Read the Full Video Transcript

Andrea Miyahira: Hi everyone. Welcome to another UroToday episode on Translational Prostate Cancer Research. I'm Andrea Miyahira here at the Prostate Cancer Foundation. With me today is Dr. June Chan, a professor and Dr. Rebecca Graff, assistant professor, both at UCSF. We'll be discussing two of the group's recent papers, one published last year and this year, both in the British Journal of Cancer titled, "Post Diagnostic Health Behavior Scores and Risk of Prostate Cancer Progression and Mortality." And, "Post Diagnostic Health Behavior Scores in Relation to Fatal Prostate Cancer." Thank you both for joining me today.

Rebecca Graff: Thank you so much for having us here today. I'm going to kick us off and then pass us over to Dr. Chan about midway through this slide deck. And as Dr. Miyahira mentioned, we'll be talking about post diagnostic health behavior scores and risk of prostate cancer progression and mortality. As published in a couple of different articles in the British Journal of Cancer, in the past year or so. Many health behaviors among cancer-free men or pre-diagnostic health behaviors have been evaluated in relation to the incidence of both overall and lethal prostate cancer. So smoking and obesity, for example, have been associated with an increased risk of lethal prostate cancer. While increasing physical activity is associated with a decreased risk. In the pre-diagnostic setting, a 2015 study from Stacy Kenfield found that a score comprised of the first six behaviors in this table, from smoking through processed meat, all oriented toward more points for healthier practices was associated with a substantially reduced risk of incident lethal prostate cancer among cancer-free men.

Among men who've already been diagnosed with prostate cancer, post diagnostic health behaviors have also been associated with prostate outcome. But before our research, prior studies hadn't assessed the impact of a combination of post diagnostic health behaviors. But before getting into prostate cancer specific behaviors, I'll quickly address recommendations for cancer more generally. The World Cancer Research Fund and American Institute for Cancer Research collaborated on recommendations for cancer prevention in 2018. Those are the recommendations on the right of the slide. And then six years before in 2012, the American Cancer Society developed nutrition and physical activity guidelines for cancer survivorship.

Both sets of guidelines recommend maintenance of a healthy weight, regular physical activity, and a diet rich with fruits, vegetables, and whole grains. And the WCRF AICR guidelines take the dietary recommendations further, addressing limited consumption of processed foods, red and processed meats, sugar sweetened beverages and alcohol. But whether practices consistent with recommendations such as those from these national organizations, are associated with better outcomes among men with prostate cancer specifically, was previously unknown before our research.

So based on the prior literature and national recommendations, we created six scores comprised of multiple factors that allow for assessment of the combined impact of health behaviors. In 2021, we conducted a review of the literature on post diagnostic behaviors and prostate cancer progression and prostate cancer specific mortality. And we identified three behaviors, smoking, physical activity, and body mass index that had the strongest existing evidence of associations with outcomes among men who have the disease. And then we also created a version of the 2021 score that added four dietary factors that have demonstrated some evidence of an association with progression or mortality. The 2015 score refers to the six factor score that Dr. Kenfield developed in the context of primary prevention of lethal prostate cancer among cancer-free men. But for these analyses, we evaluated the score in the post diagnostic setting among men with prostate cancer.

The operationalization of the ACS survivorship recommendations was based on the principles of Marjorie McCullough and colleagues, and we also created a version of the score with and without alcohol. And the operationalization of the WCRF AICR prevention recommendations was based on published guidelines from the NCI. We studied the relationships between the scores and prostate cancer outcomes in two cohorts comprised of men with localized prostate cancer. In the health professionals follow-up study, we evaluated prostate cancer specific mortality among roughly 4,500 participants, among whom 219 died from the disease.

But there were fewer prostate cancer specific mortality events in the captured diet and lifestyle substudy. And so we also looked at a composite outcome of progression, which was comprised of biochemical recurrence, secondary treatment, bone mets or prostate cancer specific mortality. And for those analyses of progression, we implemented parametric survival models with the Weibull distribution to accommodate interval censoring associated with uncertainty in the actual date of prostate cancer progression. And then for analyses of prostate cancer specific mortality in both cohorts, we implemented Cox proportional hazard model. And Dr. Chan will take us into the results.

June Chan: Great. Thank you so much, Rebecca, for providing that background on the rationale and the methods. And so what is being shown right now here on the screen are the results or at least one version of the results from the health professionals follow-up study. Just to help orient people on what's being shown on the screen these are the multi-variable hazard ratios with the 95% confidence interval bars associated with each of the scores that Dr. Graff just went over. An association of 1.0 means no association. If you're above the 1.0 line, that would indicate an increased risk, whereas if you're below the line, that would indicate a decreased risk.

So what you can see here for each of the scores or for most of the scores, there's mainly a null association with the exception of the 2015 score. I don't know, Rebecca, if you want to highlight that with the pointer or something. But the 2015 score there was associated with approximately a 19% reduction in prostate cancer mortality for each one point increase in the score. So what all of the figures here are showing is what is the association for a one point change in that specific score. So it was null for five of the scores, but inversely associated for the 2015 score.

Okay, so next we are moving on to the capture results. In this analysis, as Dr. Graff indicated, we didn't really have as many prostate cancer specific death events because the study has had less follow-up over time. So we mainly looked at prostate cancer progression, as she described, on this graphic. It's the same kind of setup as on the previous slide. One is a null association. If you're below one, that means there's an inverse association. On this graph the gray bars, the lighter color bars are for the outcome of prostate cancer progression, whereas the black or darker colored bars are for prostate cancer death. So as you can see here, a general trend was seen as for an inverse association for all the scores and for both outcomes, prostate cancer progression as well as prostate cancer death.

It was not statistically significant for every single one of the scores. As you can see, it was statistically significant mainly for the progression outcome for the 2021 plus diet and the ACS plus alcohol and borderline there for the 2015 score. What is interesting, but we would interpret cautiously is that the results for the scores and the outcome of prostate cancer death were also quite inversely associated in some cases even suggestively stronger. But again, we interpret this cautiously since again, this population only observed about 73 prostate cancer specific death outcomes.

So a couple of things to keep in mind when interpreting these data. Both populations, the health professionals follow-up study and the capture population are comprised mainly of Caucasian men who tended to be more highly educated. And as Dr. Graff indicated, they are all lower prognostic risk people by the way we selected the study population, we focused on people who were T3A or below. As you could see from the graphs, there were some differences in associations for the progression versus the prostate cancer specific death outcome. One interpretation or one caveat we can glean for that is that maybe there really are different mechanisms that drive recurrence versus death. There's also some differences between the two studies with regards to the actual samples themselves when exposure data were collected. And then as mentioned, there was a difference in the outcome focus. So that could account for some differences in the results between the two studies.

So trying to step back and put these together. If you look at both of these paper, I think one of the overarching key takeaways is that not smoking, regular exercise, and maintaining a healthy body weight likely impart benefit for men with prostate cancer in deterring both prostate cancer progression and prostate cancer death. Those three factors are pretty consistent across the scores. So whenever we see an inverse association, even though there's variability from one score to the next, those three things are pretty consistent. If you were to look at the individual results from the two papers separately from the capture paper, we would say that the 2021 score plus diet was associated with a 24% lower risk of prostate cancer progression and a 35% lower risk of prostate cancer death.

And from the health professional study, as mentioned, there was a 19% lower risk of prostate cancer death. And again, all those metrics are for a one unit increase in the score. So overall, we would say that men with non-metastatic prostate cancer may potentially improve their survivorship by following the general recommendations of not smoking, getting regular exercise, maintaining a healthy body weight, and then perhaps also following specific dietary recommendations. Even though we would say those results were a little less conclusive. But overall modifiable health habits could improve these clinical outcomes in these men with prostate cancer.

So lastly, it may be hard to read on the slide, but we did try to acknowledge all our co-authors from both the papers as well as some other colleagues who contributed to the mentoring of some of the trainees who worked on these studies. And then of course the many different funding sources and all the participants. And also acknowledge of course, the Prostate Cancer Foundation for their support of ourselves and our work overall.

Andrea Miyahira: Thank you so much, Dr. Graff and Dr. Chan for sharing that. So some follow-up questions, your newest score, the 2021 plus diet along with the WCRF and AICR score showed significant associations with reduction in prostate cancer, specific mortality, but the other scores did not. All of these scores measured lifestyle and dietary factors differently. So what do you think are the specific major factors that contribute to reduction in prostate cancer, specific mortality? And this is in the post-diagnosis space.

Rebecca Graff: I'll take a first stab at that question, which really aligns with the conclusions that Dr. Chan just mentioned, which is that BMI, physical activity in particular were shared by all the scores that showed significant inverse associations. And so I think it's fairly safe to say that those probably have an association with prostate cancer outcomes. The ACS and WCRF AICR scores didn't include not smoking, not because they don't think that that's important, but because they were focused more on dietary and other lifestyle behaviors.

But I think not smoking is also pretty safe to say, is an important part of a healthy lifestyle following a prostate cancer diagnosis. We're a little bit more hesitant to make conclusions about dietary factors with respect to prostate cancer specific mortality. The reason being that the scores included different factors and we didn't see super consistent results across the score. The possible exception is that red and processed meat were in all of the scores for which we saw significant inverse associations, and so perhaps there could be something there. But certainly there are other dietary factors that could potentially be associated with worse or better outcomes.

Andrea Miyahira: Okay, thank you. So then overall, what are your recommendations for patients and what should doctors be telling their patients? And then also, how would your recommendations differ for individuals without cancer versus those who've already been diagnosed with prostate cancer?

June Chan: I'll take that one. So actually, just sort of piggybacking on what Dr. Graff just shared, just would like to underscore that the data, there's more opportunities to study. There have been in the literature more opportunities to study factors such as body size, body metrics, and physical activity than diet. So first of all, there's just not as much data or not as many studies that have collected really detailed data on diet. So that's an area that we're working on. So sometimes we also have to keep in mind that part of the reason why I think we feel like we can stand behind the recommendations on exercise and body size is because there's just more literature there. We're really hopeful that there'll be more studies in the future where we can really delve into the dietary questions more. But it's quite a magnitude of scale.

If you look at one of the review articles, I think Dr. Graff mentioned, there've been dozens of papers that look at BMI. Probably similar, maybe slightly less for physical activity, probably a similar amount. But for diet, we're talking small handful, most of them from these two cohorts. And so we are trying very hard to find opportunities to look at that. But that's just so people understand when we say that these are stronger and more consistent, it's also there's just been more work done on those two areas. So, be interesting to see what the future brings. The results for physical activity are also... I'm aware of increasing amounts of data that have just corroborated the findings from these two cohorts about exercise. So I just wanted to point that out. It has been shown in probably like five other distinct populations, observational data of the benefits of exercise on prostate cancer death.

So then with that bigger backdrop, when you talk about what are the recommendations for patients or the doctors, I would say with what Dr. Graff just said. And to the question about does that differ for people who've been diagnosed with cancer or not, I'm going to ask if I can, Dr. Graff, to pull back up one more slide. So here in this graphic, which our team actually made for a publication that was recently in European Urology Focus, this is a summary of dietary recommendations from other organizations around primary prevention for heart disease, primary prevention for cancer, that's the middle column.

And then what we have done is added on what one might recommend for prevention of lethal prostate cancer in prostate cancer survivors. And so this is one summary just again, focused on diet. The little graphics on the side are supposed to indicate that of course, this is after we're already recommending don't smoke, get physical activity, and maintain a healthy body weight. That's what... So this is really just focused about the diet because those three are consistent across all of the three conditions on this slide already.

What this graphic is trying to illustrate though is even among the dietary recommendations, there's a good amount of overlap. One of the things I would want to point out is that heart disease remains the leading cause of death in general in the population and among men with prostate cancer is also a leading cause of death. And so there was a study done using prostate cancer registry data from Sweden and the United States. It is about 10 years ago, so please keep that in mind. But in that study, it was estimated that about 16% of US men diagnosed with prostate cancer would die from their disease. Which one can then take away that's where you get the statement that you hear that most men diagnosed with prostate cancer will die of something else. Of that something else, the top thing is heart disease.

So I think it's appropriate if you're going to ask what should be recommended, what should doctors take away, what should patients think about? That's why we start on this graphic here with heart disease. And so those are all the dietary factors. The green checkbox means those things are recommended. The Xs are supposed to indicate those things are not recommended. Hoping people will take away that between heart disease and cancer, there's quite a number of similarities. And then for lethal prostate cancer, what we've done is just added a couple nuances of interpretation. I think the most notable one might actually be... And I just realized this is a one slide outdated.

But on alcohol, to clarify, no one's recommending that you start drinking alcohol. But it's if someone is already drinking alcohol for heart disease, it says to limit to one to two drinks per day. I would just, for cancer, general cancer, primary prevention, you would not recommend any alcohol at all. For lethal prostate cancer there's still quite a bit of research going on in this area. Some studies have actually suggested possible protective effects, but we would interpret those to say, "If you are already drinking alcohol, try to limit to three to five drinks per week." But I did want to point out that those two limits are different. So it's more restrictive in the prostate cancer survivorship space.

Andrea Miyahira: Thank you for that. And how do you think other factors such as genetic ancestry, race, or comorbidities could impact the effects of lifestyle on risks for prostate cancer progression and mortality?

Rebecca Graff: I'll take another turn at trying to answer, though this one is a bit tough because as Dr. Chan said, there are limited numbers of populations in which these questions can be studied, and some of them don't have all of the data that we might want in order to assess those interactions. But in the prediagnostic setting, Anna Plim at Harvard recently published a paper in European urology that showed that even among men at the highest genetic risk for prostate cancer, that healthier lifestyle as defined by the factors in the 2015 score that we talked about, was associated with close to half the risk of prostate cancer that went on to be lethal relative to men with an unhealthy lifestyle.

So basically that's suggesting that even if you're at a high genetic risk for prostate cancer in the first place, that there are ways that you can potentially undermine that risk by engaging with healthy lifestyle practices. How other factors interact with lifestyle in the post diagnostic setting is less well established. But I don't think that there's any reason that lifestyle wouldn't still be important, even in the context of other factors at play.

Andrea Miyahira: Thank you. And what are the next steps for your studies?

June Chan: I'll take that one. So thanks for that question. I think I'd just like folks to be aware, we have a couple randomized clinical trials that are ongoing. One of them is the prostate A2 randomized controlled trial, which is led by our close colleague, Dr. Stacy Kenfield. And in that study, that's a forearm trial, in that study we're all examining different combinations of diet and or exercise intervention versus the control group to examine effects on PSA recurrence and other prostate cancer biomarkers. So the study, I believe, fingers crossed, is almost complete with enrollment, literally the last nudge across the finish line in the next few weeks. And so as sort of a preview, hopefully we're looking forward to having data, hopefully to analyze in about another year. Because once you finish enrollment, we still need to have a minimum of 12 months follow up, I believe, for that trial.

So that's ahead of us. And then the other one, actually, again, a shout-out to Dr. Kenfield. She is leading our team on an analysis of looking at very similar questions to looking at scores or certain dietary practices within the multi-ethnic cohort study. And we're doing that in collaboration with co-investigators of that large cohort. And then Dr. Graff and myself, we're also looking at studies related to metabolomics. Currently, most of our data looking at metabolomics is in the exercise space, but we're investigating various biomarkers, other sort of plasma metabolomic indicators to try to understand what are the underlying biological pathways that are being triggered by these health habits. As I mentioned, right now we mostly have data in the exercise space. We're looking forward to hopefully also successfully applying for funding so that we can analyze it from some of our studies, but we'll have dietary data as well.

Andrea Miyahira: Okay. Awesome. Well, congratulations on all of this work, and thank you so much, Dr. Graff and Dr. Chan for joining me today and sharing this.

June Chan: Thanks again for having us.