Show Summary Details
Page of

Disease caused by environmental mycobacteria 

Disease caused by environmental mycobacteria

Disease caused by environmental mycobacteria

J. van Ingen

and P.D.O. Davies



Chapter heavily revised.

Updated on 31 May 2012. The previous version of this content can be found here.
Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2015. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy).

Subscriber: null; date: 25 February 2017


Introduction—there are over 130 species of mycobacteria; species other than M. tuberculosis complex and M. leprae are collectively referred to as the nontuberculous or environmental mycobacteria. Nontuberculous mycobacteria (NTM) are divided into two groups, the slow growers and the rapid growers. The most common organisms causing human disease are the slow-growing species M. avium complex and M. kansasii and, less commonly, M. marinum, M. xenopi, M. simiae, M. malmoense, and M. ulcerans. The rapid growers that are human pathogens are M. abscessus, M. fortuitum, and M. chelonae.

Ecology and epidemiology—NTM are ubiquitous in the environment and have been isolated from water, soil, domestic and wild animals, milk, and food products. Transmission to humans is though inhalation, ingestion, or traumatic inoculation. The prevalence of NTM infections is likely to have been underestimated, and appears to be increasing in developed countries.

Clinical features—four clinical syndromes have been described: (1) pulmonary disease; (2) lymphadenitis; (3) postinoculation mycobacteriosis; (4) disseminated disease. Cervical lymphadenitis is the most common presentation in children whereas chronic pulmonary disease is more frequent in adults.

Diagnosis—microscopic examination using acid fast stains and culture on appropriate media remain the cornerstone of diagnosis. The use of techniques such high performance liquid chromatography (HPLC), polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) analysis and 16S ribosomal DNA sequencing have enabled more accurate speciation of nontuberculous mycobacteria.

Treatment—this depends on the site and severity of the infection, the presence of predisposing conditions, and the species of mycobacterium. Therapy of disease due to slow growers is usually based on regimens containing clarithromycin or azithromycin; that for rapid growers is largely empirical. Antiretroviral therapy is more beneficial than antimycobacterial agents in patients with AIDS-related disease.


Owing to the advent of molecular tools for identification, the genus Mycobacterium is now known to host over 130 species. The species other than the causative agents of tuberculosis and leprosy (Hansen’s disease) are collectively referred to as nontuberculous mycobacteria (NTM) or environmental mycobacteria. The latter nomenclature reflects the habitats of these mycobacteria and the source of human infections. The environmental mycobacteria are subdivided into slow and rapid growers, according to their rate of growth on subculture.

A small subset of the environmental mycobacteria is capable of causing opportunistic infections in humans; most of these are slow growers. The bacteria of the M. avium complex (MAC, a complex that includes M. avium, M. intracellulare, and several rarely isolated species) are the most frequent causative agents of human infections, followed by M. kansasii, M. ulcerans, M. marinum, M. malmoense, M. xenopi, and M. simiae. Among the rapid growers, only the M. abscessus group, M. chelonae, and M. fortuitum are commonly associated with human infections. The relative frequency of disease caused by these species differs by geographical region. The principal pathogenic environmental mycobacteria and the diseases associated with these species are listed in Table

Table Principal pathogenic environmental mycobacteria and assoicated diseases

Slow growers

M. avium complex

Pulmonary disease, lymphadenitis, disseminated disease

M. kansasii

Pulmonary disease

M. xenopi

Pulmonary disease, spondylodiscitis in HIV-infected patients

M. malmoense

Pulmonary disease, lymphadenitis

M. simiae

Pulmonary disease

M. szulgai

Pulmonary disease

M. marinum

Cause of fish tank granuloma or swimming pool granuloma

M. ulcerans

Cause of Buruli ulcer disease

M. haemophilum

Swimming pool granuloma

M. ulcerans

Buruli ulcer

M. haemophilum

Lymphadenitis, skin disease in transplant recipients

M. terrae complex

Wound infections after soil contamination, tenosynovitis

M. gordonae

Common in the environment, rare cause of disease

Rapid growers

M. abscessus

Pulmonary disease, disseminated skin disease

M. chelonae

Pulmonary disease, disseminated skin disease (both rare)

M. fortuitum

Pulmonary disease, postinoculation localized skin infections

Environmental mycobacteria cause two named diseases with characteristic features: swimming pool granuloma caused by M. marinum and Buruli ulcer caused by M. ulcerans. Disease due to other environmental mycobacteria is much less specific, often resembles tuberculosis, and requires identification of the causative organism for diagnosis.

Ecology and epidemiology

The environmental mycobacteria are particularly associated with soil and water. They have been isolated from various natural waters, varying from swamps to oceans, as well as from treated tap water. NTM have also been isolated from domestic and wild animals, milk, and food products. Transmission to humans is by aerosol inhalation, ingestion, or traumatic inoculation. Skin test surveys have revealed that human infection is widespread and common, though overt disease is rare. Infection by environmental mycobacteria may give rise to false-positive tuberculin skin test results and may affect the efficacy of BCG vaccination. This may explain, in part, the diversity of protection by BCG seen in various trials.

The incidence of overt disease likely results from an interplay between host susceptibility, virulence, and load of the various environmental mycobacteria in the local environments and opportunities for infection. Human transmission of overt disease is thought not to occur or to be highly exceptional.

The frequency of disease caused by different species of NTM is unknown; this is because, unlike tuberculosis, reporting of cases is not mandatory. In the United States a nationwide survey of 32 000 mycobacterial isolates conducted in 1979 to 1980 indicated that 2/3 were NTM. The estimated prevalence of NTM was 1.8 per 100 000 population in the 1980s. Clinical and laboratory studies from the United States, Canada, and Australia indicate that the burden of NTM has been underestimated and is increasing in developed countries. This may be a result of increased clinical attention, improved laboratory techniques for detection, and a growing number of people at increased risk because of immunosuppressive drug use, chronic pulmonary diseases, and HIV infection.

Clinical features

The NTM cause four main types of disease: pulmonary, lymphadenitis, postinoculation, and disseminated.

Pulmonary disease

Chronic pulmonary disease

Chronic pulmonary infections are the most frequent disease manifestation of NT|M. Estimates of the incidence of pulmonary disease caused by environmental mycobacteria differ from 1 per 100 000 population per year in Denmark to 4.3 per 100 000 population per year in Ontario, Canada. In many regions, the incidence of environmental mycobacterial disease in the middle-aged and elderly white population exceeds that of tuberculosis.

Two distinct disease entities exist; the cavitary disease type, radiologically similar to tuberculosis (see Fig., affects patients with pre-existent pulmonary diseases, especially chronic obstructive pulmonary disease. As a result, it is more common among men and usually appears in their late 50s or 60s. The nodular-bronchiectatic disease type (see Fig. Fig. is a more subtle disease that mostly affects the lingula and middle lobe. This disease type is more common among female lifetime nonsmokers with no significant pulmonary history.

Fig. Chest radiograph of a patient with right upper lobe cavitary M. avium disease.

Chest radiograph of a patient with right upper lobe cavitary M. avium disease.

Fig. CT image of nodular bronchiectatic M. intracellulare pulmonary disease.

CT image of nodular bronchiectatic M. intracellulare pulmonary disease.

The symptoms of cough, malaise, weight loss, and reduced exercise tolerance develop over months or even years. Especially for the cavitary disease type, clinical distinction from tuberculosis is difficult, though its course is more prolonged. Diagnosis relies on isolation and accurate identification of the causative agents. Because these are environmental organisms, a single culture yielding environmental mycobacteria is insufficient for diagnosis. Positive cultures from nonsterile samples such as those from the respiratory tract can result from accidental presence after environmental exposure or contamination during sample acquisition or handling. Hence, clinical and radiological as well as microbiological (i.e. multiple positive cultures yielding the same species) signs of infection must be obtained and other disease rigorously excluded to make a diagnosis of true environmental mycobacterial disease. Especially in the nodular-bronchiectatic disease type, bronchial washings and CT imaging are often required for diagnosis and follow-up.

Acute pulmonary disease

Environmental mycobacteria, especially MAC, can cause a hypersensitivity pneumonitis. Exposure is often from indoor spas, hence the name ‘hot tub lung’. This acute or subacute disease results from either inflammation after antigen exposure, or true infection, or both. Dyspnoea, cough, and fever are the most common symptoms. Occasionally, hypoxemic respiratory failure may occur and require intervention. CT reveals diffuse infiltrates with prominent nodularity of all lung fields. The optimal treatment remains controversial and corticosteroids, antimycobacterial treatment, or both can be successful. Interrupting exposure to the mycobacteria is the most important intervention.


Lymphadenitis is the second most frequent environmental mycobacterial disease. It predominantly, though not exclusively, affects immunocompetent children under the age of 8 years. Cervicofacial lymph nodes are most frequently affected, although infection of axillar and inguinal lymph nodes has been reported. Disease that involves the abdominal lymph nodes is observed in HIV-infected patients. In these patients, as well as in otherwise immunocompromised patients, lymphadenitis can be a sign of disseminated disease (see below).

Lymphadenitis is generally caused by slow-growing environmental mycobacteria, mostly M. avium complex, M. haemophilum, M. malmoense, and M. kansasii. The different species seem to affect children of different ages, with M. avium affecting the youngest. The risk is reduced by neonatal BCG vaccination. Surgical treatment is curative and lymph node excision is preferred over incision and drainage, which may lead to sinus formation. A 3 month regimen of rifabutin and clarithromycin or a wait-and-see policy can be successful in selected cases.

Postinoculation mycobacterioses

Postinoculation mycobacterioses affect the organs that have immediate interactions with the environment, i.e. the skin and the eyes. It remains unknown whether the mycobacteria are permanent members of the human skin microbiome. Skin disease caused by NTM need not be a postinoculation disease; it may be a sign of disseminated disease (see below).

Localized skin infections

NTM cause two named postinoculation skin diseases with characteristic clinical features: Buruli ulcer disease is a severe skin infection by M. ulcerans, presenting as nodular or, in later stages, ulcerative lesions and is endemic to parts of West Africa, Australia and Latin America, with minor pockets in East Asia. The source of M. ulcerans infection remains controversial, although water insects may be vectors. This disease is covered in Chapter 7.6.28. The swimming pool granuloma or fish tank granuloma is a localized nodular or pustular, sometimes ulcerative, skin lesion resulting from local infection of an existing skin abrasion by M. marinum. The infection is acquired during swimming or fish tank cleaning activities. There may be ‘sporotrichoid’ spread of lesions along the draining lymphatics. The disease can be self-limiting, but chemotherapy accelerates resolution. Local spread of the infection can occur and lead to tenosynovitis, osteomyelitis or even disseminated disease.

Most other cases of postinoculation environmental mycobacterioses are caused by rapid-growing M. fortuitum and M. chelonae. These include injection site abscesses and footbath-associated furunculosis. These diseases present as sporadic cases, though miniepidemics may be noted as a result of reusing of contaminated drug vials or needles or suboptimal hygiene measures in nail salons or other spas. Injection site abscesses may take months to develop and are either localized abscesses or multiple abscesses with spreading cellulitis. The latter occurs in patients who inject frequently, e.g. insulin-dependent diabetics. Surgical excision or drainage cures localized disease; 2 to 4 months of antibiotic treatment can be warranted for multiple or spreading lesions.

Tenosynovitis caused by environmental mycobacteria is rare (Fig.; gardeners seem to be at increased risk and inoculation occurs in wounds from thorns or other plant material. Bacteria of the M. terraeM. nonchromogenicum complex are the most frequent causative agents and related to wound contamination with soil. In rare cases, M. kansasii, M. malmoense, and rapid growers have been isolated.

Fig. Erythematous swelling in tenosynovitis caused by M. malmoense.

Erythematous swelling in tenosynovitis caused by M. malmoense.

Eye infections

Trauma to the cornea can lead to infection by rapid-growing M. fortuitum or M. chelonae. These localized infections respond well to topical treatment with combinations of macrolides, quinolones and aminoglycosides. Corneal grafting and systemic therapy may be warranted in severe cases.

Accidental inoculation may occur during surgery with contaminated materials and can lead to severe infections. Osteomyelitis of the sternum and endocarditis with septicaemia has been reported after cardiac surgery. Again, causative agents are mainly rapid growers.

Disseminated disease

Prior to the HIV pandemic, disseminated infections by environmental mycobacteria were rare and restricted to patients with congenital immune deficiencies. Disseminated disease caused by M. avium (or, less frequently, M. genavense or M. simiae) was an important and frequently lethal clinical entity during the early phase of the HIV pandemic, before the advent of highly active antiretroviral therapy (HAART). This was particularly true for countries with a low tuberculosis burden. Disseminated M. avium infection was far less frequent in HIV-infected patients Africa. Dissemination of the causative mycobacteria was thought to start from the intestines, as many patients were known to harbour M. avium in their faeces before the onset of disseminated disease.

Since the introduction of HAART disseminated environmental mycobacterial disease has become infrequent in HIV-infected patients. At the same time, notification of this disease has not diminished, as more cases are now diagnosed in patients who are treated with immunosuppressive drugs, mostly after solid organ transplantation or in patients with haematological malignancies. In these ‘new’ patient categories, the dominant causative agents are M. avium, M. genavense, M. haemophilum, and M. chelonae. Disseminated disease presents with two distinct clinical syndromes. M. avium and the difficult to culture M. genavense cause a nonspecific disease with symptoms of fever, weight loss, night sweats, malaise, and anaemia (or, in M. genavense disease, pancytopenia); diarrhoea, abdominal lymph node enlargement and abdominal pain are frequent, especially in patients with HIV infection. The diagnosis is usually made by culture of bone marrow, liver or other biopsies, or by blood culture. M. haemophilum and the rapid growers cause a disseminated disease with subcutaneous abscesses, nodular lesions, or skin ulceration. This disease appears to be more common in patients with haematological malignancies. Their localization to the skin has been related to these species’ preferences for lower temperatures. Diagnosis is usually made by culture and histological examination of biopsies of lesions, or blood cultures. Disease caused by M. haemophilum can be difficult to diagnose as the bacteria need an external iron source (e.g. blood, hence its name) for in vitro growth.


Microscopic examination using acid fast stains and culture on appropriate media remain the cornerstone of diagnosis. Specimens may be stained with the Ziehl–Neelsen stain or one of its modifications, e.g. Kinyoun stain, and appear pink as a result of staining with carbol-fuschin. Microscopy is relatively insensitive as it requires at least 10 000 organisms per ml of sputum for smear positivity. The sensitivity of microscopy may be improved by use of a fluorochrome stain such as auramine-O or auramine-rhodamine and examination by fluorescence microscopy.

Mycobacterial culture is more sensitive but more time-consuming than microscopy as it requires specialized equipment and a containment level 3 facility. Nonsterile specimens such as sputum should be decontaminated before culture in order to eliminate more common bacteria or fungi that would overwhelm growth of mycobacteria. Sterile samples such as serous fluids, blood, or cerebrospinal fluid can be inoculated directly on to appropriate solid media (e.g. Middlebrook 7H11) or liquid media (e.g. BACTEC 12B broth or Mycobacteria growth indicator tube, MGIT).

Once cultures have grown speciation is performed by high performance liquid chromatography (HPLC), polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) analysis, or 16S ribosomal DNA sequencing, which has enabled more accurate speciation of NTM.

Susceptibility testing of NTM is done in specialist reference laboratories.


The choice of therapy depends on the causative agents and their in vitro susceptibility, the predisposing conditions and their prognosis, and the site of disease as well as its severity. In general, there is a lack of evidence for the efficacy of regimens as very few clinical trials have been performed.

For skin disease caused by M. marinum, drug susceptibility tests have a limited role as the disease usually responds to monotherapy with doxycycline, minocycline, or trimethoprim-sulfamethoxazole, or the combination of rifampicin and ethambutol. Multidrug therapy may be indicated in severe, spreading disease. Surgical excision, curettage, or drainage cures localized skin disease caused by rapid growers (see above) and surgical excision is the treatment of choice for lymphadenitis and even single nodular pulmonary lesions. For extrapulmonary disease by rapid growers where chemotherapy is needed, results of drug susceptibility tests should guide the selection of a regimen. A minimum of two active drugs is needed, based on the severity of disease and a treatment duration of 3 to 6 months may be indicated; timing of clinical improvement guides the treatment duration. For extrapulmonary and disseminated disease caused by slow-growing species, mainly M. avium, treatment regimens should include a macrolide (clarithromycin, azithromycin), a rifamycin (rifampicin, rifabutin), and ethambutol.

Pulmonary disease by environmental mycobacteria is difficult to treat; the long treatment duration and drug toxicities are a significant burden for patients. For disease caused by slow growers, mainly MAC, drug susceptibility results are only helpful for the macrolides. In case of macrolide susceptibility, most clinicians have adopted the use of macrolides, combined with rifampicin and ethambutol, despite limited evidence for additional efficacy of macrolides (Table These regimens should be used for a total duration of 24 months or up to 1 year after culture conversion. The notable exception is M. kansasii for which short (9 month) regimens of rifampicin and ethambutol are highly effective. The role of quinolones in pulmonary disease by slow growers seems limited.

Table Recommended regimens for treatment of pulmonary infections caused by the more usually encountered slow-growing environmental mycobacteria in HIV-negative patients



Areas of uncertainty

M. avium complex

18–24 months of rifampicin, ethambutol and a macrolide


24 months of rifampicin and ethambutol

Role of macrolides, role of aminoglycosides in severe disease

M. kansasii

9 months rifampicin and ethambutol

M. xenopi

24 months rifampicin and ethambutol

Role of macrolides and quinolones

M. malmoense

24 months rifampicin and ethambutol

Role of macrolides and quinolones

For pulmonary disease by rapid growers, mostly the M. abscessus group and M. fortuitum, drug susceptibility results guide the selection of drug regimens. For M. abscessus group infections, a macrolide combined with amikacin and cefoxitin, tigecycline or imipenem is often used. For M. chelonae, cefoxitin is inactive and tobramycin is more active than amikacin. Macrolides may not be effective against M. fortuitum owing to natural resistance and a multidrug regimen that combines a quinolone with doxycycline, trimethoprim-sulfamethoxazole, an aminoglycoside or imipenem can be used. Treatment duration in pulmonary disease by rapid growers is usually 4 to 6 months.

Cure rates of pulmonary disease by environmental mycobacteria are limited, in the 50 to 70% range; M. kansasii disease has more favourable outcome. Adjunctive surgical resection of the affected areas of the lung improves outcomes in selected cases.

To achieve success in the treatment of environmental mycobacterial disease, optimal treatment of the underlying and predisposing conditions is vital.

Further reading

Falkinham JO III (2009). Surrounded by mycobacteria: nontuberculous mycobacteria in the human environment. J Appl Microbiol, 107, 356–67.Find this resource:

Griffith DE, et al. (2007). An official ATS/IDSA statement: diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J Respir Crit Care Med, 175, 367–416.Find this resource:

Lindeboom JA, et al. (2007). Surgical excision versus antibiotic treatment for nontuberculous mycobacterial cervicofacial lymphadenitis in children: a multicenter, randomized, controlled trial. Clin Infect Dis, 44, 1057–64.Find this resource:

Research Committee of the British Thoracic Society (2008). Clarithromycin vs ciprofloxacin as adjuncts to rifampicin and ethambutol in treating opportunist mycobacterial lung diseases and an assessment of Mycobacterium vaccae immunotherapy. Thorax, 63, 627–34.Find this resource:

Subcommittee of the Joint Tuberculosis Committee of the British Thoracic Society (2000). Management of opportunist mycobacterial infections: Joint Tuberculosis Committee guidelines 1999. Thorax, 55, 210–18.Find this resource:

van Ingen J, van Soolingen D (2011). Cervicofacial lymphadenitis caused by nontuberculous mycobacteria; host, environmental or bacterial factors? Int J Pediatr Otorhinolaryngol, 75, 722–3.Find this resource:

Wolinsky E (1979). Nontuberculous mycobacteria and associated diseases. Am Rev Respir Dis, 119, 107–59.Find this resource: