Detection of tuberculosis-associated compounds from human skin by GCxGC-TOFMS

Area of Science:

• Biochemistry
• Analytical Chemistry
• Medical Diagnostics

Background:

• Tuberculosis (TB) remains a significant global health challenge.
• Accurate and accessible diagnostic methods are crucial for TB control.
• Current diagnostic approaches can be invasive or have limitations.

Purpose of the Study:

• To explore the potential of human skin volatile organic compounds (VOCs) as non-invasive biomarkers for TB diagnosis.
• To identify specific VOCs associated with TB infection.
• To evaluate the diagnostic accuracy of these VOCs using predictive modeling.

Main Methods:

• Utilized a wearable silicone rubber band for non-invasive VOC sampling from human skin.
• Employed comprehensive two-dimensional gas chromatography-time of flight mass spectrometry (GCxGC-TOFMS) for detailed chemical analysis.
• Applied chemometric techniques, including targeted and untargeted screening, and artificial neural network (ANN) modeling for data analysis.

Main Results:

• Identified 27 known TB-associated VOCs and 16 novel potential biomarkers on human skin.
• ANN models demonstrated high diagnostic accuracy: 92% AUROC for targeted analysis (two compounds) and 98.5% AUROC for untargeted analysis (16 features).
• Achieved high sensitivity (100% and 83%) and specificity (94% and 88%) in differentiating TB-positive from TB-negative individuals.

Conclusions:

• Human skin VOCs represent a promising, non-invasive biomarker source for TB diagnosis.
• The identified VOC profiles offer a potential alternative to existing diagnostic methods like breath analysis and GeneXpert.
• Further validation could lead to a simple, accessible TB screening test.