Enzalutamide Decreases Cognitive Function vs Darolutamide in Prostate Cancer

Data from the phase 2 ARACOG trial (NCT04335682) demonstrated that enzalutamide (Xtandi) showed significantly greater cognitive score decline compared with darolutamide (Nubeqa) in patients with prostate cancer, according to Alicia Morgans, MD, MPH.¹

“The ARACOG trial [investigators] tried to understand whether there was a difference in change from baseline cognitive function between patients who were treated with darolutamide vs enzalutamide; both of these treatments are highly utilized,” Morgans said in an interview with OncLive. “They're both highly important. Enzalutamide has a much broader label [compared with] darolutamide, but this was really an important question for us to ask and to understand.”

Morgans is a genitourinary medical oncologist and medical director of the Survivorship Program at Dana-Farber Cancer Institute, and an associate professor of medicine at Harvard Medical School in Boston, Massachusetts.

In the interview, Morgans discussed the background of the ARACOG study, the notable findings that she presented from the trial during the 2026 American Society of Clinical Oncology Annual Meeting, and how these findings affect her clinical practice.

OncLive: What were the key design characteristics of ARACOG?

Morgans: ARACOG was designed with a primary end point of change in cognitive function.¹^,² We have assessed things like adverse effects and cognitive function within other trials, but having a trial designed specifically to identify this, measure it, and compare between treatment arms is unique. It is also important as we're thinking about the statistical rigor that goes into trial design, especially as we think about that P value that everyone is so excited about at the end. When we design the P value around the primary end point of cognitive testing, we ensure and strengthen our ability to really compare things around that specific issue, and I think that's one of the most important things about this study.

Patients were [randomly assigned] to treatment either with darolutamide or with enzalutamide, and they were followed for 48 weeks of total treatment. In this analysis, we are reporting to the primary end point, which was at 24 weeks after initiating treatment, and we were trying to understand whether there was a difference between treatment arms in terms of the change in cognitive function as assessed by computer-based cognitive functional assessments between baseline and 24 weeks.

Recognizing that many patients with advanced prostate cancer who are going to get these different treatments may start from a different place, we had to understand the change from where they started, rather than an absolute difference at that time point. What this helped us to evaluate was that maximally changed cognitive domain, which was our primary end point. So whichever treatment, whichever domain of cognitive testing, or whichever module of cognitive testing was most significantly impacted, we would compare between the treatment arms to make sure that we were understanding the worst effects of each treatment and then comparing those, rather than just comparing this cognitive domain vs that cognitive domain.

What were the key findings that were presented during ASCO?

When we compared those maximally changed cognitive domains between the different treatment arms, we found that there was a greater change, or decline, in cognitive function by these CANTAB computer-based cognitive tests for patients treated with enzalutamide than for patients treated with darolutamide, and that was statistically significant.¹

We also found that when we looked at individual modules assessing different aspects of cognitive function, 4 out of 5 modules statistically significantly favored darolutamide treatment over enzalutamide treatment in that assessment. Ultimately, these are nonclinically used cognitive tests—that should be very clear. They're research-assessed cognitive tests, but they did suggest that there might be differences between these treatments in terms of the effect of therapy on cognitive function for people with prostate cancer.

How would you contextualize the clinical significance of the difference between these effective cognitive test rates?

Patients come into any clinical visit with their own needs, expectations, and preferences, and, importantly, sometimes that is: Which treatment works the best? And sometimes that is: How is that treatment going to treat me, my body, and my function? How is it going to affect my life and my ability to care for myself or other people?

Most patients come in with needs and expectations that marry those things, and they sit somewhere between wanting to maximize treatment efficacy and really maximize quality of life. When we have a better understanding of the effects of different treatments on cognitive function, we are able, as clinicians, to provide information to the best of our ability to patients so that they can try to sort through their decision between this androgen receptor pathway inhibitor [ARPI] and that ARPI and find the one that makes the most sense for them.

There are always going to be these trade-offs, and we inherently help make some of those trade-off decisions as we engage in treatment decisions. But having the information, at least to say that this treatment may affect a particularly vulnerable population more than that treatment in terms of cognitive function, will be valuable as we sit with our patients and try to sort through all the options that they face during their treatment.

Are there certain patient characteristics or disease factors that help you weigh these risks and benefits?

The analyses that are still to be performed include understanding patients with a particularly high risk related to dementia because of family history, older age, or perhaps comorbidities that can contribute to the development of dementia. We are doing these analyses within the ARACOG data set and are grateful to patients for contributing all the rich information that they have so that we can sort through those subgroup analyses and understand who might be most vulnerable.

We are also doing analyses to understand whether there's a genetic component to all of this by integrating a polygenic hazard score to try to sort that out. But I think that, at least at ASCO 2026, our primary goal was to report on the primary end point and make sure that we are sharing that, when we look at objective cognitive test scores from these computer-based cognitive tests, we do see a difference between treatment arms.

There is just so much more to learn and investigate. Even understanding how these drugs may act in particularly vulnerable or particularly invulnerable populations is something that we hope to report at a future meeting.

Do you think these cognitive test assessments may change your clinical approach to monitoring patients?

Every clinician has a different approach to cognitive assessments. For me personally, I talk with patients and, perhaps more importantly, with their loved ones, whether it's a spouse, a partner, or an adult child, to try to understand their general cognitive function over time. I continue to have those conversations and check-ins to understand how any of our treatments may be affecting patients' ability to think clearly and smoothly and to take care of themselves and be fully independent.

These are ongoing conversations that I don't think necessarily will change based on the information from this trial, at least from my perspective in clinical practice. But certainly, if we have patients who we think need to be on enzalutamide, perhaps in the back of my mind, whether a patient is having any decline in cognitive function will be something extra that I ensure I ask about and really dig into.

I would say that most clinicians are doing this anyway, perhaps not in a standardized way. We ask patients how they're feeling, we ask patients what's changed, and we ask their spouses, ‘How are they doing?’ Of course, there are other loved ones [we consult] as well, and I think we do get a sense of cognitive function.

Particularly for individuals who are at high risk because of older age or who may be [having difficulty maintaining] their independence, we might be a little more thoughtful, regardless of what treatment the patient is on, just to make sure that things are going smoothly and that we can help patients both treat their cancer and care for themselves at the same time.

References

1. Morgans AK, Bobek O, Kwon DH, et al. Cognitive effects of darolutamide vs enzalutamide: results of ARACOG (AFT-47), a randomized clinical trial from the Alliance for Clinical Trials in Oncology. J Clin Oncol. 2026;44(suppl 16):5005. doi:10.1200/JCO.2026.44.16_suppl.5005
2. Androgen receptor directed therapy on cognitive function in patients treated with darolutamide or enzalutamide (ARACOG). ClinicalTrials.gov. Updated November 4, 2025. Accessed June 16, 2026. https://clinicaltrials.gov/study/NCT04335682