Which Systemic Therapy for Which Patient with Newly Diagnosed Metastatic Prostate Cancer Presentation - Chris Sweeney
September 17, 2019
Christopher Sweeney, MBBS, Professor of Medicine, Harvard Medical School, Medical Oncologist, Dana-Farber Cancer Institute, Boston, Massachusetts, USA.
Written Coverage: APCCC 2019: Which Systemic Therapy for Which Patients with Newly Diagnosed Metastatic Prostate Cancer?
Chris Sweeney: So I've been given the charge of which systemic therapy for which patient with newly diagnosed metastatic prostate cancer. Here are my disclosures, because I'm going to be talking about a number of different drugs. So my goal is to have a high level summary of the eight systemic trials in this space, docetaxel and the new hormones, highlight the balancing act that we as physicians and nurses and clinicians take, choosing between one treatment versus another, co-morbidities, patient cancer-related prognosis, emerging data, and gaps in our data.
Focus on overall survival, quality of life and treatment burden. And the working premise is that overall survival is still the most important endpoint in these studies, hormone-sensitive prostate cancer because it accounts for treatment burden, including treatment-related deaths and treatment benefit, including the impact of salvage systemic therapy for castration-resistant prostate cancer.
This is the spectrum of the disease that we call castration-sensitive disease, hormone-sensitive prostate cancer, castration naïve and whatever you want to call it. Some patients present de novo and others present after a prostatectomy or radiation metachronous. Some are fit and young, some are frail and elderly and every iteration between. Some have minimal disease on conventional scans and some have widespread disease. Some will have had prior adjuvant testosterone suppression with radiation, or with a prostatectomy and possibly abiraterone or docetaxel.
Now take these two patients for a second, the 55-year-old with no comorbidities, high volume de novo metastatic disease. This is the 82-year-old with congestive heart failure, coronary artery disease and two bony metastases 10 years after prostatectomy. These are two different disease settings. These are two different patients and we have to be selective about how, which drug, and when and why.
We've all heard about this definition of CHAARTED low and high volume and I can come back to that if anyone ever wants to hear how we came up with that prospectively in the trial. And I give great credit to the GETUG and the STAMPEDE team going back and doing it retrospectively because our data does evolve and we have to not be stagnant. We'd looked at this in both CHAARTED and GETUG prior to local therapy and low volume disease. Your meeting time from overall survival from the time you've started your testosterone suppression is eight years. If you have one or the other, it's about four and a half years. So clearly there's a watershed between the two different disease states and three years for the de novo high volume. We then looked at this in a hospital registry called the Dana Farber Cancer Institute. Basically identical, both clinical trial and routine practice.
But there are limitations, and I won't dwell on this because I know Dr. Clarke is going to do this later, but CHAARTED's got some misclassifications. Lung only isn't only an indolent disease, but in some cases, and we call it high volume, large bulky nodes with two, three, bony metastases, some of those patients do poorly. We do not account for de novo versus prior local therapy, and some patients may have many bony metastases but never spill out to the ribs and the like and we would've called them low volume. That's a rare scenario but there are some misclassifications. Latitude has some misclassifications. Again, similar to what we did with CHAARTED note, lung only would have been visceral. It only accounts for de novo presentation. None of those patients were prior metachronous and it uses Gleason and not everyone with prostate cancer metastatic disease gets a Gleason score, because they may have a metastatic biopsy or get treated based on clinical features alone.
Dr. Clark and his team have talked about node only versus bone only and when you look at that, if you look at those patients in the SEER data, most of those patients who get called M1 are de novo metastatic M1 and have a median survival of about three years in that paper. Clearly we need to improve upon what we have.
So I'm going to go through the current scorecard for docetaxel by volume. I just want to go back and say as part of this spectrum of metachronous disease is the rising PSA state and Mike Morris is doing a yeoman's task of trying to get the registry updated of the patients who got docetaxel on TAX3503. Rising PSA, post-prostatectomy or radiation. No delay in the time to PSA progression, and Stefan Oudard and the team recently published the results of docetaxel in the same patient population rising PSA, post-prostatectomy, radiation, possibly PSMA PET-positive. So very low volume metachronous disease, no benefit.
We prospectively looked at this in the CHAARTED and we, this slide is just to emphasize interim analyzes do not report the overall story. It's a snapshot in time and it reports the outcomes of those who have a poor prognosis, and on the left overall survival 13-month improvement in median, hazard ratio of 0.6 and then it gets diluted with time when we looked at the longer-term followup. The overall survival in the high volume didn't change at all. 17 months hazard ratio 0.6, but it changed with the low volume because the indolent patients diluted the overall survival curve over time. You can see it was trending possibly with the early followup for the low volume patients, but we lost that in the end. I do believe there are patients who are low volume and specifically the low volume de novo metastatic who are just short of being high volume and may get a treatment benefit, but right now I do not know and we need more data to clearly annotate this.
Here we have the test for heterogeneity showing that there was a difference between the high volume and the low volume patients in the CHAARTED study. And that could be because of the patient mix, and patient mix for CHAARTED is both metachronous low volume and synchronous low volume. Whereas STAMPEDE, let's be very clear, nearly all of those patients are synchronous de novo metastatic. So patient mix may impact how we see the results from our trials. Quality of life, we didn't see a quality of life benefit for the low volume patients and we did see one for high volume and Alicia Morgans published this last year. Again, evolving data informing us on how to treat individual patients and we do get a quality of life benefit for the high volume patients.
So this is the big snapshot and recognizing this is evolving over time, we see a very clear signal for an overall survival benefit even in the so-called negative study of GETUG-15, hazard ratio going the same direction as CHAARTED and STAMPEDE docetaxel. You can see in the high volume you do a test for homogeneity and actually the high volume in both studies, GETUG-15 and CHAARTED with a similar, and we see a clear benefit of about 0.68 when you pull the data for the high volume, and we have a no clear benefit in the low volume with a hazard ratio of 1.03, 479 patients. But again, many of these patients are both metachronous or synchronous low volume. Test for heterogeneity showed that the high volume and the low volume were different between the two studies, but as many people are aware we'll have a larger data set of low volume, but nearly all de novo metastatic presented at ESMO this year. This is where there is a bit of a variation in interpretation of the data. Let's recognize when STAMPEDE was started, it was started as a big M0-M1 combined analysis, but recognizing M0 includes hybrid localized to with ADT and radiation, rising PSA post-local therapy, and then one is both low volume and high volume.
With time, the team is passing this out and we now have 11,000 data Nick James was just alluding to, and there's much more granularity evolving with time. The initial analysis was based on statistical inferences, but kudos to the team to going back and actually get greater granularity. So they basically, their inference was, because M0 and M1 in their earlier look was not heterogeneous on the test for heterogeneity. But, I would argue that I'd need more data for that bottom right curve there to show that there's a benefit in the non-metastatic and that is evolving with time.
So two studies with direct overall survival benefit, documented quality of life improvement, no direct evidence yet and this is evolving. And the next question is how much of a benefit is needed in the low volume? So 2.9% overall survival benefit with the non-steroidals and the combined androgen blockade was not enough to change practice and become a standard of care. How much do we need it for docetaxel, where yet the benefit is less in routine practice, toxicity is greater in routine practice. Volume is prognostic for outcome and for me, I think it is predictive for docetaxel benefit in high volume but not low volume. Is there biological underpinning to be determined? Work in progress.
Next scorecard, abiraterone by volume of disease. High volume, low volume, clearly longterm follow-up. You can see the benefit and the median survival is again three years for this high volume group. De novo metastatic, it's incredibly consistent. Low volume, less of a signal, but going in the right direction. Quality of life benefit, slower decline in deterioration in quality of life with the early abiraterone. All good news, all going in the same direction.
And then we have the longterm followup. So not the longterm followup, the volume analysis but that's just about to be published, I believe, in the high risk versus non-high risk. And again we see a clear survival benefit in the high risk with a median survival of about three years for the high volume and a less of a treatment benefit, but still going the right direction, for abiraterone. But remember, nearly all the patients in STAMPEDE are again de novo low volume. So the three-year point estimates do help patient counseling. With relative risks, I think are less intuitive for patients. So I'd like to be able to say at three years your chances are X versus Y and we can actually quantitate that to patients, quantify that for patients: 20% for high volume, 5% for low volume and it's an easier counseling question. We do need longer-term overall survival benefit to see if the survival benefit is greater with early use in these indolent patients, or do some patients get salvage with subsequent therapy? And again this is a conversation about the de novo low volume patients.
Now moving onto the score card for amides. Isn't it great we've got scorecards for three different types of drugs. We didn't have this, it was all ADT, ADT, ADT, ADT and more ADT. But now we've got new data. ENZAMET, I'm going to be the first to say, it's a mixed bag. We see an overall survival benefit in the hazard ratio of 0.67 again on an interim analysis. So this is driven by the early poor prognosis patients, but it's the mixed bag of high and low volume, de novo and metachronous, concurrent docetaxel in every permutation.
Just to note in the early analysis, we do not see that the patients who got early docetaxel, which was mostly the high volume, an overall survival benefit, but we do see a major progression-free survival benefit. We do need longer follow up for this group. Notably, look at this. Those who did not get docetaxel, mostly low volume, you can see a clear survival signal in those patients.
These are the three-year point estimates because this is where the medium follow up is. This is where we have the most information in longer-term follow-up and you can see no clear benefit if you've got early docetaxel, a clear benefit if he didn't get early docetaxel and a benefit in high volume, but the benefit is diluted because our control arm, many of them got docetaxel. But there is a clear benefit in the low volume and now we're talking about a 90% overall survival peak for patients with low volume metastatic disease at three years. I'm just going to take this for a second. The patients on ENZAMET got standard care on both arms, about half the patients who had progressed had got early docetaxel, life-prolonging. What you can see, the main difference in the subsequent therapy was those who got testosterone suppression with the weak non-steroidal 79% of them, almost 80% had subsequent amide or abiraterone. Good news. There was no difference in the other drugs.
The other thing to notice, some patients who have not had the non-steroidal, sorry, have not had enzalutamide or abi, patients who've had an anti-androgen withdrawal phenomenon. One of my patients on the study who's nine months into an anti-angiogenic withdrawal and responding, and the other patients may be those who rapidly progressed to not small cell cancer or anaplastic who would be futile to give the amide. I'd say 80% is probably about right.
Fortunately, this study is bolstered by the results from the ARCHES study. Again, one thing that we see is progression-free survival benefit, but we need longer-term follow-up for the overall survival analysis, and the TITAN study presented by Kim Chi in the New England Journal, also showing something and the same message, I think, but the main thing I take away from this, this is mostly high volume de novo metastatic. Few patients had prior radiation, but there are some metachronous patients, but even in those who had prior docetaxel, there was less of an overall survival signal. We need to pull that data and analyze.
Now I've been harping on about this de novo late relapsing patient and these are the only two data sets where we actually have a look at this with the hormones. ENZAMET and TITAN forest plots analysis are showing a benefit in these patients who presented with M0 and relapsed later. That's the extent of that data in this unique patient population, but it's going in the right direction for the amides, we just don't have it for abiraterone.
So a brief word on skeletal-related events per zoledronic acid and denosumab because if Silke and Aurelius says, Chris does. So there is two studies that show no overall survival benefit. Is that 15 or the...? Perfect, excellent. [inaudible 00:14:52], no benefit. First-line hormone-sensitive disease, we do not give these agents because the first-line therapy is probably working well. Do we add it on early for those not responding well? Possibly, work in progress. Osteoporosis management, a completely different conversation, especially with the patient with long term abiraterone, on prednisone and super low testosterones. Think about monitoring them and probably monitor their radius rather than their lumbar vertebra.
Here is the final scorecard. You ready? This used to be all ADT alone, but this is what I see. Chemo-fit, high volume old agents an option. Not chemo-fit, high volume, abi, apa, and enzalutamide. There's some direct evidence, chemo fit and not chemo fit, low-volume, de novo metastatic. All those options are there and maybe docetaxel for the not chemo fit in the low volume de novo metastatic, to be determined with time. Chemo-fit and not chemo-fit, prior local therapy, weak data but some for apa and enza.
And the concluding slide. Balancing act where we need to look for patterns in the available data to make treatments for individual patients in front of us. Balancing disease burden, co-morbidities, treatment benefit, treatment risk. We need to harmonize and improve our clinical definitions and this is a work in progress. I think we need to revisit treatment breaks in intermittent therapy and we are doing biomarker work from this, and between Gert, myself, STAMPEDE team, ENZAMET team, CHAARTED team, more than 3000 patients with blood and tumor specimens to help with the biological characterization and proving how we prevent emergence of CRPC in the future. Thank you for your time.