| Leonard Evens 2004-10-04, 2:21 am |
| George Conklin wrote:
> "Leonard Evens" <len@math.northwestern.edu> wrote in message
> news:FeidnUtXYe9FSsfcRVn-tw@comcast.com...
>
>
>
> What are you talking about? Men develop prostate cancer after they die?
Of course not. I explained what I meant further in my comments.
>
> I looked around the Web for the article, but it is not going to be there.
> There were about 6 different reports all saying the same thing as the one
> published here.
That is because they are all based on the same press release from Stamey
through Stanford.
>
>
>
>
> Let me point out that the only way to establish this, using
>
>
>
> So you are saying that the diagnosis is better at autopsy?
No, I definitely didn't say that. I said there is a difference between
a cancer detected by examining the entire prostate at autopsy and a
cancer discovered in a living man by current (or any plausible)
techniques. We can presume that if you just removed prostates from
living men and examined them in the same detail you would in an autopsy,
the results would be the same if the populations were closely matched.
In fact, that is an assumption. A stickler for precision might argue
that it is possible that men in their 20s who die of any cause may be
more or less likely to show prostate cancer on autopsy than would
prostates removed from healthy living men. But let's assume that isn't
a problem. To make sure it isn't you could restrict the autopsies to
men who were killed by accident or through external violence rather than
disease. Then the thing you might compare the autopsy data to is how
many men in an otherwise healthy population with no indications of
anything would show prostate cancer if you did biopsies on all of them.
But since no one has ever seriously recommended that---at any age,
certainly not for men in the 20s---a more appropriate comparison is if
you did PSA testing for some groups of men in the 20s matched as
carefully as possible to the men in the autopsy group, and then did
biopsies on those who had abnormal PSA levels. You could then compare
the detection rates done by these entirely different procedures. In
older men, it is known that it is something like 25 percent. There is
no reason to believe that in younger men it would be higher, and there
are a lot of reasons for believing it would be considerably less.
> So what!!
> Men in their 20s still have cancer of the prostate. They have it all their
> lives. It may even come and go. Why does this bother you? Men in their
> 60s, same thing. By 80 we all will have it.
Again you are mixing up two different things. First of all the autopsy
studies have shown rather different results. So it has not yet been
established how many men in their 80s would show evidence of prostate
cancer if they died and had autopsies. It is certainly not 100 percent.
That is just hyperbole. But I agree it is probably fairly high,
almost certainly well over 50 percent. However, as I've noted several
times before, and which you keep ignoring, only 16 percent of living men
in the US are diagnosed by current techniques with prostate cancer.
Your argument is that current techniques are diagnosing a lot of
prostate cancer than doesn't need aggressive treatment. There is a lot
of truth in that. And in fact a lot of prostate cancer in older men is
not treated except by watchful waiting and hormone therapy when
necessary. It is even true that many primary care doctors and
urologists recommend not doing PSA testing at all on men past a certain
age, particularly men past 80. It is important to determine by better
methods which men with longer lifespans need aggressive treatment, and
there is a lot of work going on to solve that problem. But you seem to
have concluded that no attempt should be made to treat early prostate
cancer at any age. This seems an article of faith for you, and your
arguments for it don't hold water.
In politics, simple minded reasoning seems to win the day. But science
is complicated and you have to make fine distinctions often to get it right.
> So? That is life.
>
>
>
>
>
>
> I am skeptical about procedures which are not evaluated for 100 years and
> I see your very emotional reactions as one reason why this happens. Maybe
> some result will emerge after 10 years.
I admit that I am emotional about some things. So are you. I
understand that having made a decision to treat my prostate cancer, I
may be biased in looking at the evidence to think positively about
evidence which seems to confirm my decision. But there is nothing all
that new about that. In my professional life, when trying to do
research, some ideas about the mathematics were more congenial
emotionally than others. It would have been nice if an idea which was
original with me proved to be very important so everyone would see how
bright and inspired I was. Unfortunately, often that wasn't the case,
so over the years I learned to be suspicious of my own beliefs and to
try to see the arguments against them. I learned to question
everything, especially my own work. So I think I am capable of
introspection in these matters and separating out what I would like to
believe from what I know is true. Are you sure you are?
>
> As the article said, men with lung cancer all die. Few men with prostate
> cancer do, and it looks as if a lot of men have all the way from the 20s on.
> Just facts.
You really have a problem with the precise use of scientific terms.
First, not all men with lung cancer die of it. My mother wasn't a man,
but she had lung cancer and she didn't die of it. She lived for 15
years after her diagnosis and surgery. There is a difference between a
high disease specific mortality rate and all people who have the disease
dying of it.
Second, it is not enough to just look at the number of men who are
diagnosed with prostate cancer who die of it (or the same for lung
cancer.) Prostate cancer is diagnosed in a large majority of cases in
older men who die of other things before they can die of prostate
cancer. Indeed if those men are old enough, they are generally not
treated aggressively for prostate cancer. You have to look at the
expected number of deaths from both diseases as a function of age at
diagnosis and take into consideration other health factors. If we lived
in a society where no one took up smoking until age 65, it is likely we
would also find that few people would be dying of lung cancer compared
to the number diagnosed with it.
Be all that as it may, I readily admit that lung cancer in most cases
is much more likely to kill people diagnosed with it than prostate
cancer. But you are ignoring another point here. Generally, in the
US, prostate cancer is treated, and the treatments may be effective. So
the lesser disease specific death rate among people diagnosed with
prostate cancer may be due, in part at least, to the difference in
treatment rather than the difference in the lethality of the disease in
typical patients. As far as I can see, it is not known currently how
much of the low relative death rate is due to each of the possible
factors: life expectancy at diagnosis, virulence or mildness of the
disease, effectiveness of treatment.
Note by the way that people are treated for lung cancer and sometimes
the treatment is successful, as in my mother's case. Stamey's argument
is actually more of an argument not to treat lung cancer than it is an
argument not to treat prostate cancer.
Prostate cancer for men in their 20s is a red herring, as I've shown.
Diagnosis of prostate cancer is very rare in men that age. Trying to
draw some conclusions about how much prostate cancer in older men needs
treatment from autopsy results in men in their 20s is pretty pointless.
Research is better directed in more fertile areas.
>
>
|