Targeted Therapy Options in Advanced CRPC

Rahul Aggarwal, MD

Director, UCSF STAND Clinic

UCSF Helen Diller Family Comprehesnive Cancer Center

San Francisco, CA

INVESTIGATING THE USE OF OLAPARIB

AGGARWAL: The PROfound study [NCT02987543]…led to the initial FDA approval of olaparib.¹ This is a study that was randomized 2:1 in an open-label fashion. Patients had to have had 1 prior novel hormonal agent [NHA], so they could have had either abiraterone [Zytiga] or enzalutamide [Xtandi], and they had to have a mutation of either [BRCA1, BRCA2, ATM, BARD1, BRIP1, CDK12, CHEK1, CHEK2, FANCL, PALB2, PPP2R2A, RAD51B, RAD51C, RAD51D, and/or RAD54L]. This was done in partnership with Foundation Medicine, which is the companion diagnostic, and the study was designed with 2 cohorts.

Cohort A included patients with BRCA1, BRCA2, and ATM mutations in their disease; cohort B was everyone else. The primary end point was radiographic progression-free survival [rPFS] in cohort A—so the prioritized gene mutations [to target]—and the key secondary end point was rPFS across both cohorts.

The control arm was randomly assigned to allow physician’s choice of treatment being the other novel hormonal therapy, and these were patients [who] were already resistant to 1 NHA, so these were basically second-line NHAs in the control arm.

The number of patients [randomly assigned in cohort A] was 162 [to 300 mg of olaparib] vs 83 patients to physician’s choice. [In cohort B], the split was 94 patients to [olaparib] vs 48 [patients on physician’s choice]. This was not a large study, but it screened [more than] 3000 patients to be able to enroll in this study, given that these patients were genomically selected.

HIGHLIGHTS FROM THE SUBGROUP ANALYSIS

AGGARWAL: The key [takeaway from the study] is that not all DNA repair mutations are created equal…. [We know now that] the mutations that are most sensitive to PARP inhibition tend to be those that [harbor] BRCA2 and then BRCA1, perhaps to a slightly lesser degree, whereas mutations like ATM have limited activity. In fact, [there was] no difference in [rPFS].² The caveat is [that this] study wasn’t specifically powered to look at the specific subsets. Nevertheless, it’s hard to imagine there’s much of a benefit there with an HR of approximately 1 [Table²].

[The same thing was seen for] patients who harbored a CDK12 gene mutation; [it was a] small group [that had] much less benefit. The same goes for those with a PPP2R2A and RAD54L. There’s a possibility that there’s some benefit there, although the sample size is quite small, so [there was] a wide CI [range]. [This was] another small subset, but [a mutation] that we do think is sensitive to PARP inhibition is PALB2. These can be part of the inherited mutations we see in a small subset of [patients with] prostate cancer, and these do tend to be quite sensitive to PARP inhibition.

COMPARING RUCAPARIB’S ROLE WITH OLAPARIB’S

AGGARWAL: The rucaparib FDA approval was based on the phase 2 TRITON2 trial [NCT02952534].³ This [study had a] similar gene list in screening to the PROfound study, although not completely overlapping…and these were patients who had to have had both prior abiraterone [and] enzalutamide, or apalutamide [Erleada]⁴—so at least 1 NHA [and] 1 prior line of chemotherapy—whereas PROfound was [approximately] 65% of patients [who] had prior docetaxel. So [it was a] slightly different setting but [was] overall similar.

The major difference, of course, [is that] this was a single-arm study, so the primary end point was objective response rate [ORR] criteria based on measurable disease and those [who] had soft tissue disease, [as well as] a confirmed [prostate-specific antigen] response of a 50% or greater decrease. We try to avoid cross-trial comparisons, but [in this case, there are] similar drugs and a similar drug class with a similar patient population. In terms of the various features we look at when we think about PARP inhibitor trials…both allowed [patients with a] biallelic or monoallelic somatic or germline deleterious homologous recombination deficiencies.⁴

This is something we always think about in terms of [whether] that explains the difference in response. If we’re seeing 2 hits in 1 of these genes vs another, it’s hard to determine that with our usual [next-generation sequencing] report. Again, the primary end point was rPFS in the PROfound trial vs the composite response rate in TRITON2.

If you look at the rPFS and ORR data [between both trials]…the responses were largely driven by patients [who] harbored [some sort] of BRCA mutation in their disease. Overall survival [OS] was also significantly improved in the BRCA and ATM subgroups in cohort A, at 19.1 vs 14.7 months [HR, 0.69; 95% CI, 0.50-0.97], whereas there was no difference in OS in cohort B.

In TRITON2, [the ORR breakdown by mutation was] a similar story to PROfound, that not all mutations are created equal. [Breaking this down] further for patients with somatic vs genomic BRCA1/2 alterations, the response rates are quite comparable, [at 43.9% and 42.9%, respectively]. However, for patients with an ATM mutation, [there was] a much lower response rate at 10.5%, and it was even lower in the CDK12 group, which we know now is quite insensitive to PARP inhibition, at [a 0% ORR and 6.7% proportional reporting ratio].

Based on these 2 studies, olaparib gained a full approval in the whole gene set…using the Foundation Medicine companion diagnostic, whether it be tissue or blood, and [for those with] mutations in any one of those genes, use of olaparib is on-label. However, there are differences in clinical activity between these mutations. Conversely, for rucaparib, the label is much narrower. It’s specifically for BRCA mutations, germline or somatic, and patients had to have a prior androgen receptor-targeted therapy [and] a taxane therapy, so [it has] a narrower label based on the numbers of patients.

RESPONSIVENESS TO PARP INHIBITORS WITH NON-BRCA HOMOLOGOUS RECOMBINATION REPAIR MUTATIONS

AGGARWAL: Again, [patients whose disease harbors a] BRCA1/2 mutation has a response rate in the 40% to 60% range, whereas with ATM, CDK12, and CHEK2, it’s in that 10% to 15% range; it’s not 0%, but it is quite low.

[This then] gets to the question of prioritization. Let’s say [you had] a [patient with a] CHEK2 mutation; you would wait and perhaps try a PARP inhibitor after they exhausted all other standard-of-care options. I tell patients that it [has] chemotherapy-like adverse events, with cytopenias, nausea, and decreased appetite, but perhaps a little better ability to tolerate it and adjust the dose [compared with chemotherapy]. It’s just easier [as an oral therapy, too].

Another caveat [to consider], especially with liquid biopsies like Foundation and Guardant, [is that they] will pick up clonal hematopoiesis of indeterminate potential [CHIP] alterations, and in ATM, and particularly in CHEK2, to some degree. You have to take those mutations with a grain of salt and try to delve into the mutation allele frequency and what the specific mutation [there is]. Ideally, you will have the ability to consult with someone with a little more molecular pathology expertise, because those might be more clonal hematopoiesis. There are ways to try to filter out those mutations, but it’s not perfect, so it’s just something to keep in mind.

DIFFERENTIATING THE IMPORTANCE OF CERTAIN GENE MUTATIONS

AGGARWAL: When I see an ATM mutation [in a patient’s disease], especially if it’s coupled with other CHIP alterations, I try to verify them in the tissue even if it’s an archival sample from a few years ago. That still gives me more confidence than [whether] that’s a real ATM mutation.

I’m still not that excited about using a PARP inhibitor because the response rate is low, but at least you know [whether it’s] tumor specific. So it sometimes does require additional testing of a different sample. I do use sequential NHAs selectively, but I will say that at each juncture where you’re switching therapy, you want to see the overall risk/benefit and [whether] the response rate is 50% to 60% in the [patients harboring a] BRCA2 mutation with a PARP inhibitor.

With a second-line NHA, [it’s approximately] 15% and often not durable. I would say that for [most] patients, the data would probably say to go with a PARP inhibitor over a second-line NHA, but it is still done, because it’s easy to switch from one hormonal agent to another.

_References

_{1. FDA approves olaparib for HRR gene-mutated metastatic castration-resistant prostate cancer. FDA. May 19, 2020. Updated May 20, 2020. Accessed July 12, 2023. https://bit.ly/3rB7xf9}

_{2. 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}

_{3. FDA grants accelerated approval to rucaparib for BRCA-mutated metastatic castration-resistant prostate cancer. FDA. May 15, 2020. Accessed July 12, 2023. https://bit.ly/3NXDR3r}

_{4. 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. 2020;38(32):3763-3772. doi:10.1200/JCO.20.01035}