The smell of a tumor

Is it possible to smell hidden cancer and what does cyborg insects have to do with it

Oleg Lischuk, N+1

Medicine has the tools to fight almost all diseases. We identify and treat cardiovascular diseases, diabetes mellitus, infections and much more at the population level. The techniques and technologies that we use are still very far from perfect, but they are there. With a timely diagnosis of cancer, everything is much worse. Existing screening methods require the collection of biomaterials, are complex, expensive and limited in their capabilities. Therefore, while some scientists are improving them, others are looking for completely new ways of diagnosis — and sometimes come back from their searches with rather unusual ideas. Recently, American researchers used the brain and antennae of live locusts to diagnose tumors by smell. The article on this has not yet been reviewed and is available only as a preprint (Farnum et al., Harnessing insect olfactory neural circuits for noninvasive detection of human cancer). In its current form, the technology is, of course, unsuitable for clinical use, but its prospects are worth understanding. In addition, this is not the first attempt to "sniff out" cancer (of course, potential non-invasive cancer screening!), and a lot of interesting things have already accumulated in this area.

Diagnosis of diseases by smell has been known at least since ancient Greece. Some conditions are accompanied by the release of characteristic volatile compounds, distinguishable by the human nose. For example, diabetic ketoacidosis smells of fruit or acetone, liver failure smells of stale fish, kidney failure smells of urine, and anaerobic lung abscess smells of sewage. However, all this concerns quite severe, neglected diseases. There is no question of early detection here.

Sum of technologies

In the XX century, the development of ideas about the metabolism in the body, sample collection technologies and high-precision methods of chemical analysis (primarily mass spectrometry), naturally, pushed researchers to create a method of non-invasive diagnosis of diseases by human exhaled air.

The first significant attempt was made in 1971 by the Stanford laboratory of Nobel laureate Linus Pauling. Using gas-liquid chromatography, the scientists conducted a quantitative assessment of about 250 substances in the breath and about 280 in the urine vapors of patients. The subjects had to prepare for tests for several days, adhering to a strict diet, and complex and expensive equipment was used to work with their breathing and urine vapors. The technique was not suitable for mass use. In addition, olfactory profiles of specific diseases were not developed then.

The research, however, continued. Now all of them are referred to a special section of metabolomics, which is called "breathomics" (from the English breath — "breath"), or "volatolomics" (volatolomics, from the English volatile compounds — "volatile compounds") — these names do not have a well-established translation into Russian yet. The very concept of this discipline is based on the idea that all cells of the body, both normal and pathologically altered (especially malignant ones, which extract energy not by glycolysis, but by oxidative phosphorylation), in the process of metabolism secrete a certain set of low molecular weight volatile organic compounds (VOCs) peculiar only to them. These substances enter the bloodstream and in small but detectable concentrations (from billionths to trillionths of volume parts) are excreted with exhaled air, and are also present in sweat, saliva, urine and feces.

The place of respiratory metabolomics among other "omics". Maryam Khoubnasabjafari et al. / Critical Reviews in Analytical Chemistry, 2021

However, in addition to these endogenous VOCs, exogenous impurities are contained in human respiration: from food, beverages, certain medicines, tobacco and atmospheric chemicals, as well as products of microbiome metabolism (primarily oral cavity, respiratory tract and gastrointestinal tract). The total number of known "breathing" VOCs is already more than 3000 (read about the number of odors and difficulties with determining the basic dimensions of the olfactory space in the material "Phenomenology of the Spirit"). All this significantly complicates the determination of the odors of individual diseases.

Nevertheless, these techniques already exist and continue to be developed. Their implementation consists of three stages.

• At the first stage, samples for analysis are collected in chemically inert containers and concentrated using condensation chambers, cryoprecipitation, thermal sorption tubes, solid-phase microextraction or needle traps.
• At the second stage, their chemical analysis is carried out by various types of chromatography/mass spectrometry, as well as nuclear magnetic resonance, fluorescence, terahertz spectroscopy and "electronic noses" based on sensor nanoparticles, which are gaining popularity.
• At the third stage, the "smells" of individual diseases are isolated from the totality of the detected VOCs using statistical algorithms — and a conclusion is drawn up.

Instrumental diagnostics by smell. Oluwasola Lawal et al. / Metabolomics, 2017

Today, predicting the smell of a tumor, knowing its metabolism, is an extremely difficult task. However, leaving aside the difficult task of modeling odor, in practice it is possible to follow a purely empirical path: creating in the future replenished and updated databases with VOC profiles of specific tumors by analogy with databases of genetic cancer markers.

It is obvious that the existing methods require separate laboratories, are complex, expensive and are still far from being widely used for screening, although they undoubtedly have prospects for development.

Live detectors

While some researchers are dealing with the instrumental difficulties of such diagnostics, others are trying to use natural "detectors" — animal noses, which have already proven themselves well in the search for explosives and narcotic substances.

For the first time, the staff of the British Royal College Hospital Hywel Williams (Hywel Williams) and Andres Pembroke (Andres Pembroke) thought about it. In 1989, they described a case in which a dog constantly sniffed and tried to bite a mole on the leg of a 44-year—old hostess - and thus helped to identify the early stage of melanoma.

Since then, various research teams have been trying to use dogs to diagnose lung cancer, breast, prostate, intestines, skin pigment cells and other malignant tumors. In addition to oncological diseases, hypoglycemia with an overdose of insulin, as well as tuberculosis, malaria, covid and other infections came to the attention of "medical trainers". Optimism was added by the fact that the animals in such experiments did not seem to be disturbed by the background smells in the room.

Over time, however, there were also disadvantages in canine diagnostics. As the data accumulated, it became clear that the method has low sensitivity, experiments sin with the non-reproducibility of the results (read the article "The Dog ate the protocol" about the problems of reproducibility in oncobiology), the degree of false-positive diagnoses by the nose is great, and in general the technique is not effective enough. It even came to lawsuits against overly optimistic entrepreneurs who rushed to commercialize it.

If we also take into account the fact that breeding, training and keeping dogs are very expensive, it becomes clear that without serious technological breakthroughs (for example, the development of unified training protocols that are maximally protected from errors, the use of genetically modified animals), the prospects for using the method are extremely limited.

Cyborgs Exit

Nevertheless, the diagnosis of diseases with the help of "live instruments" has not ceased to interest researchers. A new work by American scientists, however, describes a very exotic solution in an already young and non-trivial research field.

First, they decided to turn instead of "standard" mammals to much simpler creatures — insects. They are unpretentious, reproduce quickly and without problems in artificial conditions and, most importantly, have a remarkable sense of smell. However, in order for it to work in the way necessary for scientists, they had to cybernetically improve their wards.

The staff of the University of Michigan used adult (after the fifth molt) individuals of the North American locust Schistocerca americana, whose odor recognition system is well studied (strictly speaking, "locust" in everyday life is called a variety of insects, including Schistocerca Americana, but in general this species does not swarm and does not migrate, so it is more correct to call them fillies). These insects pick up chemical stimuli by olfactory receptor neurons in the antennae. Using a combinatorial coding scheme, ⁵⁰ of these neurons are able to recognize up to 2,50 individual VOCs — which gives a total of several trillion odors. Signals from the antennae enter the olfactory lobes of the brain, where they are processed by a network of projection neurons.

The insects were fixed on a surgical platform, their antennae were fixed and direct access to their brains was obtained to place 16-channel silicon electrodes in the olfactory lobes. The electrodes recorded the activity of individual projection neurons and sent this data to recording equipment.

In the experiment, cyborgized insects "sniffed" cells of three types of oral cancer (Ca9-22, HSC-3 and SAS) and healthy HaCaT epithelium grown on the same nutrient media.

At the training stage, VOC cell cultures were collected and fed to the fillies' antennae using a standard olfactometer. At the same time, the researchers recorded and statistically processed the spatio-temporal picture of the excitation of neurons of the olfactory lobes in response to each smell.

The principle of operation of the technology. Alexander Farnum et al. / bioRxiv, 2022

The revealed patterns turned out to be distinct and individual (p<0.05). With their help in the experiment, it was possible to distinguish all types of cancer from each other, from healthy cells, from a pure nutrient medium and control VOCs (undecane and hexanal) in 100 percent of cases. It took only 250 milliseconds to detect the smell.

The reaction of individual projection neurons of locusts to different odors. Alexander Farnum et al. / bioRxiv, 2022

Additional experiments have shown that the system remains effective at different stages of the life cycle of cancer cells, when the chemical composition of the environment changes under the influence of their metabolism.

A stable picture of the reaction of neurons in changing environmental conditions. Alexander Farnum et al. / bioRxiv, 2022

 "In this work, we have demonstrated the technological feasibility and functional stability of such an approach for non—invasive cancer diagnosis," the authors conclude.

Locusts for the doctor

In the current state, cyborg locusts are not yet ready, of course, for clinical use, or even clinical trials. During the experiment, in order to obtain a clear signal, the activity of 40 neurons was required to be recorded. Taking into account the use of 16-channel sensors, the researchers had to simultaneously engage from 6 to 10 insects, maintaining the viability of the brain of each of them, and combine their reactions to the smell to obtain a complete picture. This is an excessively complex and cumbersome design.

The main advantage of the technology is that such a sensor captures the smell of a particular tumor entirely, regardless of its specific chemical components. The neuronal "fingerprint" recorded in this case is quite distinct, unique and, as experiments have shown, can "shine through" through the noise of other olfactory signals. In this case, the result is given immediately.

But there are still a lot of unresolved problems. And not only technological. For example, if a cyborg locust - be it fillies, grasshoppers, crickets or representatives of real locusts — is turned into a full—fledged diagnostic system, difficulties will inevitably arise during its legal registration. How to classify it? There is no such type of sensor as "insect brain" in any national register of medical devices. In addition, it is a living biomaterial, the handling of which is subject to special regulation. In order to overcome such purely bureaucratic obstacles, the methodology will have to demonstrate exceptional screening (and commercial) potential — because for its application it will have to make edits to a rather massive body of legal documentation in the field of healthcare.

Ethical questions will also inevitably arise, because technology literally uses the brain of a living being as a consumable. However, scientists have something to say in her defense. Firstly, we still do not know for sure whether insects feel pain and whether it can be described precisely as suffering. The threat of extinction of locusts, to put it mildly, is not threatened — even Schistocerca americana periodically cause significant damage to the fields of Florida — pesticides and biological methods are used to combat these insects. Moreover, people have been eating insects for years and are used for a variety of experiments — from school biology lessons to academic and commercial research. In general, it's time to get tangible benefits from pests.

So far, the authors continue to improve the technology. Now they are working to increase its resolution, using high-precision electrodes, so that the brain of one insect is enough for diagnosis. In parallel, a portable platform is being developed, in which only the brain of a locust in a nutrient medium and its antennae serve as a sensor. In this form, the technique may be suitable for clinical trials.

So it is possible that in the foreseeable future, something like a "hybrid bioelectronic olfactoanalytic mobile system for screening diagnostics of certain types of oncological diseases in various age groups within the framework of medical examination programs" will appear in the arsenal of physicians. Translating into human language, this means that at a doctor's appointment you will be able to breathe into the device and immediately find out if it's time to see an oncologist. Before it's too late.

Portal "Eternal youth" http://vechnayamolodost.ru