Periodic Reporting for period 1 - PreBiomicsPMT (A metagenomic-based precision-medicine tool for personalized diagnosis, prognosis and treatment of oral diseases)
Reporting period: 2023-05-01 to 2024-04-30
Every year, around 20 million dental implant procedures are performed worldwide to replace one or more teeth with “artificial roots” made of titanium. Unfortunately, dental implants are also exposed to specific diseases caused by dental plaque microorganisms. More than 50% of patients with dental implants develop a condition called mucositis, an inflammation of the soft tissues surrounding the implant. If left untreated, mucositis can progress to peri-implantitis, a severe infection that affects both the soft and hard tissues, potentially leading to implant failure. Currently, there are no reliable tools for early detection of these diseases, assessing their severity or predicting their progression, let alone treating them by specifically targeting the causing agents. To date, dentists can only provide general treatments without knowing the specific composition of the plaque microbiome triggering and sustaining the disease.
Over 10 years of research, PreBiomics has identified a set of bacteria whose presence and abundance are crucial in predicting the severity and progression of implant infections. Exploiting metagenomics and AI, we are developing a precision-medicine tool called PreBiomicsPMT that analyzes the microbial composition of the plaque collected around dental implants to predict disease progression and provide personalized treatment recommendations for each patient.
This EIC-Transition project aims to a) automate the testing process using robotics to help reduce costs and speed up results delivery; b) create a comprehensive database to enhance our decision support system (DSS) for dentists; c) introduce and validate an antibiotic resistance gene identification feature to support dentists performing informed therapeutic choices; d) validate the DSS ability in predicting disease and treatment outcomes; d) build a solid business and exploitation plan for our product.
Over 10 years of research, PreBiomics has identified a set of bacteria whose presence and abundance are crucial in predicting the severity and progression of implant infections. Exploiting metagenomics and AI, we are developing a precision-medicine tool called PreBiomicsPMT that analyzes the microbial composition of the plaque collected around dental implants to predict disease progression and provide personalized treatment recommendations for each patient.
This EIC-Transition project aims to a) automate the testing process using robotics to help reduce costs and speed up results delivery; b) create a comprehensive database to enhance our decision support system (DSS) for dentists; c) introduce and validate an antibiotic resistance gene identification feature to support dentists performing informed therapeutic choices; d) validate the DSS ability in predicting disease and treatment outcomes; d) build a solid business and exploitation plan for our product.
During the project's first year, we fully implemented the automatization of sample processing through robotic solutions and optimized the laboratory procedures by reducing reagent volumes and tailoring different sample processing steps. These improvements have led to a significant cost reduction (32% in consumables and 80% in personnel cost per sample) and improved standardisation of the results. The current time for sample processing from reception to sequencing library preparation has decreased from 10 to 2 working days.
As an additional feature of great interest to our dentist customers, we have moreover curated a panel of resistance genes against antibiotics and antiseptics routinely used in dental practice, with the final aim of providing professionals with information regarding the presence and abundance of resistance genes that may hinder antibiotic or antiseptic therapy efficacy. By combining existing software with our curated resistance genes panel, we developed a computational workflow to identify and quantify the genes of interest in plaque microbiome samples and contextualize such resistance levels with respect to those found in samples of the PreBiomics proprietary database, which includes both healthy and diseased samples. This information is then summarised in a beta version of the report, providing the dentist with useful information for personalized antibiotic or antiseptic selection. The antibiotic resistance identification tool has been validated in-vitro by testing its ability to recapitulate the antibiotic resistances of a given strain by comparing the tool’s results with those of bacteria grown on a plate and subjected to antibiotic pressure.
We have started the expansion of our proprietary database with new samples provided by the growing number of clinics that decided to participate in the study. The expanded database will be crucial for the full development of the DSS and its validation on real-life data. The PreBiomics bioinformatic team started developing the machine learning algorithm providing information on the risk of peri-implant disease progression based on the plaque microbiome composition (microbiome score) and certain clinical data (clinical score). The algorithm will then be expanded to include suggestions for the best treatment according to the extensive PreBiomics database and the treatment outcomes of the PreBiomicsPMT project.
As an additional feature of great interest to our dentist customers, we have moreover curated a panel of resistance genes against antibiotics and antiseptics routinely used in dental practice, with the final aim of providing professionals with information regarding the presence and abundance of resistance genes that may hinder antibiotic or antiseptic therapy efficacy. By combining existing software with our curated resistance genes panel, we developed a computational workflow to identify and quantify the genes of interest in plaque microbiome samples and contextualize such resistance levels with respect to those found in samples of the PreBiomics proprietary database, which includes both healthy and diseased samples. This information is then summarised in a beta version of the report, providing the dentist with useful information for personalized antibiotic or antiseptic selection. The antibiotic resistance identification tool has been validated in-vitro by testing its ability to recapitulate the antibiotic resistances of a given strain by comparing the tool’s results with those of bacteria grown on a plate and subjected to antibiotic pressure.
We have started the expansion of our proprietary database with new samples provided by the growing number of clinics that decided to participate in the study. The expanded database will be crucial for the full development of the DSS and its validation on real-life data. The PreBiomics bioinformatic team started developing the machine learning algorithm providing information on the risk of peri-implant disease progression based on the plaque microbiome composition (microbiome score) and certain clinical data (clinical score). The algorithm will then be expanded to include suggestions for the best treatment according to the extensive PreBiomics database and the treatment outcomes of the PreBiomicsPMT project.
In the next two years of the project, we will focus on expanding our current proprietary database with 1,500 new samples collected by our cohort of dental clinics involved in the PreBiomicsPMT study. Newly sequenced microbiome data will be integrated and harmonized with previous data. The DSS will be trained on this extended microbiome database, baseline clinical information, treatment details (as independently chosen by the dentist), and the clinical outcomes of such treatment. Once the DSS is fully functional, we will start its validation on an additional 1,500 newly collected microbiome samples from both peri-implantitis and mucositis patients. The validation of the DSS will focus on proving its predictive power in real-life settings.
Moreover, we will further validate the antibiotic resistance detection software by exploiting synthetic metagenomes. We will bioinformatically build metagenomes starting from genomes of known bacteria usually found in the plaque microbiome, and we will add a test bacterial species genome carrying known resistance genes at decreasing relative abundance to test the limit of detection of our method. This will help us optimize the sequencing depth for our plaque microbiome samples to account for the software limitations and offer the best value-for-money test to clients.
We will test product interest in our reference markets by conducting independent surveys for dental professionals and patients and further collaborating with our partner Geistlich, which is already well-established in all three countries. We will expand collaboration with IPR consultants to ensure the best protection for our innovative products, and with regulatory consultants to ensure adherence to ethical, legal, and administrative guidelines and rules.
Moreover, we will further validate the antibiotic resistance detection software by exploiting synthetic metagenomes. We will bioinformatically build metagenomes starting from genomes of known bacteria usually found in the plaque microbiome, and we will add a test bacterial species genome carrying known resistance genes at decreasing relative abundance to test the limit of detection of our method. This will help us optimize the sequencing depth for our plaque microbiome samples to account for the software limitations and offer the best value-for-money test to clients.
We will test product interest in our reference markets by conducting independent surveys for dental professionals and patients and further collaborating with our partner Geistlich, which is already well-established in all three countries. We will expand collaboration with IPR consultants to ensure the best protection for our innovative products, and with regulatory consultants to ensure adherence to ethical, legal, and administrative guidelines and rules.