Read the Full Video Transcript

Rashid Sayyid: Hello, everyone, and thank you for joining us today in this UroToday Journal Club recording. I'm Rashid Sayyid, a robotic urologic oncology fellow at USC, and I'm delighted to be joined today by Zach Klaassen, associate professor and program director at WellStar MCG Health. We'll be discussing the recently published JAMA Oncology meta-analysis looking at cardiovascular events and androgen receptor signaling inhibitors in advanced prostate cancer. This study, as mentioned, was published in JAMA Oncology with Dr. El-Taji as the first author.

And so, androgen receptor signaling inhibitors were initially thought mainly to be used in patients with M1 CRPC. We know in 2011 and '12 we had very important data emerging in the post-docetaxel setting. But we've seen over the last decade that the indications for this class of drugs have really grown. And if we look at the real-world data, the use of this drug is expanding, with more than 500,000 men using this drug since 2012. And even this figure is probably an underestimation of the true magnitude of the effect of this drug in our practice.

If we look at patients in the M1 CRPC setting, we have these drugs approved in the post-docetaxel setting and then moved up in the pre-docetaxel with abi and enza. And we're seeing them now being used even in combination in the first-line setting with the PARP inhibitors with TALAPRO-2, PROpel, and MAGNITUDE. We also have them indicated in the M0 CRPC setting, whereby the second-generation antiandrogens, enza, darolutamide, and apalutamide, are approved by the FDA for the treatment of these patients. We saw them as well being approved in the M1 HSPC setting, either in the doublet regimen setting or the triplet regimen setting with PEACE-1 and ARASENS. And then more recently, we're even seeing them being moved up to the non-metastatic HSPC setting, whereby we have enzalutamide approved based on the data from EMBARK and the higher risk biochemical recurrent patients.

And also, we have them being approved now in the primary treatment setting, whereby the STAMPEDE data showed the survival benefit when added to radiotherapy for patients with clinically node-positive disease or high-risk patients. And so this is not really a comprehensive list, but what it serves to show us is that these drugs are everywhere. And many physicians, including urologists and radiation oncologists, are going to have to be comfortable prescribing them and understanding their effect, both in terms of a survival benefit as well as the side effect profile.

Which leads us to this point. When we talk about cardiovascular disease in prostate cancer patients, this really has to be at the forefront of our thought process. When we look at different data, and these figures are going to vary, about 16% of prostate cancer patients die of their disease. But conversely, up to a quarter of them die of ischemic heart disease. They're more likely to die—in a subset of patients, of course, not the highest risk patients—they're much more likely to die of other conditions rather than their prostate cancer.

What further complicates it is that advanced prostate cancer patients have cardiovascular risk factors, which include, among many, increased BMI, dyslipidemia, and preexisting cardiovascular disease. And really, this survivorship aspect from the cardiovascular standpoint is an area of increased interest. One prime example of this is the data from the HERO trial that looked at relugolix or ORGOVYX and really looked at this aspect, whereby patients who were taking this drug as opposed to a regular LHRH agonist had improvement in their cardiovascular outcomes, particularly if they had a history of a prior major adverse cardiovascular event, whereby the risk was lowered from about 18% to three and a half percent going from an LHRH analog to the oral ORGOVYX. And so really, again, as we prescribe these drugs, we can't have tunnel vision thinking about the oncologic aspect, but this survivorship aspect should be at the very forefront of our thought process.

And so, you may ask, "Well, why are we doing this meta-analysis? We already know that these drugs have adverse cardiovascular effects as well as looking at comparative evidence." But the limitations of the prior studies are really twofold. First, they really only focus on oncologic endpoints and not the adverse event aspect, including the survivorship cardiovascular aspect. And also, the meta-analysis of prior studies did not include those studies of patients in earlier-stage prostate cancer, such as the M0 HSPC, focusing on the M1 HSPC and CRPC settings. And so, really, the external generalizability or the validation of these studies cannot be applied to a wider aspect when focusing on the prior non-inclusive studies. And so, based on these limitations, the objective of this study was to evaluate the risk of cardiovascular events with these androgen receptor signaling inhibitor intensifications across all subgroups, including M0 HSPC, the M1 HSPC, and then M0 CRPC and M1 CRPC patients.

And so, this was a systematic review and meta-analysis of all trials of ARSI agents for prostate cancer treatment. And the investigators conducted a systematic search of all relevant databases from inception to May of 2023. They did not include those studies which were non-randomized, as well as the trials with an ARSI in the control arm. Essentially, you don't want to compare ARSIs in the intervention and control groups because then there's really no comparison. Essentially to simplify, you want an androgen receptor signaling inhibitor in the experimental arm and then a placebo or standard of care or the equivalent of a control arm to compare the effects. Now what's notable is they included those studies that had patients that had received prior or concurrent docetaxel.

The primary study outcome was all grade and grade three or higher composite cardiovascular events. And then in terms of the secondary outcomes, they looked at grade three or worse hypertension, acute coronary syndromes, cardiac dysrhythmia, cerebrovascular events, venous thromboembolism, and cardiovascular-related death. And so, these were essentially the individual components of the primary outcome, which is a composite cardiovascular event.

In terms of their statistical methods, they used a pairwise meta-analysis—pairwise meaning comparing one to the other, intervention versus control—using a random effects model. And just to clarify, when we say random effects, that's opposed to fixed effects. You use random effects in a meta-analysis when you anticipate that the studies that you will include have differences in terms of patient characteristics, namely here the disease setting, is it M0 HSPC versus M1 CRPC, the choice of the drug that you'll use. And so, using this statistical method allows for flexibility to account for these differences in order to account for the heterogeneity that you know will be in your study. And the outcome measure that this generates is a pooled risk ratio, and that's because we here have a dichotomous outcome of cardiovascular event, yes or no.

The authors performed subgroup analysis by the disease state and the type of ARSI and essentially made sure that this effect that we observe is consistent across subgroups. And then, to further account for this, they conducted a meta-regression analysis, which is kind of along the same lines as a logistic or linear regression to quantify the effect of these variables on the study outcomes, namely age, the performance status, follow-up duration, the combination of the ARSI treatment, and the docetaxel use. Essentially just trying to gain some granularity and understanding of what variables might impact the cardiovascular toxicity of each drug. At this point, I'll turn it over to Zach, who'll go over the results as well as the discussion and highlight the key takeaways from this study.

Zach Klaassen: Thanks so much, Rashid, for that great introduction of this study. This is the PRISMA flow diagram. You can see that there were 1,912 records identified. Ultimately, 1,036 records were screened, 62 were assessed for eligibility, and ultimately 22 reports were included in the review, including 24 studies in the review. And the reason for this discrepancy was that four trials from the STAMPEDE trial platform were reported across two reports. And so this gives us a sample size of 22,166 patients.

The next two slides will look at the included studies and highlight some of the important aspects. And so when we look at the M0 HSPC disease space, there were three RCTs for 2,171 patients. For M1 HSPC, there were nine RCTs for 8,994 patients. And for M0 CRPC, there were three RCTs for 4,256 patients.

For the M1 CRPC trials, there were nine RCTs for 6,745 patients. And when we look at a couple of additional highlights from table one, we see that the median follow-up was 3.9 to 96 months. The median age was 63 to 77 years. 75% or 18 studies reported cardiovascular events, and 67% of studies, or 16 studies in total, reported grade three plus cardiovascular events.

This is the risk of bias assessment. We'll focus on the right side of the slide. The summary shows that there was very little risk of bias arising from the randomization process or bias due to deviations from the intended interventions. We do see a little bit of bias due to missing data outcome and bias in the measurement of outcome. And overall, the risk of bias was essentially low risk for about two-thirds of the studies, some concern in about 20 to 25% of the studies, and somewhat high risk in about 10% of the studies.

Let's get into some of the data. This is the all-grade cardiovascular events. And we see at the bottom here, highlighted in the green box, the relative risk for the entire cohort for ARSI versus control group for cardiovascular all-grade events was significant, 1.75 with a statistically significant 95% confidence interval of 1.50 to 2.04. What's important is that when we look at this by disease state, we see there's increased risk of cardiovascular all-grade events across all these disease spaces for the combination of ADT plus ARSI. For M0 HSPC, relative risk of 2.26. For M1 HSPC, relative risk of 1.85. For M0 CRPC, relative risk of 1.79. And for M1 CRPC, a relative risk of 1.46.

When we look at the all-grade cardiovascular events by ARSI, this is a subgroup analysis, and so we take these results in the context of being careful how we compare across treatments. We do see that when looking at all-grade cardiovascular events, the best tolerated or the least number of cardiovascular events was seen with darolutamide, relative risk of 1.30. Second was apalutamide, relative risk of 1.43. The third was abiraterone, relative risk of 1.58. Fourth, enzalutamide, relative risk of 1.93. And fifth was a combination of abi and enza with a relative risk of 2.92.

Switching gears to the grade three plus cardiovascular events, this is the second part of the primary outcome in this study. And again, we see across all these studies a relative risk for increased grade three plus cardiovascular events for ADT plus ARSI versus control was increased at a relative risk of 2.10 and a 95% confidence interval of 1.72 to 2.55. Again, as we saw for all-grade cardiovascular events, this is consistent across all these prostate cancer disease spaces. M0 HSPC, relative risk of 3.801. M1 HSPC, relative risk of 2.060. M0 CRPC, relative risk of 1.59. And M1 CRPC, relative risk of 2.03.

Similarly to the all-grade, we look at grade three plus cardiovascular events by ARSI. In this cohort, the least number of grade three plus cardiovascular events, the lowest risk ratio was for apalutamide at 1.39. Second was darolutamide at 1.91. Third was enzalutamide at 1.98. Fourth was abiraterone at 2.04. And again, the most toxic for cardiovascular was the combination of abiraterone plus enzalutamide with a relative risk of 4.08.

This is sort of the raw numbers or the incident rates for cardiovascular toxicity. We can see that any grade cardiovascular toxicity with an ARSI is 36.6%. This is one in three of our prostate cancer patients receiving ADT plus an ARSI compared to 22% of those without an ARSI. When we look at the grade three plus cardiovascular toxicity, 15.6% for those with an ARSI compared to 7.8% for those without an ARSI.

Moving to the secondary outcomes, this is looking at specific cardiovascular events. We see an increase in grade three plus hypertension for ARSIs, risk ratio of 2.25. Increase in grade three plus acute coronary syndrome, risk ratio of 1.93. Increase in grade three plus dysrhythmia, risk ratio of 1.64. Increase in grade five cardiovascular events, so this is cardiovascular mortality, risk ratio of 2.02. We also see an increased grade three plus cerebrovascular accident or stroke, risk ratio of 1.86, and no difference in the risk of grade three plus venous thromboembolic events.

When we look at these incident rates for secondary outcomes, we do see 9.1% of patients with grade three plus hypertension with the ARSI compared to 4.6% without ARSI. And then you can see the remaining breakdown for the rest of these secondary outcomes. Generally, the ARSI group had roughly one to 1.7% grade three events compared to less than 1% without the ARSI.

The key findings of this study are several. The first is that there's a twofold increase in cardiovascular morbidity associated with an ARSI. The second is there's a fourfold increase in cardiovascular morbidity with two ARSIs, namely abiraterone plus enzalutamide in this analysis. We also saw an increased risk of grade three plus adverse events, namely hypertension, acute coronary syndrome, dysrhythmias, and cardiovascular events. And finally, there's a twofold increase in the risk of cardiovascular death. Although as we just saw in the last slide, the absolute risk of cardiovascular death is still quite low.

What's important is when we apply this to everyday clinical practice, these drugs may affect patients with preexisting cardiovascular conditions compared to the healthier clinical trial population. We know that these patients in clinical trials are generally probably healthier than those in the real world. And we know from the real world that 67% of patients receiving ARSI have at least one cardiovascular comorbidity.

And so what this means is that a comprehensive baseline cardiovascular assessment and implementation of preventative strategies are crucial prior to initiating ARSI treatment among those with cardiovascular risk factors. And this may include up to two-thirds of our patients. This highlights the importance of a multidisciplinary approach, particularly coordinating with the primary care physicians, as well as the importance of cardio-oncologists.

Several limitations of this study, the first is substantial heterogeneity despite inclusion of only RCT data and stratification by disease type. And this may be secondary to variations in the collection of adverse event data, follow-up frequency, and the follow-up duration of trials. Secondly, cardiovascular outcomes were not consistently pre-specified across all studies, including two studies that reported using the CTCAE version 3.0. However, amongst these trials, we would think that these would be well-balanced among the two treatment groups, those with an ARSI and those without.

Thirdly, improved survival with ARSIs could manifest as a higher captured incidence of CV events. However, this is impossible to account for with no time to event. And this really does highlight the potential importance of an individual patient data meta-analysis. And finally, we're unable to account for established baseline cardiovascular risk factors in the majority of these trials.

The key take-home messages are that this up-to-date systematic review and meta-analysis provides evidence of increased cardiovascular events in patients receiving an ARSI plus ADT across multiple disease spaces of prostate cancer. As the use of combination therapy becomes more widespread, particularly in the early disease settings, there's a greater need to evaluate and optimize baseline cardiovascular risk prior to commencing ARSI therapy. We thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion of this recently published systematic review and meta-analysis in JAMA Oncology.