Covid-19: in the saliva the possible predictive identikit of the seriously ill patient

Discovered three new markers for Covid-19 related to severe or mild disease conditions. Thanks to an innovative analysis using Artificial Intelligence and the study of the salivary microbiota and metabolome, it will be possible to direct therapeutic choices in situations of difficult clinical prediction.

With the same symptoms, some people affected by Sars-CoV-2 need hospitalization while others can be treated at home. What distinguishes the two groups and how to identify them quickly to guide the therapeutic choices of doctors is the subject of a Humanitas research, published in Gastro Hep Advances, which describes a new method based on the analysis of saliva and blood.

Coordinator of the study is Prof. Maria Rescigno    head of the Laboratory of Mucosal Immunology and Microbiota of Humanitas and professor of General Pathology at Humanitas University, who with her team of researchers joined Dr. Antonio Voza – head of the Emergency Department of Humanitas, and Dr. Elena Azzolini – head of the Vaccine Center of Humanitas.

Faced with the difficulties of the first pandemic waves, when thousands of patients poured into the emergency room and there was still little knowledge about the course of the disease, the team of researchers used their expertise on the microbiota and mucous membranes to identify new markers of severity that would work early. Maria Rescigno and Chiara Pozzi, immunologist-researcher at Humanitas, focused on the microbiota of saliva and on the set of metabolites, i.e. products resulting from a chemical process related to digestion or ingestion of food.

“Through a retrospective study, we analyzed saliva and blood of hospitalized and home-treated patients to find what distinguished the two groups, comparing the data with those collected from healthy and cured subjects – explains Prof. Maria Rescigno. A machine learning approach was essential: our data scientists, led by Riccardo Levi, helped us to eliminate confounding parameters and the age factor, arriving to isolate two metabolites, Myoinositol and 2-pyrrolidinic acetic acid. These, together with a protein present in the blood (Chitinase 3-L1), have been shown to correlate with the severity of Covid, thus with the need or not for hospitalization.”

The combination of these 3 saliva and blood parameters describes the identikit of the critically ill patient and thus would be able to distinguish Covid patients based on the expectation of their clinical course.

“Subsequently, we saw that these two metabolites correlate with certain groups of bacteria in the salivary microbiota – continues Prof. Rescigno. Those who have altered metabolites also have altered bacteria. The result is not surprising: the microbiota has an important role in infection because it prepares the immune system and can have anti-microbial activities. And saliva, where part of the microbiota is located, is one of the points where the virus penetrates. It is also important to note that the protein detected in the blood is involved in the regulation of the ACE2 receptor, the Sars-CoV-2 virus receptor. This means that if the protein is already high to begin with, the person has more receptors and therefore could ‘let in’ more virus.”

The next step could therefore be the development of a diagnostic test, which is currently not available in the Analysis Laboratories. The methodology based on the analysis of metabolites – metabolomics – is a novelty that is emerging in the diagnostic landscape. A revolution speeded up by Covid-19, because during the pandemic it was possible to analyse data from a large number of patients very quickly.

“The results of this study give us hope – concludes Prof. Rescigno. In the future, it will be possible to design these tests based on salivary and blood tests also for other dangerous and difficult to predict diseases, such as sepsis”.