Understanding the Tumor Microenvironment to Unlock the Key to Oncology Drug Development With Jesus Garcia, PhD

This interview features Jesus Garcia, PhD, a tissue and liquid biopsy expert who analyzes trial sponsors’ clinical protocols and aids them in developing relevant biomarker strategies for their trials. Dr. Garcia discusses his passion for understanding the tumor microenvironment, its implications in drug development, and breakthrough strategies in pathology and technology that are leading to expanded therapeutic options for patients.  

Oncology Data Advisor: Welcome to Oncology Data Advisor. Today, I have the pleasure of being joined by Dr. Jesus Garcia. Thanks so much for coming on today.

Jesus Garcia, PhD: Thank you.

Oncology Data Advisor: Would you like to introduce yourself and share a little bit about what your work and your research focus on?

Dr. Garcia: Sure. So, I’m originally from Mexico. That’s where I went to college. I originally came to the United States for grad school. I did all my research in collaboration with the Houston Methodist Research Institute. I rotated through different labs, and I ended up in the pathology lab there. After graduating with my PhD, I got the opportunity to work at MD Anderson Cancer Center as a scientist there in one of the laboratories under the umbrella of Dr. James Allison, which was a privilege. It was the immuno-oncology platform over there. That’s where I started learning and developing skills as a scientist, specifically in the biomarker space, working in assays for clinical trials. I dove into new technologies, not only immunohistochemistry and multiplex immunofluorescence, but I got to learn about image analysis, digital pathology, and more.

After a few years there, I switched into industry, into a small laboratory which provided lab services for clinical trials, mainly for biotech or biopharma. It was mainly specialty lab services around tissue and liquid biopsy assays. I was a scientist there for about a year or two before I transitioned into more of a subject matter expert role. And that’s what I’ve been doing for the last four or five years. The part that I play in the research we do at Precision for Medicine is to really dive into the sponsor’s clinical protocols and see how we can help them develop an appropriate biomarker strategy. Even though it’s not exclusive, the majority of our projects are oncology, and a big part of that is immuno-oncology. That’s why a lot of my work has been in the tumor microenvironment. So, that’s a short introduction.

Oncology Data Advisor: Awesome, I’m excited to hear more about this research too. For background, what is the tumor microenvironment, and what is the role that it plays in tumor growth?

Dr. Garcia: Historically, we have been very simplistic when approaching cancer and developing therapies—and I’m talking some time ago. It has been a very simplistic approach, taking into account understanding a tumor cell and how it responds to X or Y therapy. That’s why we came up with different, not necessarily targeted therapies, but therapies that were not taking into account the cancer or tumor as a whole or as a complex system. That’s when research starting gearing towards, “Okay, why are a lot of these therapies failing?”

Then we started trying to understand that the tumor is not an isolated system. It exists in a more complex environment with a tumor microenvironment where there are a lot of things that play a role. Some of those things can be different proteins like cytokines or things like how much oxygen is present, the pH or acid involved in the tumor microenvironment, and obviously a lot of different, potentially impactful cell types—not only tumor cells, since that’s what we’ve been focusing more on for the last couple of decades, but also other cells like cytotoxic immune cells, suppressive immune cells, and different subtypes of these cells.

We’ve learned to look at the tumor and the tumor microenvironment as a sort of super-Darwinian ecosystem where survival of different clones of these tumor cells matters and how they survive to their environment maybe even impacts therapies or the immune system. So, that’s a little bit of what the tumor microenvironment is. It’s everything around the tumor involving other cells, necrosis around the tumor, and blood vessels around the tumor. And all of that is going to have an impact on not only whether the therapy is effective, but whether it’s delivered, whether it’s going to have unintended consequences, et cetera.

Oncology Data Advisor: Thank you, that was a really helpful overview in understanding it. So, what are some of the current unmet needs regarding understanding the tumor microenvironment?

Dr. Garcia: One of the main things is that the tissue availability is always an issue. That’s because in order to obtain tissue to analyze it, a patient has to undergo an invasive procedure. Some of those can potentially present a significant risk depending on where the tumor is and the adjacent structures. Lung biopsy can be very, very dangerous just by itself. So, it’s always a challenge to obtain enough tissue to explore the tumor microenvironment.

Then there are limitations of technology. Immunohistochemistry (IHC) is one of the gold standards that has been used by pathologists and clinicians. But even though that’s still the gold standard for many purposes, it’s very limited, and it’s potentially wasteful because you’re looking at one protein, and you’re using one whole slide. Depending on the size of the tissue, you might have just a handful of slides available for you to get a better understanding of the tumor. Now, even though it’s still very useful, a limitation has been, “Okay, how can we make better use of the limited amount of tissue that we sometimes have?”

Even though we’re making great progress, speaking again about proteins, for example, we have multiplex immunofluorescence approaches where now we’re able to look at many more proteins at a time. Immunofluorescence is not a new technology, but the number of proteins that we can analyze at a time, the sensitivity of the assays, the throughput, and just the quality in general of these technologies keeps increasing. Now we have multiple technologies where you can, in theory, measure hundreds of proteins. It’s like doing 100 IHC assays in one slide.

Now, the challenge that presents then is data analysis. Before, scoring an IHC slide was relatively simple, for a pathologist, of course. They had to look at a handful of areas, quantify the expression of these proteins, see how many cells, assess the intensity of the expression of this protein, and then come up with a scoring, whether it’s a H score or a custom scoring system. But it was relatively simple. Once you turn that into having not only 1, but 2 or 10 or dozens of proteins in the same slide, we want to do the same thing but on a different scale. And it’s not only doing the same thing times 20 or 30 times, but now we also want to assess, “Now that we have those 30 markers, how do they relate to each other?” Now we can define potentially more complex phenotypes of cells. Before, we could say, “We have this score for CD8 cells.” Now we can say, “We have a CD8 cell, but that is expressing marker X, Y, and Z.” We can assess those cell types and in context of other complex phenotypes.

It really becomes not only the complexity of the data and the analysis that needs to be done, but it’s exponential. It’s not like scoring 30 IHC slides. The complexity of the relationship among all those different proteins or cell types becomes much, much, much more complex than the sum of 30 proteins. That’s one of the main limitations right now. The field is starting to come up with solutions on how to even ask the right questions, because when you have just a handful of proteins to analyze, it can be not as difficult to ask how many CD8 cells there are. But once you have all those 30 markers or more, then phrasing a meaningful question that is going to be useful for your patient ultimately becomes a real, real challenge. So, I think that’s one of the main limitations we’re in now.

The same goes for genomics-based assays. Genomics produces a lot of data. But now that we’re also starting to use next-generation sequencing (NGS) or other molecular tools to analyze the tumor microenvironment in the spatial context, because we have always done bulk sequencing, but now we’re starting to add the spatial variable to it, the same thing happens. Now that we’re sequencing all these different cell types and we are trying to take the variable of spatial context into account, then it just becomes exponentially more complex. Those are probably the main limitations right now.

Oncology Data Advisor: In light of all these complexities and the solutions that are being made to address them, how can an improved understanding of the tumor microenvironment impact oncology drug development, and ultimately patient outcomes?

Dr. Garcia: Where we can see that understanding of the tumor microenvironment can be more impactful is in immunotherapy. That’s the textbook example right now, because, of course, immunotherapy uses the immune system to treat the cancer. In order to understand how the immune system is going to respond once the therapy is given to the patient, we need to understand the spatial environment and the spatial context of the tumor. We classify the immune infiltration in the tumor as either hot or cold, but it’s really a continuum. It’s not just hot or cold. It’s the degree of immune cells involved in this region in the tumor microenvironment or the lack of these immune cells. Obviously, if we see that there are no immune cells and our therapy needs immune cells, we either need to recruit immune cells to the tumor or decide that this therapy might not work for this patient.

Sometimes, it’s even helpful to know that we’re not even going to select this patient for this therapy because it’s very unlikely that it will work for them. Of course, we can save costs, but we can also save the potential secondary effects that might impact the patient without actually giving them any benefit, hopefully so they can move on to the next possible alternative. Understanding the immune cells is the easiest example—where they are and if we have the target phenotypes that we are interested in, because immune therapies can target different proteins or different cell types. They might want to involve more regulatory cells or more cytotoxic cells or other cell types that might be involved, like macrophages.

We have to understand how many of my cells of interest are there, and where are they? If they are outside of the tumor, if there’s no infiltration, we might need to say, “Maybe it’s not going to work because something is preventing the immune cells from actually working where they should be working.” But if we see the right conditions either before or after therapy, we can see that we have the correct cell types. We see the protein and the targets of interest in the appropriate regions. Then we have a better chance of patient responding better to these therapies.

In our case, at Precision, we are hoping to help sponsors with developing assays that they’re going to use in their exploratory end points at the beginning. We’re going to provide this type of information or data so the sponsors can learn, or it can help the sponsors explain the results of what they are seeing after therapy, whether it’s correlating with survival or with other measures. We are trying to provide them with that information. Hopefully, if things go well in the trials, we can develop assays that are really going to help them achieve this precision medicine and help them come up with appropriate subjects of patients who are more likely to respond to their therapies. This is one way of how assessing a tumor microenvironment can really help with precision medicine.

Oncology Data Advisor: Awesome. Speaking about these trials that are underway or that are planned, what is some of the progress that you hope to see in this field within the next few years?

Dr. Garcia: Right now, of course, there’s a lot of traction with the typical checkpoint inhibitors: programmed death-ligand 1 (PD-L1), cytotoxic T-lymphocyte–associated protein 4 (CTLA-4), et cetera. But for the typical therapies, for the first couple of checkpoint inhibitor therapies, the tumor microenvironment was not assessed with the technologies that we have right now. Now we see a lot of the sponsors trying to use these technologies to improve therapies that have already been doing very well but have somehow failed to provide the response that we were expecting at the beginning, at least with some indications of cancers.

Right now, for your typical PD-L1 therapy, we still use a companion diagnostic that is just a single chromogenic IHC, so one target. The way we score that is relatively still old-fashioned, with immunohistochemistry, which is still the gold standard. But it lacks the tools or the technologies that we now have available. I see a lot of clients or sponsors using multiplex immunofluorescence. Maybe they want to understand not only PD-L1 expression in tumor cells also how it relates to regulatory cells around the tumor and cytotoxic cells around the tumor. It’s trying to use more complex ways of selecting patients that are more likely to respond to the therapy—which is already good just by doing an IHC, but if you use the current technologies to do a better selection of those patients, we’re more likely to have better results.

There’s also human epidermal growth factor receptor 2 (HER2). There are very good advancements with the low-expression HER2s that were previously classified as negative, but a group made some awesome advancements where they saw that some patients that had low expression of HER2 could potentially still get benefit from targeted HER2 therapy, again, using better technologies or more advanced technologies and digital pathology.

Even in IHC, another thing that has been implemented is helping the pathologist with the scoring system. There are some companies that have used artificial intelligence (AI) to help the pathologist with better scoring. For example, the pathologist might be focusing on just a few fields of view to score, while the AI algorithms can do the entire slide. Sometimes AI can distinguish smaller differences in expression or counts, or they can count every single cell in this slide versus the pathologies that rely on a smaller sample. There have been some studies where, retroactively, they say that if they had used the AI to help the pathologist enroll patients, they would have enrolled more patients that were going to respond to the therapy versus what they originally did, which was enroll with the pathologist scoring method. We’ve already seen those improvements and how using more technology to better assess the tumor microenvironment is going to help with, mainly, immunotherapies, but not limited immunotherapies.

Oncology Data Advisor: Thank you, that’s another really great overview of all the progress and the upcoming advances. Anything else you would like to share on this topic?

Dr. Garcia: No, just thank you so much for having me. This is a great opportunity. I always love talking about what I do, in not only cancer, but all the advancements in the tumor microenvironment, digital pathology, and even AI, and how it’s helping us come up with better treatment for patients. I always like the opportunity to discuss and talk about these things, so I just wanted to say thank you for having me.

Oncology Data Advisor: Yes, absolutely, and we look forward to hearing more about the research in the future as well.

Dr. Garcia: Awesome. Thank you very much.

About Dr. Garcia

Jesus Garcia, PhD, is a Biotechnology Scientist with expertise in oncology, liquid and tissue biopsy assays, and digital pathology. He currently works as a tissue and liquid biopsy expert at Precision for Medicine, where he analyzes trial sponsors’ clinical protocols and aids them in developing relevant biomarker strategies. In addition, Dr. Garcia partners with biopharma to implement digital pathology and artificial intelligence into the drug development process.

Transcript edited for clarity. Any views expressed above are the speaker’s own and do not necessarily reflect those of Oncology Data Advisor. 

Related Articles


Your email address will not be published. Required fields are marked *