Progressing Prostate Cancer Responds to Multikinase Inhibitors

Mayer N. Fishman, MD, PhD

During a Targeted Oncology Case Based Peer Perspective event, Mayer N. Fishman, MD, PhD of Tampa General Hospital Cancer Care, Tampa General Hospital, discussed the case of a 60-year-old man with progressing prostate cancer.

Targeted Oncology^TM: What are your thoughts on this patient’s case?

FISHMAN: He received cabazitaxel, a second course of taxane, [which is] consistent with the CARD [NCT02485691] trial. Then he got carboplatin therapy. I’m not sure I would’ve gone for carboplatin [at progression]. He’s used 3 conventional cytotoxic therapies and antiandrogen therapy. He had molecular testing on his original specimen, and they found a BRCA2 mutation and they found that it was germline. This seems a bit late to be checking this on the fifth-line therapy.

What are your thoughts on the poll results?

We have 2 votes for germline alone, which you can test off a blood specimen; somatic alone, which you need a tumor specimen for; and simultaneously, which would typically be a tumor sent for general testing.

With this high-risk presentation, I definitely would have been tempted to get it sooner rather than the point of time [that they did the testing].

What has your experience been with getting biopsies in the setting of prostate cancer?

My experience is that the primary specimens—you might as well never be able to get them. There was even an incentive against getting the BRCA testing before the PARP drugs were available because it seemed like a setup for disappointment. You would check somebody, then you’d say, “Well, you know, we’ve got these drugs with approval for other diagnoses, but we’re just going to check you in case they get an approval.” Now that there are the approvals, I think it’s much easier [to get BRCA testing].

I don’t think anybody doubts that the future of oncology is [heading toward more extensive] molecular testing, and prostate cancer is right in the thick of that.

Do you agree with using liquid biopsies then, especially in the setting of disease progression?

Somatic mutations that occur as the disease gets more aggressive are out there, and they’re probably more interesting in terms of catching somebody that could [benefit from a treatment].

Approximately what percentage of patients test positive for a BRCA mutation?

I think the experience was 10% or 20%. When I was at Moffitt [Cancer Center], we did not see many somatic mutations. We had the rucaparib [Rubraca] trial open. There would be one patient after another. We’d check them, and check the next one, and check the next one—and nobody made it past screening. So it was a low percentage with those somatic mutations. I think, once you add a few other things together, you can get to about [10%].

What data support the use of BRCA as an actionable mutation in prostate cancer?

So there are 2 trials here. The rucaparib [TRITON2 (NCT02952534)] trial [included all] your favorite homologous recombination [repair (HRR)] genes.¹ So we have BRCA1, BRCA2, and ATM. My last patient who had a problem, he had a PALB2 germline mutation, for which the B stands for BRCA2. That’s a BRCA2-associated protein. And so if the patient had this, either somatic or germline, and had had progression on at least an androgen receptor drug and at least 1 taxane—and no other therapies—then they got treatment. This was a single-arm treatment, which reflects the confidence. They tested [the patients] every several weeks and treated to progression.

The end points [were] the overall response rate [ORR] and PSA response rate. Not everybody was assessable for the ORR. They had to have measurable disease.

In TRITON2, the phase 2 trial, 1781 [patients were] screened and 200 were enrolled, so just over 10%. Some of these you have to take with a little perspective because of those 1781 patients screened, sometimes they were known cases and then they were referred in for the trial. And then they were screened when they showed up. But, 10% is not far off. But I have to tell you, the experience when you have a trial with 10% of the markers—and you’ve screened 11 patients and nobody had it—you have to wonder if it’s ever going to happen. But they found [the patients]—115 with BRCA alterations. So that’s definitely the dominant [alteration]. This was a single-arm trial, but a quarter of the patients were still on treatment.

Between germline and somatic, we see...maybe 60:40 in favor of somatic. Then, as far as how many alleles were involved [31.3% biallelic, 7.8% monoallelic], most of them [60.9%] were unknown, I suppose. So BRCA2, 102 dominated the whole picture. BRCA1, there [are some]. Between the 2 of them: 115 out of the 209 [enrolled].

The ORR was 43% of the patients with target lesions that had a decrease.² Then if you add in the patients who are least stable, you see most of the patients were at least stable and then there were 3 dramatic, confirmed complete responses. A fair number of them had ongoing responses, either BRCA1 or BRCA2. Either way, you could see a good response.

I would make all the same statements about the PSA [response rate], just there are more patients evaluable. Fifty-four percent had [a decrease] greater than 50%, but if you go down to at least being [stable], it’s more like about 80%. I don’t think there’s a pattern there to say that it works better for [BRCA1] or the other.

There were no big changes between investigator and central review, but we have a limited number of patients and an exciting drug. They want to do a deep dive on that. So that’s the BRCA1 and BRCA2 subpopulation.

In the progression-free survival [PFS] and PSA survival, you see about 9 months among the responders until there was radiologic progression and about 6 or 7 months until there was PSA progression, so [it was] not mature but still considered a decent amount for this refractory population.

If we think back to the CARD trial,³ also a refractory population, I think it was 4 or 5 months for the median PFS.

In the non-BRCA [population], ATM was the most frequent [mutation], which is always reassuring to me because those are the ones that are easy to remember, and then CDK12 and CHEK2. I don’t think there’s a big pattern to see [in PSA response in this group], except to say that [complete responses] are quite rare [only 1], and some stabilization and a 6-month benefit rate. At 12 months, a few patients are still [benefiting].

PSA response was not that exciting. [There were] much more disappointing results for how many patients had PSA reductions. In the non-BRCA [group], some are responsive but it’s not as exciting.

As far as safety, the patients were on [treatment for] about 4 months. The BRCA1 and BRCA2 patients, a little bit longer, and 95% had at least [some adverse event (AE)]. Only 5% or 6% led to discontinuation. These are oral-type chemotherapy [AEs]. Asthenia/fatigue, nausea, constipation—typically low-grade and manageable, but definitely some patients may need to interrupt for this.

So that led to the rucaparib approval,4 [and the FDA indication] says that they had to have had androgen receptor therapy and a taxane, so that’s just part of the label for rucaparib.

Turning now to the olaparib [Lynparza] phase 3 study PROfound [NCT02987543]⁵: The patients were split into 2 cohorts: BRCA1, BRCA2, and ATM—kind of an “A list,” we could call that—and cohort B with all the other [HRR] alterations. [Physician’s choice or olaparib] was the randomization. Those who didn’t get the drug were allowed to cross over in kind of a phase 2-type format. PFS in cohort A was the main end point. But cohort B, the other alterations, those were also of interest. They used FoundationOne for [the genetic testing].

Is there a difference in outcome according to the alterations found in these 2 trials?

Here’s the [list of genes from the] TRITON2 and the PROfound trial [TABLE]. [In PROfound], they had 28% of the 2792 samples, so a bit higher percentage. But I’m skeptical of these percentages because I think that some patients will have been referred into a trial like this knowing they already have one of these changes. I think what happened is people would prescreen their patients and then send them in.

If we go down the list here, you can see that they’re similar, and the main difference was PPP2R2A. Those patients did worse, but otherwise, a relatively similar population. It’s just split out into the populate group with BRCA1, BRCA2, and ATM.

The forest plot [of PFS by subgroup from the PROfound study] was almost the first time in many papers you can see there’s a problem, and the problem is ATM is not especially any different [with olaparib (HR, 1.04)], whereas all the other ones that they checked were better [with olaparib treatment]. Then [for patients with] PPP2R2A, don’t give this drug because they did worse with that [HR, 6.61].

What was the survival for the overall populations on the PROfound trial?

It turns out using something that’s not the target drug is not as good as the targeted drug. [There was] a good hazard ratio [for PFS] at 0.34, 7.4 versus 3.6 months. This is a phase 3 trial, so a little more powered.

The A-plus-B cohorts total [hazard ratio for PFS was] still positive if you add the A and the B together, at 0.49.

Now the overall survival for cohort A [showed a median of 19.1 months with olaparib versus 14.7 months (HR, 0.69)], and then they did a crossover adjusted sensitivity analysis, so patients got censored when they crossed over (67%); then the curves split more [HR, 0.42]. It didn’t turn out to be necessary to see a difference, but if you’re trying to see the real difference in a trial where patients were allowed to cross over, that’s the technique that does it, and no surprise there. It made for a bigger difference.

On the other hand, for cohort B, it was a lot harder to tell that there was a difference [14.1 months vs 11.5 months (HR, 0.96)], and even the crossover [63%] sensitivity analysis couldn’t make a big difference [HR, 0.83]. So not a lot of enthusiasm for cohort B, which is everything except for those top 3 alterations when using olaparib.

What is the tolerability like with these PARP inhibitors?

With our PARP inhibitors, there are some patients who have some major AEs. Thus, some dose reductions [22.3%] were needed there. Patients were on [olaparib] treatment longer [7.4 months vs. 3.9 months], so a little more chance of having an AE.

_Reference:

_{1. Abida W, Bryce AH, Vogelzang NJ, et al. Preliminary results from TRITON2: a phase 2 study of rucaparib in patients (pts) with metastatic castration-resistant prostate cancer (mCRPC) associated with homologous recombination repair (HRR) gene alterations. Ann Oncol. 2018;29(suppl 8):viii271-viii302. doi:10.1093/annonc/mdy284.002}

_{2. Abida W, Patnaik A, Campbell D, et al; TRITON2 investigators. Rucaparib in men with metastatic castration-resistant prostate cancer harboring a BRCA1 or BRCA2 gene alteration. J Clin Oncol. Published online August 14, 2020. doi:10.1200/JCO.20.01035}

_{3. de Wit R, Kramer G, Eymard J, et al. CARD: randomized, open-label study of cabazitaxel (CBZ) vs abiraterone (ABI) or enzalutamide (ENZ) in metastatic castrationresistant prostate cancer (mCRPC). Ann Oncol. 2019;30(suppl 5):v851-v934. doi:10.1093/annonc/mdz394.040}

_{4. FDA grants accelerated approval to rucaparib for BRCA-mutated metastatic castration-resistant prostate cancer. FDA. Updated May 15, 2020. Accessed October 30, 2020. https://bit.ly/2HGYMcy}

_{5. de Bono J, Mateo J, Fizazi K, et al. Olaparib for metastatic castration-resistant prostate cancer. N Engl J Med. 2020;382(22):2091-2102. doi:10.1056/NEJMoa1911440}