Nontuberculous mycobacterial pulmonary infections in children

INTRODUCTION — Nontuberculous mycobacteria (NTM) are a miscellaneous collection of acid-fast bacteria that are widespread in the environment [1]. They have been isolated from numerous environmental sources including water, soil, food products, and domestic and wild animals [2]. Health care-associated transmission has occurred with medical equipment [3-5].

This topic will provide an overview of NTM pulmonary infections in children. NTM lymphadenitis, skin and soft tissue infection, disseminated infection, and bacteremia in children are discussed separately.

●(See "Nontuberculous mycobacterial lymphadenitis in children".)

●(See "Nontuberculous mycobacterial skin and soft tissue infections in children".)

●(See "Disseminated nontuberculous mycobacterial (NTM) infections and NTM bacteremia in children".)

MICROBIOLOGY — More than 170 species of NTM have been identified, not all of which have been documented to cause disease in humans [6-9].

●Classification – NTM pathogens are classified as rapidly growing or slowly growing (table 1). Rapidly growing species grow within seven days and include Mycobacterium fortuitum, Mycobacterium abscessus, and Mycobacterium chelonae. Slowly growing species require several weeks to grow and include Mycobacterium avium complex (MAC), Mycobacterium marinum, and Mycobacterium kansasii. (See "Microbiology of nontuberculous mycobacteria", section on 'Classification'.)

●Disease associations – In children, NTM cause four main clinical syndromes: lymphadenopathy, skin and soft tissue infection, pulmonary disease (predominantly in children with underlying pulmonary conditions), and disseminated disease (predominantly in immune-compromised children). The type of infection varies with the species of NTM and host characteristics.

There is little information about which species of NTM cause pulmonary disease in children. In one series of 17 children with NTM pulmonary disease, MAC was more common in previously healthy children (n = 5) and M. abscessus or M. chelonae in children with cystic fibrosis (n = 8) [10]. MAC (M. avium and M. intracellulare), M. abscessus, and M. kansasii are the most common NTM species that cause pulmonary disease in adults in the United States [11]. In countries with a high burden of tuberculosis, up to 40 percent of mycobacterial isolates are NTM species, most commonly MAC, M. abscessus, and M. fortuitum [12,13].

In a registry of over 16,000 pediatric and adult patients with cystic fibrosis, 20 percent had an NTM species isolated at least once during 2010 to 2014, with the prevalence increasing each year [14]. MAC (M. avium and Mycobacterium intracellulare) and M. abscessus are the species most frequently isolated from children with cystic fibrosis [15-20]. In a retrospective cohort study that was limited to children slow-growing NTM species (eg, MAC) were detected more often in patients who were younger at the time of cystic fibrosis diagnosis (median age 1.2 months, range 0.1 to 2.8 months), and rapidly growing NTM species (eg, M. abscessus) were detected more often in patients diagnosed with cystic fibrosis at an older age (median age 4.5 months, range 1.5 to 84.3 months) [21].

EPIDEMIOLOGY — Estimates of the true burden of NTM infections in children are unavailable, in part because NTM infections may be asymptomatic and because NTM infections are not communicable; reporting of NTM infections is not required in the United States or many other countries [22]. Cystic fibrosis registries provide some of the more robust pediatric data. The overall prevalence of NTM disease appears to be increasing with time (possibly as a result of enhanced detection) [14,23-25]. (See "Epidemiology of nontuberculous mycobacterial infections".)

In case series, the prevalence of NTM among respiratory specimens from patients with cystic fibrosis ranges from 4 to 20 percent [15,17,22,26]. However, isolation of NTM from a respiratory specimen does not necessarily indicate NTM disease. In a prospective study of 682 sputum specimens from 106 children with cystic fibrosis, NTM were isolated from 6.6 percent, but only 1.9 percent had NTM disease (ie, clinical and functional decline in addition to isolation) [26]. Among patients with cystic fibrosis, NTM isolation is more common in teenagers and young adults than in younger patients [15,16,18,26-28]. There is geographic variation in species detection. In one large cystic fibrosis registry, M. abscessus was most prevalent in Hawaii, Florida, and Louisiana, whereas MAC was most prevalent in Nevada, Kansas, Hawaii, and Arizona [14].

NTM is transmitted through environmental sources. They are ubiquitous in the environment and have been isolated from water, soil, food products, and domestic and wild animals [22]. An observational study found near-identical isolates of M. abscessus in cystic fibrosis clinics, suggesting the possibility of person-to-person transmission [19]. However, this is controversial. In a multicenter cohort, when isolates were analyzed with whole genome sequencing, unique strains were detected, even after intensive contact with other patients [29]. The respiratory or gastrointestinal tract is the usual portal of entry for NTM pulmonary disease. The incubation periods are variable [9].

RISK FACTORS — Cystic fibrosis is the most common risk factor for pulmonary NTM disease in children in industrialized nations [18,30]. In adults with cystic fibrosis, highly effective modulator therapy, which improves the function of the cystic fibrosis transmembrane conductance regulator protein, may decrease the risk of pulmonary NTM infection [31]; the impact in children is less clear. Screening patients with cystic fibrosis to make sure that NTM cultures are negative before initiation of anti-inflammatory therapy with a macrolide antibiotic is discussed separately. (See "Cystic fibrosis: Overview of the treatment of lung disease", section on 'Azithromycin'.)

Other risk factors for NTM pulmonary disease include polymorphisms in the cystic fibrosis transmembrane conductance regulatory genes without classic cystic fibrosis, polymorphisms in the natural resistance-associated macrophage protein 1 (NRAMP1) gene, hematopoietic stem cell transplantation, and congenital defects in interferon gamma and interleukin (IL)-12 synthesis and response pathways [32-39]. While children with severe combined immunodeficiency, DiGeorge syndrome, chronic granulomatous disease, and hyperimmunoglobulin M syndrome also have defects in portions of the IL-12 pathway, they are not prone to disseminated infection with NTM species [40]. (See "Mendelian susceptibility to mycobacterial diseases: Specific defects".)

Although the mechanism is unclear, vaping-related lung injury also may be a risk factor for NTM pulmonary infection. There is a case report of three patients who developed NTM pulmonary infections in association with vaping-related lung injury [41].

CLINICAL FEATURES — Clinically significant NTM pulmonary disease is most often described in children with preexisting lung disease (eg, cystic fibrosis) and in children with non-human immunodeficiency virus (HIV) immune deficiency who have disseminated NTM disease; it is less common in patients with HIV infection who have disseminated NTM disease [42,43]. (See "Disseminated nontuberculous mycobacterial (NTM) infections and NTM bacteremia in children", section on 'Risk factors'.)

The clinical features of NTM pulmonary disease are variable and nonspecific [22]. Most patients have chronic or recurrent cough [44]. Other pulmonary symptoms and signs include increased sputum production, dyspnea, hemoptysis, chest pain, rhonchi, crackles, wheezing, and stridor. In patients with cystic fibrosis, use of transmembrane conductance regulator modulator therapy may alter the clinical manifestations of pulmonary NTM [45].

The clinical features of NTM pulmonary disease may depend, to some extent, upon the underlying medical problem(s).

●Constitutional symptoms (eg, fever, weight loss, fatigue, malaise) are common in children with NTM pulmonary disease, whether or not they have medical comorbidities or underlying lung disease [46,47]. However, they are more prevalent in immunodeficient children and children with advanced disease [44,48].

●Pulmonary symptoms of NTM lung disease in children without underlying lung disease include wheezing or stridor that does not respond to bronchodilator therapy; these symptoms may be related to endobronchial granulomas and/or hilar lymphadenopathy [49,50].

In children with underlying lung disease, NTM pulmonary disease is associated with worsening of existing pulmonary symptoms (eg, cough, increased sputum production, exercise intolerance, decline in pulmonary status) [47].

The clinical manifestations of NTM pulmonary disease in patients with cystic fibrosis (including disease due to infection) are discussed separately. (See "Cystic fibrosis: Clinical manifestations of pulmonary disease".)

EVALUATION AND DIAGNOSIS

Diagnostic criteria — The diagnosis of NTM pulmonary disease is established based on the combination of compatible clinical and radiographic findings, the isolation of an NTM species in culture, and exclusion of other conditions in the differential diagnosis (table 2) [2,22]. The clinical, radiographic, and microbiologic criteria must be fulfilled. (See 'Differential diagnosis' below.)

When NTM is isolated from a patient with underlying lung disease, it can be difficult to determine whether NTM represents a true pathogen, colonization, or a contaminant. Consultation with an expert in infectious diseases and/or pulmonary medicine may be necessary. (See "Diagnosis of nontuberculous mycobacterial infections of the lungs".)

Clinical suspicion — NTM pulmonary disease may be suspected in [22,30,49] (see 'Clinical features' above):

●Children with underlying pulmonary disease with worsening of pulmonary symptoms/function, weight loss, or low-grade fever, in whom other causes have been excluded

●Children with cystic fibrosis and exacerbation of pulmonary symptoms that does not respond to routine treatment and antipseudomonal antibiotics (see "Cystic fibrosis: Overview of the treatment of lung disease" and "Cystic fibrosis: Antibiotic therapy for chronic pulmonary infection")

●Children with HIV immune deficiency and disseminated NTM disease

Imaging studies — Imaging of the chest is necessary to diagnose NTM pulmonary disease [22]. In children with underlying pulmonary disease, it is important to compare radiographs and computed tomographic (CT) findings with previous studies to determine if there has been a change. It can be difficult to differentiate between worsening of the primary pulmonary disorder and NTM pulmonary disease. In addition, in patients with cystic fibrosis, use of transmembrane conductance regulator modulator therapy may alter the radiographic manifestations of pulmonary NTM [45].

Radiographic findings in NTM pulmonary disease include:

●Nodular or cavitary opacities on chest radiograph (image 1), or

●Multifocal bronchiectasis with multiple small nodules on high-resolution CT (image 2)

Other radiographic features of NTM pulmonary disease may vary with the clinical scenario. The most common radiographic findings in otherwise healthy children include hilar lymphadenopathy and a tree-in-bud appearance associated with endobronchial spread [49]. Infiltrates are more likely in children with constitutional symptoms [46]. Pleural effusions are uncommon, and cavitary lesions in young children are rare.

Radiographic features of NTM pulmonary disease also may vary depending on the causative species [22]:

●M. avium complex (MAC) – Apical fibrocavitary disease or nodular and interstitial nodular infiltrates often involving the right middle lobe or lingula

●M. kansasii – Cavitary infiltrates in the upper lobes

●M. abscessus – Multilobar, patchy, reticulonodular or mixed interstitial-alveolar opacities with an upper lobe predominance; cavitation may occur

Microbiologic testing — When NTM pulmonary disease is suspected, pulmonary specimens should be sent for histopathology, acid-fast bacillus (AFB) staining, and mycobacterial culture and susceptibility testing (as well as bacterial and fungal stains and cultures as clinically indicated). A positive NTM culture is necessary for diagnosis. Compatible histopathology, positive AFB staining, rapid diagnostic tests, and tests that exclude Mycobacterium tuberculosis support the diagnosis.

It is helpful to discuss specimen collection, transport, and culture processing with the microbiology laboratory before sending tissue or fluid to the laboratory. Most NTM species will not grow in traditional bacterial culture media; rapidly growing NTM species (eg, M. abscessus and M. fortuitum) are an exception [51]. Techniques for maximizing the yield of cultures are discussed separately. (See "Microbiology of nontuberculous mycobacteria", section on 'Culture'.)

●AFB staining – AFB staining provides supportive information but is neither necessary nor sufficient to make the diagnosis of NTM disease. A positive AFB stain increases the likelihood that an isolate is clinically significant but cannot distinguish among NTM species or between NTM species and M. tuberculosis [22]. Negative AFB stains do not exclude a diagnosis of mycobacterial disease. In a prospective series of 104 positive NTM cultures from patients with cystic fibrosis, the AFB smear was positive only in one-third [15]. (See "Microbiology of nontuberculous mycobacteria", section on 'Microscopy' and 'Clinical significance' below.)

●Culture – A positive NTM culture (from sputum, bronchial wash or lavage, transbronchial or other lung biopsy) in a patient with compatible clinical and radiographic findings establishes the diagnosis of NTM pulmonary disease (table 2). In addition, the culture identifies the NTM species and the presence or absence of M. tuberculosis; these results are necessary to determine the clinical significance of the infection and appropriate antimycobacterial regimen.

Mycobacterial cultures may take days to several weeks to grow. The time (in days) to detection of mycobacterial growth may help to distinguish between NTM species.

•The following species typically grow within seven days:

-M. abscessus

-M. fortuitum

-M. chelonae

-Mycobacterium smegmatis

-Mycobacterium mucogenicum

-Mycobacterium peregrinum

•The following species require several weeks for growth:

-MAC

-M. kansasii

Obtaining adequate specimens for culture and proper processing are essential. The yield of cultures may be reduced by insufficient quality (eg, nasopharyngeal swab rather than expectorated sputum) or quantity of a respiratory specimen [52]. (See "Microbiology of nontuberculous mycobacteria", section on 'Culture'.)

●Rapid diagnostic tests – Rapid diagnostic tests confirm the presence of NTM but do not distinguish between colonization and NTM-related disease. They do not satisfy the microbiologic criteria for diagnosis (table 2) [22]. Rapid methods for identification of NTM species include polymerase chain reaction (PCR) and high-pressure liquid chromatography (HPLC).

•PCR – PCR may identify NTM species before identification in culture results are available. Nucleic acid probes for MAC and M. kansasii are commercially available. However, there is little experience with PCR and other molecular modalities for the diagnosis of NTM pulmonary disease in children. In adult patients, real-time multiplex PCR (eg, Xpert MTB/RIF) is increasingly used to differentiate between M. tuberculosis and NTM species such as MAC and M. abscessus [53]. Newer multiplex PCRs can also detect mutations conferring macrolide and aminoglycoside resistance [54,55].

•HPLC – HPLC examines the mycolic acid fingerprint patterns that differ among most species or complexes of mycobacteria and can be used to speciate NTM, including rapidly growing mycobacteria [56]. However, HPLC cannot reliably differentiate between M. abscessus and M. chelonae.

●Excluding M. tuberculosis – Although culture is necessary for definitive diagnosis, interferon gamma release assays (IGRAs) and tuberculin skin testing (TST) may be helpful in differentiating NTM from M. tuberculosis.

Pulmonary NTM and M. tuberculosis coinfection may occur in immunocompromised children. In observational studies in settings where M. tuberculosis is hyperendemic, NTM species were isolated in 3 to 15 percent of positive mycobacterial cultures. In another observational study in a tuberculosis-endemic setting, 2.8 percent of adults with confirmed pulmonary M. tuberculosis had coinfection with NTM [57].

•IGRAs – IGRAs are the optimal test to exclude M. tuberculosis in children with NTM pulmonary disease. They are far more specific than TST. IGRAs measure interferon production by sensitized lymphocytes after exposure to antigens that are predominantly, but not exclusively, found in M. tuberculosis. (See "Tuberculosis infection (latent tuberculosis) in children", section on 'Interferon-gamma release assays'.)

Among the NTM species that most commonly cause pulmonary disease, the IGRAs do not include antigens shared by MAC or M. abscessus, but do include antigens shared by M. kansasii [58,59]. Thus, in a child with clinical and radiographic findings of pulmonary disease, a positive IGRA supports a diagnosis of M. tuberculosis or M. kansasii, a negative IGRA supports a diagnosis of NTM (other than M. kansasii) or another condition in the differential diagnosis, and an indeterminate IGRA is not helpful.

•TST – The TST contains numerous antigens shared by both M. tuberculosis and NTM species. NTM infections may account for many positive PPD reactions [58].

Interpretation of TST depends upon risk factors for M. tuberculosis and receipt of bacille Calmette-Guérin (BCG) vaccination.

-In children without risk factors for M. tuberculosis and who have not received BCG vaccination, TST with ≥5 mm induration at 48 hours supports a diagnosis of NTM infection.

-In children with risk factors for M. tuberculosis or who have received BCG vaccination, TST ≥5 mm does not distinguish between NTM, M. tuberculosis, or reaction to BCG vaccination [22,60-62]. TST ≥15 mm may be more suggestive of M. tuberculosis than NTM but does not exclude NTM. (See "Tuberculosis disease in children: Epidemiology, clinical manifestations, and diagnosis".)

-Independent of risk factors for M. tuberculosis, TST with <5 mm of induration at 48 hours is not helpful in excluding mycobacterial disease. If NTM continues to be suspected in a child with negative TST, specimens should be obtained and sent for microbiologic and histopathologic studies.

Clinical significance — When NTM is isolated from a child with pulmonary symptoms and signs, the clinical significance of the isolate must be determined. The pathogenicity of the various NTM species varies substantially. The possibility that NTM was isolated as a result of contamination or represents colonization of diseased lung also must be considered [22]. Consultation with a specialist in infectious diseases may be helpful in determining the clinical significance of the isolate.

Factors that increase the likelihood of clinical significance include recovery from multiple specimens, recovery in large quantities (ie, AFB-positive specimens), and recovery from a normally sterile site [22].

DIFFERENTIAL DIAGNOSIS — Important considerations in the differential diagnosis of NTM pulmonary disease in children include:

●Progression of primary pulmonary disease (eg, worsening bronchiectasis in patients with cystic fibrosis) – In children with underlying lung disease, it can be difficult to determine whether pulmonary symptoms are related to the progression of the underlying lung disease or NTM pulmonary infection. Mycobacterial cultures of respiratory specimens (eg, expectorated sputum, bronchial wash/lavage, lung biopsy) are necessary.

●M. tuberculosis – The cavitary appearance and endobronchial spread of NTM can mimic radiographic findings of M. tuberculosis. It is important to distinguish tuberculosis from NTM pulmonary disease because they require different treatments; in addition, M. tuberculosis requires contact tracing and airborne isolation precautions, whereas NTM does not.

In a child with risk factors for M. tuberculosis and acid-fast bacillus-positive sputum, NTM and M. tuberculosis must be differentiated by mycobacterial culture and speciation; interferon gamma release assays and tuberculin skin testing cannot reliably differentiate NTM from M. tuberculosis. In a series of respiratory specimens from 17 children in Texas who met criteria for diagnosis of NTM pulmonary disease (table 2), M. tuberculosis was isolated more frequently than all other mycobacterial species (23 versus 17 specimens) [10]. (See 'Microbiologic testing' above.)

It may be necessary to initiate empiric therapy before the distinction between NTM and M. tuberculosis is made. In such cases, coverage for both pathogens is provided in a multidrug regimen. (See 'NTM and M. tuberculosis' below.)

●Other forms of nodular lung disease, including disseminated fungal disease (table 3) – Clinical features (eg, exposures) may help to differentiate NTM from other nodular lung diseases, but microbiologic studies are necessary for definitive diagnosis. (See "Diagnosis and treatment of pulmonary histoplasmosis" and "Primary pulmonary coccidioidal infection" and "Clinical manifestations and diagnosis of blastomycosis".)

●Hypersensitivity reactions, as seen in vaping (ie, use of electronic cigarette devices to inhale aerosolized substances) [63]. (See "E-cigarette or vaping product use-associated lung injury (EVALI)".)

●Pneumocystis jirovecii pneumonia – The ground-glass appearance occasionally seen in NTM pulmonary disease can mimic P. jirovecii pneumonia. Microbiologic studies can be helpful in differentiating between NTM and P. jirovecii pneumonia. (See "Epidemiology, clinical manifestations, and diagnosis of Pneumocystis pneumonia in patients without HIV", section on 'Diagnosis'.)

TREATMENT

Decision to treat — Antimycobacterial treatment generally is warranted for children who meet clinical and microbiologic criteria for NTM pulmonary disease (table 2). However, the potential risks and benefits of antimycobacterial therapy disease (table 4) must be considered before initiating treatment.

As an example, in an adolescent with cystic fibrosis who is awaiting lung transplantation in whom M. abscessus is isolated, the benefits of decreased burden of M. abscessus and potential decreased risk of recurrence posttransplant may outweigh the adverse effects of parenteral multidrug therapy with ototoxic and nephrotoxic drugs [64]. On the other hand, the risks of therapy may outweigh the benefits in a young child in whom a potentially nonpathogenic species or contaminant (eg, M. mucogenicum) is isolated from a respiratory specimen. Similarly, an asymptomatic child with cystic fibrosis with stable pulmonary function testing and chest imaging in whom sputum samples demonstrate an NTM species may not benefit from treatment at that time. (See 'M. abscessus' below.)

Susceptibility testing — Susceptibility testing is recommended for clinically significant NTM pulmonary isolates [22]. The antibiotic susceptibilities of NTM species vary substantially. However, with the exception of macrolides in the treatment of M. avium complex (MAC), drug susceptibility testing for NTM has less correlation with in vivo response than for tuberculosis or other bacterial pathogens.

Given that the diagnostic criteria for NTM pulmonary disease require sequential sputum samples (table 2), speciation and drug susceptibility results often are available before initiation of antimicrobial therapy [22]. However, it may be necessary to initiate empiric therapy before speciation is complete, particularly when M. tuberculosis cannot be excluded. (See 'Diagnostic criteria' above and 'Differential diagnosis' above.)

Antimycobacterial therapy — Suggested antimycobacterial regimens for NTM pulmonary infections are summarized in the table (table 5).

NTM and M. tuberculosis — If the possibility of M. tuberculosis cannot be excluded after clinical, radiographic, and initial laboratory evaluation (eg, children with risk factors for M. tuberculosis and positive smears for acid-fast bacilli [AFB]), we suggest empiric treatment with the following regimen until speciation is available (table 4 and table 5) [65]:

●Isoniazid, plus

●Rifampin (rifampicin), plus

●Ethambutol, plus

●Azithromycin (preferred) or clarithromycin, plus

●Pyrazinamide

Based on in vitro susceptibilities, this regimen provides initial empiric coverage for M. tuberculosis and the two NTM species most frequently isolated from children with NTM pulmonary disease (MAC and M. kansasii). It is consistent with the recommendations of the American Thoracic Society/Infectious Diseases Society of America (ATS/IDSA) and the American Academy of Pediatrics (AAP) Committee on Infectious Diseases [9,11]. This initial regimen can be tailored according to speciation and susceptibility testing when the results are available. (See 'Microbiologic testing' above.)

For children in whom the high pill burden of this five-drug regimen is too difficult, it is reasonable to exclude pyrazinamide. However, if M. tuberculosis is confirmed, the course of therapy should be nine months (rather than the six-month course that can be used for isolated noncavitary pulmonary disease in most immunocompetent children). (See "Tuberculosis disease in children: Treatment and prevention".)

We generally prefer azithromycin to clarithromycin for pulmonary NTM infection. Azithromycin can be administered once daily, and azithromycin suspension is more palatable than clarithromycin suspension [66], which may increase adherence. In addition, there is the potential for drug interaction between clarithromycin and rifamycins – rifamycins can decrease the serum concentration of clarithromycin [67,68]. It is unclear the degree to which the benefit of combination therapy is outweighed by the risk of selecting for clarithromycin resistance. However, use of azithromycin rather than clarithromycin permits combination therapy without potentially sacrificing the minimal inhibitory concentration for the macrolide. A systematic review found no evidence of a difference in the effectiveness of different antimicrobial regimens for NTM pulmonary infection in patients with cystic fibrosis [69].

M. avium complex

●Children without HIV

•Nodular or noncavitating bronchiectasis – We suggest that children without HIV infection children with nodular or noncavitating bronchiectatic MAC lung disease be treated with a three-drug regimen that includes (table 5 and table 4) [9,22]:

-A rifamycin (rifampin [rifampicin] or rifabutin), plus

-Ethambutol, plus

-Azithromycin (preferred) or clarithromycin

We generally prefer azithromycin to clarithromycin for pulmonary NTM infection. (See 'NTM and M. tuberculosis' above.)

The medications are administered three times per week until the sputum cultures have been negative for at least one year [22]. Three-times-weekly dosing for patients with nodular or nonsevere bronchiectatic MAC disease is supported by several prospective studies in adult patients demonstrating effectiveness [70-72] and a systematic review demonstrating increased adherence [73].

•Fibrocavitary or extensive bronchiectasis – We suggest that children without HIV infection children with extensive bronchiectasis or fibrocavitary MAC lung disease be treated with a regimen that includes (table 5 and table 4) [9,22]:

-A rifamycin (rifampin [rifampicin] or rifabutin), plus

-Ethambutol, plus

-Azithromycin (preferred) or clarithromycin), plus

-An aminoglycoside (amikacin or streptomycin)

For the macrolide, we generally prefer azithromycin to clarithromycin. The first three drugs should be administered daily to avoid the interval development of macrolide resistance. The aminoglycoside is also administered daily in children; however, in older adolescents, amikacin or streptomycin may be administered three times per week. In a randomized study in adult patients with tuberculous or nontuberculous mycobacterial infections, the administration of amikacin 25 mg/kg three times per week was not associated with increased toxicity compared with daily dosing [74].

The aminoglycoside should be discontinued after eight weeks; the remainder of the drugs should be continued until the child's sputum cultures have been negative for at least one year [22].

The recommendations for treatment of MAC pulmonary disease in children without HIV infection are based upon in vitro susceptibilities and observational studies in HIV-negative adult patients, in which multidrug treatment with a macrolide-containing regimen was associated with clearance of MAC from sputum in >60 percent of patients [70,75,76]. They are consistent with the recommendations of the ATS/IDSA and the AAP Committee on Infectious Diseases [9,11].

●Children with HIV infection – Respiratory symptoms are uncommon among children with HIV infection with disseminated MAC, and isolated pulmonary disease is rare [43]. We suggest that children with HIV infections and MAC who meet the diagnostic criteria for pulmonary disease (table 2) be treated in the same manner as those with disseminated infection. (See "Disseminated nontuberculous mycobacterial (NTM) infections and NTM bacteremia in children", section on 'Antimycobacterial therapy'.)

M. abscessus — For children with M. abscessus pulmonary disease, we suggest initial treatment with (table 5 and table 4) [9,11]:

●Azithromycin (preferred) or clarithromycin, plus

●Amikacin, plus

●Cefoxitin or a carbapenem (eg, imipenem, meropenem)

Another drug (eg, linezolid, tigecycline) should be added if M. abscessus subspecies abscessus or M. abscessus subspecies bolletii is isolated [11]. These subspecies have inducible macrolide resistance although resistance to other antibiotics remains constant [77]. Although the macrolide may be continued for its immunomodulatory effects, it should not count as one of the active drugs in the regimen.

Although there is limited evidence to determine the optimal combination, the initial regimen should include at least three drugs to which there is documented susceptibility, bearing in mind that in vitro susceptibilities may not correlate with clinical response [22,78]. For children with substantial antibiotic exposures, it may not be possible to find three antibiotics to which the isolate has documented susceptibility. When this occurs, although there is little evidence for guidance, we start with the agent(s) to which M. abscessus is apparently susceptible and then add one or more bactericidal agents, trying to choose those with the fewest adverse effects and the most convenient dosing regimens (eg, oral rather than parenteral medications).

While both azithromycin and clarithromycin have been used to treat M. abscessus lung disease, we prefer azithromycin for several reasons: decreased drug-drug interactions, increased palatability of the liquid formulation, and an observational study suggesting that sustained culture conversion was more common in adults with M. abscessus lung disease who received azithromycin than in those who received clarithromycin [79].

The goal of M. abscessus therapy is not microbiologic eradication. Instead, symptomatic and radiographic improvements should be the endpoints of therapy [80]. There is limited evidence to determine the optimal duration of treatment. In addition to the difficulty in eradicating M. abscessus, it is possible for patients to be reinfected by different M. abscessus genotypes during or after completion of antimicrobial therapy [81]. In a retrospective review of adult patients with M. abscessus lung disease, >40 percent relapsed within a year of treatment cessation; AFB sputum smear positivity was the only consistent predictor of treatment outcome [81]. In another study, lower body mass index, bilateral lung involvement, and fibrocavitary disease were associated with poor prognoses [79].

After four to eight weeks of intensive therapy with at least three drugs with documented susceptibility, two drugs may be used to complete therapy (usually for a total duration of 6 to 12 months). The duration of therapy may be limited by the patient's ability to tolerate the medications [22,78]. M. abscessus has a predilection for infecting diseased lungs and is difficult to eradicate [82]; cure is generally possible only with surgical resection when lung disease is limited [44]. (See 'Surgical therapy' below.)

Parenteral agents with in vitro activity against M. abscessus include amikacin, cefoxitin, tigecycline, and imipenem. Oral drugs with some in vitro activity include macrolides, clofazimine, linezolid, and fluoroquinolones. Linezolid and fluoroquinolones can be used for patients with refractory M. abscessus pulmonary disease. However, the utility of linezolid in children may be limited by bone marrow suppression or optic neuritis, and high rates of fluoroquinolone resistance have been reported in children with no previous exposure to fluoroquinolones [10].

Bedaquiline and delamanid are available for the treatment of drug-resistant M. tuberculosis in adults and may have potential in the treatment of M. abscessus infections [83-86]. However, there is little experience with these agents for the treatment of M. tuberculosis or NTM pulmonary disease in children [87,88]. Consultation with an expert in mycobacterial disease is warranted. (See "Tuberculosis disease in children: Treatment and prevention".)

M. kansasii

●Rifampin (rifampicin)-susceptible M. kansasii – We suggest that children with rifampin (rifampicin)-susceptible M. kansasii pulmonary disease be treated with (table 5 and table 4):

•Rifampin (rifampicin), plus

•Ethambutol, plus

•A macrolide (azithromycin [preferred] or clarithromycin) or isoniazid

If isoniazid is included in the initial regimen, children should receive therapy daily; if a macrolide is used, therapy can be given three times a week. All children with cavitary M. kansasii pulmonary disease should receive daily therapy.

We generally prefer azithromycin to clarithromycin for pulmonary NTM infections. (See 'NTM and M. tuberculosis' above.)

Treatment for rifampin (rifampicin)-susceptible M. kansasii should continue until sputum cultures have been negative for at least one year [22].

●Rifampin (rifampicin)-resistant M. kansasii or intolerance to a first-line medication – We suggest that children with rifampin (rifampicin)-resistant isolates or intolerance to any of the first-line medications be treated with three drugs based upon in vitro susceptibilities. Potential agents include macrolides, fluoroquinolones, aminoglycosides, and trimethoprim-sulfamethoxazole (TMP-SMX) (table 5 and table 4). Treatment for rifampin (rifampicin)-resistant M. kansasii should continue until sputum cultures have been negative for 12 to 18 months [22].

There have been no randomized trials of treatment for M. kansasii. In in vitro studies, M. kansasii is susceptible to macrolides, TMP-SMX, amikacin, fluoroquinolones, rifamycins, bedaquiline, and delamanid [89-93]. In observational studies, multidrug regimens that included rifampin (rifampicin) were associated with clearance of M. kansasii within four months in all patients [94-97]. Relapse, which was rare, was associated with development of rifampin (rifampicin) resistance [94].

Surgical therapy — Surgical resection may be required as an adjunct to antimycobacterial therapy for children with MAC with isolated cavities, M. abscessus, failure to convert sputum cultures after six months, or patients who cannot tolerate antimycobacterial therapy [22]. (See 'Monitoring response to therapy' below.)

In patients with limited pulmonary disease, surgical resection can be curative. In patients with extensive pulmonary disease, surgical resection may be performed in an attempt to decrease organism burden.

Monitoring response to therapy — Children who are being treated for pulmonary NTM disease should be monitored by monthly symptom screening, sputum analyses, and pulmonary function tests. Monitoring for adverse effects of therapy is discussed below. (See 'Monitoring adverse effects' below.)

We do not routinely monitor response to therapy for pulmonary NTM disease with serial CT of the chest, although repeat imaging may be warranted one to two months after initiation of therapy and for children who have clinical deterioration during therapy. Children with pulmonary NTM disease, by virtue of their underlying medical conditions, are likely to have higher cumulative radiation exposure at baseline than children without such conditions [98].

Successful treatment of patients with MAC pulmonary disease is indicated by clinical improvement within three to six months and negative sputum cultures within 12 months of initiation of therapy [75]. Possible reasons for failure to respond within these timeframes include poor adherence/drug intolerance, macrolide resistance, or anatomic limitations (eg, focal cystic or cavitary disease) [22]. An alternative medication regimen or surgery may be warranted. (See 'M. avium complex' above.)

The endpoint of therapy for M. abscessus is symptomatic improvement and prevention of disease progression rather than microbiologic eradication. Microbiologic eradication is potentially possible only with surgical resection of limited lung disease [44]. (See 'M. abscessus' above and 'Surgical therapy' above.)

Although not yet available for clinical use, studies suggest that certain biomarkers may be used to monitor response to therapy and distinguish exacerbations from disease. Levels of antiglycopeptidolipid core immunoglobulin A antibody (anti-GPL-core IgA) fall during treatment and are correlated with symptomatic improvement; anti-GPL-core IgA levels are also higher in patients with MAC and bronchiectasis than those with MAC without bronchiectasis [99]. Urine lipoarabinomannan (LAM) is a point-of-care assay best used to detect tuberculosis in persons living with HIV, but urine LAM may be a marker of progression from pulmonary colonization to disease in patients with MAC [100].

Monitoring adverse effects — The antimycobacterial agents that are used to treat NTM pulmonary infections often are difficult to tolerate and some have important toxicities (table 4). It may be challenging to determine which of the multiple medications required for treatment of NTM pulmonary infections is responsible for a given reaction. In addition, children with comorbid medical conditions may require additional medications that can interact with antimycobacterial therapy, particularly with rifamycins [101]. They also may have baseline hepatic or renal dysfunction, which may decrease medication tolerance.

Although most adverse effects are monitored clinically by assessing the patient for symptoms (eg, dizziness, vomiting), some specific parameters should be monitored at baseline or periodically as indicated below, even in children without comorbid medical conditions [22].

For children receiving combination rifampin (rifampicin), ethambutol, and macrolide therapy (with or without other agents), we suggest baseline complete blood count (CBC), electrocardiogram (ECG), alkaline phosphatase, aspartate aminotransferase (AST), and alanine aminotransferase (ALT). Thereafter, the need for subsequent laboratory evaluation should be driven by symptoms and/or examination findings; scheduled laboratory evaluation may not be needed unless patients are receiving other marrow-suppressive or potentially hepatotoxic medications. If the regimen includes other agent(s), the specific parameters for those agents should be monitored as indicated below. Ethambutol is rapidly metabolized in children and ocular toxicity is extremely rare in children with normal renal function [102,103]. For the treatment of M. tuberculosis in children, the World Health Organization recommends vision screening if possible, but unavailability of vision screening is not a contraindication [104].

For children receiving other regimens, we suggest baseline CBC, alkaline phosphatase, AST, and ALT followed by periodic monitoring of specific parameters for particular agents as follows:

●Rifamycins (rifampin, rifabutin, rifapentine) – Periodic CBC to evaluate granulocytopenia and thrombocytopenia (see "Rifamycins (rifampin, rifabutin, rifapentine)")

●Macrolides (eg, azithromycin, clarithromycin) – Baseline ECG before initiation of therapy to evaluate prolongation of the QT interval; periodic alkaline phosphatase, AST, and ALT during the first three months of therapy to evaluate hepatotoxicity (see "Azithromycin and clarithromycin", section on 'Adverse reactions')

●Isoniazid – Periodic monitoring of AST and ALT may be warranted in children with pulmonary NTM who are at increased risk for liver disease (eg, those with cystic fibrosis) or are receiving other medications that are metabolized in the liver (see "Isoniazid hepatotoxicity")

●Aminoglycosides (including amikacin and streptomycin) – Periodic hearing evaluations and renal function tests to monitor ototoxicity (both sensorineural hearing loss and vestibulitis) and nephrotoxicity; for children receiving amikacin, periodic amikacin levels

●Cefoxitin – Periodic CBC to monitor bone marrow suppression

●TMP-SMX – Periodic CBC to monitor bone marrow suppression

●Linezolid – Periodic CBC to monitor bone marrow suppression; periodic eye examinations by an ophthalmologist to evaluate optic neuritis

Bedaquiline or delamanid may be used in certain children with M. abscessus or M. kansasii pulmonary disease. Adverse effects of bedaquiline include nausea, hepatitis, prolongation of the QT interval, and pancreatitis; obtain baseline alkaline phosphatase, AST, ALT, ECG, and lipase before initiation and periodically during treatment. Adverse effects of delamanid include nausea, vomiting, hepatitis, anemia, eosinophilia, dizziness, peripheral neuropathy, and prolongation of the QT interval; obtain baseline CBC, alkaline phosphatase, AST, ALT, and ECG before initiation.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Nontuberculous mycobacteria".)

SUMMARY AND RECOMMENDATIONS

●Risk factors and microbiology – Nontuberculous mycobacterial (NTM) pulmonary disease in children usually affects patients with underlying lung disease, particularly cystic fibrosis. Mycobacterium avium complex (MAC) and Mycobacterium abscessus are the species most frequently isolated from children with NTM pulmonary disease. (See 'Risk factors' above and 'Microbiology' above.)

●Clinical features – Clinical features of NTM pulmonary disease are variable and nonspecific. Most patients have chronic or recurrent cough. Other findings include increased sputum, dyspnea, hemoptysis, chest pain, abnormal auscultatory findings (eg, crackles, wheezing, stridor), fever, fatigue, and weight loss. (See 'Clinical features' above.)

●Diagnosis and differential diagnosis – Diagnosis of NTM pulmonary disease requires a combination of clinical symptoms, compatible radiographic findings (eg, nodular or cavitary opacities on chest radiograph (image 1), multifocal bronchiectasis with multiple small nodules on high-resolution CT (image 2)), isolation of NTM in culture, and exclusion of other conditions in the differential diagnosis (table 2). (See 'Evaluation and diagnosis' above.)

It is particularly important to exclude Mycobacterium tuberculosis because its treatment and public health implications differ from those of NTM. Additional considerations in the differential diagnosis include progression of primary pulmonary disease, other forms of nodular lung disease (table 3), and Pneumocystis jirovecii pneumonia. Microbiologic studies are necessary to distinguish among the possibilities. (See 'Differential diagnosis' above.)

●Indications for treatment – Although antimycobacterial treatment generally is warranted for children who meet clinical and microbiologic criteria for NTM pulmonary disease (table 2), the risks and benefits of treatment and clinical significance of the isolate must be considered. (See 'Decision to treat' above and 'Clinical significance' above.)

●Treatment regimens – Treatment regimens for NTM pulmonary infections in children are summarized in the tables (table 5 and table 4).

•NTM and M. tuberculosis – For children in whom the distinction between NTM and M. tuberculosis remains uncertain, we suggest empiric treatment with isoniazid, rifampin (rifampicin), ethambutol, pyrazinamide, and a macrolide (azithromycin or clarithromycin) (table 4) rather than other combinations pending speciation and results of antimycobacterial susceptibility testing (Grade 2C). It is reasonable to exclude pyrazinamide if the five-drug regimen is too difficult. (See 'NTM and M. tuberculosis' above.)

•MAC – Treatment of MAC pulmonary disease in HIV-negative children depends upon the severity (see 'M. avium complex' above):

-For children with nodular or noncavitating bronchiectatic disease, we suggest treatment with a rifamycin, ethambutol, and a macrolide (azithromycin preferred) (table 4) rather than other combinations (Grade 2C). Therapy is administered three times per week and continued until sputum cultures have been negative for at least one year.

-For children with cavitary or extensive bronchiectatic disease, we suggest treatment with a rifamycin, ethambutol, a macrolide (azithromycin preferred), and an aminoglycoside (table 4) rather than other combinations (Grade 2C). Therapy is administered daily; thrice-weekly therapy has been used in adolescents and adults. The aminoglycoside is discontinued after eight weeks; the other drugs are continued until the child's sputum cultures have been negative for at least one year.

Treatment of MAC pulmonary disease in HIV-positive children is the same as that for disseminated NTM infections and is discussed separately. (See "Disseminated nontuberculous mycobacterial (NTM) infections and NTM bacteremia in children", section on 'Antimycobacterial therapy'.)

•M. abscessus – For children with M. abscessus pulmonary disease, we suggest initial treatment with a macrolide (azithromycin preferred), amikacin, and either cefoxitin or a carbapenem (table 4) rather than other combinations (Grade 2C). The endpoint of therapy for M. abscessus is symptomatic improvement and prevention of disease progression rather than microbiologic eradication. (See 'M. abscessus' above.)

•M. kansasii

-Rifampin-susceptible – For children with rifampin-susceptible isolates, we suggest combination therapy with rifampin plus ethambutol plus either a macrolide (azithromycin preferred) or isoniazid (table 4) (Grade 2C). Treatment is continued until sputum cultures have been negative for at least one year.

-Rifampin-resistant or intolerance to a first-line agent – For children with rifampin-resistant isolates or resistance to a first-line agent, we suggest treatment with three drugs based on in vitro susceptibilities (Grade 2C). Potential agents include macrolides, fluoroquinolones, aminoglycosides, and trimethoprim-sulfamethoxazole (table 4). Treatment is continued until sputum cultures have been negative for 12 to 18 months.

●Adjunctive surgery – Adjunctive surgical resection may be necessary for children with MAC with isolated cavities, M. abscessus, failure to convert sputum cultures after six months, or patients who cannot tolerate antimycobacterial therapy. (See 'Surgical therapy' above.)

●Monitoring response and adverse events – Children who are being treated for pulmonary NTM disease should be monitored by monthly symptom screening, evaluation for medication-associated adverse events (table 4), sputum analyses, and pulmonary function tests. (See 'Monitoring response to therapy' above and 'Monitoring adverse effects' above.)