Natural Killer Cell Infiltration in Prostate Cancers Predict Improved Patient Outcomes - Nicholas Zorko

March 20, 2024

Nicholas Zorko discusses the promising role of natural killer (NK) cells in prostate cancer, described his team’s study published in Prostate Cancer and Prostatic Diseases. The research reveals that increased NK cell infiltration in prostate tumors correlates with improved patient outcomes, suggesting NK cells as a potential next frontier in cellular therapies for prostate cancer. The study utilized patient biopsies and advanced sequencing techniques to identify and correlate immune populations, including NK cells, with patient survival across various cancer types. Notably, prostate adenocarcinoma showed one of the higher levels of NK cell infiltration. The findings also highlight that the site of NK cell infiltration, specifically in primary prostate tissue, significantly impacts patient survival. This comprehensive study not only sheds light on the pivotal role of NK cells in prostate cancer but also paves the way for future explorations into NK cell-based therapies.


Nicholas Zorko, MD, PhD, Assistant Professor of Medicine, Division of Hematology, Oncology and Transplantation, University of Minnesota, Minneapolis, MN

Andrea K. Miyahira, PhD, Director of Global Research & Scientific Communications, The Prostate Cancer Foundation

Read the Full Video Transcript

Andrea Miyahira: Hi everyone, I'm Andrea Miyahira at the Prostate Cancer Foundation. With me is Dr. Nicholas Zorko, an assistant professor at the University of Minnesota. He will discuss the paper "Natural Killer Cell Infiltration in Prostate Cancers Predicts Improved Patient Outcomes" that his group recently published in Prostate Cancer and Prostatic Diseases. Dr. Zorko, thanks for joining us today.

Nicholas Zorko: Thank you for having me today, Andrea, and with UroToday. We're very excited about this paper that was just published in the last few weeks here, and it really focuses on our interest in natural killer cells and natural killer cell therapeutics to treat prostate cancer and other GU malignancies here at the University of Minnesota. This was done with a team at the University of Minnesota, but also with the Caris Precision Oncology Alliance as well. So, just a little bit about natural killer cells. Many people are unfamiliar with them, but we think that they're going to be the next wave of cellular therapies. So, they're approximately five to 15% of circulating lymphocytes. Importantly, they're not major histocompatibility complex-restricted. They don't have clonotypic receptors, so they are not antigen-specific themselves and they don't require priming. They have preformed granules and they're ready to go as soon as they encounter an antigen.

They're typically defined by CD16, CD56+, and CD3-. And the advantages of NK cells and why we really like NK cells are there's no graft versus host reaction, minimal cytokine release syndrome risk, and we can make them antigen-specific with monoclonal antibodies or other engagers. So, let me talk a little bit about the workflow for this paper in particular. So, this started with a patient biopsy, whether that was a prostate or a metastatic site biopsy. So, that was then sent to Caris Life Sciences where tumor microdissection was done. So, this was done on an enriched tumor sample. And then whole exome and whole transcriptome sequencing was completed. And with that, then we used a technique called immune deconvolution using quanTIseq, which had been developed and validated extensively to determine the immune populations in this particular study. And then we correlated this immune deconvolution with DNA and RNA mutations from the whole exome and whole transcriptome sequencing that we obtained.

So a little bit about the data here, moving into what would be the first figure. Our first question was, are NK cells actually infiltrating these solid tumors using this technique? We had to answer that question first to see if NK cell therapies are worth pursuing, which we believe they are based on this data, but also if there is a better target tumor that we should be looking at as these therapies are starting to roll out. And this is taking a look at just the median natural killer cell fraction by cancer type. We went through 44 different cancer types plus a pan-cancer, and we found that prostate adenocarcinoma was actually among the higher groups of infiltration, with approximately 5% of the total nucleated cells being natural killer cells. There are also multiple other GU malignancies that we won't go over, but this is a huge dataset that we're really excited to delve into more as well.

When we looked at the upper half of NK cell infiltration, so the upper 50% in this particular case, we wanted to see if those patients benefit in terms of overall survival, and in fact, they did. Increased tumoral natural killer cells improved overall survival in 28 of the 44 patient groups, which was really interesting. And we saw a very strong effect for prostate adenocarcinoma as well, as we can see with a hazard ratio of approximately 0.6. And this was highly significant, as we can see, and that's highlighted by the red arrow. Importantly, this was a very large group. It was 3,365 patients in total. So it was the largest dataset so far of prostate cancer patients that has been investigated for NK cell infiltration. So, we moved to the next step to look, does it matter that the site of NK cell infiltration in prostate tissue, and it does appear to matter the site of biopsy here.

So when we look in the upper left corner, now we're looking at quartiles rather than the median or above the median. So, red is the upper quartile of NK cell infiltration, and we saw that there was only a statistically significant survival benefit in patients who had under or had a sample from the primary prostate. We looked at bone, liver, and other metastatic sites and lymph node; we did not see a statistically significant difference. But there comes a big caveat with this too because we're really at the mercy of the dataset. So many patients who have primary metastatic disease still undergo a prostate biopsy. We were not able to determine the extent of the patient's disease or staging but were only able to determine the site of the biopsy that was reported in that case. So while we're seeing this right now, there are many patients who likely had metastatic disease who actually had a prostate biopsy submitted to the dataset here.

What was really interesting is that we found in patients that had increased natural killer cell infiltration—so again, this is looking at the upper quartile versus the lowest quartile here. So, upper quartile is red, lowest quartile is in blue—we saw that within the prostate itself, there was a counterregulatory immune checkpoint inhibitor signal. So we see all of these different immune inhibitors or immune co-stimulating molecules, and we see they're all upregulated in the group that had the highest number of NK cells. This really made us believe that increased NK cell infiltration led to this counterregulatory calming effect almost on the immune system and is leading us to delve into what other effects or what other mechanisms we can try to use to calm this down and increase the NK cell activity. We didn't see this so clearly in the metastases, but many of these were bone as well. We just didn't see a strong signal. But this also goes with the caveat that I mentioned before. We're not sure of the actual extent of patient staging.

So finally, with the last figure that we'll talk about today, there's increased counterregulatory cell infiltration as well. So when we look at primary prostate populations here, and this was using quanTIseq immune deconvolution, we look at the NK low group, which is the left column in each of these groups in prostate or NK high. And what we see is that there was a significant difference or significant increase in B cells, M2 macrophages, and regulatory T cells in these populations. So this very much fits with a counterregulatory immune calming effect that is going on within the tumor itself that seems to be countering this NK cell infiltration. That was also seen in the metastatic sites as well. But what was really interesting is when you take a look at the total percent of cells, the natural killer cells are approaching 5% when we look at the red bars on the left, but the total number of T cells from the immune deconvolution is only about one to 1.5%. So this really points to the fact that NK cells may be much more predominant even in our low group compared to T cells.

So I'd just like to acknowledge the rest of the team here at the University of Minnesota, including my mentors Emmanuel Antonarakis and Jeffrey Miller. From Caris Life Sciences, our data analytics team with Andrew Elliott, and then the members of the Caris Precision Oncology Alliance who all contributed data and analysis and were authors on this paper. And then my funding sources, I'm really grateful to be a Prostate Cancer Foundation young investigator from 2022, sorry, not 2002. I'm also funded by the Department of Defense, the University of Minnesota Urologic Philanthropy Fund, Masonic Cancer Center, and the ASTCT. I was able to be a part of the Clinical Research Training course as well there, which was really important. So that's everything I have and thank you very much.

Andrea Miyahira: Thank you so much for this presentation, Nick. So, are you able to tell if tumor-infiltrating NK cells associated with improved outcomes are activated and what they are doing?

Nicholas Zorko: Based on this dataset and one of the things that I forgot to mention as I was going over the pipeline, this was not single-cell RNA-Seq. This was bulk RNA-Seq. So, what we can't do is we can't determine the features of individual cells. We would love to do single-cell RNA-Seq, and we're actually actively working on that in our lab or the Greater Miller lab and Antonarakis groups at the University of Minnesota along with Justin Wong's informatics team. So unfortunately, we can't tell. We're based on just whole tumor signals right now from that. And that's why we had to use immune deconvolution as part of that strategy with quanTIseq.

Andrea Miyahira: Okay. Thanks. So, in your dataset, the NK cells were not associated with improved outcomes in the samples from metastases, but they were in the samples from localized disease, and you mentioned this has something you guys are not sure about the staging for all the patients. But do you know why you might see the impact of NK cells specifically in the prostate? And do you think this data suggests that neoadjuvant therapy or early treatment with NK cell drugs might be a better treatment option?

Nicholas Zorko: Yeah, that's a great question. That's something that we were very interested in trying to figure out as we were going through the original dataset. Part of it, as you mentioned, really depends on the staging. We're looking at soft tissue as well. It's far more likely that NK cells and other immune cells are going to infiltrate into soft tissue such as the prostate, potentially lymph node or liver. The total number of patients that were sampled for lymph node and liver were also quite small within those groups. So, I'm wondering if we had a larger dataset there, if we would maybe see some significance within those groups. But I believe that within bone, bone is just difficult to infiltrate at baseline for any immune cell, T or let alone natural killer cells.

In terms of earlier treatment or treatment with neoadjuvant, I think that that's really an area of interest for us because what we'd really like to do, number one, is to treat this more with a curative intent. Once we, knock on wood, show that NK cells are safe and effective in later stages of prostate cancer, we would really like to move these therapies forward to a point where they can be the most effective and potentially treat with a curative intent.

Andrea Miyahira: Okay. Awesome. Have you evaluated whether NK cell infiltration is associated with any responses to immune checkpoint inhibitors?

Nicholas Zorko: Unfortunately, we have not been able to do that yet. That is something also very much of interest for us. As we saw that there was upregulation of PD-L1 in the groups that had increased NK cell infiltration, we're wondering if there's a signature that we can look at to predict maybe patients that have better NK cell infiltration who may also respond to an immune checkpoint inhibitor. Or if a combination of NK cell engagers or engineered NK cell products in combination with a PD-L1 inhibitor or immune checkpoint inhibition may increase the efficacy as well. That's what we don't know. I think it's an intriguing combination because we know for the most part that treating with immune checkpoint inhibitors is not effective in prostate cancer. But we're looking to piece out what else there are specific applications where this might be applicable, especially as we're developing more and more immune engagers, whether they're T cell or NK or engineered cell products.

Andrea Miyahira: Okay. Thank you. And do you know what might explain the associations that you saw between higher levels of NK cells and responses to ADT but not docetaxel?

Nicholas Zorko: That's a great question, and that's really what we're wondering. One of the things that we've looked at, and this really goes back to Amy Moran and her group's paper at Oregon Health & Science University, they found that blocking of AR specifically with enzalutamide in combination with ADT and pembrolizumab, there was an increased T cell effect. We're wondering if there's a similar NK cell effect as well with that. So, whether it's ADT or just the lack of testosterone itself or lack of androgen, or whether it's actually blockade, we're very interested in pursuing that, and there's been other evidence for T cells as well. I believe the group at Fred Hutch is now looking at combining enzalutamide with their CAR-T product as well to boost the CAR-T efficacy. And that would be great to be able to use an existing approved medication to augment these cell therapies, whether they're NK or T cells. It's still a win regardless, and we learn more and are able to apply existing therapies.

Andrea Miyahira: Okay. Thanks. And what are your next steps in these studies?

Nicholas Zorko: So the next step for these studies, we have the other 44 tumor types that we are very interested in. There are three other GU oncology subsets that are in there: bladder, kidney, testicular cancer that we're very interested in. So we want to analyze that data as well. But we're also in the process of developing an NK cell engager targeting B7-H3, which is going to hopefully get to a clinical trial in late 2024, early 2025. So we are using this to show that there was evidence that treatment with NK cells may work for prostate cancer, and then we'd really like to delve into the immune engagers and engineered NK cells as well to treat prostate cancer. So this is just good preliminary evidence and helps point us in a direction of which tumors may be the most responsive and which ones to include in our trials moving forward.

Andrea Miyahira: Okay. Well, I look forward to it, and thanks for joining us and sharing this with us today.

Nicholas Zorko: Yeah. Thank you for having me.