Examining Survival Rates in Prostate Cancer: Exploring the STAMPEDE Platform, Journal Club - Rashid Sayyid & Zachary Klaassen
June 5, 2023
Rashid Sayyid, MD, MSc, Urologic Oncology Fellow, Division of Urology, University of Toronto, Toronto, Ontario
Zachary Klaassen, MD, MSc, Urologic Oncologist, Assistant Professor Surgery/Urology at the Medical College of Georgia at Augusta University, Georgia Cancer Center
Rashid Sayyid: Hello everyone, this is Rashid Sayyid. I'm a Urologic Oncology Fellow at the University of Toronto, and along with Zach Klaassen, Assistant Professor and Program director at Augusta University, we'll be discussing one of the recent publications from the STAMPEDE platform, looking at abiraterone acetate plus prednisolone, with or without enzalutamide for patients with metastatic prostate cancer starting ADT, the final results from two randomized Phase III trials. This study was recently published in Lancet Oncology.
So, we know that the historic median overall survival for patients with metastatic hormone-sensitive prostate cancer with ADT plus or minus of first generation non-steroidal, such as bicalutamide, is about only three and a half years. So, 42 months, based on historic data. But since 2015, we've seen numerous RCTs in this disease space that have demonstrated overall survival benefits, with doublet treatment intensification. We know that we have docetaxel addition from STAMPEDE and CHAARTED, shows an improvement in the median overall survival from 43 months, about three and a half years, to 59, about five years, with a median follow-up of 78 months.
We know that abiraterone has been shown to improve overall survival, based on the results from STAMPEDE and LATITUDE, with STAMPEDE improving a three-year OS from 76% to 83%, and we also know that enzalutamide, in the ENZAMET and ARCHES trial, and apalutamide and TITAN trial, have also showed an OS benefit in this disease space.
So, the objective of this study was to report separately the long-term outcomes of abiraterone plus standard of care, versus standard of care alone, and in a second trial, so it's not one trial with two arms, it's a completely separate trial to report the outcomes of abiraterone plus enzalutamide, plus standard of care, versus standard of care alone. Of note, the second trial was initiated after the completion of accrual to the abiraterone trial. And what was the rationale for combining abiraterone and enzalutamide? So we know that androgen receptor blockers, enzalutamide may increase androgen synthesis, and then conversely, residual hormones in patients receiving androgen synthesis inhibitors, such as abiraterone, may activate AR signaling. So, by including both drugs, in theory, you would counteract the escape mechanisms that each drug suffers from, and minimize the resistance that each one may have.
And after doing both trials, the authors had a preplanned fixed effect meta-analysis to indirectly compare the results of these two separate trials using individual patient data, and the hypothesis here was that there was an estimated 15% reduction in risk of death, with combination abi/enza, versus abi alone. So, what they're trying to say is the additive benefit of enza with abi, compared to abi alone, would be about a 15% reduction in risk of death.
And so, to do this, the authors performed two open label, randomized controlled Phase III trials of the STAMPEDE platform protocol. And as we know, this was conducted across 117 sites in the UK and Switzerland, and these trials included patients with metastatic hormone-sensitive prostate cancer, per conventional imaging. These patients could've had either de novo or recurrent MHSPC. When we say they're recurrent, they will need to ensure that they haven't had ADT for more than 12 months, and completed it at least at 12 months prior to enrollment. All patients who had fairly good performance status, zero to two, mostly are going to be zero to one. No age restrictions were applied. All patients at inclusion had to have normal hematologic renal liver labs, and no history of clinically significant cardiac disease. So, as is standard with these RCTs, this is a healthier patient profile.
Patients were randomized one-to-one in the first trial to either abiraterone plus standard of care, versus standard of care alone, and this was between November 2011 and January 2014. Of note, standard of care included ADT plus prednisone, or prednisolone. Prednisolone was given in the Swiss sites, and essentially was five milligrams daily. In the second trial, patients received abiraterone plus enzalutamide, plus standard of care, versus standard of care alone, and as we previously mentioned, the accrual for this trial started after the first one was done. So we see July 2014, so about six months after the first one was done. Accrual for the second started, and this continued until March 2016. Standard of care, of note, was a little different in this trial. So, ADT was considered standard of care until December 2015, and then as we know, we have the data that showed that docetaxel intensification in this setting conferred a survival benefit. So, the authors pivoted at this point, and altered the protocol to include docetaxel for six cycles, plus prednisolone 10 milligrams daily, as the new standard of care, in addition to ADT, concurrent with standard of care at that time.
Of note, radiation was allowed for symptom palliation. Again, I want to emphasize the point, these are two separate trials, and so, there are no overlapping controls. So, there are two separate control populations for each trial, and randomization was performed using what we call the minimization approach, with stratification by numerous factors, and this is to ensure that the two arms of each trial were fairly balanced for prognostic factors, and minimize again, any selection bias that may confound the results that we see.
Study treatments were administered until patients had evidence of radiographic progression, or any reason that required drug discontinuation. Of note, in the combination trial of enza and abi, patients received full dose enza of 160 daily, and abi 1000 daily, and of course, those modifications were permitted per protocol. For participant assessment, all patients underwent PSA safety lab tests and adverse event evaluation every six weeks during the first six months, so fairly commonly, and then every 12 weeks or three months until year two, and then once they reached year two, every six months until year five. The PSA nadir was defined as the lowest PSA value within 24 weeks of randomization, and imaging was performed per local guidelines, and again, the outcome was investigator-assessed, as opposed to central review.
The primary outcome was overall survival for both trials, and this was defined from randomization to any-cause death. They also looked at secondary outcomes, such as prostate cancer mortality, failure-free survival, progression-free survival, metastatic progression-free survival, whether a patient had symptomatic skeletal related events, and also looked at toxic effects and adverse events as well.
So, just a quick note about sample size calculations. So, the study required for both total, 1,800 patients with either high-risk localized or metastatic disease, and then this study had 90% power. So, usually we see about 80% power in these trials, so this was fairly well-powered to assess the outcome, which was anticipated to have a hazard ratio of 0.75 at a two-sided P-value of 0.05, again, which is pretty standard, and this was assuming a four-year median overall survival with standard of care. It's important to note that both trials recruited their target sample size, and standard efficacy analyses was performed in an intention-to-treat population, so, based on where you randomized initially. And then toxic effects and adverse events were analyzed in the safety population, so meaning any patient starting treatment within the randomly assigned groups.
To account for different durations of treatment exposure, so if you have a better response on the first drug and you receive it longer, you would expect, just by virtue of increased cumulative exposure, to have worse adverse events. So, to adjust and account for that, the authors looked at time to any Grade 3 or 5 adverse event, as opposed to just looking the proportion of adverse events. So, again, a nice little adjustment by the authors to account for differential exposure time to each drug.
Standard survival analysis methods were used with Kaplan-Meier curves, and then on the Cox proportional hazards models, which allows for multi-variable modeling. The authors and their models adjusted for all stratification factors, except treatment center and ADT type, and then they also assessed the proportionality assumption, which essentially looks at if the hazard of an event is constant over time. And so, you would assume with these models that the hazard remains pretty constant, but we know in reality, that that's not always the case. So, if it was a non-proportional hazard, so it violated this assumption, the authors had decided that they would use flexible parametric models, with time-varying treatment effects to account for the fact that the hazard changes over time. They also looked at prostate cancer-specific mortality and a competing risks model, which adjusts for the competing risk of other-cause death, which may preclude the occurrence of prostate cancer-specific mortality.
For the meta-analysis part, looking at whether there was benefit with adding enzalutamide to abi, versus abi alone, they used a fixed effect meta-analysis using individual patient data, and heterogeneity between the two trials was calculated using the Cochran's Q test, again, making sure that the use of the fixed effect model was reasonable in this setting.
They also looked at pre-specified subgroup analyses to evaluate for consistency of treatment effect between both trials, and at cross-randomization factors such as the age, nodal stage, the performance status, whether patients used NSAIDs or aspirin, and then planned docetaxel as standard of care, and they also importantly, looked at the outcomes by the CHAARTED volume steps. And at this point, I'll turn it over to Zach to go over the results and discussion and conclusions of this report.
Zach Klaassen: Thanks so much, Rashid, for that great introduction. So, this is the trial profile. We'll spend a little bit of time walking through this. As Rashid mentioned, there were two separate randomized controlled trials. So overall, there was 5,488 patients randomly assigned, and this included 1,003 with metastatic disease allocated to the abiraterone trial, and this included 502 to standard of care, as well as 501 to standard of care plus abiraterone, and additionally, during the same time period, 916 patients with metastatic disease were allocated to the abi plus enza trial, including 454 assigned to standard of care, and 462 assigned to standard of care plus abi plus enza.
This is the baseline characteristics, and you can see here this is broken down by the two trials, and the stratifications within those trials. We can see here that as expected, the median age was late 60s for these patients. The median PSA was quite high for all four groups of patients roughly, high 90s PSA. Time from diagnosis to randomization was roughly 71 to 78 months. The most common T stage amongst these patients was T3, more than 50%, and roughly 50 to 60% were N-positive in terms of metastatic volume, high volume, and low volume was roughly 50/50. And the Gleason score, most commonly, was Gleason 9 disease, at roughly 40 to 50%.
With regards to performance status, roughly three quarters of patients were performance status of zero, pain from prostate cancer was not present in roughly 80% of patient. And when we flip over to the right side of this slide, this is the continuation of the baseline characteristics table, we see that the overwhelming majority of patients, nearly 100% of patients had LHRH agonists or antagonists. Palliative therapy was planned as standard of care, and only about one to 5% of patients, majority of patients did not have previous treatment to the prostate, and docetaxel was planned as standard of care in only 9% of patients.
The next several slides will look similar to this one. So, this is the breakdown between the abiraterone trial on the left, and the abiraterone plus enzalutamide trial on the right. For the abiraterone trial standard of care is in red, standard of care plus abi is in blue, and for the abi/enza trial, standard of care is in red, and standard of care plus abi/enza is in blue. So, this is overall survival by trial. We can see that for both of these trials, there was a benefit to treatment intensification in the abi trial hazard ratio of 0.62, 95% confidence interval of 0.53 to 0.73. In the abi/enza trial, hazard ratio of 0.65, so very similar to the abi alone, with a 95% confidence interval of 0.55 to 0.77.
This is overall survival by low volume metastatic disease. In the abiraterone trial, we see a benefit for abi, hazard ratio of 0.58, 95% confidence interval of 0.44 to 0.75. For the abi plus enza trial, we also see a benefit to the addition of treatment intensification, with a hazard ratio of 0.65, 95% confidence interval, 0.48 to 0.87.
This is overall survival by high volume metastatic disease. Again, we see a benefit for both of these trials. Abiraterone hazard ratio of 0.66, abiraterone plus enzalutamide, hazard ratio of 0.68.
This is the overall survival subgroup analysis, just for the abiraterone trial, and we see that generally, all of these subgroups did have a benefit to adding abiraterone, with the only one that didn't, was the NX patients, as well as perhaps some heterogeneity in patients over the age of 70. But generally, the majority of these patients did benefit from the addition of abiraterone. This is the overall subgroup analysis for abi plus enza. Again, we see the majority of these patients did have a benefit to treatment intensification, with the addition of abiraterone plus enzalutamide.
Specific to the abi plus enza trial, this was for planned docetaxel use, no versus yes, and we can see overall, as well as in the substratification, there was no benefit to the addition of abiraterone plus enza, for those planned or not planned to have docetaxel use.
This is the overall survival by subgroup analysis, by volume of disease. Split by trial, we can see here low versus high, also split by abi trial alone versus abi plus enza, and we see that for both trials, in both low and high volume disease, we see benefit to treatment intensification.
This is prostate cancer-specific survival. Again, using the same color scheme as we previously had discussed. We see a benefit to abiraterone versus standard of care, with a hazard ratio of 0.56. Statistically significant 95% confidence interval. Again, we see a benefit to standard of care plus abi, plus enza, on the right of the screen, with a hazard ratio of 0.61.
Similarly, for failure-free survival, early and consistent splitting of the curves for both trials. In the abi trial, hazard ratio of 0.34. In the abi plus enza trial, hazard ratio of 0.36.
Similarly, for metastatic progression-free survival, again, an early and consistent splitting of the curves for Both of these trials. The abiraterone trial has a ratio of 0.50, and for the abi plus enza trial, 0.52.
Where this is important here is the adverse event profile. So, you can see that on the left is the abi trial, on the right is the treatment intensification with abi plus enza. I've highlighted three specific adverse events, because we do see with the addition of abi plus enza, there is more cardiac toxicity, specifically when we're looking at Grade 3, 4 events, at 2% specifically, compared to 2 and 1% with the abi trial. As you would expect, fatigue is quite a bit higher with the treatment intensification of abi plus enza, 78% Grade 1-2 events, 8% Grade 3 events compared to 65, and 3% respectively. Again, at the bottom, we see for hypertension, more Grade 3 adverse events for enza plus abi, 16.3% versus 3%, and Grade 1-2 events, 42% versus 13%. So, we are seeing specific to some of the adverse events, more adverse events, and more significant adverse events for abi plus enza.
So, by way of discussion, combining abiraterone plus enzalutamide with ADT is more effective than ADT alone, but from indirect comparison, this study showed no improvement over abiraterone plus ADT. Of note, STAMPEDE has not formally compared the abi plus enza trial with trials testing enzalutamide, and increasing toxicity with the combination of abi plus enza does not justify further exploration in unselected patients. Follow-up of patients allocated to abiraterone was longer than other trials in this disease space. The survival benefit of starting abi plus ADT is maintained with longer follow-up, and 23% of patients in the abi trial, and 22% of patients in the abi plus enza trial continued abiraterone, with no indication for a treatment change, after more than 82 months of follow-up.
So, in conclusion, enzalutamide plus abiraterone should not be combined for patients with prostate cancer, starting long-term ADT. However, clinically important improvements in overall survival, from the addition of abiraterone to ADT, are maintained for longer than seven years. We thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion.