Drug Resistant Tuberculosis

In early 2008, the World Health Organization (WHO) reported an unexpectedly large increase in the number of cases of drug-resistant tuberculosis (1). At present, an estimated 5% of the more than 9 million persons who develop tuberculosis (TB) around the world every year are infected with a multi-resistant strain of tuberculosis (multidrug-resistant tuberculosis, MDR-TB), i.e., a strain that is resistant to (at least) the two most powerful anti-tuberculosis drugs that are currently available, isoniazid and rifampicin. The current WHO report (1) also contains data on extensively drug-resistant tuberculosis (XDR-TB), which was first -described in 2006. By definition, XDR-TB is MDR-TB that is additionally resistant to at least one of the fluoroquinolones and to one of the three injectable second-line anti-tuberculosis drugs, amikacin, kanamycin, and capreomycin (2, e1).

In the 1970’s, TB was thought to have been nearly vanquished, yet it is now the most deadly bacterial (and curable) infectious disease around the world. The main reason for this is co-infection with tuberculosis and HIV, a condition that is most common in sub-Saharan Africa but is also on the rise in other regions of the world, including Europe (e2).

Despite positive epidemiological developments in Germany, the effects of the worldwide trend are making themselves felt here as well, as half of the tuberculosis patients in Germany were born abroad (3, e3).

This article is based on a selective review of the literature.

Its learning objectives for readers are

• to obtain insight into the problem of resistance development in tuberculosis;
• to master the fundamentals of the diagnosis and treatment of drug-resistant tuberculosis;
• to be able to assess epidemiological developments in this area.
Epidemiological developments

The WHO estimates that there were 9.27 million new cases of tuberculosis and 1.78 million deaths from tuberculosis worldwide in 2007 (4). Figure 1 (gif ppt) shows the regional incidences of tuberculosis, while Table 1 (gif ppt) contains statistics on the 22 “high-burden” countries (1, 4).

In the WHO European region—a geographical region defined by the WHO, consisting of 53 countries—nearly 500 000 new cases of tuberculosis were registered in 2006, with a clear east-west gradient: the incidence ranged from 282 per 100 000 population in Kazakhstan to 5.5 per 100 000 population in Sweden (e4).

Germany, with 6.1 new cases of tuberculosis per 100 000 population per year (2007; n = 5020), is one of the low-incidence countries, with a mortality rate of 0.2/100 000 (3). 43% of these patients were born outside the country. In 2007, the incidence of tuberculosis among citizens of foreign countries living in Germany was five times that among German citizens (22.8 versus 4.2 per 100 000 population) (3). Resistance rates rose slightly from 2001 to 2005, then fell again somewhat in 2006 and 2007, perhaps because of a decline in immigration, especially from the newly independent countries that formerly belonged to the Soviet Union (3) (Figure 2 gif ppt). Among patients from these countries, the rate of MDR was 12.8%, and that of “any drug resistance” (i.e., resistance to any one of the five standard drugs) was 34.6%; the corresponding figures for persons born in Germany were 0.6% and 7.5%, -respectively (3). 53 of the 66 patients diagnosed with MDR tuberculosis in 2007 were born outside of the country, 38 of them in countries that formerly belonged to the Soviet Union (3). Precise data on XDR tuberculosis are not available, because findings regarding resistance to second-line drugs are not routinely reported; XDR is thought to account for fewer than 5% of MDR cases (5, e5, e6).

Resistance to anti-tuberculosis drugs has been observed around the world, yet valid and comprehensive data are often unavailable (2, 6, e7). Data from 116 countries and regions for the year 2006 (2 509 545 tuberculosis patients) (1) yield a 2.9% rate of MDR -tuberculosis among new cases of the disease (Figure 3 gif ppt) and a 15.3% rate among persons who have received at least one month of treatment for tuberculosis. The “combined” resistance rate of 5.3% corresponds to nearly half a million patients with MDR tuberculosis around the world.

Data on resistance to second-line drugs in the year 2006 were collected from international reference laboratories in 49 countries. Among 17 000 bacterial isolates, the rate of MDR was 20% and that of extensive resistance was 2% (e8), although selection bias cannot be excluded. The percentage of XDR varies markedly; it is currently highest in Estonia, at 24% (1). XDR tuberculosis had been diagnosed in 45 countries by June 2008 (eFigure gif ppt). Tuberculosis with extremely extensive resistance (XXDR tuberculosis), i.e., resistance to practically all of the anti-tuberculosis drugs, has been seen in only a few cases to date (e9).

Reasons for the development of resistance
Impressive therapeutic outcomes were seen when streptomycin, the first anti-tuberculosis drug, was introduced in 1944. There were, however, many recurrences of tuberculosis thereafter, because of the selection of streptomycin-resistant bacterial strains by monotherapy (e10, e11). The more widespread tuberculosis is in the patient’s body, the greater the number of bacteria that are present, and the more likely it is that some of the pathogenic organisms will contain spontaneous mutations conferring drug resistance (e12).

The need for combination therapy against tuberculosis was recognized after the introduction of para-aminosalicylic acid in 1944 and that of isoniazid in 1952 (e13), as the rate of mutations conferring resistance to multiple drugs is very low. Furthermore, combination therapy can better reach bacteria with different levels of metabolic activity at multiple sites in the body. The treatment must be continued long enough to kill quiescent bacteria (“dormant persisters”) as well.

The next drugs to be introduced were pyrazinamide and cycloserine in 1952, capreomycin in 1960, ethambutol in 1961, and rifampicin in 1966. The introduction of rifampicin and pyrazinamide enabled a marked shortening of the duration of therapy, from 18–24 to 6 months (“short-term chemotherapy”), provided that the patient’s tuberculosis is fully drug-sensitive. The recurrence rate after such treatment is less than 5% in patients who take all their medications correctly every day as prescribed (e13).

Faulty prescriptions, treatment compliance problems, inadequate intestinal resorption of drugs, and poor drug quality are factors that can promote the development of resistance (1, 2).

Drug resistance was first recognized as a major problem in 1992, when 12% of the tuberculosis patients in New York City were found to have MDR tuberculosis (e14). MDR tuberculosis spread around the world (1) because of the lack or inadequacy of tuberculosis control programs, insufficient resources, and inadequate protective measures against infection, as well as delayed diagnosis of tuberculosis (7).

The following are special risk factors for MDR tuberculosis:

• prior treatment with anti-tuberculosis drugs
• immigration from an area where MDR tuberculosis is highly prevalent, or contact with MDR tuberculosis patients
• imprisonment
• possibly, HIV infection (1, 3, 6, 8, e15, e16).

Prisons require special attention, particularly in the Newly Independent States of the former Soviet Union (9). In the WHO European region, the mean case-reporting rate for tuberculosis among prisoners is 232 per 100 000 population, while the country-specific rates are highest in Kazakhstan and Azerbaijan, at 17,808 and 3944 per 100 000 population, respectively (e15). Even though there is a downward trend in some countries (Russian Federation, 1999–2005: from 4000 to 1591 cases per 100 000 population), the high rates of MDR—sometimes accounting for more than 30% of overall incidence—and the rising prevalence of HIV are causes for concern (9). Prisoners have been found to have higher rates of MDR in Western, industrialized countries as well (e15, e17).

In some regions of the world, the so-called Beijing genotype of Mycobacterium (M.) tuberculosis is associated with a high resistance rate and, in particular, with a high MDR rate (the “W strain”) (e18). These strains may be more virulent, and/or more likely to mutate, and/or able to spread more easily because of poorer tuberculosis control in the areas to which they are endemic.

Resistant tuberculosis and co-infection with HIV
It is estimated that, in 2007, 1.37 of the 9.27 million persons with new tuberculosis infections were co-infected with HIV (14.8%), and 456 000 persons died of tuberculosis in the presence of an HIV infection (4). In some countries of sub-Saharan Africa, the -tuberculosis/HIV co-infection rate has risen dramatically, to 50–80% (4, 10, e19). HIV-positive persons carrying a latent M. tuberculosis infection are at markedly higher risk of developing tuberculosis (10). Next to malaria, tuberculosis and HIV are the deadliest infectious diseases worldwide, and tuberculosis is one of the main causes of death in HIV-infected persons (10, e20). It is unclear whether HIV infection is a risk factor for drug-resistant or multidrug-resistant tuberculosis (1, 6, 8). Higher resistance rates might be explicable as the result of higher susceptibility to resistant bacterial strains, which are often less virulent than non-resistant strains, as well as of the higher percentage of new infections (8). Other factors that can promote the development of resistance include malabsorption, drug intolerance, drug interactions, and noncompliance among IV drug abusers (8). Hospitalization also -increases the risk of exposure (6, e16). There was a catastrophic development in South Africa in 2006, when XDR tuberculosis was transmitted from infected patients to members of a village community with a high prevalence of HIV. The affected patients were hospitalized, whereupon a large number of patients and hospital employees died within a few weeks (11). The main causes for the persistent transmission of XDR-TB in South Africa are, aside from the high prevalence of HIV, delays in diagnosis and treatment and the inadequate availability of modern diagnostic procedures, second-line drugs, and precautions against infection.

Another current cause for concern is the rising rate of HIV infection in Eastern Europe, particularly in the Russian Federation and the Ukraine (e2, e21). Prisons in these countries are high-risk areas for dual infections because of an increasing rate of IV drug abuse, combined with a high prevalence of MDR tuberculosis (e21).

There are no reliable figures on the rate of tuberculosis/HIV co-infection in Germany, because HIV infection is reported anonymously. The co-infection rate is estimated to be below 5% (5, e2, e22).

The diagnosis of drug-resistant tuberculosis
Drug resistance is suspected when one or more of the risk factors discussed above are present. It can be definitively confirmed only with the aid of standardized, quality-controlled bacteriological sensitivity testing. Because directed therapy is possible only on the basis of the resistance findings, bacteriological proof of tuberculosis infection should always be attempted, even in types of pulmonary or extrapulmonary infection where relatively few bacteria are present.

The gold standard is resistance testing with culture techniques; such testing previously took eight to twelve weeks, but its duration has been shortened recently to two to three weeks with the aid of liquid cultures and radiometric methods (12). More rapid molecular biological methods for the detection of genetic mutations that confer resistance to various drugs (rifampicin, isoniazid) are an outstanding recent advance in this area (12, e23, e24). Microscopic observation of drug susceptibility (MODS) is one of several promising new techniques (e5, e23, e25).

Resistance testing for second-line drugs is highly demanding and requires the expertise of specialized laboratories (2). Moreover, in vitro results often do not accurately reflect actual drug efficacy (2). A rapid test for tuberculosis that could be performed directly on a sputum sample, with which the pathogens could be simultaneously detected and comprehensively tested for resistance, would certainly be a milestone in the fight against tuberculosis (e5, e25, e26).

The treatment of drug-resistant tuberculosis
Tuberculosis must be treated with a combination of antibiotics (e13). The currently recommended standard chemotherapy of non-resistant tuberculosis consists of the initial administration of four first-line drugs (isoniazid, rifampicin, pyrazinamide, and ethambutol or streptomycin) in combination for two months, followed by a four-month stabilization phase with a combination of isoniazid and rifampicin (13, e27).

Sensitivity testing should be performed as rapidly as possible, particularly when drug resistance is suspected, so that the development of further resistance will not be promoted by nonspecific therapy (e28). A single drug should never be added to an existing regimen, as this creates the danger of monotherapy (2).

No randomized trials or evidence-based data are available on the treatment of resistant tuberculosis (14, e5). The WHO recommends that patients who were previously treated for tuberculosis should be treated with at least three drugs that they have not received before. When multiple resistance is suspected, at least four drugs that are still potentially effective should be given (2). As a rule, complex cases of resistant tuberculosis should be treated by physicians with special experience in this area.

The new WHO classification of first- and second-line anti-tuberculosis drugs is shown in Table 2 (gif ppt). The fluoroquinolones are among the main types of second-line drug. In Germany, linezolid is often given in cases with complex resistance patterns (5), even though it has not been recommended by the WHO for routine use. It should be used only in specifically chosen, individual cases, in view of its potential toxicity (5)—in particular, marked changes in blood counts and peripheral polyneuropathy—and high cost.

The treatment takes up to two years and is often poorly tolerated. It thus requires a high degree of patient cooperation, and the rate of premature termination of treatment is higher than in non-resistant tuberculosis (up to 30%) (e5, e29). Thus, extensive patient education is needed, and the patient should take the medications under professional supervision, if possible. The possibility of contagion necessitates adequate preventive measures against infection. Patients who are unwilling or unable to comply with such measures may need to be involuntarily quarantined; such decisions are to be taken on a case-by-case basis (15).

The success rate of treatment is lower for MDR tuberculosis than for less resistant or non-resistant tuberculosis, and it is lower still for XDR tuberculosis (2, 5, 11, 15–17, e5, e30). The Robert Koch Institute reports a current 52% cure rate for MDR tuberculosis (3). This figure accords well with other figures on the same subject from Germany (5, e22) and other countries (14, 17, e5, e31–e33).

The relevant percentage of patients whose therapeutic outcome is unknown, or whose treatment has not yet been completed, can substantially diminish the success rate, depending on how this rate is defined (5). Furthermore, the therapeutic outcome may be difficult to categorize: for example, when the treatment has been changed or interrupted for a long time (e34). Nonetheless, effective surveillance of resistance findings and therapeutic outcomes is very important if the quality of tuberculosis control is to be accurately judged (3).

No official statistics are available on therapeutic outcomes for XDR tuberculosis in Germany, but studies of the relevant data have revealed, in accordance with international studies (17, 18, e5, e31, e35), that the outcomes are worse than in MDR tuberculosis, with markedly longer duration of illness and hospitalization, delayed bacteriological conversion, and higher cost (5, 15, e6). Seven patients treated for XDR tuberculosis in Germany had, in comparison to 177 patients with MDR tuberculosis, both a longer duration of hospital stay (202 vs 162 days) and a longer time to conversion of sputum cultures (141 vs 82 days), and these differences were statistically significant (5). The mean duration of treatment in four patients treated for XDR tuberculosis was 2.2 years (15). When previous treatments before the diagnosis of XDR tuberculosis were also taken into account, the duration of hospitalization in these patients (among whom compliance was also a major problem) ranged from eleven months to six years.

Improved nutrition and improvement of the patient’s social environment are among the most important interventions that can supplement drug therapy for tuberculosis (2, 14). Operative treatment, in addition to drug therapy, is indicated in cases of MDR or XDR tuberculosis if not enough medications are available, and also in cases of non-conversion of serum cultures, persistent caverns, and/or mainly localized disease, as long as there are no functional contraindications to surgery (14, 16). Good results have been described, but often with quite high complication rates (14, 16, 18, e35). There have been no controlled trials on this subject; it seems likely that the operability criteria that were applied led to a selection of prognostically more favorable cases.

The cost of treatment in cases with complex drug resistance is several times higher than that of drug-sensitive tuberculosis (e8, e36). Moreover, the indirect costs, including that of prolonged inability to work, are often substantial. When these costs are taken into account, the cost of some cases of MDR tuberculosis in the USA is found to be in excess of one million dollars (e37). The cost of treating XDR tuberculosis is even higher. In Germany, the direct medical cost of two years of treatment for XDR tuberculosis alone amounts to 170 000 euros (15).

Strategies against drug resistance
In 2006, the WHO announced an ambitious global plan to lower the rate of new infections with tuberculosis and the death rate from tuberculosis to half of their 1990 levels by the year 2015 (19). A further goal is the eradication of the disease by the year 2050, i.e., lowering its incidence to less than one new case of tuberculosis per one million population per year. The overall financing plan envisions 56 billion dollars of financial support for the period 2006–2015 (20). More than one billion dollars are budgeted for the successful treatment of MDR and XDR tuberculosis cases in the year 2009 alone, in addition to the necessary overall expenditures for global tuberculosis control, which amount to 5.3 billion dollars. A basic prerequisite for the prevention of drug-resistant tuberculosis is adherence to the stated principles of treatment, in the setting of an effective national tuberculosis control program, whenever possible (7, 21, 22). The DOTS strategy (Directly Observed Treatment Short Course”) is recommended for implementation (19); in recognition of the problem of resistance, the DOTS strategy has been extended to the so-called DOTS-plus strategy, and to other action plans that build upon it (22, 23, e8, e38–e39). The implementation of these plans is difficult, however, not just because of inadequate financial means, but often also because the necessary infrastructure is lacking, e.g., adequately trained personnel.

The Green Light Committee established by the WHO provides technical support to poorer countries and negotiates reduced prices for quality-controlled second-line drugs. A functioning national tuberculosis control program is a prerequisite (e40).

It is currently debated whether standardized treatment regimens for non-responders to therapy (e27) should be replaced by individualized regimens based on (rapid) resistance testing (e5, e7, e24, e41) in regions of the world where MDR tuberculosis is highly prevalent.

Research also needs to be substantially intensified in this area (21, e42). Alongside better diagnostic techniques for tuberculosis, including techniques for the determination of drug resistance, there is an urgent need for the development (or further development) and testing of highly effective anti-tuberculosis drugs (24). Over the long term, there are high hopes that effective vaccines can be developed through the improvement, supplementation, or replacement of BCG (Bacille Calmette-Guérin), a vaccine based on an attenuated strain of M. bovis (25).

The “Global Stop TB Partnership” (e43), founded in 2000 and now with more than 700 private and governmental partners, serves to merge common interests and capabilities. It receives major financial support from the Global Fund to Fight AIDS, Tuberculosis, and Malaria (e44), as well as other organizations.

All of these approaches need to be continuously maintained and vigorously supported. Even in a country like Germany, in which there are adequate high-quality laboratory services and all second-line drugs are available, success rates in the treatment of MDR and XDR tuberculosis remain unsatisfactory (3, 5, 15). It follows that the worsening resistance situation all over the world can only be combated and alleviated through a common effort (7, 21, e38, e39). The political will of the industrialized countries and their assumption of responsibility for health policy, which were expressed in a declaration by a WHO European ministerial forum organized by the German government and held in Berlin in 2007 (e45), must now be practically implemented through support for research on the national and international levels, and through adequate financial contributions.

We thank the German Federal Ministry of Health for its support.

Conflict of interest statement
The authors state that they have no conflict of interest as defined by the guidelines of the International Committee of Medical Journal Editors.

Manuscript submitted on 19 May 2009; revised version accepted on
15 July 2009.

Translated from the original German by Ethan Taub, M.D.


Corresponding author
Prof. Dr. med. Dr. h.c. Robert Loddenkemper
Deutsches Zentralkomitee zur Bekämpfung der Tuberkulose
Stralauer Platz 34
10243 Berlin, Germany
rloddenkemper@dzk-tuberkulose.de

@For e-references please refer to:
www.aerzteblatt-international.de/ref0110
eFigure available at:
www.aerzteblatt-international.de/article10m0010