Isoniazid

Background: Isoniazid-resistant tuberculosis (INHr-TB) can be treated successfully with several treatment regimens. However, the optimal regimen and duration are unclear.

Study objective: To analyze the efficacy of treatment regimens used for INHr-TB in the southeastern Texas region.

Design: Retrospective cohort study.

Setting: Health-care facilities reporting tuberculosis (TB) patients in the Houston and Tyler areas.

Subjects: All patients reported to have INHr-TB from 1991 to 1998. Exclusion criteria included poor compliance, additional first-line drug-resistance (except aminoglycosides), and death before completion of 1 month of treatment.

Measurements and results: Main treatment outcomes were treatment failure, relapse, and TB-related death. Fifty-three of 83 patients were included in the study; aminoglycoside resistance coexisted in 37.5% of isolates. Seven types of treatment regimens were identified. Eighteen patients (34%) received rifampin, pyrazinamide, and ethambutol thrice weekly for 9 months. Four patients (7.5%) had a total effective treatment duration of [is less than] 9 months. Thirty patients (56.6%) and 16 patients (30.2%) received thrice-daily and daily treatment regimens, respectively. Forty-nine patients achieved sputum conversion. Treatment failure and death occurred in one patient (1.9%). Three patients (5.7%) experienced relapses. There was a significant difference in total effective treatment time between patients with and without relapses (8.3 [+ or -] 1.1 months vs 11.1 [+ or -] 2.1 months; p [is less than] 0.02). Twice-weekly treatment regimens were associated with relapse (p = 0.05).

Conclusions: Several treatment regimens were prescribed for INHr-TB in southeastern Texas. INHr-TB treatment durations were [is greater than] 7 months, and treatment regimen efficacy was adequate. Twice-weekly treatment was associated with relapse, whereas thrice-weekly and daily treatments performed similarly. A prospective study with different treatment durations is needed to determine the optimal treatment regimen for patients with INHr-TB. (CHEST 2001; 119:1730-1736)

Key words: drug resistance; isoniazid; isoniazid resistance; relapse; treatment; tuberculosis

Abbreviations: ATS = American Thoracic Society; BMRC = British Medical Research Council; CDC = Centers for Disease Control and Prevention; DOT = directly observed therapy; EMB = ethambutol; INH = isoniazid; INHr-TB = isoniazid-resistant tuberculosis; PZA = pyrazinamide; RIF = rifampin; SM = streptomycin; TB = tuberculosis

Effective short-course chemotherapy for tuberculosis (TB) has been available worldwide for [is greater than] 20 years.[1] Despite this, drug-resistant TB continues to threaten the efforts to control the disease in many parts of the world. A recent global surveillance study[2] assessing the resistance to first-line anti-TB agents depicted the magnitude of the problem. In particular, primary and secondary isoniazid-resistant TB (INHr-TB) strains were highly prevalent, particularly in areas with a high incidence of TB.[2] Some developed countries, including the United States, have also shown high rates of isoniazid (INH) resistance, especially in patients with a history of prior TB.[2,3] In addition to the INHr-TB problem, a significant number of TB patients are unable to use INH in their regimens. INH intolerance can be a significant problem, especially in adult populations, and although INH treatment is uncommonly associated with hepatitis, the use of this drug is problematic in patients with chronic liver disease.[4,5]

Several regimens have been used successfully to treat patients with INHr-TB in the past.[6] Treatment failures are rare when regimens containing rifampin (RIF) and two other first-line drugs for [is greater than or equal to] 6 months are used. A review[6] of 12 controlled trials conducted in Africa, Hong Kong, and Singapore described reduced relapse rates with the use of four or five drug treatment regimens, including RIF. Despite the numerous patients included in those studies, the optimal treatment regimen and duration for INHr-TB have not been exclusively addressed in a prospective, randomized fashion. The only exception is a clinical trial conducted in Kenya.[7] This study compared 6-month and 9-month regimens of daily direct observed therapy (DOT) with RIF plus ethambutol (EMB), including an initial 8-week daily treatment regimen of pyrazinamide (PZA) and streptomycin (SM). Treatment failures were low ([is less than] 1%). However, the relapse rates were relatively high (up to 21%) when the isolates were INH resistant and SM resistant, and/or the treatment was given for only 6 months.[7] A joint statement of the American Thoracic Society (ATS) and the Centers for Disease Control and Prevention (CDC) has recommended the use of a "6-month four-drug" treatment regimen as an effective regimen for INHr-TB,[8] based on a randomized trial performed in Singapore.[9] This clinical trial, performed about 20 years ago, included 33 patients with INHr-TB (with or without coexistence of SM-resistant TB) and had only 6 months of follow-up. The ATS/CDC statement also recommends the discontinuation of INH treatment, and the use of PZA for the entire 6 months of treatment. In cases in which PZA is not used, the treatment duration is extended to 9 months. When RIF and EMB are used solely, a 12-month regimen is recommended. Although the efficacy of intermittent short-course treatment has been well established for the treatment of pansensitive Mycobacterium tuberculosis,[10-12] the optimal treatment dose frequency for INHr-TB has not been elucidated at the present time.[8] The main aims of our study are to retrospectively describe and to attempt to evaluate the efficacy of the different treatment regimens used for INHr-TB in the southeastern Texas region.

MATERIALS AND METHODS

Patients

The study included all known patients who were reported to have INHr-TB isolates from different health facilities and clinics located in southeastern Texas to the health departments of the Houston and Tyler areas between 1991 and 1998. Additional INHr-TB patients were prospectively included from referrals to the Chest Clinics at Ben Taub General Hospital and the University of Texas Health Center at Tyler (referral centers for cases of drug-resistant TB in the southeastern Texas region) since early 1997. In addition, the existing patient information was complemented by the use of a comprehensive database from the Houston TB Initiative. The Houston TB Initiative is an ongoing, population-based, active surveillance and molecular epidemiology project encompassing all known reported cases of TB (approximately 2,580 cases up to January 1999) in Harris County, Texas since 1995. Treatment and follow-up information was gathered from the local TB control program records, patient charts, patient interviews, and telephone calls.

Exclusion Criteria

Exclusion criteria were as follows: inadequate clinical information, evidence of poor compliance during INHr-TB treatment course, uncompleted treatment, coexistence of resistance to other first-line drugs (excluding aminoglycosides), coinfection with mycobacteria other than tuberculosis, silicotuberculosis, and death after [is less than] 1 month of treatment. Inadequate clinical information was defined as insufficient patient data after searching all possible sources of case information related to patient TB-focus, clinical history, bacteriology, treatment, and posttreatment patient information.

Specimen Processing and Drug-Susceptibility Testing

All clinical specimens were processed similarly in all locations of the study according to the procedure manual of the Texas Health Department. Briefly, smear specimens were decontaminated with 4% NaOH solution and processed using the Truant's staining for fluorescent microscopic acid-fast bacilli determination.[13] All positive smear results were confirmed by overstaining with Ziehl-Neelsen stain unless oil was placed on the smear.[14] Primary isolation was performed using the BACTEC 460 instrument technique (Becton Dickinson Microbiology Systems; Sparks, MD) and culture in Middlebrook 7H10 or Lowenstein-Jensen media.[13,15] All isolates were niacin positive and nitrate reductase positive, as well as AccuProbe (GenProbe; San Diego, CA) positive.

Drug-susceptibility testing was performed locally twice and routinely repeated at the Texas Health Department for all patients reported as having smear-positive and/or culture-positive disease. Susceptibility testing was performed using the direct and indirect BACTEC methods (Becton Dickinson Microbiology Systems) and by an indirect method using Middlebrook 7H10 agar. The drugs (and breakpoint concentrations) tested by the BACTEC method included INH (0.4 [micro]g/mL), RIF (2.0 [micro]g/mL), EMB (2.5 [micro]g/mL), SM (2.0 [micro]g/mL), and PZA (50 [micro]g/mL) in acidic media.[15] The agents (and breakpoint concentrations) tested using Middlebrook 7H10 agar included INH (1.0 [micro]g/mL), RIF (1.0 [micro]g/mL), EMB (5.0 [micro]g/mL), SM (2.0 [micro]g/mL), kanamycin (5.0 [micro]g/mL and 6.0 [micro]g/mL), ethionamide (5.0 [micro]g/mL), rifabutin (2.0 [micro]g/mL), capreomycin (10.0 [micro]g/mL), and ciprofloxacin (2.0 [micro]g/mL). No INH dilution was performed with the liquid media. Drug-susceptibility testing was conducted by laboratories certified under the Federal Clinical Laboratory Improvement Act.[13] Drug resistance was defined as any degree of resistance reported to the local public health authorities according to the Texas Department of Health laboratory retesting results described above.

Clinical Outcomes and Definitions

All available clinical patient information was recorded, including patient demographics, clinical symptoms, medical history, comorbidities (including HIV status), social history, bacteriologic information, radiologic data, detailed treatment information, side effects, treatment outcomes, and posttreatment follow-up information.

Failure during treatment was defined in bacteriologic terms as the presence of at least two positive culture findings obtained during the last part of the patient treatment or after 4 months of adequate treatment. One positive culture finding associated with worsening of the patient's symptoms or TB-related death during treatment was also included in the definition of treatment failure. Relapse after stopping treatment was defined as the presence of at least two positive culture findings obtained in different months in any consecutive 3-month period. Relapse was also considered by the presence of one positive culture finding associated with recurrent patient symptoms after completing treatment. Adequate treatment was defined by a minimum of 6 months of DOT, the use of one of the currently ATS/CDC recommended regimens[8] for treatments started after 1994 when the statement was published, and the opinion of the patient's responsible physician attesting to appropriately completed treatment. Adherence to treatment was assessed by comparing the number of administered treatment doses against the number of treatment doses scheduled each month. Although there are no specific guidelines to define poor adherence in retrospective treatment studies in TB, poor compliance with treatment was defined as consistently missing [is greater than or equal to] 35% of monthly scheduled doses of medication in [is greater than or equal to] 2 months of treatment. Other end points measured, based on available information, included death during treatment, bacteriologic response assessed by routine monthly sputmn/specimen smear for acid-fast bacilli and culture as described above, radiologic response based on available chest radiograph reports, and the presence of side effects to treatment. Sputum conversion was defined as the first negative follow-up sputum culture finding.

The efficacy of the various anti-TB regimens was assessed based on (1) the prevention of emergence of drug resistance, treatment failure, and death, and (2) sterilization of the lesions, measured as the proportion of patients with (a) sputum conversion and maintenance as such, and (b) the prevention of relapse.

Statistics

Relevant patient clinical data including treatment regimen characteristics were analyzed against described clinical end points. Statistical significance was tested using either a [chi square] or Fisher's Exact Test according rules for parametric data.[16] Data were analyzed using software (SAS version 6.12; SAS Institute; Cary, NC).

RESULTS

Fifty-three of 83 patients (64%) reported to have INHr-TB were included in the study. The main reasons for patient exclusion are described in Table 1. Baseline patient characteristics are listed in Table 2. SM resistance coexisted with INH resistance in 20 patients (37.5% of isolates). There was a great variation between treatment regimens prescribed for INHr-TB patients in the study population, especially with regard to anti-TB drug regimen composition, frequency of administration, dosage, and length of treatment. Although the treatment regimens in general were different between patients, we attempted to group them arbitrarily into seven types according to drug regimen composition, dose frequency, and frequency of utilization (Table 3). Eighteen patients (34%) received RIF, PZA, and EMB thrice weekly for at least 9 months. Thirty-six patients (67.9%) had a duration of specific treatment, after INH-resistant determination, of [is greater than or equal to] 9 months. Only four patients (7.5%) had a total effective treatment duration of [is less than] 9 months. However, no patient had a total effective treatment duration of [is less than] 7 months, except one patient who died after 1 month of treatment. Thirty patients (56.6%) and 16 patients (30.2%) received thrice-weekly and daily treatments, respectively.

All 49 patients who had pulmonary TB involvement and documented sputum acid-fast bacilli smear achieved sputum conversion (Table 4). Treatment failure occurred in one patient (1.9%) who died and did not achieve a sputum-negative culture finding. This patient had AIDS with a CD4 count of 72, and received treatment with daily INH, RIF, PZA, and EMB for 1 month in the hospital. The patient died of respiratory insufficiency due to a combination of TB, aspiration pneumonia, and sepsis. Three patients (5.7%) experienced relapses: one patient had 7 months of treatment of undetermined dose frequency, and the other two patients had 11 months and 14 months of treatment, respectively, administered twice weekly. In the first case, the patient received PZA for the first 2 months, in addition to RIF and EMB for a total of 7 months, 5 months of which the patient received treatment in a chest hospital and the last 2 months by DOT. The second patient initially received twice-weekly treatment with INH, RIF, and PZA for 2 months followed by twice-weekly treatment with RIF, PZA, and EMB by DOT for the next 9 months, in addition to SM for the first 2 months of the 9 months. Except for missing 35% of the doses in his fourth month of treatment, his overall compliance was good. The third patient was initially treated in 1991 with INH, RIF, and PZA twice weekly for 5 months, followed by 3 months of treatment with RIF, PZA, and EMB twice weekly, and 6 additional months of twice-weekly RIF and EMB treatment. This third patient had virtually no missing scheduled doses of treatment. All patients with relapses were HIV-negative. Their adherence to treatment did not meet the definition of poor compliance according to the study exclusion criteria. The first two patients developed RIF resistance in addition to the original INH resistance. Follow-up time posttreatment ranged from 0 to 51.75 months with a mean of 7.21 months. Six patients (11%) developed side effects during treatment, four of which were attributed to PZA. In three of these patients, PZA treatment was discontinued.

Analysis of relevant patient clinical data, including treatment characteristics, against main clinical end points showed a statistical difference between total effective treatment time in patients with and without relapses (8.33 [+ or -] 1.15 months vs 11.06 [+ or -] 2.08 months, p [is less than] 0.02; Fig 1). Twice-weekly treatment was associated with relapse (p = 0.05; Table 5). The likelihood of relapse was not associated with any demographic feature, history of prior TB, lack of treatment before availability of drug susceptibility results (pretreatment period), HIV status, comorbidities, alcohol abuse, bacteriologic or radiologic feature, or other treatment characteristics. Treatment failure was not associated with any clinical or bacteriologic end point. No other clinically relevant and/or statistically significant association was found.

[GRAPH OMITTED]

DISCUSSION

Despite the introduction of modern chemotherapy decades ago, TB continues to be a global public health problem. Moreover, primary and secondary INH resistance has became a significant issue that appears to be worsening in different parts of the world.[2] In addition, INH intolerance is not an uncommon challenge facing physicians who treat patients with TB. These problems emphasize the need to define a standard, cost-effective, and well-proven regimen for the majority of INH-resistant and INH-intolerant patients.

Extensive and extraordinary clinical work has been done led mainly by the British Medical Research Council (BMRC) TB and Chest Disease Unit and collaborators in East Africa, Hong Kong, and Singapore, which sustains the foundation of our modern short-course TB therapy. Yet, the optimal treatment for INHr-TB patients is unclear and has not been exclusively evaluated in a prospective, randomized, and controlled fashion. Nevertheless, different treatment regimens for INHr-TB have been shown to be effective (by having reasonable low failure and relapse rates) according to data from INHr-TB patients from these trials and the prospective trial from Kenya.[6,7] However, a post hoc analysis from the BMRC trials included mostly primary INHr-TB patients, and the Kenya trial included only secondary INHr-TB patients. Both studies included patients treated under ideal protocol conditions, and excluded patients who had poor compliance or side effects that precluded protocol continuation without alterations in their treatment regimens. In addition, none of these studies has addressed the impact of the frequency of treatment dose in their analysis.

We attempted to retrospectively describe the treatment regimen characteristics and to analyze the efficacy of these schemes in normal working conditions in southeastern Texas. Houston, the largest metropolitan area in the study region, has a large Mexican, Central American, and Asian immigrant population. These immigrant groups have the highest prevalence rates of INHr-TB in the United States.[3] This may explain the high proportion of Hispanic and Asian persons found in the study sample (75%). Although a substantial amount of information was obtained retrospectively, a considerable portion, especially regarding treatment outcome and follow-up data, was obtained prospectively. Nevertheless, the sample size is similar to the extracted number of INHr-TB patients (with or without SM resistance) from each of the BMRC trials, which ranged from 14 to 86 patients.[6,9,11] Notably, 5 of 30 patients who were excluded from the study died, 3 of them before achieving i month of treatment. However, it is unclear whether TB was the main cause of death, and no further information was available to elucidate this issue.

Despite the aforementioned limitations, the relative heterogeneity of the treatment regimens, and the usual shortcomings of a descriptive retrospective cohort study, the study sample has shown an overall adequate treatment efficacy measured by the sputum conversion rate at 2 months (64.6%), treatment failure (1.9%), and relapse rate (5.7%). These rates are similar to those in the post hoc BMRC trials study that extracted patients with INHr-TB and/or SM resistance (86%, 4%, and 7%, respectively) and in which patients received at least four drugs including RIF (with or without PZA) for a minimum of 6 months of treatment.[6] The difference in sputum conversion at 2 months may be due to the probably greater number of patients with secondary INHr-TB (approximately 20%) and cavitary disease (45%) in our study population. Therefore, the slightly diminished sputum conversion rate in our study may be due to a slower treatment response in more chronic and extensive disease forms, compared to the population of the BMRC trials. Otherwise, the final sputum conversion rate was close to 100%.

In spite of the adequate treatment efficacy of the study population overall, there was great variation in the composition, duration, and dose frequency of the study treatment regimens. Therefore, it is virtually impossible to identify superior or inferior regimens given the type of study and the sample size. Nevertheless, almost 90% of treatment regimens included at least three drugs, such as RIF, PZA and EMB, that were administered under DOT for at least 9 months. However, no treatment regimen was given for [is less than] 7 months despite the extensive evidence of short-course treatment efficacy in pansensitive TB and the ATS/CDC recommendations.[8] Almost 87% of patients in the study received mostly thrice-weekly or daily treatment regimens. None of the patients who received daily and thrice-weekly RIF, PZA, and EMB in their treatment regimens for at least 9 months (30 of 53 patients) had negative outcomes.

Remarkably, two of four patients who received treatment twice weekly experienced relapses. A comparison with all other treatment dose frequencies yielded statistically significant association between relapse and twice-weekly treatment. If we eliminate the relapsing case of unknown treatment regimen from the analysis, the p value became 0.02 (Table 5). Thrice-weekly and daily treatment regimens were not associated with relapse. Although there were some compliance issues in one of these patients with a relapse, and treatment in the third patient with a relapse did not follow completely the ATS/CDC recommendations (his treatment was initiated 3 years before those were published), these shortcomings were not believed to be severe enough for exclusion from the study. Nevertheless, the twice-weekly treatment results and the association with relapse should be interpreted cautiously, given the type of study and the shortcomings presented. Although to our knowledge there is no prospective study comparing the efficacy of twice-weekly vs thrice-weekly treatment in INHr-TB, virtually all treatment regimens that included RIF (with or without PZA) for [is greater than or equal to] 4 months in all the BMRC-sponsored trials received daily and/or thrice-weekly treatment.[6] Those treatment regimens had the smallest treatment failure and relapse rates for INHr-TB with or without SM resistance. When RIF was given for 2 months and the continuation phase was given twice weekly, such as in two arms of the second BMRC/Hong Kong trial, then the relapse rate increased up to 11% and 36% with the use of PZA and EMB, respectively, in conjunction with SM and INH.[17] Even though relatively weaker regimens did not include RIF in the continuation phase, the findings may still suggest that twice-weekly treatment may not be as effective as daily and thrice-weekly treatments for INHr-TB. This is not the case for pansensitive TB patients treated with twice-weekly regimens, which have been shown to be at least as effective as the daily regimens.[10]

Interestingly, there was a statistical difference between total effective treatment time in patients with and without relapses (Fig 1). This result should be interpreted with caution because of the sample size effect, the shortcomings of a retrospective study, and the heterogeneity of the treatment regimens as a confounding factor in our study. Nevertheless, it is reasonable to postulate that a shorter duration of treatment may increase the chances of relapse in INHr-TB patients if the treatment regimen associated with relapse is not highly effective. To our knowledge, the Kenya trial was the only prospective study that compared treatment times in patients with INHr-TB. This trial, which included 226 patients with only INHr-TB with or without additional SM resistance, may support the hypothesis of treatment for [is greater than] 6 months.[7] If we rearrange the Kenya trial data to group all the patients who received 4-month RIF and EMB treatment (n = 86) and 7-month RIF and EMB treatment (n = 92) in the continuation phase mixing INHr-TB patients and SM-resistant TB patients, then the resulting new proportions of relapse cases are 6 of 86 patients (6.9%) and 2 of 92 patients (2.1%) for the 6-month and 9-month treatment periods, respectively (p [is less than] 0.05). However, the third Hong Kong study included 78 patients with INHr-TB with or without SM resistance.[11] The trial tested five treatment regimens: INH-RIF-SM-PZA-EMB (thrice weekly), INH-RIF-SM-PZA (thrice weekly), INH-RIF-SM-EMB (thrice weekly), INH-RIF-PZA-EMB (thrice weekly), and INH-RIF-PZA-EMB (daily), all for 6 months. The only regimen that showed a higher relapse rate compared to all the others was the one using INH, RIF, SM, and EMB thrice weekly (4 of 21 patients; 21%). These findings also suggest that using less effective regimens such as RIF and EMB thrice weekly as the only effective treatment for INHr-TB for [is less than] 6 months is inadequate. In addition, the same study points out the importance of receiving PZA, in addition to RIF, throughout the treatment,[11] or at minimum during the intensive phase, as shown by others.[6]

Finally, the overall rate of side effects (11.4%) was within the expected number, compared to similar values reported in the third (6-month thrice-weekly and daily) Hong Kong/BMRC trial.[11] Interestingly, in our study as well as in the BMRC trial, PZA was the drug most frequently associated with interruption of treatment and termination of the drug treatment from the regimen. However, the rate of side effects related to PZA was not high enough to preclude the use of this important drug.

CONCLUSION

Several types of treatment regimens for INHr-TB have been used in southeastern Texas since 1992. INHr-TB treatment regimens have for the most part been for [is greater than] 6 months, and the overall treatment efficacy was adequate. Both twice-weekly and once-daily treatments had similar treatment outcomes. A prospective study with different treatment durations and comparing twice-weekly against thrice-weekly and/or once-daily dose frequency is needed to determine the optimal treatment for INHr-TB patients.

ACKNOWLEDGMENT: We thank all personnel at the Ben Taub General Hospital and at the University of Texas at Tyler Chest clinic for their kind and helpful collaboration. We also kindly acknowledge the assistance of the workers at the Tuberculosis Control Unit of the City of Houston, the Houston TB Initiative, and the Texas Department of Health laboratory personnel.

REFERENCES

[1] Iseman MD, Sbarbaro JA. Short-course chemotherapy of tuberculosis: hail Britannia (and friends) [editorial]! Am Rev Respir Dis 1991; 143:697-698

[2] Pablos-Mendez A, Raviglione MC, Laszlo A, et al. Global surveillance for antituberculosis-drug resistance, 1994-1997. N Engl J Med 1998; 338:1641-1649

[3] Moore M, Onorato IM, McCray E, et al. Trends in drug-resistant tuberculosis in the United States, 1993-1996. JAMA 1997; 278:833-837

[4] Molding TS, Redeker AG, Kanel GC. Twenty isoniazid-associated deaths in one state. Am Rev Respir Dis 1989; 140:700 -705

[5] Ormerod LP, Horsfield N. Frequency and type of reaction to antituberculosis drugs: observation in routine treatment. Tubercle Lung Dis 1996; 77:37-42

[6] Mitchison DA, Nunn AJ. Influence of initial drug resistance to short-course chemotherapy of pulmonary tuberculosis. Am Rev Respir Dis 1986; 133:423-430

[7] Babu Swai O, Aluoch JA, Githui WA, et al. Controlled clinical trial of a regimen of two durations for the treatment of isoniazid resistant pulmonary tuberculosis. Tubercle 1988; 69:5-14

[8] American Thoracic Society. Treatment of tuberculosis and tuberculosis infection in adults and children. Am J Respir Crit Care Med 1994; 149:1359-1374

[9] Singapore Tuberculosis Service/British Medical Research Council. Clinical trial of six-month and four-month regimens of chemotherapy in the treatment of pulmonary tuberculosis. Am Rev Respir Dis 1979; 119:579-585

[10] Castelo A, Jardim JRB, Goihman S, et al. Comparison of daily and twice-weekly regimens to treat pulmonary tuberculosis. Lancet 1989; 2:1173-1176

[11] Hong Kong Chest Service/British Medical Research Council. Controlled trial of four thrice-weekly regimens and a daily regimen all given for 6 months for pulmonary tuberculosis. Lancet 1981; 1:171-174

[12] Hong Kong Chest Service/British Medical Research Council. Controlled trial of 6-month and 9-month regimens of daily and intermittent streptomycin plus isoniazid plus pyrazinamide for pulmonary tuberculosis in Hong Kong: the results up to 30 months. Am Rev Respir Dis 1977; 115:727-735

[13] Centers for Disease Control and Prevention. Laboratory practices for diagnosis of tuberculosis-United States, 1994. MMWR Morb Mortal Wkly Rep 1995; 44:587-591

[14] Truant JP, Brett WA, Thomas W Jr. Fluorescence microscopy of tuberculosis bacilli stained with auramine and rhodamine. Henry Ford Hosp Med Bull 1962; 10:287-296

[15] Heifets LB, Good RC. Current laboratory methods for the diagnosis of tuberculosis. In: Bloom BR, ed. Tuberculosis: pathogenesis, protection, and control. Washington, DC: ASM Press, 1994:85-109

[16] Dawson-Saunders B, Trapp RG. Basic clinical biostatistics. 2nd ed. Norwalk, CT: Appleton & Lange, 1994; 143-161

[17] Hong Kong Chest Service/British Medical Research Council. Controlled trial of 6-month and 8-month regimens in the treatment of pulmonary tuberculosis: the results up to 24 months. Tubercle 1979; 60:201-210

(*) From the Sections of Pulmonary and Critical Care (Drs. Escalante and Awe) and Infectious Diseases (Dr. Graviss), Institute for the Study of Human Bacterial Pathogenesis, Baylor College of Medicine, Houston, TX; the Center for Pulmonary and Infections Disease Control (Dr. Griffith), University of Texas Health Center at Tyler, Tyler, TX; and the Laboratory of Human Bacterial Pathogenesis (Dr. Musser), Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Hamilton, MT.

Support was provided by US Public Health Services Grant DA-09238 (to J.M.M.).

Manuscript received March 23, 2000; revision accepted January 10, 2001.

Correspondence to: Patricio Escalante, MD, Assistant Professor, Division of Pulmonary and Critical Care, Department of Medicine, Keck School of Medicine, University of Southern California, Los Angeles County Medical Center, 1200 N. State St, GNH 11900, Los Angeles, CA 90033; e-mail: Patricioe@aol.com

COPYRIGHT 2001 American College of Chest Physicians
COPYRIGHT 2001 Gale Group