Sugar in saliva or blood could aid early detection

According to Dr. Daniel Bojar, associate senior lecturer in bioinformatics at the University of Gothenburg and lead author of this study, glycans are complex sugar molecules that are attached to proteins and fats in our bodies.

“They are intricate chains of different sugar units linked together in various ways,” Dr. Bojar explained to Medical News Today. “If those connections are changed, the function of the glycan changes. In cancer, several processes can change glycans.”

“Mutations in the tumor may change the proteins that build up these sugar chains, leading to altered glycans,” he continued. “Additionally, inflammation and various other systemic conditions that may accompany a tumor also have a known impact on which glycan structures are being produced.”

While scientists know that glycans are systematically changed in cancer, aiding the tumor in its development, Dr. Bojar said most of this knowledge is not based on rigorous statistical analysis, compared to standards in other scientific fields.

“We were convinced that we could gain more insights from these molecules with newly developed methods,” he continued. “As the most information-rich type of molecule in our bodies, it is a natural proposition for a data scientist to use this information for predictions.”

“Further, glycans are present on proteins and particles secreted by the tumor and can thus be found in convenient locations such as saliva or blood, which allows us to remotely monitor the tumor without having to perform an actual biopsy,” Dr. Bojar added.

For this study, Dr. Bojar and his team analyzed tumor and healthy tissue data from about 220 people with diagnosed cancers, focusing on gastric, skin, liver, prostate, colorectal, and ovarian cancers.

“Our main concern here was to select types of cancer for which high-quality glycan data was available so that our results would be more reliable,” he said when asked why they decided to focus on those types of cancer.

Using a new method of studying the substructures of glycan using AI, the scientists were able to identify differences in the substructure of the glycan depending on the type of cancer.

“Glycans are structurally very complex, much more so than proteins or DNA, and are best understood with advanced analytical methods such as AI,” Dr. Bojar explained. “Further, the way that glycans are currently measured — via mass spectrometry — typically leads to very heterogeneous data, including missing data points due to a lack of sensitivity. This has really hindered the field from making robust assessments from this type of data in many cases.”

“Methods such as AI allow us to improve the data quality, which we have shown in the paper describing this method, and this allows us to identify these relevant substructures with high statistical significance.”

– Dr. Daniel Bojar

While this study focused on specific types of cancer, Dr. Bojar said there is no reason to think this test would not translate to other types of cancer.

“Especially the glycans that we find to be present in most or all of the types of cancer we analyzed should also be present in other types of cancer,” he added.

Dr. Bojar also said a blood or saliva test developed through this research could possibly result in a faster detection of cancer in its early stages, although further research would be required to determine conclusively.

“Another benefit of this technique is that cancer can be monitored, since samples such as saliva or blood are minimally invasive, compared to biopsies for example,” he continued. “This could also extend to monitoring for re-emergence of a treated tumor, which could be rapidly done using the method we present.”

For the next steps in this research, Dr. Bojar said they plan to collect more cancer data, specifically from saliva samples of lung cancer patients, to extend and improve their results.

“In parallel, we plan to use the universal as well as specific cancer markers we have identified so far and develop tests for them, using specific glycan-binding proteins that allow for rapid and reliable measurement of patient state,” he continued. “This would then be also much cheaper than mass spectrometry and is only possible because we now know what we are looking for.”

“While it is hard to judge when such a test would be commonly available to patients, we are planning to validate these tests on clinical samples of patients within the next four to five years,” he added.

MNT also spoke with Dr. Richard Reitherman, a board-certified radiologist and medical director of breast imaging at MemorialCareBreast Center at Orange Coast Medical Center in Fountain Valley, CA, about this study. Dr. Reitherman was not involved in this research.

“When cancers are detected, they’re detected by multiple means,” Dr. Reitherman explained. “The blood usually carries the usual components of red blood cells, platelets, proteins, [and] white cells, but it also carries DNA, RNA, proteins, glycoproteins, glycans, [and] lipids that are not normal, and therefore they test to recognize those. They are shed from the cancer cells elsewhere in the body because the bloodstream happens to go everywhere.”

“It’s kind of like panning for gold,” he noted. “You have the river running down, but you have to put your dish in there with its proper sieve, and it has to be the right size and the right place in the river, or you’re never going to detect the gold nuggets. And once you detect the gold nuggets, you can start to develop strategies of therapy for a particular cancer at whatever stage it is, so this is really exciting.”

Dr. Reitherman also said that he was very excited about the use of AI in this study.

“AI is becoming increasingly essential in the analysis of metadata, which is what this is,” he told us. “These studies generate so many data points and metrics of what they’re measuring, whether it’s glycans or proteins, DNA, [and] RNA in the blood or the saliva, that we can no longer rely upon little spreadsheets where smart people can look and say, oh, this means this, and this means that.”