Discovery of urinary biosignatures for tuberculosis and nontuberculous mycobacteria classification using metabolomics and machine learning.

Nontuberculous mycobacteria (NTM) infection diagnosis remains a challenge due to its overlapping clinical symptoms with tuberculosis (TB), leading to inappropriate treatment. Herein, we employed noninvasive metabolic phenotyping coupled with comprehensive statistical modeling to discover potential biomarkers for the differential diagnosis of NTM infection versus TB. Urine samples from 19 NTM and 35 TB patients were collected, and untargeted metabolomics was performed using rapid liquid chromatography-mass spectrometry. The urine metabolome was analyzed using a combination of univariate and multivariate statistical approaches, incorporating machine learning. Univariate analysis revealed significant alterations in amino acids, especially tryptophan metabolism, in NTM infection compared to TB. Specifically, NTM infection was associated with upregulated levels of methionine but downregulated levels of glutarate, valine, 3-hydroxyanthranilate, and tryptophan. Five machine learning models were used to classify NTM and TB. Notably, the random forest model demonstrated excellent performance [area under the receiver operating characteristic (ROC) curve greater than 0.8] in distinguishing NTM from TB. Six potential biomarkers for NTM infection diagnosis, including methionine, valine, glutarate, 3-hydroxyanthranilate, corticosterone, and indole-3-carboxyaldehyde, were revealed from univariate ROC analysis and machine learning models. Altogether, our study suggested new noninvasive biomarkers and laid a foundation for applying machine learning to NTM differential diagnosis.

Genitourinary non-tuberculous mycobacterial (GU-NTM) infections: A single institution experience in South India.

INTRODUCTION: Non-tuberculous mycobacterial (NTM) infections are rarely reported, and more so with genitourinary infections. This retrospective study was designed to understand the proportion and behaviour of genitourinary non-tuberculous mycobacterial (GU-NTM) infections compared with genitourinary mycobacterial tuberculosis (GU-MTB) treated at a tertiary care hospital in South India. MATERIALS AND METHODS: The hospital records of every bacteriologically proved GU-MTB and GU-NTM infections treated at this centre from 2010 to 2016 were retrospectively reviewed. RESULTS: There were ten patients of GU-NTM and 15 patients of GU-MTB. There was no significant difference in presentation other than lesser frequency of irritative lower urinary tract symptoms (LUTS) among patients with GU-MTB. Urine smear for AFB was positive in 60% and 47% of GU-NTM and GU-MTB patients. 40% of GU-NTM patients had history of urinary tract instrumentation. Mycobacterium abscessus was grown in four patients and one had Mycobacterium fortuitum/chelonae complex; all the rest were rapid growers. No patient had multi-drug resistant tuberculosis. Imaging studies of GU-NTM patients were indistinguishable from GU-MTB with renal, ureteral and bladder involvements, and stone formation. The drug sensitivities varied among the NTM patients but all showed sensitivity to clarithromycin uniformly. Need for varieties of surgeries in the early and late phases were also comparable. CONCLUSIONS: GU-MTB and GU-NTM infections are indistinguishable from their clinical presentation and imaging studies. All cases of suspected genitourinary mycobacterial infections must be subjected to nucleic acid testing. Treatments based on clinical and radiological features without culture studies may misdiagnose GU-NTM infections as MDR GU-MTB, thereby delaying the appropriate treatment.

Urine lipoarabinomannan as a marker for low-risk of NTM infection in the CF airway.

BACKGROUND: Individuals with Cystic fibrosis (CF) are the most vulnerable population for pulmonary infection with nontuberculous mycobacteria (NTM). Screening, diagnosis, and assessment of treatment response currently depend on traditional culture techniques, but sputum analysis for NTM in CF is challenging, and associated with a low sensitivity. The cell wall lipoarabinomannan (LAM), a lipoglycan found in all mycobacterial species, and has been validated as a biomarker in urine for active Mycobacterium tuberculosis infection. METHODS: Urine from a CF cohort (n = 44) well-characterized for NTM infection status by airway cultures was analyzed for LAM by gas chromatography/mass spectrometry. All subjects with positive sputum cultures for NTM had varying amounts of LAM in their urine. No LAM was detected in subjects who never had a positive culture (14/45). One individual initially classified as NTM sputum negative subsequently developed NTM disease 657 days after the initial urine LAM testing. Repeat urine LAM testing turned positive, correlating to her positive NTM status. Subjects infected with subspecies of M. abscessus had greater LAM quantities than those infected with M. avium complex (MAC). There was no correlation with disease activity or treatment status and LAM quantity. A TB Capture ELISA using anti-LAM antibodies demonstrated very poor sensitivity in identifying individuals with positive NTM sputum cultures. CONCLUSION: These findings support the conclusion that urine LAM related to NTM infection may be a useful screening test to determine patients at low risk for having a positive NTM sputum culture, as part of a lifetime screening strategy in the CF population.

Urine lipoarabinomannan point-of-care testing in patients affected by pulmonary nontuberculous mycobacteria--experiences from the Danish Cystic Fibrosis cohort study.

BACKGROUND: The urine lipoarabinomannan (LAM) strip test has been suggested as a new point-of-care test for active tuberculosis (TB) among human immunodeficiency virus (HIV) infected individuals. It has been questioned if infections with nontuberculous mycobacteria (NTM) affect assay specificity. We set forth to investigate if the test detects LAM in urine from a Danish cystic fibrosis (CF) population characterized by a high NTM prevalence and negligible TB exposure. METHOD: Patients followed at the Copenhagen CF Center were comprehensively screened for pulmonary NTM infection between May 2012 and December 2013. Urine samples were tested for LAM using the 2013 Determine™ TB LAM Ag strip test. RESULTS: Three-hundred and six patients had a total of 3,322 respiratory samples cultured for NTM and 198 had urine collected (65%). A total of 23/198 (12%) had active pulmonary NTM infection. None had active TB. The TB-LAM test had an overall positive rate of 2.5% applying a grade 2 cut-point as positivity threshold, increasing to 10.6% (21/198) if a grade 1 cut-point was applied. Among patients with NTM infection 2/23 (8.7%) had a positive LAM test result at the grade 2 cut-point and 9/23 (39.1%) at the grade 1 cut -point. Test specificity for NTM diagnosis was 98.3% and 93.1 for grade 2 and 1 cut-point respectively. CONCLUSIONS: This is the first study to assess urine LAM detection in patients with confirmed NTM infection. The study demonstrated low cross-reactivity due to NTM infection when using the recommended grade 2 cut-point as positivity threshold. This is reassuring in regards to interpretation of the LAM test for TB diagnosis in a TB prevalent setting. The test was not found suitable for NTM detection among patients with CF.

Genitourinary tract infections due to atypical mycobacteria.

A patient with chronic prostatitis caused by Mycobacterium kansasii is reported, demonstrating that these organisms occasionally cause urinary tract infections. Criteria useful in evaluating the clinical significance of atypical mycobacteria isolated from the urine are suggested. Application of these criteria to 19 patients encountered in routine clinical practice demonstrates their usefulness.

[The importance of Scotochromogenous mycobacteria in the urine]. / La importancia de las micobacterias escotocromógenas en la orina. Análisis de 155 casos

In order to find the significance of the presence of Scotochromogenous Mycobacteria in the urine, 155 cases with this finding and without other Mycobacteria were analyzed. Most had urinary symptoms (91.6%), abnormal urine analysis (62.5%), negative urine culture (71.7%), and changes in intravenous pyelograms (89.5%). Clinical and laboratory improvement was noticed in 70% of those who received antituberculous drugs and in 30% of those who did not receive them (P less than 0.01). All of the foregoing characteristics were present in each 23 patients with repeated isolations of Scotochromogens in the urine, and could be cases of urinary-tract disease due to Scotochromogens. In other patients this Mycobacteria could coexist with M. tuberculosis, or be a contaminant, although Scotochromogens were not found in the urine of 100 control subjects, a significant difference to these patients (P less than 0.0001).