The Uromigos Episode 136: STAMPEDE Abiraterone for High-Risk Non-Metastatic Prostate Cancer

Dr. Gert Attard describes this practice-changing plenary data and the STAMPEDE study presented at ESMO 2021.

Episode Transcript

Tom:
Welcome to another ESMO 2021 podcast. It’s a prostate cancer podcast with Gert Attard, our friend and colleague from the UK. Gert, do you want to introduce yourself, talk about any major conflicts and then maybe give a quick overview of what you did?

Gert Attard:
Yeah, so Gert Attard. I’m a medical oncologist in the Cancer Institute in University College London, UK. In terms of conflict, I’m going to discuss data related to abiraterone, which is a drug developed at the Institute of Cancer Research, where I was employed up until 4 years ago.

I’m on the list for rewards to discoveries for abiraterone. Abiraterone’s license to Janssen, for whom I’ve consulted and received travel support and speaker’s fees from. I’m also presenting data related to enzalutamide, which is licensed by Pfizer and Astellas in the US and Europe, respectively. Again, I’ve consulted and received speaker’s fees and travel support, and I’ve also consulted several other companies that have drug development in prostate cancers.

Tom:
That’s cool Gert, let’s go to data.

Gert Attard:
Okay. Let’s move on. So, data. Do you want a background, Tom?

Tom:
Yeah, go a little bit.

Gert Attard:
Little bit of background. So, the majority of men who die from prostate cancer in Europe and North America were non-metastatic at diagnosis. This has increased as a proportion over the past 10 to 15 years. In this population, we know that for men with high-risk features, if we treated ADT for 3 years and local radiotherapy, we have improved outcomes compared to either alone.

However, of course, the post-treatment failure rates remain high. So, the question we sought to address was whether 2 years of abiraterone plus or minus enzalutamide in combination with standard of care improves metastasis free survival and overall survival and then other efficacy outcomes. We present new data from 2 comparisons conducted in the STAMPEDE platform protocol.

One comparison where men were randomized between ADT, ADT plus abi and the second independent comparison where men were randomized between ADT and ADT plus abi plus enzalutamide. There are no overlapping controls, same protocol, same eligibility criteria, all received abiraterone for the same regimen and period of time, as I said 2 years …

Tom:
And Gert hit us with the patient population once more apologies.

Gert Attard:
1974.

Tom:
And these patients had radiotherapy and essentially M0 disease and that was because STAMPEDE included quite a broad subgroup of patients. Is that right?

Gert Attard:
Correct. So, when STAMPEDE started accrual 15 years ago, any man with advanced prostate cancer was recruited. That included men with metastasis on conventional imaging and a non-metastatic man on conventional imaging and there’s no metastatic or high risk. Either node positive or node negative. Two of any of stage T3 or T4, PSA 40 or greater, or Gleason sum 8 to 10. Yes. So that …

Tom:
This is a classic radiotherapy type population. So, essentially, you’ve got radiotherapy, was 2 randomizations, but you got radiotherapy with LHRH, radiotherapy LHRH abi, radiotherapy LHRH abi and Enza together. That’s almost a sort of a spectacular combination. What was the rationale for combining those two together?

Gert Attard:
So, they’re not three-way randomizations. We don’t have abi versus abi plus Enza, they’re 2 separate comparisons.

Tom:
Yeah.

Gert Attard:
But we do perform a pre-specified meta-analysis or subgroup analysis in our meta-analysis of the 2 comparisons to look for a difference in treatment effect. The rationale for the combination is that although they both target the androgen receptor signaling axis, they hit 2 different parts. Of course, abi blocks androgen biosynthesis, enzalutamide binds to the androgen receptor and is a potent AR antagonist.

Resistance to abi can occur due to residual androgens and increased androgens could overcome enzalutamide blockade of the androgen receptor or there’s a number of studies, I guess, 10 years ago that generated data to this effect.

Two large phase 3 trials were conducted in metastatic castration resistant prostate cancer. One was a US cooperative group study presented by Mike Morris about 3 years ago. The other was an industry sponsored study sponsored by Janssen ACIS. Mike Morris’s presentation was of a trial Enza plus abi versus Enza and the ACIS trial is apa plus abi versus abi. Now both trials, I guess let’s not go into the detail on that, because we’ll get detoured, but neither showed a discernible benefit for overall survival.

Tom:
Nice. Gert, so tell me it’s a plenary session. I’m guessing the results are good.

Brian:
Good guess Tom.

Gert Attard:
Good guess. They spectacular really. Really notably much greater than we expected. In our plan, we targeted a hazard ratio of 0.75 that required 300 events on the control arm for a power of 90%. We preset the one-sided type 1 error to 1.25% because of prior reporting, bottom line metastasis-free survival hazard ratio of 0.53. Very significant P value 10 to the minus 11, the 95% confidence interval is 0.44 to 0.64.

Tom:
Tell me about metastasis-free survival and why you picked that endpoint? It’s not a standard endpoint or hasn’t been historically, but there’s been some work with ICECaP, which is validated to that. Is that a fair comment?

Gert Attard:
Exactly. Really should be a standard endpoint now for this population so beautiful work from ICECaP and our friend, or led by our friend, Chris Sweeney, show that MFS for this non- metastatic population is a valid surrogate. You’ll see the data. Our data really shows the same relationship between MFS and OS as predicted by-

Tom:
You haven’t given us the OS ratio yet. What is the OS?

Gert Attard:
OS hazard ratio 0.6.

Tom:
That’s not bad.

Gert Attard:
That is impressive.

Tom:
Yeah. This is adjuvant prostate cancer, essentially after radiotherapy, you’ve got 40% reduction in the risk of death in a thousand-patient trial.

Gert Attard:
2000.

Tom:
Wow.

Gert Attard:
Yeah, 2000 in all. The 6 year – if you are interested in absolute improvements – the six-year improvement in survival is from 77% to 86%. So exactly as predicted by ICECaP after 6, 7 years follow up, we’d have about 25% death from any cause.

Actually, we should stress that Tom and Brian, MFS uses death from any cause and that’s a really important concept in this group. This metastasis confirmed on imaging or death from any cause …

Tom:
Every time a friend of mine succeeds, Gert, a part of me dies. This is …

Gert Attard:
Dies?

Tom:
It’s actually, it’s an Oscar Wilde quote. I’ll tell you it’s fantastic. I mean, this is really spectacular, isn’t it?

Gert Attard:
And it’s a team effort, you referred to ICECaP. They really, I guess in retrospect, MFS is irrelevant here because of this big OS benefit.

Tom:
No, but it does validate that as an endpoint for the future, which is good. Isn’t it?

Gert Attard:
Correct, which makes a big difference to future studies.

Tom:
I agree.

Brian:
And so, Greg, did you say that what’s the role of giving both drugs then? abi and Enza.

Gert Attard:
Yeah. So, this is a footnote ultimately because we do not see a discernible difference in treatment effect between the 2 comparisons and in a way, we have 2 independent groups that validate each other. Each is a thousand patients. So, 2 effectively large independent studies, but no discernible difference.

The interaction P values are not significant, but we do see increased toxicity with adding enzalutamide to abiraterone. So overall, we don’t see a justification for adding Enza to abi in this population. So, I was just going to say, Brian, as you and your listeners will know there are ongoing trials of Enza alone and other AR antagonists. So, we’ll see what that data shows, but we cannot really talk about AR antagonists in this context of these data.

Brian:
And you chose 2 years of abiraterone, is that right?

Gert Attard:
Correct. And that’s …

Brian:
I was going to ask, what was that based on and did most patients make it through 2 years?

Gert Attard:
So that was a pragmatic decision. 10 years ago, midway between 18 months, 3 years that were evaluated or 6 months, 3 years that were evaluated 15 years ago for ADT alone. Shorter maybe as effective, longer maybe more effective.

The median time to permanently stopping treatment is not far from 24 months, but the interquartile ranges are broad. And of course, the discontinuation rate is greater when abi and Enza are given together. In terms of how many the proportion of patients who report stopping prior to planned completion. It’s about half with abi alone and about 40% of the combination.

Brian:
Why they stopping Gert? What’s the tox?

Gert Attard:
So tox is fatigue, hypertension. The other grade 3 toxicities are liver transaminitis, but that’s not that common. And erectile dysfunction is the other grade 3 toxicity.

Brian:
What percent getting grade 3 or 4 tox?

Gert Attard:
So, 33% on abi. And that’s compared to 30% on ADT alone so not far off.

Tom:
But 50% are stopping? Five? Zero?

Gert Attard:
About 50% are stopping. Yeah.

Tom:
So that’s quite a lot. Isn’t it?

Gert Attard:
It is. We’re stopping before planned completion of therapy.

Brian:
but not long before the median’s almost 2 years then, right?

Gert Attard:
Correct. I think you get, as we’ve seen previously with these types of studies asymptomatic man, that was a lot of tablets, 4 tablets of abi, 1 of pred, this was the 250 dose. Uncertain benefits. So, I think there’s clear increased toxicity of the combination and that contributed to some discontinuation. With abi, there is some increased toxicity, but not major in terms of at least reported grade 3 and 4 adverse events.

Tom:
The last podcast that we did with Karim Fizazi with his plenary session, I wasn’t able to talk in that session. So, I’m getting over excited in your podcast. This is …

Brian:
It was a great podcast by the way.

Tom:
This seems to be spectacular. Is there subset analysis, which suggest the benefit is, I mean the benefits of 0.6 is clearly it’s an all-comers approach, but are there subsets which may benefit even more? Or is there anything robust there in any of the correspondent analysis?

Gert Attard:
So, we have subgroup analysis by randomization stratification factors. We don’t think any of the heterogenetic seems clinically relevant. There are 2 factors of the P value for interaction is significant. The first is performance status.

So, the patients who are PS 1 or 2, and this is really a tiny, small proportion are 1 or 2, so about 15%. They’re hazard ratio is 0.86. I think we see this commonly across studies, the PS 1, 2, not because of mets because of disease, of course, but because of concomitant illness. They have a higher death rate. So, we see less effect there.

I don’t think we’ve looked in the subgroup. Actually, I know we haven’t looked where they have more deaths from any other cause, but that’s what I would suspect.

The second interaction Tom is I guess we need to be careful how we interpret it, but it may prove to be interesting in the future. It’s regular NSAID or aspirin were used at baseline. And this is, of course, because STAMPEDE is a platform trial, NSAID have been a stratification factor since the comparison that included celecoxib. So that’s why that sits there. And we have a significant interaction with the patients taking NSAIDs at baseline having a lower hazard ratio.

So, suggestion of a greater effect. I remember Matt Sides and Nick had published an editorial a few years ago called Thursday’s child has far to go, where they looked at date of birth and benefit from abiraterone and STAMPEDE. And the men who were born on a Thursday got no benefit. And I think it was the ones on Tuesday.

Tom:
I’ve always been worried about that group.

Gert Attard:
Yeah. The ones on Tuesday. So, I like to refer to that we need to be careful. Now, if this is seen repeatedly across studies, if anyone else generates biological data that supports this interaction, then yes. But there could be other explanations, or it could be an artifact.

Tom:
Gert, how does this change practice?

Gert Attard:
So, I believe, and all my colleagues on the TMG, I think are in agreement that abi for 2 years should now be standard of care for this population. That is non-metastatic high risk. We’ve already had discussions with NHS England. And I hope that this will be widely implemented globally.

Brian:
Gert, do you think this same approach could move into the salvage radiotherapy setting where patients are commonly getting hormones for 6 months or some short course? Are there studies going on or do you think people might just adopt it in that setting as well?

Gert Attard:
So, salvage. So, they’ve had prostatectomy and then they’re having radiotherapy due to a rising PSA.

Brian:
Yeah. Adjuvant or salvage, I suppose. But yeah.

Gert Attard:
So, we allowed relapsed patients with high-risk features. However, they were only 3% of the total population. So, I think if it’s a man who has high risk disease and high-risk features, so he’s sitting somewhere between the newly diagnosed patients who constitute the majority here and the high volume metastatic at the other end. And we’re talking about a population in the middle, I think it’s clear they will derive benefit. We’ve seen it.

We’re seeing this effect now consistently across the spectrum of the disease. I think Brian, it’s going to be down to identifying patients who are high risk and are going to die from prostate cancer and the signal will be lost in a population where the chance of dying from prostate cancer is low. We only have data for high-risk patients. I cannot comment on a population that does not have high risk features.

Tom:
Gert, this is a triumph for the STAMPEDE group and the investigators who took part in and the patients and what a wonderful result. Is there anything that you want to add to what you’ve said so far?

Gert Attard:
You asked about subgroup, Tom. So that’s what we’ve done on the clinical features. We’re now analyzing the tumors. We’ve succeeded in retrieving tumor samples from 60% of these men, a major effort by the CTO and really the research teams up and down the country who do this timelessly with limited or no reimbursement. So that analysis is ongoing and hopefully in the next year or so, we’ll answer your question about subgroups that may benefit more than others.

Tom:
This is super cool. Brian, anything from you?

Brian:
No, really great work. STAMPEDE continues to produce practice changing data. So, congrats on these.

Gert Attard:
Thank you.

Tom:
Gert, looking forward to seeing you soon.