Smear-negative, Culture-positive Pulmonary Tuberculosis Six-month Chemotherapy with Isoniazid and Rifampin1.2
ASIM K. DUTT, DORY MOERS, and WILLIAM W. STEAD
Introduction
SUMMARY We have shown in Arkansas that 9 months of therapy with Isoniazid (INH) and rifampin
Nine months of chemotherapy with iso(RIF)can achieve lasting success in 95% of cases with sputum-smear-positlve pUlmonary tuberculoniazid (INH) and rifampin (RIF) have sis. It seemed likely that when the tubercle bacilli were less numerous, i.e., could not be seen on proved adequate for patients with admicroscopy, less therapy would suffice. Thus, in January 1980, we began giving only 6 months of treatment to patients in whom at least one sputum culture showed M. tuberculosis but at least three vanced pulmonary tuberculosisevenwhen sputum smears showed no organisms. The regimen for adults is INH 300 mg and RIF 600 mg daily Mycobacterium tuberculosis organisms for 1 month followed by INH 900 mg and RIF 600 mg twice weekly for another 5 months. To date, are present in large enough numbers to 286 patients with an average age of 68.2 yr have been treated in this manner. Associated medical be found on simple microscopy of the conditions were present as "risk factors" in 23.7%. The full course oftherapy could not be completed sputum. This is true in patients with susin 75 patients (26.2%), largely because of side effects of the drugs and non-TB deaths In this group ceptible organisms (1, 2) whether mediof elderly patients. Side effects of the drugs requiring change of drug(s) occurred in 33 patients cation is administered on a daily basis (11.5%), but major side effects occurred in only eight (2.8%), four (1.4%) with toxic hepatitis and for the entire time or changed to twicefour With hematologic toxicity. The side effects in 25 patients (8.7%) were not life-threatening and weekly administration at the end of the were due to drug intolerance. Treatment failed during therapy in only one patient. The full 6-month first month. However, several investigacourse of therapy was completed by 211 patients. During follow-up from 3 to 107 months (median, tors have found that reducing the dura45 months), five of 211 patients (2.4%) relapsed, all with drug-susceptible organisms. Life table analysis did not show significant difference In survival between patients treated for 6 months (three tion of treatment to 6 months increases negative smears) and those treated for 9 months (smear-positive). An overall success of 97% was the number of both failures and relapses achieved, which is comparable to the success of 9 months in smear-positive cases. Thus, shorten(3-5). ing of therapy for less serious cases can be achieved without loss of effectiveness. In 1979, we observed that patients AM REV RESPIR DIS 1990; 141:1232-1235 whose initial sputum smears were negative experienced a much more rapid conversion of sputum cultures to negative than did patients in whom sputum smears with an interest in tuberculosis treat patients and serologic studies for fungi, bronchoscowere positive and that both failures and at 40 county health department chest clinics. py including bronchial washing, brushing and relapses were less common. It occurred They are ably assisted by local public health trans bronchial biopsy, transthoracic needle to us that it would be reasonable to re- nurses in the supervision of compliance and aspiration, or even open lung biopsy. When duce the duration of therapy from 9 to monitoring for drug side effects. When indi- such investigations are nondiagnostic, the 6 months in carefully selected patients. cated, patients may be hospitalized in 16 physician may elect to start antituberculosis Therefore, starting in 1980,we requested general hospitals around the state under con- therapy with INH and RIF while awaiting cultract with the Arkansas Department of Health. ture results. If any of the specimens yield M that the chest clinicians working with us The drugs are generally self-administered tuberculosis on culture, but all smears are make a point ef always submitting at except when noncompliance is suspected, in negative, therapy is stopped at 6 months and least four specimens of sputum for smear which case therapy is directly supervised by the patient is followed carefully for another and culture of tubercle bacilli before or a nurse or by a responsible person. Surveil- 18 months. As reported recently, therapy is within 2 wk after starting therapy with lance of patient compliance and mycobacteri- stopped at 4 months if all smear and cultures INH and RIF for a projected course of al examination in the program has been de- are negative in a patient with a clinical and scribed previously (2,6). Initial and progress radiographic picture compatible with tuber9 months. Then, if sputum yielded M tuberculosis on culture but all smears reports are sent to the central office where culosis (6). were negative, indicating a small popu- the data are maintained in a microcomputer Treatment Protocol to facilitate case management and analysis lation of organisms, the course of thera- of results. The therapeutic protocol consists of INH 300 py should be reduced to 6 months, with a careful post-therapy follow-up of 18 Selection of Patients months. The results of our nine years of In the program, at least four specimens of experience with this shortened regimen sputum are sent for bacteriologic examina- (Received in original form May 9, /989 and in revised form October 26, /989) are presented herein. tion for each patient suspected of tuberculoMethods The tuberculosis program of the Arkansas Department of Health has been described elsewhere (2). Statewide, about 45 physicians 1232
sis. If a sputum smear is positive for acid-fast bacilli, antituberculosis chemotherapy is started without delay. In smear-negative patients, additional investigations may be performed to establish the cause of radiographic abnormalities, i.e., smears and culture of sputum
I From the Tuberculosis Program, Arkansas Department of Health, Little Rock, Arkansas. 2 Correspondence and requests for reprints should be addressed to Tuberculosis Program, Arkansas Department of Health, 4815 West Markham, Little Rock, AR 72205-3867.
1233
SMEAR-NEGATIVE, CULTURE-POSITIVE PULMONARY TUBERCULOSIS
TABLE 1
therapy was changed to eliminate the offending drug.
AGE, SEX, AND RACE DISTRIBUTION AMONG 286 PATIENTS' Age Group (yr) 18-39 40-59 60-79 80-100 Total
White
Black
Other
Male
Female
Male
Female
8 22 66 26 122 (42.7)
11 13 41 41 106 (37.1)
5 7 14 9 35 (12.2)
1 4 7 7 19 (6.6)
Male
Incomplete Therapy
Female
2 1 (0.3)
3 (1.0)
Total 27 46 130 83 286
Protocol therapy was not completed for reasons other than side effects in 42 patients (14.7%) (table 4). Death from nonTB causes occurred in 21 patients, mostly as a result of cardiac and cerebrovascular disease. Death occurred in four patients soon after initiation of therapy, and tuberculosis may have contributed to their demise. During therapy, eight patients moved to another state and four patients were either lost or refused therapy. Primary INH resistance was found in four patients (1.4%) for which two drugs were added to cover for drug resistance. Thus, therapy could be completed in 211 patients. However, treatment failure occurred in one patient when a sputum smear was positive in the fifth month. Pretreatment culture was fully drugsusceptible. Two additional drugs were added under supervision pending susceptibility tests. When the organisms were later cultured and reported susceptible to both drugs, a full course of 9 months was completed with INH and RIP. The patient has remained well since then.
(9.4) (16.1) (45.5) (29.0)
... Percentages are shown in parentheses.
mg and RIF 600 mg (given as two combination capsules of Rifamate) in a single daily dose for I month followed by INH 900 mg and RIF 600mg (two Rifamate capsulesplus two INH 300 mg tablets) in a single dose, twice-weekly for another 5 months, i.e., a total course of 6 months. Pyridoxine 50 mg is given withevery dose for the duration of therapy to patients who are elderly or poorly nourished. Monitoring of Bacteriology After the several initial sputum specimens, one specimen is examined every month during therapy plus another 6 months and thereafter every 3 months for another 12months. The patient is discharged from observation 2 yr after initiation of therapy with instructions to report if symptoms recur or develop. Monitoring of Drug Side Effects In everypatient, baseline liver function tests and complete blood count (CBC) are performed beforeinitiation oftherapy. Monitoring of drug toxicity is largelyclinical. On development of symptoms suggestive of drug toxicity, the drugs are discontinued until results of aminotransferase, alkaline phosphatase, and serum bilirubin determinations are available. On recovering from the symptoms, return of liverfunction to normal, and at the discretion of the treating physician, the patient may be challenged with reduced doses of each drug separately for severaldaysin order to identify the offending drug. In about half of the patients, challengedoes not evoke a reaction and therapy with both drugs can be reinstituted. When the patient is unable to toleratea drug, a suitable bactericidalregimen is substituted. Patient Characteristics From January 1980to September 1988,286 patients met the criteria and weretreated with the describedregimen. Age,race,and sexdistributions are shown in table 1. The average age was 68.2 yr (range, 18 to 99 yr), which is about 8 yr older than the mean age of other groups we have reported. Three-quarters were older than 60 yr of age. Presence of Risk Factors A significant "risk factor" (diabetes, alcoholism, etc.)waspresent in 68 (23.7070) of the
TABLE 2 PRESENCE OF OTHER MEDICAL CONDITIONS AS "RISK FACTORS" IN 68 (24%) PATIENTS Patients' (n)
"Risk Factors" Various present or past malignancies Excess alcohol use Diabetes mellitus Long-term corticosteroid therapy Gastrectomy Chronic renal disease Collagen vascular disease
27 18 15
5 5
3 2
• Seven patients with more than one condition.
286patients (table2).Seven patients had more than one "risk factor."
Follow-up and Relapses
Results
Treatment protocol for 6 months was successfully completed in 211 patients. Follow-up ranged from 3 to 107 months (table 5). The median follow-up was 45 months for a total of 4,815 patient-months. During this period of follow-up, five patients (2.4%) relapsed bacteriologically. These relapses occurred 3, 29,37,60, and 73 months after completion of therapy. All relapses were associated with drugsusceptible organisms, and the patients were retreated with the same drugs for a full 9-month course. During followup, 34 patients died of nontuberculous causes, mostly as a result of cardiac and cerebrovascular disease, and six patients moved out of the state. The remaining
Side Effects of Drugs Reported side effects of the drugs were clinical impressions of the 45 treating physicians. When the side effects were life-threatening (hepatic and hematologic abnormalities), they were considered major; otherwise, they were considered minor. Drug side effects required change of regimen in 33 (11.5010) of 286 patients during therapy (table 3). However, major side effects occurred in only eight patients (2.8%): hepatitis in four (1.4%) and hematologic abnormalities (petechiae, thrombocytopenia, or anemia) in four (1.4%). Although the side effects were minor in the remaining 25 patients (8.7%),
TABLE 3 DRUG SIDE EFFECTS IN 33 OF 286 PATIENTS' Side Effects Major, 8 Hepatitis Hematologic Minor, 25 "Flu-like syndrome" Fever, myalgia Rash Gastrointestinal intolerance Insomnia, nightmares
RIF
INH
1 2
2 1
2 3 2 2
4 4 5
• Side effects necessitated withdrawal of drug(s) and institution of new regimen.
Unknown
1234
DUTT, MOERS, AND STEAD
1.0
TABLE 4
-r--..::""------------..,
PROTOCOL THERAPY NOT COMPLETED AMONG 286 PATIENTS Fig. 1. Survival of patients with tuberculosis: 6 months (closed squares) versus 9 months (open squares) of chemotherapy.There was no significant difference between the groups. The insignificant divergence in the lines is due to greater mean age (8 yr) in the 6-month therapy group.
Patients (n)
Reason Drug side effects Non-TB death TB death Relocation during therapy Refusal of treatment or lost Initial INH resistance Treatment failure during therapy Total
0.9
33 21 4 8 4 4 1
o 8 -+-..,.....-r-...-,-...-,--..-.,,......---,-...--,.........--r....,...-i 16 4 6 8 10 12 14 o 2
75 (26.2%)
Six Month Intervals
(2). Reduction of the duration to 6 months with INH and RIF in unselected patients has been shown to result in increased failures (5, 7). The large bacterial populaOverall Success oj Therapy tion in smear-positive cases of disease Two hundred twelve patients with three cannot be reliably eliminated with less or more negative initial sputum smears than 9 months of treatment with only two but at least one culture positive for M. bactericidal drugs. However, we observed rapid convertuberculosis completed 6 months protocol therapy with INH and RIP. An over- sion of the sputum culture to negative all success was achieved in 206 patients in patients with smaller populations of (97070) during the median follow-up of organisms shown to be negative on sev45 months. Treatment failed in one pa- eral sputum smears. In smear-negative tient during therapy, and late relapses oc- patients, cultures converted to negative curred in five after completion of thera- in 93 of 101 (93%) within 2 months as py, all with drug-susceptible organisms. compared with 282of 415 (68%) in smearSurvival of patients, using the life table positive patients. This was presumably method, did not show significant differ- due to a smaller bacterial population in ences between patients treated with a the former (4). One would assume from 6-month (smear-negative) or a 9-month this that smear-negative patients would (smear-positive) regimen (figure 1). require lesstreatment than smear-positive ones (8), but no study has been done with Discussion INH + RIF to determine the appropriIn Arkansas, where the incidence of pri- ate duration of therapy. We postulated mary drug resistance is quite low (about that smear-negative, culture-positive pul2%), the routine of therapy since 1976 monary disease should require somewhat has been a 9-month short-course chemo- less therapy. Preliminary studies from the therapy regimen consisting of INH 300 British Medical Research Council had inmg and RIF 600 mg daily for 1 month dicated that 2 or 3 months of treatment followed by INH 900 mg and RIF 600 with four drugs consisting of streptomymg twice weekly for another 8 months cin (SM), INH, RIF, and pyrazinamide 166patients werefollowed for 18months, with a sputum culture every month for 6 months and every 3 months thereafter.
TABLE 5 FOLLOW-UP OF 211 PATIENTS AFTER COMPLETION OF THERAPY Follow-up (months)
1-12 13-24 25-36 37-48 49-60 61-72 73-107 Total
Patients (n)
38 50 25 31 31 14 22 211
Relapse (n)
(months)
3
Non-TB Deaths 17
14 29 37 60 1 5 (2.4%)
Moved
5 1
3
73 34
6
Remaining Disease-free
15 35 24 27 30 14 21 166
(PZA) werenot adequate for smear-negative but culture-positive disease (9). In early 1980, on the basis of our observations, the duration of therapy was reduced to 6 months for such patients and to 4 months for smear-negative, culturenegative patients thought to have tuberculosis on clinical and radiologic grounds (6). During this period, 286 patients qualified for the protocol therapy with INH and RIP. Many of the patients wereelderIy, with an average age of 68.2 yr (range, 18to 99 yr); 74% of them wereolder than 60 yr of age, which was somewhat greater than the 60-yr mean of patients in our earlier reports (2, 10). Also, 24% of patients had associated medical disorders as "risk factors." Although approximately 26% of patients did not complete the full course of therapy, this was largely due to minor side effects of the drugs and death from cardiac and cerebrovascular disease. Neither is unexpected in such a group of elderly patients. Drug side effects were reported in the present series in 11.5% of the patients, which is higher than we have reported in our overall experience (2, 10, 11). This is not surprising when one considers that the mean age was 8 yr greater in this group than in our general patient population. Major life-threatening toxicity in only 3% was almost the same as our overall experience. Hepatic toxicity was reported in 1.4% and hematologic disorders consisting of petechiae, thrombocytopenia, and anemia in 1.4% of patients. The minor side effects, however, were more frequent (8.7%) as reported by 45 treating physicians. This may be attributed to a greater incidence of drug intolerance in elderly patients and interphysician variability in reporting. Most of the time the drugs could be reinstituted with minor adjustment of time of ad-
1235
SMEAR-NEUATIVE, CULTURE-POSITIVE PULMONARY TUBERCULOSIS
ministration. In others, therapy had to be changed to elminate the offending drug. Initial INH resistance was found in four patients (1.4010), which indicates low prevalence of drug resistance in Arkansas. Thus, the use of an additional drug(s) is not necessary, as should be the case in patients from areas with a greater incidence of drug resistance (e.g., large cities and countries with high incidence of drug resistance). During therapy, treatment failed in one patient when smears and cultures were found to be positive in the fifth month of treatment. Pretreatment culture had drug-susceptibleorganisms. Becausetherapy had not been directly supervised, two additional drugs were added to the regimen because of suspicion of development of drug resistance. When full susceptibility to drugs was reported, therapy with INH and RIF was continued for 9 months. It is likely that the failure was due to noncompliance. Supervision of therapy was necessary in less than 1% of the patients in this study. Five of 211 patients (2.4%) relapsed during a median follow-up of 45 months. Relapses occurred evenly throughout the period of observation from 3 to 73 months. The relapse rate is comparable to that of smearand culture-positive cases after 9 months of therapy in INH/RIF (2). The bacilli in the relapsed cases were fully susceptible to the drugs, and the patients were retreated with the same drugs for a full course of 9 months. The life table method of analysis of survival of patients with tuberculosis did not show any significant differences between patients with smear-negative and culture-positive disease treated for 6 months and patients with smear- and culture-positive disease treated for 9 months (figure 1). The insignificant divergence in the lines is due to the 8-yr greater mean age of the 6-month therapy group. The present 6-month regimen appears to be adequate for most cases with smear-negative, culture-positive pulmonary tuberculosis. It appears superior to the 2 or 3 months of treatment with SM/INH/RIF/PZA reported by the British Medical Research Council, which was
followed by relapse rates of 32% in the 2-month group and 13% in the 3-month group during an observation of 60 months (12). It is comparable to 4 months of daily or intermittent treatment in the regimen, with 2% relapse rate (13). Results of a 6-month regimen of therapy with INH/RIF in smear-negative, culture-positive tuberculosis show an overall reduced duration of treatment and appear to be a reasonable approach to shortening the course of therapy for patients who meet the criteria for a small bacillary population (~3 negative smears with positive culture). As we reported recently, when both smears and cultures are all negative and when the incidence of drug resistance is low, as it is in Arkansas, we found that 4 months of INH/RIF therapy generally suffice (6). Among the 211 sputum-smear-negative, culture-positive successfully treated patients with tuberculosis, 1,266 person-months of treatment sufficed instead of 1,899,which represents a very considerable saving of resources with no significant increase in failures of relapses in this carefully selected group of patients. The cost for 70 doses of INH/RIF in this regimen was approximately $50.00 compared with $225.00 for a 6-month regimen consisting of daily INH/RIFI PZA for 2 months followed by INH/RIF for 4 months or to $380.00 for 4-month daily therapy with INH/RIF/PZA/SM. Although bacteriologic monitoring is the best method to evaluate response and to detect early failure of therapy, the number of sputum examinations in our protocol was greater than is generally needed because we were blazing a trail. Because the success of the regimen is now known, bacteriologic examinations of sputum may be reduced quantitatively in treating patients under routine program conditions. Also, follow-up of patients after cessation of therapy for 6 to 12months is desirable with examination of one or two specimens during the period before discharging the patient from care. Acknowledgment The writers thank all the treating physicians throughout Arkansas, all the public health
nurses for their dedicated services to the patients, Mr. David Bradleyof the mycobacteriology laboratory, Ms. Nancy Brannon, Ms. Mary Bryant, and Ms. Aliene Martin of the central registry, and Ms. Debbie Turner for her secretarial support. Specialthanks are due to Cassandra R. Shack for invaluable help in analyzing the data. References I. American Thoracic Society/Centers for Disease Control. Treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Respir Dis 1986; 134:355-63. 2. Dutt AK, Moers D, Stead Ww. Short course chemotherapy for tuberculosis with mainly twiceweekly isoniazid and rifampin: community physicians seven-year experiencewith mainly outpatients. Am J Med 1984; 77:233-42. 3. Fox W.The chemotherapy of pulmonary tuberculosis: a review. Chest 1979; 76(Suppl:785-96). 4. Fox W. Whither short course chemotherapy? Br J Dis Chest 1981; 75:331-57. 5. Second East African/British Medical Research Council study. Controlled clinical trial of four short course (6 month) regimens of chemotherapy for the treatment of pulmonary tuberculosis. Am Rev Respir Dis 1976; 114:471-5. 6. Dutt AK, Moers D, Stead Ww. Smear and culture negative pulmonary tuberculosis: four-month short course chemotherapy. Am Rev Respir Dis 1989; 139:867-70. 7. Snider DE, Long MW, Cross FS, Farrer LS. Six months isoniazid-rifampin therapy for pulmonary tuberculosis. Am Rev Respir Dis 1984; 129:573-9. 8. Anon. Smear-negative pulmonary tuberculosis. Tubercle 1980; 61:113-5. 9. Hong Kong Chest Service, Tuberculosis Research Center, Madras, India and British Medical Research Council. Sputum-smear-negative pulmonary tuberculosis, controlled trial of 3 month and 2 month regimens of chemotherapy. First report. Lancet 1979; 1:1361-3. 10. Dutt AK, Moers D, Stead WW. Short-course chemotherapy for extrapulmonary tuberculosis: nine years experience. Ann Intern Med 1986; 104:7-12. 11. Dutt AK, Moers D, Stead WW. Undesirable side effects of isoniazid and rifampin and largely twice-weeklyshort-course chemotherapy for tuberculosis. Am Rev Respir Dis 1983; 128:419-24. 12. Hong Kong Chest Service/Tuberculosis Research Center, Madras/British Medical Research Council. A controlled trial of 2-month, 3-month, and 12-month regimens of chemotherapy for sputum-smear-negative pulmonary tuberculosis. Results at 60 months. Am Rev Respir Dis 1984; 130:23-8. 13. Hong Kong Chest Service/Tuberculosis Research Center, Madras/British Medical Council. A controlled trial of 3-month, 4-month, and 6month regimens of chemotherapy for sputumsmear-negative pulmonary tuberculosis: results at 5 years. Am Rev Respir Dis 1989; 139:871-6.
ASIM K. DUTT, DORY MOERS, and WILLIAM W. STEAD
Introduction
SUMMARY We have shown in Arkansas that 9 months of therapy with Isoniazid (INH) and rifampin
Nine months of chemotherapy with iso(RIF)can achieve lasting success in 95% of cases with sputum-smear-positlve pUlmonary tuberculoniazid (INH) and rifampin (RIF) have sis. It seemed likely that when the tubercle bacilli were less numerous, i.e., could not be seen on proved adequate for patients with admicroscopy, less therapy would suffice. Thus, in January 1980, we began giving only 6 months of treatment to patients in whom at least one sputum culture showed M. tuberculosis but at least three vanced pulmonary tuberculosisevenwhen sputum smears showed no organisms. The regimen for adults is INH 300 mg and RIF 600 mg daily Mycobacterium tuberculosis organisms for 1 month followed by INH 900 mg and RIF 600 mg twice weekly for another 5 months. To date, are present in large enough numbers to 286 patients with an average age of 68.2 yr have been treated in this manner. Associated medical be found on simple microscopy of the conditions were present as "risk factors" in 23.7%. The full course oftherapy could not be completed sputum. This is true in patients with susin 75 patients (26.2%), largely because of side effects of the drugs and non-TB deaths In this group ceptible organisms (1, 2) whether mediof elderly patients. Side effects of the drugs requiring change of drug(s) occurred in 33 patients cation is administered on a daily basis (11.5%), but major side effects occurred in only eight (2.8%), four (1.4%) with toxic hepatitis and for the entire time or changed to twicefour With hematologic toxicity. The side effects in 25 patients (8.7%) were not life-threatening and weekly administration at the end of the were due to drug intolerance. Treatment failed during therapy in only one patient. The full 6-month first month. However, several investigacourse of therapy was completed by 211 patients. During follow-up from 3 to 107 months (median, tors have found that reducing the dura45 months), five of 211 patients (2.4%) relapsed, all with drug-susceptible organisms. Life table analysis did not show significant difference In survival between patients treated for 6 months (three tion of treatment to 6 months increases negative smears) and those treated for 9 months (smear-positive). An overall success of 97% was the number of both failures and relapses achieved, which is comparable to the success of 9 months in smear-positive cases. Thus, shorten(3-5). ing of therapy for less serious cases can be achieved without loss of effectiveness. In 1979, we observed that patients AM REV RESPIR DIS 1990; 141:1232-1235 whose initial sputum smears were negative experienced a much more rapid conversion of sputum cultures to negative than did patients in whom sputum smears with an interest in tuberculosis treat patients and serologic studies for fungi, bronchoscowere positive and that both failures and at 40 county health department chest clinics. py including bronchial washing, brushing and relapses were less common. It occurred They are ably assisted by local public health trans bronchial biopsy, transthoracic needle to us that it would be reasonable to re- nurses in the supervision of compliance and aspiration, or even open lung biopsy. When duce the duration of therapy from 9 to monitoring for drug side effects. When indi- such investigations are nondiagnostic, the 6 months in carefully selected patients. cated, patients may be hospitalized in 16 physician may elect to start antituberculosis Therefore, starting in 1980,we requested general hospitals around the state under con- therapy with INH and RIF while awaiting cultract with the Arkansas Department of Health. ture results. If any of the specimens yield M that the chest clinicians working with us The drugs are generally self-administered tuberculosis on culture, but all smears are make a point ef always submitting at except when noncompliance is suspected, in negative, therapy is stopped at 6 months and least four specimens of sputum for smear which case therapy is directly supervised by the patient is followed carefully for another and culture of tubercle bacilli before or a nurse or by a responsible person. Surveil- 18 months. As reported recently, therapy is within 2 wk after starting therapy with lance of patient compliance and mycobacteri- stopped at 4 months if all smear and cultures INH and RIF for a projected course of al examination in the program has been de- are negative in a patient with a clinical and scribed previously (2,6). Initial and progress radiographic picture compatible with tuber9 months. Then, if sputum yielded M tuberculosis on culture but all smears reports are sent to the central office where culosis (6). were negative, indicating a small popu- the data are maintained in a microcomputer Treatment Protocol to facilitate case management and analysis lation of organisms, the course of thera- of results. The therapeutic protocol consists of INH 300 py should be reduced to 6 months, with a careful post-therapy follow-up of 18 Selection of Patients months. The results of our nine years of In the program, at least four specimens of experience with this shortened regimen sputum are sent for bacteriologic examina- (Received in original form May 9, /989 and in revised form October 26, /989) are presented herein. tion for each patient suspected of tuberculoMethods The tuberculosis program of the Arkansas Department of Health has been described elsewhere (2). Statewide, about 45 physicians 1232
sis. If a sputum smear is positive for acid-fast bacilli, antituberculosis chemotherapy is started without delay. In smear-negative patients, additional investigations may be performed to establish the cause of radiographic abnormalities, i.e., smears and culture of sputum
I From the Tuberculosis Program, Arkansas Department of Health, Little Rock, Arkansas. 2 Correspondence and requests for reprints should be addressed to Tuberculosis Program, Arkansas Department of Health, 4815 West Markham, Little Rock, AR 72205-3867.
1233
SMEAR-NEGATIVE, CULTURE-POSITIVE PULMONARY TUBERCULOSIS
TABLE 1
therapy was changed to eliminate the offending drug.
AGE, SEX, AND RACE DISTRIBUTION AMONG 286 PATIENTS' Age Group (yr) 18-39 40-59 60-79 80-100 Total
White
Black
Other
Male
Female
Male
Female
8 22 66 26 122 (42.7)
11 13 41 41 106 (37.1)
5 7 14 9 35 (12.2)
1 4 7 7 19 (6.6)
Male
Incomplete Therapy
Female
2 1 (0.3)
3 (1.0)
Total 27 46 130 83 286
Protocol therapy was not completed for reasons other than side effects in 42 patients (14.7%) (table 4). Death from nonTB causes occurred in 21 patients, mostly as a result of cardiac and cerebrovascular disease. Death occurred in four patients soon after initiation of therapy, and tuberculosis may have contributed to their demise. During therapy, eight patients moved to another state and four patients were either lost or refused therapy. Primary INH resistance was found in four patients (1.4%) for which two drugs were added to cover for drug resistance. Thus, therapy could be completed in 211 patients. However, treatment failure occurred in one patient when a sputum smear was positive in the fifth month. Pretreatment culture was fully drugsusceptible. Two additional drugs were added under supervision pending susceptibility tests. When the organisms were later cultured and reported susceptible to both drugs, a full course of 9 months was completed with INH and RIP. The patient has remained well since then.
(9.4) (16.1) (45.5) (29.0)
... Percentages are shown in parentheses.
mg and RIF 600 mg (given as two combination capsules of Rifamate) in a single daily dose for I month followed by INH 900 mg and RIF 600mg (two Rifamate capsulesplus two INH 300 mg tablets) in a single dose, twice-weekly for another 5 months, i.e., a total course of 6 months. Pyridoxine 50 mg is given withevery dose for the duration of therapy to patients who are elderly or poorly nourished. Monitoring of Bacteriology After the several initial sputum specimens, one specimen is examined every month during therapy plus another 6 months and thereafter every 3 months for another 12months. The patient is discharged from observation 2 yr after initiation of therapy with instructions to report if symptoms recur or develop. Monitoring of Drug Side Effects In everypatient, baseline liver function tests and complete blood count (CBC) are performed beforeinitiation oftherapy. Monitoring of drug toxicity is largelyclinical. On development of symptoms suggestive of drug toxicity, the drugs are discontinued until results of aminotransferase, alkaline phosphatase, and serum bilirubin determinations are available. On recovering from the symptoms, return of liverfunction to normal, and at the discretion of the treating physician, the patient may be challenged with reduced doses of each drug separately for severaldaysin order to identify the offending drug. In about half of the patients, challengedoes not evoke a reaction and therapy with both drugs can be reinstituted. When the patient is unable to toleratea drug, a suitable bactericidalregimen is substituted. Patient Characteristics From January 1980to September 1988,286 patients met the criteria and weretreated with the describedregimen. Age,race,and sexdistributions are shown in table 1. The average age was 68.2 yr (range, 18 to 99 yr), which is about 8 yr older than the mean age of other groups we have reported. Three-quarters were older than 60 yr of age. Presence of Risk Factors A significant "risk factor" (diabetes, alcoholism, etc.)waspresent in 68 (23.7070) of the
TABLE 2 PRESENCE OF OTHER MEDICAL CONDITIONS AS "RISK FACTORS" IN 68 (24%) PATIENTS Patients' (n)
"Risk Factors" Various present or past malignancies Excess alcohol use Diabetes mellitus Long-term corticosteroid therapy Gastrectomy Chronic renal disease Collagen vascular disease
27 18 15
5 5
3 2
• Seven patients with more than one condition.
286patients (table2).Seven patients had more than one "risk factor."
Follow-up and Relapses
Results
Treatment protocol for 6 months was successfully completed in 211 patients. Follow-up ranged from 3 to 107 months (table 5). The median follow-up was 45 months for a total of 4,815 patient-months. During this period of follow-up, five patients (2.4%) relapsed bacteriologically. These relapses occurred 3, 29,37,60, and 73 months after completion of therapy. All relapses were associated with drugsusceptible organisms, and the patients were retreated with the same drugs for a full 9-month course. During followup, 34 patients died of nontuberculous causes, mostly as a result of cardiac and cerebrovascular disease, and six patients moved out of the state. The remaining
Side Effects of Drugs Reported side effects of the drugs were clinical impressions of the 45 treating physicians. When the side effects were life-threatening (hepatic and hematologic abnormalities), they were considered major; otherwise, they were considered minor. Drug side effects required change of regimen in 33 (11.5010) of 286 patients during therapy (table 3). However, major side effects occurred in only eight patients (2.8%): hepatitis in four (1.4%) and hematologic abnormalities (petechiae, thrombocytopenia, or anemia) in four (1.4%). Although the side effects were minor in the remaining 25 patients (8.7%),
TABLE 3 DRUG SIDE EFFECTS IN 33 OF 286 PATIENTS' Side Effects Major, 8 Hepatitis Hematologic Minor, 25 "Flu-like syndrome" Fever, myalgia Rash Gastrointestinal intolerance Insomnia, nightmares
RIF
INH
1 2
2 1
2 3 2 2
4 4 5
• Side effects necessitated withdrawal of drug(s) and institution of new regimen.
Unknown
1234
DUTT, MOERS, AND STEAD
1.0
TABLE 4
-r--..::""------------..,
PROTOCOL THERAPY NOT COMPLETED AMONG 286 PATIENTS Fig. 1. Survival of patients with tuberculosis: 6 months (closed squares) versus 9 months (open squares) of chemotherapy.There was no significant difference between the groups. The insignificant divergence in the lines is due to greater mean age (8 yr) in the 6-month therapy group.
Patients (n)
Reason Drug side effects Non-TB death TB death Relocation during therapy Refusal of treatment or lost Initial INH resistance Treatment failure during therapy Total
0.9
33 21 4 8 4 4 1
o 8 -+-..,.....-r-...-,-...-,--..-.,,......---,-...--,.........--r....,...-i 16 4 6 8 10 12 14 o 2
75 (26.2%)
Six Month Intervals
(2). Reduction of the duration to 6 months with INH and RIF in unselected patients has been shown to result in increased failures (5, 7). The large bacterial populaOverall Success oj Therapy tion in smear-positive cases of disease Two hundred twelve patients with three cannot be reliably eliminated with less or more negative initial sputum smears than 9 months of treatment with only two but at least one culture positive for M. bactericidal drugs. However, we observed rapid convertuberculosis completed 6 months protocol therapy with INH and RIP. An over- sion of the sputum culture to negative all success was achieved in 206 patients in patients with smaller populations of (97070) during the median follow-up of organisms shown to be negative on sev45 months. Treatment failed in one pa- eral sputum smears. In smear-negative tient during therapy, and late relapses oc- patients, cultures converted to negative curred in five after completion of thera- in 93 of 101 (93%) within 2 months as py, all with drug-susceptible organisms. compared with 282of 415 (68%) in smearSurvival of patients, using the life table positive patients. This was presumably method, did not show significant differ- due to a smaller bacterial population in ences between patients treated with a the former (4). One would assume from 6-month (smear-negative) or a 9-month this that smear-negative patients would (smear-positive) regimen (figure 1). require lesstreatment than smear-positive ones (8), but no study has been done with Discussion INH + RIF to determine the appropriIn Arkansas, where the incidence of pri- ate duration of therapy. We postulated mary drug resistance is quite low (about that smear-negative, culture-positive pul2%), the routine of therapy since 1976 monary disease should require somewhat has been a 9-month short-course chemo- less therapy. Preliminary studies from the therapy regimen consisting of INH 300 British Medical Research Council had inmg and RIF 600 mg daily for 1 month dicated that 2 or 3 months of treatment followed by INH 900 mg and RIF 600 with four drugs consisting of streptomymg twice weekly for another 8 months cin (SM), INH, RIF, and pyrazinamide 166patients werefollowed for 18months, with a sputum culture every month for 6 months and every 3 months thereafter.
TABLE 5 FOLLOW-UP OF 211 PATIENTS AFTER COMPLETION OF THERAPY Follow-up (months)
1-12 13-24 25-36 37-48 49-60 61-72 73-107 Total
Patients (n)
38 50 25 31 31 14 22 211
Relapse (n)
(months)
3
Non-TB Deaths 17
14 29 37 60 1 5 (2.4%)
Moved
5 1
3
73 34
6
Remaining Disease-free
15 35 24 27 30 14 21 166
(PZA) werenot adequate for smear-negative but culture-positive disease (9). In early 1980, on the basis of our observations, the duration of therapy was reduced to 6 months for such patients and to 4 months for smear-negative, culturenegative patients thought to have tuberculosis on clinical and radiologic grounds (6). During this period, 286 patients qualified for the protocol therapy with INH and RIP. Many of the patients wereelderIy, with an average age of 68.2 yr (range, 18to 99 yr); 74% of them wereolder than 60 yr of age, which was somewhat greater than the 60-yr mean of patients in our earlier reports (2, 10). Also, 24% of patients had associated medical disorders as "risk factors." Although approximately 26% of patients did not complete the full course of therapy, this was largely due to minor side effects of the drugs and death from cardiac and cerebrovascular disease. Neither is unexpected in such a group of elderly patients. Drug side effects were reported in the present series in 11.5% of the patients, which is higher than we have reported in our overall experience (2, 10, 11). This is not surprising when one considers that the mean age was 8 yr greater in this group than in our general patient population. Major life-threatening toxicity in only 3% was almost the same as our overall experience. Hepatic toxicity was reported in 1.4% and hematologic disorders consisting of petechiae, thrombocytopenia, and anemia in 1.4% of patients. The minor side effects, however, were more frequent (8.7%) as reported by 45 treating physicians. This may be attributed to a greater incidence of drug intolerance in elderly patients and interphysician variability in reporting. Most of the time the drugs could be reinstituted with minor adjustment of time of ad-
1235
SMEAR-NEUATIVE, CULTURE-POSITIVE PULMONARY TUBERCULOSIS
ministration. In others, therapy had to be changed to elminate the offending drug. Initial INH resistance was found in four patients (1.4010), which indicates low prevalence of drug resistance in Arkansas. Thus, the use of an additional drug(s) is not necessary, as should be the case in patients from areas with a greater incidence of drug resistance (e.g., large cities and countries with high incidence of drug resistance). During therapy, treatment failed in one patient when smears and cultures were found to be positive in the fifth month of treatment. Pretreatment culture had drug-susceptibleorganisms. Becausetherapy had not been directly supervised, two additional drugs were added to the regimen because of suspicion of development of drug resistance. When full susceptibility to drugs was reported, therapy with INH and RIF was continued for 9 months. It is likely that the failure was due to noncompliance. Supervision of therapy was necessary in less than 1% of the patients in this study. Five of 211 patients (2.4%) relapsed during a median follow-up of 45 months. Relapses occurred evenly throughout the period of observation from 3 to 73 months. The relapse rate is comparable to that of smearand culture-positive cases after 9 months of therapy in INH/RIF (2). The bacilli in the relapsed cases were fully susceptible to the drugs, and the patients were retreated with the same drugs for a full course of 9 months. The life table method of analysis of survival of patients with tuberculosis did not show any significant differences between patients with smear-negative and culture-positive disease treated for 6 months and patients with smear- and culture-positive disease treated for 9 months (figure 1). The insignificant divergence in the lines is due to the 8-yr greater mean age of the 6-month therapy group. The present 6-month regimen appears to be adequate for most cases with smear-negative, culture-positive pulmonary tuberculosis. It appears superior to the 2 or 3 months of treatment with SM/INH/RIF/PZA reported by the British Medical Research Council, which was
followed by relapse rates of 32% in the 2-month group and 13% in the 3-month group during an observation of 60 months (12). It is comparable to 4 months of daily or intermittent treatment in the regimen, with 2% relapse rate (13). Results of a 6-month regimen of therapy with INH/RIF in smear-negative, culture-positive tuberculosis show an overall reduced duration of treatment and appear to be a reasonable approach to shortening the course of therapy for patients who meet the criteria for a small bacillary population (~3 negative smears with positive culture). As we reported recently, when both smears and cultures are all negative and when the incidence of drug resistance is low, as it is in Arkansas, we found that 4 months of INH/RIF therapy generally suffice (6). Among the 211 sputum-smear-negative, culture-positive successfully treated patients with tuberculosis, 1,266 person-months of treatment sufficed instead of 1,899,which represents a very considerable saving of resources with no significant increase in failures of relapses in this carefully selected group of patients. The cost for 70 doses of INH/RIF in this regimen was approximately $50.00 compared with $225.00 for a 6-month regimen consisting of daily INH/RIFI PZA for 2 months followed by INH/RIF for 4 months or to $380.00 for 4-month daily therapy with INH/RIF/PZA/SM. Although bacteriologic monitoring is the best method to evaluate response and to detect early failure of therapy, the number of sputum examinations in our protocol was greater than is generally needed because we were blazing a trail. Because the success of the regimen is now known, bacteriologic examinations of sputum may be reduced quantitatively in treating patients under routine program conditions. Also, follow-up of patients after cessation of therapy for 6 to 12months is desirable with examination of one or two specimens during the period before discharging the patient from care. Acknowledgment The writers thank all the treating physicians throughout Arkansas, all the public health
nurses for their dedicated services to the patients, Mr. David Bradleyof the mycobacteriology laboratory, Ms. Nancy Brannon, Ms. Mary Bryant, and Ms. Aliene Martin of the central registry, and Ms. Debbie Turner for her secretarial support. Specialthanks are due to Cassandra R. Shack for invaluable help in analyzing the data. References I. American Thoracic Society/Centers for Disease Control. Treatment of tuberculosis and tuberculosis infection in adults and children. Am Rev Respir Dis 1986; 134:355-63. 2. Dutt AK, Moers D, Stead Ww. Short course chemotherapy for tuberculosis with mainly twiceweekly isoniazid and rifampin: community physicians seven-year experiencewith mainly outpatients. Am J Med 1984; 77:233-42. 3. Fox W.The chemotherapy of pulmonary tuberculosis: a review. Chest 1979; 76(Suppl:785-96). 4. Fox W. Whither short course chemotherapy? Br J Dis Chest 1981; 75:331-57. 5. Second East African/British Medical Research Council study. Controlled clinical trial of four short course (6 month) regimens of chemotherapy for the treatment of pulmonary tuberculosis. Am Rev Respir Dis 1976; 114:471-5. 6. Dutt AK, Moers D, Stead Ww. Smear and culture negative pulmonary tuberculosis: four-month short course chemotherapy. Am Rev Respir Dis 1989; 139:867-70. 7. Snider DE, Long MW, Cross FS, Farrer LS. Six months isoniazid-rifampin therapy for pulmonary tuberculosis. Am Rev Respir Dis 1984; 129:573-9. 8. Anon. Smear-negative pulmonary tuberculosis. Tubercle 1980; 61:113-5. 9. Hong Kong Chest Service, Tuberculosis Research Center, Madras, India and British Medical Research Council. Sputum-smear-negative pulmonary tuberculosis, controlled trial of 3 month and 2 month regimens of chemotherapy. First report. Lancet 1979; 1:1361-3. 10. Dutt AK, Moers D, Stead WW. Short-course chemotherapy for extrapulmonary tuberculosis: nine years experience. Ann Intern Med 1986; 104:7-12. 11. Dutt AK, Moers D, Stead WW. Undesirable side effects of isoniazid and rifampin and largely twice-weeklyshort-course chemotherapy for tuberculosis. Am Rev Respir Dis 1983; 128:419-24. 12. Hong Kong Chest Service/Tuberculosis Research Center, Madras/British Medical Research Council. A controlled trial of 2-month, 3-month, and 12-month regimens of chemotherapy for sputum-smear-negative pulmonary tuberculosis. Results at 60 months. Am Rev Respir Dis 1984; 130:23-8. 13. Hong Kong Chest Service/Tuberculosis Research Center, Madras/British Medical Council. A controlled trial of 3-month, 4-month, and 6month regimens of chemotherapy for sputumsmear-negative pulmonary tuberculosis: results at 5 years. Am Rev Respir Dis 1989; 139:871-6.
