May 1, 2018 (Vol. 38, No. 9)

Silicon Chip Technology Can Reduce the Cost, Time, and Complexity of the Workflow

CAR T-cell therapy is a promising new cancer treatment, based on a patient’s own immune system. The workflow starts with taking blood from the patient, typically in a hospital, selecting the white blood cells, and from these the T cells.

Next, the T cells are transported to a manufacturing facility. Here, they are re-engineered to produce receptors on their surface that allows the cells to recognize and destroy cancer cells. The code for the new receptor is inserted into the T cells through viral vectors or electroporation. These re-engineered T cells are then cultured in bioreactors to hundreds of millions of cells. Finally, the re-engineered cells are re-introduced into the patient, where they have the ability to further multiply and kill cancer cells.

Next to the biological issues (e.g., side effects, efficacy for solid tumors, new tumor targets), a key challenge is related to the manufacturing process. In patient-specific cell therapy, the complexity of the end product—living cells with a specific product profile—and the variability of the starting material (the patient’s own cells) make it challenging to ensure a cell therapy product that is comparable in quality, safety, and efficacy. Also, there is the complex logistics chain with blood being taken from a patient at a hospital; cryopreserved; shipped to a facility where the re-programming and manufacturing of T cells occurs; and then shipped back for infusion into the patient.

Finally, there is the need for a short turnaround time due to the patient’s situation and evolution of his or her disease. And the therapy should be affordable by keeping cost of goods (CoGs) down and, more importantly, have a scalable, sustainable, manufacturing process.

Chip-Centered Workflow

The complexity, cycle time and cost of the CAR T-cell therapy workflow is a major challenge to overcome to achieve clinical implementation of this revolutionary new cancer treatment. Chip technology can help (Figure 1). What follows are some examples of relevant chip-based solutions.

Figure 1. Workflow for CAR T-cell therapy with indications (in blue circles) of potential nanoelectronics impacts. (Based upon a figure from the Regenerative Medicine Innovation Platform.)

A Bubble-Based Jet Flow Sorting Technique

We developed an alternative sorting method—a microfluidic FACS technique—based on the use of micro-vapor bubbles generating a jet flow for fast but gentle cell sorting in microfluidic channels (Figure 2). The sorting speed is 5,000 cells/sec per single microfluidic channel, with a >90% cell sorting yield, >99% purity, and well-preserved cell viability. It’s a generic cell sorting technique that is independent of any physical characteristics of cells.

Advantages over the current cell therapy standard for selection (immunomagnetic separation, or IMS) include multimarker support, a more compact and automated system, low disposable cost, and no post-sorting step.

Figure 2. Overview of the vapor bubble jet cell sorter technology developed by imec. From left to right: the cell sorter chip; animated detail of this chip with focus on microheaters that create micro-vapor bubbles; image of cells being sorted inside the microfluidic channels on the chip.

A Micro-Electrode Array for Single-Cell Electroporation

Using chip technology, one can make large-scale microelectrode arrays (MEAs). These silicon chips, containing thousands of small electrodes, are covered with surface chemistry to make them compatible with cell cultures. When a small voltage is applied to the cell (via the electrode underneath), the cell membrane opens, and molecules in solution enter the cell.

The voltage used in these electrodes is so small that it has no negative effect on the cell. Also, there is a very precise control on the electroporation parameters for each individual cell. This will increase the yield and reproducibility, minimize the potential toxic effect, and increase the ultimate efficacy of the cell therapy in the patient.

Ion Sensors for Inline Monitoring

During the cell multiplication step, it is key to monitor and control the microenvironment of the cells inside these bioreactors. Based on chip technology, multi-ion sensors for fluid monitoring can be developed to measure, e.g., pH, Cl, Na, K, Ca, and NO3. It is a generic platform that can be tailored toward specific applications: if the selective membranes on the electrodes are changed, the sensor can be adapted to detect other ions.

The sensors outperform current systems in terms of performance, are easy to mass-produce, have a wireless connection, and are energy optimized and extremely miniaturized. The latter makes it possible to integrate the sensors inside the bioreactors.

A Miniaturized Lens-Free Microscope

Also, a visual inspection of the cells in the bioreactor is indispensable. Normally, this is performed by a process operator by taking a sample out of the bioreactor and inspecting it with a microscope. Based on chip technology, a lens-free imaging cytometer was developed. It could be integrated on top of microfluidic channels or into bioreactor walls. The lens-free holographic-imaging system uses a light source and a CMOS imager to capture the light that is diffracted off small objects.

The captured diffraction pattern is reconstructed into an in-focus image by custom software algorithms. It’s a very compact and low-cost solution, with a large field of view and very good resolution. A machine learning–based image analysis and classification pipeline is also being developed to evaluate the images and differentiate between specific cell types using powerful classification algorithms.


With a microfluidic-based and chip-centered approach (Figure 3), cell therapy manufacturing is expected to have a much shorter turnaround time. All process steps (selection, electroporation) are performed accurately, resulting in a higher yield of CAR T cells from the same amount of blood. This would eliminate the differences in efficacy and toxicity as seen today in clinical trials. With the use of microfluidics, processes occur much faster, and chip processing enables a high level of parallelization, speeding up the process enormously.

Figure 3. Vision of a manufacturing site with single-patient, miniature, smart bioreactors that each have integrated multianalyte sensors and imaging cytometers as well as a wireless connector for online and real-time monitoring of relevant parameters.

Liesbet Lagae, Ph.D. ([email protected]), is co-founder and program director of life science technologies at imec.

Previous articleConsortium Identifies 44 Variants as Risk Factors for Major Depression
Next articleFour Tips for Identifying Microbes in Your Facility