Nanoscale Testing Could Save Lives
IBM Research and others are collaborating on technology that helps doctors discover potential cancers even before symptoms appear.
Image by IBM Research
By Jim Utsler02/01/2018
If caught early enough, many cancers can be successfully treated. That’s why Gustavo Stolovitzky, program director of Translational Systems Biology and Nanobiotechnology at IBM Research, and researchers in the pathology department at the Icahn School of Medicine at Mount Sinai are collaborating on technology that helps doctors discover potential cancers even before symptoms appear.
Significantly, this type of testing could be done at a doctor’s office, with results being available within hours, or even at patients’ homes, with the data being uploaded to a clinic’s cloud for further examination. And if successful, this technology could save the lives of many people who may have otherwise succumbed to their disease.
IBM Systems Magazine (ISM): What was the genesis of your current research?
Gustavo Stolovitzky (GS): I wanted to use traditional IBM microchip technology and apply that to biology. At one point, in 2005, 2006, one of my goals was to contribute to the push to develop a technology to sequence a genome at the cost of less than $1,000. A few colleagues and I started that work by imagining a nanopore and nanochannel-based DNA sequencer leveraging our silicon technology and our computational biology expertise. That effort led to a number of applications to move DNA through pillars, and eventually, we recognized that we could apply an existing technology called “deterministic lateral displacement” (DLD) to the nanoscale, which gave rise to the nanoDLD technology that we are developing. Our current project involves the use of exosomes, which are extracellular vesicles that carry markers that have the potential to diagnose cancer earlier than other existing biomarkers.
ISM: How is IBM chip technology used in this scenario?
GS: Imagine a river with rocks in it. Smaller particles will follow the river flow without hitting any of the rocks. Big particles that approach the rocks at a particular distance, however, will hit the rocks. Now imagine that you have an array of rocks, or in our case, the pillar arrays on the chip. Small particles will not collide with the pillars, whereas bigger particles will be redirected when they collide with the pillars. In this way, particles of different sizes can be separated in our array and isolated for detection or analysis. In this way, we can separate exosomes from smaller particles. The information extracted from the exosomes isolated with the help of these chips can then be used to determine whether a disease is present.
ISM: Could you describe what exosomes are?
GS: Exosomes are little vesicles that are basically little bags coated with the same material that separates the outside and the inside of the cell, what’s called a lipid bilayer—two layers of fat. Because exosomes are formed inside the cells, they’re made of the same materials found in the cell. When exosomes were discovered in the 1980s, researchers thought they were basically garbage bags that helped cells get rid of the things that weren’t useful anymore, such as unneeded proteins or degraded RNA. Later on in the ’90s, people started realizing that exosomes weren’t just garbage bags, but also a means by which cells, both neighboring and distant, communicate with each other.
So in the late ’90s, early 2000s, people began wondering if exosomes could be used for diagnostics. Because exosomes contain what’s in the cells from where they originated, they might indicate the cancerous state of parent cells.
So here we are today. There’s a big need to understand the best ways to isolate exosomes from bodily fluids from, for example urine or blood, to detect prostate, bladder, pancreatic cancer, etc. We now have the hope that if you catch the exosomes that have the cargo of the cells that are sick, you can detect very early on—possibly before there are any symptoms—if there’s a disease lurking in someone’s body.
ISM: Do specific markers determine what type of cancer it might be?
GS: Yes, but the markers depend on the cancer you want to detect. For example, there recently was a scientific article reporting an exosome-based gene expression signature that helped determine whether a patient had an aggressive prostate cancer versus a possible indolent cancer (bit.ly/2oLpB5v). The clinical implication of this are big because for a patient it could mean the difference between doing a radical prostatectomy versus doing just active surveillance of the disease. That study was done by looking at patients’ urine. Another study a few years ago showed that if a cancer is going to metastasize, some exosomes will prepare the terrain for that cancer in another place in the body (go.nature.com/2yYMYIw). Some exosomes with particular proteins landed, for example, in the liver, with the proteins on the exosome surface determining where the metastases will occur. This is a great reason to research exosomes in detail at a very small scale.
ISM: Does this technology involve filtering exosomes from other materials found in bodily samples?
GS: Yes, but filtering is only one of the steps in the process. Our technology focuses on two levels. First, we’re working on a system that uses the nanoDLD technology system to separate and isolate exosomes from bodily fluids like urine or blood serum. Then, we analyze which markers are in those exosomes. So after separation, you have a pure, more concentrated sample of exosomes and you start to interrogate the exosomes. The nanoDLD technology will allow us to separate exosomes that contain particular surface markers and, in one of the applications we envision, run computations on the markers to determine whether a person is above or below normal thresholds.
Say a patient goes to the doctor for a PSA (Prostate-Specific Antigen) prostate cancer screening, which typically has a high rate of false positives. Many people get frightened if they may have elevated PSA levels, but it shouldn’t really be an immediate cause for concern because other factors will instead indicate that the prostate isn’t cancerous. Maybe the prostate is just inflamed. So the question is, can we find a marker that complements or maybe replaces PSA tests and is more specific and sensitive?
ISM: How does the physics of fluids play into this new work?
GS: Exosomes have been found in virtually every fluid in the body, therefore the physics of fluids can play an important role in separating the different components, including exosomes, from those fluids. As it turns out, the nanoDLD lends itself very well to separating, identifying and quantifying exosomes from fluid samples. There are technical difficulties with the isolation of exosomes in that they’re very small, and we believe our technology can solve those difficulties. When it does, we’ll make big step towards using exosomes in a platform that could be used at the doctor’s office, which would make a big contribution towards liquid biopsy analysis.
ISM: So this will be used in clinical settings?
GS: That’s the idea and vision behind what we’re doing: A person goes to a doctor and gives a blood or urine sample. This sample then goes into our device, where the solution gets processed in such a way that the exosomes are separated and interrogated with respect to the markers they have. In the best case, you don’t have any markers that should be of concern. In the worst case, a few markers show levels that are in a range that indicates that we should do more testing. So the advantage of this is that you could look at many markers simultaneously corresponding to several cancers or even other diseases in the same assessment very fast and cheap.
ISM: How quickly would the results come back?
GS: Ideally, within hours. So on the same day of the appointment—and maybe in the doctor’s office itself—we should know the results. Another possibility, and this is a long-term vision, is that patients could test themselves at home every so often, and the information is transmitted wirelessly to your local cloud or to the cloud of your doctor or your doctor’s clinic where it’s stored and then mined against other information that might determine that it’s time to visit your doctor. Of course, many hurdles exist to make this a reality, but it’s not inconceivable.
ISM: Are you collaborating with cancer researchers to determine the significance of particular exosome-derived markers?
GS: Yes. For example, we’re collaborating with researchers in the pathology and urology department at the Icahn School of Medicine at Mount Sinai, where I’m an adjunct professor. We’re working on trying to find markers of prostate cancer with them. They have a lot of input in terms of what to look for and gathering a group of patients that will eventually be needed to test whether we’ll find more of a particular marker in patients that have more aggressive forms of cancer compared to other patient samples. We’re contributing the technology that will enable those markers to be fleshed out using exosomes and then quantifying those markers. We share some lab space over there where I spend two days a week. I don’t think we could understand the very complex field of cancer biology and clinical practice without their assistance.
ISM: Are the medical professionals you’re working with excited about this type of technology in relation to their own research?
GS: Yes. Absolutely. The momentum behind the concept is looking at people when they’re healthy, and not when they are sick, so that you have time to act when you find something that might be troubling. One of my collaborators at Mount Sinai says, “We don’t do healthcare. We do sick care. It would be nice if instead of sick care we can do healthcare.”
Jim Utsler, IBM Systems magazine senior writer, has been writing for IBM since the mid-1990s.