Periodic Reporting for period 5 - RAIL (Radiomics of lung cancer (RAIL): non-invasive stratification of tumour heterogeneity for personalised cancer therapy)
Reporting period: 2019-06-01 to 2020-03-31
Lung cancer is the most common cause of death from cancer worldwide with 1.59 million deaths annually. It also places the highest economic burden of all cancers on the EU with EUR 18.8 billion. Non-small cell lung cancer (NSCLC) comprises 85% of all lung cancer cases. The current clinical routine to guide treatment for NSCLC patients is primarily based on the TNM staging system founded in 1958, mainly to look at surgical operability. However, TNM-based evaluation, nor doctors’ prediction do not provide an accurate prognosis. The heterogeneity between patients (inter-patient), tumours (inter-tumour) and even within one tumour (intra-tumour) underlies these highly variable prognoses and presents a major clinical challenge. There is a lack of easy to perform, ‘actionable’ biomarkers to stratify NSCLC patients which results in i) undertreatment in 20% of NSCLC patients leading to disease progression and ultimately death, ii) over-treatment in 30% of NSCLC patients which reduces the quality of life of patients and places an economic burden on the healthcare system and iii) ineffective clinical trial design, due to lack of optimal stratification, which requires very large, costly clinical trials to be performed in order to bring new therapeutic strategies to the market. ptTheragnostic has developed breakthrough technology, called “Radiomics”, which was recently published in Nature Communications, which enables patient stratification through the use of imaging biomarkers acquired from routine CT & PET imaging (animation: https://youtu.be/Tq980GEVP0Y).
In Radiomics for lung cancer (RAIL) we will validate and qualify imaging biomarkers together with a ready-to-use application to deliver more accurate prognostic information, personalise treatment for NSCLC patients, reduce healthcare costs by EUR 500+ million and enable efficient clinical trial design. During the project we will implement the clinical benefits together with a CE-certified computer-aided prognostic and predictive application that converts medical images into a clear and concise report based on Radiomic signatures to support treatment decisions. Radiomic imaging biomarkers will be validated based on the highest standards in clinical validation (level I) in a phase II multi-centre clinical study. Based on these needs, we have formulated the Radiomics for Lung cancer (RAIL) project with the following key objective:
“To validate and qualify a multi-site level I imaging biomarker (RIB1) together with a prognostic application (RadioCAP) to deliver a better prognosis and personalise treatment for NSCLC patients.”
In order to achieve our main goal, we have implemented the following sub-objectives:
1. To develop a radiomics-based prognostic application (WP1)
2. To validate clinical utility of RadioCAP and RIB1 (WP2)
3. To validate cost-effectiveness and obtain CE-certification (WP3)
In Radiomics for lung cancer (RAIL) we will validate and qualify imaging biomarkers together with a ready-to-use application to deliver more accurate prognostic information, personalise treatment for NSCLC patients, reduce healthcare costs by EUR 500+ million and enable efficient clinical trial design. During the project we will implement the clinical benefits together with a CE-certified computer-aided prognostic and predictive application that converts medical images into a clear and concise report based on Radiomic signatures to support treatment decisions. Radiomic imaging biomarkers will be validated based on the highest standards in clinical validation (level I) in a phase II multi-centre clinical study. Based on these needs, we have formulated the Radiomics for Lung cancer (RAIL) project with the following key objective:
“To validate and qualify a multi-site level I imaging biomarker (RIB1) together with a prognostic application (RadioCAP) to deliver a better prognosis and personalise treatment for NSCLC patients.”
In order to achieve our main goal, we have implemented the following sub-objectives:
1. To develop a radiomics-based prognostic application (WP1)
2. To validate clinical utility of RadioCAP and RIB1 (WP2)
3. To validate cost-effectiveness and obtain CE-certification (WP3)
Working towards achieving the objectives, we have completed the following deliverables within this first reporting period.
User input software RadioCAP
We performed a market research by interviews under a group of professionals, representative of our future costumers. Based on the results of this market research we identified a list of key features and functionalities for RadioCAP, as well as the desired structure of RadioCAP.
RadioCAP software and IT infrastructure
The (alpha phase) RadioCAP software is able to fully process CT and/or PET imaging and associated structure contours (DICOM). Our core Radiomics software is capable of extracting different classes of Radiomic features, describing various characteristics of the tumor (or other regions of interest): (1) intensity, (2) shape, (3) texture and (4) multi-scale or multi-level features. The software provides three main functionalities: (1) DICOM data management, (2) Radiomic workflow management, and (3) management and export of results.
Dedicated IT infrastructure
The (alpha phase) IT infrastructure for RadioCAP is implemented using a DICOM processing pipeline, making use of proven open source components, which provides a stable DICOM management framework for the Radiomic analysis to be plugged in. The IT infrastructure allows us to access (i.e. query) and process DICOM imaging and DICOM RTSTRUCT objects directly from the PACS (Picture Archiving and Communication System), as well as query already obtained results from any configured source (e.g. MSSQL, PostgreSQL, SPARQL).
Online version of public web site
We have launched a new, separate public website for Radiomics in the context of the RAIL project, describing our mission and the Radiomics related products we want to offer. For the branding of Radiomics products we have chosen the name OncoRadiomics, designed a logo and registered an appropriate domain for our website. See: http://www.oncoradiomics.com
Detailed communication and dissemination plan
For communication and dissemination, we will make use of our website(s), e-mail accounts, social media and animations (YouTube). We also plan to set up a Wikipedia page regarding Radiomics. With these tools we expect to have a solid basis to promote and exploit the RAIL project results and associated research and development efforts and make research results (and at a later stage clinical results) available to the general public (including external stakeholders).
Dissemination material and updates
We developed a new, responsive (i.e. cross-device compatible for large screens, smartphones, etc.) website for OncoRadiomics (http://www.oncoradiomics.com/test/) and use the domain oncoradiomics.com to send and receive e-mail. Furthermore, we created social media accounts (Twitter, Facebook and LinkedIN), flyers/pamflets (e.g. for conferences) and business cards.
User input software RadioCAP
We performed a market research by interviews under a group of professionals, representative of our future costumers. Based on the results of this market research we identified a list of key features and functionalities for RadioCAP, as well as the desired structure of RadioCAP.
RadioCAP software and IT infrastructure
The (alpha phase) RadioCAP software is able to fully process CT and/or PET imaging and associated structure contours (DICOM). Our core Radiomics software is capable of extracting different classes of Radiomic features, describing various characteristics of the tumor (or other regions of interest): (1) intensity, (2) shape, (3) texture and (4) multi-scale or multi-level features. The software provides three main functionalities: (1) DICOM data management, (2) Radiomic workflow management, and (3) management and export of results.
Dedicated IT infrastructure
The (alpha phase) IT infrastructure for RadioCAP is implemented using a DICOM processing pipeline, making use of proven open source components, which provides a stable DICOM management framework for the Radiomic analysis to be plugged in. The IT infrastructure allows us to access (i.e. query) and process DICOM imaging and DICOM RTSTRUCT objects directly from the PACS (Picture Archiving and Communication System), as well as query already obtained results from any configured source (e.g. MSSQL, PostgreSQL, SPARQL).
Online version of public web site
We have launched a new, separate public website for Radiomics in the context of the RAIL project, describing our mission and the Radiomics related products we want to offer. For the branding of Radiomics products we have chosen the name OncoRadiomics, designed a logo and registered an appropriate domain for our website. See: http://www.oncoradiomics.com
Detailed communication and dissemination plan
For communication and dissemination, we will make use of our website(s), e-mail accounts, social media and animations (YouTube). We also plan to set up a Wikipedia page regarding Radiomics. With these tools we expect to have a solid basis to promote and exploit the RAIL project results and associated research and development efforts and make research results (and at a later stage clinical results) available to the general public (including external stakeholders).
Dissemination material and updates
We developed a new, responsive (i.e. cross-device compatible for large screens, smartphones, etc.) website for OncoRadiomics (http://www.oncoradiomics.com/test/) and use the domain oncoradiomics.com to send and receive e-mail. Furthermore, we created social media accounts (Twitter, Facebook and LinkedIN), flyers/pamflets (e.g. for conferences) and business cards.
Together with the French SME Aquilab (www.aquilab.com) and Maastro clinic (www.maastro.nl) ptTheragnostic has successfully applied for the Eurostars grant E-10116, with acronym DART. The total funding for this project is 1.29M €. Additional investments needed for this project by ptTheragnostic will, in part, be funded through EU (SME instrument RAIL¬–the current project) and own funds.
DART stands for “DeltaRadiomics: non-invasive stratification & response prediction tool to personalise cancer therapy”. DART will expand the applicability of Radiomics by developing and validating delta Radiomics to enable longitudinal treatment response monitoring. The ultimate ambition is that delta Radiomics becomes a first line diagnostic tool for tumour phenotyping and for clinical decision making, reaching thousands of patients. It will be offered as standalone software, web-based and as plug-in integrated into Aquilab’s imaging software (non-exclusive license agreement). DART opens up possibilities to maintain our position as a market leader in Radiomics. We have already obtained promising preliminary results for which we have filed patent applications (Netherlands Patent Application No. 2016278, patent application number PCT/NL2014/050728) and our aim is to extend our studies to further optimize Delta Radiomics and fully integrate it in our clinical application.
DART stands for “DeltaRadiomics: non-invasive stratification & response prediction tool to personalise cancer therapy”. DART will expand the applicability of Radiomics by developing and validating delta Radiomics to enable longitudinal treatment response monitoring. The ultimate ambition is that delta Radiomics becomes a first line diagnostic tool for tumour phenotyping and for clinical decision making, reaching thousands of patients. It will be offered as standalone software, web-based and as plug-in integrated into Aquilab’s imaging software (non-exclusive license agreement). DART opens up possibilities to maintain our position as a market leader in Radiomics. We have already obtained promising preliminary results for which we have filed patent applications (Netherlands Patent Application No. 2016278, patent application number PCT/NL2014/050728) and our aim is to extend our studies to further optimize Delta Radiomics and fully integrate it in our clinical application.