Periodic Reporting for period 1 - ArtIST (Pre-clinical evaluation and feasibility study for Artificial Inteligence-based Strain Typing and Bacterial Resistance analysis software)
Période du rapport: 2019-06-01 au 2019-11-30
An estimated 4.1 million people are affected by HAIs yearly in the EU, with an economic impact of €7 billion. Particularly worrisome is when these infections are caused by multi-resistant bacterial strains. Reducing the impact of these diseases is one of the Goals for Sustainable Development of the United Nations (UN).
The quick identification of the resistance mechanisms present in an infection is key to enable fast treatment with the right antibiotic, saving lives, reducing costs and preventing the disease from spreading.
MALDI-TOF Mass Spectrometry (MS) is a standard tool for the accurate, rapid and cost-efficient identification of pathogens. However, current routine-used systems give very limited to no information on the resistance mechanisms. Other technologies available are based on expensive and time-consuming.
The main goal of ArtIST is to identify, develop and commercialise specific solutions for the rapid and cost-efficient identification of resistance mechanisms in bacteria in clinically relevant subspecies. Our project contributes to the implementation of the EU One Health Action Plan against Microbial Resistance.
During the action involved in the SME Instrument Phase 1 grant, we have performed a feasibility study to prepare for clinical validation of our solutions. The main milestones achieved during this action are:
1) Very promising identification of top clinically relevant resistance mechanisms, such as Vancomycin-resistant Enterococcus, Methicillin-resistant Staphylococcus aureus and mechanisms against beta-lactam antibiotics.
2) Design of easily integrable protocols, being able to read different vendor data and simplifying the sample preparation protocols to minimum impact.
3) Increasing our presence in the microbiology field. Papers have been submitted to international peer-reviewed scientific publications and our work has been presented in international conferences.
4) Our software UI have undergone major changes to improve usability by microbiology users, not usually expert in Artificial Intelligence algorithms. Algorithms for identification of blind samples have been included.
5) New collaborating hospitals have been included to our network, and we have started initial steps to enter regulatory processes and CE marking.
To conclude, the feasibility study has proven that our solution shows good results and is ready to be clinically validated. Also, we have confirmed the marketability and scalability of our business model.
The quick identification of the resistance mechanisms present in an infection is key to enable fast treatment with the right antibiotic, saving lives, reducing costs and preventing the disease from spreading.
MALDI-TOF Mass Spectrometry (MS) is a standard tool for the accurate, rapid and cost-efficient identification of pathogens. However, current routine-used systems give very limited to no information on the resistance mechanisms. Other technologies available are based on expensive and time-consuming.
The main goal of ArtIST is to identify, develop and commercialise specific solutions for the rapid and cost-efficient identification of resistance mechanisms in bacteria in clinically relevant subspecies. Our project contributes to the implementation of the EU One Health Action Plan against Microbial Resistance.
During the action involved in the SME Instrument Phase 1 grant, we have performed a feasibility study to prepare for clinical validation of our solutions. The main milestones achieved during this action are:
1) Very promising identification of top clinically relevant resistance mechanisms, such as Vancomycin-resistant Enterococcus, Methicillin-resistant Staphylococcus aureus and mechanisms against beta-lactam antibiotics.
2) Design of easily integrable protocols, being able to read different vendor data and simplifying the sample preparation protocols to minimum impact.
3) Increasing our presence in the microbiology field. Papers have been submitted to international peer-reviewed scientific publications and our work has been presented in international conferences.
4) Our software UI have undergone major changes to improve usability by microbiology users, not usually expert in Artificial Intelligence algorithms. Algorithms for identification of blind samples have been included.
5) New collaborating hospitals have been included to our network, and we have started initial steps to enter regulatory processes and CE marking.
To conclude, the feasibility study has proven that our solution shows good results and is ready to be clinically validated. Also, we have confirmed the marketability and scalability of our business model.
The EU grant has helped us pushing forward our collaboration with hospitals and finding the group of bacterial pathogens where we need to focus to have the greatest clinical impact in terms of identifying resistance markers. These are known as the ESKAPE group: Escherichia coli, Staphylococcus aureus, Klebsiella pneumoniae, Pseudomonas aeruginosa, Enterococcus faecium/faecalis. They are responsible for two thirds of all healthcare associated infections .
On the one hand, our RUO platofrm, targetting microbiology and mass spectrometry research labs, is commercialized (TRL 9), and during this year we have extended its user base to more than 40 users from more than 25 institutions (see Figure 1). Main updates on the software are:
- Simplified data preparation workflow.
- Improved usability and language close to microbiology experts.
- New methods and modules (Heatmap, reproducibility analysis, FTIR data, SVM, PLS-DA cross validation and more).
- Sample identification and prediction, allowing building customised detection methods.
- Social features, allowing sharing research results with your contacts within or external to your own organisation.
More importantly, our clinical-IVD version are already in TRL 6-7 with some of the main resistance mechanisms in the clinically relevant ESKAPE group, with technology demonstrated with real samples in hospitals and moving towards demonstration in real clinical settings integrated with the laboratory workflow and multi-centre studies. Several publications have been submitted to international journals and relevant conferences. Figure 1 shows
After the feasibility study, our company is now ready to formally start clinical evaluation and regulatory phases. We have identified the most severe pathogen strains causing most of the infections for which we have a potentially good biomarker identified. The next step will be then to obtain both analytical and clinical validation for our method. Analytical validation is being done together with our collaborator in the Medical University of Vienna. Clinical validation will be done by multicentre studies coordinated by our collaborating hospitals and under the umbrella of the formal agreements we are signing with them currently. After validation, we will implement our quality management system and obtain certification ISO 13485 and probably 9001. This will lead to the corresponding notified body audits to obtain compliance with European IVDR. We plan to be in the clinical market during 2022. In the meantime, we’ll also use the experience to obtain USA FDA approval to be open to other markets besides Europe.
During the project we have also performed market study and optimise our market entry and commercialisation strategy. The project allowed us as well to attract talent, as we hired two Applications Specialists and partnered with a sales executive consultant.
On the one hand, our RUO platofrm, targetting microbiology and mass spectrometry research labs, is commercialized (TRL 9), and during this year we have extended its user base to more than 40 users from more than 25 institutions (see Figure 1). Main updates on the software are:
- Simplified data preparation workflow.
- Improved usability and language close to microbiology experts.
- New methods and modules (Heatmap, reproducibility analysis, FTIR data, SVM, PLS-DA cross validation and more).
- Sample identification and prediction, allowing building customised detection methods.
- Social features, allowing sharing research results with your contacts within or external to your own organisation.
More importantly, our clinical-IVD version are already in TRL 6-7 with some of the main resistance mechanisms in the clinically relevant ESKAPE group, with technology demonstrated with real samples in hospitals and moving towards demonstration in real clinical settings integrated with the laboratory workflow and multi-centre studies. Several publications have been submitted to international journals and relevant conferences. Figure 1 shows
After the feasibility study, our company is now ready to formally start clinical evaluation and regulatory phases. We have identified the most severe pathogen strains causing most of the infections for which we have a potentially good biomarker identified. The next step will be then to obtain both analytical and clinical validation for our method. Analytical validation is being done together with our collaborator in the Medical University of Vienna. Clinical validation will be done by multicentre studies coordinated by our collaborating hospitals and under the umbrella of the formal agreements we are signing with them currently. After validation, we will implement our quality management system and obtain certification ISO 13485 and probably 9001. This will lead to the corresponding notified body audits to obtain compliance with European IVDR. We plan to be in the clinical market during 2022. In the meantime, we’ll also use the experience to obtain USA FDA approval to be open to other markets besides Europe.
During the project we have also performed market study and optimise our market entry and commercialisation strategy. The project allowed us as well to attract talent, as we hired two Applications Specialists and partnered with a sales executive consultant.
During the development of the project, and together with our collaborations with hospitals, we have made very significant progress in many of the pathogens in the ESKAPE group, which are responsible for two-thirds of every health-care associated infections. Some of these results have been submitted to publication in scientific journals (see Fig. 2). Others, like our work with subspecies identification of E. coli or resistant mechanisms in Klebsiella pneumoniae and Enterococcus faecium has been submitted as abstract to the Annual Conference of the European ESCMID.
Our target is to take at least two of these solutions to the IVD market for 2022. To achieve this, we are starting more advanced clinical evaluation and we are starting the appropriate regulatory and certification processes. This is very important becuase antimicrobial resistance is an increasing burden on the health-care systems, and the threat is becoming now a first-priority issue for our society, with more than 33000 deaths per year. Moreover, it is estimated that resistance to antibiotics will grow in a range from 22% in western economies to 82% in developing countries, to and estimated 10 millions deaths in 2050.
Besides, no new major classes of antibiotics have been approved in many years to treat common, deadly Gram-negative infections. Preventing the infection in the first place, stopping the spread, and rapid diagnosis and action against detected cases is our best change to prevent patients to be harmed. Identifying resistance allows healthcare providers to promptly use effective antibiotics and to prevent spread. Diagnostics can be just as critical for fighting infections as antibiotics . Nowadays, most current antibiotic resistance identification technologies are either:
• Effective but costly and time-consuming
• Do not detect emerging resistance markers
• Do not rapidly discriminate bacterial or fungal infections
Our work will help to place improved diagnostic tools which are fast, cost-efficient, fully integrated with the laboratory workflow and vendor-agnostic.
Our target is to take at least two of these solutions to the IVD market for 2022. To achieve this, we are starting more advanced clinical evaluation and we are starting the appropriate regulatory and certification processes. This is very important becuase antimicrobial resistance is an increasing burden on the health-care systems, and the threat is becoming now a first-priority issue for our society, with more than 33000 deaths per year. Moreover, it is estimated that resistance to antibiotics will grow in a range from 22% in western economies to 82% in developing countries, to and estimated 10 millions deaths in 2050.
Besides, no new major classes of antibiotics have been approved in many years to treat common, deadly Gram-negative infections. Preventing the infection in the first place, stopping the spread, and rapid diagnosis and action against detected cases is our best change to prevent patients to be harmed. Identifying resistance allows healthcare providers to promptly use effective antibiotics and to prevent spread. Diagnostics can be just as critical for fighting infections as antibiotics . Nowadays, most current antibiotic resistance identification technologies are either:
• Effective but costly and time-consuming
• Do not detect emerging resistance markers
• Do not rapidly discriminate bacterial or fungal infections
Our work will help to place improved diagnostic tools which are fast, cost-efficient, fully integrated with the laboratory workflow and vendor-agnostic.