Read the Full Video Transcript

Stephen Freedland: Hi, I'm Steve Freedland, a Urologist at Cedars-Sinai Medical Center in Los Angeles, California, and a Staff Physician at the Durham VA Medical Center in Durham, North Carolina. This Beyond the Abstract commentary summarizes key findings from article recently published in the World Journal of Urology. This article describes real-world data of patient characteristics, treatment and treatment patterns and survival outcomes of patients with localized or locally advanced prostate cancer in the PANTHER study. The study was funded by Janssen. Disclosures of all authors are listed on the manuscript.

We know that for patients with localized prostate cancer or locally advanced prostate cancer who undergo treatment, that there is a significant risk, up to one out of three developing early progression to metastatic disease, and once patients survive and are able to progress to metastatic disease, if they do that, often their survival can be very short. And so we really need better treatments for these patients. But currently, there's a lack of consensus on how to do that. And so that was our goal, was to use real-world data to better understand this disease space, understand the implications of early progression, and ultimately to develop potential options to better treat these patients.

Specifically, the objectives of our PANTHER study was to determine the potential impact of patient demographics and characteristics, treatment patterns, and really importantly, the timing of disease recurrence and how this affects survival outcomes in real-world data using both insurance claims and electronic health records. To do this, we used the Optum database, both insurance claims data as well as the EHR part of Optum. We have key inclusion criteria. Obviously, they had to have localized prostate cancer, locally advanced adults, and they needed to be treated with aggressive treatment, whether it be primary ADT, which is really not our focus, but we can include it as a control group, surgery or radiation.

So those that got active surveillance walked in the door with metastatic disease were not eligible for the study and they couldn't have had other primary cancers as well. We collected key demographic information, diagnosis, procedures, treatment patterns, and really importantly, data on death, which was one of the primary outcomes.

When looking at outcomes, as I said, overall survival was one of the primary outcomes we looked at. We also looked at, and I'm not going to go into it here just for the sake of time, but event-free survival, progression-free survival, metastasis-free survival, evidence of disease progression, which we'll see some data on, but many other outcomes as well as the treatments that they ultimately received down the road.

The overall study was really large numbers of patients. If we combine it, it was over 60,000 patients in both the claims data set and the EHR data set. Really large numbers. If we look at the baseline characteristics, as you would expect the ADT-only patients were significantly older. Radiation was older than surgery, and the Charlson Comorbidity Index, most patients had zero and a sizable percent had one. It's not the majority of patients, but unlike what you see in clinical trials, this real-world data set, we do have patients with up to four or higher Charlson Comorbidity Index, which as expected was higher in the ADT set, that was the highest and then higher in the radiation and lowest in the radical prostatectomy. But we did see actually even in the radical prostatectomy, some patients had significant comorbidities. Again, just how our real-world data vary from some of the clinical trials that have been done.

If we looked at what was done more frequently, radiation or surgery, it really depended on the data set. But if we combine the two data sets together, really it's pretty similar percentage had surgery or radiation and about 10% of patients had ADT alone in both data sets.

If we look at the real-world overall survival, what we see is regardless of the data set we see, it's worse in the ADT alone group, and as we said, they're older, they're sicker, there's reasons someone wouldn't be eligible for surgery or radiation. So not completely surprising. Median survival somewhere in the neighborhood of five to six and a half, seven years, somewhere in there. If we look at the surgery versus radiation, again, many data sets have shown this. Ours no different, that the surgical patients were doing better, but again, they were older, less comorbidities. We didn't adjust for all that because we didn't have the richness of data we'd really want. So we're certainly not making claims at all that surgery is better than radiation, but notably, the median survival, if we look at the radiation patients, was about 11 years and was even longer with surgery. So it is a long natural history. I think that's the key take-home that these patients are living for quite a while.

If we look at metastasis-free survival, we've seen the same story. The ADT alone patients somewhere in that four to six years, somewhere in that neighborhood, and the surgery and radiation, even medians had not been reached even after 10, 11 years. So again, overall patients are doing relatively well. Yes, surgery patients do a little bit better, but they're younger and healthier and that likely explains why they're doing better.

But really important, we next ask the question, what happens if you recur early? What happens if during that first year you develop a recurrence? How does that affect your overall survival? And importantly, we made sure everyone lived to one year so you couldn't have a recurrence and then die within that one year because that would bias things to make that group look of course worse, but everyone had to live to the one-year mark. And from there on we looked at survival and what you see is with hazard ratios in the 2.6 to 2.7 range, that recurring early is associated with significantly much worse overall survival. It over doubles the risk that someone would die thereafter.

If we look at the three-year mark, so saying recur within three years, yes, no, everyone has to live to three years. We see the same answer. Early recurrences linked with higher risk of death. I think that's the important take home message here. Early recurrence, higher risk of death. Hazard ratios, again in that 2.5 ballpark. If we do a four-year landmark, we see the same thing. Hazard ratios again around 2.5 in that range. Early recurrence, higher risk of death. If we do five years, we see the same thing again, hazard ratios 2.2 to 2.5. Take home message here, early recurrence, higher risk of death, whether it's within one year, three years, four years, five years, early recurrence, higher risk of death. And so it really presents a critical, now unmet need to try and prevent these early recurrences.

Our study was limited by an availability of pertinent baseline characteristics, so we couldn't easily do multivariable analysis to adjust for things, so we're certainly not making any claims about surgery versus radiation. And again, take home, early recurrence, higher risk of death suggests maybe in future validated, this early recurrence may be a valuable endpoint in clinical trials to identify those high risk patients either for trial design or actually as an endpoint of study. Obviously that needs confirmation, but I think clearly early recurrence, higher risk of death, and now we just need better treatments to prevent those recurrences.

On that note, I'd like to thank you all for your attention as well as my co-authors for all their assistance in putting this together. So thank you.