Read the Full Video Transcript

Neeraj Agarwal: Hello. Welcome to UroToday. Today we have our guest, Dr. Niven Mehra, who is an associate professor and a medical oncologist at the Radboud University Medical Center in the Netherlands.

Welcome, Niven.

Niven Mehra: Thank you, Neeraj, for the invitation. So I will be briefly discussing our phase II trial we presented to ESMO as a late-breaking abstract on nivo 3 milligrams per kilo and ipilimumab 1 milligram per kilo in molecularly selected patients with metastatic castration-resistant prostate cancer.

As we know, immune checkpoint inhibitors have shown limited efficacy in phase III trials in unselected patients. Nevertheless, we have all seen patients who have responded, which retrospectively mainly were patients with mismatch repair deficiency, patients with a high tumor mutational burden, but also patients with DNA damage repair alterations, such as CDK12 inactivation or BRCA alterations. So our prospective trial investigated these four subgroups, which represent approximately 15% of our patients with the combination nivo-ipi.

What we did, we selected patients with mCRPC, ECOG zero. We set the thresholds for TMB on 7, 7.1, and we wanted patients with biallelic inactivation of CDK12 or BRCA2 mutations.

We screened in our molecular screening protocol called PROMPT. We screened almost 500 patients, and together with some referrals in the Netherlands, we screened 77 patients, of which 69 initiated our protocol. So the main cohort was Cohort A, which were patients who were naive for checkpoints, either measurable per RECIST or measurable per prostate cancer working group criteria. Patients could receive up to four cycles of combination followed by 10 cycles of nivo monotherapy.

Our primary endpoint was disease control rate at six months. Secondary endpoints were objective responses, biochemical responses, PFS, OS, and toxicity.

Here you can see the patients included. We had 21 patients with mismatch repair deficiency, mainly MSH2 and MSH6 mutations. Eight patients with high tumor mutational burden were microsatellite stable. So two patients had a really high TMB over 20 non-synonymous mutations per megabase, and the rest was under 10.

Then we had 16 patients with biallelic inactivation of CDK12, and 20 patients with BRCA2 mutations, of which 16 out of 20 were biallelic. Most patients were previously treated with taxanes and with ARPIs, and the median number of treatments in the mCRPC setting was one.

Going to the results, so for our primary endpoint, 38% of patients in Cohort A had a disease control, which was above our threshold, which we set as our threshold for efficacy in the trial, so we met our primary endpoint. Nevertheless, you can see over the four different molecular subtypes that there was a difference. So 81% disease control in the mismatch repair group, 25% in the high TMB, 19% in CDK12, and 15% in the BRCA-mutated subgroup.

Focusing on mismatch repair, so 75% of our patients with RECIST measurable disease had objective response, all of it partial responses, and the PFS was 32.7 months versus four in the entire cohort.

Here you can also see the progression-free survival curves on the left and the PSA responses on the right. Looking at the PSA, you can see that there are deep and durable responses where 90% plus were seen over all molecular subtypes, but mainly seen in the mismatch repair-deficient subgroup.

To summarize, I think this trial showed clearly in the molecularly selected subgroup that there was efficacy, so 38% of our patients met our primary endpoint of disease control at six months. Nevertheless, I think limited efficacy was still seen in patients with high tumor mutational burden, CDK12, and BRCA2. But nevertheless, clear efficacy was seen in the mismatch repair-deficient subgroup with the dual checkpoints, demonstrating a disease control of 81%, objective responses of 75%, and PSA90 responses of 86%, reflecting a PFS of 33 months with a median follow-up of 19 months.

My take-home message would be is now that monotherapy is reimbursed by the FDA—in Europe, it's not EMA reimbursed—nevertheless, it's available in some countries, and looking at the efficacy of monotherapy in mismatch repair-deficient prostate cancer, we see that objective responses are 48% and the PFS is eight months. So I think the findings from this INSPIRE trial underscore the need for early testing and treatment of mismatch repair-deficient prostate cancer with dual checkpoints.

I would love to discuss this further with you, Neeraj.

Neeraj Agarwal: Thank you, Niven, for sharing this exciting data. Congratulations.

So just to summarize for our audience, you tested ipilimumab plus nivolumab combination in patients with metastatic castration-resistant prostate cancer after prior disease progression on an ARPI and a taxane. You had four subgroups or four cohorts of patients: high TMB, biallelic losses in BRCA2 and CDK12, and mismatch repair-deficient tumors.

You saw striking results in patients who had mismatch repair deficiency in their prostate tumors with 75% of patients experiencing objective responses, which is definitely higher than what we have seen with monotherapy PD-1 inhibitors. You did not see as robust responses in other subgroups, such as biallelic losses in BRCA2 or CDK12 and high TMB patients.

So Niven, first of all, this is great to have such a biomarker-targeted study looking at the efficacy of checkpoint inhibitors in our patients with metastatic CRPC, a population with a very high unmet need after disease progression on an ARPI and taxanes.

Please tell me how do you see these data moving forward, how they can be applied moving forward, and what are the implications for our colleagues out there in the real world?

Niven Mehra: Thanks, Neeraj. It was a very complex study to do because these patients, you are looking for sometimes almost a needle in a haystack. So mismatch repair-deficient patients, I think, are the patients who should be sought after and be treated with checkpoints. I think there's clearly now sufficient evidence that immunotherapy does work with checkpoints in prostate cancer, but it's 4 or 5% of our population.

I think our message is you should screen these patients early. The NCCN guidelines, they recommend screening in the mCRPC setting. I think patients should be screened early, but maybe even moving forward, even to earlier settings. Once this patient is becoming CRPC, we know he's mismatch repair-deficient, and we should start him as soon as possible on the checkpoint.

Now we have FDA approval, I think there's almost sufficient evidence to show that even monotherapy checkpoints probably would be better than giving a taxane or maybe even PSMA lutetium or ARPI switch.

Even though 48% respond with monotherapy after three years, approximately 25% of our patients are still free from progression, which is something that we know from checkpoint inhibitors, that can happen. So I think this is something very important.

I think we should try to move forward to see if we can get dual checkpoints reimbursed. This is not the case yet, but there's another trial, the NEPTUNES trial from my colleague Mark Linch in the UCL in London, who has done a comparable trial, phase II trial. I think this data, together with the CheckMate 650 data, might be sufficient to get EMA or FDA to look at reimbursement because I believe that the dual checkpoints really show extremely good efficacy and a good efficacy to toxicity ratio. Many of the patients I've treated, which responded, I think are after two, three years free from disease. Even on new molecular imaging, we don't see any metastasis, so these patients may be cured.

I think that's something we should do for our patients with MSI. So we should screen for them as early as possible, treat them as early as possible, and then try to move forward with probably novel combinations.

Neeraj Agarwal: So Niven, these are very pertinent data, even though it may apply to maybe less than 5% of patients with mCRPC. But looking at the overall disease burden of this disease, it's still maybe a significant number of patients out there in the real world.

So the message you gave that we need to test these patients early on has been an issue, not only regarding looking at mismatch repair deficiency, but also looking at even homologous recombination repair mutations. For example, olaparib has been approved for five years now, and we got recent approval for PARP inhibitor combinations last year, but still we have found, we have seen in the real world, less than 40% of patients are being tested for these mutations.

So one quick question for our colleagues out there. When should we test? Should we wait for mCRPC? I know you addressed this question, but I just want to emphasize this. Or should we just go ahead and test them when we see them for the first time?

Niven Mehra: Yeah, great question. I think you're right. If we reflex test all our patients when they come in as a metastatic patient, metastatic hormone-sensitive, it will make sense. We do pathology, we do a diagnosis. In those patients, we do NGS and do a broad panel for BRCA, HRR genes, and MSI. I think that would make the most sense so we would not miss any patients. That would be a step moving forward.

I think the guidelines need to step up and then maybe change this. But I think they will only do that if we show that there's a survival benefit if testing is done early compared to late. But I think that would be, in my opinion, the best tool.

Neeraj Agarwal: And this is supported by our experience from the PROfound trial when 30% of patients were not even eligible to be screened because they did not have sufficient quantity or quality of tissue by the time they had developed mCRPC, so we lose tissue, either their quality or quantity.

So the message here is let's test them with comprehensive genomic profiling to look for these targetable mutations as soon as we see them, so that we have those test results ready when they need it, when they have disease progression. And, of course, this may change next year when we have multiple trials reading out, looking at PTEN-deficient patients or homologous recombination repair mutations in the metastatic hormone-sensitive prostate cancer.

Any concluding remarks, Niven?

Niven Mehra: So I think testing, that's the most important thing. I look forward to expanding the precision medicine beyond HRR immunotherapy and, indeed, see if we can expand to PI3 kinase pathway, then we could test, we could treat 40 to 50% of our patients. So I think that's the way forward.

Neeraj Agarwal: Thank you very much for taking the time to join us today.

Niven Mehra: Thank you.