Approach to the diagnosis and treatment of non-tuberculous mycobacterial disease.

Non-tuberculous mycobacteria (NTM) is a collective name given to a group of more than 190 species of Mycobacterium. The clinical presentation for most NTM infections is non-specific, often resulting in delayed diagnosis. Further complicating matters is that NTM organisms can be difficult to isolate. Medications used to treat NTM infection can be difficult for patients to tolerate, and prolonged courses of anti-mycobacterial therapy are often required for adequate suppression or eradication. Herein, we review different NTM syndromes, appropriate diagnostic tests, and treatment regimens.

Introduction

Non-tuberculous mycobacteria (NTM) is a collective name given to a group of more than 190 species of Mycobacterium other than Mycobacterium tuberculosis and Mycobacterium leprae [1]. These organisms are ubiquitous in the environment, in particular in soil and water sources, and are increasingly recognized as a cause of human disease. Non-tuberculous mycobacteria cause infection not only in immunocompromised populations, but also in otherwise healthy individuals. A substantial number of challenges are associated with the diagnosis and management of NTM infections.

The clinical presentation for most NTM infections is non-specific, often resulting in delayed diagnosis. Once NTM is suspected, growing NTM from biological specimens for identification can be difficult secondary to variable growth time and requirements. Furthermore the actual identification of the organism often requires molecular methods that may not be widely available. Identification of NTM from biological specimens is not equivalent to a diagnosis of NTM infection; the ubiquitous presence of NTM in the environment creates the dilemma of distinguishing between colonization and true infection.

The treatment of NTM infections can also be challenging. Antimicrobial-susceptibility testing (AST) is not available for all NTM species and drugs of interest. Where it is available, it is not always reliable and the clinical applicability is uncertain. Optimal dosing of recognized anti-mycobacterial drugs is unknown, and guidelines for the optimal number of drugs and treatment duration are largely opinion-based. Furthermore, current guidelines for NTM infections lump NTM into a small number of categories. Thus differences among NTM species factors such as virulence, relevant host factors, resistance profile, and response to treatment are inadequately addressed. Herein, we present our approach to the diagnosis and treatment of NTM infections. We do not intend for this paper to serve as a detailed guidance for the management of individual NTM infections or species, but rather as a description of a higher level overall approach to the recognition, diagnosis, and management of these infections.

When should one suspect NTM infections?

The clinical manifestations of NTM infections are typically non-specific. In general, NTM infection is suspected in an individual with persistent symptoms (for example fever, respiratory symptoms, skin lesions) that do not respond to traditional anti-bacterial therapy. Alternatively, an understanding of common NTM infection syndromes and the context/clinical scenario in which they occur, can heighten the suspicion for the presence of an NTM infection. Pulmonary NTM infection continues to be the most common clinical presentation [2], but there are several other distinct clinical syndromes described with NTM. These are classified by organ system of involvement and summarized in Table 1. Additionally, several NTM species have unique clinical syndromes associated with them. These should be considered in appropriate clinical scenarios and are summarized in Table 2.

Table 1

Common non-tuberculous mycobacterial infection syndromes.

Clinical Syndrome	Most Common Species	Other Species	Risk Factors/ Associations
Pulmonary44, [110]	MACM. kansasii	M. abscessusM. fortuitumM. szulgaiM. xenopiM. celatumM. asiaticumM. shimoidei	α1-Antitrypsin deficiencyCiliary dyskinesiaCystic fibrosisGERDBronchiectasisPulmonary histoplasmosisLow BMI with scoliosis, pectus excavatum, mitral valve prolapseSmoking or heavy alcohol use (M. malmoense)Lymphopenia
		M. malmoense (Scandinavian Peninsula; Northern Europe)
Cervical Lymphadenitis3, [4], [111]	MACM. scrofulaceum	M. malmoense	Children; 1–5 yrsMen with lung cancer
Skin, soft tissue infection	M. ulceransM. kansasii	M. szulgaiMAC	Bathing in water from bore holes insect bitesPossible association with pedicures
Tenosynovitis	M. marinum*MACM. abscessus	M. terraeM. szulgaiM. malmoenseM. xenopi	*Exposure to fish tanks or other contaminated marine water sources
Bone disease [112], [113]	M. kansasiiM. fortuitum	MACM. xenopiM. chelonaeM. abscessus	May occur via direct trauma or hematogenous seeding
Disseminated disease [114], [115]FUOBSIAbscessesPeritonitis	MAC	M. kansasiiM. chelonaeM. abscessusM. haemophilum M. scrofulaceumM. malmoense	Severe Immunosuppression:HIV infection/AIDS (CD4 count <50 cells)SOT or stem cell transplantlong-term, high dose steroids

Abbreviations: BMI – Body Mass Index, BSI – Blood Stream Infection, FUO – Fever of Unknown Origin, HIV – Human Immunodeficiency Virus, MAC – Mycobacterium Avium Complex, SOT – Solid Organ Transplant.

Table 2

Clinical syndromes associated with non-tuberculous mycobacteria.

Clinical Syndrome	Common Species	Features	Risk Factors/Associations
Prosthetic valve endocarditis [116], [117]	M. chimaera	Presents months to years after index surgery	Cardiac surgery or LVAD
Buruli Ulcer (cutaneous/subcutaneous granulomatous lesion)	M. ulcerans	Areas of body with skin temperatures <30 °C	Common in Central Africa, Asia, South America
Fish tank or swimming pool granuloma (Fish Fancier’s Finger)	M. marinum	Preferential growth in cooler temperatures (27–32 °C)	Exposure to fish tanks or other contaminated marine water sources from direct inoculation
Lady Windermere Syndrome [44] (fibronodular pulmonary disease)	MACM. fortuitumM. chelonaeM. abscessus	“Tree-in-bud” pattern preferentially affecting RML or lingula on CT chest	Women over 50 years; Non-smokers
Hot tub lung [35] (Hypersensitivity lung disease)	MACM. fortuitumM. immunogenum	CT chest mimics hypersensitivity pneumonitis	Indoor hot tub use
CLABSI [118]	M. mucogenicum		Long term central catheters; contaminated water source
Postoperative wound infection [119]	M. fortuitumM. chelonaeM. abscessus		Cosmetic surgery
Peritoneal dialysis associated infection [120]	M. fortuitumM. chelonaeM. abscessus

Abbreviations: CLABSI – Central Line Associated Blood Stream Infections, CT – Computed Tomography, LVAD – Left Ventricular Assist Device, MAC – Mycobacterium Avium Complex, RML – Right Middle Lobe.

Diagnosing non-tuberculous mycobacteria infections

There are no pathognomonic tests for diagnosis of NTM infections. Diagnosis requires correlation of the clinical syndrome with radiographic findings, and ultimately microbiologic confirmation of the etiologic agent. Specimen collection and testing may need to be repeated on several occasions to confirm or exclude a diagnosis of NTM infection. In view of the complexities involved, tests should be interpreted in consultation with specialists experienced in management of NTM infections, especially if the microbiologic results are not congruent with the clinical picture.

Microbiologic diagnosis

NTM are ubiquitous in the environment and are commonly found in soil and water, including municipal and household water supply systems. Therefore, caution must be exercised to avoid environmental contamination of biologic specimens to prevent false positive results. Specimens can be obtained from nearly any tissue or organ suspected to be infected.

The guidelines provided by American Thoracic Society (ATS) and Infectious Disease Society of America (IDSA) for diagnosis of pulmonary NTM infection require positive sputum culture results on at least two separate occasions [3], [4]. Alternatively, a single positive culture from a lower respiratory specimen such as bronchial lavage or washing is considered adequate for diagnosis. If clinical suspicion for pulmonary NTM infection is high and initial tests remain non-diagnostic, then additional sputum cultures should be obtained [3], [4]. Induction of sputum or bronchoscopy may be necessary in patients unable to expectorate sputum. The optimal interval between sample collections is unknown. Separating specimen collection by at least a week may be a reasonable timeframe to consider as the interval between sample collections, such that transient environmental contamination does not confuse the diagnosis [5]. Lung biopsy is rarely described as necessary for diagnosis, but may provide additional information beyond positive stains and cultures, demonstrating histopathology findings of mycobacterial infections, such as granuloma formation.

In non-pulmonary disease, source of culture is determined by site of involvement. For example, lymphadenitis may require excisional biopsy or fine needle aspiration for microscopy and mycobacterial culture. When NTM disease is disseminated, mycobacterial blood cultures may prove diagnostic. In the presence of cytopenias, bone marrow or liver biopsy may be useful for diagnosis [6]. It is important to note that skin infections are best evaluated with biopsy rather than superficial swabs since swabs may represent environmental exposure rather than infection [6].

When clinical suspicion exists, the requisite biologic samples should be obtained and sent to the microbiology laboratory, with notification of the suspicion for NTM infection, so that the laboratory personnel process these specimens appropriately. Different staining and culture techniques may be required to maximize yield. If NTM species with special culture requirements are suspected, such as growth at lower temperatures for skin related pathogens or nutritional supplementation of media, these should be informed to the laboratory personnel as well. Time to growth and colony counts should be recorded as well.

Some clinical specimens, such as sputum, have a high potential for contamination or bacterial overgrowth. When processing these specimens, concentration and decontamination techniques are applied to improve the microbiologic yield [3], [4]. Decontamination with N-acetyl-l-cysteine-NaOH-oxalic acid (NALC-NaOH-OxA) is commonly utilized, though some reports have expressed concern that it may affect the viability of mycobacteria [7], [8]. Another decontamination method that does not affect mycobacterial viability uses chlorhexidine [9]. Both of these methods are approved by the Clinical Laboratory Standards Institute (CLSI).

There are no established recommendations for staining of NTM; most recommendations were derived from staining methods established for M. tuberculosis. The cell walls of NTM, similar to other mycobacteria, have high lipid content and are designated as acid-fast bacilli (AFB) because of a unique ability to bind a carbol-fuchsin dye that cannot be decolorized by acid-alcohol. Several staining methods have been used for staining AFB (Table 3). It is not possible to distinguish tuberculous and non-tuberculous Mycobacterium species on the basis of staining alone. Sensitivity of staining likely depends on numerous factors, including burden of pathogen and tissue processing methods. Processing of sputum with bleach and concentrating the specimen before staining has been shown to increase sensitivity of detection of AFB in some studies [10], [11]. However, no data is available to accurately quantify minimum NTM pathogen burden to yield a positive AFB sputum smear.

Table 3

Commonly used stains for detection of acid fast bacilli.

Stain	Specimen type	Appearance	Other Findings	Mycobacteria
Ziehl-Neelson [121]	Any clinical specimen	Bright red with blue background		All
Modified Kinyoun [121]	Cytospin or solid culture media	Bright red	Weakly acid fast organisms do not appear bright red	M. fortuitumM. chelonaeM. abscessus
			All organisms appear identical with variable staining	MAC
			Filament-like or with slight branching	Dry colonies of MAC
Rapid Modified Auramine O Fluorescent [122]	Any clinical specimen	Yellow to orange or bright green with dark background	Some rapid growing mycobacteria may not fluoresce well	All
Fite [123]	Preferred for tissue specimens	Pink with blue background		Preferred stain for M. leprae
Gomori-methenamine Silver Stain [124]	AFB in tissue specimens	Brown to black		M. kansasiiMAC
Periodic Acid Schiff Stain [124]	AFB in tissue specimens	Pink		M. ulceransM. cheloneiM. kansasiiMAC from skin

*Reduced with formalin.

Abbreviations: AFB – Acid Fast Bacilli, MAC – Mycobacterium Avium Complex.

All clinical specimens should be cultured and then further testing modalities applied to identify species. Establishing the diagnosis of NTM infection to the species level needs to be pursued as treatment varies depending on species. Mycobacteria are fastidious organisms, often requiring special growth media and prolonged incubation in a carbon dioxide rich atmosphere. Specimens should be inoculated in liquid (Mycobacteria Growth Indicator Tube (MGIT)) and/or solid media (Middlebrook 7H10 or 7H11 agar or Lowenstein-Jensen agar (L-J)) and incubated for a variable period of time to allow isolation of these organisms. Several organisms need special culture conditions to improve yield. Time to growth is usually earlier in liquid media than solid media.

NTM are commonly classified into rapid and slow growers based on time to growth of mature colonies on solid culture media [12]. Rapid growing NTM grow in less than seven days; slow growing NTM take more than seven days to grow. The characteristics of rapid (RGM) and slow-growing NTM species are summarized in Table 4.

Table 4

Classification of non-tuberculous mycobacteria based on time to growth.

Classification	Time to Growth	Major Groups	Common Species	Growth Requirements	Common Clinical Syndromes
Rapidly Growing	<7 days	M. fortuitum[125]	M. fortuitumM. peregrinumM. senegalenseM. setenseM. porcinumM. houstonenseM. boenickeiM. brisbanenseM. neworleansense	35°–37 °C	SSTISkeletal infection
		M. chelonae/abscessus[126]	M. chelonaeM. salmoniphilumSubspecies of M. abscessus:M. bolletiM. massilienseM. abscessus		SSTISkeletal infectionDisseminated infectionPulmonary disease
		M. mucogenicum[127]	M. mucogenicumM. aubagnenseM. phocaicum		Bacteremia in ICH
		Late pigmenting RGM	M. smegmatisM. goodii		PneumoniaProsthetic valve endocarditis
		Early pigmenting RGM	M. flavescensM. neoaurumM. vaccaeM. canariasenseM. cosmeticumM. monacenseM. thermoresistibileM. bacteremicumM. iranicum		BSIProsthetic valve endocarditisSSTISkeletal infection
		Non-pigmenting RGM	M. mageritenseM. wolinskyi		SSTICIED Infection
Slow Growing [128]	>7 days	MAC	M. intracellulareM. avium	35°–37 °C	Pulmonary diseaseLymphadenitisSkeletal infection Disseminated infection Hypersensitivity pneumonitis
			M. szulgaiM. malmoenseM. scrofulaceum		SSTIPulmonary diseaseSkeletal infectionDisseminated infectionCervical lymphadenopathy
			M. kansasii		Skeletal infectionsDisseminated infectionPulmonary disease
		M. simiae	M. simiaeM. lentiflavum		Disseminated infection
		M. terrae/nonchromogenicum	M. arupenseM. kumamotonenseM. heraklionenseM. longobardum		Skeletal infections
			M. haemophilum	28°–30 °C requires hemin or iron	SSTILymphadenitis
			M. genavense		Colitis in ICH
			M. xenopi	42°–45 °C	Pulmonary diseaseSkeletal infection
			M. ulcerans		SSTI: “Buruli ulcer”
			M. marinum	28°–30 °C	SSTISkeletal infection“Fish tank Granuloma”
			M. gordonae	35°–37 °C

Abbreviations: BSI – Blood Stream Infection, CIED – Cardiovascular Implantable Electronic Device, ICH – Immunocompromised Host, MAC – Mycobacterium Avium Complex, RGM – Rapid Growing Mycobacteria, SSTI – Skin and Soft Tissue Infection.

Slow growing mycobacteria colonies are typically first seen at 2 weeks and mature colonies are expected at 3 weeks on solid media such as L-J agar [13] and the Middlebrook agars [12]. In contrast, the rapid growing mycobacteria colonies are typically seen by 4 days. Middlebrook 7H10 is a non-selective medium that can be utilized for quantitative reporting and to differentiate chromogenic from nonchromogenic mycobacteria (Table 5) [14]. MGIT is a liquid broth medium that allows automated detection of mycobacterial growth based on emission of fluorescence (MGIT 960 instrument; BD) [15], [16]. The sensitivity and specificity for detection of NTM for this system is 98.8% and 100%, respectively [17].

Table 5

Reporting of non-tuberculous mycobacterial growth on middlebrook 7H10 medium.

Colony Count	Interpretation
None	Negative
<50	Actual count
50–100	+
100–200	++
200–500 (Almost confluent growth)	+++
>500 (Confluent growth)	++++
Appearance of Colonies
White, Cream, Buff	Non-chromogenic
Lemon, Yellow, Orange, Red	Chromogenic

Identification of NTM species is essential in the evaluation of a patient with NTM disease. Species identification allows for assessment of clinical significance, prognosis, and expected antimicrobial resistance. Unlike M. tuberculosis, no widely adopted techniques for species identification directly from clinical specimens exist. Therefore, culture remains the first step in identification. Once an organism has been isolated in culture, a number of diagnostic modalities can be employed to identify the organism to species level. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) is very effective at identifying most NTM. It is usually cheaper and simpler than commercial nucleic acid amplification assays or whole genome sequencing. However, it requires pure isolates and may not be able to differentiate between closely related mycobacteria such as M. chimaera and M. intracellulare. It is also limited by the breadth of the spectra library available to the instrument [18].

Molecular diagnostics has revolutionized the practice of species identification, but can be applied only after identification of growth on culture media [6]. A variety of Food and Drug Administration (FDA)-approved DNA probes for species specific RNA is available. If MALDI-TOF and species-specific probes are unable to identify an organism, sequencing of the 16s RNA may be useful. Two regions of the 16s RNA can be helpful in identifying NTM species. Several commercially available systems that offer out-of-the box sequencing are available. However, several species may not be differentiated as there is not enough genomic divergence to allow for accurate diagnosis beyond species complex level. Similar to MALDI-TOF, test characteristics depend on the library of available sequences [19].

Identifying organisms directly from clinical specimen is an area of ongoing research. Polymerase chain reaction (PCR)-based assays of respiratory samples for rapid identification of NTM species are currently in development. Assays used to diagnose TB such as Xpert MTB/RIF Ultra and TrueNAT (MolBio) may have some cross reactivity with NTM species, but they are not a primary means for diagnosis of NTM disease [20]. Line probe assays do have the ability to identify some species of NTM and test for aminoglycoside and macrolide resistance, but they are not widely available. MALDI-TOF may also be effective at identifying NTM species directly from clinical specimens [18]. However, these techniques are at present experimental, but hold the promise for a more rapid identification of NTM infections in the future.

Histology. In the majority of patients, diagnosis of pulmonary NTM disease can be made without a tissue sample. When other microbiological criteria are met, biopsies are not required [3], [4], [21]. In contrast to this, the diagnosis of extra-pulmonary NTM infections typically requires obtaining a tissue sample. The histologic hallmark of mycobacterial disease is necrotizing or non-necrotizing granulomatous inflammation with the presence of an NTM organism. A single tissue sample with these features is sufficient to establish NTM disease [22]. In immunocompetent patients, NTM histologic findings can mirror that of tuberculosis [23]. Granulomatous inflammation alone is not specific for NTM. Lung biopsy may be culture negative due to the small tissue sample, and an NTM diagnosis can still be made with evidence of granulomatous inflammation and one or more cultures with NTM growth [3], [4], [24].

Immunocompromised patients may not show granulomas or detectable organisms on tissue biopsy. In cases of disseminated NTM infection, the lack of hallmark histologic findings does not exclude the diagnosis [24]. Histologic findings in this patient population may include foamy histioctyes containing mycobacteria, poorly formed granulomas, or no clear inflammatory response [23].

Antimicrobial susceptibility testing

CLSI and ATS/IDSA have published criteria for antimicrobial susceptibility testing (AST) [25]. The gold standard for determining antimicrobial susceptibility is culture and growth via broth microdilution [3], [26]. In general, NTM AST should be performed only for clinically significant isolates. CLSI recommends against susceptibility testing for NTM species that are rarely pathogenic, such as M. gordonae, M. mucogenicum, or M. terrae [25], [26].

Clinicians should be cognizant of the fact that correlation of clinical response to in-vitro susceptibility patterns has not been established for each species and anti-mycobacterial drug. For example, for Mycobacterium avium complex (MAC) isolates, there is no correlation of clinical response with in vitro minimum inhibitory concentrations (MIC) except for the macrolides and amikacin [27]. Thus, AST prior to initiation of treatment for MAC should be limited to clarithromycin and amikacin susceptibility testing. Similarly, failure of treatment for Mycobacterium kansasii is generally associated with rifampin resistance; MIC values for ethambutol (EMB) and isoniazid (INH) do not correlate well with clinical response [25], [26]. For this reason, AST prior to initiation of treatment for M. kansasii -include rifampin susceptibility testing first and expanded to other agents only if rifampin resistance is present. For these two NTM organisms, AST against a wider panel of antimicrobials is recommended only in the presence of macrolide-resistance (for MAC) and rifampin resistance (for M. kansasii).

For rapidly growing mycobacteria, susceptibility testing against a wider panel of antimicrobials upfront is recommended. However, it is important to note that the purpose of this testing is “to guide rather than dictate therapy”, as there is no established MIC cutoff for susceptibility or resistance for most of these agents [25], [26]. This wider panel of antimicrobials includes clarithromycin, amikacin, tobramycin, cefoxitin, ciprofloxacin, clarithromycin, doxycycline (or minocycline), imipenem, linezolid, moxifloxacin, tigecycline, clofazimine, and trimethoprim-sulfamethoxazole. Additionally, all M. abscessus and M. fortuitum isolates should be routinely tested for inducible macrolide resistance due to the presence of erythromycin ribosomal methylase (erm) gene [25], [28]. Testing for inducible macrolide resistance involves incubating the organism with sub-inhibitory concentrations of clarithromycin for 14 days [29].

Radiographic diagnosis of NTM infections

No radiographic pattern is pathognomonic for pulmonary NTM disease. Two radiographic patterns are commonly encountered in pulmonary NTM disease: fibro-cavitary disease and nodular bronchiectasis. A combination of these two radiographic patterns or a ‘mixed’ pattern may also be seen.

Fibro-cavitary NTM pulmonary disease: Fibro-cavitary disease usually occurs in middle age or older men with underlying structural lung disease such as COPD or pneumoconiosis. Fibro-cavitary disease can be identified on chest X-ray and is difficult to distinguish from tuberculosis (TB) and endemic fungal infections with X-ray alone (Fig. 1). This presentation is characterized by upper lobe predominant, thin or thick walled cavities with surrounding pleural thickening [30]. Whereas TB usually has calcified lymphadenopathy and associated parenchymal infiltrates, NTM is not typically associated with these findings [30]. Nonetheless radiographic findings alone should not be used to discern between TB and NTM infections. Pulmonary malignancies (particularly squamous cell carcinoma), fungal infections and pulmonary vasculitides may have similar radiographic findings making clinical, radiographic, microbiologic, and possibly histopathologic correlation necessary.

Fig. 1

Radiographic images of patients with fibro-cavitary MAC. A. Chest roentgenogram demonstrating a left apical thin walled cavity with surrounding pleural reaction. The background lung architecture has increased reticular markings consistent with scarring. B. Chest computed tomography, axial section, demonstrating a thick-walled cavitary lesion in the left upper lobe.

Nodular bronchiectatic NTM pulmonary disease: Nodular bronchiectasis usually manifests in elderly, thin women and is associated with pectus excavatum, scoliosis, and mitral valve prolapse [31]. Nodular bronchiectasis may be visualized as ‘tram tracking’ with micro nodularity on chest roentgenogram but is best observed on high resolution computed tomography of the chest (HRCT). On HRCT, this appears as cylindrical bronchiectasis with branching centrilobular nodules (i.e. “tree-in-bud”) (Fig. 2) [32]. These changes are usually bilateral but preferentially involve the lingula and right middle lobe. Overall these radiographic findings are non-specific and can be seen in aspiration and other bacterial infections. When the cause of tree-in-bud opacities can be determined, less than half are attributed to NTM infection [33].

Fig. 2

High resolution chest computed tomography of nodular bronchiectasis affecting the lingula. Lingula and right middle lobe are thought to be more often involved in nodular bronchiectatic NTM.

Hypersensitivity pneumonitis: Hypersensitivity pneumonitis secondary to inhalation of MAC is another variation of NTM associated lung disease that is worth mentioning. This entity is better known as “hot tub” lung disease [34] and is presumed to be secondary to a hypersensitivity reaction to aerosolized MAC rather than true infection since clinical symptoms and radiographic changes improve with antigen avoidance and/or glucocorticoids [35]. However, NTM infection and hypersensitivity features can coexist [35], [36]. HRCT findings are consistent with acute hypersensitivity pneumonitis: upper lobe predominant, bilateral ground glass opacities with air trapping on expiration and centrilobular nodularity.

Extra-pulmonary NTM disease: Radiographic features of extra-pulmonary NTM infection are non-specific and variable, depending on the site of infection. For skin and soft tissue NTM infections, radiographic imaging studies are not routinely required for diagnosis. However, imaging studies such as sonography, CT, and MR can be useful to localize an abscess or determine if infection extends into deeper soft tissues or adjacent bone, in particular in situations where the response to treatment is sub-optimal. Radiographic features of NTM lymphadenitis include decreased echogenicity on ultrasound, corresponding to intranodal liquefactive/cystic necrosis, as well as an enlarged nodal mass with central necrosis on CT or MRI [37], [38]. Common radiographic features of NTM osteomyelitis include permeative osteolysis, bone sequestration, and periostitis. Articular NTM infections manifest radiographic findings ranging from regional osteopenia to marginal and subchondral osseous erosions [39].

Checking interleukin-2 and interferon gamma pathways

Certain patients are more susceptible to NTM disease than others. This is illustrated in other mycobacterial diseases by the large number of patients with positive tuberculin skin testing who never progress to active disease, and the small number of patients who receive bacille Calmette-Guerin (BCG) vaccination and develop disseminated BCG-osis [42]. Protection from mycobacterial infection is provided by the type I cytokine pathway; investigations have shown that defects in the genes encoding IL-12, IL-23, and interferon- γ (INF-γ) receptors lead to reduced immunity to mycobacteria [43]. Patients with these primary immunodeficiencies are said to have ‘Mendelian susceptibility to mycobacterial disease’.

While not addressed in the 2020 or prior ATS/IDSA guidelines [4], [44], it is reasonable to assess patients for primary defects in cell mediated immunity who present with disseminated NTM infection without known secondary cause, especially if family members are also ill. This information informs prognosis and may affect treatment decisions [45]. These patients are also susceptible to other infections including cytomegalovirus, human herpes virus-8, herpes simplex virus, Listeria, and Histoplasma infections [46].

Distinguishing infection from colonization

Non-tuberculous mycobacteria are ubiquitous in the environment; isolation of NTM species from non-sterile samples does not equate to infection. Most NTM isolated from a respiratory source may not meet currently established criteria for NTM infection [47]. The respiratory tract can be colonized with NTM, especially if underlying structural airway and/or lung disease are present. MAC is the most commonly described NTM airway “colonizer” [48], but it is unclear if this is true colonization or low grade infection. Colonization without infection is an unproven concept in pulmonary NTM disease; however, this does not imply that all patients with NTM respiratory isolates require antimicrobial treatment. The current criteria for NTM pulmonary infections were established by a joint guideline from ATS and IDSA in 2007 but were more recently updated in 2020 [3], [4]. These guidelines by necessity focus on the most commonly encountered and best described NTM species—MAC, M. kansasii, and M. abscessus. Presumably these guidelines apply to other less common and lesser-known NTM species, but data and experience with these NTM species are limited. The diagnosis of pulmonary NTM disease requires clinical, radiographic, and microbiologic correlation. These criteria are of equal importance, and all must be met to diagnose pulmonary NTM disease [3].

Clinical manifestations of NTM infections

Pulmonary disease is the most common manifestation of NTM infections [48], but NTM can also cause localized infection of the skin, bones and soft tissues; disseminated infections in patients with severely compromised immune systems can also occur. The current ATS/IDSA guidelines review these alternate manifestations but do not specifically address the diagnosis and treatment of extra-pulmonary NTM [3], [44]. Common NTM infection syndromes are summarized in Table 1.

The clinical manifestations of NTM infections are variable and non-specific. Most patients with pulmonary NTM disease will have chronic or recurring cough with or without sputum production [49]. This may be accompanied by constitutional symptoms (e.g. fever, fatigue, weight loss, and night sweats), chest pain, or shortness of breath. These latter symptoms may only be present with advanced disease. Because many patients with pulmonary NTM disease have underlying lung disease such as bronchiectasis or chronic obstructive pulmonary disease (COPD), it can be very challenging to determine if NTM infection is the cause of the presenting symptoms.

Patients with skin, soft tissue, and musculoskeletal NTM infections may present with skin nodules, ulcers, wound drainage or abscesses in addition to the constitutional symptoms noted above. Musculoskeletal NTM infections may present with joint pain, stiffness, and swelling as well as muscle wasting and draining sinuses. Mycobacterial lymphadentis presents typically with enlarged lymph nodes, often times non-tender. Disseminated NTM infection may present with scattered skin lesions. In patients with advanced HIV infection, disseminated MAC infection presents as a systemic febrile illness with abdominal pain and diarrhea.

Physical findings reflect the system or organs affected but are variable and nonspecific. For example, features of skin and soft-tissue NTM include or pain, erythema, warmth, local pain, swelling, ulcers, and abscesses. It should be noted that cosmetic surgeries are common precipitant for NTM skin and soft tissue infections. A firm, painless mass with overlying skin changes is a common physical examination finding in NTM lymphadenitis. Hepatosplenomegaly is a common physical examination finding in disseminated MAC disease.

Antimicrobial treatment of non-tuberculous mycobacterial infections

As described in the preceding sections, the treatment of NTM infections is complex and limited by several issues, including the need for multidrug regimens, prolonged treatment durations, and the use of drugs that are often associated with adverse reactions and not well-tolerated. Co-morbidities that are often present in patients with NTM infections add another layer of complexity to the management. The optimal drug doses, drug combinations, and treatment durations have not been reliably established for most, if not all, NTM infections. It should be noted that several new medications used for the treatment of multi-drug resistant TB such as, bedaquiline, pretomanid and delamanid have in-vitro activity against NTM species but in vivo data is lacking at this time [50], [51]. Consensus guidelines for the management of NTM infections exist. However, they are based on limited data and expert opinion rather than evidence from adequately powered controlled trials with sufficiently long follow-up period. This notwithstanding, the following general considerations can be noted.

In general, anti-mycobacterial treatment should be offered only when the clinical, radiographic, and microbiologic criteria for the diagnosis of NTM infections are met.

The optimal management of NTM infections requires a well-coordinated multidisciplinary approach, involving infectious diseases physicians, pulmonologists, pharmacists, microbiologists, radiologist, nurses, respiratory therapists, and dietitians.

The antimicrobial drugs currently used for treating NTM infections were not specifically developed for that purpose. Their utility in treating NTM infections was derived from demonstration of activity in-vitro or in animal models, drug susceptibility testing, and observations from clinical practice. Antimicrobials that have demonstrated activity against NTM species and/or infections in vitro are highlighted in Table 6.

Table 6

Drugs with microbiologic activity for NTM demonstrated in human clinical disease or in vitro.

Isolate (s)	Drugs
Slow Growing NTM
MAC	Human: AMK [129] AZ [130] CFZ [131] CLR [132] BDQ [133] EMB [130] RIF [130] RFB [130] SM [134] LZD [135] MOX [136]Micro: CTZ/AV [137] ETO [138] TIG [139] TZD [140]
M. kansasii	Human: CLR [141] EMB [142] INH [142] LFX [143] RIF [142] RFB [144] SM [142]Micro: AMK [145] AZ [146] BDQ [146] CFZ [146] CIP [145] DLM [146] LFX [146] LZD [146] MOX [146] SMX [145] TIG [146] TZD [140]
M. marinum	Human: AMK [147] CLR [148] DOX [148] EMB [149] MIN [148] RIF [149] RFB [150] SMX/TMP [151]Micro: CPX [152] IMI [152] INH [152] LVX [152] LZD [153] MOX [152] TZD [140]
M. leprae	Human: CFZ [154] CLR [155] DAP [156] MIN [154] MOX [157] OFX [154] RIF [158]Micro: LFX [159]
Rapid Growing NTM
MABC	Human: AMK [160] AZ [160] BDQ [133] CFX [160] CFZ [161] CLR [160] CPX [162] DOX [163] IMI [161] LZD [120] MOX [162] TIG [164]Micro: ERA [165] OMA [165] LZD [166] TZD [140]
M. chelonae	Human: AMK [167] AZ [168] CFZ [169] CLR [170] CPX [171] DOX [172] IMI [173] LFX [174] LZD [175] SMX [176] TIG [164] TOB [177]Micro: MIN [125] OMA [178] TDZ [140]
MFCa	Human: AZ [179] AMK [180] CLR [181] DOX [182] IMI [173] LFX [183] MIN [184] OFX [180] SMX/TMP [185]Micro: CFX [186] MOX [187] LZD [166] OMA [178] TIG [188] TOB [189] TZD [140]
MSGa	Human: AMK [190] CFX [190] DOX [190] IMI [190] SMX/TMP [190] TOBb[190]Micro: CLR [191] CPX [191] EMB [191] CFZ [192] TZD [140]
M. immunogenum	Human: AMK [193] CLR [194]Micro: AMK [195] CLR [195] LZD [140] TIG [188] TZD [140]
M. mucogenicum	Human: AMK [196] AZ [196] CIP [196] CLR [196] LFX [197] MFX [198]Micro: AMX [127] AMX/CLV [197] CFX [199] CPX [199] DOX [199] IMI [199] LZD [198] MIN [199] SMX/TMP [199]

Abbreviations: AMK – amikacin, AMX/CLV – amoxicillin/clavulanate, AZ – azithromcyin, BDQ – bedaquiline, CFX – cefoxitin, CFZ – clofazimine, CLR – clarithromycin, CPX – ciprofloxacin, CTZ/AV – ceftazidime/avibactam, DAP – dapsone, DOX – doxycycline, DLM – delaminid, EMB – ethambutol, ERA-eravacycline, ETO – ethionamide, IMP – imipenem, INH – isoniazid, LFX – levofloxacin, LZD – linezolid, MIN – minocycline, MFX – moxifloxacin, OFX – oflaxacin, OMA-omadacycline, RFB – rifabutin, RIF – rifampin, SM – streptomycin, SMX – sulfamethoxazole, TIG – tigecycline, TOB – tobramycin, TMP – trimethoprim, TZD – tedizolid.

MABC – Mycobacterium Abscessus Complex (M. abscessus, M. massiliense, M. bolettii), MAC – Mycobacterium Avium Complex (M. avium, M. intrellulare, M. chimaera), MFC – Mycobacterium Fortuitum Complex (M. alvei, M. boenickei, M. conceptionense, M. farcinogenes, M. fortuitum, M. mageritense, M. neworleans, M. peregrinum, M. porcinum group (includes M. porcinum, M. bonickei, M. houstonense, M. neworleansense), M. senegalense, M. septicum)), MSG – Mycobacterium Smegmatis Group (M. smegmatis, M. wolinski, M. goodii).

aM. fortuitum, M. abscessus subspecies abscessus, and M. abscessus subspecies bolletii, and M. smegmatis have the erm gene for inducible macrolide resistance which should be tested for if considering macrolide [200].

bM. wolinksy isolates intrinsically resistant to tobramycin [191].

Treatment of NTM infections, even treatment of limited skin or soft tissue infection, typically requires the use of at least 2 drugs in combination. The decision of which drugs and how many drugs to use, while guided by susceptibility testing, is often based on clinical experience and expert opinion rather than evidence from controlled trials. The typical dosing for anti-mycobacterial drugs is reviewed in Table 7. The number of antimicrobials needed to treat an NTM infection is dependent on the severity of illness. Using macrolide-susceptible MAC pulmonary disease as an example, initial therapy for nodular bronchiectatic disease is a three-drug therapy including a macrolide, ethambutol, and rifamycin. However, severe bronchiectactic or fibro-cavitary disease would generally warrant the addition of an injectable aminoglycoside as a fourth agent for the first two to three months of therapy. Due to difficulties with adverse effects in certain target populations (i.e. elderly, low body weight pulmonary NTM patients), the drugs are often initiated and titrated to full doses in a staggered format to enhance tolerability and compliance to the drug therapy regimen.

Table 7

Typical drug dosing for NTM disease.

Drug	Standard dose	Renal adjustment
Amikacin	15 mg/kg1 IV daily	Est crcl < 30 ml/min:12–15 mg/kg1 three times weekly
Azithromycin	250 mg once daily	No adjustment
Bedaquiline Fumarate	Week 1–2: 400 mg once daily with foodWeek 3–24: 200 mg three times weekly with food	No adjustment
Cefoxitin	200 mg/kg/day IV in 3 divided doses2 (maximum 12 g/24 h)	Est crcl 30–49 ml/min:1–2 g q 8–12 hEst crcl 10–29 ml/min:1–2 g q12-24 hEst crcl < 10 ml/min:1–2 g q24-48 h
Ciprofloxacin	500–750 mg twice daily	Est crcl 10–49 ml/min:250–500 mg twice dailyEst crcl < 10 ml/min:250–500 mg daily
Clarithromycin	500 mg twice daily or 1000 mg once daily (extended release)	Est crcl < 30 ml/min:1) 500 mg once daily-OR-2) 500 mg once, then 250 mg twice daily
Clofazimine	100 mg daily	No adjustment
Delaminid	100 mg twice daily	Est crcl < 30 ml/min:Not recommended
Doxycycline	100 mg twice daily	No adjustment
Ethambutol	15–20 mg/kg (maximum 1600 mg) daily	Est crcl < 30 ml/min:20–25 mg/kg three times weekly
Ethionamide3	15–20 mg/kg/day (max 1000 mg) daily in divided doses	No adjustment
I/C	Imipenem 1000 mg/Cilastin 1000 mg given intravenously twice daily	Est crcl 20–40 ml/min: 750 mg −750 mg (I/C) every 12 h.Est crcl < 20 ml/min: 500 mg −500 mg (I/C) every 12 h)
Isoniazid3	5 mg/kg (max 300 mg) daily	No dose adjustment
Levofloxacin	750 mg daily	Est crcl 20–49 ml/min: 500–750 mg every 48 hEst crcl < 20 ml/min: 250–500 mg every 48 h
Linezolid3	600 mg daily	No adjustment4
Minocycline	200 mg once, then 100 mg twice daily	Est crcl < 10 ml/min:Standard dose-OR-200 mg once, then 100 mg once daily
Moxifloxacin	400 mg daily	No dose adjustment
Rifabutin	5 mg/kg (max 300 mg) daily	Est crcl < 30 ml/min:Consider standard dose4
Rifampin	10 mg/kg (max 600 mg) daily	No dose adjustment
Streptomycin	15 mg/kg1 daily	Est crcl < 30 ml/min:12–15 mg/kg1 3 times weekly
SMX/TMP	SMX/TMP 800 mg/160 mg twice daily	Est crcl 15–29 ml/min:SMX/TMP 800 mg/160 mg once daily
Tedizolid	200 mg once daily	No adjustment
Tigecycline	100 mg IV once, then 50 mg twice daily2	No adjustment

Abbreviations: Est crcl – Estimate creatinine clearance, I/C – Imipenem/ Cilastin, IV – intravenous, kg – Kilogram, mg – Milligram, SMX/TMP – Sulfamethoxazole/Trimethoprim.

1Use an adjusted body weight (IBW + 0.4(ABW − IBW) for empiric dosing in the setting of obesity and follow serum levels.

2Some providers use reduced dose or schedule to optimize long-term tolerability.

3Consider Vitamin B6 25–50 mg daily to decrease risk of toxicity.

4Renally-cleared metabolite accumulation can occur. Monitor closely for toxicity.

aRifabutin: May consider decreasing to 50% of standard dose and monitoring serum drug levels to avoid toxicity. Standard dosing appropriate for patients on dialysis.

Treatment of NTM is often provided in 2 phases: An initial phase with more drugs (often including intravenous antimicrobials), followed by a maintenance phase with fewer drugs, and when possible, an all oral drug regimen. The duration of the initial phase of therapy is variable, often in the range of 1–3 months. This recommended duration is not evidence based; in clinical practice, it is often shortened because of intolerance or treatment-emergent drug adverse events.

Empiric therapy: Species level identification is necessary to determine the most appropriate treatment regimen. For most NTM infections, susceptibility-based treatment is preferred over empiric therapy. Notwithstanding the poor in vitro susceptibility correlation with clinical outcomes, baseline susceptibility testing, to guide but not to dictate treatment, is recommended for NTM isolates from patients believed to have definite disease. However in disseminated and or severe infections, waiting for the full microbiologic results prior to initiation of anti-mycobacterial therapy may endanger patients’ lives. In such cases, empiric therapy based on the prevailing understanding of the usual antimicrobial susceptibility profile of the putative causative organism is recommended. For example, for infections by slow-growing NTM, a regimen consisting of an oral macrolide, rifamycin, ethambutol, and amikacin may be considered as empiric therapy; whereas, a regimen comprising of amikacin, imipenem, and a macrolide may serve as appropriate empiric therapy for infections presumed to be caused by rapid-growing NTMs. Empiric therapy should be tailored once the pathogen has been fully identified and AST results are available.

For some patient populations, initiating a full dose multi-drug regimen all at once may be intolerable. In such cases, clinicians may wish to consider initiating therapy sequentially, adding a drug every few days until the desired multidrug regimen is achieved. Sequential drug addition with dose titration during therapy initiation (Table 8) often facilitates tolerance of a difficult drug regimen. This approach affords minimal risk of drug resistance due to the slow replication of NTM compared to other bacterial species. This approach can be used for both pulmonary and non-pulmonary NTM infections. Moreover given the slow replication of NTM species in general and prolonged duration of therapy, this approach can also be used to improve tolerance even in severe infections.

Table 8

Example of MAC medication titration schedule.

DAY 1	DAY 2	DAY 3
Ethambutol 400 mg (1 tablet) once daily	Ethambutol 800 mg (2 tablets) once daily	Ethambutol 1200 mg (3 tablets) once daily
DAY 4	DAY 5	DAY 6
Ethambutol 1200 mg (3 tablets) once dailyRifampin 300 mg (1 capsule) once daily	Ethambutol 1200 mg (3 tablets) once dailyRifampin 300 mg (1 capsule) once daily	Ethambutol 1200 mg (3 tablets) once dailyRifampin 600 mg (2 capsules) once daily
DAY 7	DAY 8	DAY 9 and ongoing
Ethambutol 1200 mg (3 tablets) once dailyRifampin 600 mg (2 capsules) once dailyAzithromycin 125 mg (0.5 tablet) once daily	Ethambutol 1200 mg (3 tablets) once dailyRifampin 600 mg (2 capsules) once dailyAzithromycin 125 mg (0.5 tablet) once daily	Ethambutol 1200 mg (3 tablets) once dailyRifampin 600 mg (2 capsules) once dailyAzithromycin 250 mg (1 tablet) once daily

The optimal duration of treatment for individual NTM species and infections is not known. In general, treatment is prolonged, ranging from 3 to 4 months for limited skin and soft tissue infections to 12 months and longer for severe lung and disseminated infections. Factors that guide the decision to discontinue antimicrobials include clinical, microbiologic, and radiologic response to treatment, as well as the tolerability of the treatment regimen. For example, drug therapy for pulmonary NTM disease should generally continue for 12 months after confirmed culture conversion.

Chronic suppressive therapy: For some NTM infections, no antimicrobial regimen has been shown to effect long-term microbiologic or clinical cure. In situations where curative treatment is not possible, the goal of therapy shifts from definitive eradication of infection to minimization of symptoms. For example, if source control is not possible (e.g. infected retained prosthetic material, severe structural lung abnormalities) then chronic suppressive therapy may be warranted. Non-pharmacologic adjunctive interventions (discussed below) such as aggressive pulmonary hygiene should be optimized in these patients.

Monitoring for clinical response to treatment

Clinical, radiographic, and microbiologic assessments are typically used to determine if a patient is responding to therapy.

Clinical symptoms should be assessed at each clinical encounter. Improvement in symptoms can be a key indicator of response to treatment, but clinical symptoms do not always correlate with changes in radiographic or microbiologic status. Furthermore, complete resolution of symptoms at the end of treatment is not universal given the multifaceted nature of NTM disease and comorbid illness that can contribute to symptomatology. Oftentimes clinical monitoring is dictated by the need to assess for the presence of drug toxicities in order to determine if drug and dose changes are necessary.

For pulmonary NTM infections, serial sputum cultures serve as a marker for treatment response. Current guidelines recommend monitoring sputum cultures every 1–2 months; confirmed sputum culture conversion (subsequent negative cultures one month apart) provides a basis for microbiologic cure and is a key factor in determining treatment duration. In situations where spontaneous expectorated sputum specimens cannot be collected, induced sputum should be obtained; obtaining deep respiratory samples via bronchoscopy for the purpose of monitoring response to therapy is not usually required.

For extra-pulmonary NTM infections, treatment response monitoring based on serial collection of microbiologic data is less well proscribed and is usually performed at the discretion of clinicians based on clinical progress.

Radiographic imaging studies are also surrogates for treatment response. Imaging is often performed during therapy and at the end of a planned treatment period to confirm treatment response. The optimal intervals for imaging performance are not well established.

Safety monitoring during anti-mycobacterial therapy

Adverse drug reactions (ADRs) are very common during the treatment of NTM infections and are associated with treatment interruption or early treatment cessation [27], [52], [53]. Not surprisingly, some evidence exists that ADRs also impact treatment outcome [54]. Therefore, educating patients about the likely ADRs that they may encounter with the drug-regimen that they have been prescribed, and monitoring them for treatment-emergent ADRs is a key component of the management of the treatment of NTM infections. This monitoring is conducted in a variety of ways; there is no evidence basis for the optimal frequency of monitoring of patients with NTM for adverse drug reactions. At a minimum, active assessment for the development of ADRs, through questions or physical examination, should be part of each clinical encounter. A high potential for drug interactions exists with many of the antimicrobials used to treat NTM infections; therefore, a close review of the medication list is warranted at every clinical encounter.

Safety monitoring for antimicrobial toxicity will vary based on multiple factors: co-morbid medical conditions, baseline functional status, liver function, kidney function, drug-drug interactions, concomitant medications with additive toxicity profile, intravenous versus oral therapy, drug dose, and the availability of laboratory resources. Nonetheless, it is important for clinicians to anticipate potential adverse reactions. Table 9 highlights the more notable adverse effects, while providing guidance on clinical and laboratory monitoring for extended therapy. These recommendations are meant to provide general guidance. More detailed monitoring may be warranted based on patient age, comorbidities, concurrent drugs, overlapping drug toxicities, and availability of healthcare resources.

Table 9

Adverse drug reactions of NTM therapy and monitoring guidance [201], [202].

Drug	Adverse drug reactions	Laboratory/Clinical monitoring
Amikacin (parenteral)	Kidney toxicity, electrolyte abnormalities (K, Mg, Ca), vestibular, toxicity, hearing loss	Clinical: consider audiometry, balance assessmentsLaboratory: BMP w/SCr, BUN, K (2x/weekly), TDM for durations >1 week
Amikacin (inhalation)	Sore mouth/throat, bronchospasm, cough, wheezing/SOB, dysphoniaRare: kidney/vestibular/auditory toxicity	Clinical: pre-/post administration respiratory statusLaboratory: SCr (monthly), serum trough level (to confirm lack of systemic accumulation) after 1 week of therapy
Amoxicillin- Clavulanate	Abdominal discomfort, diarrhea, nausea, vomitingRare: hypersensitvity, rash	Clinical: GI effectsLaboratory: CBC w/differential, ALT and SCr (1 month, 3 months, then annually)
Azithromycin	GI (abdominal pain, nausea, vomiting, diarrhea, dyspepsia), headache, visual disturbanceRare: hearing loss, QT prolongation	Clinical: consider audiometry, GI effects, consider ECGcLaboratory: CBC w/differential , ALT, SCr (1 month, 3 months, then annually)
Bedaquiline Fumarate	Nausea, arthralgia, headache, chest pain, QT prolongation, hepatotoxicity	Clinical: ECG (baseline, 2 weeks, 12 weeks, and 24 weeks)Laboratory: serum Ca, Mag, and K (baseline); liver function tests: AST, ALT, bilirubins, alk phos (baseline, then monthly)
Cefoxitin	Injection site reactions, hypotension, transaminitis, leukopenia, thrombocytopeniaRare: aplastic anemia	Clinical: rash and hypersensitivity symptomsLaboratory: CBC w/differential, ALT, and BMP w/SCr, BUN, K (weekly)
Ciprofloxacin	GI (abdominal pain, nausea, vomiting, diarrhea, dyspepsia), hypoglycemia, CNS effects (agitation, nervousness, disorientation, memory impairment), rashRare: peripheral neuropathy, tendon rupture, QT prolongation	Clinical: consider ECG, signs of hypoglycemia/ CNS effectsLaboratory: CBC w/differential, ALT, and SCr (1 month, 3 months, then annually)
Clarithromycin	GI (taste disturbance, abdominal pain, nausea, vomiting, diarrhea, dyspepsia)Rare: hearing loss, QT prolongation, hepatotoxicity	Clinical: consider audiometry, GI effects, consider ECGc, Monitor DDILaboratory: CBC w/differential , ALT, SCr (1 month, 3 months, then annually)
Clofazimine	Orange/red discoloration of body fluids/feces, skin manifestations (hyperpigmentation ichythosis, xeroderma ,pruritis, phototoxicity, rash), GI disturbance (nausea, vomiting, diarrhea, anorexia, abdominal pain)Rare: QT prolongation, hepatitis, depression due to skin discoloration	Clinical: ECG (baseline, 2 weeks, 12 weeks, and 24 weeks), GI adverse effects, psychological disturbances (depression/anxiety) related to skin effectsLaboratory: liver function tests: AST, ALT, bilirubins, alk phos (baseline, then monthly)
Doxycycline	GI effects (nausea, vomitting, diarrea, dysphagia), photosensitivity, rashRare: esophageal ulceration, benign intracranial hypertension, hepatitis	Clinical: photosensitivity, esophageal discomfortLaboratory: CBC w/differential, ALT, and SCr (1 month, 3 months, then annually)
Ethambutol	optic neuritis, GI effects (nausea/vomitting)Rare: peripheral neuropathy	Clinical: monthly visual acuity assessment and red-green color-discrimination assessment (Ishihara), fundoscopic exam (baseline and every 3 months)Laboratory: Scr (baseline and monthly)
Ethionamide	GI (nausea, vomiting, anorexia, taste disturbance, diarrhea), endocrine (gynecomastia, hair loss, acne, impotence, menstrual irregularity, reversible hypothyroidism), hepatotoxicity, neurotoxicity (peripheral neuropathy, optic neuritis, depression, and psychosis)	Clinical: GI, endocrine, and psychiatric symptomsLaboratory: creatinine and LFTs: AST/ALT, alk phos, bilirubins (baseline and monthly), TSH (baseline and every 3 months)
Imipenem/Cilastin	GI effects (nausea, vomiting, diarrea), rashRare: seizure, cytopenias	Clinical: GI effects and rashLaboratory: CBC w/differential, ALT, and BMP w/SCr, BUN, K (weekly)
Isoniazid	Peripheral neuropathy, asymptomatic transaminase elevations, hepatitisRare: rash, hypersensitivity reactions	Clinical: clinical signs of hepatotoxicity (nausea, abdominal pain, jaundice) and neuropathyLaboratory: LFTs: AST/ALT, alk phos, bilirubins (baseline and monthly) if underlying liver compromise or concomitant hepatotoxic drugs
Levofloxacin	GI (nausea, diarrhea) hypoglycemia, CNS effects (dizziness, agitation, nervousness, disorientation, memory impairment, insomnia)Rare: peripheral neuropathy, tendon rupture, QT prolongation	Clinical: Consider ECG, signs of hypoglycemia/CNS effectsLaboratory: CBC w/differential, ALT, and SCr (1 month, 3 months, then annually)
Linezolid	GI effects (nausea, vomiting, diarrhea), headache, hematologic myelosuppression (thrombocytopenia, leukopenia, anemia)Rare: lactic acidosis, peripheral neuropathy, seizure, optic neuropathy	Clinical: GI effects, clinical thrombocytopenia (unusual bleeding, bruising) and peripheral neuropathyLaboratory: CBC w/differential (weekly × 4 weeks, then bi-weekly), LFTs: AST/ALT, alk phos, bilirubins (baseline and monthly)
Minocycline	GI effects (nausea, vomiting, diarrea), skin effects (rash, photosensitivity, skin pigmentation), dizziness, headache	Clinical: cutaneous, neurologic, and GI effectsLaboratory: CBC w/differential, ALT, and SCr (1 month, 3 months, then annually)
Moxifloxacin	GI (nausea, diarrhea) hypoglycemia, CNS effects (dizziness, agitation, nervousness, disorientation, memory impairment, insomnia)Rare: peripheral neuropathy, tendon rupture, QT prolongation	Clinical: consider ECG, signs of hypoglycemia/CNS effectsLaboratory: CBC w/differential, ALT, and SCr (1 month, 3 months, then annually)
Rifabutin	Discoloration (orange/red) of body fluids, cutaneous effects (pruritus, rash), GI effects (nausea, vomitting), hepatitis, anterior uveítis, hematologic toxicity (leukopenia, neutropenia), arthralgias/myalgias	Clinical: Monitor DDI, GI/ cutaneous effectsLaboratory: CBC w/differential, ALT, and SCr (1 month then every 3 months)
Rifampin	Discoloration (orange/red) of body fluids, cutaneous effects (pruritus, rash), GI effects (nausea, vomitting), hepatitis, hematologic toxicity (leukopenia, hemolytic anemia)Rare: flu-like syndrome	Clinical: Monitor DDI, GI/ cutaneous effectsLaboratory: CBC w/differential, ALT, and SCr (1 month then every 3 months)
Streptomycin	Kidney toxicity, electrolyte abnormalities (K, Mg, Ca), vestibular, toxicity, hearing loss	Clinical: consider audiometry, balance assessmentsLaboratory: BMP w/SCr, BUN, K (2x/weekly), TDMb for durations >1 week
Tigecycline	GI effects (nausea, vomiting, diarrea, anorexia, dyspepsia), dizziness, rash	Clinical: nausea/vomitting, dizzinessLaboratory: BMP w/SCr, BUN, K; LFTs: AST/ALT, alk phos, bilirubins; CBC w/differential (weekly)
Tobramycin	Kidney toxicity, electrolyte abnormalities (K, Mg, Ca), vestibular, toxicity, hearing loss	Clinical: consider audiometry, balance assessmentsLaboratory: BMP w/SCr, BUN, K (2x/weekly), TDM for durations >1 week
Trimethoprim/Sulfamethoxazole	GI effects (nausea, vomiting, diarrea), cutaneous effects (pruritus, rash),	Clinical: nausea/vomiting and rashLaboratory: CBC w/differential, ALT, and SCr (1 month, 3 months, then annually)

Abbreviations: Alk Phos – alkaline phosphatase, ALT – alanine aminotransferase, AST – aspartate aminotransferase, BMP – basic metabolic panel, BUN – blood urea nitrogen, Ca – calcium, CBC – complete blood count, CNS – central nervous system, DDI – drug drug interaction, ECG – electrocardiogram, GI – gastrointestinal, K – potassium, Mag – magnesium, SCr – serum creatinine, SOB – shortness of breath, TDM – therapeutic drug monitoring, TSH – thyroid stimulating hormone.

Therapeutic drug monitoring (TDM) is the clinical practice of measuring drug concentrations in serum during a treatment course to ensure that target drug concentrations are achieved. The routine use of TDM other than to manage aminoglycoside therapy for NTM infections is not common. However, TDM of other drugs should be considered in situations where there are concerns for inappropriate drug exposure. Low therapeutic exposure could lead to treatment failure and/or resistance development while high exposure can lead to exposure dependent toxicity and intolerability of the drug program. Common reasons to consider TDM would include: sub-optimal treatment response, drug-drug interactions, exposure dependent toxicity development, renal/hepatic compromise, or malabsorption [4], [27].

We recommend 2 h and 6-h serum drug levels following the end of infusion for amikacin, streptomycin, and tobramycin to calculate a back-extrapolated Cmax. We recommend the following targets: goal amikacin or streptomycin Cmax of 35–45 mcg/ml (for 15 mg/kg once daily dosing) [27]. A Cmax of 65–80 mcg/ml may also be considered for a 25 mg/kg dose given three times weekly [55]. We recommend a goal tobramycin (for M. chelonae) Cmax of 20–30 mcg/ml on 7 mg/kg daily dosing. If serial levels are not feasible, consideration may be given to a single serum level taken 1-h following the end of infusion for amikacin or streptomycin targeting a serum range of 25–35 mcg/ml. For tobramycin, a single serum level taken 6–14 h following the start of infusion can be used, applying the Hartford nomogram [56]. After attainment of appropriate goal targets with stable renal function, we recommend weekly trough monitoring with a goal of ‘undetectable’ to ensure no drug accumulation.

Adjunctive therapies for NTM infections

Reduction of immunosuppression. Immunosuppressive therapies, such as oral and inhaled corticosteroids and TNF-alpha inhibitors, increase the risk for NTM disease. Prednisone equivalent doses of greater than 15 mg per day or inhaled fluticasone doses more than 800 mg per day are associated with a higher risk of NTM disease [57]. For patients taking TNF-alpha inhibitors, a United States’ study showed a 5–10 fold higher rate of NTM pulmonary disease as compared to the general population and rheumatoid arthritis patients on other therapies [58]. Infliximab showed the highest rates of NTM lung disease. However, Etanercept therapy was associated with more serious infections, including fatal MAC lung disease, fatal pulmonary M. abscessus infection, M. chelonae endophthalmitis, and M. xenopi spinal osteomyelitis [59].

Unlike with M. tuberculosis, routine screening for NTM disease prior to the initiation of TNF-alpha inhibitors is not recommended. However, certain patients, such as those with refractory, chronic cough may benefit from CT imaging and sputum testing for NTM disease prior to the start of immunosuppressive therapy [57]. When NTM disease is diagnosed, decreasing or discontinuing immunosuppressive medications can beneficial. For patients that need to continue on immunosuppressive medications, non-biologic agents such as methotrexate and low risks therapies such as abatacept should be considered [57].

Surgical resection. For extra-pulmonary NTM infections, such as skin/soft tissue, musculoskeletal, and lymphadenitis, surgical interventions such as debridement, excision, and drainage of abscesses are often a necessary and important component of the management of these infections [60], [61], [62], [63], [64], [65].

The situation is vastly different when it comes to pulmonary NTM infections. Here, surgical resection plays a role only in select situations and in facilities where a surgical and multidisciplinary team experienced in managing such interventions is available. Typical indications for surgical therapy include failure of medical therapy and alleviation of life-threatening symptoms such as hemoptysis. In areas of extensive lung damage, antimicrobial penetration is low leading to treatment failure [3], [4]. In a trial of 371 patients with mostly cavitary NTM disease, only 36% of patients achieved a cure after 24 months of multi-drug therapy [66]. In such cases, de-bulking surgery may slow disease progression as an adjunct to medical therapy. Drug therapy and medical interventions should precede surgery and continue afterward. Nutrition status should also be optimized prior to surgery. Studies reporting outcomes of surgical therapy in pulmonary NTM infections are mostly observational, heterogeneous in their patient and mycobacterial characteristics, and suffer from inherent selection bias. That notwithstanding, most reports indicate that adjunct surgical therapy can lead to improved patient outcomes with a low rate of complications [67], [68], [69], [70], [71].

Interferon-γ. INF-γ, a cytokine produced by CD4 lymphocytes, has been administered exogenously for the treatment of a variety of syndromes ranging from chronic granulomatous disease to hematologic malignancies [72], [73]. INF-γ plays a critical role in host defenses against mycobacteria through the activation of macrophages [74], [75]. The identification of low INF-γ production in certain individuals susceptible to NTM disease led to exploration of exogenous administration of various formulations of INF-γ for NTM treatment [76], [77], [78]. A randomized, double blind, placebo-controlled study performed in Cuba assessed 32 patients assigned to either placebo or recombinant, intramuscular INF-γ once daily for a month then three times per week up to 6 months as adjuvant to traditional antimicrobials. At the final evaluation, 72% of patients in the treatment group and 36% in the placebo group were deemed complete responders [78]. In contrast, a brief report from 2001 describes no sustained benefit of subcutaneous INF-γ therapy and an inability of INF-γ to improve the production of Th1 cytokines [79]. Flu-like symptoms including fevers, chills, and malaise are common with INF-γ [80]. Treatment with INF-γ may be a consideration in the treatment of select non-HIV patients with progressive, refractory NTM disease associated with a demonstrated INF-γ deficiency.

Rituximab. A specific syndrome of lymphadenitis associated with disseminated, rapid-growing NTM organisms has been identified in persons of Asian descent without HIV infection [81], [82]. Co-infection with endemic fungi, herpes viruses, salmonella as well as reactive skin diseases is common [81], [82]. High titers of antibodies neutralizing interferon-γ are seen resulting in increased susceptibility to these infections; thus, immune modulating therapies have been employed as adjunctive therapy to augment the body’s natural defenses [83], [84], [85], [86], [87], [88], [89].

Rituximab is an anti-CD20 monoclonal antibody that rapidly depletes circulating and tissue B-cells which has been used in the treatment of several autoantibody diseases including pemphigus vulgaris, pulmonary alveolar proteinosis, myasthenia gravis and red cell aplasia [90], [91], [92], [93], [94], [95]. Rituximab was employed to halt progressive NTM disease by neutralizing anti-interferon-γ antibodies in a small number of cases in Asian patients without HIV infection. In both reports, patient’s anti-interferon-γ levels were shown to be elevated at baseline. Subsequent to rituximab treatment, autoantibody titer was reduced and IFN-γ signaling was improved. In conjunction with laboratory marker improvement, all patients showed clinical improvement and none experienced significant adverse effect from rituximab [96], [97]. Rituximab dosing was based on lymphoma regimens at 375 mg/m2 given every 7 for at least 4 doses [90]. In one of the case reports, methylprednisolone was also given at 100 mg daily [96]. There is a small body of evidence that rituximab may serve as an important adjunctive to anti-mycobacterials for immune restoration purposes in patients with progressive, severe anti-interferon-γ associated NTM disease.

Non-pharmacological interventions. Beyond antimicrobial therapy, non-pharmacological interventions can contribute to the alleviation of symptoms and improvement in the quality of life of patients with NTM infections, and as such should be a component of the overall management of NTM infections [98]. These interventions include tobacco cessation (where applicable), pulmonary rehabilitation/pulmonary hygiene, exercise programs, and nutritional support.

Pulmonary hygiene. Many patients with NTM pulmonary disease also have bronchiectasis and signs of airway mucous plugging. In patients with concomitant bronchiectasis, bronchial hygiene regimens with active airway clearance interventions can be used as adjunctive therapy to theoretically facilitate clearance of NTM and prevent superimposed bacterial infections [4], [98]. Retrospective data suggests that pulmonary hygiene alone with active surveillance may be appropriate in patients with mild nodular bronchiectatic MAC pulmonary disease. Active airway clearance interventions have not been prospectively studied in NTM patients with bronchiectasis, but potential benefits are suspected based on expert opinion [99]. These regimens can include mucous clearance devices, such as flutter valves and chest vest, and inhaled medications such as, hypertonic saline nebulizations. An individualized approach and careful introduction of these treatments by a respiratory therapist or experienced nurse are recommended since these can cause bronchospasm and/or airway irritation. Short acting bronchodilators should precede administration of nebulized hypertonic saline in patients with airway hyper-reactivity.

Exercise training. The evidence supporting the benefit of exercise in the management of NTM infections is largely an extrapolation of studies conducted in patients with bronchiectasis, a common comorbidity in patients with NTM lung disease. These studies were of variable duration (3–8 weeks) and of variable design (inpatient or outpatient; with or without supervision; with or without respiratory physiotherapy). Overall, these programs resulted in improvements in exercise capacity and health-related quality of life, as well as decreases in the frequency of acute exacerbations [100], [101], [102], [103].

Nutrition. The linkage between infection and nutrition is not new. Poor nutrition can predispose to infection, infection in turn can contribute to malnutrition [104]. Weight loss is a common symptom in patients with NTM infections and may be further exacerbated by decreased appetite, nausea, early satiety, related to the treatment of the NTM infection itself [105], [106]. Poor nutritional status has been associated with poor outcomes in MAC lung disease [107], [108]. Vitamin A deficiency was strongly associated with the prevalence and risk for NTM pulmonary disease [109]. Dietitians and nutritionists are important members of the NTM management team. Nutrition assistance can ensure adequate protein and caloric intake, while providing guidance regarding other nutritional supplements, vitamins, micronutrients, and antioxidants.

Summary

Non-tuberculous mycobacterial infections are an increasingly important contributor to human disease, causing a wide range of infection from localized soft-tissue infiltration to disseminated life-threatening conditions. Having a ubiquitous presence in nature can make it challenging to discern NTM biological sample contamination from a true NTM infection. The clinical and radiographic features are non-specific; microbiologic diagnosis is cumbersome and insufficiently informative. Treatment is complex, requiring multi-drug regimens with substantial associated toxicity, administered for a prolonged period. Defining treatment milestones is difficult; clinical outcomes are often disappointing.

Given these considerable challenges, a thorough systematic approach to the diagnosis and treatment of NTM infections should be adopted to maximize the likelihood of positive treatment outcomes. Early diagnosis is key and requires a high index of suspicion and an understanding of the clinical scenarios suggestive of the presence of NTM infection. Identification of the causative organism to the species level should be performed; antimicrobial susceptibility should be tested on all isolates where suspicion of a true NTM infection exists. Treatment should be directed by AST results, in compliance with prevailing guidelines, and with careful attention paid to patient education and proactive monitoring for treatment responses as well as treatment-emergent adverse drug reactions. Antimicrobial therapy should be complemented by adjunct therapies such as pulmonary hygiene measures, nutritional support, and exercise training programs to improve clinical outcomes and quality of life.

Ethical statement

This paper has been written, revised, and reviewed by all co-authors.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.