The Rapid Diagnosis of Drug-Resistant and Susceptible Tuberculosis Using Mass Spectrometry and Machine Learning

*Important Notice: Preprints with The Lancet / SSRN publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be considered conclusive, guide clinical practice/health-related behaviors, or treated as established information.

In a recent study published on Preprints with The Lancet SSRN* server, a team of researchers in China has developed a machine learning-based diagnostic tool to detect tuberculosis and drug-resistant tuberculosis using nanoparticle-enhanced laser desorption/ionization mass spectrometry (NPELDI MS ) to determine metabolic fingerprints from serum samples.

Study: Machine Learning of Serum Metabolic Fingerprints for the Diagnosis of Tuberculosis and Drug-Resistant Tuberculosis: An Observational Study. Image Credit: Kateryna Kon/Shutterstock

Background

After coronavirus disease 2019 (COVID-19), tuberculosis is the most common cause of death from an infectious agent, and the emergence of drug-resistant diseases Mycobacterium tuberculosis heightened public health concerns. Unfortunately, statistics indicate that nearly 40% of TB cases are not diagnosed in time for early treatment.

Current methods for detecting tuberculosis include immunological examinations, detection of the etiological agent from sputum smears and molecular biological methods. While sputum smear tests are rapid, their specificity and sensitivity in detecting tuberculosis are low, while mycobacterial cultures, which are accurate, require longer processing times and are difficult for the tests to large-scale and same-day treatment of TB patients in general hospitals. Additionally, molecular diagnosis of TB and rifampicin-resistant TB from sputum specimens presents a challenge due to high costs and variable susceptibility to extrapulmonary TB. Therefore, a feasible test method with high sensitivity and specificity that can be used to rapidly detect TB and drug-resistant TB is essential.

About the study

In the present study, researchers designed an NPELDI MS platform with machine learning algorithms to simultaneously detect TB and drug-resistant TB based on metabolomic fingerprints of serum samples. When Mr. tuberculosis parasitizes host macrophages, affects host metabolism, and metabolites formed from various metabolic reactions indicate the response to environmental, proteomic, and genomic changes. The identification and quantification of these metabolic fingerprints can be used to detect and diagnose various diseases.

For this observational study, researchers recruited 110 patients with pulmonary tuberculosis and 118 healthy people between 2020 and 2021. Tuberculosis was diagnosed based on positive sputum smears, Mr. tuberculosis cultures, Mr. tuberculosis nucleic acid screening, chest X-ray and pulmonary histopathological diagnostics. Drug susceptibility testing or Gene Xpert testing for rifampicin resistance Mr. tuberculosis was used to classify TB patients.

Serum samples were collected for NPELDI MS analysis to determine metabolic fingerprints, which were processed using machine learning algorithms to identify biomarkers of drug-susceptible and drug-resistant tuberculosis.

Results

The results indicated that the MS-based machine learning method NPELDI could differentiate TB patients from healthy individuals with 85% sensitivity and 100% specificity. The method was also able to distinguish between rifampicin-susceptible and rifampin-resistant tuberculosis patients with a sensitivity of 87.5% and a specificity of 85.7%.

Metabolite biomarkers used to detect tuberculosis and drug-resistant tuberculosis included lipids such as monoglyceride, phosphatidylcholine, ceramide, triglyceride, cholesterol ester, amino acids, phosphates, octacosanoic acid and d other basic compounds. Biomarker analyzes revealed that Mr. tuberculosis infections disrupt lipid metabolism pathways such as sphingolipid and glycerophospholipid metabolism. Phosphates such as nicotinamide adenine dinucleotide phosphate and all-transheptaprenyl diphosphate were abnormally expressed, and glutathione concentrations were lower in tuberculosis patients.

Biomarkers used to differentiate between rifampin-susceptible and rifampin-resistant tuberculosis patients included uric acid, taurine, ascorbic acid, and homocysteine, which were elevated in rifampicin-resistant tuberculosis patients. Researchers believe that sulfur amino acids such as homocysteine ​​and taurine may indicate drug-resistant tuberculosis because they are associated with antioxidant and membrane-stabilizing activity. These amino acids are thought to protect the liver from the toxic effects of anti-tuberculosis drugs such as rifampin and isoniazid. Homocysteinemia, the increase in serum or plasma homocysteine ​​levels, is often reported during anti-tuberculosis treatment.

The study had some limitations, such as the small sample size of rifampicin-resistant TB patients, as only 58 serum samples were included in the analysis to determine biomarkers to differentiate between drug-resistant TB patients. of those susceptible to drugs, of which 28 were from rifampicin-susceptible patients. This could have affected the accuracy of the diagnosis.

conclusion

In summary, the results identified biomarker panels and serum metabolic fingerprints to diagnose tuberculosis from serum samples with high specificity and sensitivity, and to differentiate rifampicin-resistant tuberculosis patients from those who were susceptible. to rifampicin. Machine learning algorithms and the NPELDI MS method using optimized ferric particles could help in the early and accurate detection of tuberculosis and improve the prognosis and care of tuberculosis patients.

*Important Notice: Preprints with The Lancet / SSRN publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be considered conclusive, guide clinical practice/health-related behaviors, or treated as established information.