Read the Full Video Transcript

Kim N. Chi: Good morning, everyone. Thank you, as everyone else has said, to Silke and Aurelius for inviting me here and giving the opportunity for me to swim in the Rhine. That was quite an experience. Also, thank you for giving me this topic, PARP inhibition, which is very exciting. It already has proven efficacy in patients with ovarian cancer and breast cancer with BRCA2 alterations, as well as maintenance therapy for patients with ovarian cancer that are cisplatinum sensitive.

As we heard earlier this month in the press release, the PROfound trial, which tested olaparib in patients with mCRPC and BRCA1/2 or ATM gene mutations, was also positive for its primary endpoint of radiographic progression-free survival. We're all very excited to see the details of these results, which we presented at ESMO, as we heard from Johann. This study was led by Maha Hussain and Johann de Bono.

These are my disclosures.

I'm going to review PARP inhibitors in prostate cancer, starting with the rationale, go over the mechanism of action, then also look at some of the data that's already known and presented from monotherapy studies as well as combination studies, and look at some of the ongoing studies that are already occurring.

As we heard yesterday, metastatic prostate cancer frequently harbors alterations in DNA damage repair pathway genes. BRCA2 tends to be the most common, followed by CDK12 and ATM. The exact prevalence depends on the population studied, the genes looked at, and the assays used. As an example in the MSKCC population and a report from Dr. Abida, the frequency of alterations in metastatic prostate cancer was close to 30%, with 19% of the patients having a germline alteration. Our British Columbia data in a population-based analysis, we found 18%, so a much lower rate, predominantly because of a lower rate of germline alterations. In our population, it's only about 6% with the remainder being the somatic alterations. BRCA2 tends to be about a third of the patients, CDK12 and ATM another third, and then there's a long tail of other alterations in these patients.

The frequency of these alterations and our ability to identify them allows us to actually action it. Of course, this is referring to PARP inhibition and inducing synthetic lethality, particularly in those patients that have homologous recombination repair deficiency. PARP, for the uninitiated, plays a role in single-cell DNA repair via base excision repair. It binds to sites of DNA damage and then serves to recruit other DNA repair proteins. Inhibition of PARP enzymatic activity prevents this and then you get replication for collapse. The single-strand DNA breaks progress to double-strand breaks. In a normal cell, this would be repaired by homologous recombination. In a cell that has default or a defective homologous recombination, these double-strand breaks persist and then you end up getting cell death, so that's the synthetic lethality hypothesis, which of course subsequently was proven in BRCA2-deficient breast and ovarian cancer.

In addition to the catalytic inhibition, PARP inhibitors also work by trapping PARP. This is through an allosteric interaction between the drug and PARP trapping the PARP at sites of DNA damage. This also leads to replication fork damage, but additional repair pathways are required for this. This potentially increases the number of lesions that could sensitize to PARP inhibitors. Now, this isn't of interest because of the various PARP inhibitors in clinical testing, their inhibitory concentrations to inhibit the catalytic activity are very similar, and they're very low nanomolar concentrations. However, their ability to trap PARP is differentiating with veliparib having a low activity and talazoparib having the highest activity to trap PARP.

The first clinical trial to demonstrate the potential use in prostate cancer was the TOPARP A trial, now published back in 2015. Certainly, Johann de Bono, who led this trial, has to be admired for his insight because when he started this trial at the time, we didn't realize the prevalence of the DNA repair alterations. Perhaps it's insight or perhaps it's just good luck. In any case, the TOPARP A Trial demonstrated that the responses were really only in biomarker-positive patients that had existing DNA repair defects. Notably, BRCA2 patients all responded, four to five of the ATM patients responded, including those with just monoallelic alterations. 

This response was robust. Biomarker-positive patients median time to radiographic progression-free survival, 9.8 months. Recognizing that these patients were heavily pretreated, 80% of them had four lines of prior therapy for metastatic CRPC.

Joaquin Mateo just recently presented the TOPARP B Trial at ASCO. Now, look at these numbers. Those 711 patients screened ultimately to enroll 98, so 14% of patients. Notably, biallelic events we're not required to get onto this study. Roughly a third of the patients were BRCA2, a little more than a third were ATM or CDK12, and a smattering of other PALB2 and so on. What they saw was that indeed the trial was positive in the predefined criteria, particularly with the 400-milligram arm of olaparib showing high activity overall.

But, a secondary endpoint was looking at the various genomic alteration patients, with the highest response rate occurring in the BRCA2 patients, BRCA1 and 2 patients. A very low response in the ATM and CDK12 patients. If you look at the resist and PSA response rate, roughly only just one patient in the ATM group and no patients in the CDK12 group. A moderate response in the other alterations.

Looking at that though, in the waterfall plots, we do see various lesser responses across these gene alterations. Yet, these patients may not have met a response criteria, they might have had a minor response, so it doesn't preclude benefit in patients with other alterations than BRCA2. I have a patient with ATM mutation that's had a minor response and has had progression-free survival of over two years. He has a biallelic ATM alteration.

Other trials that have reported out are the TRITON2 study, which looked at rucaparib again in a heavily treated patient population with prior AR pathway inhibitors as well as prior taxane therapy. This is a preliminary report, 85 patients were enrolled, 45 of them who had BRCA1 and 2, and then a smattering of other including ATM and CDK12 again.

We again see high responses in the BRCA1/2 patients. No responses in ATM patient, one response in CDK12 patient, and then a smattering of other activity in the other gene alterations.

Another preliminary report from the GALAHAD study. This is a little bit different. This study only used at ctDNA to screen for the patients with a turnaround time of approximately two weeks. Patients were also heavily pretreated with prior AR pathway inhibitor as well as prior taxane chemotherapy.

The report at ASCO was only in 50 patients, half approximately BRCA1 and 2. The other half were divided into non-BRCA patients. What we see is again, a high response rate in the BRCA1/2 patients and a lower response rate but not non-existent in the non-BRCA altered patients.

Putting this together to summarize up these phase II trials in monotherapy, we can see a high response rate in BRCA patients. Now, I think we should just look at this qualitatively rather than quantitatively given the preliminary results and the short follow-up in the rucaparib and the niraparib trials. We see a lower response rate in ATM and CDK12 patients and then a modest or moderate response rate in the long tail of the curve of other alterations. But again, lack of response or achieving the predefined criteria of response does not necessarily mean lack of benefit. We really need to see the randomized trials to understand what that is.

Now, just to be complete, I'm going to throw up the selected adverse events across all these trials. These treatments are not without their side effects but certainly well-tolerated, and patients have had these treatments for years and tolerating them quite well, although anemia, thrombocytopenia are noted as well as fatigue and GI side effects.

These are the other ongoing PARP inhibitor trials that will report out, including TALAPRO1, which is a phase II trial with talazoparib. Then, we have two phase III trials, TRITON3 and; of course, PROfound, as I've just talked about at the beginning, looking at radiographic progression-free survival in patients that are post-AR-pathway inhibitors.

Going on to combinations, the combination of a PARP inhibitor plus an AR pathway inhibitor has a lot of preclinical rationales. We know that androgen receptor promotes DNA damage. Then, conversely androgen deprivation therapy upregulates PARP-mediated repair pathways, inducing a kind of BRCAness, and that synthetic lethality can be induced and demonstrated between ADT and PARP inhibition. In addition, going the other way, PARP inhibitor also regulates AR-mediated transcriptional activation, hence giving a good rationale for combining the two in both patients that are BRCA altered or DNA repair altered, and those that are not.

The first look at this was from Dr. Hussain looking at abiraterone plus veliparib. This trial was negative although the DRD patients seem to do better overall whether they had abiraterone or abiraterone plus veliparib. But, however, there's some methodologic issues with this, and it may be because of the veliparib of having less trapping activity.

As opposed to the olaparib plus abiraterone trial where we saw a benefit of the combination across all patients whether they're mutated or not mutated.

Nevertheless, there are three phase III trials looking at this combination of PARP inhibitor plus an AR pathway inhibitor in all comers. But in the MAGNITUDE study, specifically, there is a predefined primary endpoint looking at both DRD patients as well as non-DRD patients.

Briefly touching on immunotherapy, just a couple of slides. This has been tested looking at PARP inhibition plus immunotherapy, and there is a rationale for this. Here's a trial with patients that was enriched for DRD deficiency showing a response.

Here's another trial (KEYNOTE-365) showing patients where there was a very low number of patients with DRD deficiency showing actually a low response. It's kind of hard to understand from these trials what the efficacy will be. Nevertheless, there is a phase III trial that's underway, the KEYLYNK-010.

To summarize, somatic and germline DNA repair damage gene defects are common in metastatic prostate cancer. We need to identify these patients. That's not necessarily an easy task. PARP inhibitors as a monotherapy, high levels of antitumor activity with identified alterations in DNA repair genes, particularly BRCA2. It's clear that defective HRR is required for PARP inhibitor activity. What this will translate to for specific predictive biomarkers remains to be defined beyond what the trials will tell us what to do. Then, phase III trials of combinations are underway and will help us decide in a few years about additional ways we can use these agents. Thank you very much.