Nontuberculous mycobacteria are ubiquitous in the environment. They exist in food, soil and water, on the surface of many plants, and in buildings, particularly within water pipes. For many years they were thought to be implicated in human disease only as saprophytic contaminants in tuberculous lesions. Now, several species are recognized to be facultative parasites capable of causing chronic granulomatous diseases that can be pathologically indistinguishable from tuberculosis.
These infections are not readily identified because the causative bacteria can be distinguished from M. tuberculosis only by specialized methods. However, they have latterly attracted attention for two reasons. Firstly, wherever tuberculosis has declined within the population at large, they now account for a greater proportion of cases of granulomatous disease. Secondly, like tuberculosis, they have emerged as common secondary infections among patients with acquired immunodeficiency syndrome (AIDS).
Most localized infections result from inoculation of organisms into the skin. Pulmonary infection usually occurs only in patients with predisposing disease, while disseminated infection is confined almost exclusively to patients with impaired immune responses. Thus far, there is no evidence of case-to-case transmission.
Clinically, four types of disease are described:
Localized cutaneous lesions
Inoculation of organisms into superficial abrasions and into puncture wounds can result in the formation of localized nodular or ulcerative lesions. The organisms most commonly implicated are M. marinum, which colonizes swimming pools and fish aquaria, and M. ulcerans, which is largely restricted to Australia and some tropical regions and causes deep necrotic lesions known as Buruli ulcers. M. haemophilum has more recently been associated with similar lesions in immunosuppressed patients. Abscesses resulting from contaminated injections have most frequently been attributed to two rapidly growing species, M. chelonei and M. fortuitum. More such cases are to be anticipated among drug addicts who are immunosuppressed as a result of AIDS but, as yet, most have occurred either among diabetics or following the injection of contaminated drugs and vaccines.
The lung is the most frequent site of opportunistic mycobacterial infection and the lesions are clinically and radiologically indistinguishable from pulmonary tuberculosis. Predisposing conditions include chronic bronchitis, occupational dust-induced diseases, residual tuberculous lesions, cystic fibrosis, carcinoma, AIDS and other causes of immunosuppression. Most recorded cases have been attributed to the M. avium-intracellulare complex, M. kansasii and, to a lesser extent, M. xenopi, but in some regions M. scrofulaceum, M. chelonei, M. szulgai and M. malmoense are also of significance. Any of the organisms implicated in pulmonary disease can also cause lymphadenitis and other forms of nonpulmonary disease more commonly associated with M. tuberculosis.
The lesions are usually unilateral and self-limiting and most cases occur in children under 5 years old. However, lymphadenitis is also sometimes a prominent feature of disseminated disease in adults. Most reported cases have been attributed to the closely related M. avium-intracellulare complex and M. scrofulaceum (known as the MAIS complex).
Single or multiple foci of granulomatous disease can occur in virtually any system or organ and, when cellular immunity is depressed, dissemination of the infection can occur as rapidly as in miliary tuberculosis. Most such cases have been attributed to the M. avium-intracellulare complex and to M. chelonei.
Diagnosis is dependent upon the clinical characteristics of the disease and identification of the causative organism, when this is possible. Although all mycobacterial infections are presumed to give rise to a positive result in the tuberculin test, seroconversion is most likely to have resulted from previous infection with M. tuberculosis. The test is thus of little, if any, practical value in the diagnosis of nontuberculous infections.
It is not known whether BCG vaccines protect against infection by any of the nontuberculous mycobacteria and, as yet, no specific vaccines have been developed. The management of established infection is determined by the anatomical focus of the disease, the identity of the organism, the age of the patient, and the competence of the immune system.
Deep and widely disseminated infections can be treated only by chemotherapy. However, the response, even when treatment is prolonged, is uncertain, and surgical resection-now rarely employed in tuberculosis-remains of value in localized nontuberculous pulmonary disease. Surgical excision is also frequently used to hasten resolution of localized lymphadenopathy and of solitary skin lesions, even though these lesions are likely to be self-limiting.
The isolation of nontuberculous mycobacteria from biopsy of a chronic granulomatous lesion is generally evidence of a causal association. However, the identification of these ubiquitous facultative parasites in sputum or urine requires guarded interpretation as their presence does not necessarily imply a pathogenic role. Only when tuberculosis has been rigorously excluded and positive cultures have been consistently obtained over a period of several weeks should the patient be committed to the prolonged, costly and sometimes hazardous courses of chemotherapy required. Whenever possible, the identity of the causative organism and its sensitivity to candidate antibiotics should be established within a specialized reference laboratory. However, in vitro sensitivity tests can be misleading, and treatment may need to be determined empirically on the basis of published case reports and retrospective surveys.
Selection of chemotherapeutic agents
Most experience has been gained in the treatment of localized pulmonary disease caused by the more prevalent and relatively slowly growing M. kansasii, M. malmoense and, in immunocompetent hosts, the M. avium-intracellulare complex. These species are usually ultimately responsive to standard antituberculosis chemotherapy-even though the M. avium-intracellulare complex can be relatively resistant in vitro - but it is often necessary to administer a combination of rifampicin, ethambutol and isoniazid for at least 18 to 24 months.
Several other antibiotics not normally used in antituberculosis therapy have been claimed to be of value. These include erythromycin in infections due to M. kansasii, M. scrofulaceum and, less reliably, the M. avium-intracellulare complex, and the combination of sulfamethoxazole and trimethoprim in infections attributed to the M. avium-intracellulare complex, M. chelonei, M. marinum and M. xenopi. Reports also exist of M. chelonei and M. fortuitum infections responding to a combination of amikacin and doxycycline, and of M. marinum infections responding to minocycline. However, the evidence supporting the use of these drugs is largely anecdotal, and firm recommendations cannot be made.
Clofazimine, which is concentrated in fatty tissues and in cells of the reticuloendothelial system, is claimed to be of particular value in the suppression of disseminated disease due to the M. avium-intracellulare complex. It has been advocated in combination with rifamycin in the management of opportunistic infections in patients with AIDS, but insufficient information is currently available to determine whether this regimen has significant effect on morbidity and survival time.