Early Interception: Integrating Tissue-Based Insights to Optimize ARPi+PARPi Combination Therapy in HSPC
SPECIAL LECTURE: Early Interception: Integrating Tissue-Based Insights to Optimize ARPi+PARPi Combination Therapy in HSPC
Patrick Pilié (University of Texas MD Anderson Cancer Center)
Introduction by Christopher Logothetis (University of Texas MD Anderson Cancer Center)
View the Transcript Below:
Early Interception Integrating Tissue-based Insights to Optimize ARPi + PARPi Combination Therapy in HSPC
Christopher Logothetis, MD [00:00:12] So, it’s a real privilege to introduce Patrick Pilié, who’s joined us at MD Anderson and has brought some clarity in bridging that critical difference between mechanism of pathways driving the lethality of this disease with clinical outcomes, and very importantly, trying to do this in the context of moving drugs up earlier in therapy, which has been one of our biggest challenges. As we go earlier in treatment that has served us so well in other disease types, we’ve had a unique problem in prostate cancer that we’ve actually reduced population-based effectiveness because of the noise and the number of patients who do well despite us. So, this will be an interesting talk trying to walk that distance. Patrick?
Patrick Pilié, MD [00:01:15] Thank you so much to Howard and Andrea for the opportunity to speak today. I know I am the last talk between dinner, so I’ll jump into it, but I’m very excited to present today our early data from our adaptive clinical trial for men with high-risk localized disease, where we are using neoadjuvant androgen receptor pathway inhibition and sequential tumor biopsies to select patients for escalation of therapy. These are my disclosures. A brief outline of the talk today. So first, I’ll go over existing clinical and preclinical evidence supporting the use of this combination beyond BRCA mutant prostate cancer and in earlier disease settings in the hormone-sensitive localized disease space. Then I will go over current results from our adaptive trial of men with high-risk localized disease, where we are using on-treatment biopsies to guide therapy escalation. And finally, touch on future directions and correlative plans for the tissues and non-invasive biomarker development. First, for the clinical and biological rationale. As many in this room are aware, there is clinical benefit of combination ARPi and PARP inhibitor beyond just BRCA mutant prostate cancer. While the benefit is greatest in BRCA, there is extension to other HRR mutations as well as beyond HRR deficiency. In addition, the recently published AMPLITUDE study has shown that the earlier application of this combination may improve benefit even further in both the BRCA mutant as well as the HRR broader mutation population. In terms of mechanisms of this combination beyond just canonical homologous recombination deficiency, here highlighted on the upper left is a PDX model that is HRR wild type, AR positive, and through multiple transplantations develops enzalutamide resistance and eventual neuroendocrine differentiation. Application of ARPi on an early transplant of this PDX model, looking at gene set enrichment analysis, we see significant upregulation of MYC targets and E2F plasticity and replication stress pathways. That with the application of combination ARPi and PARP inhibitor, we see downregulation of the MYC targets E2F and DNA repair. This same pathway changes have been seen across model systems, including immunocompetent models and seen in various other tumor types, but whereby AR pathway inhibition is inducing DNA damage, leading to upregulation of PARP 1 and C-MYC with crosstalk between PARP1 and C-MYC and subsequent tissue tumor microenvironment remodeling and immunosuppressive milieu. With the application of the combination, leads to both tumor cell intrinsic cell death through PARP trapping as well as TME remodeling and an interferon response. Clinically, we have seen these pathway changes in the metastatic castrate-sensitive space. This is patients on the combination therapy of Saruparib and Darolutamide in the active phase two PETRANHA clinical trial, whereby a cohort of patients is getting baseline metastatic tumor biopsies and four-week on treatment combination tumor biopsies. With gene set enrichment analysis showing the most significantly downregulated pathways are the MYC target genes, E2F and other cell cycle and replication stress pathways, with significant and quick upregulation of an interferon and immune response at four weeks on combination therapy. Summary in support of an adaptive trial using this combination and localized disease, again, benefit is beyond BRCA with earlier escalation. ARPi and ADT can induce a PARP dependency and PARP vulnerability, and there is opportunity in the high-risk localized disease space. To use post ARPi tissue to select patients for this combination. How we are using ARPi exposed tissue to select patients for this combination in the localized setting. Prior work has shown that androgen responsiveness is a strong prognostic indicator in patients with advanced disease. On the left highlights a presurgical study of apalutamide, an androgen deprivation therapy, whereby patients had prostatectomy following six months of treatment. Nearly all patients on the study achieved undetectable PSA by the time of prostatectomy. However, when looking at retained tumor cell density and morphology on prostatectomy specimens, those patients with higher levels of retained tumor cell density and morphology, indicating less hormone ablative treatment effect, had significantly worse relapse survival compared to those with a more favorable response to the antiandrogen therapy. Under the direction of Patricia Troncoso and Miao Zhang, they then translated these prostatectomy findings to ex vivo prostate biopsies to allow for the application of this information in high-risk localized disease in patients receiving standard, gold standard systemic therapy, and radiotherapy. Into the current active adaptive clinical trial, whereby men with high-risk localized prostate cancer are treated neoadjuvantly with three months of apalutamide and ADT. At the three-month time point, they have an MRI guided fusion biopsy of the prostate, go on to receive definitive gold standard radiotherapy, and following radiotherapy in the adjuvant setting, those patients with an unfavorable morphologic treatment effect to the antiandrogens are randomized to receive escalation of therapy with combination PARP and androgen ablative therapy versus continued gold standard androgen ablative therapy. And those patients with a very favorable on-treatment morphology are not randomized and continue gold standard therapy. Baseline prostate biopsies, three-month on-treatment biopsies, and at end of treatment, liquid sampling is also obtained. Thus far, with the first cohort of patients on this study with 99 patients, that baseline indicators that we traditionally are using in clinic are not predicting outcomes. Approximately 60% of the patients thus far have favorable on-treatment morphology, with approximately 40% having an unfavorable on-treatment morphology. But no differences in baseline between these two groups in terms of T stage, PSA at diagnosis, or grade group. In addition to baseline factors, PSA response is almost universal by the point of radiotherapy and did not correlate with morphologic response to the antiandrogens. A tale of two patients to start to deconvolute this heterogeneity that we’re seeing in localized prostate cancer. These two patients started out almost identical presentation, both grade group 5, both with PSA ranging in 10 to 20, both with T3B disease, both even the same age at diagnosis in their mid-60s, both with similar tumor content on their baseline prostate biopsy. However, at the three-month on treatment prostate biopsy, we have significant divergence, with the patient on the left showing almost absent nucleoli and significant hormone ablative effect, with the patient on the right showing high levels of retained viable tumor, including intraductal morphology. And subsequently, with follow up, the patient on the left is now almost three years from completing treatment with continued no evidence of disease and undetectable PSA and a eugonadal testosterone. While the patient on the right unfortunately developed metastatic relapse nine months after completing therapy. At the point of relapse, PSA was undetectable. And on biopsy of liver lesion, the patient had high-grade neuroendocrine morphology. Rounding out the preliminary results thus far of this adaptive trial with the 99 patients. So far, we have seen no relapse events in the patients with the favorable on-treatment morphology and have seen four relapses in patients with unfavorable on-treatment morphology. With further follow-up will confirm the comparison of apalutamide versus the combo of the ARPi and PARP inhibitor. It is notable that two of the four patients in the apalutamide unfavorable arm did have progression of disease at a PSA of undetectable with again high-grade neuroendocrine on metastatic biopsy. So, the summary of the preliminary trial results is that the on-treatment morphology does inform course correction and allows for patient selection for therapeutic escalation in the localized setting, that the baseline characteristics, including T stage PSA and grade grouping, did not predict for adverse on treatment morphology or the progression events to date. The cancer progression events thus far are exclusive to the unfavorable on-treatment morphology group and have occurred within one year of therapy completion and not all detected by rise in PSA. The diagnostic and on-treatment tissue biopsies with paired serial liquid sampling are banked for extensive correlative analysis with focus on connections between AR and DNA damage response signaling and the tumor microenvironment. In order to take this beyond sequential tissue sampling, we are planning to look at non-invasive monitoring. Dr. Panaretakis at MD Anderson has developed a monitoring tool with EVs, whereby under the pressure of AR treatment, EVs can be monitored for changes in gene expression. Down on the bottom here is example from his prior work showing that with enzalutamide resistant LNCaP in vivo, with further time on enzalutamide, the rise in GR expression in the EVs. And this was recapitulated in clinical samples from patients that were on a presurgical study of apalutamide and ADT, where we see rise in GR expression obtained from EVs at the end of treatment. The application of the extracellular vesicles in localized disease is pertinent because they are present in earlier disease settings and they allow for evaluation of transcriptomic signatures, gene expression changes, that are driving downstream progression. In addition, we have seen in prior PARP inhibitor studies, this is from PETRA, the single agent PARP inhibitor, PARP1 selective inhibitor and metastatic castrate-resistant prostate cancer patients on the study, that ctDNA did predate PSA 50 and resist responses. Looking at it from the other angle, we will assess liquid biopsy sampling to see if ctDNA can be used as a marker of early relapse where PSA may fail and identify the molecular changes leading to therapy resistance, particularly applicable with the great talks earlier today on plasticity is monitoring for that in the localized disease setting and early interception strategies. So, in conclusion, prostate cancer exposed to ARPi upregulates key survival pathways, including the PARP1-C-MYC axis. That the combination results in down regulation of this axis and an anti-tumor immune TME remodeling in both model systems as well as patients with metastatic castrate-sensitive prostate cancer on clinical trial. That biopsies following short-term ARPi exposure do provide prognostic and predictive information for early interception and therapy escalation in the localized disease. And also looking at the favorable on-treatment morphology, those patients are primed for opportunities for toxicity-sparing approaches downstream. With robust match tissue and liquid sampling, we’re able to confirm the biological cascade of events following ARPi exposure in the tumor cell and the tumor microenvironment and hope to validate novel predictive biomarkers to identify more durable therapy combinations in this space and develop the non-invasive liquid genomic and transcriptomic strategies to bring this strategy to a broader application. Many people to thank who are doing this work. Christopher Logotheris, pioneer in adaptive trial designs. This trial was brainstormed actually six years ago here at PCF Retreat at the dinner on Thursday night. Many thanks to the pathology team, Patricia Controso and Miao Zhang, as well as the radiation and urology collaborators. And a big call out to patients on this study. They are not only willing but eager to participate in these adaptive correlative-rich trials for the betterment of patients. Beyond just themselves. And this study is open not just at main campus but at our community affiliates and it shows that it’s feasible to do correlative rich studies even in community settings. Thank you for your time and happy to take questions.
Unknown [00:16:18] Yeah. Hi, Patrick. Thank you for the talk. So, one question. We many of us are looking at transcriptomic data on similar settings to try to understand this crosstalk. But then the question I have, and I don’t expect you to have an answer now, but it’s like what does it mean in terms of the function in the genome? Are we just reading something that has happened already, or is it true driver that a biological effect that can be targeted? How do we prove that from these studies?
Patrick Pilié, MD [00:16:48] Yeah, I think you know, with the pathway changes, how do we prove that it’s directly linked? In the upregulation of PARP1 and C-MYC crosstalk, it does seem to be both in the preclinical models and in the clinical samples a very early event. In terms of the targetability, you know, at the end of it, is it gonna be just PARP? Maybe not. But with the evidence from the transcriptomic signatures that we can see responses from the combination beyond just HRD. For example, with CDK12 and CHEK2, non-canonical, you know, HRR, they don’t display traditionally HRD signatures, but with the combination seeing response rates hovering around 30 percent with single agent PARP, you know, negligible. So, in terms of explicit mechanism rationale, I’m not sure that we have that fine detail, but in terms of these pathway changes that are occurring as an early event and ARPi resistance, the PARP inhibitor you know combination has the potential to turn those around. Thank you so much.