Final Report Summary - HISOCELL (High Throughput Multimodal Microfluidic Sorting and Biological Analysis of Circulating Tumour Cells)
The main goal of the HiSoCell project was development of a microfluidic platform for high throughput and multi-modal capturing of circulating tumour cells (CTCs).
Cancer is a field of major societal importance with high demand of progress in bio-analytical diagnosis. One of the challenges for cancer treatment is development of metastases, arising from cancer cells that transiently travel in the blood as circulating tumor cells (CTC). These cells, capable to invade further organs are responsible for 90% of cancer casualties. They also represent a valuable biomarker and an important diagnostic tool for patient follow-up after primary treatment, leading to the concept of “liquid biopsy”. Their specific identification is however difficult, since blood contains numbers of different cells with concentration 102-109 times higher than CTCs. Therefore, the development of efficient pre-concentration techniques is required to exploit the potential interest of CTC for diagnosis and prognosis purpose.
Currently, the “gold standard” for CTC enumeration and analysis is the FDA approved J&J Veridex platform, which performs batch sorting of cancer cells in blood samples up to 7.5 ml. The capture is based on Epithelial Cell Adhesion Molecule (EpCAM) positive selection with magnetic particles. Although Veridex has considerably improved the automation of CTC sorting, the capture specificity is limited to only one antibody; therefore only EpCAM positive CTCs are detected. Furthermore, system does not allow easy retrieving of captured cells and downstream molecular characterization.
Microfluidics for CTC sorting is now widely recognized as very promising technology family for the future of CTC analysis. There are two main approaches, either cell capture using antibodies or cell sorting based on different size or deformability of cell populations. Microfluidic antibody-based systems require bio-functionalization separately for each chip, what is expensive and difficult to reproduce. The host laboratory already developed a microfluidic CTC capture system, EPHESIA based on columns of bio-functionalized superparamagnetic beads, self-assembled in a microfluidic channel onto an array of magnetic traps. This technique offered flexible solution for bio-functionalization, but as all the current microfluidic CTC sorting solutions is limited by too low throughput. Total volume of blood processed is in the range of 3 ml in two hours.
Aim of the HiSoCell project was to develop a microfluidic platform allowing increase more than 10 times the volume of sample analysed at equal time, while keeping the characterization capabilities of a microfluidic approach. Progress towards this goal was realized along two separate lines. Both strategies relied on the deterministic lateral displacement (DLD) method, allowing size-sorting of objects by steric interaction under flow in an asymmetric array of posts. Value of the critical diameter determines whether is object transported longitudinally or laterally and depends on the parameters of the array (post size, distance and shift between lines of posts). Stereolithography was selected as an optimal fabrication technique; it allows fast and flexible fabrication of complex 3D architectures in variety of materials and can be therefore used to produce systems based on innovative concepts difficult to prepare using standard technologies. Within HiSoCell project we developed a way to address software limitation of the standard stereolithography machine and decrease the achievable post size from 90 µm to 32 µm.
A first research track involved a microfluidic implementation of the Rosette-Sep protocol, allowing CTC enrichment by negative selection using tetrameric antibody complexes directed against unwanted red and white blood cells, in combination with magnetic guidance using magnetic beads to form aggregates. This negative selection approach allows us to retrieve CTC, without any a priori limitation on expressed antigens.
We developed and optimized a protocol to prepare planar microfluidic device with incorporated magnetic rails (ferromagnetic material locally modify magnetic field distribution, resulting in a strong magnetic field gradient, which laterally deviates cells of interest aggregated using magnetic beads) and DLD post array in the channel placed over the rails. The system was so far validated with regards to a positive enrichment strategy using SKBR3 cell line and anti-EpCAM magnetic particles.
We were able to prepare “magnetic” blood aggregates by modification of the magnetic beads with CD45 and GYPA antibodies, and control these aggregates in the microfluidic device. Validations of the system using blood spiked with cancer cell lines were performed. However, this approach did not allow achieving a satisfactory purity. Further optimization is needed.
The second research track involved a 3D representation of the deterministic lateral displacement concept. The validity of the proposed geometry was first checked by Comsol simulations. A stereolithography protocol including the use of two different materials (mechanically stable material to form posts, transparent material to form walls) was developed and optimized for fabrication of a device with the critical diameter 120 µm. A fluidic environment was also designed and optimized, in order to accommodate the fluid volumes treated in this device, significantly higher than in standard microfluidics. The experimental proof of concept was achieved with polystyrene beads, to evaluate the potential of the system without being dependent of the variability of geometries of cancer cells. The device was able to sort polystyrene beads with 100 and 150 µm diameters with more that 99% efficiency, at a sample flow-rate 300 µl/min.
These results must now be transferred to the sorting of CTC. This will require further progress in the resolution of microfabrication, to shift the critical size to the 10-20µm range. These results, however, show that stereolithography offers a potential breakthrough in micro-fabrications techniques. While pushing the system resolution boundary and developing protocols incorporating multiple materials we showed that stereolithography can be used to produce microfluidic devices with high flexibility and with a significantly higher throughput than conventional systems, what was a main goal of the project. The production of the two systems described above would be extremely challenging using standard microfabrication techniques. Although further progress is still required, we believe that 3D DLD provides a new powerful strategy towards high throughput CTC sorting. Maintaining of the system throughput, while decreasing the system resolution and/or optimizing the formation of blood aggregates without magnetic beads would lead us towards the possibility to process 7.5 ml of the blood sample within 25 minutes.
Cancer is a field of major societal importance with high demand of progress in bio-analytical diagnosis. One of the challenges for cancer treatment is development of metastases, arising from cancer cells that transiently travel in the blood as circulating tumor cells (CTC). These cells, capable to invade further organs are responsible for 90% of cancer casualties. They also represent a valuable biomarker and an important diagnostic tool for patient follow-up after primary treatment, leading to the concept of “liquid biopsy”. Their specific identification is however difficult, since blood contains numbers of different cells with concentration 102-109 times higher than CTCs. Therefore, the development of efficient pre-concentration techniques is required to exploit the potential interest of CTC for diagnosis and prognosis purpose.
Currently, the “gold standard” for CTC enumeration and analysis is the FDA approved J&J Veridex platform, which performs batch sorting of cancer cells in blood samples up to 7.5 ml. The capture is based on Epithelial Cell Adhesion Molecule (EpCAM) positive selection with magnetic particles. Although Veridex has considerably improved the automation of CTC sorting, the capture specificity is limited to only one antibody; therefore only EpCAM positive CTCs are detected. Furthermore, system does not allow easy retrieving of captured cells and downstream molecular characterization.
Microfluidics for CTC sorting is now widely recognized as very promising technology family for the future of CTC analysis. There are two main approaches, either cell capture using antibodies or cell sorting based on different size or deformability of cell populations. Microfluidic antibody-based systems require bio-functionalization separately for each chip, what is expensive and difficult to reproduce. The host laboratory already developed a microfluidic CTC capture system, EPHESIA based on columns of bio-functionalized superparamagnetic beads, self-assembled in a microfluidic channel onto an array of magnetic traps. This technique offered flexible solution for bio-functionalization, but as all the current microfluidic CTC sorting solutions is limited by too low throughput. Total volume of blood processed is in the range of 3 ml in two hours.
Aim of the HiSoCell project was to develop a microfluidic platform allowing increase more than 10 times the volume of sample analysed at equal time, while keeping the characterization capabilities of a microfluidic approach. Progress towards this goal was realized along two separate lines. Both strategies relied on the deterministic lateral displacement (DLD) method, allowing size-sorting of objects by steric interaction under flow in an asymmetric array of posts. Value of the critical diameter determines whether is object transported longitudinally or laterally and depends on the parameters of the array (post size, distance and shift between lines of posts). Stereolithography was selected as an optimal fabrication technique; it allows fast and flexible fabrication of complex 3D architectures in variety of materials and can be therefore used to produce systems based on innovative concepts difficult to prepare using standard technologies. Within HiSoCell project we developed a way to address software limitation of the standard stereolithography machine and decrease the achievable post size from 90 µm to 32 µm.
A first research track involved a microfluidic implementation of the Rosette-Sep protocol, allowing CTC enrichment by negative selection using tetrameric antibody complexes directed against unwanted red and white blood cells, in combination with magnetic guidance using magnetic beads to form aggregates. This negative selection approach allows us to retrieve CTC, without any a priori limitation on expressed antigens.
We developed and optimized a protocol to prepare planar microfluidic device with incorporated magnetic rails (ferromagnetic material locally modify magnetic field distribution, resulting in a strong magnetic field gradient, which laterally deviates cells of interest aggregated using magnetic beads) and DLD post array in the channel placed over the rails. The system was so far validated with regards to a positive enrichment strategy using SKBR3 cell line and anti-EpCAM magnetic particles.
We were able to prepare “magnetic” blood aggregates by modification of the magnetic beads with CD45 and GYPA antibodies, and control these aggregates in the microfluidic device. Validations of the system using blood spiked with cancer cell lines were performed. However, this approach did not allow achieving a satisfactory purity. Further optimization is needed.
The second research track involved a 3D representation of the deterministic lateral displacement concept. The validity of the proposed geometry was first checked by Comsol simulations. A stereolithography protocol including the use of two different materials (mechanically stable material to form posts, transparent material to form walls) was developed and optimized for fabrication of a device with the critical diameter 120 µm. A fluidic environment was also designed and optimized, in order to accommodate the fluid volumes treated in this device, significantly higher than in standard microfluidics. The experimental proof of concept was achieved with polystyrene beads, to evaluate the potential of the system without being dependent of the variability of geometries of cancer cells. The device was able to sort polystyrene beads with 100 and 150 µm diameters with more that 99% efficiency, at a sample flow-rate 300 µl/min.
These results must now be transferred to the sorting of CTC. This will require further progress in the resolution of microfabrication, to shift the critical size to the 10-20µm range. These results, however, show that stereolithography offers a potential breakthrough in micro-fabrications techniques. While pushing the system resolution boundary and developing protocols incorporating multiple materials we showed that stereolithography can be used to produce microfluidic devices with high flexibility and with a significantly higher throughput than conventional systems, what was a main goal of the project. The production of the two systems described above would be extremely challenging using standard microfabrication techniques. Although further progress is still required, we believe that 3D DLD provides a new powerful strategy towards high throughput CTC sorting. Maintaining of the system throughput, while decreasing the system resolution and/or optimizing the formation of blood aggregates without magnetic beads would lead us towards the possibility to process 7.5 ml of the blood sample within 25 minutes.