Study will test targeted PCa therapies based on genomic alterations

A trial in prostate cancer known GUNS (Genomic Umbrella Neoadjuvant Study) uses a multi-arm, multistage adaptive design to test targeted therapies in patients with high-risk localized disease by matching neoadjuvant therapies to baseline genomic alterations. At the Society of Urologic Oncology annual meeting, Martin Gleave, MD, of the University of British Columbia in Vancouver, sat down with Urology Times to discuss the rationale behind the trial, its unique multistage design, and other biomarker-driven trials currently underway.

Explain what the GUNS (Genomic Umbrella Neoadjuvant Study) trial is and the rationale behind it.

What we've learned from several decades of neoadjuvant studies prior to radical prostatectomy is that when we use androgen deprivation therapy, even for long periods of time (out to 8 months), when we combine androgen deprivation therapy with more potent androgen receptor (AR) pathway inhibitors out to 6 months, or even when we combine androgen deprivation therapy with chemotherapy, all of these regimens have been shown to prolong life in advanced disease. But when we combine them in the neoadjuvant space, our ability to get complete responses continued to be less than 10%.

Also see: 5-year cancer control comparable for focal therapy, RP

So despite using combinations of regimens that are life prolonging in the advanced space, these failed to achieve a high level of complete response, which is quite different from what we see with systemic therapies in breast cancer or in bladder cancer, where neoadjuvant regimens are associated with a 30% complete response rate when they improve survival in advanced disease.

That's a conundrum. Why is it that in prostate cancer, despite having systemic therapies that are very active, can't we achieve complete response rates similar to those of other solid cancers? One of the rationales for the GUNS trial is understanding, using genomics, the heterogeneity of localized prostate cancer, where different subgroups may have variable responsiveness to different regimens. Hence, using genomic segmentation of localized disease to identify subgroups that may be more or less susceptible to AR or androgen deprivation-type therapies, may be more susceptible to chemotherapy, or may be more susceptible to a PARP inhibitor will allow us to then potentially use combination regimens to push the complete response rate higher.

How is the GUNS trial being conducted?

GUNS stands for Genomic Umbrella Neoadjuvant Study, in which men with high-risk localized prostate cancer are enrolled. We then sequence their needle biopsies and during the first 8 weeks of therapy, they're treated with androgen deprivation therapy plus apalutamide [ERLEADA], an AR antagonist. It takes us about 8 weeks to get that sequencing done.

Based upon their sequence, they are then assigned to one of four different groups. If their tumor has genes that would predict for increased androgen responsiveness, they would be assigned to Group 1 and then be randomized to more intensive therapy by adding abiraterone [ZYTIGA] on top of apalutamide in a randomized fashion. Group 2 enrolls patients who have an aggressive tumor. They've lost p10, they've lost p53, and their tumors are associated with poor response to androgen deprivation therapy. Group 2 patients are then randomized between AR pathway inhibitor therapy alone, plus or minus chemotherapy, in the hope that the addition of chemotherapy would increase benefit in that genomic subpopulation.

Group 3 would capture about 6% to 8% of patients who have alterations in DNA repair-BRCA, FANCA, and others that have been associated with sensitivity to PARP inhibition-and they would receive the master protocol therapy, ADT plus apalutamide, with a PARP inhibitor, niraparib. Group 4 comprises the 5% of patients who have an immunogenic type of cancer because of alterations in MSI, Lynch syndrome, or CDK12. They receive a PD-L1 inhibitor with that therapy.

Again, GUNS uses a multi-arm, multistage adaptive design that allows certain arms to identify what we call conditional lethality. So based upon patients' tumor group, their genomic sequence, and their therapy, if we can increase complete response rates above 20%, that would be of interest and that arm gets expanded. If they don't, then the arm gets dropped in the first 20 patients, which is an early "go" or "no-go" signal. It allows certain arms to be dropped off early, other arms to grow, and new arms to be added as new targeted therapies and new understanding of genomic markers emerge over the coming years.

What other biomarker trials are currently underway?

At the University of Washington, there is a neo-PARP study, which is looking at PARP inhibitor monotherapy in the small segmented population that is germline DNA repair altered. In breast cancer, PARP inhibitors are associated with up to a 30% complete response rate. Whether or not we see that with monotherapy in prostate cancer will be tested in that trial. There are other trials being conducted using various immunotherapy and neoadjuvant strategies based upon immune infiltrate or that characterize the immune infiltrate changes with hormone therapy and with PD-L1 inhibitors as an example.

The challenge has been the time and cost required for getting a genomic signature. Ultimately, many alterations occur in small subgroups of patients. To try and capture 10% of the population in one trial is very inefficient and costly. What the GUNS trial tries to do is allow us to capture as many subtypes as possible, bundle them in one trial, test them in an adaptive way, drop those that are not promising, and expand on those that are.