In a recent paper published in Proceedings of the Royal Society, researchers trained ants to become olfactory cancer tumour bio-sensors. Quite literally, the ants could sniff out cancer– and researchers just had to follow the ants to find the tumour.
Sniffing out cancer
Cancer is a leading cause of mortality worldwide, and early detection is an effective way to improve the cancer survival rate. However, most early diagnosis methods are often invasive (e.g. colonoscopy) or expensive (MRI) making them inaccessible to many people. A cheaper alternative may be the use of animals which have been trained to detect cancer via scent.
Dogs are leading the pack in the use of animal olfaction for early cancer detection by recognising volatile organic compounds (VOCs) produced by cancer cells. Other species, including mice, nematodes and fruit flies, have also shown the ability to detect cancer through scent. Other species have been less successful. For instance, honeybees, despite using a well-known olfactory conditioning paradigm based on the proboscis extension response, were not successful in detecting cancer through scent.
But what about other insects?
Insects are attractive as a detection tool as they are relatively easy to handle, are available in large numbers, are not expensive to look after and care for, and can be trained to recognise a specific odour in very few trials. And among insects, ants (specifically the species Formica fusca) have previously demonstrated an exceptional ability to learn odours – one training trial was enough to remember the scent long-term, lasting days. The same team of researchers showed in a previous study that ants could correctly sniff out a cancerous cell line from a healthy cell line, as well as tell the difference between two different types of cancerous cell lines (ovarian and breast cancer cell lines).
Follow the ants… and find the tumour
In this study, researchers used urine samples from patient-derived xenograft mice (PDX) carrying human tumours as the olfactory stimuli. Compared to cell lines (which grow in a stable, controlled environment), PDX mice represent “real” tumours better, as the tumours are growing within a live organism – an inherently complex environment. The PDX mice carried a “triple negative” breast tumours – a particularly aggressive subtype which leads to poor outcomes.
The researchers used a fairly simple conditioning paradigm to train and test whether ants could discriminate between PDX mice urine and healthy mice urine. To characterise changes in VOC composition in the urine due to the presence of cancer, they used solid-phase micro-extraction and gas chromatography coupled with mass spectrometry (GC-MS). They trained individual ants to associate the odour of mouse urine with a reward of 30% sugar solution. During three conditioning trials, the time it took the ant to locate the reward was measured.
The ants are marching towards… the clinic?
The findings of this study showed that not only did ants reliably distinguish between tumour-bearing and tumour-free mice, they could also detect volatile organic compounds (VOCs) at low concentrations. Furthermore, after 15 minutes of completing the conditioning test, a single ant took around only 37 minutes to discriminate between tumour and tumour-free samples, and only 24 ants were needed to reveal a statistical difference. Moreover, Formica fusca ants are very resistant to memory extinction, and after training they can be tested up to nine times without any reward before their responses start to decline.
Ants could be a potentially cost-effective and non-invasive tool for early tumour diagnoses in humans. However, further research using human-origin tumours of various types is needed before ants can be considered a routine cancer screening method. Additionally, some patients might not like the idea of getting ants in their pants during the annual health check-up… but if these findings are substantiated, the benefits of using ants as a diagnostic tool could be astounding.