Discussing our VOC Atlas Paper: Interview with Wisenave Arulvasan

Can you tell us about your background and why you chose to research breath in particular?

I have been fortunate enough to contribute to research in a diverse range of medical areas in my career, all the way from neurological disease research during my University of Cambridge summer scholarship and human DNA epigenetic modification during my fellowship in Vienna,  to prenatal birth defects research at University College London (UCL) in collaboration with Great Ormond Street Hospital, and metabolomic research studies in blood, urine and stem cells for 2 years at the University of Cambridge after graduating from UCL.

Breath research is an area I’m particularly fascinated by as I believe there’s so much untapped potential to provide non-invasive real time diagnostics for various diseases. This can facilitate early diagnosis and precision medicine, which can significantly improve patient outcomes, and save lives through increased efficacy of existing treatments, for many conditions.

Both concrete evidence, such as the current utilization of the lactulose breath test as a gold standard diagnostic tool in the clinic for small intestine bacterial overgrowth (SIBO), along with tantalizing clues into the immense diagnostic potential of breath, through studies of trained dogs demonstrating the ability to reliably distinguish between the breath profiles of humans with lung or colorectal cancer vs. those without, hugely motivated me to pursue breath research.

What was the motivation and aims behind embarking on the VOC Atlas project?

One of the key motivations underpinning the VOC Atlas project was the desire to advance the field of breathomics by building up knowledge of what VOCs are truly “on-breath” (i.e., those reliably distinguished from background signals), through the development of a robust identification workflow.

This facilitates on-breath VOCs’ targeted optimization through the breath research pipeline from pre-analytical considerations all the way to untargeted feature extraction, thus improving capture and measurement of VOCs truly originating from breath; the more accurately and reliably we can do this, the higher the likelihood of finding the all-important on-breath VOC biomarkers, for development into clinical diagnostic tools.

What makes this project different from work that has previously looked at compounds in breath?

This work’s uniqueness is related to the combination of processes implemented including:  

• Equipment blank samples are collected prior to every breath sample, to reliably distinguish truly on-breath VOCs from VOCs originating from the sampling hardware or background contamination. 
• Three metrics for statistical determination of whether a VOC is on-breath by comparing the signal observed in breath samples vs. the equipment blank samples have been utilized: 1) standard deviation 2) paired-test and 3) ROC-AUC (Receiver Operating Characteristic-Area Under Curve). Each metric has unique advantages, and using all three in combination allows a wide net to be cast – ideal for early discovery work. 
• A robust multi-step identification workflow was developed to allow confident, traceable assignment of chemical identities to on-breath features. 

Were there any surprises in the results?

Yes. Chlorinated compounds typically wouldn’t feature as likely on-breath compounds if most people were to take a guess. However, we’ve identified Chlorodibromomethane as an on-breath VOC in a large proportion of breath samples. The principal source of human exposure to this compound is thought to be ingestion of chlorinated tap water, along with inhalation/dermal exposure during showering and swimming in chlorinated pools. Chlorodibromomethane is metabolized by CYP450 enzymes in the liver into CO2/CO and rapidly eliminated via exhalation, and therefore can be measured in breath. 

What is an example of a breath VOC identified in this work and what could it tell you?

One example of an identified on-breath VOC is indole, which is generated by the catabolism of tryptophan, mediated by the human gut microbiome. Elevated levels of indole observed in cirrhosis patients could be explained by impaired hepatic clearance, providing a plausible mechanistic relationship between a microbiome-related VOC and a clinical diagnosis of cirrhosis. 

What is the significance of identifying these compounds in the wider context of breath research/analysis?

Targeted research activities: knowing what’s truly on-breath allows the targeted development of analytical methods that can reliably and accurately measure these on-breath VOCs, along with targeted optimization of pre-analytical factors, such as storage time, and dry purge settings to minimize analytical variability and accurately measure biological variability. 

Development of custom panels for discovery work: knowing what’s truly on-breath allows for biological interpretation, such as deciphering the potential origins of the VOC, which biological pathways they may be involved in, and associations to disease in the literature. This enables the design of custom panels of VOCs known to be on-breath and pertinent to the disease population being studied.

For example, butyric acid was successfully identified on-breath, and biological interpretation reveals it’s produced by the gut microbiome and associated with human gut health. Thus, recommendations could be made to consider inclusion of butyric acid in a custom VOC panel for analysis alongside breath samples in studies looking at gut health (e.g. irritable bowel syndrome (IBS)).  

Analysis of calibration curves of panel VOCs alongside samples has the added benefit of improved quantitation of levels of the VOC on-breath within the study population, in comparison to retrospective calibration, improving the likelihood of biomarker detection.

Following these findings, are there any next steps planned to further advance the breath analysis field?

One of the next steps for the VOC Atlas is targeted characterization and optimization of pre- and analytical factors to reliably capture and measure identified on-breath VOCs in breath studies. 

We also want to increase the number and chemical diversity (e.g. chemical class, polarity, volatility) of identified on-breath VOCs by systematic modification of analytical factors (such as GC column phases, TD sorbents) and sampling of different populations.