Periodic Reporting for period 1 - AI CUrES (AI to predict Cancer metastasis using Ultra-Echo-Sono imaging)
Berichtszeitraum: 2022-11-01 bis 2024-04-30
Cancer metastasis is the leading cause of death within tumor disease and is currently an unmet clinical need. This dramatic clinical evidence requires immediate action in terms of early diagnosis and then pre-emptive therapies, not only for the primary tumor, but especially at the first stages of metastasis formation. Metastasis is the result of a complex mechanism, still under study, which involves the trafficking of biological materials, in particular extracellular vesicles (EVs), from the primary tumor site towards other tissues. It is thus imperative to develop novel and high-resolution biological techniques to understand the metastatic behaviour of tumour-derived EVs, with the aim to propose early diagnosis of metastasis and set timely pre-emptive therapies , empower metastasis treatments and patients’ outcome. To tackle these challenges, much more knowledge has to be gained on the metastatic potential of tumore-derived EVs and their molecular components, i.e. surface proteins and RNA contents, to be further generalized to other tumors.
For the above reasons, the project AI CUrES has developed novel methodologies for understanding the pathway and mechanism of metastasis spread led by EVs and has successfully implemented Artificial Intelligence algorithms to analyse these mechanisms and the early diagnostic images.
AI CUrES thus started from a Key Research Question: Is it possible to predict with high resolution the role of circulating EVs in cancer metastasis?
The answer was obtained by exploiting innovative research methodologies for high resolution and real-time bioimaging of circ-EVs and obtaining a powerful and novel digital diagnostic tool thanks to the use of Artificial intelligence (AI) algorithms. These breakthrough techniques has allowed gaining a knowledge on the role of EVs secreted by the primary tumor, but also empowering the research with new imaging tools.
This work has the future vision of predicting possible patient’s outcomes and helping to set timely pre-emptive therapies. The potential impact on oncological patient’s health is to reduce cancer-related deaths and treatment costs of the healthcare system, while opening up also breakthrough perspectives at the technological and industrial level for high-tech and high-resolution imaging techniques and e-tools in diagnostics.
In AI CUrES, EVs produced by a metastatic colorectal cancer (CRC) have been tagged by specific nanosized-contrast agents and their trafficking in biological media has been reported unravelling real-time, time-lapsed and molecular characteristics. The collected images of EVs trafficking towards recipient cell were achieved from in vitro cell cultures, 3D organoids and in vivo models, being refined and further elaborated by artificial intelligence algorithms, with the future aim to predict the possible metastasis development from a starting CRC tumor mass.
The main aim was thus to validate a breakthrough technology able to identify circulating EVs originated from a tumor mass of colorectal cancer; understand their trafficking mechanisms in spreading metastasis; early diagnose the possible insurgence of metastasis. These ambitious objectives comprise several goals in different sectors:
1. as scientific end goal, to unravel the circ-EVs trafficking mechanism and thus metastasis insurgence;
2. as a social and economic goal, to reduce cancer-related deaths, saving lives and public health money, empowering Europe in nanomedicine, med-tech tools and digital health;
3. as technological end goal, to develop a diagnostic method of circ-EVs surveillance and an innovative artificial intelligence e-tool to analyse the gained information and use it to obtain predictive values to fight cancer progression and metastasis, enabling early diagnosis;
4. as commercial end goal, to promote a Technology Transfer strategy to rapidly come up with a clinical innovation, introducing this new diagnostic technology in the clinical market.
For the above reasons, the project AI CUrES has developed novel methodologies for understanding the pathway and mechanism of metastasis spread led by EVs and has successfully implemented Artificial Intelligence algorithms to analyse these mechanisms and the early diagnostic images.
AI CUrES thus started from a Key Research Question: Is it possible to predict with high resolution the role of circulating EVs in cancer metastasis?
The answer was obtained by exploiting innovative research methodologies for high resolution and real-time bioimaging of circ-EVs and obtaining a powerful and novel digital diagnostic tool thanks to the use of Artificial intelligence (AI) algorithms. These breakthrough techniques has allowed gaining a knowledge on the role of EVs secreted by the primary tumor, but also empowering the research with new imaging tools.
This work has the future vision of predicting possible patient’s outcomes and helping to set timely pre-emptive therapies. The potential impact on oncological patient’s health is to reduce cancer-related deaths and treatment costs of the healthcare system, while opening up also breakthrough perspectives at the technological and industrial level for high-tech and high-resolution imaging techniques and e-tools in diagnostics.
In AI CUrES, EVs produced by a metastatic colorectal cancer (CRC) have been tagged by specific nanosized-contrast agents and their trafficking in biological media has been reported unravelling real-time, time-lapsed and molecular characteristics. The collected images of EVs trafficking towards recipient cell were achieved from in vitro cell cultures, 3D organoids and in vivo models, being refined and further elaborated by artificial intelligence algorithms, with the future aim to predict the possible metastasis development from a starting CRC tumor mass.
The main aim was thus to validate a breakthrough technology able to identify circulating EVs originated from a tumor mass of colorectal cancer; understand their trafficking mechanisms in spreading metastasis; early diagnose the possible insurgence of metastasis. These ambitious objectives comprise several goals in different sectors:
1. as scientific end goal, to unravel the circ-EVs trafficking mechanism and thus metastasis insurgence;
2. as a social and economic goal, to reduce cancer-related deaths, saving lives and public health money, empowering Europe in nanomedicine, med-tech tools and digital health;
3. as technological end goal, to develop a diagnostic method of circ-EVs surveillance and an innovative artificial intelligence e-tool to analyse the gained information and use it to obtain predictive values to fight cancer progression and metastasis, enabling early diagnosis;
4. as commercial end goal, to promote a Technology Transfer strategy to rapidly come up with a clinical innovation, introducing this new diagnostic technology in the clinical market.
The AI CUrES project has been divided in different WPs to achieve of the above-mentioned objectives.
We have first implemented and optimized the wet-chemical synthesis for safe-by-design, biodegradable and cytocompatible ZnO NPs (according to Carofiglio Nanomaterials 2021). Then we have set cell cultures from metastatic colorectal cancer models and isolated the produced EVs and fully characterized them. We have then performed a convenient labelling of the EVs with and without NPs obtaining a nanosystem able to be visualized with different imaging modes (fluorescence, echografic and sonoluminescent imaging).
By using the NPs as a nano-sized contrast agent we were then able to establish and optimize independent imaging techniques such as fluorescent imaging, and most innovatively echographic B-mode imaging (Ancona Ultrason. Sonochem.2020) and NP-assisted sonoluminescence (SL) imaging (Vighetto ACS Omega 2022) first in water, then cell culture media and finally in-vitro. All the acquired in vitro images have been also used as input for our proprietary AI algorithm. This work has thus achieved a key intermediate goal, leading to NP-assisted echographic B-mode and sonoluminescence imaging in high-resolution of in-vitro cells and of extracellular vesicles.
The tests further evolved to study the tumour-derived EV tropism and biodistribution in 3D models of colorectal cancer and in vivo mice model to assess the preferred organs of homing, allowing to achieve another key scientific goal of understanding the role of tumor-derived EVs in spreading metastasis. AI CUrES has also demonstrated an innovative technology enabling bioimaging and light-mediated excitation nanotools to allow further exploration of therapies and diagnostic capabilities in CRC and other tumor models.
Concerning the development of AI algorithms the hundreds of images collected from the EVs and cells bioimaging were stored, classified, labelled and segmented according to the site of imaging, conditions of acquisition, use of EVs, NPs or control experiments. Starting from literature data and our experience, two proprietary algorithms were written and adopted to work on the collected images and video frames features. The obtained results allow to track nanoobject in in vitro conditions, like EVs trafficking and recognition routines to identify tumor versus healthy cells.
We have first implemented and optimized the wet-chemical synthesis for safe-by-design, biodegradable and cytocompatible ZnO NPs (according to Carofiglio Nanomaterials 2021). Then we have set cell cultures from metastatic colorectal cancer models and isolated the produced EVs and fully characterized them. We have then performed a convenient labelling of the EVs with and without NPs obtaining a nanosystem able to be visualized with different imaging modes (fluorescence, echografic and sonoluminescent imaging).
By using the NPs as a nano-sized contrast agent we were then able to establish and optimize independent imaging techniques such as fluorescent imaging, and most innovatively echographic B-mode imaging (Ancona Ultrason. Sonochem.2020) and NP-assisted sonoluminescence (SL) imaging (Vighetto ACS Omega 2022) first in water, then cell culture media and finally in-vitro. All the acquired in vitro images have been also used as input for our proprietary AI algorithm. This work has thus achieved a key intermediate goal, leading to NP-assisted echographic B-mode and sonoluminescence imaging in high-resolution of in-vitro cells and of extracellular vesicles.
The tests further evolved to study the tumour-derived EV tropism and biodistribution in 3D models of colorectal cancer and in vivo mice model to assess the preferred organs of homing, allowing to achieve another key scientific goal of understanding the role of tumor-derived EVs in spreading metastasis. AI CUrES has also demonstrated an innovative technology enabling bioimaging and light-mediated excitation nanotools to allow further exploration of therapies and diagnostic capabilities in CRC and other tumor models.
Concerning the development of AI algorithms the hundreds of images collected from the EVs and cells bioimaging were stored, classified, labelled and segmented according to the site of imaging, conditions of acquisition, use of EVs, NPs or control experiments. Starting from literature data and our experience, two proprietary algorithms were written and adopted to work on the collected images and video frames features. The obtained results allow to track nanoobject in in vitro conditions, like EVs trafficking and recognition routines to identify tumor versus healthy cells.
This project AI CUrES sets a clear value proposition: to increase the knowledge on tumor-derived EV trafficking mechanisms enabling to undestand the insurgence of cancer metastasis. It has given rise to different innovation potential: (i) tagging the EVs, produced by a primary CRC in order to unravel their mechanism of action in spreading metastasis using multimodal imaging techniques; (ii) developing novel NP-assisted bio-imaging techniques , i.e. echography and sonoluminescence, thus being highly resolved, real-time and highly penetrating in tissues; (iii) use Artificial Intelligence algorithms to elaborate the acquired images and video frames and set a high-throughput e-tool about the EVs and cell behaviour.
To further push the multiple developed technologies to a marketable level further applied research in real environment is running, aiming to construct a demonstrator. Intellectual Property Rights have to be further made stronger to protect the design of the set up, and regulatory and standardisation framework have to be put in place for the use of nanoparticles and AI tools in clinics.
To further push the multiple developed technologies to a marketable level further applied research in real environment is running, aiming to construct a demonstrator. Intellectual Property Rights have to be further made stronger to protect the design of the set up, and regulatory and standardisation framework have to be put in place for the use of nanoparticles and AI tools in clinics.