A breath test for cancer? Let’s combine the best of analytical techniques
Wednesday, 15 April 2015 at 3:47:PM
Breath analysis was in the news yesterday morning (14 April 2015), with a BBC News article highlighting the potential of analytical methodology to identify life-threatening diseases, and articles also on ITV News, The Telegraph, The Independent and Medical News Today. This illustrates the increasing public relevance of this topic, confirming a trend I highlighted on this blog two years ago.
The new study, published in the journal Gut, is a further significant step down the road to widespread clinical adoption of breath-monitoring technology and focuses on gastric cancer. This disease is all too often only diagnosed at a late stage, when the survival outlook is poor. However, early diagnosis requires a gut biopsy, so the search is on for a quicker, non-invasive way of determining whether a patient is at risk of developing cancer.
The team behind the new research, led by Hossam Haick at the Israel Institute of Technology, Haifa, have shown how analysing breath samples by two techniques could lead to a practical solution to this problem.
In the study, they took breath profiles of 484 subjects with various stages of intestinal disease, including 99 patients with gastric cancer. Breath samples were collected in bags, and then pumped through sorbent-packed tubes to capture the volatile organic compounds (VOCs).
The first part of the analysis was an ‘information-gathering’ exercise, and used thermal desorption (TD) with GC–MS to determine the overall VOC profile.
The second part of the analysis showed how this information could be used to develop a simple diagnostic test for gastric cancer. Volatiles captured on the sorbent tubes were transferred to a small chamber containing an array of eight nanomaterial-based sensors, and conductivity readings were taken for 5 minutes. 70% of the subjects were used to ‘train’ the sensor array on the basis of the TD–GC–MS data, and the remainder were used to validate the algorithms developed.
Distinguishing patients with cancer
In the GC–MS part of the study, a total of 130 VOCs were identified and quantified, but statistically significant differences between the patient groups were noted for just eight of them – propenenitrile, furfural, 2-butoxyethanol, hexadecane, 4-methyloctane, 1,2,3-trimethylbenzene, α-methylstyrene and butan-2-one.
The nanoarray work gave equally impressive results. A key finding was that they could accurately distinguish between patients with gastric cancer and those in the ‘high-risk’ group – 36 out of 40 subjects had their condition correctly identified.
The authors are understandably very positive about all this, saying that it “demonstrates the feasibility of the sensor nanoarray for distinguishing between malignant and non-malignant conditions […] by a breath test”. I would absolutely agree, and would add that it is particularly encouraging because it tests the technology in a ‘real-world’ scenario – vital to make real progress towards the routine use of breath screening by clinicians.
Benefitting from multiple analysis techniques
However, I have one criticism of something the authors said in their paper. As justification for using nanoarrays for routine screening, they say that “GC–MS technology cannot be applied for screening purposes since it requires expensive devices, high-level expertise and long sample analysis times”.
This, I think, is a rather casual, out-of-date statement that doesn’t reflect the latest developments in the field, which have considerably reduced processing times. For example, multi-tube TD autosamplers, ‘fast GC’ techniques with sub-10-minute runs and highly sensitive mass spectrometers are already changing the way people use GC–MS, while more powerful software is eliminating the need for extensive experience of data analysis.
As for cost, advanced technologies need to repay the cost of research, development and production – a logic that will also apply to these nanoarrays, at least until they’re produced for a mass market. Until then, is it not reasonable to view the financing of life-saving analytical technologies in the same way as that of novel drugs?
The authors’ comments also underplay the contribution of GC–MS to this type of research. As a long-established, well-validated analytical protocol, GC–MS is vital in providing the baseline quantitative data that later research relies upon – as the authors have amply illustrated in this study.
So let’s continue to acknowledge the value that a variety of analytical techniques bring to the challenge of dealing with life-threatening diseases, and use the best available method at every stage. We should all look forward to a future where novel technologies are used for rapid, routine disease diagnosis – supported throughout the cycle of research, development and deployment by the analytical rigour of established techniques like GC–MS.
For lots more examples of studies into monitoring VOCs in breath using TD, have a look at the “Disease diagnosis” section of our literature compendium, Application Note 004 (“Publications and presentations citing Markes’ products”).
David Barden received his Ph.D. in Organic Chemistry from Cambridge University in 2004, and during his time as an editor at the RSC wrote news pieces for Chemistry World on various scientific topics. He is now Technical Copywriter at Markes International, where he draws on the expertise of his colleagues to explain how new thermal desorption and mass spectrometry technologies can be applied to analyse volatile organic compounds in a wide variety of situations.