SLIDES & TRANSCRIPTS
Friday, December 5, 2003

Prostate Cancer

Otis Brawley, M.D.

Slide 1:

[No text is associated with this slide]

TOP

Slide 2:

Thank you, Elizabeth . It is wonderful to be here. I want to acknowledge Drs. Klein and Thompson and Linehan and thank them as well for inviting me to be here.

I am in 15 minutes going to talk about what Elizabeth is a professor in, and that is the epidemiology of prostate cancer. We are going to look at it with some international trends and do some comparisons.

Indeed the epidemiology of prostate cancer perhaps creates more questions than it does answers, but it does give us some information through which we can draw some conclusions and create some hypotheses.

This is a well-known slide from the early 1990s and unfortunately some of these countries only gather data every 10 to 15 years. So, this is about the most current data one can get find in terms of mortality for certain countries, but you can see the vast difference in mortality from Norway, Sweden, all the way through countries in the Far East.

This is probably evidence that environmental differences have something to do with prostate cancer mortality and because I am going to talk about differences in incidence later and how, whether one looks for it, one may find it interferes with incidence I would prefer to look at mortality.  

TOP
Slide 3:

Now, prior to the PSA screening era in the United States Japanese migrants to the United States were shown to have a marked increase in prostate cancer rates just within one generation of migration. This has actually been shown also for Polish immigrants to the United States as well and even people moving from Eastern Europe to Australia increased their rates of prostate cancer and rates of prostate cancer death simply by migration.

This probably indicates that there is something environmental, not inherent biology but something extrinsic in environmental that actually increases risk of prostate cancer

TOP
Slide 4:

and indeed Dr. Greenwald will talk to you about the prostate cancer prevention trial in a bit where one hypothesis is removal of some of these forces might decrease prostate cancer risk and perhaps prostate cancer death.

TOP
Slide 5:

Now, Armstrong and Doll years ago and this has been well talked about so I won't go into it tremendously have talked about differences in dietary fat amongst various countries and we are all very familiar with the hypothesis that increased amounts of dietary fat in the western world increase prostate cancer risk and indeed the previous talk talking about diet and especially pork brings up some interesting questions about prostate cancer.

TOP
Slide 6:

Now, the precise mechanisms through which diets high in fat might increase risk are not totally known and it has something to do most likely with increase in bioavailability of steroids in some way.

TOP
Slide 7:

Now, this is a slide that is well known to many people. This is the incidence for black and white Americans and indeed the black/white difference may very well be more due to environmental than to any kind of inherent genetic differences.

Indeed I prefer to use area of geographic origin instead of race as my labels for populations but the increase in incidence throughout the 1990s and this is incidence per 100,000 age adjusted to the new 2000 standard by the way, the increasing incidence is noted and then the decline and what is really called the Betty Ford effect in epidemiology. When Betty Ford announced that she had breast cancer many women went out and got screened and artificially raised the incidence of breast cancer for some time and then there was a decline.

This is the mortality rate, the number of deaths per 100,000 again age adjusted for blacks and for whites, and we are going to blow these up and look a little bit at these differences.

TOP
Slide 8:

This is to remind me to tell you that this is a disease of the elderly. The incidence and mortality rates are very low for people age 50 to 54 or 55 and below and the incidence rates rise dramatically for blacks here and then for whites. Mortality, also, rises dramatically as one gets into the seventies.

TOP
Slide 9:

Indeed in the United States the median age at diagnosis has gone down just a little bit in the PSA era. It was 72 in 1980 and 71, really 70.8 or so in 1995.

In Europe where there is very little PSA screening median age at diagnosis generally tends to be around 75 to 77 years of age.

TOP
Slide 10:

Now, here is the slide again and we wanted to point out that the

TOP
Slide 11:

increasing use of TURPs throughout the 1970s and 1980s actually caused much of this initial increase and increased screening caused additional increase later.

TOP
Slide 12:

Now, in terms of stage trends there is an increasing number of the proportion of men diagnosed with apparently localized prostate cancer and increasing number and proportion of men diagnosed with apparently, I am sorry, a decreasing number and proportion diagnosed with apparently distant disease.

TOP
Slide 13:

This is again rates per 100,000. This is to show that that incredible increasing rise throughout the 1990s was localized disease and then

TOP
Slide 14:

here we have distant disease, stage D disease and we have actually a decline starting in the 1990s.

These are actual rates, and these are actually proportions of men diagnosed. You see 70 percent of men diagnosed by 1995 had apparently localized disease.  

TOP
Slide 15:

I prefer to use the phrase "apparently localized disease" because a substantial proportion of men undergoing radical prostatectomy for apparently localized disease 2, 3, 4, 5 years later actually have a measurable PSA. So, they are categorized in our data sets as localized disease at diagnosis but they actually have apparently localized disease.

TOP
Slide 16:

Now, by looking and comparing populations and my favorite population to compare because their data is so good is the United Kingdom . Screening is much less common in the United Kingdom versus the United States

TOP
Slide 17:

and incidence among white Americans is much higher while mortality rates are very similar and this is actually thought to be proof of a concept in epidemiology called over diagnosis or the diagnosis of individuals who don't need to be treated because the disease is not a threat to them.

Here you have an incidence for white Americans in the United States from 1986 through 1999. Here you have the incidence in the United Kingdom for again white Brits and here you have the mortality rates and let us say that these mortality rates are essentially the same, and we are going to blow that up and look at it a little bit more in a bit;  

TOP
Slide 18:

Etzioni has suggested that perhaps the over diagnosis rate is 29 percent for white Americans and because of competing causes of death 44 percent for black Americans. The prostate cancer prevention trial suggests that the over diagnosis rate may be even higher.

TOP
Slide 19:

Ian Tannic recently published this paper in the Lancet saying that we have cured a symptomless disease, that being symptomless prostate cancer.

TOP
Slide 20:

Now, in the PCPT which you will hear more about later, I actually think one of the most important findings in the prostate cancer prevention trial is not about finasteride in prostate cancer prevention. I am actually very interested in the findings of screening in the people in the placebo arm. A large group of men, nearly 9000 with a PSA less than 3, median age of 62 were followed for 7 years until the median age of 69.

This means that this was a group of men at relatively low risk for prostate cancer who were given rigorous, very rigorous, more rigorous than the norm in the United States PSA and DRE screening for 7 years with all men ultimately biopsied at the end of the study.

As many of you know 24 percent of men in the placebo arm were found to have prostate cancer versus 18.4 percent in the intervention arm.

TOP

Slide 21:

Now, 12.2 percent of men were diagnosed with prostate cancer due to an abnormal PSA or DRE during the 7 years of screening and 12.2 percent were diagnosed at the end of the trial due to a biopsy after having normal PSA and DRE for 7 years.

It is important to realize whether one focuses on the 12.2 percent from PSA and DRE screening or the 24-plus percent overall that the NCI would predict that for men in their sixties who have not been diagnosed with prostate cancer they have a 4 percent chance of ever dying from prostate cancer.

So, this suggests that the over diagnosis rate for prostate cancer screening could be at least two out of every three men don't benefit from the diagnosis to perhaps as high as five out of every six or seven out of every eight.

TOP
Slide 22:

Now, to blow up this mortality figure that we looked at earlier, this is from 1975 through the year 2000, and this is again rate per 100,000, death rate for blacks in pink, death rate for whites in the United States in black. Please note that there was this increasing rise in the death rate in both groups throughout the 1980s and an increasing rate then a little unsettling and still not exactly clear why it is. It might be an attribution bias. It might be something that is actually a real finding.

TOP
Slide 23:

This is a comparison of the mortality rate in the United States in boxes versus the mortality rate in the United Kingdom in diamonds, and this is only for men age 50 years and over. It differs a little bit from the slide that we showed you before and that was all men, and this is men 50 years of age and older.

You will note that there was a rise in the number of deaths among men in Britain over the age of 50 throughout the 1970s and into the 1980s and then of course you have essential parity in terms of deaths.

It is also important to note that the number of radical prostatectomies done in the United States is approximately 60 to 80 times the number done in Great Britain in the year 2000.

TOP
Slide 24:

The increase in British mortality has been largely attributed to an increase in older men and I will show you in just a second, and this might actually be due to improvement in treatment of other disease, lowering cardiac mortality such that men in their seventies and eighties can actually live to die of prostate cancer, and

TOP
Slide 25:

here you see the increase in mortality rate for men in their seventies, something that is relatively flat for men in their fifties and sixties.

TOP
Slide 26:

Here it is actually broken down even more to show that the rise in death rate from prostate cancer for men 85 and above was much more significant for men in their early eighties.

TOP
Slide 27:

This is perhaps one of the most important papers and one of the most important observations that has been done in prostate cancer. It is an observational paper from Tyrol and Georg Bartsch is here in the audience

TOP
Slide 28:

and he demonstrated when he went back to Tyrol in 1990 a small area in Western Austria and implemented really all things American, screening, aggressive treatment and actually very good high-quality treatment. I cannot over stress that because in the United States as a whole, unfortunately treatment is not still as rigorously controlled as it is in Western Austria . He saw a decrease in the number of men who had distant disease from 1988 through 1998, an increase in the number of men who actually had stage 1 and stage 2 or localized disease,

TOP
Slide 29:

so, the positive effects of screening that we normally like to see in terms of stage migration and change in stage, and these are actually the mortality figures for Austria as a whole in black and the mortality figures for Tyrol, the State of Tyrol in white, and we have a decline in mortality again with the implementation of everything American.

TOP
Slide 30:

Now, the Tyrol observation I really want to stress may show us that screening saves lives. It also may show us that rigorous treatment, rigorous and aggressive treatment saves lives that is the use of hormones as well as localized treatment perhaps without the use of screening.

TOP
Slide 31:

Now, the increasing proportions of patients receiving aggressive local therapy in the United States is a dramatic change and the increasing use of hormonal therapy early in the course of the disease is a dramatic change over the last 20 years.

All of these may be part of the reason for the decline in mortality and

TOP
Slide 32:

can the decline in mortality be seen without screening and its inherent over diagnosis? That is an important question.

Epidemiology legitimizes that question but really cannot answer it at this point. Given the current status of data and without a randomized clinical trial having been completed a small advantage to screening for prostate cancer actually cannot be excluded right now but truly there are implications that might exist but it really cannot be excluded or truly affirmatively included and what we have really got from epidemiology is clear evidence that we really truly need some kind of mathematical model that might be based on genetics and population work or proteomics and population work, life style and population work that actually would allow us to say, "Sir, you have prostate cancer, but you are in that six out of seven that don't need to be treated and need to be observed," or "Sir, you have prostate cancer, and you are in that minority where this prostate cancer is actually dangerous to your health,"

TOP
Slide 33:

and so with that I am going to conclude.

Thank you very much. It has been a pleasure to talk to you and it is a pleasure to be here.


TOP