Quality Control Approaches for Volatile Molecule Analysis Using GCxGC-HRTOFMS, Case Study for TB Sputum Volatiles

GC×GC-HRTOFMS: (1) QC approaches for volatile molecule analysis and (2) A case study on the detection of TB using sputum volatiles

Grant S. Ochoa, Graham E. Browse, Jane E. Hill

In this presentation we discuss (1) the need for quality control of breath and other volatile sample analysis and (2) present a case study of the volatile analysis of sputum.

(1) Biological volatile samples carry intrinsic variation, thus external methods of quality control which assess the instrument and sampling variation are necessary to assess data integrity. In this presentation, I will share common quality control approaches such as the use of external standards, internal standards, check samples, and continuing calibration verifications, which can be utilized to demonstrate instrument reliability throughout sample analysis.

(2) Tuberculosis (TB) is the world’s deadliest infectious disease, claiming more than 1 million lives each year, largely in low- and middle-income countries. The lack of cheap, rapid diagnostics stands in the way between curbing transmission as well as between patients and treatment in many communities. Volatile analysis represents the potential for a rapid, non-invasive test, with the APOPO organization demonstrating rats can distinguish TB by sniffing the headspace of human sputum (N = 24,000 patients[1]). To identify the biomarkers associated with this approach, the headspace of 100 sputum samples obtained from the Foundation for Innovative New Diagnostics (FIND) biobank were analyzed with SPME-Arrow – GC×GC-HRTOFMS. First, we optimized headspace extraction using SPME Arrow. Using that method, from the full set of volatile metabolites detected, partial least squares – discriminant analysis (PLS-DA), was applied to select a panel of 14 features which were optimal for distinguishing culture-positive sputum from negative. The cross-validated PLS-DA model yielded an AUC of 0.92 and a sensitivity of 0.9 and a specificity of 0.84. These results suggest that volatile biomarkers which distinguish tuberculosis are present in sputum and might contribute to a breathprint for tuberculosis that we are actively exploring. Specifically, this work continues through the creation of the Shirley Diagnostics’ TB Breath Atlas which is based on the analysis the volatile molecules breath from over 1,000 patients.

References

[1] Kannan R, et al. Scientific Reports 2021, 11:1, https://doi.org/10.1038/s41598-021-81086-x