Trends in PC DX and PC deaths, very puzzling numbers.

I took a look at the SEER data on from 1975 to 2011 and here's what is shows:
All data is in cases per 100,000 men.

1975 94 new cases, 31 deaths, 32% died

1992 237 new cases, 39.2 deaths, 16% died

2000 183 new cases, 30.4 deaths, 16% died

2011 139.9 new cases, 20.8 deaths, 15% died.

Here what I see from the data:

There was a significant rise in PC diagnosis in the early 90's and has been declining since then.
The death rate has remained very consistent among those DXed for the last 20 years.
The reduction in deaths is primarily due to the reduction in new PC cases.
In the era of PSA testing, why would the number of PC diagnosed go down?
In the era of better testing, treatments and early detection why has the death rate of those DXed not significantly changed?

Hopefully others will have better answers to this data and the questions raised by it than I do.

Robert,
I have no explanation for these trends. I looked at this because of MS WWs thread about the oncologist who saw no value in the psa test. At this level he seems to be correct.
This is a 1,000 ft view and it is always important to look at the big picture.
The decrease in PC cases is a long established trend and common sense says that this should have increased with psa testing.
The survival rate has stayed relatively constant and this also defies common sense with early detection and much improved treatments; this is very disturbing.
The first thought is that maybe we are just chasing our tails; doing a lot of work and not accomplishing very much in the overall scheme of things. Maybe we just can't affect the nature of this disease and some types of PC will kill us regardless of what we do in either early detection or treatment. I think it is also wise to question conventional thinking and look at the big picture from time to time in order to get a different perspective.

Jim,
I believe the numbers reflect the total of those DXed and the total deaths for that year.
Obviously the deaths would come some years after DX, but the death rate didn't really change from 1992 to 2011 so that would be more than enough time to offset any time differences fro DX to death.

JohnT,

I think the percentage died number is a bit misleading, because it was derived by dividing the number of men who died that year (and presumably been diagnosed with PCa in previous years) by the number of men diagnosed in that year. This is mixing apples with oranges.

I think you would do better focusing on the number of men who died in that year, which shows a clear decrease since 1992 ... evidence supporting the theory that PSA testing is decreasing the number of PCa related deaths.

In the early years (1992) weren't lower Gleason grades (ie. below gl 6) considered cancerous (I had thought I had heard that)? Could that be the reason for the higher number of diagnoses in 1992? Also, weren't there changes in the Gleason Method made in 2005 that may account for a decrease in the number of men diagnosed?

Post Edited (RCS) : 9/12/2014 5:20:18 PM (GMT-6)

John,

For me the problem is in the denominator (diagnoses) apparently the definition of what is considered cancerous has changed .... Similar to what is being proposed today .... Not considering a Gleason 6 as cancerous. I will admit that I am not positive of this change, but from the little I know/ remember it appears that what was considered cancerous in 1992 (high number) is not considered cancerous today (low number).

John,

I can not agree with the conclusions you are drawing until I am sure that the number of cancers diagnosed in each year are reported using the same basis (ie. that the earlier years are not overstated based on today's definition of PCa).

Also, I disagree on the method of calculating what you call cancer rate (in your first post you used % Died). As I stated in an earlier post, to divide the number of PCa deaths this year by the number of PCa diagnoses this year, is mixing apples and oranges. The number of deaths this year is not dependent on the number of diagnoses this year .... Especially if this rate uses a denominator whose basis may be changing with time.

Ps . I thought you were looking for possible explanations of why the numbers look funny ... It looks to me like you are trying to make an argument for not doing PSA testing ... Am I wrong?

Post Edited (RCS) : 9/12/2014 7:33:17 PM (GMT-6)

John,

I am pleased that you agree that the % Died calculation is not entirely valid.

The reason the % Died is staying the same is because the denominator (number diagnosed) is falling ... I think we both agree to that ... So why is it dropping (why do the numbers look funny .... That is your original question that you wanted answered?). What I have proposed to you is that maybe the number is getting smaller due to the definition of PCa being changed over time .... for example, where Gleason 5s included in the earlier definition of PCa making the earlier numbers look large compared to today's numbers? It would help explain the decrease, but I do not know if it is true.

I apologize if this does not fit your theories/conclusions of these numbers ... I only offer it because I thought you were looking for help understanding the numbers.

I guess I would need to look at the 20 years of data to agree with your slow and steady conclusion. The data you presented only has three data points for the 20 years ... with two of these points being end points (start and finish).

Well, I am headed off to bed ... early golf in the morning ... after waiting on repair men today, I plan on hammering the ball tomorrow ... will probably hit a bunch of worm burners ... Have a good night!

RCS said...
............. The number of deaths this year is not dependent on the number of diagnoses this year .... Especially if this rate uses a denominator whose basis may be changing with time.

Ps . I thought you were looking for possible explanations of why the numbers look funny ... It looks to me like you are trying to make an argument for not doing PSA testing ... Am I wrong?

These numbers are quite stunning. But the 1st thing I am trying to figure out is how the number being disgnosed could possibly be dropping each year and how the lower numbers dieing relates to that number. I think, RCS, what you say in the 1st line in above quote has something to do with it. Obviously(or so it seems to me), the number of men dying from PC this year has little to do with the number of men diagnosed this year. These deaths are almost completely from men diagnosed from a few to as much as 10-20 years earlier. I think we would all agree here that very few of us are going to die from PC in the year we are diagnosed with PC, or even a year or two after. That would be a tiny % of us that happens to. So what we are left with is: the absolute numbers dieing(in this country?) has dropped steadily since 1975. And has had a precipitous drop- about 30% fewer deaths- from 2000 thru 2011.

So do we believe that, even with vastly increased testing, that there are actually fewer men getting PC than there were in 2000 and 1992? The big jump from 75 to 92 makes total sense: PSA testing led to a huge increase in detection of PC before men could die from other causes or die from undiagnosed PC.

But then all of a sudden, fewer men getting PC every year? That is a hard pill to swallow. I do not quite understand all of this, but could it some how be that, once PSA testing became wide spread in the USA, that a vast pool of men between 40 and 90 who were already living with undiagnosed PC, were "suddenly" (that is over a 10 year period or so), diagnosed with PC? Leaving a smaller pool of NEW cases of PC that had not already been diagnosed, even though about the same number of new cases of PC were occurring every year?

So just say that since 1975, on average, 40,000 men get PC every year. PSA testing comes on line and for the next 10 years, there is a huge increase not in the rate of men getting PC but in the numbers diagnosed of PC. Many of whom, in years past, would have never been diagnosed before they died of a heart attack or stoke or colon cancer. But after that 10 years(or however long it takes), the majority of older men have had a PSA test and therefore diagnosed if they have it. Then after whatever period of time it takes for most men over 40 to be tested, the numbers of NEW diagnoses would start dropping. Not the number of men getting PC, but the number being diagnosed would drop.

If that is the case, and who knows, and the number of men getting PC is more or less the same as 10 or 20 years ago, then there has been a drop in deaths from 39.2 deaths per 100,000 in 1992 to year 2000 with 30.4 deaths and finally in 2011 "only" 20.8 deaths per 100,000. So the death rate, if the actual numbers of men getting PC is about the same, has dropped by 1/2 since 1992. And that is some very good news indeed.

I suppose the other thing that could be happening is that men are no longer getting PC near as often as they did even in 2000 or 1992? Hence the lower numbers of new cases being diagnosed each year? Seems unlikely to me, but it would still be good news, especially if the trend towards fewer men getting PC each year continues right down to zero!
Bill

EDIT: Oh, forgot: as has been mentioned, we also need to know if these were deaths of PC patients from all cause death, or deaths that were PC specific(probably is the latter). But if all cause death, the PC death rate might have dropped at an even more impressive rate.

Post Edited (BillyBob@388) : 9/13/2014 9:09:58 AM (GMT-6)

JohnT,

You are absolutely correct, there has been a slow steady decline since 1991-1992 in the number of diagnosed prostate cancers in the US. I could find no explanation as to why there is a decrease or why there was such a spike in 1991-1992.

My guess is that there was an environmental issue(s). In this time period a possible issue may have been new diagnoses based on Vietnam agent orange exposures and subsequent declines as veterans were diagnosed and exposures to agent orange were halted. Also, in the same time period EPA banned the use of several herbicides (one of which was used in agent orange). This ban could explain the slow decline in diagnoses as new exposures were halted, especially among agriculture workers and consumers of food.

These are just guesses on my part .... but that big spike in the early nineties and the long slow drop has to have some sort of rational reason.

Well of to the links.

John T said...
I took a look at the SEER data on from 1975 to 2011 and here's what is shows:
All data is in cases per 100,000 men.

1975 94 new cases, 31 deaths, 32% died

1992 237 new cases, 39.2 deaths, 16% died

2000 183 new cases, 30.4 deaths, 16% died

2011 139.9 new cases, 20.8 deaths, 15% died.

I'm thinking that in 1975, the higher death rate was because of the cancer being more advanced at the time of diagnosis. A cancer diagnosis was probably made after looking into the symptoms the patient reported. All too often, there are no symptoms until the cancer is too far advanced for successful treatment. And, of course, 40 years ago cancer treatment was not as refined as it is now.

In 1992, testing became available that made detection of cancer possible before symptoms were noticed as well as after symptoms were noticed. With early stage and greater cancer being detected, there is an upward bump in the number of cases showing up because this was the first time men were tested. Early stage cancer would probably not be detected before this time.

The death rate later stabilized because the men who were tested likely had their cancer detected and treated at a point in time that gave them a much better chance of stopping the cancer. On the down side, we know that cancer treatment isn't always successful and that some men are not tested at all or tested too late.

SpecialLady said...
To add another angle - when I talked to my father´s urologist, he told me (I forgot in which context now), that they are seeing many more agressive type of kidney cancer, which was rare before. I believe that the above numbers are almost impossible to interpret. Are we seeing more aggressive PC today than before? Are we seeing less aggressive cases today, but perhaps our bodies/environment have changed in such a way to allow cancer to march on? Are we seeing PC more in younger people today? (this is true, I believe) Is the death rate of aggressive PC in younger people offsetting the treatment benefit among 60-yr-olds? Is the fact that we are living longer contributing to the more-or-less constant death rate?

And it is true that some bad chemicals are off the market, but each year thousands of new ones are released into market, without much studies on their health or enivornment effect - chemical comapnies are not required to do this, almost all bans are retroactive. Some of those eventually find their way into the food chain, as was the case with BPA : baby bottles, cashier receipts, pizza boxes. And BPA changes prostates in mice, as it has a hormone-like structure. Now, all baby bottles say "BPA-free" but what is in them instead?

I agree with your points about the environment. I suspect there at least as many reasons for the numbers of new PC cases per year to increase, at least a bit, as there are for them to decrease. For example, just look at what ocean pollution had done to a food source that used to be considered a huge health booster. Now the warnings are to not eat too much of certain fish. Then there are all of the many new sources of estrogen exposure. Etc, Etc.

Which is why I doubt there is any significant decrease in the #s of men coming down with PC each year. Rather, as I stated above, there is a big decrease in the #s being diagnosed each year, simply because such a large part of the population has already been tested. Therefore a huge hunk of the older population of men with PC have already been diagnosed. about the only ones left to be diagnosed now are the brand new cases as the younger guys pass 40. As opposed to the huge pool of men between 40 and 90 who already had PC and did not know it, PSA testing becomes common and there is a huge spike in diagnosis. No one was getting PC at an increased rate, they just already had it and all of a sudden there is a huge increase of diagnosis of all of the guys, young and old, who already had it but did not know it.

So, the more I think about that, I am left wondering how you could NOT have a drop in new cases diagnosed after that? The numbers of men per 100,000 could stay about the same, but you would almost have a guaranteed drop in new cases being diagnosed.

But in the meantime, even if the numbers of guys newly getting PC remain about the same( and I have no reason to think it would not), the numbers of deaths per year drops steadily by 1/2. And I would not be surprised to find it has dropped even more since 2011. Hopefully anyway. That all sounds like good news to me. It also sounds like a good reason to keep doing PSA testing and trying to find better treatments.

Again though let me ask: Do Y'all actually believe that the numbers of men in the USA coming down with PC has actually dropped by 1/2 in 20 years? (as opposed to simply a decrease in diagnosis because so many men have already been diagnosed) Because if fewer men are actually getting PC, then it makes sense that fewer men would be dieing from PC, and all this treatment has been worse than useless, but actually harmful.

Which do Y'all think it is?

All those ideas are good reasons. The rates are per 100,000 age adjusted (don't know what age adjusted means). Someone from the cancer society said that in the 70s if you were ever Dxed with PC and later died, (maybe of something else) it was usual to put PC on the death certificate as a cause of death. This could explain the higher rates in the 70s and 80s.. Environmental reasons could also be in play as well as the intro of the psa test skewing the numbers in the 90s.