Culture independent markers of nontuberculous mycobacterial (NTM) lung infection and disease in the cystic fibrosis airway

Abstract

Nontuberculous mycobacteria (NTM) are opportunistic pathogens that affect a relatively small but significant portion of the people with cystic fibrosis (CF), and may cause increased morbidity and mortality in this population. Cultures from the airway are the only test currently in clinical use for detecting NTM. Culture techniques used in clinical laboratories are insensitive and poorly suited for population screening or to follow progression of disease or treatment response. The lack of sensitive and quantitative markers of NTM in the airway impedes patient care and clinical trial design, and has limited our understanding of patterns of acquisition, latency and pathogenesis of disease. Culture-independent markers of NTM infection have the potential to overcome many of the limitations of standard NTM cultures, especially the very slow growth, inability to quantitate bacterial burden, and low sensitivity due to required decontamination procedures. A range of markers have been identified in sputum, saliva, breath, blood, urine, as well as radiographic studies. Proposed markers to detect presence of NTM or transition to NTM disease include bacterial cell wall products and DNA, as well as markers of host immune response such as immunoglobulins and the gene expression of circulating leukocytes. In all cases the sensitivity of culture-independent markers is greater than standard cultures; however, most do not discriminate between various NTM species. Thus, each marker may be best suited for a specific clinical application, or combined with other markers and traditional cultures to improve diagnosis and monitoring of treatment response.

1. Introduction

Nontuberculous mycobacteria (NTM) are important pathogens in cystic fibrosis (CF) and other obstructive lung diseases. The CF population has the highest known risk for NTM, and poses unique challenges with regards to diagnosis and treatment [1]. The lack of sensitive and specific markers of NTM in the airway makes significant improvement of patient care difficult. Markers are needed to screen for infection, and to help distinguish between indolent infection and NTM disease. In those receiving treatment, markers are needed to assess the response to therapy within the individual, and to determine if an endpoint has been achieved which could signal end of treatment. Currently, airway cultures combined with an appropriate clinical presentation are the “gold standard” by which all diagnosis and treatment decisions are based. Limitations of culture include slow growth of the organisms and low sensitivity due to required decontamination procedures. In addition, sputum samples are now frequently unavailable in a large proportion of the CF population due to the introduction of CF transmembrane-conductance regulator (CFTR) modulator therapy. Furthermore, the subspecies of M. abscessus (Mab) may be particularly difficult to culture from CF sputum due to concurrent treatment with antibiotics against P. aeruginosa or S. aureus, many of which have sufficient activity against Mab to suppress growth in culture.

The objective of this brief review, which collates data presented at the 2022 Colorado NTM Conference at Colorado State University in Ft. Collins, CO (May 31-June 4, 2022), is to provide an overview of current and emerging culture-independent markers of NTM, and the strengths and weakness of each to complement traditional airway cultures. The potential of these markers, alone or combined, to replace and possibly surpass cultures, will be reviewed in the context of CF airway disease.

2. NTM airway infection and disease

NTM are well-recognized pathogens in the CF population, and are widely viewed as one of the most challenging infectious complications of the disease [1,2]. People with CF (pwCF) have the highest prevalence of pulmonary NTM cases compared to other disease states [3], but pulmonary NTM disease is also a significant complication of non-CF bronchiectasis and chronic obstructive pulmonary disease (COPD). Longitudinal data from U.S. CF Foundation (CFF) Patient Care Registry reveals that 20% of patients who were cultured over a 4-year span (2012–2016) had one or more NTM species isolated [4,5]. Due to the high prevalence of NTM positive cultures, annual screening of the CF population using sputum culture is recommended in the absence of clinical suspicion [6]. The genetic distribution of NTM species within the CF population and care centers in the U.S. has thus been well-studied [7-10]. M. avium complex (MAC) and subspecies of Mab account for the overwhelming majority of NTM species recovered in CF samples [8, 10], and accurate identification of the pathogen dictates the treatment course [6].

Reaching a diagnosis of NTM pulmonary disease in the setting of CF is uniquely complex as clinical manifestations of infection can range from nondetectable to severe, and the majority of patients have only transient or indolent infection [6,11]. Thus, isolation of an NTM from a respiratory specimen is not synonymous with disease, nor is it necessarily an indication to initiate treatment. Current guidelines call for the presence of >1 positive culture, in the setting of characteristic clinical symptoms and radiographic findings not attributed to other co-infections and co-morbidities [6,12]. The extensive overlap of both clinical features and radiographic manifestations of NTM infection with underlying manifestations of CF makes diagnosis challenging, and requires that clinicians first ensure typical CF co-pathogens and co-morbidities are adequately treated before making a diagnosis of NTM pulmonary disease [6,13]. Initiating NTM treatment in patients with indolent infection and no clinical decline will not result in clinical benefit and places the patient at risk from treatment-related side effects. On the other hand, inability to make a timely and accurate diagnosis can delay treatment when needed, resulting in accelerated clinical decline [11,14].

In those diagnosed with NTM disease and initiated on antibiotics, the decision regarding when to end treatment can be even more challenging than the decision to initiate treatment. Often the patient and/or provider is motivated to “clear” the infection, and there is real concern that too short of a treatment span could result in reoccurrence of a more resistant strain. Conversely, an unnecessarily long treatment course may result in drug toxicity, undo expense, and little long-term benefit protecting against a new infection by a different NTM isolate [11]. In this setting, reliance on cultures is particularly problematic given that the antibiotics utilized may significantly reduce the yield of the cultures, independent of clinical response. Improved methods are needed for detecting NTM in clinical specimens, and for determining if NTM in the airways are bystanders or causing active disease.

3. Limitations of airway cultures and ideal qualities of culture-independent markers

NTM screening, identification, diagnosis, and therapeutic management are currently all based solely on traditional clinical culture methods (Fig. 1). Experts in the field universally recognize that sputum-independent, non-culture-based biomarkers are needed to improve all aspects of NTM disease detection and clinical care [15]. NTM sputum cultures have low sensitivity and poor negative predictive value, require up to 6 weeks to detect slow growing NTM, and are costly. Decontamination protocols for co-infection by P. aeruginosa and S. aureus significantly reduce NTM viability in samples [16]. Consensus guidelines recommend a two-step approach to CF sputum decontamination using NALC-NaOH followed by oxalic acid prior to mycobacterial culture [6,17-19], which may decrease NTM recovery by a log [20], potentially reducing low concentrations of mycobacteria below the threshold of detection [21]. Culture sensitivity is further reduced by the concurrent use of antibiotics by patients such as azithromycin, linezolid, amikacin, imipenem, and tedizolid for the treatment of co-infections, resulting in suppression of NTM growth ex vivo. Sputum cultures are also problematic in patients who do not produce sputum, which is a growing population since the introduction of CFTR modulator therapy, and these individuals may require guided bronchoscopy and the acquisition of bronchoalveolar lavage (BAL) fluid to assess for NTM [15,22].

Fig. 1. Clinical utilization of NTM markers.

A. Screening for presence of infection in the majority of individuals who have never had NTM (white figures) and are at low annual risk for NTM infection. B. Species or complex identification among those found to have positive cultures. C. Diagnosis of indolent infection (yellow figures) versus individuals with NTM disease who may benefit from treatment (right column). D. Within subject determination of treatment response or eradication.

Ideally, culture-independent markers will be available from samples that could be collected noninvasively during routine care. Desirable qualities for novel diagnostic tests include rapid results, relatively low cost, and a quantitative or semiquantitative measure of bacterial burden. Given the ubiquitous nature of NTM in the environment, an ideal marker would likely not be overly-sensitive, which could lead to potential false positive results, and increased uncertainty over treatment decisions. In those diagnosed with NTM disease there is a need for species identification and drug susceptibility testing. No single test will likely be optimal for screening, determining the presence of disease, identifying the species and resistance patterns, and monitoring treatment response. More likely, a variety of tests will be utilized for specific purposes, possibly in combination, and with at least one culture prior to treatment.

4. Compartments for culture-independent testing of airway NTM

As we seek to identify and validate sensitive and specific markers for NTM airway infection and disease, a number of sample types and sources can be considered. In each case, advantages and disadvantages must be weighed, as well as the relative abundance and stability of the marker. The following samples are most commonly assayed:

Sputum:

Sputum is considered a direct sample of the airway environment, and historically, copious sputum production was a hallmark of CF, as well as most forms of bronchiectasis. Through improved treatment strategies, the proportion of pwCF capable of expectorated sputum production has been steadily declining. CFTR modulation is now standard of care for up to 90% of pwCF in the U.S. These molecules directly modulate the activity and trafficking of the defective CFTR protein [23,24]. Of particular interest is “highly effective CFTR modulator therapy” (HEMT) which has resulted in exceptional improvement in mean lung function, body mass index, quality of life, and rate of pulmonary exacerbations. HEMT commonly refers to use of ivacaftor in patients with CFTR gating mutations [25,26], and the triple combination elexacaftor, tezacaftor and ivacaftor (E/T/I) in pwCF who are either homozygous [24] or heterozygous for F508del mutation [23]. A reduction in sputum production is a nearly universal benefit of HEMT, and the vast majority of pwCF produce sputum rarely or only during exacerbations. This is particularly true in pediatric cohorts and adults with relatively mild lung disease. Sputum induction can increase the proportion of pwCF that can be sampled directly from the airway [27], although yield from this technique has fallen as well, and was curtailed in many centers during the COVID-19 pandemic. Bronchoscopy with BAL is increasingly used, but is invasive and not well-suited for serial sample collection. Although sputum sampling has disadvantages, when available it has the advantage of directly accessing pathogenic mycobacteria and commensal flora. In our experience, individuals with NTM disease are among those pwCF most likely to produce sputum. However, with successful treatment, sputum production may become scant and verification of treatment response more challenging, especially in the setting of a clinical trial.

Saliva:

Collection of saliva is easy and noninvasive, and can be accomplished in very young children. The mouth is part of the upper airway, and airway pathogens can be detected in saliva, although in much lower quantities than sputum. The quantity of NTM recovered from oropharyngeal swabs is typically below the level of detection for standard cultures [6]. The main disadvantage for molecular detection is the significant contribution from oral flora, which dwarfs the presence of genetic material from lower airway pathogens. However, saliva is a promising site for immunoglobulin testing.

Breath:

Breath sample collection is non-invasive and feasible for all subjects. Samples may reflect infection occurring within all pulmonary compartments, as well as host response. Some training and special equipment is required.

Blood:

Blood collection is minimally-invasive, feasible in all subjects, and a familiar clinical technique at all testing centers. Samples may reflect infection occurring within all pulmonary compartments. Components to be tested include whole blood, leukocytes, serum, and plasma.

Urine:

Urine collection is non-invasive and feasible for all subjects. Samples may reflect infection occurring within all pulmonary compartments.

4.1. Types of markers

Culture-independent markers can be broadly categorized as biologic components of NTM or markers of the host response to the bacteria. Assays that directly test for bacterial products are often limited by very low quantities of marker, but in some cases may be too sensitive and not distinguish between living or dead bacteria. Markers that test for a host response are dependent on an intact host immunity. Some markers, such as volatile metabolites, detect products that may be derived from both host and pathogen. Many culture-independent markers for NTM can be assessed from more than one sample type (Table 1).

Table 1.

Proposed culture independent markers of NTM.

Compartment	Advantages	Challenges	Markers
Sputum	Direct sampling of the airways	Limited availability	Molecular markers [38,41,42,44]
		Heterogenic sampling	LAM
		May require invasive techniques	Immunoglobulins [83]
			Volatile metabolites [60]
Saliva	Direct connection to airways Readily available Noninvasive	Oral pharyngeal contamination	Molecular markers [107]
		Low yield of NTM products	Immunoglobulins [64,70,72]
Breath	Readily available Noninvasive	Specialized collection	Volatile metabolites [60]
		Nonstandardized analysis	Condensate
Blood	Systemic marker Readily available Minimally invasive	Low yield of NTM products	Immunoglobulins [83]
			RNA signatures [84]
			DNA signatures [48]
			LAM
Urine	Systemic marker Readily available Noninvasive	Low yield of NTM products	LAM [37]
		Nonstandardized analysis
High-resolution computed tomography (CT)	Noninvasive	Nonstandardized analysis	Characteristic features [51,52]
		Radiation exposure

4.2. Lipoarabinomannan (LAM)

LAM is a cell wall lipoglycan found in all mycobacteria species which is being proposed to be used as a biomarker of mycobacterial infection. During infection, LAM is released from metabolically active or degrading mycobacteria into the circulation with subsequent filtration by the kidneys, passing into urine [28,29]. LAM is a validated biomarker for active Mycobacterium tuberculosis (TB) when antigen detection rather than antibody measurement is applied [30-35]. Antigen detection assays based on LAM are available both as ELISAs and as a point-of-care test. Qvist et al. tested the utility of a commercially available TB-specific LAM antigen kit in a Danish CF clinic and determined it was not suitable for diagnosis of NTM in pwCF as the quantity of detected urine LAM antigen in NTM positive pwCF was quite low compared to patients with TB [36]. The urine LAM antigen test had a high specificity (99%), but even by lowering the threshold far below that recommended for diagnosis of TB, they reported a very low sensitivity (39%) for identifying pwCF with positive NTM sputum cultures [36].

Recently we reported the utility of gas chromatography–mass spectrometry (GC-MS) analysis of urine LAM from a cross-sectional analysis of pwCF (n = 44) with well-documented NTM status [37]. All subjects were followed for ≥5 years, with results of ≥5 NTM sputum cultures available at enrollment. All patients were co-infected with at least one other typical CF pathogen, including Pseudomonas aeruginosa, Staphylococcus aureus, or Aspergillus species. The presence or absence of specific monomeric components of urine LAM by GC-MS corresponded perfectly with previous NTM culture history. All pwCF with a past history of one or more positive NTM sputum cultures had a positive urine LAM, however, LAM was undetectable in all subjects never culture positive for NTM [37]. For those individuals with a positive urine LAM, the values were quite variable. Among pwCF whose most recent infection was a subspecies of Mob, the quantity of LAM derivatives were greater than those infected with MAC. The potential to distinguish between M. avium, M. intracellulare and M. chimaera by the quantity and/or pattern of LAM subtypes has not yet been tested. There was no clear association between quantity of detected LAM and clinical features or culture status, as subjects who had positive cultures within a year of urine collection had a similar range of LAM as subjects who had cleared their sputum culture for >1 year, supporting the conclusion that once present, NTM clears slowly from the CF airway, if at all.

LAM may also serve as a marker of effective treatment. In one of the few cases ever reported where Mab was eradicated with certainty, we tested the utility of urine LAM as a marker of treatment response [38]. In 2022 we described airway culture conversion in a 26 y/o man with CF after approximately 100 days of phage therapy [38]. Control of his infection allowed for successful lung transplant 379 days after phage, with no growth of Mabs in the explanted lungs or subsequent airway cultures. Phage and antibiotic treatment were discontinued after 551 days, and he remained culture negative under close observation, despite immunosuppression. Urine LAM measurements were available >2 years prior to phage, and with intensive antibiotic treatment a modest reduction in values was observed. With addition of phage, urine LAM values rose dramatically, presumably due to Mab lysis by the phage, followed by a precipitous drop to below the level of detection (LOD) after approximately 3 months. This drop in urine LAM corresponded to conversion of his sputum cultures to negative, and airway qPCR for Mab also dropping below the LOD [38]. In two pilot studies the percent of TB patients with a positive rapid urine LAM test declined after treatment, supporting the potential utility of urine LAM as a marker of treatment response of NTM, if NTM-specific assays were available [39,40].

4.3. Molecular methods of identification

Molecular testing based on DNA amplification from sputum offers advantages over traditional sputum culture in that it bypasses sample decontamination and can be applied directly to total DNA extracted from raw sputum. Two general methods of quantitating bacterial DNA have been evaluated: quantitative PCR (qPCR) and 16S gene sequencing. NTM nucleic acids have previously been detected in clinical samples by these methods with varying success. Targeted qPCR assays for MAC and Mab nucleic acids were applied to over 100 clinical specimens (respiratory and non-respiratory) with sensitivities of 83% and 90%, respectively, compared to culture [41,42] In contrast, a pilot study that applied qPCR to 15 culture positive CF respiratory samples showed poor detection of NTM nucleic acids (in ~20% of samples) [43], though detection was improved to ~40% with a modified DNA extraction method. Another small study analyzed CF sputum samples (n = 23) by qPCR and demonstrated 100% correlation of qPCR and Mab culture results (only four samples were culture positive), with a detection limit of 10³ cfu/mL [44]. We have applied qPCR to longitudinally collected sputum samples from a CF subject undergoing bacteriophage therapy and detected Mab DNA in 89% of culture positive samples [38]. We also observed a dramatic decline in Mab burden after phage administration suggesting that qPCR is a potentially useful tool for assessment of treatment response.

Conventional 16S rRNA gene sequencing has been widely used for detection of common CF pathogens such as Pseudomonas and Staphylococcus, but this approach has been less successful for detection of Mycobacterium in CF and non-CF respiratory samples in the context of the entire microbiome [43,45,46]. This is largely attributed to the fact that NTM have only a single copy of the 16S rRNA compared to higher copy numbers in other microbiome bacteria resulting in sporadic detection of Mycobacterium reads. Furthermore, partial sequencing of the 16S rRNA does not allow for classification of Mycobacteria beyond the genus level and thus does not distinguish pathogenic species from benign colonizers (e.g. M. gordonae). Modifications to this approach including nested amplification [46] or multiplex amplicon sequencing using species-specific primers would improve sensitivity of detection and allow for simultaneous identification of MAC and Mab in clinical samples such as sputum and saliva. Key features of a successful molecular test include 1) the ability to distinguish NTM species or subspecies and 2) a detection threshold suggestive of active infection.

4.4. Circulating DNA signatures

Diagnostically informative microbial cell-free DNA (cfDNA) can be detected from blood plasma during fulminant infections such as sepsis [47]. A modified “liquid biopsy” technique has now been reported to detect circulating DNA of NTM and other CF pathogens [48]. The approach uses a panel of oligonucleotide capture probes that are complementary to regions of the highly divergent bacterial 16S rRNA gene, and can be used to assign taxonomic classifications, often to the species-level. The oligonucletoide panel spans the entire length of the 16S rRNA gene, with probes having various degrees of sequence diversity redundantly designed such that one or more of the individual oligonucleotides have close homology with each bacterial species [48]. Cell-free DNA from CF patients has been compared to sputum cultures, with correlation between sequence data and sputum culture, consistent with the detection of bacterial DNA originating from the lungs of CF patients.

4.5. High-resolution chest computed tomography (CT)

High-resolution chest CT scans are an important and sensitive measure of pulmonary disease in patients with and without CF [49,50]. Cavitary lesions, nodules, tree-in-bud opacities, and consolidation have been described as characteristic features of NTM pulmonary disease on chest CT [51,52]. In a pilot cohort study, these 4 features combined with the validated Brody CF bronchiectasis score [53,54] did not sensitively detect a difference between indolent NTM and NTM pulmonary disease [52]. However, this study was limited by the fact that NTM pulmonary disease was defined retrospectively. Because established scoring systems, such as the Brody score, are focused on bronchiectasis and air trapping rather than opacities and consolidation observed in this disease, NTM-specific chest CT criteria need to be defined and a scoring system needs to be developed. Data published from our team show that use of automated, quantitative imaging chest CT scores can both classify disease [55] and determine risk of disease progression and mortality in other pulmonary disorders such as idiopathic pulmonary fibrosis [56]. However, this has not been tested in patients with NTM and CF. A manual, quantitative NTM chest CT score that differentiates indolent infection and NTM pulmonary disease would be a useful, objective clinical tool to improve accuracy and efficiency of NTM pulmonary disease diagnosis. Establishing a manual NTM chest CT score may also be useful for development of future automated scoring algorithms. This is currently underway as an ancillary trial to the CF Foundation (CFF)-sponsored PREDICT (PRospective Evaluation of NTM Disease In CysTic Fibrosis, NCT02073409) and PATIENCE (Prospective Algorithm for Treatment of NTM in Cystic Fibrosis, NCT02419989) Trials.

4.6. Volatile metabolites

Analysis of volatile compounds from the breath or sputum sample headspace is a promising diagnostic approach for infectious diseases in the pulmonary system, and several pilot studies have successfully used breath biomarkers for identification of tuberculous disease [57-59]. In 2022, a pilot study investigated whether exhaled breath can differentiate between CF patients with active NTM lung disease compared to those with indolent infection and those without any history of a positive NTM culture from their airway [60]. Seventeen breath molecules were identified that together discriminate between pwCF with active NTM disease and those who do not, and demonstrated an example where breath analysis identified a subject with NTM infection at the time of first culture conversion [60]. Work is ongoing to test this approach in a variety of CF cohorts, including participants in the PREDICT and PATIENCE Trials.

4.7. Immunoglobulins

The use of immunoglobulins for detection of pathogens, based on a robust humoral immune response to infection, is established in CF. Induction of plasma antibodies can occur within days of exposure to pathogens and are generally maintained for the duration of infection. Detection of pathogens-specific antibodies correlates well with culture results for P. aeruginosa [61-65] and circulating antibody levels have been used to predict the onset of infection and response to treatment in pwCF [61-64,66]. Pathogen-specific antibodies can also confirm CF sputum culture results for B. cepacia [67], and A. fumigatus [68,69]. Sputum IgA also correlated with culture results [64,70-72], future infection status [61-63,66,73,74] and treatment outcomes [61-63,66,68,69,73,74] suggesting these may be ideal markers of host response in the respiratory tract. A few studies have addressed the role or appearance of saliva IgA in pulmonary infection [64,70,72].

NTM infection status has been analyzed by measuring anti-MAC antibodies in plasma [75-79]. These studies consistently detected samples from culture-positive subjects and discriminated between subjects with and without pulmonary disease and with treatment outcome [75-77,79], but were not able to discriminate Mab from MAC infections [78,79], likely due to the use of common mycobacterial antigens. Measurement of anti-Mab antibodies has been reported less frequently but also used common mycobacterial antigens such as the glycopeptidolipid core or antigen A60 which failed to discriminate Mab from MAC infections [78-81]. Recent publications described serum IgG and IgA recognizing a protein fraction of rough Mab and recombinant PLC with an improved ability to separate Mab from MAC infections [82,83]. We developed an ELISA to detect plasma anti-Mab antibodies using bacterial lysates to assess Mab infection status. Mean titers of anti-Mab IgM were not different between cohorts with a history or Mab, MAC or never culture positive, likely reflecting frequent environmental exposure to the bacteria. Mean levels of anti-Mab IgG3 correlated well with culture results and had highest specificity for Mab [83]. However, with all antibodies tested, a portion of pwCF did not mount an antibody response, reducing the negative predictive value of the test. Using this method we observed a decline in anti-Mab IgG and IgA after phage administration [38].

4.8. Circulating RNA signatures

Circulating leukocyte transcriptome profiles have the potential to help differentiate between latent and active disease in mycobacterial lung infection, by identifying key host responses associated with effective anti-mycobacterial immune responses. The application of transcriptional blood signatures revealed type I IFNs as key drivers of inflammation during acute pulmonary TB infection [84]. In the setting of pulmonary inflammation, transcriptional signatures have identified subjects at risk for developing clinically symptomatic disease as well as predicting outcomes from treatment by measuring changes in gene expression, thus capturing rapidly changing host inflammatory responses [85-94].

We have applied this method to predict reduced pulmonary infection in the treatment of CF pulmonary exacerbations, utilizing a 10-gene panel [95,96]. In a 5-year follow-up, the panel also identified subjects at increased risk for mortality and disease progression, based upon an inflammatory signature which is predominantly neutrophilic versus lymphocytic [97]. The method requires collection of whole blood samples into PAXgene^™ tubes, with RNA isolation and purification followed by first strand cDNA synthesis. Log-transformed transcript abundance is normalized by a housekeeping gene and is then quantified and compared before and after treatment. A computational algorithm determines prediction probability identifying likelihood of therapeutic response. This method could be applied to pwCF and NTM lung infections to help determine latent vs active disease and efficacy of treatment protocols.

4.9. Airway immune cell signatures

Methods for phenotyping the host immune response are being developed to distinguish between indolent and active infection for many pathogens and opportunistic infection [98,99]. Particularly, for organisms that can survive within mononuclear phagocytes (including mycobacterial species), characterization of the transcriptomic, proteomic, or metabolomic state of airway macrophages can reveal whether inhaled microbes are being eradicated or contained by host cells, or rather are replicating and causing active disease [100,101]. Early investigations have detected unique macrophage transcriptional profiles and sputum metabolites that discriminate between control of NTM infections and active bacterial replication [102,103]. Application of single cell RNA sequencing to induced or spontaneously expectorated sputum [104], to bronchoalveolar lavage fluid [105], or to sorted sputum immune cell populations [106], promises to reveal unique signatures of permissive and restrictive immune cell phenotypes correlating with transient/indolent NTM infection and active disease.

5. Conclusions

A variety of metabolic and molecular markers have now been identified with the potential to monitor different aspects of NTM infection and treatment response. For each, formal validation trials are needed, as well as a better appreciation for the appropriate application, alone and in combination with orthogonal markers and standard cultures. In studies to date, urine LAM and volatile metabolites appear particularly well-suited for screening populations for initial infection, and for monitoring response to treatment. Molecular markers can distinguish between species and identify antibody resistance, and could be used to guide treatment decisions. Immunoglobulin titers have a high specificity for infection when positive, but false negative results are common, and the test may be best suited for within-subject monitoring of response to treatment and surveillance for reinfection following successful treatment. Likewise, HRCT studies are likely most useful in assessing response to treatment within subjects. As discussed above, we have recently combined results from urine LAM, sputum qPCR, serum anti-Mab immunoglobulins and traditional cultures to demonstrate eradication of Mab pulmonary disease in response to bacteriophage treatment over several years [38]. Combining culture-independent markers is feasible, and may be the only approach to follow treatment response with certainty. While it seems unlikely that culture-independent markers will completely replace traditional cultures, various markers may someday be relied on for clinical decisions, such as prompting the need for bronchoscopy in a child unable to produce sputum, or signaling the end of treatment with high certainty that the burden of infection has reached a baseline.