and the nursing, radiographic, technical,secretarial, and administrative staffs of the Wanchai, Sai Ying Pun,Shaukeiwan, Yaumati, Kowloon, East Kowloon, Shek Kip Mei, and SouthKwai Chung Government Chest Clinics, the Government Pathological Insti-tute, Ruttonjee Sanatorium, Grantham, Haven of Hope, Kowloon, and WongTai Sin hospitals cooperated in the study. The Medical Research Council Unitfor Laboratory Studies of Tuberculosis (Professor D. A. Mitchison, Mr V.Aber, and Mr B. W. Allen) was the bacteriological laboratory. Dr W. G. L.Allan, Mr W. Y. Chiu, and Mr Wong Tit Tong coordinated the study in HongKong; Dr Dip Singh and Dr W. G. L. Allan, being independent assessors. DrD. J. Girling of the Medical Research Council Tuberculosis and ChestDiseases Unit coordinated the study in London; Dr J. R. Bignall was the in-dependent radiological assessor. The report was prepared by Dr D. J. Girling,Prof. Wallace Fox, Mr A. J. Nunn, and Prof. D. A. Mitchison. Ciba-Geigy ofBasle and Gruppo Lepetit of Milan generously provided the rifampicin. Allthose cooperating in the study thank the Director of the Medical and HealthServices in Hong Kong, Dr K. L. Thong, for his interest and encouragement.
Requests for reprints should be addressed to the Director, Medical ResearchCouncil Tuberculosis and Chest Diseases Unit, Brompton Hospital, FulhamRoad, London SW3 6HP, or to the Government Senior Specialist, WanchaiChest Clinic, Kennedy Road, Wanchai, Hong Kong.
REFERENCES
1. British Thoracic and Tuberculosis Association. Short-course chemotherapy in pul-monary tuberculosis. Lancet 1976; ii. 1102-04.
2. Singapore Tuberculosis Service/British Medical Research Council Clinical trial of six-month and four-month regimens of chemotherapy in the treatment of pulmonarytuberculosis. Am Rev Resp Dis 1979; 119: 579-85.
3. Hong Kong Chest Service/British Medical Research Council. Controlled trial of6-month and 8-month regimens in the treatment of pulmonary tuberculosis theresults up to 24 months. Tubercle 1979; 60: 201-10.
4. East African/British Medical Research Councils Controlled clinical trial of four6-month regimens of chemotherapy for pulmonary tuberculosis Second report. AmRev Resp Dis 1976; 114: 471-75.
5. East African/British Medical Research Councils. Controlled clinical trial of five short-course (4-month) chemotherapy regimens in pulmonary tuberculosis. Lancet 1978,ii: 334-38.
6. Hong Kong Chest Service/British Medical Research Council Controlled trial of6-month and 8-month regimens in the treatment of pulmonary tuberculosis. Firstreport. Am Rev Resp Dis 1978; 118: 219-27.
7. Worlledge S. The detection of rifampicin-dependent antibodies. Scand J Resp Dis 1973,suppl 84: 60-63.
8. Ellard GA, Greenfield C. A sensitive urine-test method for monitoring the ingestion ofisoniazid. J Clin Pathol 1977; 30: 84-87
9. Hong Kong Chest Service/British Medical Research Council. A service program ofantituberculosis chemotherapy with 5 drugs for 4 months in the treatment of drugaddicts and prisoners with pulmonary tuberculosis in Hong Kong. Am Rev Resp Dis1980; 122: 417-24.
10. Tuberculosis Chemotherapy Centre Madras. A concurrent comparison of isoniazidplus PAS with three regimens of isoniazid alone in the domiciliary treatment ofpulmonary tuberculosis. Bull World Hlth Org 1960; 23: 535-85
11. Simon G. Radiology in epidemiological studies and some therapeutic trials. Br Med J1966; ii: 491-94.
12. Girling DJ. Adverse reactions to rifampicin in antituberculosis regimens. J AntimicrobChemother 1977, 3: 115-32
13. Somner RA. Six month regimens in pulmonary tuberculosis. Presented at the
Congressus Phthiseologicus et Pneumologicus Cechoslovacus, Prague, July 8-11,1980.
14. Fox W. The chemotherapy of pulmonary tuberculosis: a review. Chest 1979, suppl.76S: 785S-96S.
15. Fox W. The current status of short-course chemotherapy Bull Int Union Against Tuber-culosis 1978; 53: 268-80.
INCREASED SENSITIVITY AND ACCURACY OFPHAEOCHROMOCYTOMA DIAGNOSIS
ACHIEVED BY USE OF PLASMA-ADRENALINEESTIMATIONS AND A PENTOLINIUM-
SUPPRESSION TEST
M. J. BROWND. A. JENNER
D. J. ALLISONP. J. LEWIS
C. T. DOLLERY
Departments of Clinical Pharmacology and Radiology, RoyalPostgraduate Medical School, Du Cane Road, London W12 OHS
Summary 26 patients with suspected phaeochromo-cytoma underwent selective venous sampl-
ing: a tumour was identified and removed in 18. 4 of these pa-tients did not have sustained hypertension, and the only per-sistent abnormality was an elevated plasma-adrenaline level.The increased plasma-catecholamine concentrations in thetumour patients were not suppressed by pentolinium 2· 5 mgi.v., whereas the levels in the 8 non-tumour patients werelowered to the normal range. Calculations based on thearteriovenous differences of adrenaline and noradrenaline
suggest that adrenal medullary secretion contributed lessthan 2% of circulating noradrenaline in the non-tumour pa-tients. Plasma-adrenaline, however, is derived almost entirelyfrom the adrenal gland, and its estimation appears superior tothat of noradrenaline in the detection of a small adrenalphaeochromocytoma.
