Read the Full Video Transcript

Christopher Wallis: Hello and thank you for joining us for this UroToday Journal Club. Today, we are discussing a recent publication entitled, An Adaptive Biomarker-Directed Platform Study of Durvalumab in Combination with Targeted Therapies in Advanced Urothelial Cancer. I'm Chris Wallis, a Fellow in Urologic Oncology at Vanderbilt, and with me today is Zach Klaassen, Assistant Professor in The Division of Urology at The Medical College of Georgia. This is the citation for this recent publication in Nature Medicine, led by Dr. Powles and his colleagues.

Advanced urothelial carcinoma remains a disease associated with a relatively poor prognosis. Platinum chemotherapy is the most commonly used first-line approach for patients with advanced urothelial carcinoma with a cisplatin backbone for those with adequate renal function and no toxicities.

However, in the past five years or so, we've seen a rapid proliferation of immunotherapy agents as well as targeted agents in this disease space, particularly in second and third-line approaches.

As a result, the current standard of care is now a platinum-based chemotherapy backbone as first-line treatment for those who are eligible, followed by PD-1 or PD-L1 blockade as either maintenance or second-line therapy with targeted therapies reserved for the third-line and later on.

These more novel approaches could alternatively be used earlier in the disease space and so things like FGFR inhibitors, PARP inhibitors, and TORC1 and 2 inhibitors have recently been developed and analyzed in this disease. Combinations of targeted or immune therapy can improve response rates and the durability of response, particularly when using a biomarker selected approach to target patients with relevant mutations.

The authors sought to assess the hypothesis that durvalumab could be combined with one of three targeted agents in a safe fashion, which may have improved clinical activity. For patients who had FGFR mutations, the FGFR inhibitor AZD4547 was used in combination with durvalumab. For those initially with homologous recombination repair mutations, the PARP inhibitor olaparib was used. And this was additionally opened up to patients without biomarker selection later on. For those who had TSC1 or 2 or RICTOR gene alterations, the TORC1/2 inhibitor vistusertib was used. In addition to these combinations, the study further assessed monotherapy with AZD4547 and with durvalumab.

This is the schema of this study. It is a novel design in solid tumor oncology, employing a multi-arm, personalized approach with an adaptive design. The goal of this study was to rapidly identify combinations of agents that were active and there was the opportunity to add new arms or new biomarkers and increase enrollment to arms where there was evidence of activity. As a result, this is considered an open-label, multi-drug, biomarker-directed, multi-arm phase 1b study.

The authors included patients with locally advanced metastatic urothelial carcinoma who had disease progression following platinum, which was either given in the metastatic setting or as an adjuvant or neoadjuvant setting as long as that was within one year of enrollment, patients that had measurable disease, according to RECIST criteria, adequate performance status, and archived tissue to allow for biomarker analysis. Patients were excluded if they had inadequate organ function, prior exposure to any of the study drugs, prior auto-immune disease, or ongoing immunosuppression, which would affect the use of immunotherapy or if they had an active infection.

As discussed, this is a biomarker-driven trial and so biomarker selection was employed and genetic alterations were measured using FoundationOne testing on archived tissue. And this here shows the association between the agents, biomarkers, and prevalence. As we can see, FGFR1 to 3 fusions or FGFR3 activating mutations were present in approximately one-fifth of the screened patients. Homologous recombination repair deleterious gene alterations were present in about 14%, and RICTOR amplification or TSC1 or 2 loss or inactivation mutations were present in about 15%.

We are now going to go through each of the arms in turn. And so this is the first group with FGFR mutations. And so this was the only randomized comparison within the study and the randomization was to AZD4547 monotherapy or this targeted agent in combination with durvalumab. These were administered as standard doses as you can see here.

The next arm was among those with homologous recombination repair genes. And this was a single-arm assessment of the combination of durvalumab and olaparib. This was initially performed, as you can see here, among this biomarker selected population, and then they also added it in an unselected population in which patients could receive either durvalumab or the combination of olaparib and durvalumab. In patients with TSC1 or TSC2 mutations or RICTOR aberrations, we see only the use of combination therapy. And so this again was a single armed assessment.

For tumors that had more than one targeted mutation, patients were allocated to the arm with the least common mutation, except for those where there was an FGFR mutation, and patients were prioritized for this as it was a randomized comparison.

Here we can see the endpoints and assessment. The primary objective of this phase 1b study was to assess the safety of the combination approach and secondary objectives considered objective response rate, progression-free survival, and overall survival. They further considered the effect of established biomarkers, including a PD-L1 and tumor mutational burden. In terms of assessments, ctDNA was measured at baseline and this was correlated with tissue-based assessment and then used longitudinally to track dynamic changes in FGFR3 mutations.

In terms of statistical analysis, this was done with no preplanned efficacy analysis but a planned enrollment of 26 patients per arm to demonstrate the safety of these approaches. As highlighted with the outcomes, these are predominantly descriptive analyses and confidence intervals were calculated using the Clopper-Pearson method. For time-to-event analyses, these were assessed using the Kaplan-Meier approach.

At this point, I will now pass it over to Zach to walk us through the results of this trial.

Zachary Klaassen: Thanks, Chris. This is the CONSORT diagram for the BISCAY trial. You can see that there were 391 patients screened, ultimately 159 were allocated to treatment and there were 135 patients that had mature data. And as Chris already highlighted, these are the arms of the trial with the one randomization you can see at the top of the figure here with AZD plus durvalumab versus AZD alone and in the subsequent arms, olaparib plus durvalumab, olaparib plus durvalumab unselected, and vistusertib plus durvalumab and durvalumab alone. And so you can see the breakdowns here of the number of patients allocated to each arm.

This is a Venn diagram depicting prevalence and overlap in terms of the mutations. On the right side, in the dark purple, you can see FGFR mutations, in the green DDR mutations, and in the light blue, TR mutations. And so moving over to the left part of this slide, between the BISCAY enrollment biomarkers in tissue utilizing the FoundationOne population, the most common mutation in 61 patients was FGFR. In the middle, this is between all enrolled and treated patients, most prevalent again were FGFR mutations. And in the right of this part of the figure between the BISCAY enrollment biomarkers and plasma ctDNA for treated patients, once again, most commonly FGFR mutations at 25 and TR mutations at 20.

This figure looks at the comparison of FGFR3 expression between patients. On the left here in the purple figure, is AZD monotherapy or plus durvalumab had the most frequent FGFR expression. And then similarly, between the two on the right, we see olaparib plus durvalumab or durvalumab monotherapy. So the most common FGFR was in the AZD arms.

This figure looks at the confirmed response rates by RECIST criteria. The most prevalent or most robust response rate was in the olaparib plus durvalumab biomarker selected arm in blue in the middle of the figure at 35.7%. The second most impressive response rate was in the AZD monotherapy arm at 31.3%. And to highlight here, you can see that the addition of durvalumab actually had a lower response rate at 28.6% when compared to AZD alone. On the far right, you can see durvalumab alone, a respectable response rate at 27.6%. Vistusertib plus durvalumab 24.1% And the worst response rate was olaparib plus durvalumab in the biomarker unselected group.

These are the waterfall plots assessing maximum reduction in target lesion for the study arms. The two parts on the left here are AZD monotherapy and AZD plus durvalumab. And you can see here responses in about three-quarters of patients, but very similar when you look at AZD plus durvalumab versus AZD alone. On the right here shows the olaparib plus durvalumab unselected and selected waterfall plots.

This is a summary of the best radiological outcomes with treatment. In the dark green is complete response with only a handful of complete responses in the durvalumab and the vistusertib plus durvalumab arm. In terms of partial responses, the most robust is in the olaparib plus durvalumab biomarker selected arm and comparable partial responses between the AZD monotherapy arm and the AZD plus durvalumab arm. Moving over to progressive disease, we see a lot of patients in the olaparib plus durvalumab unselected arm in terms of progressive disease, as well as in the durvalumab monotherapy arm and a decent number of progressive disease patients also in the olaparib plus durvalumab biomarker selected arm.

These figures are looking at the swimlane plot showing duration of therapy and timing of confirmed responses and progression. You can see here on the bottom left is durvalumab in orange, in the gray is vistusertib plus durvalumab, in the yellow is olaparib plus durvalumab biomarker selected, below that in the light green is olaparib plus durvalumab unselected.  Then moving to the top left, AZD plus durvalumab in dark blue and AZD alone. And you can see here that the single-arm versus combinations are pretty much exactly the same here in the AZD plus durvalumab or AZD monotherapy, which is also very similar to the bottom left here in the durvalumab arm. So this is highlighting that the combination therapies specifically did not add much to the monotherapy arms.

In terms of progression-free survival for the study arms using the Kaplan-Meier method, in the top left, you can see the AZD six-month progression-free survival rate was 28.1% compared to a lower 20% for durvalumab plus AZD. To the top right, durvalumab plus vistusertib, 31.3% for the six-month PFS rate. Durvalumab alone, the six-month PFS rate was 32.9%, and in the bottom left, durvalumab plus olaparib, 42.9% in the biomarker selected group, and at 20.5% in the durvalumab plus olaparib unselected group.

Similar, these are the OS figures for these arms, comparable one-year OS rate of 42.3% for AZD monotherapy compared to 56% for AZD plus durvalumab. Moving to the top right, durvalumab plus vistusertib, one year OS rate of 49%, and durvalumab monotherapy at 44.6%. And in the bottom left, durvalumab plus olaparib in the biomarker selected group, 49.4% compared to 49.2% in the durvalumab plus olaparib unselected group.

Several discussion points from the BISCAY trial. We know that running multiple traditional randomized phase two studies to address drug activity is not always feasible. In BISCAY, they pursued an alternative approach with an adaptive design, a modest number of patients in each arm, and the potential to expand or change arms during the trial. This strong biomarker platform facilitated the potential reinforcement of efficacy signals and decision-making. As we showed on the previous slide, monotherapy arms were used as indirect benchmark comparators to combination therapy. However, this does lack accuracy and the benefit of direct comparisons and formal statistical analysis, but it does have the advantage of addressing multiple research questions concurrently. As mentioned previously, none of the combination arms showed adequate efficacy or compelling biomarker data to move into the randomized trial setting. AZD4547, which was FGFR monotherapy, showed activity supported by biomarker analysis but the addition of durvalumab did not enhance this activity.

In conclusion, this biomarker-directed, multi-arm study, combining durvalumab with three promising molecular targets shows that this new design of trials is feasible. The strong biomarker platform in BISCAY, including circulating biomarkers, helped reinforce efficacy signals, which further adds to the utility of this approach. However, it is important to note that the combination of immune therapy and targeted therapy in biomarker-driven populations does not appear to improve efficacy with implications for further drug development in urothelial cancer and beyond.

Thank you very much. We hope you enjoyed this UroToday Journal Club discussing the recently published BISCAY trial.