Multidrug-resistant tuberculosis: An update on the best regimens

April 1, 2006

Abstract: Multidrug-resistant tuberculosis is defined as tuberculosis caused by strains that have documented in vitro resistance to isoniazid and rifampin. Treatment involves a regimen consisting of at least 4 or 5 drugs to which the infecting strain has documented susceptibility. These agents may include ethambutol, pyrazinamide, streptomycin, a fluoroquinolone, ethionamide, prothionamide, cycloserine, and para-aminosalicylic acid. In addition, an injectable agent, such as kanamycin, amikacin, or capreomycin, should be used until negative sputum cultures have been documented for at least 6 months. If the patient has severe parenchymal damage, high-grade resistance, or clinically advanced disease, also consider clofazimine, amoxicillin/clavulanate, or clarithromycin, although there is little evidence supporting their efficacy in this setting. Routine monitoring includes monthly sputum smear and culture testing, monthly assessment of renal function and electrolyte levels, and liver function tests every 3 to 6 months. (J Respir Dis. 2006;27(4):172-182)

Multidrug-resistant tuberculosis (MDRTB) is a growing problem globally and can be difficult to manage clinically. It is especially difficult to treat in areas where adequate drug susceptibility testing, appropriate second-line drugs, and support services are unavailable. Unfortunately, it is in these set-tings that the incidence of MDRTB is increasing.

Even in resource-rich settings, MDRTB presents clinical and logistic challenges, because the treatment is long and requires the use of several relatively toxic drugs. Over the past 2 decades, experience in the treatment of MDRTB has grown, even if new drug development has not kept pace.

In this article, we will explore the epidemiology, diagnosis, and management of MDRTB, based on our collective experience managing this disease in the United States, Peru, Haiti, and Russia.1


MDRTB is defined as tuberculosis caused by strains of Mycobacterium tuberculosis that have documented in vitro resistance to at least isoniazid and rifampin.2 Such resistance is clinically significant because these 2 drugs represent the most powerful antituberculous agents currently available.

Once the mycobacterium is no longer susceptible to killing by these drugs, management of the disease becomes more complicated, because the remaining drugs are less effective and more toxic.3 Resistance to other combinations of drugs--polyresistance--is generally easier to eradicate.

Resistance to antituberculous drugs can occur in 1 of 2 ways. Acquired drug resistance occurs when the patient is infected with a strain that is initially susceptible but becomes resistant during the course of inadequate therapy. Organisms that have acquired resistance to an antituberculous drug (such as isoniazid) are not sufficiently suppressed by other companion drugs (such as rifampin and pyrazinamide) to prevent resistant organisms from predominating, which leads to treatment failure. These companion drugs may be unavailable, or they may be prescribed inappropriately, taken incorrectly, or malabsorbed.

Organisms already resistant to one drug can acquire additional resistance mutations to another drug. If these organisms are not suppressed by other agents, resistance increases in a stepwise fashion.

The other type of resistance, known as primary resistance, occurs when a patient becomes infected with a strain that is already drug-resistant. Both primary and acquired resistance are responsible for the increase in MDRTB seen globally.4 The contribution of primary resistance to the global pandemic is probably underestimated, since most patients do not undergo drug susceptibility testing before receiving therapy for tuberculosis.5


MDRTB is a significant problem in many countries and has been documented in all countries that have been surveyed.6 In some areas, known as "hotspots," there is a high prevalence (more than 3% of tuberculosis cases) of MDRTB. These areas include the Dominican Republic, Latvia, and parts of the former Soviet Union.7 In the United States, persons from these areas are at increased risk for MDRTB.

In addition, immigrants, students, and visitors from areas such as Mexico, India, the Far East, and especially the Philippines are at higher risk for MDRTB.Other risk groups include persons with HIV infection and those who are known contacts of persons with MDRTB.8

The incidence of MDRTB appears to be declining in the United States and in other countries where the incidence of tuberculosis is declining; however, the incidence of MDRTB appears to be increasing in many other parts of the world where tuberculosis is less well controlled.9 However, since an increasing proportion of persons with tuberculosis in the United States, Canada, Western Europe, and Australia are foreign-born, the problem of MDRTB is universal, and an international approach is needed.10


The only way to diagnose MDRTB is by testing an isolate from the patient to determine whether the mycobacteria can grow in the presence of isoniazid, rifampin, and other antituberculous agents.11 Conventional testing methods usually require the growth of mycobacteria in solid culture media, a process that can take up to 2 months. Growth characteristics are then compared between cultures that contain drugs and those that do not.

More rapid radiometric methods can also be used, in which mycobacteria are inoculated into a radiolabeled broth, with and without drug, then growth in the broth is measured.12 Genotypic methods for detecting specific genetic mutations responsible for certain types of drug resistance are available and allow for more rapid testing to be performed.13 Genotypic methods are especially useful for rifampin resistance. However, not all known genetic mutations are accounted for in resistance testing, thus limiting this modality.


The treatment for MDRTB is more difficult than that for its pan-susceptible counterpart.14 Because isoniazid and rifampin cannot be used, less powerful agents--the second-line drugs--are required. These drugs are less potent, so a larger number of drugs are needed for a longer time. Table 1 lists the drugs used for treating MDRTB and the usual dosages.

To cure patients of MDRTB, multidrug regimens consisting of a minimum of 4 or 5 drugs to which the infecting strain has documented susceptibility ideally should be used for a minimum of 18 to 24 months.15 A daily injectable medication should be used until negative sputum cultures have been documented for at least 6 months. Our strategy for building a regimen is as follows16:

• Use ethambutol or pyrazinamide if susceptibility to the drug has been documented.

• Include an injectable agent until sputum cultures are negative for a minimum of 6 months.

• Use a fluoroquinolone (especially moxifloxacin or gatifloxacin) whenever possible.

• Add other second-line agents to reach a minimum of 4 or 5 drugs.

• If there is severe parenchymal damage, high-grade resistance, or clinically advanced disease, consider the use of reinforcing agentsthat show in vitro evidence of antimycobacterial activity. Surgery is also used for select patients with localized lung destruction.

Because successful treatment is critical, therapy must be directly observed and should be given a minimum of 6 days per week.In addition to antituberculous therapy, all patients should receive oral pyridoxine, 50 mg/d.

While awaiting results of drug susceptibility testing, it is often necessary to institute empiric therapy, depending on the clinical status of the patient. When initiating empiric therapy, it is important to build on a solid foundation of at least 4 or 5 agents to which the infecting strain is likely to be susceptible. Knowledge of local resistance patterns and the patient's known contacts, and avoidance of drugs that the patient has previously received are important principles in the design of empiric therapy. Regimens are usually adjusted once drug susceptibility test results are available.17


Designing a treatment regimen

Case 1: LR was a 41-year-old man with tuberculosis, who presented with cough, fever, and weight loss. His chest radiograph demonstrated biapical cavitation, fibrosis, and infiltrates (Figure 1).He was started on a regimen of isoniazid, rifampin, pyrazinamide, and ethambutol, but he continued to have symptoms, and his sputum smears remained positive.

A sputum specimen was sent for drug susceptibility testing. While awaiting the results, the patient was given empiric MDRTB treatment with streptomycin, ciprofloxacin, ethionamide, cycloserine, and para-aminosalicylic acid.

Two months later, his symptoms had improved and his sputum specimen was smear-negative. Drug susceptibility tests confirmed resistance to isoniazid and rifampin; pyrazinamide and ethambutol were added to his regimen, and para-aminosalicylic acid was discontinued. He continued to respond well, and streptomycin was stopped after 6 negative sputum cultures were obtained. He completed a total of 24 months of therapy, during which sputum smears and cultures remained negative.

Adverse effects

Second-line agents and multidrug regimens for MDRTB often result in adverse events (Table 2). When managing these events, it is important to weigh the adverse effect against the potential compromise of treatment outcomes. Research has shown that most adverse events can be managed in an outpatient setting without compromising the efficacy of treatment.18

Case 2: PP was a 36-year-old woman with tuberculosis that showed resistance to isoniazid, rifampin, pyrazinamide, ethambutol, and streptomycin. She was started on a regimen of kanamycin, ciprofloxacin, cycloserine, ethionamide, and para-aminosalicylic acid.

During the first 2 weeks of treatment, the patient complained of burning epigastric pain, nausea, and vomiting. She had normal liver function test results. She was given an H2 blocker to take 2 hours before her antituberculous medications and an antiemetic to be taken 30 minutes before treatment. Her symptoms improved, and by month 2, she was weaned off the H2 blocker and the antiemetic. Subsequent sputum smears and cultures were negative.

During her ninth month of therapy, she complained of fatigue, dysmenorrhea, dry skin, and hair loss. Her serum thyroid-stimulating hormone (TSH) level was elevated as a result of the ethionamide and para-aminosalicylic acid. Thyroid hormone replacement therapy was started, and the antituberculous therapy was continued.

The patient's TSH level normalized, she completed 24 months of MDRTB therapy, and thyroid hormone replacement therapy was continued until 2 months after antituberculous medications were stopped. A subsequent TSH level was normal.

Monitoring during therapy

Close follow-up of patients during treatment is essential. Patients should be seen daily to receive directly observed therapy. This can usually be done by a trained community health worker, who can also learn to screen for adverse events.19 Monthly clinical visits should occur while the patient is receiving injectable therapy, and then every 2 to 3 months thereafter.

Table 3 lists the baseline laboratory tests that should be evaluated during treatment. All patients should have sputum smear and culture testing monthly.20

HIV infection

Since poor outcomes from MDRTB are more likely to occur in patients with HIV infection,21 aggressive treatment of both diseases should be pursued. Careful monitoring is needed to avoid drug-drug interactions and the added toxicities of therapy. Community-based antituberculous and antiretroviral therapy should be considered for all patients with CD4+ cell counts of less than 200/µL; this approach has been demonstrated to be successful in some settings.22

Case 3: RS was a 23-year-old man with tuberculosis that was resistant to isoniazid, rifampin, and pyrazinamide. HIV test results were positive, and his CD4+ cell count was 187/µL. His chest radiograph is shown in Figure 2. He was given trimethoprim/sulfamethoxazole prophylaxis and antituberculous therapy with ethambutol, streptomycin, ciprofloxacin, cycloserine, and ethionamide. Smears and cultures of the sputum were negative after the first month of therapy.

One month after antituberculous therapy was initiated, the patient began antiretroviral therapy with lamivudine, stavudine, and nevirapine. Two weeks later, a rash developed and the nevirapine was replaced by efavirenz; the rash resolved. The patient tolerated this therapy well and continued to have negative sputum smears and cultures. He was monitored closely for psychiatric adverse effects because of the combined use of cycloserine and efavirenz.

Pediatric MDRTB

MDRTB in children is a serious global problem. The diagnosis in children may be more difficult than in adults because of the difficulty in obtaining sputum samples. Although some second-line antituberculous agents are not recommended for use in children, we have had little problem using them in some children. In general, we have found that the best approach in children is an aggressive, weight-based dosing schedule that relies on the same principles as the treatment of MDRTB in adults.23

Case 4:TJ was a 4-year-old girl whose uncle had tuberculosis with documented resistance to isoniazid, rifampin, and pyrazinamide. She was noted to have kyphosis and received a diagnosis of Pott disease. Her chest radiograph is shown in Figure 3.Based on her contact's resistance pattern, she was started on an empiric MDRTB treatment regimen of ethambutol, ciprofloxacin, ethionamide, cycloserine, and capreomycin, using weight-based dosing.

Gastric aspirates obtained when therapy was initiated confirmed MDRTB caused by strains resistant to isoniazid, rifampin, and pyrazinamide. The treatment regimen was continued for 24 months, with regression of the patient's spinal lesions. Repeated testing of gastric aspirates yielded negative results.


In pregnant women, there is an increased risk of disease progression, fetal loss, and postpartum contagion if MDRTB is not treated. This risk must be weighed against the potential teratogenic effects of some second-line antituberculous agents and the lack of data regarding the safety of most second-line drugs. We have had success in a small population of pregnant women with MDRTB using a strategy that defers the use of potentially toxic medications until after the first trimester while at the same time advocating an aggressive regimen that will ensure sputum sterilization and clinical improvement by the time of delivery.24

Case 5:ES was a 19-year-old woman who received a diagnosis of tuberculosis during her first trimester of pregnancy. Her chest radiograph revealed a small, right upper lobe infiltrate. Several of her family members had MDRTB; therefore, her sputum specimen was sent for drug susceptibility testing. She was clinically stable, and empiric therapy was deferred.

During her second trimester, however, she experienced hemoptysis; a repeated chest radiograph revealed right upper lobe cavitation and more extensive bilateral disease (Figure 4). Drug susceptibility testing showed resistance to isoniazid, rifampin, ethambutol, and streptomycin, and she was started on a regimen of pyrazinamide, ciprofloxacin, para-aminosalicylic acid, ethionamide, and amoxicillin/ clavulanate.

The patient's sputum cultures and smears converted to negative in the second month of therapy. She gave birth to a healthy boy. After delivery, kanamycin was added to her treatment regimen.

Surgical therapy

Surgery, including collapse therapy (therapeutic pneumothorax) has long been advocated as a complement to the medical treatment of tuberculosis. Surgery may be particularly beneficial in patients with high-grade resistance or complications of the disease (such as hemoptysis), or when sputum cultures and smears fail to convert.25 In such patients, portions of the lungs are often extensively destroyed by chronic disease.

In general, surgery is more beneficial when performed earlier rather than later in the course of disease. The role of nonresective procedures, such as therapeutic pneumothorax and thoracoplasty, and surgery in the setting of bilateral disease are still being debated.

Case 6:NM was a 26-year-old man with MDRTB and documented resistance to isoniazid, rifampin, pyrazinamide, ethambutol, streptomycin, kanamycin, and ciprofloxacin. He was started on a regimen of capreomycin, ethionamide, cycloserine, para-aminosalicylic acid, amoxicillin/clavulanate, and clofazimine, but sputum smear and culture results failed to convert by the third month of treatment. The patient's chest radiograph showed extensive left lung destruction (Figure 5).

Given the patient's high-grade resistance, localized disease, and failure to convert, he was referred for adjuvant surgery. In month 4 of treatment, he underwent a left upper lobectomy. Sputum became smear- and culture-negative, and his medical regimen was continued for an additional 18 months.

Treatment setting

In most resource-rich countries,patients with MDRTB are treated by specialists in tertiary care centers. MDRTB treatment requires experience with drugs that are not commonly used in low-burden countries. Treatment centers must also be able to ensure adherence to therapy and good follow-up.

Patients are often hospitalized, and their contacts protected by use of negative-pressure airflow rooms and particulate respirators. In poor countries, however, isolation is often not possible and nosocomial spread of the disease can occur.26 Furthermore, the monetary costs of hospitalization often make treatment of MDRTB prohibitive in poor countries.

We have had great success in 3 countries treating patients with MDRTB using a community-based approach, with cure rates approaching 80%.27 The bulk of the work is performed by trained community health workers, who not only provide MDRTB medications but also provide a wide range of social, nutritional, and economic support.28 This comprehensive community-based care has allowed us to implement MDRTB treatment in settings where it had been deemed "impossible" to achieve cure.



1. Farmer P, Furin J, Shin S. Managing multidrug-resistant tuberculosis.

J Respir Dis.

2. Iseman MD, Madsen LA. Drug-resistant tuberculosis.

Clin Chest Med.

3. Iseman MD. Management of multidrug- resistant tuberculosis.


1999; 45(suppl 2):3-11.
4. Chevrel-Dellagi D, Abderrahman A, Haltiti R, et al. Large-scale DNA fingerprinting of

Mycobacterium tuberculosis

strains as a tool for epidemiologic studies of tuberculosis.

J Clin Microbiol.

5. Crofton J, Chaulet P, Maher D.

Guidelines on the Management of Drug-Resistant Tuberculosis.

Geneva: World Health Organization; 1996.
6. Pablos-Mendez A, Raviglione MC, Laszlo A, et al. Global surveillance for antituberculosis drug resistance, 1994-1997. World Health Organization-International Union against Tuberculosis and Lung Disease Working Group on Anti-tuberculosis Drug Resistance Surveillance.

N Engl J Med.

7. World Health Organization. Anti-tuberculosis drug resistance in the world: the WHO/ IUATLD global project on antituberculosis drug resistance surveillance 2000. Geneva: World Health Organization; 2000.
8. Centers for Disease Control. Meeting the challenge of multidrug-resistant tuberculosis: summary of a conference.


1992; 41(RR-11):49-57.
9. Program in Infectious Disease and Social Change.

The Global Impact of Drug Resistant Tuberculosis

. Boston: Harvard Medical School and the Open Society Institute; 1999.
10. Farmer P, Bayona J, Becerra M, et al. The dilemma of MDR-TB in the global era.

Int J Tuberc Lung Dis.

11. Heifets LB, Cangelosi GA. Drug susceptibility testing of



a neglected problem at the turn of the century.

Int J Tuberc Lung Dis.

12. Grange JM, Festenstein F. The human dimension of tuberculosis control.

Tuber Lung Dis

. 1993;74:219-222.
13. Drobniewski FA, Wilson SM. New biomolecular assays must be tested by direct study in the developing world.


14. Goble M, Iseman MD, Madsen LA, et al. Treatment of 171 patients with pulmonary tuberculosis resistant to isoniazid and rifampin.

N Engl J Med.

15. Iseman MD.

A Clinician's Guide to Tuberculosis.

Philadelphia: Lippincott Williams & Wilkins; 2000.
16. Partners In Health, Harvard Medical School, the Bill and Melinda Gates Foundation.

A DOTS-Plus Handbook: Guide to the Community-Based Treatment of MDR TB.

Boston: Partners In Health; 2002.
17. Mukherjee J, Shin S, Furin J, et al. New challenges in the clinical management of multidrug-resistant tuberculosis.

Infect Dis Clin Pract.

18. Furin JJ, Mitnick CD, Shin SS, et al. Occurrence of serious adverse effects in patients receiving community-based therapy for multidrug-resistant tuberculosis.

Int J Tuberc Lung Dis.

19. Farmer P, Kim JY. Community based approaches to the control of multidrug resistant tuberculosis: introducing "DOTS-plus."


20. Partners In Health, Harvard Medical School, the Bill and Melinda Gates Foundation.

Dots-Plus at a Glance: Protocols and Norms for the Management of Multidrug- Resistant Tuberculosis.

Boston: Partners In Health; 2003.
21. Centers for Disease Control. Outbreak of multidrug-resistant tuberculosis--Texas, California, and Pennsylvania.


1990; 39: 369-372.
22. Farmer P, Leandre F, Mukherjee JS, et al. Community-based approaches to HIV treatment in resource-poor settings.


. 2001; 358:404-409.
23. Mukherjee J, Joseph K, Rich M, et al. Clinical and programmatic considerations in the treatment of MDR-TB in children: a series of 16 patients from Lima, Peru.

Int J Tuberc Lung Dis.

24. Shin S, Guerra D, Rich M, et al. Treatment of multidrug-resistant tuberculosis during pregnancy: a report of 7 cases.

Clin Infect Dis.

25. Pomerantz M, Madsen L, Goble M, Iseman M. Surgical management of resistant mycobacterial tuberculosis and other mycobacterial pulmonary infections.

Ann Thorac Surg.

26. Iseman MD. Tailoring a time-bomb. Inadvertent genetic engineering.

Am Rev Respir Dis.

27. Mitnick C, Bayona J, Palacios E, et al. Community-based therapy for multidrug- resistant tuberculosis in Lima, Peru.

N Engl J Med.

28. Palacios E, Guerra D, Llaro K, et al. The role of the nurse in the community-based treatment of multidrug-resistant tuberculosis (MDR-TB).

Int J Tuberc Lung Dis.