An article in the online version of the British Journal of Cancer suggests the possibility of developing a breath test capable of detecting risk for four different but common types of cancer: lung, colon, breast, and prostate cancers.
Peng et al. have combined the use of a highly specialized array of nanosensors with gas chromatography and mass spectrometry (GC-MS) to identify specific, volatile organic compounds (VOCs) in the breath of known cancer patients and healthy volunteers.
We know that the growth of solid tumors is associated with changes in the expression of genes and/or proteins that can lead to the emission of VOCs. Peng et al. took advantage of the ability of their nanosensor array to discriminate between VOCs in the breath of healthy men and women and men and women with the four widespread forms of cancer itemized above.
They collected alveolar breath from 177 volunteers aged between 20 and 75 years of age. Breath from the patients with cancer was collected before they received any form of treatment.
The results of their study showed that:
The nanosensor array could differentiate between “healthy” and “cancerous” breath.
The nanosensor array could also discriminate between the breath of patients with different types of cancer.
The nanosensor array could distinguish between the breath patterns of patients with different cancers in the same statistical analysis– regardless of age, gender, lifestyle, and other confounding factors.
Each cancer appears to have a unique pattern of VOCs, when compared with healthy states, but not when compared with other cancer types.
Potentially, this is a very exciting finding. It suggests the possible ability to develop a non-invasive, easily-applied, portable system that might overcome several of the deficiencies associated with currently available methods for the diagnosis of cancer or at least the analysis of cancer risk. However, the data are limited and very early. The prostate cancer analysis was based on data from only 18 patients with prostate cancer, most of whom had early stage, localized disease.
Several things will need to be established if this research is to lead to a sound test for prostate cancer risk that has a high level of specificity and selectivity:
Can we identify with accuracy the VOCs that are specific for prostate cancer risk?
Do different types of VOC predict for aggressive as opposed to indolent types of cancer?
Is the use of VOCs any more accurate than the PSA test in predicting the presence of prostate cancer on biopsy?
Whether any or all of these can be done is pure speculation at the present time, but if dogs can be trained to identify men with prostate cancer by smell, it seems highly likely that we ought to be able to identify prostate cancer through the use of GC-MR and appropriately constructed nanosensor arrays.
We do wonder, however, just how much cancer has to be present before sufficient concentrations of the relevant VOCs are exhaled in the alveolar breath of patients at risk. The test certainly appears to be both sensitive and specific for four different types of cancer but can it be developed to be sensitive and specific enough for clinical application in the “real world”?