PARP inhibitors for breast cancer — Hope S Rugo, MD

DR RUGO: It's a pleasure to talk about PARP inhibitors for breast cancer today, a really evolving area of treatment for our patients with metastatic disease.

So, we actually have one new paper that was published in Breast Cancer Research and Treatment, Joyce O’Shaughnessy was the first author, and I think it's interesting to highlight because we have an idea about from the studies that have been done so far about how many patients have BRCA mutations that are germline, but this gave a more global evaluation over 14 countries. Previous to this data, we had some information where it appeared that if you test all patients with triple-negative breast cancer, that you’ll see a reasonable frequency, in the low teens, in younger woman, and when you get to older women with no family history, it really goes down into the single digits, in the 1- to 2% range.

So, in this they had planned a much larger group of patients, but they ended up with 341 patients who actually had information, from 14 countries, and they had both germline and somatic testing, which is interesting. So the median age was 56 — 25 to 89 years of age — so it wasn’t a very young population and only a quarter were premenopausal, which is important. Almost 10% had a germline BRCA mutation and anything that we do that has 10% is interesting. These are patients who were just tested, so there wasn’t any specific criteria other than having breast cancer. About 5% had germline BRCA1 mutations, 3.5% BRCA2, and 1.5% both, which is much less common.

So, the prevalence of mutations interestingly, is similar in hormone receptor-positive and -negative disease. It was 9% in European patients and almost 11% in Asian patients. And 13% were less than or up to 50 years of age at breast cancer diagnosis versus 5.4% in patients greater than 50 years. So that suggests that we really should be testing patients who are under — up to age 50 who have a new breast cancer diagnosis if it's possible.

So, if you had any risk factor for having a germline BRCA mutation, so a family history of breast cancer and/or ovarian cancer, you were aged up to 50 years at diagnosis, or you had triple-negative disease, the rate of mutations was 10.4% versus 5.8% in those with no risk factors. They also looked at homologous recombination defect, and they found that prevalence in 14% of patients with wild-type germline BRCA. So these patients did not have a mutation, but 9 out of 64 had this homologous recombination defect. So it's very, very interesting. And that suggests that they had somatic BRCA mutations and that's not an infrequent number. So now that we have some emerging data about using PARP inhibitors in patients who have somatic mutations, that should be something we should be looking at when we’re doing next-generation sequencing in these patients.

So, what about PARP inhibitors? Well, we know now that there are 5 PARP inhibitors that are approved for various indications. In terms of breast cancer, we have 2 PARP inhibitors still, olaparib, approved first, and talazoparib, all in 2018. The way these drugs work, of course, is shown in the diagram to the right, where PARP is trapped on DNA and blocks the ability to repair DNA and you need homologous recombination. And if you have a germline BRCA or, apparently, somatic BRCA mutation, you actually can — in these patients, you have death of the tumor cell. And this is enhanced by adding drugs that cause cell death by causing DNA damage because you can’t repair DNA damage that's caused by chemotherapy. So this is quite interesting. And it's brought up a lot of interest as well, as we’ll talk about in a moment, about whether or not these platinum agents work as well as PARP inhibitors, which we really don't know the answer to as yet.

There were 2 Phase III trials that led to the approval of olaparib and talazoparib, the EMBRACA study and OlympiAD study, both trials randomized patients with known germline BRCA mutations to receive talazoparib, olaparib, or physician’s choice of chemotherapy. Patients could have either triple-negative or hormone receptor-positive disease, but they excluded patients with HER2-positive disease, and that's important. We do see a small number of patients who have germline BRCA mutations in HER2-positive disease, but because it's such a small number and we of course treat those patients preferentially with HER2-targeted therapy, they weren’t included in these studies.

You could have a history of controlled brain metastases, as you can see in the EMBRACA study design, and for the OlympiAD study, you could have received up to 2 prior chemotherapy lines in the metastatic setting. There were some little bits of differences in terms of evaluating whether or not you’d received a platinum, but neither of the trials allowed patients who clearly had disease resistant to platinum.

The primary endpoint in both studies was progression-free survival. And I’ll point out that these studies, very hard to actually power for overall survival because the numbers are small. The EMBRACA study randomized 2:1 and OlympiAD, as you can see, is 2:1 randomization similarly.

So, if you look at the progression-free survival, which was the primary endpoint, what you can see is the hazard ratios are in the .55 range for both studies, highly statistically significant, with almost a doubling of progression-free survival in patients receiving the PARP inhibitor versus chemotherapy of physician choice. This is actually quite fascinating data because the PARP inhibitors in general are better tolerated than chemotherapy, with the exception of a very small number of patients who get unusual toxicities. So, this is actually really important data. I noticed that the reference here on the OlympiAD trial will be shown on subsequent slides.

So, here you can see the data on looking at triple-negative breast cancer versus hormone receptor-positive disease, and again with the randomization looking at comparing to treatment of physician choice. And this is actually also really interesting because we tend to think of patients with hormone receptor-positive disease, regardless of BRCA status, as having very chemotherapy-sensitive disease. But, interestingly, in both of these groups we actually saw that the PARP inhibitor had better progression-free survival compared to chemotherapy of physician choice, although the dataset is small, and the numbers are small when you divide up into these subgroups.

We also looked at response rates in this population of patients. You can see that the response rates were higher for talazoparib or olaparib compared to standard therapy, chemotherapy of physician choice, and here you can see the numbers really striking, the 2:1 randomization and the lower response rates are quite nicely shown graphically.

So, what about overall survival? And this gives you this reference here for the OlympiAD study with Robson in the New England Journal in 2017, with an update on the overall survival in 2019 in Annals of Oncology. And this is interesting because overall, the overall survival showed no difference in the OlympiAD trial. But then they hypothesized that potentially this was due to global mechanisms of resistance, even though patients could not have disease that was resistant to platinum therapy, they generate — we know that by treating with chemotherapy we generate multiple mechanisms of resistance. So here, you can see that patients treated in the second- and third-line, kind of median overall survival of between 17, almost 19 months. But if you looked at patients who had no prior chemotherapy for metastatic disease, a little under 100 patients, you can see that the standard therapy was about 15 months and about 22.5 months in patients with olaparib.

So this, I think, has really shaped the way we use PARP inhibitors and treatment to patients in the metastatic setting who have known germline BRCA mutations. Because we try and treat patients as early as possible, based on the other standards of care for treatment of patients with either triple-negative or hormone receptor-positive breast cancer. Remember, this is chemotherapy. It’s not hormone therapy. And that's really important because you can use endocrine therapy with CDK4/6 inhibitors, for example, in patients with hormone receptor-positive disease first and still use the PARP inhibitor and you’re still before getting to chemotherapy for metastatic disease.

So, this is the OlympiAD Extended Follow-Up in patients who had no prior chemotherapy for metastatic disease, and that's in this publication in Annals of Oncology in 2019. Fourteen patients, interestingly, were still on olaparib and this is a small number of patients who really can stay on their PARP inhibitor for a long time without progression of their metastatic disease, as small number of these patients actually never end up progressing, at least not within the timeframe that we're used to seeing. Nine percent of patients received olaparib for more than 3 years. And if you look at these patients in the no prior chemotherapy setting, you can see that this number actually is really continued, 15 versus 22.6 months, as I showed you earlier, and there was no difference based on tissue receptor subtype, although the numbers get very small when you get, actually, to this group.

If you looked at the safety signals in that paper, which I think are really important as well, and this is true — also the EMBRACA study, there’s been no increase in the number of patients who have acute myelogenous leukemia, or other myeloid malignancies in patients receiving the PARP inhibitor. And although, the exposure to treatment is shorter than we’re going to see in the adjuvant setting when we see the adjuvant trials, it's, I think, really important because we had been really concerned that by blocking DNA repair by the only remaining major mechanism that you have, that you would have more myeloid malignancies. And these patients all received prior chemotherapy afterwards, or they had received in the early stage setting but we’re not seeing an increase in leukemia.

Now, at AACR this year, now published in Annals of Oncology, we presented the final overall survival data from the EMBRACA trial and we saw no statistically significant difference between the treatment groups, as you can see here. It was about, interestingly, the same as what was seen in the second- and third-line group and the intent-to-treat group for the OlympiAD trial. And here you can see the differences when you look at month 12 and at month 24, 36 and 48 between talazoparib and chemotherapy. There’s no big differences in any of those times.

So what could make this happen that you see really no difference. And we did look in the first-line chemotherapy group, although it's a small number, patients who are receiving this in the first-line group before any chemotherapy. But if you look at the crossover, I think that's really important. Because remember that olaparib was approved almost a year before talazoparib and that was the first data. So, in terms of patients who received chemotherapy, 33% of those patients went on to receive a PARP inhibitor, and incredibly important. Whereas, only 4.5% of the patients on talazoparib understandably, continued.

And then platinum patients —more patients with talazoparib interestingly, received the platinum therapy, although that difference was relatively modest.

So I think that the crossover has a huge impact here on evaluation of overall survival in the EMBRACA trial.

Now, what about quality of life when we use these oral agents? We don’t just want to believe that because it's a pill it's better than chemotherapy. But actually, you can see here, by looking at both the OlympiAD and EMBRACA trial, that quality of life was prolonged for longer. So there was a significant delay in the time to clinically meaningful deterioration, global health status or quality of life and regardless of how you looked at. And you can see the hazard ratio here, .44. This is really quite impressive. Our patients just have better quality of life here when they are treated with a PARP inhibitor.

So, I think that that does give us even more information about potentially treating patients before they receive chemotherapy, except for in the situations where we have treatment that can result in a major survival difference.

What about toxicities? These are shown here. It's very hard to do cross-trial comparisons because there were differences in these patient populations. We see a little bit of bone marrow suppression, as you can see here, and we’re looking at all grades. There is some nausea and headaches. There seemed to be a little bit more headaches here, shown in the talazoparib trial, than in the olaparib trial. We don’t know if that's a true difference. But I think it is actually really interesting information. And then, the other area of interest is anemia, where some patients do require transfusions. And again, there was a numerical difference here. But cross-trial comparisons are quite different and both drugs can cause anemia. Whether or not there’s a difference in needing blood transfusions or not, would require head-to-head trials, or whether or not there was a difference in actually the headaches.

Alopecia. Some patients have a little bit of hair loss with the PARP inhibitor. So, important to mention to the patients. They don’t lose all their hair like they do with doxorubicin.

So, one of the questions that comes up is, is this the same as what we’re seeing with a DNA damaging agent like carboplatin or cisplatin? And the only trial we have that looked at platinum versus chemotherapy in the metastatic setting was this first-line trial, looking at carboplatin versus docetaxel in patients with triple-negative breast cancer called TNT. The number of patients with BRCA mutations was quite small in that trial. And you can see it was 25 receiving carboplatin and 18 docetaxel. So, really small numbers. But actually, you can see that the progression-free survival is quite similar, looking at platinum versus the PARP inhibitors. But I think we would all agree that getting a PARP inhibitor in general is better tolerated than receiving chemotherapy. And because the numbers are so much greater with the OlympiAD and EMBRACA trials, we really do, I think, in general in our guidelines, prefer using PARP inhibitors where they’re indicated versus a platinum agent. So, very interesting data.

So, then the other question is what about combining the PARP inhibitor? If platinum works so well in these patients who have homologous recombination defects due to their germline BRCA status, does it help if we add in chemotherapy to a PARP inhibitor? Well, previous studies had looked at giving either talazoparib or olaparib with chemotherapy and found that it was way too toxic. And I showed you in the toxicity that there was actually bone marrow suppression with these agents given as single agents.

Veliparib is a less potent PARP inhibitor not yet approved for breast cancer, and so it can be combined with chemotherapy. And we gave it with paclitaxel and carboplatin in the neoadjuvant I-SPY trial and showed that we could give this triplet together. And it was also studied in the neoadjuvant BrighTNess trial as well.

So the BROCADE3 trial actually looked at this in the metastatic setting. Patients had to have known germline BRCA mutations, had received up to 2 prior lines of cytotoxic therapy for metastatic disease, and they couldn’t have progressed within 12 months of a platinum. So it did allow neoadjuvant or adjuvant platinum therapy. Patients were randomized again, 2:1, to receive veliparib or placebo with carboplatin and paclitaxel. And what was interesting about this trial and is really important, is that if patients were actually doing okay, they had stable or responding disease, they could discontinue their chemotherapy and continue with veliparib alone or placebo. So this was sort of a maintenance-like treatment, similar to what we’ve seen as now as standard of care in ovarian cancer. And then they used a lower dose of veliparib when it was combined with carboplatin and paclitaxel, 120 mg twice daily, versus the 300 to 400 mg that's used as a single agent when they combined with chemotherapy. So because of that, they allowed an escalation, when patients stopped the chemotherapy and continued with veliparib or placebo.

Now there’s one carrot to these patients because, of course, PARP inhibitors are approved in many countries now, you could have a crossover, an open-label cross over with progression, which I think is really important.

So this shows you the progression-free survival by investigator assessment, which was the primary endpoint. And what you can see here is actually, which is very interesting, in the top left plot here, is that the median progression-free survival was increased by about 2 months in patients receiving veliparib and chemo versus placebo and chemo. But where you see the curves separate is about 10 months from randomization, so, maybe in the 8- to 10-month range. And if you look at PFS 24- versus 36 months, it's really even more different at that time point. So you get out to 36 months, it's 26% versus 11%. So this is really a big difference. So why could that be the case?

Well, remember, that you could continue on veliparib alone after chemotherapy, but you got a higher dose. So we believe, really, that this is a role, not just of combining the veliparib with chemo — we can’t separate that effect — but that maintenance with the PARP inhibitor is a really important factor here in breast cancer, not just in ovarian cancer.

And then additional data was shown here looking at the blinded, independent Central Review, again showing a hazard ratio of .63. Remember, this was a subgroup of patients who had no chemotherapy in the metastatic setting, and you can see that the curves separate even earlier, and this was data presented at San Antonio in 2019. And the whole BROCADE study now has been published in Lancet Oncology this year, just recently, by Véronique Diéras, who was the principle investigator.

What about the toxicity? Well, when you combine the PARP inhibitor, even at a lower dose of chemotherapy, you see more thrombocytopenia. And that's a tough toxicity for us to manage. But interestingly, and encouragingly, there was no difference in neutropenia or anemia. So, it's sort of further, I think, gives weight to the idea of maintenance, where, if you need to give chemotherapy induction, you could use a PARP inhibitor as maintenance therapy and you might make a really big difference in the duration of disease control for this patient — 42.5% received monotherapy with veliparib and 32.4% received monotherapy with placebo. And survival is not going to be a good endpoint in this trial because almost 50% of patients crossed over with progression from placebo to veliparib.

So, fascinating data and gives us a lot of information that I think is important for treating our patients.

This shows you the separation of curves here when patients stopped their chemotherapy. And these patients who were treated in the first-line setting, the little red arrow points to where they stopped their chemo and continued maintenance with just the PARP inhibitor and you see this really dramatic separation of curves.

So, now at ASCO this year, we saw some really interesting data from the Cooperative Group/SWOG, presented by our colleague Sharma, from Kansas, and she’s done some really interesting work in this area, and this study I think was quite a unique approach. So, they randomized patients who either had metastatic triple-negative breast cancer, or BRCA germline mutation associated, HER2-negative metastatic disease. So those patients could also have ER-positive breast cancer. They could have received up to 1 prior chemotherapy for metastatic disease. Three-hundred and twenty one patients were randomized to receive cisplatin every 3 weeks with veliparib, and that's the bigger dose, 300 mg twice a day, or cisplatin and placebo.

Now they did a lot of correlative work here and they looked at cell-free DNA. So there’s blood testing, as well as getting a tumor biopsy before patients enrolled.

And their primary endpoint was looking at PFS in 3 different groups. So they looked at the germline BRCA group. Patients who had BRCA-like disease, so they have this homologous recombination defect, or non-BRCA-like disease. And then their secondary endpoints are, they're similar to what we would expect in any trial. And the post-randomization germline and BRCA-like biomarker testing assigned patients into these prespecified groups for their endpoint.

So, I’m sure you can see at the bottom actually the numbers of patients who were assigned. So, you had a BRCA mutation, and that included just 37 patients. So, then you have the patients who didn’t have a germline mutation and they did this BRCA-like biomarker analysis. So they have homologous recombination defect genomic instability score and they used a score of 42 or greater — I won’t get into that because that's very detailed and somewhat controversial between studies. Patients were also included who had somatic BRCA mutation. They were included if they had a BRCA1 promoter methylation which silences the BRCA gene in general, and then germline homologous recombination repair gene mutations that excluded BRCA1 and 2. Any one of any of the above 4 markers made the patient go into the BRCA-like group. That was 99 patients, so bigger group, and 110 were in the non-BRCA group.

So what did we see with this patient group? Very interesting data. So here you can see the progression-free survival and overall survival in the group of patients who have BRCA-like disease. So remember, this is about 100 patients. Because of the randomization, you can see here it was 1:1, you see about a little under or over 50 patients in each group. And what you can see is progression-free survival looks quite similar, 5.9 versus 4.2 months, but because of their statistical waiting and the small numbers, this was statistically significant with a p-value of .0006 and hazard ratio of .53. So, overall, the PFS was longer with olaparib. And strikingly, the overall survival was also longer, going from 12 to 14 months. That's quite interesting, a little under a 2-month difference. The p-value was not significant at .067, but it's borderline.

Interestingly, in this study, in the non-BRCA-like, as you might expect, they found no differences. But in the germline BRCA-population, they also saw no difference in progression-free survival. Now patients continued on the doublet until they developed progression. So there was no maintenance. And the issues about toxicity, I think are important here. It is likely that giving these drugs together may reduce the relative benefit of the PARP inhibitor so that we’re better off trying to maximize the dose and exposure to the PARP inhibitor rather than to give it with chemotherapy. So it kind of promotes the maintenance-like effect we saw in BROCADE3.

What’s really intriguing about this study is the suggestion of benefit in the small number of patients who have this BRCA-like disease that was very broadly defined. We’ll see a lot more data from this study in terms of trying to analyze the different criteria for BRCA-like disease. But I would caution that this is a small study and in order to really understand the benefit in patients who have disease that's BRCA-like, we’ll need a bigger randomized trial in the future. Of course, these studies are complicated now by the use of checkpoint inhibitors in patients who have PD-L1-positive disease and then now demonstrated survival benefit in patients who were receiving that with specific chemotherapy agents. And so, another mechanism of study is to look at PARP inhibitors in that setting, as I’ll show you in just a moment.

So now, if the drugs work, then we might as well try them in patients who haven’t received treatment yet, and of course, we saw in the OlympiA trial this sort of striking difference in benefit in overall survival suggested by a subgroup analysis in the first-line setting where they hadn’t received chemotherapy for metastatic disease.

So this small trial done at MD Anderson and published in the JCO by Jennifer Litton who led the trial, shows you what you could actually get pathologic complete response rates for just taking a PARP inhibitor for 6 months. So this was neoadjuvant talazoparib. They received it for 6 months in patients with known germline mutations, and they were able to assess the pathologic complete response rate in 19 patients, overall, in terms of RCB in 20 patients looking at no invasive disease. And what you can see is that the pCR rate was 53%, that's a residual cancer burden of zero, there were additional 2 patients who had minimal residual disease for a RCB zero on rate, of 63% in the 19 patients.

And there is a Phase II trial ongoing now. You can see that there were more pathologic complete responses in patients with BRCA1 mutations than BRCA2. And we’ve seen this — and seen a hint of this anecdotally in many settings, that for some reason in BRCA2 some of the tumors seem to be more driven by the hormone receptor status potentially than by the germline BRCA mutation. But I think the numbers here are very, very hard. You only have 3 patients with BRCA2 overall in this patient population. So I think potentially, physicians spoke with their feet in terms of when they were enrolling in this trial. The now ongoing Phase II trial will give a much better idea between BRCA1 and BRCA2.

So, what to do now? So, what I’ve showed you is that with chemotherapy there is toxicity and the toxicity may limit your exposure to the PARP inhibitor and chemotherapy, so maybe maintenance is a better approach.

Should we give an endocrine therapy along with a PARP inhibitor in patients with hormone receptor-positive disease? An intriguing idea that is being studied. But we don’t have data on it yet. We know there isn’t additive toxicity.

The NeoTala trial is looking at talazoparib in 122 patients with germline BRCA mutations. It's going to be fascinating results. And then of course, the adjuvant OlympiA trial enrolled 1800 patients with germline BRCA mutations; accrual has been completed, and we’re anxiously awaiting results which are expected in 2021.

There are other germline mutations that we’re now finding. So now we can cheaply evaluate patients for germline mutations in really any gene that's been associated with cancer risk. And that is a fascinating area, where some of these genes correlate with defects in DNA repair. And I’ll show you some data in just a moment. We also see somatic mutations.

I told you already there’s a suggestion that patients who have somatic mutations might benefit from PARP inhibitors.

The other thing that's fascinating is this idea that PARP inhibitor cause Type 1 interferon activation via the STING pathway. And so, potentially, they could activate the cancer and the host to be more responsive to immunotherapy. So it makes a lot of sense to combine them. And we’ve seen safety data and some very early evidence of efficacy data, and now this is being studied in a Phase III setting.

So this shows you the OlympiA trial. The accrual is completed. We hope to have data this year. I’m sure that the COVID-19 pandemic has delayed actually just being able to analyze data; it's really hard when people aren’t going into work to get all the data. So we’re hoping to see this next year. Included in these patients were patients who had residual disease after neoadjuvant therapy and then patients who had high risk disease who were treated in the adjuvant setting, olaparib versus placebo, and it was only given for 12 months. So it's going to be very interesting to see the invasive disease-free survival that is the primary endpoint.

In terms of the management of hereditary breast cancer, we actually do have data published by Nadine Tung in the JCO from this year, that are a combined set of guidelines from The American Society of Clinical Oncology, The American Society of Radiation Oncology, and the Cervical Society of Oncology. And it's shown where that when you’re offering chemotherapy for germline BRCA mutations with metastatic disease, platinum chemotherapy is preferred to taxane therapy, based on the TNT trial. We don’t know about other germline mutation carriers.

For germline mutation carriers who are treated for breast cancer in the neoadjuvant or adjuvant setting, we don’t have data that supports routine addition of platinum, although there are studies that will give us that information, I think, in the future to standard chemotherapy. And really, you shouldn’t just automatically add a platinum because these patients have very chemotherapy-sensitive disease.

For patients who have metastatic disease with germline mutations and HER2-negative cancer, olaparib or talazoparib should be offered as an alternative to chemotherapy in the first- to third-line settings. And there’s no way to directly compare PARP inhibitors to platinum therapy.

And then they mentioned here, and this data actually predated ASCO, that if you had mutations in moderate-penetrance genes that encode for DNA repair defects, there’s no robust data to support the use of PAPR inhibitors. And then, we don't have data. And it's interesting, because I’ve seen a few patients who were treated in the adjuvant setting, unfortunately. I think patients with triple-negative disease should all, who have disease 1 cm or greater, should be treated in the neoadjuvant setting if possible. But patients who were treated who had node-positive disease in the adjuvant setting are asking, if you have a BRCA mutation should you automatically get a PARP inhibitor? But we don’t have data. And the guidelines support there’s insufficient data at this time to recommend a PARP inhibitor. We do have data for capecitabine in triple-negative disease and we have excellent data with hormone therapy and now emerging data with CDK4/6 inhibitor, abemaciclib. So, I don’t think we should be recommending PARP inhibitors in this setting yet.

And, of course, ASCO gave their comment about clinical trials that are incredibly important for these patients. And of course they’ve led to the approval of PARP inhibitors.

Now what about those patients who actually don’t have a germline mutation in BRCA, but they have germline mutations in other genes that encode for DNA repair or somatic mutations, which we see now, that we’re NGS in these patients with metastatic breast cancer. So this trial called Olaparib Expanded, which is actually expanded further, was part of our translational breast cancer research consortium and led by Nadine Tung, was just published in the Journal of Clinical Oncology.

We knew that there’s evidence of activity of PARP inhibitors in prostate cancer and ovarian cancer in patients who have homologous recombination defects. So this study really made a lot of sense. Patients were eligible in Cohort A if they had an alternative germline mutation, and in Cohort B if they had a somatic mutation, including somatic mutations in BRCA1 and 2. And then patients received olaparib, 300 mg twice daily, and were evaluated every 3 weeks in clinic and every 6 weeks with tumor assessment. And, of course, they continued until response or toxicity.

And this shows you the really striking data from this paper, 15 patients remained on study at the time of the data cutoff. And, of course, you can see that there were great responses because of the nice waterfall plot. But how do we analyze these responses? And it actually helps down here in the little table here So ,in patients who had germline mutations in PALB2, which is on the same gene where BRCA2 mutations are, 9 out of 11 patients had a partial response, 82%. And I think that this data we all believe is sufficient to recommend PARP inhibitors for patients who have PALB2 germline mutations.

In patients who had somatic mutations, there were only 2, so really hard to assess. Neither of them responded.

Patients who had somatic BRCA1 and 2 mutations, this is 17 patients, in the 16 who were evaluable, 8 patients or 50%, had a partial response. This is also dramatic information and I think suggests that, although it not yet have regulatory approval, that we could consider the use of PARP inhibitors in patients who have somatic mutations as well.

What about ATM and CHEK2, where there 17 patients? No patients responded, so we don't have the data to encourage use of PARP inhibitors in this population.

So this is really practice changing data, and I think really impacts the guidelines that we saw before where I think we can benefit patients with treatment that actually benefit patients with less impact on quality of life, as we saw in the EMBRACA and OlympiAD trials.

What about where we look for these somatic BRCA mutations? So this is a fascinating paper published by my former fellow, Dr Vidula, who’s now a faculty member; has been for several years at Mass General, and she looked actually at cell-free DNA for somatic BRCA mutations. And this is shown here, actually blue are the wild-type genes, yellow are the known pathogenic mutations and green are these novel variants, which are the somatic mutations. And actually, 13.5% of patients out of 215 patients screened, had somatic mutations in BRCA1 and 2, in cell-free DNA. Four percent were known germline pathogenic, but the rest, 9%, were novel variants. So it's really interesting. These germline mutations were more common, as you would expect in younger patients, triple-negative disease, and they were more likely to be associated with TP53 mutations.

But, actually, they did then this really fascinating correlative study, and they did cultures with these circulating tumor cells and demonstrated sensitivity to PARP inhibitors in patients with BRCA1 but not BRCA2 somatic mutations. This is just hypothesis-generating. They also saw this APOBEC-mediated mutagenesis which is just really a signature of multiple mutations as far as I could really evaluate as a clinician. Those patients who also had these multiple mutations didn’t have disease sensitive to PARP inhibitors. PARP inhibitors were sort of a come along, passenger mutation and probably don’t really suggest sensitivity.

So, I think this is a suggestion that we definitely should be looking at cell-free DNA. We have indications to look for PI3 kinase mutations. And also, that we want to treat people as early as possible, so they don’t develop this additional APOBEC-mediated mutagenesis.

Dr Vidula is now running a trial with multiple centers, looking at PARP inhibition in patients who have these mutations found in cell-free DNA without known germline mutations. And this, which has rich correlative studies, I think will be really important in terms of better understanding that role.

I mention the PARP inhibitors. There are several trials that are going on, MEDIOLA and DORA are two, and then the Phase III KEYLYNK-009 trial, that is treating patients who have triple-negative metastatic breast cancer in the first-line setting regardless of PD-L1 status, with gem/carbo with pembrolizumab, and then randomizing patients who have stable or responding disease to receive olaparib and pembrolizumab or to continue with their chemotherapy and pembro and then they can continue with pembro alone if they have toxicity from chemotherapy. So really important study that will be looking at progression-free survival and overall survival and is currently enrolling.

So, PARP inhibitors are clearly an effective treatment option for patients with advanced/metastatic breast cancer in germline BRCA mutation. We saw a benefit in progression-free survival with talazoparib and olaparib and a suggestion that there may be an improvement in overall survival if you treat earlier before you get multiple mechanisms of resistance. There was a positive impact on quality of life symptoms and functioning relative to chemotherapy.

We need to understand the optimal treatment sequence, whether or not we should be using chemotherapy as induction, and then maintenance. The benefit of adding PARP inhibitors to checkpoint inhibitors. And also, this very fascinating area of looking at alternative mutations such as PALB2 that are found in germline testing, as well as somatic mutations and homologous recombination defect.

This is a fascinating area and there is a trial called the SUMMIT trial, that has been looking at patients who have somatic mutations or really just mutations in HER2. And I think this area is really expanding, where we’re looking at data that we obtained from NGS and tissue and now in blood to really help direct treatment for our patients as we get further to precision medicine in the treatment of breast cancer.

Case: A woman in her early 60s with dermatomyositis is diagnosed with ER-positive breast cancer and ovarian cancer and a germline BRCA1 mutation

DR RUGO: So now let me just tell you briefly a couple — a few cases that I’ve actually had in my clinic to tell you about. This actually is a fascinating patient, a 60-year-old woman who presented with a severe rash, whole body rash, in the summer of 2018. Developed muscle weakness. Difficulty swallowing. All sorts of things that I really didn’t even know you got with dermatomyositis. Anyway, she was very, very sick. She had abnormal live enzymes. You can see just diffuse, whole body failure with dermatomyositis.

So she was treated with steroids and mycophenolate mofetil and IVIG. Had waxing and waning symptoms. And in September of 2019 flared and was treated with high dose steroids and then a steroid taper.

At the time she went into see her rheumatologist, she pointed out some nodules in the skin, just to the left of her sternum at the end of her breast. So, it actually was right in the left breast, just in between her sternum and breast, and she had shortness of breath and chest tightness with back pain.

So she’s had nodules in the left inner breast and left axillary adenopathy, abnormal breast imaging, and a biopsy showed that she had an adenocarcinoma, thought to be of breast origin, weakly ER-positive, PR-negative, with a high KI67, and had positive disease in an axillary node as well.

A core biopsy confirmed this, very, I think, aggressive, 25% positive invasive breast cancer. And then she had extensive adenopathy in her chest as well as pelvic nodes and hypermetabolic peritoneal nodes. And this is fascinating because I worked really hard to get an FNA of the iliac and peritoneal nodes because I said, okay, you’ve a BRCA germline mutation that had been discovered. And what’s fascinating is, once she had this biopsy done, she was seeing the surgeon who was evaluating it, she had this strong family history for breast cancer that hadn’t been picked up in Rheumatology. And so, she had the germline testing that showed a BRCA1 mutation.

So, since you can get ovarian cancer too, I really pushed for these biopsies which had to be done by interventional radiology. And it showed probably a metastatic cancer of Mullerian origin or ovarian cancer. So, here you have a space of 2 different cancers and dermatomyositis, on steroids, and really sick.

So I actually gave her paclitaxel and carboplatin first, but her rash kept flaring and I couldn’t really tell what was going on, so I switched her to nab paclitaxel, and she continued the carboplatin. Interestingly, should any of you face this problem, each time I gave her chemotherapy her dermatomyositis flared, so she got trouble swallowing. I had to have her not only solid foods, etc. And it turns out that each time you were killing the cancer, like she got 2 doses and I couldn’t feel the cancer anymore, so it was huge tumor cell necrosis. And we were releasing whatever factors caused the dermatomyositis.

So, then she had an excellent response to the dermatomyositis, eventually calmed down. She had all the rashes, all you learned about in medical school about dermatomyositis, it all went away. She’s still on steroids. I had stopped the MMF and she continued on IVIG.

So, at that point, we said, well, look, you also have ovarian cancer. We’ll go ahead with maintenance. So I put her on olaparib, 300 mg twice a day, and also letrozole because her ER was 25% in the breast biopsy. And what’s been fascinating is, her dermatomyositis has almost completely gone away. She’s down to 5 mg of prednisone. And her last scan, which was just a month ago, shows absolutely no disease. She remains in complete remission. So, really exciting response. And I’m very hopeful that this will last a long time.

DNA repair gene alterations and response to PARP inhibitors in breast cancer

DR LOVE:  So, just a couple of questions. One is the ATM story. Do we know about ATM in general through difference cancers?

DR RUGO: In general, and I don't know this data in great detail, in general patients with ATM mutations do not respond well to PARP inhibition. And I think that that has been an overall finding. And also, that they tend to have a little bit more aggressive disease.

But I think there’s a lot of interest in trying to understand whether or not — like, for example, they’ve even looked in lymphomas on this. And I think that there is some data about — in cell lines that you actually need something else. So you maybe the preclinical data suggests that you could give a PARP inhibitor, but you probably need something else to try and activate the response to the PARP inhibitor.

DR LOVE: So, another question is you showed some data or a patient, I think it was, or maybe more than one, who had both BRCA1 and BRCA 2 germline? You can have that?

DR RUGO: I didn’t show a patient myself, but there are patients. Yes, and of course I think we know the most famous person who had both mutations who was public about it, Angelina Jolie, the movie actress, she —

DR LOVE: Really? She had both? Both?

DR RUGO: Both BRCA1 and BRCA2.

DR LOVE: Really?

DR RUGO: And at 35, she had bilateral mastectomy. She had very strong family history of breast and ovarian cancer. And then, she subsequently had her ovaries removed as well. Hence, the big, early family.

DR LOVE: Wow. Interesting. Another question is related to the trial that you were showing with veliparib, where they saw a benefit with somatic BRCA, but they didn’t see it with germline. Do you think that's for real? Does that have something to do with numbers? Why would that be?

DR RUGO: Yeah, it doesn’t make a lot of sense. I think we need to understand the exposure to the PARP inhibitor and the use of maintenance in the patients with germline BRCA mutations. And patients who went off because maybe they weren’t tolerating it and they could go on the PARP inhibitor alone, then they were off study. You really need to get into the sort of weeds to understand that. In truth, that doesn’t make a lot of sense to me. It is a little bit worrisome that you didn’t see the benefit in the germline patients, even though it was a small number, and what that means to the somatic population, I don’t know. I think the somatic population just needs to be evaluated in a little bit more detail.

As you know, in the TNT trial, there was an analysis of the homologous recombination defect, and it didn’t seem to have any impact on differential response to carboplatin versus docetaxel. So, in that study, in a larger number of patients, we didn’t see a difference in terms of platinum response. Whether or not it differentiates patients who benefit when you add a PARP inhibitor to a platinum.

DR LOVE: PARP inhibitor to platinum.

DR RUGO: A platinum to — when you add a PARP inhibitor to a platinum. Yeah.

DR LOVE: So, one final question. For a couple of years now we have consistently seen when we ask general medical oncologist about first-line therapy of, for example, a patient with breast cancer and BRCA germline mutation, a common response, sometimes even the most common response, is a platinum-based chemotherapy followed by PARP maintenance, something they do all the time in ovary as far as I know. I don’t think I’ve even seen data in breast. But what do you think about that as a strategy? It's kind of going beyond the trial data, but do you think that's a rational strategy or something that should be avoided?

DR RUGO: In patients who have germline mutations?

DR LOVE: Yep.

DR RUGO: Yeah. I think that the BROCADE trial, even though it was done differently where you got veliparib but with the chemotherapy, I think it supports that approach. You really get a lot of toxicity when you add the PARP inhibitor to the chemotherapy. Although, we don’t have data on what would have happened if they just gave chemo and then the PARP maintenance. I think it really does support that approach. And it makes so much sense to me. You take out the patients who had PD-L1-positive disease, where maybe in the future we’ll be adding the PARP inhibitor to the maintenance checkpoint inhibition. But if you take out those patients and you have a patient who has germline BRCA mutation, I think it makes a lot of sense to give an induction chemotherapy followed by PARP maintenance. And I do think that, as you pointed out, the community at large thinks that's a good idea as well, sort of following along in the ovarian cancer footsteps.

When you have hormone receptor-positive disease, it's a little bit different because I think you do want to give your PARP inhibitor earlier on in the course of treatment. And we’ve seen benefits in those patients. I have one case where I gave a PARP inhibitor with endocrine therapy. And we just have no idea if endocrine therapy adds to the PARP inhibition in that setting. But since there’s not overlapping toxicities, we do have tendency to do it.

PARP inhibitors in patients with advanced ovarian cancer (OC) — Ursula Matulonis, MD

DR MATULONIS: Welcome, I’m Ursula Matulonis. I’m a medical oncologist from the Dana-Farber Cancer Institute, and it’s a real pleasure for me to talk to you today about the use of PARP inhibitors in patients with advanced ovarian cancer. These are the PARP inhibitors that are in clinical use. Specifically the 4 on the right are the ones that are FDA approved. The 3 in the middle are the ones that we use in ovarian cancer right now.

And the mechanisms of action of PARP inhibitors really fall into 2 main categories. One is that these drugs inhibit the function of PARP, and this is an enzyme that’s critical in single-strand DNA break repair. And then secondly is the formation of PARP DNA complexes through the concept of PARP trapping.

The middle 3 PARP inhibitors that are used for ovarian cancer are rucaparib, olaparib, and niraparib. And these are all 3 very good PARP inhibitors and they’re all very good PARP trappers. The drug on the right, all the way to the right, is talazoparib, which is the most potent PARP trapper. And then the drug all the way to the left is veliparib, which is a very weak PARP trapper, and hence why it can be combined with chemotherapy, and we’ll talk about some of those trials today.

As far as we know, the middle 3 PARP inhibitors, the ones that we use in ovarian cancer, rucaparib, olaparib, and niraparib, although there are differences amongst these 3 drugs, they all are very equivalent in terms of their PARP inhibition and activity.

So I’m going to go through the trials that have been done and reported on and acted upon by the FDA for patients with advanced high-grade serous/high-grade ovarian cancer. So the first one’s called SOLO-1. And this study really is specific because it involved patients who either have a germline or somatic deleterious BRCA mutation. Most of the patients enrolled had an underlying germline BRCA mutation. All patients have advanced cancer, Stage III or IV.

As I mentioned before, all these patients have high-grade serous carcinoma, and that’s really important because this is a histologic subtype of ovarian cancer that has underlying DNA repair problems, or underlying homologous recombination deficiency. Patients can either have undergone up-front surgery or an attempt at up-front surgery, then received chemotherapy, and the chemotherapy’s obviously platinum and taxane-based regimen.

And at the completion of chemotherapy, if they’re cancer responded to the chemotherapy, so again, key eligibility criteria, advanced, Stage III/Stage IV cancer, high-grade serous carcinoma, completion of chemotherapy and showing a response to chemotherapy, so distilling out a patient population whose cancers are quite HRD, and patients were randomized to either olaparib 300 mg twice daily or placebo in a 2:1 randomization. And it’s important to note that patients were on olaparib for 2 years or a placebo.

If patients had a partial response after completion of 2 years of olaparib they could continue on. But if there was a CR or near CR the trial design was to stop at 2 years. And the primary endpoint is investigator-assessed progression-free survival, and basically how long the cancer stays in remission for.

This is the PRIMA trial design. This is using the PARP inhibitor niraparib. So this expands beyond BRCA-mutated patients to include non-BRCA-mutated patients as well. But again, patients have advanced cancer and had to have high-grade cancer, high-grade serous carcinoma. And at the completion of chemotherapy, if there’s a response to chemotherapy, were randomized 2:1 to niraparib or to placebo. Here the primary endpoint was progression-free survival by blinded independent central review.

And patients here received at least 3 years of treatment, which is slightly different than the SOLO-1 trial. And patients here were stratified based upon whether or not they received neoadjuvant chemotherapy, what their best response to treatment was, and then what their HRD status was based upon the Myriad HRD assay, either deficient or proficient or not determined.

And then patients were also dosed based upon the body weight and platelet status. So if patients were more than 77 kilos and also had a platelet count greater than 150, they started off at the 300 mg daily dose. However, if they were lighter, less than 77 kilos, and/or platelets less than 150, started off with the 200 mg dosing. And that’s now written into the FDA package insert.

PAOLA-1 is a trial that used olaparib, and these were patients who were newly diagnosed, again Stage III or IV, high-grade serous or high-grade endometrioid tumor. They received some degree of bevacizumab with chemotherapy. So this is different than the first 2 trials. This trial actually incorporated bevacizumab, and they had to have at least 3 cycles of bevacizumab before getting onto the trial. Then patients were randomized to either olaparib at the usual dose of 300 mg twice daily for 2 years or bevacizumab or then they were randomized to placebo and bevacizumab. And it’s important that the bev was given for a total of 15 months. And here the primary endpoint was investigator-assessed progression-free survival.

The veliparib trial, or VELIA trial, again, here remember, because this PARP inhibitor is a weaker PARP trapper it can be combined with chemotherapy. So these are patients with advanced ovarian cancer, but instead of distilling out a patient population whose cancers are going to be responsive to chemotherapy, patients started off on the trial at the beginning of chemotherapy.

So there’s either adding veliparib to chemotherapy, and then using veliparib as maintenance, and that’s the top group. The middle group is veliparib combination only, so adding veliparib to chemotherapy, no maintenance, and then the last group, in green, is just carboplatin/paclitaxel alone. And here primary endpoint is, again, progression-free survival.

And these are high-level results that we’re seeing, and you can see that for SOLO-1 that does have the most impressive hazard ratio of 0.3. But these are really kind of top-level results of the entire populations of these studies. And you can see that in all of the trials, and again they’re all slightly different in terms of patient population, as well as eligibility and also study design and drugs that are added. Obviously veliparib’s added to chemotherapy. PAOLA-1 adds bev to chemotherapy and then continues it as maintenance therapy. You can see that all of the results for progression-free survival favored patients receiving the PARP inhibitor.

So when you think about the different subtypes. And we’ve learned this from the recurrence trials, and we’ll talk about that as well. Those were the studies that really led us to say okay, you know what, we’re seeing the best improvement in progression-free survival when a PARP inhibitor is given to a patient with a BRCA mutation. And that’s what we see here. So these are the patients pulled out of these different studies who have an underlying BRCA mutation. Again, very impressive hazard ratios. Not dissimilar to the trials in recurrence setting, as well. So these are very impressive trials.

And for SOLO-1, this is one of the trials that we have the most updated and most mature data, so on the left are the initial results from ESMO 2018. This presentation, this New England Journal of Medicine paper, led to the rather immediate FDA approval of the use of olaparib in patients with advanced ovarian cancer, newly diagnosed and responds to chemotherapy, who have an underlying deleterious BRCA1 or 2 mutation. So here you can see the impressive results. The median progression-free survival was actually not reached at the initial presentation, and about a 14-month progression-free survival for patients who were receiving placebo.

At ESMO this year we’re now seeing more mature results, and these are approximately 5-year results, showing olaparib, the median progression-free survival’s about 56 months, and placebo is still 13.8 median PFS. So impressive, and you really get the sense that we potentially may be curing patients. And it’s actually incredibly exciting, and I think this is one of the most important pieces of kind of data that we have to treat our patients with newly diagnosed ovarian cancer.

Importantly, and I’ll talk more about this coming up, is that there are no additional cases of myelodysplastic syndrome or AML reported. And the incidence right now on olaparib is less than 1.5% in the newly diagnosed setting.

Looking specifically at the BRCA wild type, so this is going to be BRCA wild type HRD, as well as HR proficient, kind of a mix, so that’s why the patient numbers are so high here. PRIMA did not specifically study that, but PAOLA-1, yeah, there’s a trend towards improvement, but that PFS, numerically, you can see is quite impressive. The same with the VELIA trial.

However, if you pull out HRD deficient but BRCA wild type, here we start to see a difference, clearly, in cancers that are not necessarily driven by BRCA but are driven by other genomic changes that lead that cancer to be HRD. So PRIMA has a hazard ratio of 0.5. PAOLA-1 has a hazard ratio of 0.43. And VELIA, here, again, I think this really shows you the limited nature of this PARP inhibitor. That the results are just not quite as robust. And again, there’s a little bit of a difference here because at the bottom you can see how HRD was defined based upon the Myriad HRD test. For PRIMA and PAOLA HRD had to be 42 or higher. For VELIA there were more patients included, and that may have altered some of the results, certainly.

For BRCA wild-type HR proficient, for PRIMA there is a hazard ratio of around 0.7. But again numerically that’s a little bit less than 3 months total improvement in progression-free survival. For PAOLA-1 there is really no improvement, so the FDA did not give PAOLA-1 an FDA approval for adding olaparib to bev for HRP tumors. And then VELIA also quite unimpressive. And this is really a way that one can help patients make decisions about should I receive a PARP inhibitor up front, yes or no, certainly looking at underlying BRCA positivity, and then also incorporating the HRD test if necessary.

So this is the current status of PARP inhibitors for newly diagnosed patients, so olaparib has 2 approvals. One is the maintenance treatment of adult patients with a germline or somatic deleterious BRCA-mutated cancer who have a response to front-line platinum-based chemotherapy, and that’s for 2 years. The second is combined with bevacizumab for the maintenance of patients with advanced ovary cancer who again are in response to first-line platinum-based chemotherapy combined with bevacizumab. But their cancer has to exhibit HRD, so HRD positivity here, but not homologous recombination proficient, but having that Myriad score of 42 or higher.

And then specifically for niraparib, at the bottom, the PRIMA study did lead to the approval of maintenance treatment for patients with a CR or PR, so basically a response to chemotherapy, the first-line treatment, really regardless of histology. All of these are regardless of histology. We just have to remember where to use these drugs, and also regardless of BRCA status or HRD status.

So what about patients with recurrent ovarian cancer? So unfortunately most of our patients with ovary cancer who are diagnosed with advanced cancer, will eventually relapse. And these are the different strategies for platinum-sensitive ovarian cancer. So the top is basically using platinum doublet chemotherapy, so carboplatin/paclitaxel, carboplatin/liposomal doxorubicin, carboplatin/gemcitabine, et cetera. The middle is analogous to what we were talking about in the up-front setting, and is giving patients platinum-based chemotherapy and then maintaining that response with PARP inhibitor. The bottom is adding bev, so adding bevacizumab to carboplatin doublet chemotherapy, and then continuing bevacizumab maintenance.

So the decision-making needs to occur at the start of platinum-based doublet chemotherapy and really should involve a careful assessment of the risks and benefits of all 3 of these treatment paradigms and treatment strategies. So there are important differences between PARP inhibitor and bev maintenance trials for recurrent ovarian cancer. So for trials testing bevacizumab maintenance, those are the OCEANS and GOG-0213, bevacizumab is started with the chemotherapy and then is used as maintenance once chemotherapy was completed.

However, for PARP inhibitor maintenance studies, because you cannot add PARP inhibitor, or real PARP inhibitor, to chemotherapy because of the overlapping bone marrow suppression, you induce a response to the chemotherapy and then maintain that response with a PARP inhibitor afterwards. However, if the patient’s cancer does not respond well to platinum-based chemotherapy, you should not use a PARP inhibitor afterwards.

And the pathology of the cancer, all PARP inhibitors trial require high-grade serous carcinoma. Some include high-grade endometrioid, but these are really high-grade serous carcinomas. For bevacizumab trials all histologies are allowed, and the questions arise which is the best strategy for a specific individual. But I really do think that clinicians can make these decisions based upon the patient, her comorbidities, what medications she’s taking, what histology she has, what was her prior response to other treatments, and it really is complex decision-making, but very doable decision-making.

The next 3 slides talk about the randomized Phase III studies in platinum-sensitive recurrent ovarian cancer for patients, again, with high-grade serous cancers testing PARP inhibitors as PARP inhibitor maintenance. So the NOVA study was the first to report out. It was the trial that first led to a PARP inhibitor as maintenance strategy in patients with relapsed platinum-sensitive ovarian cancer, and here patients were divided into either gBRCA-mutated on the left in green, and then in the red non-germline-BRCA-mutated patients.

SOLO-2, which was a follow up to study 19, focused on patients with underlying BRCA mutations, mostly germline, occasionally a few somatic; again, advanced cancer, high-grade serous, responding to chemotherapy, then get randomized to either PARP inhibitor versus placebo.

And then the ARIEL3 trial testing the PARP inhibitor rucaparib, not dissimilar trial design of advanced patient, platinum sensitive, in response to chemotherapy, high-grade serous, randomized to either rucaparib or placebo.

And this graph really kind of shows a few points that I think are really important. Really the strikingly similar progression-free survival results regardless of the PARP inhibitor, and that’s how I’ve come to the conclusion that in terms of the ability to inhibit PARP and efficacy in that matter, these 3 drugs are identical. However, you do see increasingly less benefits with less HRD. So here, for the patients who have underlying BRCA-mutated cancer, the results, again, are strikingly similar. There’s really a great hazard ratio, great improvement in progression-free survival for patients receiving the PARP inhibitor versus placebo. And I have patients on some of these studies who are receiving these drugs for many, many years.

And then, for patients who are HRD positive, and that was seen for NOVA and ARIEL3, obviously SOLO-2 was only BRCA-mutated patients, again, there’s an improvement in PFS when using the PARP inhibitor versus placebo, but it is a little bit less so than BRCA-mutated patients. And then the third group, BRCA wild type, HRD negative, very similar to what we’ve seen for the up-front trials in terms of yes there’s a slight benefit, but it’s not overly impressive.

These are some, I think, very important results that came out from ASCO in 2020. And these are looking at some of the overall survival results from SOLO-2. And these, again, remember, these were high-grade serous, recurrent, BRCA-mutated patients. And here you see about a 13-month overall improvement. It did not reach statistical significance. You can see the p-value there. But still, there is some improvement. And then in the bottom, close to 40% of patients on the placebo arm crossed over to subsequent PARP inhibitor therapy. So in that group you’re certainly seeing an even more profound benefit, at the top, really includes those patients who crossed over.

So one of the adverse events of special interest has been this risk of myelodysplasia and acute myelogenous leukemia. And it is something that was brought up even back in 2014, 2013, when these drugs were first being presented to the FDA. And olaparib is in blue, placebo is in red, and you can see that the risk of MDS and AML for patients receiving olaparib in SOLO-2 was 8%. So it’s not trivial at all, and patients really do need to be reminded about this particular toxicity. And this is really quite in difference to the less 1.5% that we’re seeing when using olaparib in the newly diagnosed setting.

Case: A woman in her mid-60s with advanced OC and a germline BRCA1 mutation achieves an excellent response to maintenance olaparib

DR MATULONIS: Okay, so coming to some of my cases. This is a patient who’s 67. She presented with 3 to 4 weeks of abdominal bloating and constipation. CT scan revealed an omental cake, peritoneal nodules, ascites, and a 6 cm pelvic mass. Her CA-125 was around 1,200. So typical presentation.

She went to primary cytoreductive surgery and was diagnosed with Stage IIIC high-grade serous ovarian cancer. She actually started off with intraperitoneal cisplatin, I still do that for a few patients, very selectively though, 100 mg/m2, and IP and IV paclitaxel. But because of toxicities of the IP cis, and she had hyponatremia, orthostatic hypertension, actually had a syncopal episode in clinic when we were bringing her back for saline, I then switched her to IV carboplatin and paclitaxel. She received a total of 6 cycles of treatment, and during treatment she was found to have, unbeknownst to her, a germline BRCA1 mutation.

So after chemotherapy finished up, we started her on olaparib, and her counts were pretty good when we started. She had some mild nausea and fatigue during the first 2 months of treatment, very typical. And these side effects have abated. By March of 2020, which is of course when COVID was starting, her hemoglobin went down a little bit, but had been pretty okay. June 2020, she had a PET scan that showed NED. By just October, when I last saw her just a couple weeks ago, hemoglobin was 12, hematocrit was 35.4%, CA-125 is 9. She’s current NED. She’s very happy, and the plan is for 2 years of olaparib, so finishing up in November 2021 and then stopping.

Case: A woman in her early 60s with homologous recombination deficient advanced OC receives niraparib

DR MATULONIS: Second case is a lovely 61-year-old woman who presented with several weeks of increasing abdominal pain and CT scan showed peritoneal carcinomatosis, which you can see by the arrows, omental caking, small-volume ascites, and a pelvic mass. And her CA-125 was 320.

She went to laparoscopy in an attempt to determine if she was going to be an up-front cytoreductive candidate, but she was deemed not an up-front debulking candidate, and multiple biopsies showed high-grade serous carcinoma, germline and somatic BRCA wild type. And she was started on neoadjuvant IV carboplatin and paclitaxel. Her CA-125 fell down to 24 after 3 cycles of chemotherapy, and she had a nice improvement in her CT scan.

She then went to interval cytoreductive surgery, had an R0 resection, finished up 3 cycles of carboplatin/paclitaxel. HRD testing was positive. Her weight was 62 kilos, platelet count was 200, so I started her on niraparib 200 mg daily earlier this year, checked her counts weekly for the first month, as per the FDA guidelines and what we know about this drug. And by the third week of niraparib her platelet count had fallen from 264 down to 110, held the drug for a week, made sure her platelet count was okay, then restarted at 100 mg. And she’s been on this and has been fine since. And when I last saw her, she was doing well on niraparib 100 a day, had an NED CT scan in October, and her CA-125 is 8.

So thank you so much, a real pleasure to talk about this subject, which is something that is think has really impacted our patients very significantly.

Perspective on the use of neoadjuvant therapy for patients with OC; risks and benefits with PARP inhibitors for advanced OC

DR LOVE: If I could, I’d like to follow up with a couple of questions. Going back to your 2 cases, when you think about it, the first case, if that patient were to present today, and I don’t know whether you were involved in the care at that point, would you more likely have gone to neoadjuvant or sent the patient for surgery, which is what she actually had?

DR MATULONIS: Yeah, no. I think that all patients who come in with clearly advanced disease have to be seen by a gynecologic oncology surgeon. ICON8, which is a trial looking at different permutations of platinum and taxane-based chemotherapy, really shows that there’s a difference. I know that there are trials looking at up-front chemotherapy versus neoadjuvant, but this is a trial that shows us the results for patients who undergo up-front surgery versus neoadjuvant chemotherapy. And clearly there’s probably a biological difference. There has to be. But the outcomes are vastly different, with patients if they can undergo up-front surgery, they seem to do better compared to those patients who cannot undergo surgery and undergo neoadjuvant chemotherapy. So that surgeon has to make that determination.

DR LOVE: For the patient who’s going to get neoadjuvant therapy, has a lot of ascites, how do you think about bev in that situation?

DR MATULONIS: Yeah, so I have a conversation with the surgeon about if that patient has really a lot of ascites and that the bev might help the patient, as long as the surgeon who makes that decision to undergo interval surgery is okay with using bevacizumab, then that’s fine. I do think, though, that for most patients with advanced high-grade serous ovarian cancer, their cancer responds very well to up-front chemotherapy. So one might say okay, look, just like PAOLA-1 made that decision. So you start off with carboplatin/paclitaxel for a lot of patients, and then perhaps post interval cytoreductive surgery, based upon what you find at the time of surgery, then you make the decision about bev or no bev.

DR LOVE: And for those patients who you think should get bev, then how do you deal with the issue of PARP inhibitor?

DR MATULONIS: So I think PAOLA-1 gives us a lot of information about that. I think that the bevacizumab and the olaparib do not seem to be particularly synergistic. They are additive, and I think you need to think about them that way. So if the decision is to add bev, and that’s because the patient has ascites or significantly persistent disease post interval cytoreductive surgery, a CA-125 that’s really sluggish coming down, then you certainly consider adding bevacizumab. However, if the patient ends up being BRCA wild type, you do an HRD test, HRP, then I would not add a PARP inhibitor to that. But if the patient ends up being BRCA mutated, then clearly one would add olaparib for 2 years.

DR LOVE: So there’s the thought that in terms of efficacy the PARP inhibitors that we have available are very, very similar. I wonder if there’re people in practice who maybe even view them as interchangeable. The data, the indications are somewhat different. But I’m curious what your thoughts are about, even though maybe it’s not that valuable clinically from your perspective, just from a theoretical point of view, why you see a benefit. I mean you said it was not that significant, but it was, I think, in the range of 0.5, certainly not what you see, maybe a little higher, not what you see with BRCA, for example. But why you see a benefit, indirectly, comparing niraparib? You see it in the PRIMA trial, and you don’t see it in the PAOLA trial with olaparib. Do you think that could be some intrinsic thing related to actually efficacy or some other explanation?

DR MATULONIS: Well, I think it’s complicated with PAOLA because then you’ve added bevacizumab. So I just don’t think there’s a negative interaction. I just cannot believe that. I think it ends up probably being patient selection. Are patients taking the drug? Are they stopping the drug? I still think that PARP inhibitors are really similar in terms of efficacy standpoint. They’re different in terms of toxicities and how patients tolerate them, but I just can’t imagine at all that 1 drug is better than the other. There’s just no data to support that.

DR LOVE: So another thing I want to ask you about is MDS, and you mentioned the data with the SOLO-2 trial with the 8%.

So when you talk to a patient, do you say the risk is 8%?

DR MATULONIS: Well, there are a few things. So that data that’s shown there, again, also for patients receiving placebo. Remember, 40% of patients crossed over to a PARP inhibitor. So that data’s not taking out those 40% of patients who went on to receive a PARP inhibitor later on.

In SOLO-2, there are patients who have underlying BRCA mutations. And we’ll get more data on this as time goes on, but I think there’s a thought that mutation carriers are going to have a higher risk of AML/MDS, and that’s why we’re seeing such a high risk. And I think we’ll see this with NOVA and with ARIEL3 and with PRIMA, et cetera, as we get more long-term data from these trials.

These patients have all had previous chemotherapy, at least 2 rounds of chemotherapy, for SOLO-2, right? So they had their up-front carboplatin/paclitaxel for newly diagnosed disease, they recur, they have at least another round of carboplatin/paclitaxel, or carboplatin/something, and then get the PARP inhibitor. So they’re already at a higher risk of AML/MDS, at least we think so, because of the more previous treatment. Also for SOLO-2, an important difference compared to SOLO-1, SOLO-2 you can continue the PARP inhibitor indefinitely.

And I mentioned I do have patients on PARP inhibitors for many years. I have a few patients who are on it for 10 years. For SOLO-1, it should be a hard stop at 2 years, per the trial. So you’re potentially changing that risk based upon the duration of the PARP inhibitor and also how much treatment the patient has previously received.

So in answer to your question, yes there’s going to be a risk for patients who are on the placebo arm, but as I said, 40% crossed over, they’re BRCA mutation carriers, and they’ve received at least 2 prior lines of chemotherapy, which in itself increases the risk for developing AML/MDS.

DR LOVE: That’s interesting though because I haven’t heard that before. Are you saying that if somebody has a BRCA germline mutation that they have an increased risk of getting MDS like just say related to chemo? Put aside the PARP.

DR MATULONIS: Yeah. I think we’re going to get more data showing that, but you’re already getting a sense from SOLO-2 that 8% risk, compared to other data sets, at least from what NOVA presented initially, that the risk was 1.5%. And again, I think with longer follow up, and also on these trials really understanding what exactly the follow up is. Because again, I don’t think that 1 PARP inhibitor has less MDS or AML toxicities. But just in my experience, and getting back to your question, again, personal experience, it does seem that the risk for AML/MDS seems a little bit higher for patients who have underlying BRCA mutations. And again, I think we don’t have data on that, at least we don’t have randomized prospective data. But these trials will sort that out, I think.

DR LOVE: That’s really interesting. One other thing I want to ask you about is a more macro question, which is the issue that right now you have progression-free survival with these strategies, but not overall survival, correct?

DR MATULONIS: So for the up-front setting, that’s correct. We have progression-free survival. For the recurrent setting there’s SOLO-2 data showing some OS benefit of about 13 months that was not statistically significant, but still, it’s 13 months. That’s pretty impressive. But you’re correct, for NOVA and for ARIEL3 we do not have any survival data yet.

DR LOVE: So it’s interesting, I’ve been really reflecting a lot, and we’ve been talking a lot, about the ADAURA study in lung cancer that was just presented at ASCO of adjuvant osimertinib in EGFR-positive disease. They actually had a better, I think it was 0.2 or lower, slightly even better than what you all saw with PARP in BRCA germline. And there, there’s Dr Bruce Johnson and all in your group. We’ve been starting to talk to all the lung cancer people. And it’s interesting, they’re excited about it, of course it’s a fantastic advance, but a lot of people are bringing up the issue of is this going to affect survival, because we know osimertinib is very effective in relapsed disease. So the question is do you use it adjuvant or you wait. And it looks like most people are going to go forward with adjuvant.

But I’m just kind of curious about the philosophy of the gynecologic community because you all just jumped right on top of this without any survival data. And to me, in a way it’s sort of similar, I mean they’re both bad diseases when you recur. Any thoughts about that?

DR MATULONIS: Yeah, it’s a good point. I think we’re so excited when we see any significant progression-free survival benefit. That’s unique and that’s really one of the first times we’ve seen that. The FDA has definitely changed their stance from really requiring OS to a significant improvement in progression-free survival. So that’s a big change.

I think for patients with BRCA-mutated cancer, advanced disease, high-grade serous, the SOLO-1 data, you cannot ignore it. And I think it’s absolutely standard of care. And I think if we’re not giving those patients a PARP inhibitor that is wrong because you’re going to minimize the long-term toxicities of the PARP inhibitor in a patient who’s just been treated with 1 round of chemotherapy. You’re limiting the duration of the PARP inhibitor for 2 years. You’re seeing a tail on that progression-free survival curve that is highly suggestive that perhaps we’re going to see patients who are cured of this cancer, which would be incredible.

And then it also follows the biology, that the more chemotherapy that we use, the more PARP inhibitors we use, we see drug resistance. We see chemotherapy resistance, increasing platinum resistance, and then all the different mechanisms of PARP inhibitor resistance. And in a single patient it’s not just BRCA reversion mutations, but it’s other multiple mechanisms of PARP inhibitor resistance.

So then in the recurrent setting you’re treating a patient whose cancer is much more complex, and I think it’s harder to treat. And for the most part those patients, I mean I do have a few patients who have been long-term survivors of their recurrence who’ve been either just treated with chemotherapy or a PARP inhibitor. So it’s possible, but it’s really rare. I mean most of our patients with advanced disease unfortunately will succumb to their cancer.

So I do think you’re correct. Around the BRCA wild type there is a PFS benefit there. It would be great to see overall survival benefit, as well, but I think we may be changing the natural history of these cancers, perhaps. We’ll have to see. But I think for those tumors that are HRD the PFS results are impressive. I do think for HRP that it has to be a discussion with the patient, and really the risks and the benefits about the use of a PARP inhibitor.

DR LOVE: That’s so interesting putting myself in the place of a general medical oncologist listening to different groups, kind of philosophies about how they approach the disease. Just interesting philosophies.

DR MATULONIS: But it’s really taking ovarian cancer to looking at the pathology and taking that pathology and understanding the molecular underpinnings. So for example, low-grade serous carcinoma. I mean now there’s a GOG trial, or NRG trial, that’s looking at either up-front aromatase inhibitor versus chemotherapy, really taking away the chemotherapy. So you’re really basing decisions on the biology of the tumor, which is different, really different, I think, than we’ve ever thought about ovary cancer before, in a really meaningful way. And I think thinking about it this way is really going to improve survival and improve our patients’ lives because we’re going to be thoughtful about what we’re doing, and patients will benefit from that.

DR LOVE: So one final question. I’m curious about your thoughts. I know you’ve been involved with a lot of research looking at the idea of combining a checkpoint inhibitor and a PARP inhibitor. And we actually saw some data at the ESMO meeting from the MEDIOLA study, or one of the MEDIOLA studies, that was pretty interesting, particularly in the BRCA wild type. Any comment about that strategy moving forward of adding in, and there they added bev as well, to the BRCA wild-type patient?

DR MATULONIS: So I think we’ll have to find out with trials. I think preclinically, so in a mouse, in a genetically-engineered mouse, and there have been several studies that have looked at this. We’ve talked this before, you get activation of the STING pathway, and there’s a real synergy with using a PARP inhibitor and an anti-PD-1 or anti-PD-L1. I still think that for immunotherapy, IO/checkpoint inhibitors for high-grade serous ovarian cancer have yielded not such impressive results. And that’s from single-agent checkpoint inhibitors in the recurrence setting. It’s the addition of a checkpoint inhibitor to let’s say pegylated liposomal doxorubicin, as we saw in JAVELIN 200, and then more recently adding a checkpoint inhibitor to platinum-based chemotherapy for newly diagnosed disease. And in none of those instances are we seeing particularly impressive results.

So I think we’ll see. I think I personally would not do it off of a clinical trial. There’re plenty of clinical trials that are testing this concept of PARP inhibition plus anti-PD-1 or PD-L1, sometimes adding bevacizumab into it, sometimes adding other drugs into it. And when those trials report out we’ll have our answer, or at least potentially part of the answer.

DNA repair and PARP inhibitors in pancreatic cancer — Philip A Philip, MD, PhD, FRCP

DR PHILIP: What I’m going to do today is talk about pancreatic cancer, but with a special focus on DNA repair and PARP inhibitors.

One of the things which has happened is that a lot of work has been done on the molecular level to see, what is really the pancreatic cancer, what drives pancreatic cancer? What maintains pancreatic cancer? And really, what is the reason why some drugs are active and others aren’t? And why do you get resistant to treatment? So that science has really improved significantly over the last 10, 15 years. But obviously, for today, I’m not going to talk about everything, I’m just going to focus on one aspect of it, one aspect of the biology of pancreatic cancer, which is DNA repair.

It's estimated that 12- to 15% of the patients have some form of DNA repair abnormality. And that, by itself, can be utilized for treatment specific for that sort of molecular change that happens in the tumors. And the one which is really more studied at this time, is the BRCA1 and 2 mutations.

So, all of you are familiar with the concept of synthetic lethality, which we see in patients or tumors that have either BRCA1 or 2 mutations, where you can induce lethality in the cell, in the cancer cell, by using a drug, in this case would be a PARP inhibitor.

And the nice thing about this is that, although all cells in the body will have the BRCA1 or 2 mutations, that particular lethality is going to be specific to the cancer cells.

So this is something that's really very important, but at the same time, with so many drugs available and all the information we have on other cancers like breast and prostate cancer, so mostly we know that biology very well. But just to remind you that it's the concept of systemic lethality, taking into our advantage.

So there are a number of drugs have currently been tested, some of them in the market. But something you have to be aware of is that the PARP inhibition is really something which has been attacked by a number of drugs, but the potency of these drugs and the effectiveness of these drugs, and the scientific index of these drugs, is not the same.

I would like to show you this slide showing that, for example, olaparib/rucaparib, in terms of potency, they are more active than veliparib, which was the earlier drug that was being tested in patients.

So this is just to give you an idea that PARP inhibition is not equal to all the drugs that we have at this time.

Now, what about the BRCA1 and 2 mutations? How often do we see them? Certainly, there’s some variation, depending on ethnic background, in terms of, for example, the United States, where the patients are. However, I would say, on average, we see it in 5- to 7% of the patients who will have BRCA1 or 2 mutations. And I can tell you now that the BRCA2 mutations are seen more frequently than the BRCA1 mutations in those patients. And of course here I’m referring to germline mutations. However, there’s also a question on whether the BRCA1 or 2 mutations in the tumor itself, without having a germline background, is that also going to be similar when we come to biology or phenotype?

As you all know, NCCN guidelines clearly indicate that newly diagnosed patients with pancreatic cancer must be considered for germline testing. And the reason for that is that studies have shown that in patients who have BRCA1 or 2 mutations, they don’t necessarily have to have a family history. And, therefore, we could be waiting for someone who has a family history of breast cancer, ovarian or pancreatic or prostate to really go after the testing. You have to do it on all patients who are newly diagnosed. So this is something which has now become or becoming increasingly part of our standard of care in the management of pancreatic cancer.

Now, I just want to talk about the pivotal trial, which is POLO. And this was a Phase III trial which included many countries, and this was a study that randomized patients to receive olaparib, 300 mg daily, versus placebo. Now the key thing here to remember is that, number one, this was in patients who have germline mutation. Number two, patients should have received platinum-based treatment for at least 4 months. But not only that, they should have had either response to the platinum compound or stable disease.

And of course, there are things in relation to a performance status being 0, 1. All patients had metastatic disease. And ultimately, after screening for more than 3,000 patients, they ended up randomizing 154 patients who were positive for the germline mutation.

This should give you a very, very brief idea of the type of the patients who went on the study. You can see that the median age was 57, so that's younger than what we see in other trials we’ve gone through, which also goes back to the BRCA1 or 2 mutations — that can happen in the very younger patients. They don’t have to be very young, by the way, but a bit younger than the patients we see on average.

As I mentioned, most of the patients had BRCA2 mutations. And the treatment they received, although they could have received any platinum-based treatment, but, in fact, they mostly FOLFIRINOX, as you would expect. But there were some patients who received gemcitabine/cisplatin.

And, in terms of the response to the chemotherapy, which they had at the time they were enrolled into the study, was approximately 50%. Now you may argue and say they might even see a better response rate, but this is again at the timepoint of 4 months or so. I’ll come back to that in one of the studies that I’m going to also focus on with gemcitabine/cisplatin.

Now when you look at the primary endpoint of the trial, which was progression-free survival, there was a very nice separation of the curve. You can see here that the patients who had the olaparib did much better than patients who received only placebo, or, in other words, best supportive care only. So there was a significant improvement. And you can see that that continued for the duration of the follow-up and the median progression-free survival was over 7 months, versus less than 4 months. So, certainly a study that was positive. And this really led to the approval of the FDA of olaparib, almost 10 months ago, as a treatment in the maintenance after response or stability to platinum-based treatment.

Now, since the study was published, people went back and looked at the different aspects of the trial that will give us some useful information about how good this treatment can be, in terms of different subgroups of patients. But one of the things which people wanted to look at was, how does it really affect the different types of mutations? So when we say BRCA1 or 2 mutations, there are different types of mutations involving different parts of the gene. And, in fact, this treatment worked in all these different mutations the patients had. So as long as the patient has a germline mutation that's validated to be a mutation that affects the function of the gene, this treatment is really working. Very simple.

And then people looked at subgroup analyses, and you can see that in this initial look, which was part of the initial study, in fact, all the subgroups — almost all the subgroups had a favorable response to the treatment. So it wasn’t really a clinical or pathological criteria that would tell us a patient is not going to respond or respond. In fact, the response was across all the subgroups. And then they went back and did another analysis, more recently, which was presented at ESMO, again the same conclusion that if you look at the subgroup analyses here, you can see that almost all of them had benefit from the treatment compared to the placebo.

Now one of the interesting things which I find here is that patients who had the olaparib, they had not only better progression-free survival, but also, they continue to respond to treatments. So the carry-over response and augmentation of their response continued even after they stopped the chemotherapy and they continued on taking the oral pill, olaparib, compared to again, the patients who received the placebo. And the mediation duration of response was also longer. And also, interestingly, the patients had longer median time to onset of the response.

So, certainly, one would consider this as something which is the biology of the disease also changing favorably to allow patients to respond longer, if you would, on the platinum treatment.

Looking at other aspects of the disease progression, so basically this is looking at time to second progression. So, after the first progression when they put back on chemotherapy, it could be any chemotherapy that was appropriate for the patient, but when you look at the time for the progression the second time, after, obviously, they’ve been off the olaparib, again the biology appears to be different because it's favoring patients who had received olaparib in terms of the time to their second progression. Again, something which is really interesting to see. So it’s of course, a bit more than just maintenance; it's just doing something to the disease that also changing its biology.

When they looked at overall survival, they did not see a difference. And this was something which is intriguing, because why do you see that improvement in the other parameters I just showed you; whereas, when it comes to the overall survival, we're not seeing a difference.

At this point in time, it's difficult to explain. Part of it is the maturity of the data and also the possibility that patients who received the placebo may have ended up getting the olaparib. Although, looking at those numbers, there are not really that many patients who did that, probably less than 20%. But certainly, this is something which is worth seeing what happens with time when we get more mature data.

Now one of the important questions that people ask is that, okay so you’re testing the drug versus a placebo, what’s the type of the patient experience in terms of side effects, in terms of quality of life? Now certainly, testing a drug against placebo you’re going to see more side effects. And the side effects that were seen are within the range what we expect with olaparib. So basically, the fatigue, nausea, some of the GI symptoms. So you see those. And you’re going to see them because you’re treating the patient with active treatment compared to the placebo.

And what about the rate of the interruptions, in terms of dose reductions? So if you look at — I’m going to start with the median duration of treatment, so that almost doubled. So it means that the treatment was working. But if you look at any toxicity and if you look at Grade 3 or more side effects, certainly there’s more in the patients who receive olaparib. Interruptions, a third of the patients undergo interruptions. But the rate of the discontinuation was certainly less than 10%, in fact, it was 5.5%. So, an important measure in terms of putting someone on a maintenance treatment, can you really maintain them.

So, certainly, there were patients who were discontinued, but that was really a low number because the treatment was managed with interruption and restarted treatment and also by dose reduction.

So here’s another study which was presented at ASCO GI and in fact was published. And in this trial, they showed that the quality of life in fact, was well preserved and in fact better than what was seen in patients who had the placebo. So this was really an important aspect of our thinking of does this type of treatment with side effects, especially when compared to placebo, doesn’t really make patients worse. In fact, according to this, probably not. You don’t see that happening.

And then there was another analysis which was presented in ASCO in 2020, looking at again analysis of TWiST, which, as you know, it's a well-known way of looking at how the treatment affects the patient’s quality of life, in terms of the survival. And there was superiority in patients who had the treatment with olaparib compared to the placebo. So that was the POLO trial.

Now the question becomes, in the POLO trial, patients were not really heavily pretreated. What about patients who are heavily pretreated, and they get olaparib single agent? And this is a study based on the TAPUR registry. And in this trial, they had 28 patients who had received olaparib. They were heavily pretreated. And you can see that there is a response rate of 7% and disease control rate, a third of the patients had disease control rate and the 1-year overall survival was close to 50%. So there is a benefit also in patients who have been heavily pretreated. And this is single-agent activity of olaparib.

Rucaparib is another drug which has also been tested as a monotherapy in patients who have BRCA mutations. I’m not sure whether this drug is being further developed in this disease. I just wanted to show you this so that you know that there is also another possibility. But this drug is not approved by FDA, so we’re not using it. And as I said, I’m not sure what’s going to happen with this drug.

And there is another study which deals with the same drug, rucaparib, but the reason why I’m showing you this slide is simple, because in this study they also showed that patients can respond to a PARP inhibitor with a somatic mutation. So, it's not only the germline mutation maybe that's going to drive us to use these medications, but certainly if someone has a somatic mutation you do the NGS on the patient who has BRCA1/2 mutation. It looks like those patients will also respond to the treatment or have a chance of responding to the treatment. Again, the quality of response may be different. I don’t know. No one has done a randomized trial, and no one will do. But certainly something to keep in mind.

So what happens in the future development of these agents? This is a study that has been approved to start, and at this point in time this PARP inhibitor, again going back to rucaparib, is going to be probably replaced by olaparib for logistic and other reasons, to do with sponsor. And this study is looking at patients who have resected pancreatic cancer. They go on adjuvant treatment, in this case FOLFIRNOX, and after the 6 months they get randomized to receive the PARP inhibitor versus a placebo.

So, this is looking at the use of maintenance treatment in the earlier stage disease, which makes a lot of sense, to try to improve the outcome of those patients who have undergone surgical resection.

There’s another area of active research in the use of PARP inhibitors, and one of them is that the treatment of tumors with a PARP inhibitor. This is preclinical data showing that if you do that, then there is an increase in the PD-L1 expression. So that's something which is being for us in pancreas cancer, where we didn’t do that well with immunotherapy, that's something which we’d like to see improvement in our ability to use immunotherapy. So the bottom line is that with PARP inhibitors, you increase the expression of PD-L1, therefore, that gives an opportunity to test the addition of a immune checkpoint inhibitor to the PARP inhibitor.

Sure enough, the GI Steering Committee recently approved that trial. And this study is going to start hopefully within the next few weeks. This is looking at patients with, again metastatic disease, who have germline mutations and either BRCA1 or 2, and the patients will be undergoing again chemotherapy with a platinum-based treatment for at least 4 months. They respond or stable disease. Maybe randomized to receive olaparib, so that's the POLO trial, versus olaparib plus pembrolizumab.

As I mentioned, this study is going to start hopefully within the next few weeks. It will accrue just under 100 patients. So it will be interesting to see what happens.

Now I just want to circle back to the issue of the induction treatment with a platinum compound. So, as you know, gemcitabine/cisplatin was a combination that was developed like close to 20 years ago, but we never got into using it because a Phase III trial didn’t really show a benefit. And then, fast forward, we got FOLFIRINOX, a platinum-based treatment. Now we’re going to back to gemcitabine/cisplatin as an alternative to FOLFIRINOX in patients who have BRCA1/2 mutations.

Dr O’Reilly recently published this article using gemcitabine/cisplatin plus/minus veliparib. Now, veliparib wasn’t active, he didn’t add anything — added maybe toxicity. But gemcitabine/cisplatin really gave us a very interesting response rate. So, in patients with the BRCA 1 or 2 mutations, cisplatin/gemcitabine is an option for the patient. And in fact, it's an option, especially in patients who may have not so good liver function because you’re not using a drug cleared by the liver. In FOLFIRNOX, you have irinotecan. Here you don’t have that. So, again, something to keep in mind.

A recent panel, which I was part of, we looked at what is the new thing in metastatic pancreatic cancer. Again, I don’t want to go into details of this article which was recently published. It now establishes targeting BRCA1 and 2 with olaparib as one treatment option for patients who have BRCA1 or 2 mutations. In fact, this article leads us to one important point which is, when do we test? What do we test in terms of molecular profiling? Something which wasn’t being done in patients with pancreatic cancer. It's something really new for us.

So this leads me to really putting the overall picture the way at least I practice. So, in my practice, I need to get the BRCA1 and 2 mutation testing soon and early, and there are ways you can do that. And you can do it in the blood, you can do it in the tumor itself.

Because that helps me to decide on the frontline treatment.

Now, if you’re going to use a platinum-based treatment, fine. Maybe you’re not really that much stretched for time to get the test done. But certainly for the frontline treatment, you need to get some idea about the BRCA1/2 mutations.

Now I also put here PALB2 mutations, because there’s also indication or suggestion that PALB2 germline mutations may also respond to BRCA — to olaparib-like drugs. So that's something which I put in there.

Now once you’re past your first line and you go to the second line, you will need other molecular information. Unfortunately, you have a very narrow field there. But fortunately, it's expanding with some several interesting drugs.

And the key thing in the second line would be, or beyond the first line, is to see whether someone has MSI high or stable. And MSI high is only 1- to 2%. But in patients who have KRAS wild-type, which represents 10% or so of the patients, there’s an increasing list of targets that you can use for using new generation treatment that we have. And more recently, I’m sure you heard that the KRAS mutation G12C tested mainly — or is evidenced in colorectal and lung cancer, these mutations, now there’s a drug which targets that mutation. And again, when you come to other mutations like NTRK fusion, NRG1, HER2, BRAF and others, so that's really in the space of the KRAS wild-type.

Now, certainly, KRAS mutations are not routine like we do in colorectal cancer, but if you do it in patients with pancreas cancer, then you may end up having some interesting drugs that you can target those patients. And that's really something important because patients who go after front line to second line, third line, that increases your ability to search for drug for that.

So, just to conclude, platinum compounds and olaparib show activity in patients who have BRCA1 or 2 germline mutations as per FDA approval for the olaparib, but there is also indirect evidence that it may also help in patients who have only somatic mutations. And in pancreatic cancer, the first drug that has been approved based on a biomarker is olaparib. So that's something which hopefully we will be expanding on.

And again, there is a very clinically meaningful prolongation of progression-free survival, but at this point in time we don’t have evidence that the survival has been improved.

And there are future directions in what we do in patients who have these mutations: Adjuvant treatment, a maintenance treatment using olaparib. And also in combination with immunotherapy, also being tested.

So, at this point in time, you really have to think of the genomic profiling and genetic testing for the germline mutations in all patients you see. And again, please keep in mind in patients who have wild type KRAS, there’s also an increasing list of targets that you can go after.

With that, I’d like to thank you again very much.

Case: A man in his mid-70s with locally advanced pancreatic cancer and a somatic BRCA1 mutation receives modified FOLFIRINOX followed by maintenance olaparib

DR LOVE: Let’s go through your cases.

So, maybe you can present this 75-year-old man. What happened there?

DR PHILIP: So this is a 75-year-old male, who’s otherwise in good shape. He was presenting with abdominal bloating and some mild abdominal pain. He did have an ultrasound scan and then MRCP which showed a pancreatic head mass, which measured approximately 3 cm. His CA19-9 was over 1000 and his performance status was bordering on 2, 1 to 2. And his past medical history had left nephrectomy for RCC, rheumatoid arthritis, and benign prostatic hypertrophy.

So in our tumor board, we said, okay, we looked over things and the surgeon wasn’t so sure. I think coupled with probably his performance status, we decided not a surgical candidate, but we thought that he could take the modified FOLFIRINOX that he wanted to do. And at this point in time, we also did the NGS and he had somatic BRCA1 mutation. And subsequent to that, actually we did the blood test. He also had germline mutation. He had Ashkenazi Jewish heritage.

So, this patient had a very good response to the treatment. As you can see, his CA19-9 went way down.

And we had some issues with him. He had fatigue grade 3, diarrhea Grade 2, neuropathy grade 1. And in this patient, we switched earlier, so we didn’t wait for the 8 cycles like the POLO. After 6 cycles we said, let’s give him olaparib, that he wanted to be on maintenance treatment. And he now continues on it for 5 months plus. And obviously, his fatigue/diarrhea all improved. So this is a good example of someone who you can now switch them to a treatment maybe even earlier, although that's not the way the POLO trial was done. But certainly, it gives an opportunity for a patient to be on an active treatment, which is more targeted therapy.

DR LOVE: Let me just clarify, he has localized disease, right?

DR PHILIP: It was read as borderline resectable 2, locally advanced. So, it was localized, yes.

DR LOVE: But he doesn’t have mets?

DR PHILIP: No, he didn’t have mets.

DR LOVE: So my question is, normally would you integrate radiation therapy into the treatment of a patient like that?

DR PHILIP: It's possible that we may even do radiation in this guy, but he was — I mean he’s one of those people who told us, what happens if — I heard about the drug. His daughter-in-law works in one of the drug companies. And we said, we’ll go ahead. We don’t have to do radiation. But we would do radiation more in patients who we see that the disease stable for at least 4 or 5 months, just to tell us — the test of the time that the patient is not going to develop metastatic disease.

DR LOVE: So, let me ask you, do you think that most GI oncologists, investigators, would do what you did in this case?

DR PHILIP: Again, this is a unique case because the patient — we put him on FOLFIRINOX, had side effects, and we felt that we may need to move earlier to give him the other drug, the olaparib, than waiting another 4 weeks. This is just tailoring the treatment for a patient’s needs. I mean that's the way it best worked for this patient.

Now, the issue of the radiation treatment is interesting, what you say. Because is radiation a possibility in this patient? Yeah, of course it's a possibility. And I’m not sure if I’m not really even going to give him the radiation. The only thing is that normally I wait — when I give chemotherapy for example, locally-advanced disease or even borderline resectable, I wait a good number of months just to make sure that they’re fine. In this patient, I’m seeing him being stabilized by a pill of olaparib. Radiation is not going to prolong his survival. We know that. None of the radiation so far, which is given either in the borderline resectable or locally advanced, has shown to prolong survival. If I’m going to control his disease by this pill, I’ll just continue with it. If he starts to show some disease progression there, then I can give him radiation therapy. And if develops metastatic disease, then I’ll be happy that I didn’t give him radiation treatment.

DR LOVE: Right. No, I was just curious whether most people would do that. So, again, I guess your thinking was probably even with a response he still was not going to be resectable?

DR PHILIP: Correct.

So, I guess if I were going to present this case in its most simple state, we would say you have this patient with a BRCA germline. They have unresectable disease. He’s not really symptomatic from it. So you gave him chemotherapy and he sort of maxxed out on that in terms of starting to have complications. And then the question is, what do you do next? So I guess theoretically you could do nothing. You could observe him. You could do what you did, which is to give him olaparib. I guess you could give him radiation therapy. I mean I don’t know what other options there might be.

Do you think that most people would do what you did, or do you think there’d be some variation?

DR PHILIP: It would be variation. One of them will be the radiation treatment. One would be why don’t you just cut down on the chemotherapy and give him lower dose of chemo? In my opinion, I thought, why don’t I try something which is targeted.

DR PHILIP: So I have this patient with what appears to be localized pancreatic cancer, who had a CA19-9 over 1000. The CA19-9 over 1000 always is a red flag for patients not really doing well on surgery, even if you think of it that way, even if they appear to be resectable. And in a patient like this who in our tumor board we decided he’s not going to be a good candidate for surgery, we gave him neoadjuvant chemotherapy, in a sense, or induction chemotherapy or just simply chemotherapy with FOLFIRINOX. We got him a dose reduction because of his age; it was modified FOLFIRINOX. And then he had a very good response to the treatment in terms of the reduction in the CA19-9 level. And after the first, second treatment, he was developing side effects. He was developing the symptoms of — he had the diarrhea. The most important thing for him was the fatigue, in someone who’s at his age, living on his own, in this era of COVID.

So, basically, the patient had several options here, continuing with the same treatment was going to be very tough for us because of what we were seeing in him. He was not doing well. So we had options. So one option was for us to cut down on the chemotherapy and bring it down to maybe minus oxaliplatin, maybe minus oxaliplatin and irinotecan and just give him 5-FU, or give him capecitabine single agent. So that was one option.

The other option was for us to give him a break from the treatment because he wasn’t feeling that well.

The third option was, based on our knowledge of ability of the olaparib to prolong progression-free time, so why not try that agent? And he would be closely monitored, so we will know what’s going on.

The other option was to give him radiation treatment, but at that point in time it was a bit early for me to decide on giving him radiation therapy because I normally like to see patients stable and not having disease progression or metastases before I decide on radiation treatment. Keeping in mind, that radiation therapy, neither in locally advanced disease, nor in the more resectable disease, has shown an evidence for prolonging survival.

So with that in mind, I made the decision to go for olaparib, thinking that it's going to be an easier treatment on the patient, given everything which was going on with him.

Case: A man in his early 50s is diagnosed with metastatic pancreatic cancer, elevated bilirubin and mutations in BRCA1, APC, KRAS and p53

DR LOVE: Let’s go to the next case.

DR PHILIP: So the next patient is a younger patient, he was 52 when he came to see me. He worked as a pediatric emergency room doctor. At the same time, his father was also undergoing chemotherapy for pancreatic cancer. He presented with painless jaundice and he had weight loss of 15 pounds. He also had steatorrhea.

The CAT scan showed a mass in the pancreas, but also, he had multiple liver lesions. And ERCP failed in this patient and therefore, he had to have an external biliary drain. And then he had a CT-guided biopsy of the liver which showed a moderately differentiated adenocarcinoma. At that point in time, his bilirubin was just over 10, with a CA19-9 which was over half a million with a C-reactive protein. And his overall performance status was a 2 at best.

And on your right, you can see the liver metastases and also the biliary drain that was put.

So in a patient that was like this, the decision was, what do you give them? Do you give them FOLFIRNOX? Because if I want to give him FOLFIRINOX, I have to wait a long time for his bilirubin to come down to under 3 and preferably under 2 to be safe. Do I give him FOLFOX without the irinotecan? Do I give him gemcitabine/nab paclitaxel and then also risk the issue of the liver dysfunction and the nab paclitaxel? Or do I give gemcitabine alone, which I would never do? Or I give him gemcitabine and cisplatin? Obviously, I decided to give him gemcitabine/cisplatin. This was a few years ago. We didn’t have Irene O’Reilly’s data.

So, he had full doses of gemcitabine/cisplatin. He finished 6 treatments. And early on there was normalization of the serum bilirubin. Could be combination of the drainage and chemotherapy. But the key thing is that his performance status become zero. And he was planning to go back to work. And he went to his honeymoon that was postponed because of his diagnosis.

We did molecular profiling and he had BRCA1 mutation, in addition to other mutations that we see in this disease. And he also had germline mutations, so we confirmed that.

He did have an excellent response to his initial chemotherapy. You can see the scans. You can see the CA19-9 level. And then, as I say, he wanted to go back to work and try to normalize his life as much as possible. So we put him on olaparib. This was compassionate use olaparib. He was on clinical trial and this was way before FDA approved it. And then he continued on it for 10 months. And he tolerated it relatively well. As I said, he continued to work full-time. And his imaging CA19-9 continued to be stable, but then he progressed after 10 months.

When he progressed, I repeated the molecular profiling, but I couldn’t find anything which was different than the original one. His PD-L1 expression was elevated this time, but again, it's hard for me to think that I can relate it definitely to the treatment with olaparib.

DR LOVE: So if you kind of go back through that case, I guess the thing that was complicating the case, that made it more difficult was his elevated bilirubin?

DR PHILIP: Correct. The bilirubin was elevated. I mean if you were to wait until — this case was a very difficult case because you couldn’t do an ERCP. In fact, he had more than one drain, he had two drains. And for you to wait until his bilirubin was going down to the point that you can give him FOLFIRINOX or gemcitabine/nab paclitaxel would have been a long wait. Maybe you’d have never been able to give them.

So, for that reason, he needed to start treatment quickly. And gemcitabine/cisplatin was the way to go in this patient.

DR LOVE: What do you think most people would do?

DR PHILIPspan>: I think what chemo you would give is really important in this patient. Because some people might give FOLFOX frontline, although Eileen O’Reilly’s study clearly shows that gemcitabine/cisplatin is really very good and is easier to give in this patient, I don't think we even put a port when we started the treatment because we could start the treatment without having a port placed. So, it was easy.

DR PHILIP: So, this is a patient who is young. He was otherwise healthy, but now his performance status is 2. And now he comes to me with metastatic pancreatic cancer, proven histologically. And he has a high bilirubin. We can’t bring it down with ERCP. We have to do external draining; in fact, even double drainage. We couldn’t even do one. And he looked miserable. But this patient, to get him through one of the standard two treatments would be impossible because I can’t give him irinotecan, I can’t give him nab paclitaxel. So what do I do?

I don't want him to get treatment while he’s suboptimal. And therefore, I wanted to start treatment quickly. I didn’t have time to put the port on him. But I was ready to start chemotherapy with a high bilirubin. And I myself, my own experience has been that you can give gemcitabine with elevated bilirubin. Gemcitabine is not cleared by the liver. And cisplatin itself, metal doesn’t go through the liver metabolism either. You need a good kidney for it. So I started off with gemcitabine/cisplatin without a port. Started the treatment while he was feeling unwell. And that's where I was. And then after giving him several treatments, in fact, early on he showed improvement and subjectively he felt much better. And he was gaining weight.

So at that point in time, in a patient like this who’s also eager to normalize his life, what do I do? Do I continue with the chemotherapy? He has Stage IV disease. Do I continue with the chemotherapy with gemcitabine/cisplatin? Do I give him gemcitabine alone? Do I give him capecitabine maintenance? Do I switch him to something else? Or do I stop the treatment altogether with something for him more than acceptable, because he’s a young guy, he has Stage IV disease.

So the optimal treatment at that point in time was to give him a PARP inhibitor because I had him tested early on. That's the way I’ve been managing my patients for long time. And with that in mind, I went to the company and asked for compassionate use drug and I got it. Because that was the time when the POLO trial was also, I think, underway or started. And that gave us a very good way of trying to manage this patient, to take him back to full-time employment.

Optimizing the use of PARP inhibition for prostate cancer (PC) — Maha Hussain, MD, FACP, FASCO

DR HUSSAIN: Well thank you Dr Love for the opportunity to discuss how to optimize the use of PARP inhibitors in prostate cancer.

The precision medicine era in prostate cancer somewhat lagged behind other solid tumors, as you know. However, it really was rejuvenated and stimulated by the original work that was published by the Stand Up To Cancer highlighting the genomic landscape of metastatic castration-resistant prostate cancer.

For the audience, this was a prospective project where patients with metastatic castration-resistant disease underwent tumor biopsies, and the biopsies underwent next-generation sequencing. And what was interesting was the data that was published in Cell now actually 5 years ago, was that 90% of metastatic castration-resistant prostate cancer harbored potentially actionable mutations. And what was interesting is there were several new discoveries, but the highlight was actually that more than 20% of metastatic castration-resistant disease actually harbored DNA repair pathway aberrations, the BRCA1, BRCA2, ATM, and others.

Now one thing that I should point out, the whole BRCAness-like genes, there’s a long list of them. However, we published this manuscript, which was a review on the topic, a few years ago to highlight the fact that not all BRCAness genes, or not all genes that are involved in the DNA repair pathway, in fact respond similarly to PARP inhibitors. And this is basically highlighting the fact that many of these genes don’t have the same sensitivity to PARP inhibitors as outlined in here.

Now there are several PARP inhibitors that are in clinical trials, and the list of them here on the right-hand side of the slide. I’m going to focus primarily, in the beginning, on olaparib. So one of the trials that was part of the Stand Up To Cancer research project was this particular trial that was published by Dr Mateo and team from the Royal Marsden where they looked at patients who were given olaparib, and these patients were previously treated with multiple lines of therapy.

And what was interesting is that when their tissues were sequenced it does appear that the patients who were more likely to respond were those who had the DNA repair defect alterations. And in fact, 16 patients with defects in DNA repair genes had responses, in 14 of 16 patients as opposed to 2 of 2 of the 33 patients who were biomarker negative.

And as you see here, it’s reflected by radiographic progression-free survival and so on. So that really provided the rationale to proceed further with a Phase III clinical trial, which I have the privilege of also being involved in. And the first publication came out summer of this year. However, the data was presented last year at ESMO, in September of 2019.

So the PROfound trial was a randomized Phase III clinical trial that enrolled men who had metastatic castration-resistant prostate cancer with disease progression on prior abiraterone or enzalutamide, and they had to have alteration in 1 or more of any of the qualifying genes. And these genes were actually screened for by FoundationOne CDx, or companion diagnostic, as outlined in here, and there is a long list of genes there.

The patients were stratified by previous taxane exposure and the presence or absence of measurable disease. And they were assigned to 2 distinct cohorts. Cohort A, which is the primary cohort, was for patients who had what I would call canonical DNA repair defects, specially BRCA1, BRCA2, or ATM. And then the Cohort B was any of the other alterations in the pathway there. The management was very similar. In both arms patients were randomized 2:1 for olaparib versus the control, and the control was either abi or enza per treating physician choice.

At time of progression the patients could crossover from the control arm to olaparib. The primary endpoint was imaging-based progression-free survival in Cohort A, and all of the imaging progression-free survival was actually based on imaging review by an independent external reviewer. There were several secondary endpoints, including imaging-based progression-free survival in the combination cohort, confirmed radiographic response rate in Cohort A, time to progression, and then overall survival in Cohort A.

DR HUSSAIN: So patient characteristics is outlined in here, and I think it’s important to note that this trial included men of all, obviously, age groups, including some men who are in their 80s and in fact who are in their early 90s. The patients were fairly heavily pretreated, as is highlighted in here. Of course it was a requirement that all patients either have abiraterone or enzalutamide. Some have received both, and close to half of the patients have seen chemotherapy, including some of the patients receiving both docetaxel and cabazitaxel.

The other part that I would like to highlight, again to just reflect how sick this group of patients was, is the fact that a significant percentage of patients have also had visceral disease.

So this is the primary endpoint of imaging-based progression-free survival in the Cohort A, and the second graph in Cohort A and B. And as I mentioned, this data was previously presented at ESMO last year and published this summer in The New England Journal of Medicine, again highlighting the fact that olaparib resulted in not just statistically significant, but rather clinically significant improvement in radiographic progression free survival, as is highlighted in here. And this is in Cohort A, and then in the overall population also, this observation was seen.

Time to pain progression was significantly also delayed. As you know, in prostate cancer pain because of bony metastases is a critical side effect of the disease, so delaying pain progression if very much clinically relevant in there. And this was, by the way, true for both cohorts.

Now this is a forest plot demonstrating prespecified subgroup analysis of imaging-based progression-free survival, and as you see here, overall by clinical characteristics and by the different other features in both cohorts, the trend all of them favor olaparib. One thing I should point out here, when it comes by the gene type, as you see there is some variability. And I will show you some data specifically as it relates to the gene-by-gene responses.

Now at ESMO this year the overall survival data was also presented, and the paper was just published, again in New England Journal of Medicine, in September of this year. Summarizing the overall outcomes are in this slide. The overall survival in Cohort A was clinically and statistically significantly better in favor of olaparib. As you see below here, this is an adjusted analysis, adjusted for crossover, so the magnitude of benefit even is bigger than what you see in Cohort A once you adjust for the crossover.

The Cohort B, there was a trend for benefit. However, this benefit was not statistically significant. And again, it does appear to be slightly more when you adjust for the crossover. But again, not a statistically significant benefit when it comes to Cohort B for overall survival.

Now as I mentioned, we did perform a post hoc analysis gene by gene to evaluate outcomes based on the different genes with regard to overall survival. And this is basically in patients with alterations in a single homologous recombination repair gene. And what was interesting is this, is that as you see here, 7 of 15 genes had alteration frequencies that were too low for descriptive statistics, as in very few patients had it, less than 5 patients.

97% of the patients, in fact, were randomized based on alterations in 8 out of 15 single genes, as you see them listed in here. There is clearly evidence of clinical activity of olaparib in patients with alterations in genes other than BRCA1 or BRCA2, as in RAD54L and ATM and CDK12. However, I should highlight that the gene-by-gene level analysis is a complex and exploratory analysis, and comparisons may be confounded, obviously, by multiple factors.

We also did another post hoc analysis because one of the questions that I get asked, and in fact I myself was not clear about, which is why we were doing this analysis, is does it really make a difference when does one prescribe a PARP inhibitor for a patient. If you have a patient who is front-line castration-resistant disease, should they get it now if they have BRCA2 gene versus wait until they progress, like what we did in the trial and do it after an AR-targeted drug and so on? And again, I want to qualify that this analysis, again, is a post hoc exploratory analysis, and it’s not really powered for statistical significance. However, what I think is interesting is this: as we look at BRCA1 or BRCA2 patients, and we looked at prior taxane exposure versus not, and as you see, for those who have not had prior taxane exposure, compared to the control arm, there was a much bigger difference in overall survival compared to those who have had prior taxane exposure.

CDK12, interestingly, appears to be better if the patients had not seen chemotherapy. When they have seen chemotherapy the magnitude of the benefit is not there. And to my surprise, actually, ATM was a totally separate story in that front-line patients with no prior taxane, if anything the control arm, the control drug, did better, that’s abi or enza, as opposed to olaparib. However, if those had seen prior taxane it’s the reverse, olaparib did better. Again, I would caution again that this is an exploratory analysis, but it may actually give you some insight to help guide your decision with the patients that you’re seeing.

Now overall the treatment was well tolerated, and I should point out that the adverse events, there were no real surprises with regard to adverse events. It was what is expected with olaparib. The most common adverse event, specifically high-grade adverse event, was anemia. I do point out, and this is not to say that olaparib doesn’t cause anemia, because it is higher than the control arm, but this is a relatively heavily pretreated patient population. Okay.

Now I want to switch to the other PARP inhibitor, rucaparib. And rucaparib was evaluated in the second-line setting, castration-resistant disease, in a Phase II clinical trial that was published by Dr Abida and colleagues in 2 separate publications, which I will show you. I’m going to begin first with the BRCA-specific cohorts, and as you see here rucaparib led to a significant confirmed response rate, PSA declines, and very promising median radiographic progression-free survival of 9 months.

And this is the data that was seen with the other genes, specifically the ATM, CDK12, CHEK2, and other genes that they evaluated. And while there are responses, and many of them are durable, overall the rates were not as high as what was seen with the BRCA2 genes. All of it is important because it was taken into consideration by the FDA. And in fact, in May of this year the FDA issued 2 approvals for metastatic castration-resistant prostate cancer. In fact, I don’t recall we’ve ever had that many approvals compared to previously. The last few years prostate cancer has really hit the radar screen multiple times at different levels of approvals.

So the rucaparib was approved for patients with deleterious BRCA1 or BRCA2, basically BRCA1/2, so the BRCA mutations, germline and/or somatic, and in patients who have seen prior AR-targeted therapy. Olaparib was approved in deleterious or suspected deleterious germline or somatic homologous recombination repair gene mutations that had progressed on prior treatment with enzalutamide or abiraterone. And the patients are to be selected based on an FDA-approved companion diagnostic for olaparib. And in fact the FDA issued an approval for a circulating cell-free DNA companion diagnostic for use of olaparib.

The FDA actually took the PPP2R2A gene mutation from the HRR panel approval. This is really stemming from different reasons, specifically the fact that biologically it has weak evidence as a DNA repair defect contributor. And then the clinical data observed from PROfound and so on. So that’s not included in that, but all of the other genes that were included in this study, in fact, are part of the approved panel.

Now there are many lessons that I think we have learned from PROfound, and as it relates specifically to the issue of tissue collection. As you know, in prostate cancer we’re dealing with patients who have mostly bony metastases, and/or if they have primary tissue they could have had their surgery 10, 15 years ago, or their biopsies a long time ago, so the tissue quality and quantity becomes very important.

Summarized in here are the lessons that we have learned from the tissue evaluation. So the success rate overall was 69%, and I would like to remind the audience that this was a large international trial, over 4,000 men were screened internationally. And so the data that we’ve accumulated in everything highlights the fact that it’s quite applicable, irrespective of where the patients are. What we found is that newly collected tissue samples are better than archival tissue, which is really not a surprise because the quality is still there. Younger archival tissue, that is collected 5 years or less, are better than older archival tissue.

To my actual surprise, actually, while yes metastatic disease had better success rate than the primary samples, the primary samples were actually over 50% success rate in terms of properly analyzing them. And so clearly the metastatic disease, having a higher rate of success, is driven by higher cellularity, tumor content, and also that it is mostly fresh as opposed to historic tissue.

One thing that I want to point out to you to discuss carefully with your radiologist and pathologist is the method of collection and processing. So what’s important to note is that based on our experience with PROfound, what is favored is multiple samples from core needle biopsies and embedding them into 1 formalin-fixed paraffin-embedded block. The other thing to remember is to make sure that you talk to your pathologist so that they are aware that they are to avoid acid decalcification if in fact a bone biopsy is what’s been done. And when it comes to bone, as long as it’s safe, as many cores as possible, partly because the tumor content when you do a bone biopsy is not really great. And so you need many more cores.

Now I want to come and highlight the fact that we are living in an era of PARP inhibitors in prostate cancer. And there are numerous clinical trials that are being conducted, in a variety of different settings and with a variety of agents, as highlighted in here, including Phase III clinical trials, as listed in here.

The other thing that I want to feature is some selected trials only to show you the fact that the field is going into exploration in certain settings, where we have not done that before. For example, looking at rucaparib in patients with metastatic hormone-sensitive prostate cancer, so without hormonal therapy, just a PARP inhibitor if in fact the patients have the germline DNA repair mutations. Other trials looking at olaparib without hormone therapy in those who have high-risk PSA relapse disease setting, combination of cediranib, olaparib, and radium, and then carboplatin with an ATR inhibitor. So all of these trials are really exploring different venues of evaluating the PARP inhibitors.

So I want to end by highlighting the fact that metastatic prostate cancer is clearly complex. In the old days when I was in training it was assumed that the cancer is “a stupid tumor”, that it basically just responds to treatment and does nothing. And actually we’ve learned a lot about this cancer over the years, that it’s a very smart cancer with significant inter- and intrapatient heterogeneity. And not only that, it has an incredible adaptive capacity, and so therapy development has to focus on the totality of the disease biology. And clearly personalizing care is very important. PARP inhibitors have emerged as a promising therapeutic target in this disease.

Now there are many issues to be considered. The question, as I showed you, in some of the trials that I mentioned here, when is the best timing in relationship to other life-prolonging therapies? What is the impact on response to these agents post PARP inhibitor? So if you have a patient who has, let’s say BRCA2 mutation, and they have just become castration resistant, are they better off getting abi/enza or are they better off getting a PARP inhibitor?

And what happens if you give the patient a PARP inhibitor, will they respond then in second line to an abi/enza-type situation or chemotherapy? What about tumor heterogeneity and resistance mechanisms? We need to learn much more about why is it that even when you have patients have BRCA or BRCA-like genes not everybody is responding. So clearly there are tumor heterogeneity and resistance mechanisms either de novo or coming in an adaptation to treatment. And so this piece of information is really very critical to increase the durability of the benefit and increase the odds of benefit for patients.

What about liquid biomarkers? As I mentioned, there are FDA-approved companion diagnostics right now, and certainly that would make life simple. The question that I have, and I’m not clear on myself, is if the blood was negative does that mean the tumor is negative. And I had a situation with one of my patients where we were doing a study, and it turned out that the liquid biomarker was negative, but the actual tissue happened to be positive. So I would caution you about believing in 1 method of testing. To make sure that the patient is evaluated properly I would recommend that if 1 assay is negative, then flip to the other test if possible.

And of course the issue of the impact of AR targeting versus PARP inhibition versus combination treatment.

What about the role in metastatic hormone-sensitive disease in combination with hormone treatment? What about situations where patients have high-risk disease diagnosed with say a clinical T3/T4 and possibly pelvic lymph node involvement, but no metastatic disease, and they happen to have the BRCA or BRCA-like genes? Should we be beginning to look at these types of patients for trials? And of course one of the issues, clearly, is the long-term effects of PARP inhibitors in terms of safety and duration of therapy.

Case: A man in his late 60s with metastatic castration-resistant PC (mCRPC) and a germline BRCA2 mutation receives olaparib on the BRCAAway trial

DR HUSSAIN: So with that, what I’d like to do is move into a couple of examples. So the first patient is a 69-year-old gentleman who in 2008 was diagnosed with intermediate-risk prostate cancer. He underwent radical prostatectomy in December of 2008, and the pathology revealed pathologic T3N1, Gleason score 8 prostate cancer. So this is by any criteria high-risk disease.

He did okay, and then a few months later he developed a rising PSA, and so he had imaging. And the imaging showed bony metastases in the bone scan. And so he started androgen deprivation with leuprolide, and went on for many, many years of treatment, almost 6 years of treatment, when in fact he developed a rising PSA. Testosterone was checked, as it should be, and it was in the castration range, and therefore he technically entered the phase of castration-resistant prostate cancer.

He ended up getting 3 cycles of sipuleucel-T and palliative radiation therapy to the focal area in the sacrum as part of a clinical trial. And then he was stable for a while, but was developing a gradually rising PSA. However, his scans remained reasonably stable.

And I tell my patients this when I speak about the topic, a PSA alone is not necessarily an indication to immediately change treatment. I think it’s critical to look at the trend of the PSA, but definitely if the patient has symptoms and/or if there are significant imaging changes or if the PSA is significantly changing, then that would be a reason to treat. But someone’s PSA going very slowly up over a period of time, and stable by imaging, I would say it is perfectly reasonable to continue watching them.

So in 2017, continued to have a rising PSA, he underwent germline testing, and he was found to have pathogenic mutation in BRCA2. At that time he underwent imaging, which indicated progression of disease, both in bone and lymph nodes. And he enrolled in one of the trials that I mentioned to you earlier, the trial that is investigator-initiated clinical trial that I’m the PI of, and it’s being handled at multiple centers. And it’s randomizing patients who have the canonical DNA repair defects, BRCA1/2 or ATM, to either abiraterone versus olaparib versus combination. And I’m happy to explain why we decided to do this.

And he was randomized to olaparib-only arm, and he responded by PSA and also by continued stable imaging. The gentleman went on for 22 courses of treatment and then developed progression and came off.

Case: A man in his early 70s with mCRPC and a BRCA2 mutation achieves an excellent response to veliparib with abiraterone/prednisone on a clinical trial

DR HUSSAIN: The other gentleman that I would like to actually share with you is a really remarkable case. So this gentleman is what I would call an exceptional responder. This is a 72-year-old white male who in 2009 had a Gleason 9 prostate cancer. He underwent attempt at surgery, but it was aborted because he had pelvic lymph node involvement. And so he ended up getting radiation and androgen deprivation therapy per standard of care.

In April 2011 he developed castration-resistant prostate cancer, and in 2013 he came in to consult with me, in my days when I used to be at University of Michigan. And in fact, he had baseline disease in the bone, in the liver, in the lymph node, and he had pain. From a practice perspective, every time I see liver, even though it’s not rare in prostate cancer, but it’s not the most common first area of metastasis, I generally tend to biopsy these patients just to make sure that we’re not really dealing with a small-cell type or neuroendocrine-type features.

And so biopsy actually showed adenocarcinoma, PSA positive, and so in 2013 he enrolled in a clinical trial that we had at the time that was taking all comers with treatment with veliparib and abiraterone versus abiraterone alone. Veliparib is another PARP inhibitor, and so the trial was evaluating the concept of cotargeting AR and DNA repair because there was preclinical data clearly supporting the combination being more powerful than the single-agent AR-targeted drug.

And as you see here, this is just an example of his liver lesion, 1 of the liver lesions. He had multiple liver lesions. This is a 6.7 cm liver lesion in the long axis. And this was in 2013, and this is March 2016. Look at what happened to his PSA. It went down from 127 to less than 0.1, and he continued to be in response past that date. And I have to say, I call him a remarkable, exceptional responder because I don’t recall ever in any setting of castration-resistant disease where we’ve seen this remarkable response for many years. And he went on to respond also beyond 2016.

When we did the trial, actually, it was part of the Stand Up To Cancer. As it turns out, he had a BRCA2 homozygous deletion, and so clearly consistent with what I described to you is that the BRCA2 gene seems to be the flag for response to PARP inhibitors.

Role of liquid biopsy in identifying mCRPC with BRCA1/2 mutations; ongoing trials investigating PARP inhibitors in PC

DR LOVE: So yeah, just to follow up, can you kind of go through this, this new approval, but also what the different types of liquid biopsies there are and what the accuracy is?

DR HUSSAIN: Sure. I can comment. I think, as I mentioned in my presentation, is that liquid biopsies as companion diagnostic have been approved by the FDA. And in fact, there was 1 approval that came in. The issue, I think, so this one is a cell-free DNA, which I like much better, personally, than CTCs. With the CTCs, and I don’t know that there’s any assay that’s approved based on CTCs, is that with the CTCs you have to be able to pick a circulating cell to do this, whereas this one is really more like just the cell-free DNA that’s in the blood.

And so I do think that it clearly makes it much easier and simpler for us to screen patients. The issue comes up is whether a negative assay means the patient should not be given a PARP inhibitor. And this is what I was trying to make a comment on in my presentation, to say this is certainly a simple thing. Consent the patient, draw the blood, send it, but at the same time I would say if it’s negative, then I generally would also say let’s go to the actual source and make sure that the tissue is tested also.

And in that case, as long as it’s feasible, my preference is to test fresh tissue, fresh acquired tissue. If it’s difficult and not easy or whatever other confounding difficulty, then definitely archival tissue would be an option.

DR LOVE: So in your talk you mentioned CDK12, and I thought I had heard that that actually predicts for benefit from checkpoint inhibitors. Is that the case?

DR HUSSAIN: There is emerging data in that regard, but I have to say the data is, relatively speaking, limited. But certainly the last 2 years we’ve seen some data to suggest potential benefit.

DR LOVE: Could you also talk a little bit about the BRCAAway study? You mentioned it in your talk. Are patients who’ve never had hormonal therapy eligible for that?

DR HUSSAIN: So the BRCAAway trial is essentially a Phase II trial, but randomized. The randomization is for patients who have BRCA1, BRCA2, or ATM who are front-line castration-resistant prostate cancer. If they have disease progression by PSA and/or imaging, they have to have metastatic disease for sure, they’re evaluated for the NGS, and if they have BRCA1/2 or ATM they’re randomly assigned to either abiraterone/prednisone, olaparib, or combination.

And just to perhaps, if we have time, to give a brief background here. And so as I showed you in the slides about the gentleman that had the remarkable response in his liver, that was a clinical trial that I had the opportunity to design and lead. And we took all comers with metastatic front-line castration-resistant disease, and basically they were randomized to either abiraterone/prednisone or abi plus veliparib. And the rationale for the trial was driven by the fact that there was preclinical data that clearly indicated that cotargeting AR and DNA repair pathway actually gives better results. It creates like almost I would call it synthetic lethality type thing, or artificial synthetic lethality.

And in that particular trial what we did is this, is patients were required to have a tumor biopsy. And over 70 of the patients had adequate tissue to undergo NGS testing as part of the Stand Up To Cancer data set that I showed you. And to our surprise, actually, what we saw is that the patients who had DNA repair defects seem to do better no matter which arm they were on, whether they got abi alone or abi plus veliparib.

And so that kind of raised a question in my mind, well if you respond just as well with 1 treatment, then why put you through 2, and which of the 2 treatments are going to be better, which is why then we designed this trial. When we designed the BRCAAway trial, at that time, because the olaparib was not FDA approved, we had to go to the FDA and get an IND for it. And at that time the FDA required that we create the randomization only for the canonical cohort, which is BRCA1, BRCA2, ATM, partly because there was very little data on the DNA repair pathway alterations.

So we created just an exploratory cohort for anybody who was not BRCA or ATM, and so that’s really the BRCAAway trial. We’re about two thirds accrued right now. I’m hoping that we are going to be finishing by end of year/early next year. And hopefully we will have some data.

DR LOVE: Just out of curiosity, how come abi as opposed to abi or enza?

DR HUSSAIN: Well, that’s a great question, Neil. Part of it is the following. When we designed this study the key drug that was, not this particular trial, but the parent trial, abi was the drug, and so we had good experience with it in combination with a PARP inhibitor. So there is the issue of drug and drug interaction, and enzalutamide is a very powerful drug, but putting the 2 together can be problematic. And so it was a matter of the feasibility in terms of how we designed the trial. And so we chose to go with abiraterone.

DR LOVE: So just to update that, though, I believe there are trials now looking at the combination of PARP and antiandrogens. Is that correct?

DR HUSSAIN: Yes. Oh, there’s a long list of trials, which I tried to show examples. There is different, as I said, different trials in different settings. One of the trials that I should point out is a product of a Phase II trial that was conducted in the UK that looked at cotargeting AR and DNA repair, like what we did in our parent trial, but using abi plus olaparib versus abi alone. And this was in all comers, and what they did is it wasn’t an integrated biomarker analysis, but they looked at different ways, either liquid biomarker tissue and so on, but it wasn’t like required for stratification, okay? So it was all comers, but then post hoc they looked at the different tissues or liquid samples that they had.

And what was interesting, it suggested that there was potential benefit irrespective of whether the patients had DNA repair defect or not, which then led to one of the Phase III trials that was conducted. And I believe the name of it is PROpel.

Where it’s basically taking all comers without genomically defining their tumor genomics and randomizing them to combo versus abi to see actually if the combo is better. So that’s a Phase III clinical trial.

DR LOVE: What do we know about the impact of PARP inhibitors in patients who are endocrine naïve? Are there trials out there looking at that? I think I’ve heard of some.

DR HUSSAIN: Yeah, and I showed you in my slides. They were in there. There were a few trials where patients who have PSA relapse or front-line metastatic hormone-sensitive disease, where they’re looking at the role of PARP inhibitors.

I have to say, and this is my gut feeling, I could be wrong, but obviously data will prove me wrong or right, is that AR continues to be a major player in this disease, and especially in the hormone-naïve setting, so front-line metastatic hormone-sensitive disease or early stage disease. Such that I think to say that we’re going to replace AR targeting with something else is a bit, for me, hard to believe. And the reason I say this is that there is nothing to say that unless this is germline, and not everybody has germline mutations, so in non-germline settings there is no way to say that every single cell, in fact, is going to be susceptible.

And these drugs, while they are reasonably safe to give, they’re not exactly without side effects. And some have long-term issues, and this is what I was mentioning in terms of myelo issues, marrow issues, and things like that. So I do think we have to be mindful of that.

What I think, personally, is exciting to do is this is not about replacing AR therapy. My view is that the target is cure, and so let’s come up with cure strategies as opposed to shuffling around drugs, you know what I’m saying?

DR LOVE: Sure, no, and I can see where clinically it makes a lot of sense. But I was just kind of almost more thinking theoretically, after you get an endocrine intervention, does it in any affect your sensitivity to PARP, either positively or negatively? I don’t know. I guess it probably wouldn’t, I would imagine.

DR HUSSAIN: That’s right. I don’t think it does because A) the rate of having the qualifying mutations is much higher after. So I do think, again, outside a germline situation this is almost one of those progression mechanisms, as we see is that these patients with castration-resistant disease have a higher rate. The question comes up, which I think is going to be interesting, is if the combination proves to be better from the international trial that was conducted in castration-resistant disease, and let’s say our trial shows combination is better, of course ours is a Phase II trial.

I do think it’s going to be very critical to then begin to implement that, not as sequential therapy, but combination treatment with the understanding that for long-term responders there ought to be some strategy to look at finite duration of therapy to minimize long-term side effects.

DR LOVE: Getting back to your patient 2, the extraordinary responder, any clues right now to predict who these patients might be? You referred to the fact this patient had a homologous BRCA.

DR HUSSAIN: Yeah.

DR LOVE: Is that a predictor? Is that a predictor?

DR HUSSAIN: Well, it’s really hard to say because I would say potentially, but as you know from the data that I showed you from PROfound it’s not conclusive 100%. And so I do think there is a lot to be learned from this data, and one of the important parts, I think, which is what the beauty, in some way, of the liquid biomarker, is that it’s easier to actually collect the samples and look at what has developed since the patient has been on treatment. What are the potential mechanisms of progression? And I do think we need to be informed also by preclinical work so that we can begin to target our research to look at mechanisms of progression, and if they existed at baseline can we do something about them.

DR LOVE: So another question, this comes up with the other cancers where PARPs are being looked at, of course ovarian, but even breast and pancreas, which is the issues of platinums.

DR HUSSAIN: Yes.

DR LOVE: And I don’t know if we really know at this point. Do we know if there’s a correlation between these 20% of patients with DNA repair issues and sensitivity to platinum? At one point I can remember there was satraplatin being looked at. Of course obviously it’s a key in ovary. Is there any renewed interest in platinums because of this?

DR HUSSAIN: Well there is actually renewed interest, and I would say the interest has been going on for at least 5-plus years, since, again, the emergence of the PARP story. There are no what I would call definitive large-scale clinical trials. However, there are, within the VA system, there’s a planned clinical trial. There is other Phase II clinical trials that are being conducted at some centers. And there was a recent report that came in JAMA Oncology about case series from different investigators who have used platinum there.

I will tell you, in my experience, limited experience, that in the setting of patients who have seen a platinum, when they progressed, and I switched them to a PARP inhibitor it doesn’t seem to work as well and vice versa. Having said that, this is like an N of 1 or N of 2. This is not a large-scale type set of data.

I do think there is the potential advantage in that theoretically you could give a finite duration of treatment then stop. I mean no one’s going to get platinum forever. That’s a fact. The other part would be is that for some of these patients, since they are going to be getting chemo in their forecast anyway, cabazi or docetaxel and these drugs are life prolonging, one could envision a process whereby they get it up front. So imagine somebody has metastatic hormone-sensitive disease, and they are high volume, and the physician decided they’re going to offer them therapy intensification with the addition of let’s say docetaxel. Well that would be an opportunity if they were BRCA positive or CDK12 or whatever, that they could consider that. But I do think we need clinical trials, definitely.

And I honestly can’t say that I know for a fact a platinum is better or equal to a PARP inhibitor at this moment.

DR LOVE: I think what I see oncologists thinking about is they use the idea of platinum and then PARP maintenance in ovary, and of course incredible results when they’re using it up front.

DR HUSSAIN: Sure.

DR LOVE: And it kind of makes a lot of sense. I think that’s kind of what I hear people thinking about, that strategy, platinum and then PARP maintenance. Does that make sense to you?

DR HUSSAIN: The answer is no, without a clinical trial. And the reason I say that is we know different diseases are different. And the same drug gets tested in different diseases, for the same pathway, and one leads to survival advantage and the other one says it’s not working. So I would say we really need to prove it.

DR LOVE: Absolutely.

DR HUSSAIN: And if it works that way, then that’s fantastic. But remember, platinum in ovarian is a long story. It goes back probably before the PARP story. And so that’s a different situation than prostate cancer.