Module 10
Drug susceptibility testing
Part 1. Principles of resistance to antituberculosis drugs
Part 2. Performing DST on solid media
Purpose / Part 1. To provide you with knowledge of the principles of antituberculosis therapy and drug resistance, and of the relationship of drug resistance to treatment failure.Part 2. To provide you with the knowledge and skills to perform DST using the proportion method on solid and liquid media.
Prerequisite modules / Modules 1, 2, 3, 4, 6 and 7
Module time / Part 1: 35 minutes
Part 2: 1 hour and 10 minutes, plus two practical sessions in the laboratory (3 hours)
Learning objectives / At the end of Part 1of this module you will understand:
‒ the objectives and rationale of combined antituberculosis therapy;
‒ the basic definitions of drug resistance in TB;
‒ how drug resistance develops;
‒ the rationale for using a critical concentration of drug in laboratory assays;
‒ the rationale for determining the proportion of resistant bacteria that is clinically significant.
At the end of Part 2 of the module you will be able to:
‒ explain the different methods for performing DST;
‒ perform and interpret the proportion method;
‒ record and report DST results;
‒ store strains.
Content outline / Part 1
· History and objectives of antituberculosis therapy
• Compartmentalization of M. tuberculosis in infected tissue
• Bactericidal and sterilizing activity of drugs
• First-line antituberculosis drugs
• Basic definitions of drug resistance
• Development and genetic basis of drug resistance
• Relationship of critical concentrations and proportion of resistant bacteria to prediction of treatment failure
Part 2
• Drug susceptibility test methods:
‒ direct
‒ indirect
• Proportion method: technical protocol, interpretation criteria
• Recording and reporting DST results
• Storage of strains
Exercises / Practical work in the laboratory:
‒ perform proportion method
‒ read DST results.
Annexes / 10.1 Proposed protocol – drug susceptibility testing, proportion method
10.2. Proposed protocol – strain maintenance.
PART 1. PRINCIPLES OF DRUG SUSCEPTIBILITY TESTING
Introduction
Drug-resistant strains of Mycobacterium tuberculosis exist worldwide and constitute a serious threat to the efficacy of TB control programmes. The most effective strategies for limiting further spread of drug-resistant TB are rapid detection of drug resistance followed by prompt and effective therapy. This module provides an introduction to the principles of antituberculosis therapy and drug-resistance and describes the relationship of drug resistance to treatment failure.
History of antituberculosis therapy
In the 1940s, two drugs – streptomycin (SM) and p-aminosalicylic acid (PAS) – were introduced for TB therapy. Use of either SM or PAS alone in the treatment of TB was initially found to reduce deaths among treated patients. Single-drug therapy soon resulted in the emergence of drug-resistant strains in a majority of the patients; however, by combining SM and PAS, the resistance rate was significantly reduced. Eventually, more effective anti-TB drugs were introduced and used in combined therapy.
Isoniazid (INH) was introduced in the early 1950s. Use of INH in combination with SM and PAS was highly effective in preventing the emergence of resistance, but 18 months of treatment were required to ensure an adequate cure. Pyrazinamide (PZA), rifampicin (RMP) and ethambutol (EMB) were also introduced for TB treatment and were prescribed in combinations with INH.
Extensive studies to define the optimal drug combination and the minimal duration of therapy were carried out by the British Medical Research Council. The result was the 6-month short-course therapy consisting of 2 months’ treatment with INH, RMP and PZA, followed by 4 months’ treatment with INH and RMP. This protocol is still in use today, and is recommended with slight modifications (including possible addition of EMB or SM) by WHO, the International Union Against Tuberculosis and Lung Disease (IUATLD), the American Thoracic Society, the Centers for Disease Control and Prevention (CDC), and the Infectious Disease Society of America.
Objectives of antituberculosis therapy
Antituberculosis therapy has three important goals:
‒ to rapidly kill the large numbers of active bacilli that are multiplying in the infected tissues;
‒ to prevent the emergence of a clinically significant proportion of drug-resistant mutants; and
‒ to effectively eliminate bacilli that are dormant but still viable from the infected tissues.
Anatomical and metabolic compartmentalization of M. tuberculosis in infected tissue
Scientists and clinicians have postulated three populations (A, B and C) of tubercle bacilli in a patient with pulmonary TB (see Figure 1). These populations are differentiated by their locations within the lesions and by their metabolic activity.
Population A represents the large number of rapidly multiplying bacilli found in pulmonary cavities. Population B represents bacilli that multiply less rapidly because of local adverse (most often acidic) conditions. Elimination of Populations A and B through effective therapy results in negative sputum smear (usually after 2–3 weeks) and culture (usually after 2 months). Population C consists of dormant bacilli (often sequestered in granulomas) that are still capable of sporadic bursts of metabolism and replication – and it is these organisms that can be the source of potential relapses.
Anti-TB drugs such as INH play a pivotal role in killing Population A; these drugs are considered to have rapid bactericidal activity (Figure 1). Bactericidal activity is measured by the rapidity with which sputum smears and culture become negative during treatment of pulmonary TB. The drugs that are most active in killing Population A are also the most effective in preventing the emergence of drug-resistant cells (as discussed in more detail in the next section).
Anti-TB drugs that are more effective against Populations B and C are regarded as sterilizing agents. PZA is especially effective against Population B because of its unique sterilizing activity in an acidic environment, while RMP is more active against both Populations B and C (Figure 1). The potency of this activity is reflected in a high cure rate with limited relapses in patients who complete therapy.
First-line antituberculosis drugs
Isoniazid
Isoniazid is a pro-drug that requires activation by the mycobacterial enzyme, catalase. It has a potent early bactericidal effect on rapidly multiplying, metabolically active M. tuberculosis (Population A, Figure 1),. Although INH may act on several targets within the mycobacterial cell, significant evidence supports the concept that it blocks the synthesis of cell-wall mycolic acids, the major component of the envelope of M. tuberculosis (Figure 2).
Rifampicin
In addition to a significant early bactericidal effect on metabolically active M. tuberculosis (Populations A and B, Figure 1), RMP also exhibits excellent late sterilizing action on semi-dormant organisms undergoing short bursts of metabolic activity (Population C, Figure 1). This late effect of RMP and the additional effectiveness of PZA have allowed routine TB treatment to be reduced from one year to six months.
The sterilizing activity of RMP is critical for successful therapy since resistance to this drug is one of the main reasons for treatment failure and a fatal outcome in TB patients. The mechanism of action of RMP is inhibition of RNA transcription in the mycobacterial cell by targeting DNA-dependent RNA polymerase (Figure 2).
Pyrazinamide
PZA has an excellent sterilizing effect on slowly multiplying tubercle bacilli (Population B, Figure 1) especially in an acidic environment. When PZA is used in combination with RMP, the duration of TB treatment is reduced from one year to six months. PZA is a pro-drug, converted to its active form, pyrazinoic acid, by the mycobacterial enzyme pyrazinamidase; the mechanism of action is not well understood, proposed targets for pyrazinoic acid include an enzyme involved in fatty acid synthesis and the destructive acidification of mycobacterialcytoplasm (Figure 2).
Figure 1. Compartmentalization of M. tuberculosis in the infected tissue
Adapted from: Iseman M. A clinician‘s guide to tuberculosis. Philadelphia, Lippincott, Williams & Wilkins, 2000.
Basic definitions and general principles of drug resistance in M. tuberculosis
Members of the M. tuberculosis complex (MTBC) (M. tuberculosis, M. africanum, M. canettii, M. microti, M. bovis, M. bovis BCG, and M. bovis ssp caprae) share several means of natural resistance to some antibiotics. A hydrophobic, lipid-rich cell envelope surrounds these organisms and serves as a permeability barrier to many compounds. Also, drug efflux systems and drug-modifying enzymes are present. However, differences in natural resistance within the MTBC have been observed. For example, strains of M. canettii, M. bovis and M. bovis BCG are resistant to PZA, while other members of the MTBC are usually susceptible to this drug.
The following definitions are based on recommendations from WHO: drug resistance among previously treated patients (acquired drug resistance) is the most common type of resistance to any of the first-line drugs, and develops in a patient during chemotherapy. In contrast, drug resistance among newly identified patients (primary drug resistance) is defined as the presence of drug-resistant organisms in a previously untreated person, presumably because that person has been infected with a strain that had acquired resistance in another host.
Cross-resistance can occur between drugs that are chemically related and/or have the same or a similar target within the mycobacterial cell. For example, approximately 70–90% of RMP-resistant strains are also resistant to rifabutin (chemically related); some strains of M. tuberculosis with low-level resistance to INH are cross-resistant to ethionamide (similar target); and cross-resistance can develop between certain aminoglycosides, e.g. kanamycin and amikacin, or between fluoroquinolones. However, no cross-resistance has been seen between INH and PZA, even though both drugs are analogues of a common component, nicotinamide.
Figure 2. First-line anti-TB drugs and their mechanisms of action
Adapted from Somoskovi, Parsons, Salfinger. 2001, Respir. Res. 2:164-168
Polyresistance occurs when a strain of M. tuberculosis develops resistance to more than one anti-TB drug. Resistance to INH and RMP, which are the most effective anti-TB drugs, results in treatment failure and fatal outcomes more often than resistance to other drugs. Strains resistant to both INH and RMP are therefore distinguished from other polyresistant strains and defined as multidrug-resistant tuberculosis (MDR-TB). An even more serious threat is the recently described extensively drug-resistant tuberculosis (XDR-TB); these strains of M. tuberculosis are resistant not only to INH and RMP but also to at least one of the fluoroquinolones and to one injectable drug such as kanamycin or amikacin.
Development of resistance to antituberculosis drugs in M. tuberculosis
Resistance to anti-TB drugs in M. tuberculosis occurs when resistant mutants, naturally occurring in the mycobacterial population, are selected by inadequate or interrupted treatment with anti-TB agents.
Mutations occur spontaneously in the DNA of all cells, but many of these changes are deleterious and are selected against during growth of bacterial populations. However, a bacterial cell carrying a mutation that changes the structure of a protein that is the target for a particular drug, but does not compromise the protein’s functioning, will be selected for in the presence of the drug. Thus, for example, mutations that alter the amino acid sequence of the DNA-dependent RNA polymerase of M. tuberculosis such that it will no longer bind RMP will result in an RMP-resistant clone of tubercle bacilli. If single antibiotic treatment of TB were to be used, such clones would predominate in the population and, if not removed by the immune system, would cause therapy failure. This is what occurred in the 1940s when single-drug therapy with SM was used.
Mutants resistant to any given drug occur randomly and rarely, on average once in every 108 (100 million) cells. That is, there is a chance of one mutation occurring in a given gene (e.g. a gene encoding a target protein) each time 100 million bacteria divide. Given exposure to selection pressure by antibiotics, this rare event can readily give rise to substantial clones of resistant cells because of the large numbers of bacteria involved (easily 108 per lesion in lung cavities). By using two antibiotics, however, the chances that any given cell in the population will simultaneously carry resistance to both antibiotics is 10–8 x 10–8 , or 10–16.
Consistent with this theory, it was found that monotherapy led to selection of drug-resistant populations more frequently in cases of cavitary disease, in which lesions contain abundant tubercle bacilli (108–109 per lesion), than in cases with non-cavitary lesions, which contain relatively few cells (about 103–104 per lesion).
Other factors can also influence the emergence of drug resistance. Although the tubercle bacillus replicates actively in vivo, it has a long generation time and a tendency to shift its metabolism toward a dormant state (Population C, Figure 1). Such variations in metabolic activity make this organism a difficult therapeutic target. In addition, penetration of antibiotics to various body sites can vary significantly, resulting in some lesions with suboptimal concentrations of drugs and thus a greater opportunity for selection of drug-resistant mutants.
The individual patient’s influence on the development of drug resistance should also be considered. It has been found that the TB patients most likely to produce drug-resistant mutants are those who are experiencing increases in the bacterial population, together with compromised drug penetration due to underlying host conditions, particularly poor intestinal absorption of drugs due to HIV/AIDS.
Critical antituberculosis drug concentrations, proportion of resistant mutants and their relation to the accurate prediction of therapeutic failure in vivo
Early investigators of laboratory diagnosis of drug resistance faced two significant challenges. First, critical concentrations – drug concentrations necessary to eliminate susceptible (wild-type) strains of M. tuberculosis – needed to be determined. Second, the proportion of a population of M. tuberculosis that would need to be resistant to a drug in order for that strain to be interpreted as resistant, thus permitting accurate prediction of therapeutic failure, was unknown.
With regard to the first challenge, it was observed that drug-susceptible strains of M. tuberculosis that have not been exposed to anti-TB drugs (wild-type strains) do not exhibit much variation in the minimum inhibitory concentrations (MICs) to those drugs. It was thus determined that the critical concentration should be the concentration of drug that inhibits the growth of wild-type strains without appreciably affecting the growth of resistant cells. It was also found that the critical concentration varies with the media used in the different susceptibility assays, which has resulted in slight variations in the concentrations tested in laboratory assays, especially when using either solid or liquid media.