Read the Full Video Transcript

Andrea Miyahira: Welcome, everyone. I'm Andrea Miyahira, and I'm the Senior Director of Global Research & Scientific Communications at the Prostate Cancer Foundation. Today, I'm joined by Dr. Christina Guo. Dr. Guo is the medical oncologist and PhD fellow in Professor Johan de Bono Group at the Institute of Cancer Research and the Royal Marsden NHS Foundation Trust in the UK. Dr. Guo and colleagues have recently published the paper, B7-H3, as a therapeutic target in advanced prostate cancer in European neurology. Dr. Guo, thank you for joining us today and sharing your work.

Christina Guo: Thank you for your invitation to speak about a topic that's close to my heart, so B7-H3 as a therapeutic target in advanced prostate cancer. Our manuscript was published in European Urology this month, and there's also an accompanying editorial by Eugene Shenderov and Emmanuel Antonarakis. Just want to also highlight that this work was funded by the US Department of Defense. So B7-H3 is a member of the B7 family of immunomodulatory glycoproteins. Many of you will be familiar with PD-L1, another member of this family. B7-H3 is overexpressed in prostate cancers, as well as a number of other malignancies. On the other hand, PD-L1 is actually infrequently expressed in prostate cancers. B7-H3 is now the target of a number of antibody drug conjugates, antibodies, and T-cell therapies, that are in both preclinical and clinical development.

And overall, its expression is thought to be negatively prognostic in cancer. However, it has an array of functions, including immune stimulation and suppression, depending on the disease context, tumor growth, survival, and metastasis. So our study set out to characterize a pattern and level of protein expression of B7-H3 in paired treatment-naive and castration-resistant prostate cancer biopsies from a number of patients. We wanted to determine the association between B7-H3 and other treatment molecular characteristics, and finally, evaluate the anti-tumor activity of a B7-H3 targeting antibody drug conjugate, DS-7300a, in human castration-resistant prostate cancer models.

So we had 98 CRPC biopsies in total, from a number of disease sites, and 72 of these patients also had accompanying diagnostic castration-sensitive prostate cancer biopsies that were available for evaluation. And we performed IHC for B7-H3 on these samples, as well as multicolor IF for tumor-infiltrating lymphocytes. We also had genomics data from a targeted panel, as well as clinical data from a number of these patients. A subset of patients also contributed to the development of several human prostate cancer organoid and PDX models.

So this is what we found, so we showed that B7-H3 is highly expressed in metastatic, castration-resistant prostate cancers. And in the vast majority of these patients, expression was already upregulated at the time of diagnosis. So we saw no statistically significant difference between CSPC and CRPC in those patients who ultimately developed CRPC. We also showed that B7-H3 three expression was not significantly different across different disease sites, ranging from the prostate primary, to bone marrow, lymph nodes, liver, and soft tissue. Congruent with known preclinical data, B7-H3 appears to be highly expressed on tumor cells, and expression was rarely observed on stromal cells.

Interestingly, whilst there was some heterogeneity in B7-H3 expression in the tumor cells, B7-H3 positive and negative cells were generally within about a 30 to 60 micron diameter distance between each other, suggesting that drugs targeting B7-H3 positive cells may impact its B7-H3 negative neighbors through crossfire or bystander effects, depending on the mechanism of kill. Interestingly, we then went on to look at how B7-H3 expression correlated with tumor genomics, and we observed that B7-H3 expression is higher in patients, those whose tumors harbored BRCA2 bi-allaleic mutation or HomDel, as well as those with either ATM mutation or protein loss. And here are just some examples of IHCs that demonstrate this scenario.

So finally, I went on to evaluate the activity of DS-7300a in a number of prostate cancer CRPC organoids in vitro. So I tested this in three different models. Two of them had fairly high B7-H3 expression, H scores ranging from 150 to 250, and one of the model had no B7-H3 expression. And we observed substantial antitumor activity from the B7-H3 targeting ADC, compared to the antibody alone, or an isotype ADC. And this effect was only observed in the B7-H3 positive models, and was absent from the B7-H3 negative tumor model. And below, you can see this image, which demonstrates nicely that treated with the vehicle or the antibody alone, or the isotype ADC, we don't see any organoid disruption, whereas the B7-H3 targeting ADC had substantial antitumor activity.

We then went on to evaluate this in a number of PDX models derived from human CRPC biopsies. And again, we showed substantial antitumor activity with the B7-H3 targeting ADC, whereas the control compounds didn't have any activity. And consistent with our organoid data, we only see this activity in the B7-H3 positive models, and this is not as observed in the B7-H3 null PDX model. So in summary, we showed that membranous B7-H3 is frequently overexpressed in CRPC, and in those patients who develop CRPC, this expression is often present at the time of diagnosis. Membranous expression is relatively homogeneous in comparison to other targets like PSMA. And when there was heterogeneity, the positive and negative cells were generally very close to each other. B7-H3 overexpression associates with alterations of BRCA2 and ATM, and finally, a B7-H3 ADC, using a topoisomerase 1 payload, has potent and selective antitumor activity in B7-H3 positive human CRPC models.

Andrea Miyahira: Okay, thank you, Christina, for that wonderful presentation. I guess some questions, is B7-H3 necessary for prostate cancer, and do we know what role it's playing?

Christina Guo: Yeah, so B7-H3 has been shown to have a number of functions in other cancer types, whether that's promoting tumor metastasis or immune tolerance. In prostate cancer, there was a paper previously published in a TRAMP-C1 model, which has relatively low B7-H3 expression, where it was shown that B7-H3 actually promoted tumor growth. So I think there is still some contention as to what B7-H3 actually does in prostate cancer. Certainly, B7-H3 knockout in prostate cancer cell lines don't lead to marked inhibition of growth. And I think what that indicates to us is that perhaps the role of B7-H3 is more to do with the tumor cell's interaction with its stroma, which would not be evaluable in in vitro models of tumor cells only. So really, the short answer is that there has been a wide range of functions proposed for B7-H3 in cancer, in general, as far as ... however, I think there is still much work to be done as to exactly what it does in prostate cancer, specifically.

Andrea Miyahira: Okay, thank you. In this study, you compared B7-H3 expression CRPC samples with match diagnostic treatment-naive samples. Did you also look at B7-H3 expression in any primary samples from patients who did not progress to CRPC?

Christina Guo: Yeah, so we did not look at this specifically, as our primary question was really, in those patients who a number of these compounds are being developed for in the setting of treating CRPC, do these patients already have high B7-H3 at the time of diagnosis, and can the diagnostic biopsy be utilized, perhaps as a companion, so for assessment of a companion biomarker? Several other groups have looked at this question specifically, and they have shown that B7-H3 is higher in metastatic tumors, compared to the primary tumors, and it's associated with biochemical and metastatic recurrence. So I think, overall, there is fairly consistent data showing that, in tumors where the patient ultimately develops CRPC, they tend to have higher B7-H3 expression.

Andrea Miyahira: Was there any relationship between B7-H3 expression levels and other cell types in the tumors?

Christina Guo: Yeah, so we actually, in our paper, showed that patients who had high tumor cell B7-H3 expression had a lower CD3 tumor-infiltrating lymphocyte density. So we specifically look at T-cells, of course, others are also looking at how B7-H3 impacts their immune repertoire more generally, and hopefully that data will be available soon.

Andrea Miyahira: Okay, cool. We've seen some evidence that B7-H3 is regulated by AR. What do we know about the relationship between B7-H3 and AR, and does it tell us anything about when to target B7-H3?

Christina Guo: I think that's quite an interesting topic, particularly when we are thinking about combinatorial therapies. In our data, it's difficult to glean that, because between the treatment-naive biopsy and the CRPC biopsy, a number of different treatments would have been administered in addition to androgen deprivation, and therefore, we observed a bidirectional change, and it's unclear whether that's due to ADT or ADT plus, minus a number of other things. In vitro studies have shown ... one paper showed that DHT, for example, so testosterone, essentially represses B7-H3 expression. However, I've also shown other papers where this was not consistently the case. A recent publication also showed that enzalutamide-resistant cell lines have upregulated B7-H3 expression. So I think, at this stage, we don't know whether this is really going to be the case in patients.

Andrea Miyahira: I think you guys also showed that some NEPC samples were left likely to express B7-H3. Do we know, are there any immunologic differences between NEPC and adenocarcinoma CRPC, and whether B7-H3 is playing a role here?

Christina Guo: I think the immunologic difference between NEPC and adenocarcinoma has been described in a number of mouse models, and it has also been observed, I think, in some studies of human cancers. Whether B7-H3 contributes to that difference is unclear.

Andrea Miyahira: And so, if we're thinking about treating patients with the B7-H3 ADC, do you think B7-H3 should be a biomarker for selecting patients, or what other factors should we consider when it comes to patient selection?

Christina Guo: In our study, we showed that B7-H3 expression seems to be necessary for antitumor activity. I want to caveat that by saying that we only had a very small number of models, so certainly, validation in a larger cohort is necessary. Interestingly, the HER3 ADC, using the same payload in the context of patients with lung cancer, actually also showed that expression of the target was necessary for ADC activity. Now, the second question as to, in patients who expressed the target, does the level of expression matter? I think looking at the HER3 data, it seems that there are tumors that have fairly low target expression that can still have fairly impressive responses. And in our two models, there were also differences in antitumor activity from the compound, despite comparable target expression. So again, underlying tumor genomics, molecular characteristics, clearly also play a role.

Andrea Miyahira: And does your team have any other translational plans for the B7-H3 ADC or other B7-H3 targeted agents?

Christina Guo: Yeah, I mean, B7-H3 remains a key interest of our group, and I think some of the questions that you very correctly asked are the things that I would be very interested in continuing to look at.

Andrea Miyahira: All right. Well, thank you so much for sharing this work. I think B7-H3 is a really interesting target in prostate cancer. It's nice to see a new target, especially that might modulate immunotherapy in patients that's being described and being studied. So thank you for sharing this today.

Christina Guo: Thank you.