Welcome to GenCast, a sponsored podcast series brought to you by Genetic Engineering and Biotechnology News.
[00:00:07] I’m your host, Jeff Buguliskis. We often take for granted that we live in a three-dimensional world possessing the innate ability to interact and interpret the space around us. Yet for all too long in biology most researchers have had to deal with a much flatter reality while major scientific advancements have been achieved in this compressed microscopic realm in the digital age. Scientists and engineers believe we can and must begin to appreciate the cellular world closer to its unadulterated state.
[00:00:38] In order to better understand human health and disease as such, researchers have turned to new technology that can provide them with a much greater level of spatial information. Spatial information is important because a lot of times in biology you not only need to know what’s there and how much but how it relates to something next door or how communication occurs. And without the spatial information you can’t assess the communication.
[00:01:04] Dr. David Rimm, Professor of Pathology and Medicine at Yale University, in his laboratory focuses on quantitative pathology as it relates to therapy response predictions in breast and lung cancers.Whether a biomarker is in one type of cell or another can’t be distinguished necessarily just by grinding it up and looking at how much of it is present.
[00:01:24] Whereas if you use an in situ analysis (an analysis that conserves the spatial information) then you can tell which cell it came from or which subcellular compartment it’s in.
[00:01:35] David and I chatted recently about the impact of spatial information in pathology and how new technology like the GeoMx™ Digital Spatial Profiler from NanoString Technologies has been aiding and improving his research.
[00:01:47] There’s been a longtime interest in spatial information, it just hasn’t made it outside the pathology lab so well. I’ve been interested in it for the last 20 years and in fact pathologists derive their diagnostic information, which medical care is given in cancer, on the basis of spatial information. That is when pathologists look at a glass slide. It’s the architecture that we’re looking at that’s giving us our diagnosis.
[00:02:14] David feels that there is a strong link between the recent success with immunotherapies and the increased demand for spatial information platforms.
[00:02:21] As you know, that new type of therapy has really revolutionized our ability to manage patients with high stage disease, and so that therapy works based on a lot of variables- we don’t need to just look at one target. We need to understand the microenvironment around that tumor and that means we not only look at targets in the tumor, but we need to look at targets in the tumor microenvironment as well. And to tell the tumor from the microenvironmentwe need the spatial information. That’s what I would attribute to the real surge that we’ve seen of interest in spatial information over the last two years.
[00:02:55] While certainly no would argue the impact that immunotherapy has had on clinical research and the life sciences industry, I was curious to know some of the history behind spatial analysis techniques, and how David envision they might change in the near future.
[00:03:09] The big change that I think we’ll see in the future is the plex or the number of molecules we can look at one time. When Clive Taylor invented immunohistochemistry in the 70s or 80s is when we really started to see it take off in the research world. To look at not only how much was expressed but where it was expressed. The next step was quantitative fluorescence or Multiplex fluorescence where with a fluorescence microscope you could get up to five and then seven and now nine different fluorphores and then by cycling fluorescence you could get up to maybe as many as 30 or 40 different colors. But the cycling had problems of its own. So that’s become less popular. Then more recently there’s been methods of tagging antibodies with heavy metals. The MIBI technology or the IMC imaging mass cytometry technology gets us out to 60 or maybe even 70 different multiplexes and then the highest-level multiplex is a Digital Spatial Profiler instrument which has a theoretical limit of 800 or maybe even no theoretical limit if you do identification by sequencing.
[00:04:12] As a frequent user of the GeoMx platform I asked Dave what it was about the technology that initially caught his attention, and what are some of the benefits that the system has over the others that are currently available.
[00:04:24] The digital spatial profiling approach is based on molecular definition of compartments. So, what does that mean. It means that you don’t define a cell in the traditional cell segmentation way where you have a machine try to mimic a human and define what the cells are. Rather, you use a molecular interaction to define a cell instead of a shape. And so, for example, a nucleus is a DAPI positive pixel instead of a round thing in the middle of the cell. Lymphocytes might not be just considered a relatively small blue cell but would be considered CD8 positive pixels. And that is what I mean by molecular definition of compartment. And that’s what the GeoMx Digital Spatial Profiler does. You can define compartments by hand or regions of interest by just drawing them, but you can also use a molecular stain or traditional immuno-fluorescence as a mechanism to define a region of interest. That in particular caught my attention. And then also the fact that it essentially had unlimited multiplexes, and it’s always been kind of a goal to 100 plex and the highest we could get was 6 or 8 and then we could get up to 10 or 30 or 50. I think that it’s safe to say that we are beta users of the GeoMx, and I hope that before the year is out we’ve done 100 plex but nothing has ever really worked to allow us to do that for proteins. Certainly reverse phase protein analysis (RPPA) has worked in that way but without maintaining the spatial information. Now this gives us a tool to screen multiple – maybe not 20,000 yet but maybe 100 and soon 1000 different antibodies in one assay. Using our plus condition or minus condition so to speak; using two sets of conditions, treated and untreated, and then look to find the differences. That’s what the power of the tool is, for screening. We’ve had great screening tools for RNA or for DNA, but this is really one of the first screening tools that can work on protein.
And how has David’s experience with the instrument been in the short time that he’s had it?
One of the things that is interesting, I guess it surprised me a little bit, is how low the apparent background is. So, with chromogenic staining you get about one log of dynamic range- that means you can tell 1 protein from 10 proteins but once you get above 10, 12, 14, it still looks like 10. You can’t tell that the dynamic range doesn’t expand and fluorescence had a broader dynamic range maybe out to two logs. It looks like because of the very low background of this technology and the very low number of counts that are spurious counts, we’ll probably have at least four logs of dynamic range and that is really unusual to have any measurement device that gives you that much dynamic range. So, I’m pretty excited about that possibility- being on the cutting edge as well as a strong advocate of quantitative pathology.
I wondered what David saw as he looked out toward the future of this field that he has championed for more than 20 years.
[00:07:17] Well I think that we’ll see spatial information increase the sensitivity and specificity of being able to discover and then predict the two conditions, that is, who’s going to respond to therapy or who benefits from a given drug. Those sorts of things, the same kinds of things we’ve been getting from sequencing information and from RNA Seq will now yield to benefit from the capture of spatial information. I think quantitative pathology is important but it’s a tool we need to get to the point where it helps patients. And I really think that means getting the quantitative pathology tool into the CLIA lab and into those CAP certified labs, so that the assays can be used to actually improve patient care. And I would add I think that’s especially true in the adjuvant setting for cancer where the patients may well be cured by surgery, but it’s really important to decide which ones also need the drug and which ones aren’t going to benefit. Because mostly in the adjuvant setting most of the patients actually don’t benefit, but we give the drug to 10 patients when only 2 or 3 will benefit because we want to make sure we don’t miss anybody. But I would argue now that some of the drugs have more impressive toxicities, we want to make sure we don’t give drugs to patients who aren’t going to benefit. And I think that’s the main cause I’ll be championing in the next 10 years.
[00:08:44] Thanks for listening to GenCast for genetic engineering and biotechnology news. I’m Jeff Buguliskis.