Christopher Wallis: Hello, and thank you for joining us for this UroToday Journal Club. Today, we’re discussing a recent publication entitled, Outcomes Among African American and Non-Hispanic White Men With Metastatic Castration-Resistant Prostate Cancer With First-Line Abiraterone. I’m Chris Wallis, an assistant professor in the Division of Urology at the University of Toronto. With me today is Zach Klaassen, an assistant professor in the Division of Urology at the Medical College of Georgia. You can see here, the citation from this recent work published in JAMA Network Open.

We know that mCRPC is a rapidly evolving disease space, and beginning with docetaxel in 2004 and leading right up until the approval of PARP inhibitors in 2020, we’ve seen a rapid evolution here with many phase III clinical trials leading to FDA approval of novel agents. However, we need to consider the representativeness of these data when we consider how we can apply them to our patient populations.

And so, some real seminal work led by Dr. Spratt and Dr. Osborne came out, now more than 5 years ago, but highlighting the clinical trials published at that time in 2015, we can see that among the patients enrolled on randomized control trials of CRPC, we had just over 72 patients at that time, and only 3.3% of patients on trial were Black. Whereas looking at the US population in general, we would expect that this proportion should be higher, in the range of 15-16%. As a result, we have a substantial undertreatment of patients on trial and patients receiving these proven, now life-prolonging, agents in the study settings. And so we have, for example, 150 men who received trial drug out of these 7,200, when we should expect nearly 700.

As a result of this under accrual, we have difficulty assessing the validity of these new agents in racially defined subgroups. And we know that racial disparities in prostate cancer are profound in the United States and racial disparities occur both in incidence and outcomes, and are among the greatest for prostate cancer of all health conditions. So optimization of treatment approaches in African American men is critical to address these disparities, but is hampered by their underrepresenation in the clinical trials that inform our evolving guidelines.

One example of these medications is abiraterone acetate, and we know both in the Cougar 301 data, post-chemotherapy, and subsequently in the Cougar 302 data, pre-chemotherapy, that abiraterone has a significant survival benefit compared to placebo. Now, abiraterone response according to race has subsequently been assessed, and this is initial observational data, which suggested that outcomes are likely fairly comparable between Black and White men who received abiraterone. So this is the survival probabilities, and then we also can see other metrics like PSA progression.

This led to a trial led by Dr. George based out of duke that prospectively compared outcomes between Black and White men treat with abiraterone. And as you can see here, the responses, and these are obviously surrogate measures, not survival, but responses were somewhat better among Black men than White men. This translated into differences in median time to progression as well, and suggested that there may actually be a potential for actual survival differences.

The study in question we’re reviewing today, seeks to assess whether these differences in response to abiraterone, according to patient race are generalizable to real-world cohorts. And so the authors assess differences in outcomes with first-line abiraterone treatment between African American and non-Hispanic White men using a national real-world data set, and then they used enzalutamide as a comparison to assess whether there was a mechanistic difference between abiraterone and enzalutamide. They clearly act in different aspects of the androgen receptor pathway.

To do this, they performed a retrospective cohort study of the Flatiron dataset, which is an electronic health record derived de-identified database of approximately 280 cancer clinics in 800 sites of care. Using this data set, there’s a previously utilized and validated approach to identify CRPC on the basis of physician documentation and laboratory evidence of castration resistance and rising PSA values. The authors included patients with new diagnosed CRPC who received first-line systemic therapy between the beginning of January 2012 and December 2018. Patients were excluded if they had no first-line therapy, if there was a greater than 90 day delay between identification of mCRPC and EHR activity, if there was evidence of non-metastatic CRPC, or they received non-guideline concordant treatments.

Race and ethnicity in this data set were captured and characterized in keeping with the schema proposed by the US Office of Management and Budget standards for race and ethnicity, and notably, these are self-reported race and ethnicity characteristics that are documented by clinicians and are not genetically identified, but rather rely on self report. The primary exposure here was first-line treatment approach, and the authors primarily compared abiraterone with ADT, with or without other systemic therapies, to other treatment approaches that were not based on abiraterone. Primary outcome was overall survival.

In terms of statistical analysis, the authors balanced the abiraterone and non-abiraterone groups using potentially score based IPTW weighting, and so they use both patient-level and practice-level characteristics, as highlighted here, including age, race and ethnicity, comorbidity, PSA level, docetaxel use in the mHSPC space, insurance type, opiate use, treatment setting, geographic region, and at the practice-level, the prescribing rate of abiraterone. Secondarily, the authors performed the same IPTW approach for enzalutamide exposure rather than abiraterone, and then they calculated IPTW-adjusted adjusted Kaplan-Meier curves to compare overall survival on the basis of ethnicity and race, defined as African American versus non-Hispanic White. Cox proportional hazards models were then subsequently used to estimated adjusted hazard ratios, and they further performed stratified analyses, examining the overall survival effective abiraterone compared to enzalutamide within each race group, and all models included a treatment by race interaction to allow for differential effects of each treatment modality in each racial group.

I’m going to hand over to Zach at this point to walk us through the results.

Zachary Klaassen: Thanks, Chris, for that great introduction. This is the diagram for this trial. As Chris mentioned, this was the Flatiron database of which 6,473 patients with mCRPC were assessed. Ultimately, there were 3,808 patients that were included that received first-line therapy for mCRPC. This included 1,729 that received abiraterone and 2,079 that did not receive first line abiraterone, but did receive first line treatment for mCRPC.

This is the baseline characteristics of the non abiraterone and abiraterone first-line treatment groups. You can see the characteristics on the left, and we’ll focus on the weighted population on the far right. To the middle is the unweighted population. So looking at the weighted population, among men that did not receive abiraterone, 928 received enzalutamide, the median age between this group was 74 for abiraterone and 72 for non-abiraterone group. Majority of these patients, about three-quarters, were White, and then as well as 11% that were African American in the abiraterone group and 12% of patients in the non-abiraterone group.

Comorbidities and opioid use were well balanced between these groups, as was PSA. Median PSA between these groups was 57 in the abiraterone group and 54 in the non-abiraterone group. In terms of docetaxel receipt in the hormone sensitive prostate cancer setting, 9% in the abiraterone group, 11% in the non-abiraterone group. And finally, if we look at the bottom here, in terms of practice type, the Flatiron databases, it’s typically a community based database is reflected in the groups here, 92% community treated in the abiraterone group and 94% in the non-abiraterone group, and roughly one-third of these patients in each group, a commercial health plan.

When looking at baseline characteristics of African Americans and non-Hispanic Whites, you can see that the White patients were significantly older, at 74 years of age versus 69 in the African American group. Similar comorbidities as well. We can see that the median PSA at diagnosis was actually higher in the African Americans, at 126 versus 47 in the White population. Again, similar docetaxel receipt in the metastatic hormone-sensitive setting between these groups, as well as practice type. And again, similar in the payer insurance status between these two groups.

This figure, it looks at overall survival by race looking at the first-line abiraterone treatment. The African American patients are in the dark gray and the White patients are in orange. And we can see here, the median overall survival for African Americans receiving first-line abiraterone was 23 months, compared to only 17 months for White patients, with a hazard ratio of 0.76 and 95% confidence interval of 0.60 to 0.98. This is a similar figure with a similar legend looking at enzalutamide, broken down by race, and we can see here that there was no difference between these groups, African American median overall survival of 23 months and Whites 21 months, with a hazard ratio of 0.87 and a 95% confidence interval of 0.66 to 1.14.

This is the Kaplan-Meier stratified by race and treatment. You can see here in the dark gray, the African American first-line enzalutamide overall survival is 24 months, African American first-line abiraterone was 24 months, White patients with enzalutamide was 20 months, and White patients with abiraterone almost 17 months. These tables here, look at the interaction test for race and treatment, which shows a significant interaction suggesting that the relative efficacy of abiraterone and enzalutamide is different based on race.

Several discussion points from this important study. We found that abiraterone was associated with superior overall survival for African American men versus White men. These race-stratified analyses using an enzalutamide, a comparator suggests that this OS disparity may be due to decreased benefit of abiraterone in White men. Importantly, in contrast, there was no racial difference in overall survival for first-line enzalutamide. These differential survivals by race has also been reported for other first-line mCRPC treatment. So if we look at the Sipuleucel-T treatment and the PROCEED registry, which is published by Dr Sartor and colleagues a couple of years ago, this showed improved overall survival for African American men, at 35.3 months versus 25.8 months among White men. So other potential drivers of racial differences in treatment response must be assessed, and this may include socioeconomics, as well as genetic variants.

In conclusion, prior research suggests that abiraterone is associated with greater clinical benefit for African-American patients than White patients with mCRPC. This real-world cohort study affirms these findings and is the first to find that first-line abiraterone was associated with an improved overall survival among African American patients compared with White patients. This observation may be due to decreased effectiveness of first-line abiraterone relative to other first-line therapies among White patients. And finally, future prospective research in mCRPC should include greater proportions of African American men and investigate mechanisms behind potential decreased abiraterone effectiveness in White men to ultimately inform solutions to ameliorate these disparities in these outcomes.

Thank you very much for your attention. We hope you enjoyed this UroToday Journal Club discussion.