Johann de Bono: Our next session is a debate, and I'm delighted to invite Noel, Noel Clarke, who needs no introduction. We're going to debate PARP inhibition in mCRPC, lumpers or splitters, what do you want to do? So Noel is going to vote for lumpers.

Noel Clarke: Thank you, Johann. Well, this is the question. PARP inhibitors in metastatic CRPC, for all or for some? Well, you don't need to be a genius to work out that Johann is Scottish and I'm English. And for 2000 years, we've had some constructive across-conversations

Johann de Bono: Actually, I'm Maltese no, I'm not Scottish.

Noel Clarke: Yep. Well, anyway, here are my disclosures.

So debate. What is a debate? So if you go back to the Latin route batuere, to strike or to beat, well, I don't think we're going to do that, are we? If you then look at the Oxford English dictionary definition, "to contest in words, argue, discuss, thoroughly thrash out." Well, I don't think we're going to do that either. But what we might do, is engage in, think over in, reflection, and arrive at some kind of conclusion.

Now, although Johann denies his Scottish inheritance, we do have some important differences in England and Scotland in the jurisprudence system. What that is, if you're in the courts in England, you are guilty or not guilty. In Scotland, you are guilty, not guilty, or case not proven.

There's another founding principle of English common law, which is known as [inaudible]. And what that means, is let the facts speak for themselves. So let's have a fact.

This is PROFOUND. You all know this study. It clearly is positive in late stage castrate-resistant metastatic disease for patients who've got a HRR mutation. No doubt about that. So when we look at this cartoon, you can see olaparib monotherapy PARP inhibition blocks DNA repair pathways in cells, harboring HRR defects. No question.

Then that comes to this hypothesis, which is generated by work in the lab and elsewhere, is this synergy by putting together olaparib and abiraterone, and will that render previously refractory patients sensitive to treatment? On this form, the basis of what we know as study eight, which is a large Phase II study, which set out to test this hypothesis in the preliminary phase to setting, published in the Lancet Oncology in 2018.

The patients were unselected. They were heavily pretreated. All of them had docetaxel, pretty much, they were all castrate-resistant. And rather to our surprise, pleasant surprise, the combination of drugs was effective, whether or not the patient was HRR mutated or not.

And that led us on to design PROPEL, a large scale randomized international Phase III study. And we brought the patients forward in the disease cycle, so that they were first-line metastatic castrate-resistant, and strictly, the patients had not had any kind of novel hormone agent beforehand. Randomized to receive olaparib standard dose with abiraterone, against abiraterone and placebo. And without going through all the results, these were presented by Fred Saad, on behalf of the PROPEL investigators, and they're impressed with New England Journal of Medicine evidence currently, so you'll be able to read the detail, with a clear benefit for the combination of olaparib plus abiraterone, and a significant hazard ratio, significant improvement in survival.

Of course, at the same time, MAGNITUDE, an equally large scale study using niraparib, the PARP inhibitor, with abiraterone, was planned, presented at the same meeting, and the design was somewhat different. This had a futility analysis, or a futility component, for non-HRR mutated patients. 600 patients planned for, and the rest, HRR mutated.

The futility analysis, and I'll come back to this, was stopped after about 200 patients, at a quite early stage of the trial. So the HRR non-mutated were not really tested in the larger scale, but the DDR mutated patients did have a very definite benefit in radiological progression-free survival, assessed by central view and by investigator.

So why the difference in these trials? And this is something which is in active debate, and it's why we're here today. Different patient population, different trial design, different pharmacology, pharmacokinetics, the effect of prior abiraterone treatment. In MAGNITUDE, patients were allowed to have abiraterone for up to four months.

So let's address this question. Why the difference? A different patient population. Well, if you look at PROFOUND, clearly a different patient population, and a different treatment schedule. So monotherapy advanced DDR mutated. For PROPEL or MAGNITUDE, well, they are different. PROPEL, no DDR selection, no novel hormones, standard PARP dose, a lower percentage of BRCA mutated patients, and a different drug. With MAGNITUDE, DDR mutated only, a high proportion of BRCA, abiraterone for up to four months in a proportion, all comers excluded, and a different drug, niraparib, with the early stopping of non DDR mutants.

So why the difference when we look at the effect of prior abiraterone treatment and differing pharmacology and pharmacokinetics? Well, we know that, as soon as you start abiraterone, or indeed, any hormone manipulation, then there are changes in the androgen receptor in pioneer molecules, the open opening up of the molecular structure. And that may have had an effect. We don't know, but it's a possibility.

What about the differing pharmacology or pharmacokinetics? Well, here, you've got a table with the different PARP inhibitors currently, commercially either under test or available, olaparib, rucaparib, niraparib, talazoparib, and veliparib. And without looking in too much detail at the table, you can see that they've got differences, in terms of their active concentrations, their PARP trapping ability, the effect on different PARP areas.

This is a poster from the AACR 2018, and I've just encapsulated some of the results from this publication, a head-to-head comparison of the properties of five clinical PARP inhibitors. And you can see if you look at three, the cytotoxicity, that in red, you've got the non-HRD mutated in blue, the HRD mutated, that there's a different profile for each PARP inhibitor, whether it's olaparib, rucaparib, or veliparib.

If you look at four, or the novel chemical proteomic profiling of the PARP family, and there's more than part one and part two, there are things we simply don't understand in this area. That if you go across the table adjacent, that they are different. If we look at hematological toxicity, again, under six. And focus down at the left hand side, you can see a histogram which tells you that stated doses, which are usually used in monotherapy, then certain PARP inhibitors are hematologically toxic. So naraparib is more toxic than olaparib. So its dose was lowered.

So these are things which might contribute to changes in results. One important thing to consider is the differential PARP trapping and the stability of binding. Now this paper, which was in Science in 2020, explains this for the uninitiated. So if you look at PARP trapping, then the PARP molecule attaches to an area of DNA damage. And then, it detaches after a period of time.

And if you follow this round from the cartoon at the top right, down below, you'll see that there are three effective types of PARP chopping and PARP release. Type one, type two, which is olaparib and talazoparib, which is a very potent PARP binder. And then type three, naraparib, rucaparib, and veliparib. And they've got different chemistry, which means, that in type three, type two, and type one, they all bind with the same intensity. But type three releases much earlier than type two and type one. And that's shown in the green box on the left, as you see it, when you look at the retention of PARP on the DNA molecule and its release. So clearly, these PARP inhibitors have got different mechanisms and different molecular propensities, which may give one an advantage over the other.

Let me just come to this finally, which is the futility analysis in the [inaudible] trial MAGNITUDE. And my own view of this, is that it was a mistake to stop this at when it was done. If you look at the red, breakdown of competent events, 83 PSA events, 65 rPFS events, this was at nine months; far too early, actually, to see a signal in any of it.

So the question remains, PARP and NHA for all, or for some? And I put it to you, have we reached a verdict? I don't think we have yet reached a verdict. I think we're still in the Scottish guilty, case not proven setting. But if we see that overall survival come through in the PROPEL trial, then I think we have got some compelling evidence, that the combination of abiraterone and olaparib may be very effective, not withstanding DDR mutation. Thank you.