Immunotherapy impacts: Moving the needle on pediatric glioma and rare cancers with single cell and spatial tools
Immunotherapy is many things to many people. For some, it’s a treatment for the rare cancer that almost no one knows about. For others, it’s the hope that their cancer-stricken son or daughter will get the chance to be a kid—not just a patient. For clinicians and researchers, it’s gratitude: to the patients and their families. To the other researchers in the field. And to the privilege of being able to make a difference in cancer outcomes.
We were fortunate enough to be able to spend some one-on-one time with two groups of researchers who are making a difference: Drs. Abdul Rafeh Naqash, Tae Gyu Oh, and Hassan Abu Shukair (University of Oklahoma), and Dr. Sneha Ramakrishna (Stanford University). We learned more about the clinical trials that have been central to their work, how immunotherapies have already made a difference in their cancers of interest, and the tools that are helping them turn hope into action.
How Drs. Naqash, Oh, and Abu Shukair are writing the book on a rare sarcoma & why rare cancers aren’t rare
Can you tell us more about your research on rare cancers?
Dr. Naqash: I’ve been involved in a rare tumor initiative and focus on alveolar soft part sarcoma, or ASPS. ASPS is a very rare soft tissue sarcoma that didn’t have any incident-of-care therapies until 2022, where we had a multicenter study through the National Cancer Institute that led to FDA approval for an immunotherapy drug called Atezolizumab.
What was interesting was that the sarcoma benefited from single-agent immunotherapy, which most sarcomas don't. So we embarked on this multi-year project that I started a couple of years back to answer why this tumor is responsive to immunotherapy.
The trial used samples from the NCI and assessed how the immune cell changes take place before and after immunotherapy treatment. One of the concepts that came out of that research was that, since ASPS features a unique [oncogenic] TFE3 fusion with two subtypes, perhaps the two different subtypes have different functionalities and different changes in the tumor microenvironment.
In our SITC poster we were trying to expand on this work with the help of my collaborator, Dr. Tae Gyu Oh, and my post-doc Dr. Hassan Abu Shukair. We looked at some public RNA-seq data sets and tried to understand the gene expression differences. We also had a clinical trial data set from the NCI trial where we looked at the type one and type two fusions and tried to understand if there's also differential activity of the immunotherapy.
Can you tell us more about the clinical trial?
It started from a simple observation: someone treated one ASPS patient with immunotherapy because there were no standard therapies. That led to a benefit in that individual, which led to the clinical trial. The trial took 5-6 years to complete, with 55 patients included. It’s a very rare tumor, so 55 patients was a big sample size, and it led to the FDA approving this treatment.
You mentioned looking at RNA-seq databases for the basis of this work. What led you to take the next step and incorporate spatial transcriptomics?
Dr. Naqash: Sequencing is good as a broad approach to find different immune changes or expression changes, but it’s not high resolution. I didn't know a whole lot about spatial a year back, but Tae and Hassan helped me learn a lot about spatial transcriptomics.
I met Tae when I began to look for good collaborators and we kind of bonded through this project. So I'll let Hassan and Tae speak to why we used spatial transcriptomics and 10x Genomics’ tools.
Dr. Oh: My laboratory focuses on spatial transcriptomics as well as computational biology and machine learning. I’ve used 10x Genomics tools extensively, especially single cell, but there’s still one caveat to all single cell technologies: we need to dissociate the tissue sample, which can cause damage.
We want to know where our cells are spatially, but another thing that’s quite important is that spatial technologies don’t require dissociation. You just prepare the tissue, exactly the same way that pathology departments do it, and then we can visualize and digitize tissue.
For ASPS, Hassan and I are in the process of doing an integration from the public databases: we want to combine public RNA-seq databases with single cell RNA-seq and spatial tools. It’s a very exciting time, and we’re hoping to get fantastic outcomes soon!
Dr. Abu Shukair: What’s unique to ASPS, as a rare cancer, is that we don’t have single cell data for ASPS, let alone spatial single cell data. Having a spatial single cell atlas of ASPS would be very helpful to the field, especially since this type of cancer has immense tissue heterogeneity. Adding that layer of spatial data would be extremely valuable, since it would act like a microscope or a spatial lens into the immune contexture.
Speaking of tools, something that really stuck out is that we usually see one spatial tool used. You may be the first group we’ve seen that’s used Visium HD, Xenium, and Xenium Prime 5K together. What led you to use all of these technologies in the same project?
Dr. Abu Shukair: We want to get the best of both worlds: the comprehensive whole transcriptome coverage, and the subcellular high resolution with Xenium. We want all the data and, especially with rare cancers like this, it's super useful to have.
This approach could help guide potential upcoming treatment modalities, especially with spatial as some groups are marketing spatial to be a high-resolution screening tool for drug development. We’re already seeing some papers using this model.
I think combining both for ASPS, specifically with our hypothesis being that not all fusion isoforms are responding similarly, could be a very useful compass to help us delineate some potential novel combination of treatment strategies.
So what’s next? Where do you go from here?
Dr. Naqash: We want to optimize how we try to obtain single cell resolution in our spatial studies, such as on our tissue blocks or potentially the FFPE biopsy slides that we will soon acquire from the NCI trial. We want to look at pre- and post-immunotherapy to try and understand the tumor microenvironment at a subcellular and single cell level.
We’re also trying to look at novel antigens in ASPS and see if we can develop therapies against these using cell therapy, TCR approaches, or perhaps ribosome degradation approaches.
I also want to highlight that rare cancers constitute 20 to 25% of cancers. That’s not a small percentage and, an interesting thing in my experience, is that rare cancers in younger individuals usually have some actionable targets.
We’ve been able to target DNA mutations, we’ve been able to target RNA, we’ve been able to use antibody conjugates, but we’ve not been successful in targeting fusion proteins. So perhaps the next breakthrough in the space of oncology will be these oncogene fusions, which we’ve found more and more of as sequencing technologies have improved.
While you have this platform, is there anything else you’d like to draw attention to?
Dr. Abu Shukair: Donate to rare cancer research.
Dr. Naqash: Rare cancer treatments are a large unmet need. It’s important to support rare cancer researchers, not just in the US but worldwide, because there’s no big NCI biotech initiative for rare cancer biobanks. There’s an online molecular tumor board, where they will help work with a company to get compassionate use drugs for individuals with rare alterations in rare tumors, but it’s for US patients only.
It's really important to globalize these kinds of initiatives and to try to have much greater access to these therapies. It’s hard to do trials for rare cancers, ones that have an incidence of 1 in 50,000, or 1 in 100,000.
So, overall: I think more rare cancer [samples] via biobanking, more funding opportunities within the NCI or NIH for rare cancer research, and trying to build this as collaborative initiatives with private partnerships, patient advocates, and the public. This will help us get to the next part of cancer research.
How Dr. Sneha Ramakrishna is moving the needle on pediatric glioma therapies
It’s nice to see you again after your last interview! Would you mind giving a quick recap of your work?
The project that we're discussing is in the context of a GD2 CAR T-cell trial for treating pediatric patients with diffuse midline glioma, which is a universally fatal tumor. There are no curative options for these patients and, unfortunately, the median overall survival time is about 11 months.
Not only are there no treatment options, but it's really very difficult to control this tumor even for a short period of time. The only treatment that these patients get is radiation therapy and following that, there's no other treatment option available for these patients.
Thanks to the generosity of patients who have donated their brain tissue after they’ve passed, we’ve been able to learn more about the biology of these tumors. In a collaboration between Drs. Michelle Monje and Crystal Mackall, led by Drs. Chris Mount and Robbie Majzner, our team at Stanford identified that GD2 was highly and uniformly expressed in diffuse midline glioma tissues. We showed we could target these tumor cells using GD2 CAR T cells in mice. And we were able to open a clinical trial and bring this to patients.
The translation into patients has been extremely exciting, and our paper was published just this month!
Can you give an overview of what this treatment has meant for your patients?
Our patients had a much longer survival time than the 11 month average, but it's not just about time. It's about the quality of that time.
These are children. They’re going from not being able to walk to being able to run, to learn how to ride a scooter. A kid who wasn't able to walk down a hallway can now go on a one mile hike after getting these CAR T cells. These are meaningful clinical benefits that our patients are experiencing.
Would you be able to tell us a bit more about the clinical trial?
I came into the study basically saying, okay, how do we learn from our patients? How do we really understand the immune biology of what will make these CAR T cells work—or not—for our patients?
So we collected patient cerebrospinal fluid (CSF) samples at specific time points throughout the treatment course. We realized, however, that people weren't capturing the cellular population when they tried to do flow, both because of the limited cell counts from CSF and because freezing the samples reduces viable recovery.
We took the sample straight from the bedside to the bench and processed it fresh on the Chromium single cell RNA-seq platform. This allowed us to capture the cellular populations consistently over the course of the treatment and it gave us insight into what immune populations contributed to whether the CAR T cell worked in our patients.
From this robust dataset, we saw that the predominating populations are myeloid and T cells. They fluctuate over the treatment course and and we’re working on figuring out the specific immune populations that contribute to patient response versus when they stop responding and patients start to progress.
You also touched on spatial transcriptomics in your talk last year at SITC. What made you go the spatial route for your work?
When we got CSF samples, the question was whether they’re representative of what's happening in the brain and the tumor. Spatial transcriptomics is really one of the most sensitive ways to delve deep into that question, and we were looking for the same immune populations in the tumor tissue, so we collaborated with our 10x Genomics colleagues to create a custom Xenium panel based on our CSF findings.
In our initial validation of this panel, we've been able to consistently see our CAR T cells. That’s wonderful because it lets us see if there are CAR T cells in the tumor tissue at the time of collection, and it also lets us define the different immune populations at a single cell level.
We’re actually in the process of running the patient samples and ramping that up. We’re hoping that in the first quarter of 2025 we'll be able to have that data to start to really delve into what's happening.
What made you choose Xenium and Visium, specifically?
Well, as you can see from our single cell RNA-seq, 10x Genomics’ assays have been really robust, powerful, and we've been really pleased with the data quality. So when thinking about a spatial transcriptomic platform, we wanted to make sure we picked one that would provide us with similar quality and depth of data.
We started with Xenium as the first step, and now we’re superimposing Visium HD over it. We'll see, the data is coming, but we’re already able to identify these cell populations even on our first runs. We can see them, and we can also define them based on our single cell RNA-seq CSF dataset. Since both platforms are complementary, we can superimpose the CSF dataset knowledge on the spatial data outputs and understand the distinct and overlapping signatures between these two types of samples.
If we're going to learn from the patients, we have to learn everything we can from all the tissue we get. This is the best way that I can think of to honor these brave patients.
So where does the clinical trial go from here?
The paper we just published is just the first cohort of patients. In pediatric oncology, if you're not curing a patient you're not doing enough. We will never settle for less than a cure. But this trial finally moved the needle. We finally made a difference in a disease that we’ve literally never made a difference in before.
In fact, in our paper, we have a patient who was a complete responder: not only did the tumor go away, it hasn’t recurred for over three years. This is a patient who, when we met them, they weren't sure if they were going to graduate from high school. Now they're in college and living their best life. They actually gave an interview recently.
This is the dream of something that we could do for a single person. And hopefully, by taking what we learn from this individual, it becomes something we can do for all of them. That’s the power of incorporating these assays into our studies, because we can use the knowledge they give us to keep pushing that needle forward. And we won’t stop until we reach that goal.
Paving the road to a brighter future, one trial at a time
Treating, and dealing with, cancer has always been a long, scary, and dark road for both the patient and their families. But as the above researchers continue to demonstrate, immunotherapies are making the future brighter for cancer patients—and we can’t wait to see what more is coming down the pipeline.
Want to see more of how single cell and spatial tools are helping tackle the most difficult problems in cancer? See the highlights from AACR 2024!