Antimicrobial Chemotherapy, 4th Edition
The therapeutic use of antimicrobial agents
- M. Emmerson
Tuberculosis (TB) is one of the most widespread infections known to man; it is estimated that a third of the world's population harbour the pathogen Mycobacterium tuberculosis, and every year 8 million people develop new clinical disease. The pulmonary form was described by Hippocrates, and characteristic lesions of TB of the spine have been demonstrated in the mummies of ancient Egypt. It attacks both humans and animals (e.g. badgers, cattle), affects all ages and every organ in the body, and ranges from latent to hyperacute (the ‘galloping consumption’ of Victorian times), killing young and old alike, the famous and the unknown. This reservoir of infected persons results in 3.6 million people with infectious pulmonary tuberculosis (sputum-smear positive) with an overall annual mortality of about 3 million. Infection with human immuno-deficiency virus (HIV) is the most potent risk factor for development of TB: over 3 million people are co-infected with HIV and TB, according to World Health Organization (WHO) figures.
The last 50 years have seen enormous progress in the understanding of the epidemiology, prevention, and treatment of TB and, although far from eradicated, this age-old killer of man has become a preventable and curable disease. Disturbing new problems have arisen with the treatment of co-existing infection of HIV and TB.
This chapter deals principally with the treatment of pulmonary TB because it remains the commonest form of the disease (65 per cent of cases) and practically the only form responsible for human transmission.
Landmarks in the treatment of tuberculosis
Before the advent of chemotherapy during the 1940s, the treatment of tuberculosis was largely restricted to attempts to increase the patient's resistance to the
disease. Prolonged rest in hospitals and sanatoria, special diets, and avoidance of physical activity were all thought to be important. Attempts were made to immobilize affected lung tissue by artificial pneumothorax, removal of ribs, or, more radically, removal of affected lung tissue itself. Amputation of affected limbs was common, as was surgical resection of other diseased tissues. Today, most of these methods have become almost forgotten history, as the action of drugs on the tubercle bacillus itself has assumed overwhelming importance.
The main landmarks in this story are summarized in Table 26.1. The first was the discovery, in 1940, of the bacteriostatic effect of sulphonamides in guinea-pigs infected with tubercle bacilli. The most effective were diamino-sulphones such as dapsone. However, these agents were found to have little if any activity against tuberculosis in man, although they were effective in the treatment of leprosy and remain so today.
Table 26.1 Landmarks in the treatment of tuberculosis
A major advance was the introduction in 1944 of streptomycin, the first drug shown to be effective in the treatment of human tuberculosis. Its use was, however, limited by the ready emergence of streptomycin-resistant tubercle bacilli during treatment, and also by adverse reactions to the drug. In 1949, it was discovered that combined therapy with p-aminosalicylic acid (PAS) and streptomycin prevented the emergence of strains resistant to either, and since then the administration of two or more drugs in combination has been considered essential for adequate treatment of tuberculosis.
The third of the classic antituberculosis drugs, isoniazid (INH), was introduced in 1952 and, for the first time, uniformly successful primary chemotherapy of tuberculosis became possible. Initial treatment with streptomycin alone, in courses lasting from 6 weeks to 3 months, had a very high relapse rate after cessation of treatment; prolonged treatment led to a high incidence of toxicity from the drug, and emergence of drug resistance was common. The addition of PAS almost eliminated the emergence of resistant strains, and allowed treatment to be prolonged to 12 months or longer. Relapses after treatment were still common, but longer periods of treatment with streptomycin, which could only be given
by daily injection, were increasingly unacceptable to patients and it needed the introduction of INH, another oral agent, for really long-course treatment to be possible. Eventually, a regimen evolved in which streptomycin, PAS, and INH were given for 2–3 months, followed by PAS and INH for a further 18 months to 2 years. Until the late 1950s, patients were usually confined to hospital for most of this time. More than 90 per cent of patients completing the course were cured, but it was also a very exacting ordeal and patients commonly absconded from treatment. Among absconders, relapse was common.
Between 1955 and 1960 a controlled clinical trial was carried out in Madras to compare the effect of 12 months of chemotherapy in two groups of patients: one group treated under good conditions in a sanatorium, the other under poor conditions at home. The results were startling. Despite good accommodation, nursing, rich diet, and prolonged bed rest the sanatorium patients did no better than similar patients treated in overcrowded homes, who had a poor diet and much less rest, and often worked long hours under poor conditions. The risk to close family contacts was studied for over 5 years, and showed that there was no difference in the incidence of disease between the contacts of patients treated at home and those of sanatorium patients. The major risk to contacts lay in exposure to the index case before diagnosis was made and treatment initiated. Once effective treatment had been started there was little further risk to contacts. Also, the study showed that treatment in a sanatorium is no safeguard against irregularity of drug taking unless the patient is actually seen to swallow every dose or receive every injection. It was this study which caused the dramatic change from institutional to ambulatory outpatient treatment as a general policy. TB sanatoria were closed as public health authorities recognized that patients receiving combination chemotherapy did not constitute a contagious risk.
During the 1960s several new antituberculosis drugs were brought into use, of which three—rifampicin, pyrazinamide, and ethambutol—have emerged as of particular importance. Exploitation of these drugs has allowed investigation of intermittent chemotherapy regimens, in which individual drug doses are given at intervals of more than a day (in some cases only once or twice a week)—of particular importance in developing countries where fully supervised daily medication is difficult to deliver—and also, more recently, the development of much shorter regimens, several being curative in less than 12 months.
Factors involved in the response to chemotherapy
Tubercle bacilli: number, site, and activity
In human pulmonary tuberculosis, most of the tubercle bacilli are found on the walls of cavities in the lungs open to the bronchi. Here the pH is relatively high, at least on the alkaline side of neutrality. The oxygen tension is also high and the bacilli are actively multiplying. However, there is another smaller population of bacilli, dormant in closed cavities, or in caseating tissue, or inside macrophages, where the pH and oxygen tension are both low, and bacterial multiplication is slow.
Antituberculosis drugs and their mode of action
The drugs available for the treatment of tuberculosis differ in their activity against tubercle bacilli under different conditions. For example, streptomycin and INH are bactericidal against actively multiplying bacteria under alkaline conditions, whereas pyrazinamide acts largely on intracellular organisms in an acid medium. Rifampicin is active against both extracellular and intracellular organisms, and also on those dormant in caseous nodules. Bactericidal activity against actively multiplying bacteria largely determines the acute response of the sick patient to chemotherapy, but sterilizing activity against the dormant ‘persisters’ in the bacterial population is most important when considering the incidence of relapse after cessation of treatment.
Effects of the body's defences and immune response: species variation
Host factors and defence mechanisms have, in recent years, been shown to be of much less importance in determining the outcome of tuberculous infection than is the effective use of antituberculosis drugs. However, species differences in the response to tuberculosis are of considerable interest and have been the source of some confusion. In the mouse, unlike man, tubercle bacilli multiply readily inside macrophages, although there is much less multiplication in cavities and tissue spaces. In the guinea-pig, multiplication of bacilli more nearly resembles that in man but the human immune response, based on the bacteristatic effect of low pH and low oxygen tension in caseous tissue, is nearer that of the mouse than the guinea-pig. The effect of these differences on the experimental investigation of the drug treatment of tuberculosis has been to complicate an already difficult picture. For example, streptomycin has high bactericidal activity in man and moderate activity against actively multiplying bacilli in the guinea-pig, but is virtually inactive in the mouse. Pyrazinamide, bactericidal in the mouse, is without effect on tuberculosis in the guinea-pig. The practical implication is that the development of effective antituberculosis regimens in man has depended almost entirely on the use of large-scale controlled trials of different regimens in appropriate human populations. There is still no suitable in-vitro or animal model from which information can be transferred to man without reservation.
Immunity in TB is cell mediated; the key to restriction of intracellular growth of the bacillus is co-operation between macrophages and sensitized lymphocytes. Humoral reactions are thought to play little part. The generation of clones of antigen-specific T cells initiates the immune response, which is depressed in HIV-positive patients. T cell clones may facilitate elimination of the pathogen by macrophage activation and granuloma formation or by cytotoxic action; or they may interact with B cells to produce specific antibodies which, in conjunction with phagocytic cells or complement, also act as an effector arm. The mechanism by which mycobacteria survive inside activated macrophages is unknown and may be oxygen dependent or oxygen independent. The role of
iron in persistence of infection has not been elucidated. Future research is required to establish the interaction between antimycobacterial agents and the immune system, if one exists.
The antituberculosis drugs
This was the first effective drug against tuberculosis in man. Like all amino-glycosides it is not absorbed when given orally, and must be administered as an intramuscular injection. It is more actively bactericidal in the proliferative phase than in the resting phase of bacterial growth. Adequate concentrations are attained in lung, muscle, uterus, intestinal mucosa, adrenals, and lymph nodes. Diffusion into bone, brain, and aqueous humour is poor. Little normally enters the cere-brospinal fluid (CSF), although penetration increases when the meninges are inflamed. Vestibular damage occurs in about 30 per cent of cases and 24 h CSF levels may need to be monitored to keep the trough level below 3 mg/l. The plasma half-life, which is normally 2–3 h, is considerably extended in the newborn, in the elderly, and in patients with severe renal impairment. Serum levels need to be monitored in patients over 40 years of age. Hypersensitivity reactions in the form of fever and skin rashes may develop.
The injections are painful and must be administered by deep intramuscular injection; 1 per day is sufficient but barely tolerable. Duration of treatment is 1–2 months.
p-Aminosalicylic acid (PAS)
This was formerly used in combination regimens to prevent the emergence of drug-resistant organisms. Up to 10–12 g a day were given orally, in two or three doses. The drug tastes most unpleasant, and gastrointestinal intolerance was common. PAS is no longer used and is included here for historical reasons only.
Isoniazid (isonicotinic acid hydrazide; INH)
This is active only against the tubercle bacillus, not against other microbes, but against this organism it is potent and bactericidal. It penetrates rapidly into all tissues and lesions; its activity is not affected by the pH of the environment; it is well tolerated and cheap. Not surprisingly, therefore, it is the drug most widely used in the treatment of tuberculosis. The drug is given in a single daily oral dose as it is more important to achieve a high peak concentration than to maintain a continuously inhibitory level. In intermittent regimens, large doses can be used and very high peak levels attained. Isoniazid is metabolized mainly by acetylation, at a rate which varies from one individual to another. Patients can be divided into two groups—rapid and slow inactivators. This is of little clinical importance in patients treated daily, or even twice weekly, but with intermittent regimens in which the drug is given only once a week, rapid inactivators (within
1 h) fare less well than slow inactivators (within 3 h), and there is some practical value in determining, by relatively simple tests, to which group a patient belongs when an intermittent regimen is contemplated.
Adverse reactions, such as cutaneous hypersensitivity, are uncommon. Isoniazid tends to raise plasma concentrations of warfarin, diazepam, phenytoin, and carbamazepine by inhibiting their metabolism in the liver. Overdosage may give rise to vomiting, dizziness, blurred vision, and slurring of speech within 30 min–3 h.
Hepatitis is an uncommon (0.1 per cent) but potentially serious reaction that can be easily averted by prompt withdrawal of treatment. A sharp rise in serum transaminases at the outset of treatment, which may be enhanced by the concomitant use of rifampicin, is of little significance. Neverthless liver function tests should be checked before starting treatment. Isoniazid interferes competitively with pyridoxine metabolism by inhibiting the formation of the active form of the vitamin and hence often results in peripheral neuropathy, which can be-prevented by co-administration of pyridoxine.
Following oral absorption, plasma concentrations are good and CSF penetration is about 50–80 per cent of the serum levels. Urinary concentrations are high. Primary resistance in M. tuberculosis in the UK is rare (1–4 per cent) but resistance develops quickly with single-drug therapy.
This has a special sterilizing effect on intracellular tubercle bacilli in acid conditions and has therefore found an important place in short-course chemotherapy, in which the important factor is the incidence of relapse after cessation of treatment. It is given orally, and may cause anorexia, nausea, and flushing; it produces high serum levels, and penetrates freely into CSF. Adverse reactions are rare, but hepatic toxicity has been a problem; hypersensitivity reactions and photosensitivity of the skin also occur.
Treatment for longer than 2 months is not recommended. Drug interactions occur with allopurinol, causing hyperuricaemia with attacks of gout, and with oral antidiabetic agents, causing a further fall in blood sugar. The dose should be reduced in renal impairment and avoided in patients with severe liver damage and gout.
Rifampicin and other rifamycins
One of the most valuable of the antituberculosis drugs is rifampicin. It has proved a most potent drug both for primary treatment and in treatment of relapses, given daily or intermittently. However, rifampicin is not yet universally used in developing countries because it is expensive, and it has troublesome side-effects such as hepatitis and cutaneous reactions requiring special supervision and care. Curiously, these are more common on intermittent regimens than when the drug is taken daily, and typically begin 2–3 h after the single morning dose. Once-weekly
regimens give more toxicity than twice-weekly, and with daily regimens side-effects are uncommon and trivial. The side-effects are usually mild, and can usually be controlled by reducing either the dose size or the interval between doses. The exception is that the occurrence of purpura is an indication to stop the drug and not give it again.
Rifampicin is sometimes used in the treatment of life-threatening staphylococcal infections and in the prophylaxis of meningococcal andHaemophilus influenzae meningitis, but there is no evidence that this has jeopardized the value of rifampicin in the treatment of TB.
Although rifampicin has been successfully used for decades in tuberculosis, resistance remains low in most countries. However, resistance rates of 10–16 per cent are found in certain parts of the world, notably Thailand, the Dominican Republic, and some countries of the former USSR; such strains may indeed be multi-resistant.
Three newer rifamycins—rifabutin, rifapentine, and benzoxazinorifamycin—are active against M. tuberculosis. They are somewhat more active than rifampicin, accumulate within macrophages and exhibit longer half-lives, so that they may prove useful for intermittent treatment.
Ethambutol is active against M. tuberculosis and many of the atypical mycobac-teria. Resistance develops slowly during therapy, but primary resistance occurs in less than 4 per cent of strains of M. tuberculosis. It has a primarily bacter-istatic action on proliferating bacteria. It is rapidly absorbed after oral administration and high serum levels are found after 2 h, with even higher levels inside erythrocytes. Adequate levels are found in the CSF. Ethambutol is excreted in the urine both unchanged and as inactive metabolites.
Dose-dependent optic (retrobulbar) neuritis can result in impairment of visual acuity and colour vision. Early changes are usually reversible, but blindness can occur if treatment is not discontinued promptly. It is recommended that the patient's visual acuity should be tested by a Snellen chart before ethambutol is first prescribed.
Thiacetazone is one of the oldest antituberculosis drugs, having been known since before 1950. It has about the same rate of toxicity as PAS, with rashes, jaundice, bone marrow depression, and gastrointestinal upsets prominent. It is more convenient to the patient than PAS (one tablet instead of several cachets a day), is much cheaper, and is stable in tropical climates, where PAS tends to deteriorate.
Thiacetazone is bacteristatic against M. tuberculosis and is sometimes used in combination with isoniazid to inhibit the emergence of resistance to isoniazid, particularly in the continuation phase of long-term regimens. It is well absorbed
from the gastrointestinal tract and plasma levels are sustained for long periods of time. Side-effects include nausea, vomiting, diarrhoea, and skin rashes. Rare cases of fatal exfoliative dermatitis and acute hepatitis failure have been reported.
Ethionamide and prothionamide
Because of its gastrointestinal side-effects—anorexia, salivation, nausea, abdominal pain, and diarrhoea—ethionamide is one of the most unpleasant of all anti-tuberculosis drugs to take. Prothionamide is closely related in structure, but is better tolerated. The role of these drugs is almost entirely as second-line treatment of patients with tuberculosis or leprosy whose bacilli are resistant to first-choice drugs.
Both compounds are absorbed quickly when given by mouth with good tissue and CSF levels. Both are bacteristatic at therapeutic concentrations but bactericidal at higher concentrations. Complete cross-resistance occurs between both, but not with isoniazid, to which these drugs are related.
Contraindications include pregnancy, severe liver damage, and gastric complaints; care should be exercised in epilepsy and in psychotic patients. Combinations with isoniazid, cycloserine, or alcohol should be avoided.
Cycloserine, viomycin, kanamycin, capreomycin
These are four rather weak antituberculosis drugs used only in three-drug regimens as reserves for the treatment of tuberculosis resistant to the major anti-tuberculosis drugs.
Capreomycin, viomycin, and kanamycin all need to be given by injection. Primary resistance is rare but resistance develops rapidly.
Several new antituberculosis compounds. are being evaluated, notably the fluoro-quinolones ciprofloxacin, levofloxacin and sparfloxacin, but clinical experience remains sparse.
Short-course and intermittent therapy
The discovery that both pyrazinamide and rifampicin are active against dormant as well as actively dividing tubercle bacilli prompted the investigation of shorter periods of treatment with a variety of combinations of antituberculosis drugs. In the first trial streptomycin and isoniazid were used alone and with either pyraz-inamide, thiacetazone, or rifampicin. This trial clearly showed that all four regimens were effective in controlling the acute stage of the disease, but that, in addition, the regimens containing rifampicin or pyrazinamide would cure more than 90 per cent of patients in 6 months.
Further trials explored different combinations and also the effect of intermittent treatment. It soon became apparent that all regimens containing two of the three drugs, isoniazid, pyrazinamide and rifampicin would cure more than 90 percent
of patients in 6 months and virtually all in 9 months. The limitations of treatment became those of cost and of patient compliance in the face of unpleasant side-effects of the drugs. In the developing world the aim was to find acceptable fully supervised mass treatment at a cost the communities could afford.
Affluent countries have different aims. These countries need very effective unsupervised regimens for use where patient motivation is high and compliance good. Rifampicin plus isoniazid daily for 9 months, with streptomycin or etham-butol for the first 2 months, will cure virtually all patients with pulmonary tuberculosis, and these regimens have become standard treatment in Europe. The total length of treatment can be reduced from 9 to 6 months if pyrazinamide is added for the first 2 months. The British Thoracic Society now recommends a 6 month course of treatment in which rifampicin, isoniazid, and pyrazinamide are given together with either ethambutol or streptomycin for 2 months, then treatment continued for a further 4 months with the two drugs rifampicin and isoniazid (Table 26.2). There is increasing evidence that the fourth drug (etham-butol or streptomycin) can be omitted from this regimen without detriment. Almost all regimens contain an initial intensive phase with three or four drugs as it is important to bring the acute illness under control as rapidly as possible.
Table 26.2 Recommended treatment regimens for tuberculosis in adults
Consideration should be given to treating all patients with directly observed therapy (DOT), in which ingestion of the antituberculosis medications by the patient is observed by a healthcare provider, or other responsible person. DOT can be given on an intermittent schedule and can be given in the patient's home, place of employment, school, etc. It is effective in improving clinical cure rates and in decreasing rates of drug resistance. Use of fixed drug combinations (e.g. Rifinah: rifampicin + isoniazid; Rifater: rifampicin + isoniazid + pyrazinamide) reduces the risk of inappropriate monotherapy. Although such combinations may not contain exactly the same concentration as individual drugs, clinical results are favourable. It is important not to make mistakes in prescribing and dispensing because of the similarity of names.
Effective chemotherapy rapidly reduces the population of viable bacilli (sputum smears may be positive but cultures remain negative) and consequently reduces the risk of transmission, often within 2 weeks. Institutional care is necessary where supervised therapy can be given to patients who fail to respond to treatment and for those who are severely ill (e.g. massive haemoptysis or pyopneumo-thorax) or who are bedridden with severe paraparesis. Most patients can be treated satisfactorily provided that they can be relied on to take their drugs regularly or are carefully supervised.
Steroids have no place in the routine management of tuberculosis, but in hospital short courses are of value in treating patients with pericarditis, pleural effusion, tuberculous meningitis, and acute millary tuberculosis with dyspnoea.
Surgical resection is rarely necessary except in severe post-tuberculosis bronchiectasis, particularly when this is a cause of repeated haemoptysis. However, this may change if multidrug-resistant tuberculosis becomes widespread, particularly in immunosuppressed patients.
Monitoring the therapeutic response
Inadequate compliance is by far the most important cause of therapeutic failure. Most cases of chronic or relapsing tuberculosis are the result of irregular, inadequate administration of prescribed drugs or the use of less potent regimens, and often result in the selection of resistant strains. Response to treatment is most readily evaluated from the clinical response of the patient and particularly by monitoring weight changes, fever, and cough. Where resources permit, cultures and smears should be examined for acid-fast bacilli. Fixed-dose combinations that include rifampicin can be detected by observing the red-orange discoloration of body fluids, especially urine.
Drugs used in pregnancy
Treatment should never be interrupted or postponed during pregnancy or at any time when immunological resistance to the disease is reduced. Because of the risk of ototoxicity to the fetus, aminoglycosides such as streptomycin should be avoided. Patients can breast-feed normally while taking antituberculosis drugs.
Tuberculosis and HIV infection
The main consequence of HIV infection is progressive depletion and dysfunction of CD4 lymphocytes, with defects in macrophage and monocyte function. Because CD4 cells and macrophages are essential in the control of tuberculosis, it follows that HIV exerts an enormous influence on M. tuberculosis infection. Higher rates of reactivation disease occur (7–10 per cent per year, compared with 5–8 per cent per lifetime for HIV-negative patients). There are also much higher rates of acute disease (including extrapulmonary tuberculosis), skin anergy (producing a false-negative Mantoux test), and malabsorption of antituberculosis medications owing to enteropathy. Major drug–drug interactions can occur, especially between rifamycins and the protease inhibitors and non-nucleoside reverse transcriptase inhibitors used to treat HIV.
The treatment of TB in HIV-positive patients is problematic and requires expert knowledge of both anti-TB and antiretroviral drugs. Treatment regimens are often extended for 12 months and 5 or 6 drugs may be required. Multi-resistant strains are common in AIDS patients; in parts of Africa resistance to all first-line drugs occurs in one third of cases. The diagnosis of TB in HIV-positive individuals is also difficult, owing to concurrent infection with Pneumocystis carinii or atypical mycobacteria (see p. 343)
Chemoprophylaxis against tuberculosis
Isoniazid is commonly used alone in antituberculosis chemoprophylaxis because it is effective, cheap, acceptable, and has few side-effects. Fears that isoniazid-resistant organisms would emerge have so far proved unfounded.
There is wide variation between different parts of the world in the number of patients found to be infected with tubercle bacilli resistant to one or more anti-tuberculosis drugs at the time the disease is diagnosed. High rates of resistance, particularly to isoniazid, have been reported in some places, but short-course chemotherapy may be successful even in the presence of resistance to both isoni-azid and streptomycin providing the regimen includes rifampicin and pyrazinamide in the first 2 months, and rifampicin in the continuation phase. This is because initial resistance to rifampicin and pyrazinamide is low everywhere at present, and is another important justification for the use of the complex four-drug regimens already suggested. Drug sensitivity testing (see p. 107) is performed in recognized centres as part of national surveillance. Individual patient susceptibility testing should occur if the patient continues to produce culture-positive sputum after 3 months. Where adequate facilities exist for determining drug sensitivities, regimens can be devised for individual patients taking account of both initial resistances and resistances emerging during treatment.
Modern drug regimens are designed to prevent the emergence of drug resistance, but control of tuberculosis is threatened by the widespread emergence of drug resistance in M. tuberculosis in certain populations. Outbreaks, sometimes caused by multi-resistant bacilli, have occurred in hospitals, centres for the care of patients with AIDS, and prisons, posing considerable clinical and public health problems. In the UK, resistance to isoniazid alone is presently less than 4 per cent and resistance to isoniazid and rifampicin (with or without resistance to other drugs) is less than 0.6 per cent. However, with the emergence of multiple drug resistance in other parts of the world, enhanced surveillance is required.
Molecular methods that allow the rapid detection and fingerprinting of resistant strains are available in reference centres and, where facilities exist, can facilitate any necessary change of therapy or prompt patient isolation (see below). Such methods also help to differentiate reinfection from reactivation and can be used to characterize cross-infection episodes among close contacts of infectious cases.
Nosocomial transmission of multidrug-resistant M. tuberculosis
Multidrug-resistant tuberculosis is often characterized by a rapid progression from diagnosis to death and a case fatality >80 per cent. Patients are often severely immunocompromised and the clinical diagnosis of tuberculosis may be difficult. As a result there is often delay in initiating isolation of the patient and in the recognition of drug resistance. Health care workers are at great risk from multi-resistant tuberculosis, since these highly infectious patients remain untreated.
Infection control measures require the rapid identification of patients with multidrug-resistant tuberculosis and the immediate isolation of smear positive M. tuberculosis patients in isolation rooms. Negative air pressures should be maintained
in these rooms and patients should be nursed under ‘source control’. To prevent nosocomial transmission, high-efficiency particulate air (HEPA) filters may have to be fitted if there is difficulty in exhausting TB-contaminated air from buildings. Provision of personal respirators to healthcare workers may be necessary.
This form of tuberculous infection is invariably fatal if untreated, and appropriate chemotherapy is vital. It is a frequent complication of untreated miliary tuberculosis, and may vary from an abrupt and severe illness resembling other acute bacterial meningitis to a subtle and chronic disease extending over several months. Treatment should always include isoniazid and rifampicin. The isoni-azid should initially be given in larger than usual dose, with a pyridoxine supplement. During the first intensive phase, these drugs should be accompanied by pyrazinamide and ethambutol. After about 2 months, these can be discontinued, but the isoniazid and rifampicin should be continued for at least a further 8 months. Intrathecal antibiotics are unnecessary, but additional treatment with corticosteroids should be used in severe cases.
Other forms of tuberculosis
Tuberculosis can affect any system in the body including, in decreasing order of frequency, superficial lymph nodes, bone and joints, the genito-urinary tract, the abdomen, the breast, and skin. Patients with non-respiratory tuberculosis can be regarded as non-infectious, providing aerosols are not generated by intervention. Generally, the principles described for the treatment of pulmonary tuberculosis are also appropriate for the treatment of other forms, although few regimens have been given controlled clinical trials. Surgical intervention is sometimes necessary to establish the diagnosis, or to drain large abscesses, relieve pressure from tuberculous masses, or repair damage from scar tissue, but is no longer the mainstay of treatment.
The so-called ‘atypical’ species of mycobacteria that can cause human disease include M. kansasii, M. marinum, M. avium, M. intracellulare,M. fortuitum, and several others. The isolation of pathogenic atypical mycobacteria does not prove that disease is present; colonization with these organisms is not uncommon. Nevertheless, disease caused by these organisms may be indistinguishable from true tuberculosis.
Atypical mycobacteria are usually more resistant to the standard antituber-culosis agents than is M. tuberculosis. However, standard antituberculosis
chemotherapy is usually effective, although it may have to be prolonged and more often need to be accompanied by surgical resection of affected tissue.
Atypical mycobacteria of the M. avium complex have come into prominence since the appearance of AIDS. Patients with defective cell-mediated immunity such as occurs in AIDS seem particularly prone to develop infection with these mycobacteria, and the disease may be associated with a severe ‘wasting syndrome’. These mycobacteria are usually highly resistant to the common anti-tuberculosis drugs and infected patients respond poorly to antimicrobial therapy (see p. 343). Newer compound such as fluoroquinolones, rifamycins, and oxazo-lidinones are under evaluation.
This is a chronic communicable tropical disease caused by infection with M. leprae, and characterized by skin lesions and involvement of peripheral nerves causing anaesthesia, muscle weakness, paralysis, and consequent injury and deformity. Two major types are described: lepromatous and tuberculoid leprosy. In lepromatous leprosy there is diffuse involvement of skin and mucous membranes, with ulceration, iritis, and keratitis; scrapings of skin or mucous membranes contain numerous acid-fast bacilli. The tuberculoid form is more localized, but nerve involvement occurs early; only scanty bacilli are present. In both forms, progress of the disease is slow. The tuberculoid form may heal spontaneously in a few years. Death is usually due to other causes.
The mainstay of treatment has been oral dapsone (diaminodiphenylsulphone; DDS) given for up to 10 years in lepromatous leprosy. However, resistance to dapsone has become common and leprologists now recommend triple therapy, whenever possible, with rifampicin, clofazimine, and dapsone. Rifampicin is more rapidly bactericidal than dapsone, but resistance may arise if it is used alone. It is also expensive, which is a considerable constraint to its use in countries where it is most needed. Clofazimine is effective, but the compound is pigmented and some patients find the discoloration of the skin that it produces unacceptable. Ethionamide or prothionamide are useful, but more toxic, alternatives.
In multibacillary lepromatous leprosy, triple therapy with monthly rifampicin and daily clofazimine (or alternatives) and dapsone should be given for at least 2 years, or until bacilli can no longer be seen microscopically; in paucibacillary disease, 6 months' therapy with rifampicin and dapsone may be adequate. With such long and complicated regimens, patient compliance is naturally a problem, particularly in areas of the world where leprosy is most prevalent. There are preliminary indications that a single-dose triple combination of rifampicin, ofloxacin, and minocycline may be successful in lepromatous leprosy. If con-firmed, this offers an attractive regimen for use in endemic areas. Newer macro-lides, such as clarithromycin and roxithromycin, may also have a role.
It should be remembered that patients with leprosy may also have tuberculosis; regimens used for leprosy treatment are inadequate for therapy of tuberculosis.