Adverts: 0161 709 4576 - Editorial: 0161 709 4571
Mail Order: 0161 709 4578 - Subs: 0161 709 4575 - Webteam: 0161 709 4567
A dog ‘nose’ his technology

SMELL IS AN important indicator of well- being, and has been used as a diagnostic tool since ancient times. The Romans gave the distinctive odour of renal failure its own term — ‘fetor hepaticus’ — while even today many diseases are known to have a characteristic smell, such as a "sweet acetone breath" for diabetes, and "putrid breath" for streptococcal throat infections.

But now this principle is being harnessed with the latest technology to provide a sophisticated way of diagnosing disease. In one advance, researchers at the University of Rome have designed an electronic nose – called the "e-nose" - for detecting lung cancer in patients. In recent trials at the Forlanini Hospital, the e-nose correctly identified patients with lung cancer tumours simply by smelling their breath.

E-noses have been employed by the food industry for about ten years and are used for many tasks, from detecting e.coli in contaminated food to verifying that orange juice is "natural" (important for food trading standards purposes as well as for the food buyers).

The "nose" is actually a chemical sensing system — a spectrometer — and a computer system that matches patterns of smells. The technology is pretty much the same whether you are detecting lung cancer or sniffing for salmonella. Each system is "trained" to pick up certain chemical signatures, usually a complex combination of smells. People with lung cancer tend to exhale a mixture of alkanes and benzene derivatives — the University of Rome e-nose has been calibrated to pick up these chemical signatures.

The e-nose is non-invasive, quick and apparently accurate (the lung cancer patients were diagnosed in a minute). The technology is based on the olfactory model of dogs, known for their acute sense of smell (the human nose is too blunt an instrument).

In a case reported in The Lancet some years ago, a dog was said to have "discovered" a melanoma on her owner’s leg. The Border Collie/Dobermann cross kept on sniffing and even biting the mole, so finally her owner went to the doctor to have it looked at — a biopsy confirmed that the mole was in fact a malignant melanoma. "This dog may have saved her owner’s life by prompting her to seek treatment when the lesion was still at a thin and curable stage," wrote the researchers.


While it may not be practical to have a canine cancer-sniffer in every dermatology clinic, the fact that melanomas can be detected through smell means that one day an e-nose could be adapted to do the same job.

Professor Donald Broom, of the department of clinical veterinary medicine at the University of Cambridge, is currently looking to test the viability of using dogs to sniff out prostate cancer in urine samples. Again, if dogs can detect prostate cancer, then so could an electronic nose. The advantage that e-noses have over dogs is that they don’t get tired (sniffer dogs can suffer from smell "burnout’’ if they are kept working on a smell for too long).

Electronic noses could be used as a diagnostic tool for other pathologies, too, such as diabetes, liver cirrhosis and renal failure — each has its own chemical signature.

The team at the University of Rome is also studying the correlation between schizophrenia and skin odour. It was nearly 30 years ago that scientists first discovered that schizophrenic patients have a specific skin odour. Now that the technology is available, this information can be put to use.

According to Professor Corrado Di Natale, who is leading the study: "We have been measuring axilla "armpits" odour in patients and results are encouraging."

He envisages the e-nose being developed to screen smokers and other high-risk groups for lung cancer as part of a routine check-up. Of course, it may not replace the precision offered by scans and blood tests, but it could be a useful tool for early detection.

In spite of the promising applications for e-nose, Di Natale admits that the device has some limitations. Because the quantities of alkanes and benzene derivatives occur at such low levels in human breath, the sensors — an array of quartz crystals with a special coating — need to be finely calibrated to prevent false results. Furthermore, the response of the sensors fluctuates with time, reducing the accuracy of the findings. Di Natale warns that the device still has some way to go before it is "fully optimised".

But even at this stage — with a correct diagnosis in one minute — the e-nose future looks promising.

For information on e-noses: