Skip to main content

Integrated Micro-Nano-Opto Fluidic systems for high-content diagnosis and studies of rare cancer cells

Final Report Summary - CAMINEMS (Integrated micro-nano-opto fluidic systems for high-content diagnosis and studies of rare cancer cells)

Executive summary:

CAMINEMS is supported by the European Commission (EC) through the Seventh Framework Programme (FP7) for research and development (R&D) and has been running since 1 July 2009 for 42 months. It gathers 9 partners from 5 European countries and receives EUR 3.5 million of EC funding.

The project was aimed at developing new tools based on microfluidics and nanotechnologies, to improve cancer diagnosis and prognosis. Cancer causes about 13 % of all deaths in the world. It is the second cause of mortality in Europe. In particular, about 90 % of cancer deaths are due to metastases and therapeutic escape. At the origin of metastases are Circulating tumour cell (CTC)s, individual cells or small cellular aggregates issued from the primary cancer and transiently circulated in the blood. These CTC stop in specific organs and become micrometastases or disseminated tumour cells. These micrometastases can remain quiescent or dormant for months or years before developing again, leading to metastases.

It would thus be a major breakthrough for treatments, to be able to perform a detailed molecular characterisation on CTC before they develop into metastases, or as mildly invasive 'liquid biopsy' to select the best treatment. Besides this major diagnosis and prognosis application, being able to capture and to study CTC would also be a highly valuable help in research, for understanding their metabolism, and their response to existing or candidate drugs. Providing a tool to achieve these objectives was the main aim of the CAMINEMS project.

The developed Ephesia technology consists in self-assembling an array of antibody-bearing magnetic particles in a high throughput microfluidic device. These beads create a self-assembled 'micro-posts' array, with an aspect ratio much higher than that of microfabricated post arrays, allowing the use of innovative high resolution imaging with very little optical interference. Blood depleted from Red blood cell (RBC)s is then flown in the array with a uniform flow velocity, CTC are captured and complex characterisation protocols (membrane and cytosol immunophenotyping, detailed morphological analysis, genetic analysis by in situ hybridisation) can be performed in situ in a fully automated way. This reduces risk of cell loss or damage, as compared to the release and collection of the cells for delayed analysis in a separate device.

CAMINEMS' work on magnetic micro and nanoparticles has allowed several generations of particles to be synthetised. These new particles expand the range of biofunctionalisation available, as compared e.g. to state of the art systems.

New microfluidic chips allowing large blood volumes (7.5 ml) throughput were designed and developed successfully, using the new Ephesia technology. This work has overall led to several oral and poster presentations at conferences, numerous publications including two in the Proceedings of the National Academy of Science USA and in Lab on Chips, and a joint patent.

CAMINEMS system was first validated on the characterisation of lymphoma. An agreement of 100 % on 20 patients was achieved in a blind comparison with a combined use of flow cytometry and cytological observation (PNAS 2010). The capture efficiency (90 %) and specificity (x2 500 enrichment versus white blood cells) was then characterised by spiking cell lines with different expression levels. Using various cell lines (SKBR3, MCF7 for breast cancer, PC3 for prostate cancer, Raji and Jurkat as lymphoid cells), we obtained a capture efficiency up to 90.6 % for cell quantity as low as 50 cells per sample, the best result worldwide to our knowledge, and a specificity better than 99.6 %. Finally a clinical comparison with the FDA approved Veridex technology showed equal or superior capture efficiency, much high specificity and more flexibility in multimodal typing power.

The promises of the project are strong enough, that we consider future industrial exploitation. CAMINEMS is now working on a second generation pre-industrial prototype, and searching partners for industrialisation and commercialisation.

Project context and objectives:

Outline and objectives

CAMINEMS is supported by the EC through FP7 for R&D and has been running since 1 July 2009 for 42 months. It gathers 9 partners from 5 European countries and receives EUR 3.5 million of EC funding.

The project was aimed at developing new tools based on microfluidics and nanotechnologies, to improve cancer diagnosis and prognosis. Cancer causes about 13 % of all deaths in the world and thus represents a huge problem in public health. It is the second cause of mortality in Europe. Cancer is also a very painful and hard to treat disease, involving long-term and poorly predictable threats of fatal issues, so its global impact on quality of life is particularly dramatic. Cancer treatment has steadily improved in the last 100 years, but at a slow pace as compared to other diseases like infectious or cardiovascular ones, a reflection of its particularly high biological complexity.

In particular, today, about 90 % of cancer deaths are due to metastases and therapeutic escape. At the origin of metastases are CTCs, individual cells or small cellular issued from the primary cancer and transiently circulated in the blood. These CTC stop in specific organs as bone marrow, liver and / or brain, and become micrometastases or disseminated tumour cells. These micrometastases can remain quiescent or dormant for months or years be-fore developing again, leading to metastases. Quite generally, metastases develop characteristics which make them resistant to treatments that were efficient on the primary tumours, and they are often discovered when it is too late. It would thus be a major breakthrough for treatments, to be able to perform a detailed molecular characterisation on CTC before they develop into metastases. Besides this major diagnosis and prognosis application, being able to capture and to study CTC would also be a highly valuable help in research, for understanding their metabolism, and their response to existing or candidate drugs.

For all these applications, current technologies are insufficient both in sensitivity and specificity. They can detect micrometastases only in patients with advanced cancer, and they only allow the identification of a few biomarkers. Providing a tool for overcoming these limitations, based on innovations in converging sciences, was the main objective of the CAMINEMS project.

Project results:

Achievements

Overall, on the technical side the project has progressed along the planned lines, and the major expected technical deliverables have been delivered. We encountered, however, a major difficulty in an aspect that was initially expected to be a side one, and proved critical for the project, regarding sample treatment, conservation and transport. This has strongly delayed the multicentre clinical validation aspect of the project, and has required some strategic choices. Also these difficulties in validation have also delayed the integration part, by delaying some milestones.

WP1 (magnetic micro and nanoparticles) has progressed along schedule. Several generations of particles were synthetised and new biofunctionalisation routes were developed successfully. Some particles were already demonstrated as operational within the microsystems developed in WP2. These new particles expand the range of biofunctionalisation available, as compared e.g. to state of the art systems. Notably capture using HER2 surface antigen was demonstrated. This protein is a major biomarker regarding personalised medicine of breast cancer, and the possibility to specifically capture and screen CTC for this antigen opens new routes to CTC characterisation and better treatment of therapeutic escapes.

WP2 (bio-NEMS), involved the development of microfluidic chips for the capture of CTC from relatively large blood volumes (7.5 ml), using the new Ephesia technology. New microfluidic chips allowing this throughput were designed and developed successfully. A new approach for the preparation of magnetic arrays in these chips was also developed. This technique is based on an original capillary self-assembly method. This work has overall led to several oral and poster presentations at conferences, two publications so far in the Proceedings of the National Academy of Science USA and in Lab on Chips, and a joint patent. This part is thus considered highly successful.

WP3 (Nano-optics) also progressed roughly along schedule. A first generation system had been tested at month 18, showing insufficient sensitivity, and a new generation system was displayed at month 24, with a sensitivity increased by about 10. Also, T3.3 in vivo intracellular tracking, yielded an important discovery in the fundamental of Deoxyribonucleic acid (DNA) dynamics in the cell, which was not contemplated in the initial project, but will be an important 'side product' of the project.

Tools for image processing (WP4) evolved strongly towards integration and some elements towards WP5 are already validated. Denoising, which was the most innovative content of the project regarding computing, has yielded results beyond expectations (a decrease by a factor >10 of acquisition time as compared to conventional analysis). Overall, the content of technical Work package (WPs 1-4) is now considered achieved with little deviation from the initial plan, and some unexpected positive outcomes. Some optimisation and improvement are still needed, but the current state of the technology is already significantly above state of the art (see validation, WP6 below). We were also particularly satisfied by the strong team spirit and cooperation that has now arose in the consortium. Most work in the last year was multicentre, regarding both technological integration (between CURIE-CTIF, UOXF-DF, CURIE-MMBM and FLUIGENT), and clinical validation: Large samples multicentre and multi-techniques validation has been operated between LMUM, IGR, Curie-Hospital on the one hand, and Curie-MMBM on the other hand. Regarding small samples, partnership mostly involved IPATIMUP, IGR and Curie-MMBM. The transportation of Fine needle aspiration (FNA) samples from IPATIMUP and Curie-MMBM still raised problems, and this we consider it indeed as the last and only technological bottleneck that was not solved in this project. In contrast, we could finally solve the sample transportation problem between LMUM and Curie, which had stubbornly resisted for more than one year and obliged us to completely modify our sample treatment workflow.

Integration (WP5) has been delayed by the unexpected difficulty encountered when starting working with real clinical samples (see discussion above) regarding sample preparation and conservation. This required a reconsideration of the general protocol and layout of the prototype. Two operational prototypes are now in operation: a first one, in a semi-automated mode, has already been operating for now more than 6 months at Curie Institute, and was intensively used in the frame of WP6. The second one is installed at Fluigent, and was used to continue the development and maturation of automation, in preparation of WP7. The prototypes, in particular the one at Curie, continue to be intensely used, in order to pursue the clinical validation and widen the credibility of the project and its exploitation potential.

The advancement of WP6 (validation) was along schedule for cell lines. Using the new prototype for large volume samples, a systematic validation on cell lines was made and completed (deliverable 6.2). Briefly, series of experiments were performed in a blind multicentre process, using samples spiked and processed by Veridex at LMUM, then processed at Curie-Hospital by Veridex, at IGR by Veridex, and at Curie-MMBM by Ephesia. Using various cell lines (SKBR3, MCF7 for breast cancer, PC3 for prostate cancer, Raji and Jurkat as lymphoid cells), we obtained a capture efficiency up to 90.6 % for cell quantity as low as 50 cells per sample, the best result world-wide to our knowledge, and a specificity better than 99.6 %. An article was published in the international review 'Methods', and another article is under preparation. We also obtained excellent correlation between expression level and capture profile, opening the route to a new approach to perform by this technology a quantitative assessment of antigen expression by unknown cells. Clinical testing regarding small volume samples was validated on patient samples from lymphoma (FNA and pleural effusions). A 100 % agreement with state of the art methods (combination of FACS immunophenotyping and cytology) was achieved in a blind screening, using 10 times less sample than in conventional methods.

However, clinical validation on large blood samples was hindered by the same sample collection, and transportation problem as mentioned regarding WP2. Namely, due to the higher imaging performance allowed by our system, as compared to earlier ones, it was discovered that the conventional methods for sample collection and conservation induced a degradation of cells, that was not apparent in the poorly resolved stated of the art devices, but was a strong limitation for our own system. Thus, considerable effort had to be devoted to the development of a suitable sample preparation protocol. We now have essentially solved the problem, but this had two consequences. First, the transportation of samples from one site to the other remained impossible until a new preparation protocol was developed, and this limited the multi-sites aspect of validation. The problem has now been solved, and multi-sites validation could be performed, but this has overall yielded a roughly 6 months delay in the project, and a limitation in the number of samples actually treated between LMUM and Paris (we had, however, an intense multicentre screening protocol between IGR and Curie, thanks to the smaller distance allowing same day processing).

Potential impact:

Exploitation perspectives

This very positive outcome yields exploitation perspectives on the short and long term. First, the microfluidic automation technology developed by Fluigent has been industrialised into a series of products, which were launched internationally at the last microTAS conference in Japan. Second, the technological outcome of the project was used in a more clinically oriented project in the health programme, under progress (DIATOOLS).

Very encouraging results suggest that the CAMINEMS project will allow routine quantitative genetic testing by direct in situ Fluorescence hybridisation (FISH) of the captured CTC, in a much less labour-intensive and with a higher success rate than currently achieved by state of the art methods. The promises of the project are strong enough, that we consider future industrial exploitation. CAMINEMS is now working on the construction of a strategy, specification and economic analysis for a second generation pre-industrial prototype, and searching partners for industrialisation and commercialisation.

Coordinator's contact details: Dr Jean-Louis VIOVY (Institut CURIE), jean-louis.viovy@curie.fr

Project website: http://www.caminems.eu