Scientists Uncover Cancer "Road Map" 
Author Message
 Scientists Uncover Cancer "Road Map"

Scientists uncover cancer road map

Study identifies important step in how cells turn cancerous

        LONDON, July 28 Scientists Wednesday said they had unlocked the
secret of how normal human cells become cancerous in a breakthrough that could
help the search for a cure.
        UNTIL NOW, the human cancer cell has been a black box with unknown
number of regulatory changes. Now we have been able to catalog the number of
changes with precision, said Robert Weinberg, of the Whitehead Institute at
the Massachusetts Institute of Technology.
       Weinberg told Reuters in a telephone interview he and his colleagues
were able to turn normal cells into cancer cells in the laboratory.
       What (the scientists) show is that in order to get a tumor cell, there
are four steps. If we can somehow rebuild one of these steps in the tumor cell,
then we may stop the evolution of the tumor cell, Moshe Yaniv of the Institut
Pasteur in Paris told Reuters.
       The research was published in Wednesdays issue of the science journal
Nature and was reviewed by Yaniv and his colleague Jonathan Weitzman.
       Weinberg and his colleagues found that four regulatory changes or
disruptions in the normal growth of cells need to take place for a cell to
become cancerous.
       Scientists have been trying for 15 years to turn normal human cells into
tumor cells in the laboratory.
       They set out to determine the minimum number of defined genetic events
required for tumor formation, and have brought an end to an endeavor that began
more than 15 years ago, Yaniv and Weitzman wrote in Nature.

----------------

Andre



Sun, 13 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"
From Nature:

Creation of human tumour cells with defined genetic elements

WILLIAM C. HAHN, CHRISTOPHER M. COUNTER, ANTE S. LUNDBERG, RODERICK L.
BEIJERSBERGEN, MARY W. BROOKS & ROBERT A. WEINBERG

During malignant transformation, cancer cells acquire genetic mutations that
override the normal mechanisms controlling cellular proliferation. Primary
rodent cells are efficiently converted into tumorigenic cells by the
coexpression of cooperating oncogenes. However, similar experiments with human
cells have consistently failed to yield tumorigenic transformants, indicating a
fundamental difference in the biology of human and rodent cells. The few
reported successes in the creation of human tumour cells have depended on the
use of chemical or physical agents to achieve immortalization, the selection of
rare, spontaneously arising immortalized cells, or the use of an entire viral
genome. We show here that the ectopic expression of the telomerase catalytic
subunit (hTERT) in combination with two oncogenes (the simian virus 40 large-T
oncoprotein and an oncogenic allele of H-ras) results in direct tumorigenic
conversion of normal human epithelial and fibroblast cells. These results
demonstrate that disruption of the intracellular pathways regulated by large-T,
oncogenic ras and telomerase suffices to create a human tumor cell.

--------------
end of first paragraph.  Unfortunately, I currently don't have access to a
medical library, so those who are interested can go to their local medical
libraries, and read the rest of the article.

Andre



Mon, 14 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"

Quote:

>Scientists uncover cancer road map

>Study identifies important step in how cells turn cancerous

>        LONDON, July 28 Scientists Wednesday said they had unlocked the
>secret of how normal human cells become cancerous in a breakthrough that could
>help the search for a cure.
>       Weinberg told Reuters in a telephone interview he and his colleagues
>were able to turn normal cells into cancer cells in the laboratory.
>       What (the scientists) show is that in order to get a tumor cell, there
>are four steps. If we can somehow rebuild one of these steps in the tumor cell,
>then we may stop the evolution of the tumor cell, Moshe Yaniv of the Institut
>Pasteur in Paris told Reuters.

According to Weinberg in a PBS interview Wednesday evening, what they did
caused cells to become 'cancerous', but did NOT yet cause metastasis to
occur.  He said that research is continuing to identify the additional
changes needed to explain this phenomenon.

I would find it very interesting to see an analysis of the probability of
not only the four steps identified to date, but of the additional ones
needed to cause metastasis and any other as yet unexplained aspects of
cancer, based on the assumption that these are random, independent
mutations.  Such a calculation would IMO shed light on the issue of
whether or not some mechanism that negates the independence assumption,
such as a cancer-causing organism, is necessary to make the probability
calculation consistent with the observed frequency of occurrence of cancer.

-John S.,
 Wellesley Hills, MA



Tue, 15 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"

Quote:



> >Scientists uncover cancer road map

> >Study identifies important step in how cells turn cancerous

> >        LONDON, July 28 Scientists Wednesday said they had unlocked the
> >secret of how normal human cells become cancerous in a breakthrough
that could
> >help the search for a cure.
> >       Weinberg told Reuters in a telephone interview he and his colleagues
> >were able to turn normal cells into cancer cells in the laboratory.
> >       What (the scientists) show is that in order to get a tumor cell, there
> >are four steps. If we can somehow rebuild one of these steps in the
tumor cell,
> >then we may stop the evolution of the tumor cell, Moshe Yaniv of the Institut
> >Pasteur in Paris told Reuters.

> According to Weinberg in a PBS interview Wednesday evening, what they did
> caused cells to become 'cancerous', but did NOT yet cause metastasis to
> occur.  He said that research is continuing to identify the additional
> changes needed to explain this phenomenon.

c-Ras is known to be involved in the regulation of a number of genes that
are important for metastasis.  The important thing to remember about
metastasis is, that it is an excess of metastasis promoters (e.g. matrix
metalloproteinases) to their inhibitors that determines metastasis.  Other
proto-oncogenes known to be involved in regulating "metastasis genes" are
c-fos, c-jun, and c-ets, which are downstream of c-ras.  Interestingly,
these genes upregulate production of both.  It's how the metastasis
promoter-inhibitor regulation becomes uncoupled, that determines
metastatic behavior.

Quote:
> I would find it very interesting to see an analysis of the probability of
> not only the four steps identified to date, but of the additional ones
> needed to cause metastasis and any other as yet unexplained aspects of
> cancer, based on the assumption that these are random, independent
> mutations.

Thing is, they're not necessarily independent.  This study showed
essentially that for human cancers, the change in ras, p53 (due to
influence of the SV40 large T antigen) and telomerase expression by
themselves are not sufficient to induce a malignant phenotype, whereas
together they will.  They can occur independently.  However, we know from
genetic studies of patients with Li-Fraumeni syndrome, who are born with 2
inactive copies of p53, that these patients are already at higher risk for
developing cancer.  It is similar with p53 knockout mice.  These mice
develop normally, but are at extreme risk of developing tumors in early
{*filter*}hood and later.  Together these two suggest that p53, or another
similar situation such as inactivation of Rb (as seen in familial
retinoblastoma and some osteogenic sarcomas), must already be present to
initiate the process.  It also suggests that p53 inactivation allows other
mutations to occur at a higher rate than normal.  This makes sense given
p53's role in controlling the cell cycle.  The question becomes if p53
mutation or inactivation must indeed occur first.  Likely not, as other
familial cancer syndromes have mutations in different genes.  It's the
development of genetic instability that is key--but it may not matter how
that instability comes about.

Quote:
> Such a calculation would IMO shed light on the issue of
> whether or not some mechanism that negates the independence assumption,
> such as a cancer-causing organism, is necessary to make the probability
> calculation consistent with the observed frequency of occurrence of cancer.

Some cancers are known to be associated with specific viruses, but this is
not the case with all cancers.  Viruses such as HPV, are known to
inactivate p53 and/or Rb.  And even if one is infected with such a virus,
there is no guarantee that he/she will develop its associated cancer.
Tell me John, how you would reconcile the "cancer organism" theory to the
HNPCC (hereditary non-polyposis colon cancer) syndrome, where the genetic
defect is in a specific DNA repair enzyme, coupled to the normally high
rate of replication of colon epithelium cells?

T.

--
T.D. Laing

Remove "nospam" from my e-mail address in the header to reply.

My opinions are my own and do not reflect those of anyone else's.  They'd probably prefer it that way.



Tue, 15 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"
Thanks for your comments.


Quote:

>c-Ras is known to be involved in the regulation of a number of genes that
>are important for metastasis.  The important thing to remember about
>metastasis is, that it is an excess of metastasis promoters (e.g. matrix
>metalloproteinases) to their inhibitors that determines metastasis.  Other
>proto-oncogenes known to be involved in regulating "metastasis genes" are
>c-fos, c-jun, and c-ets, which are downstream of c-ras.  Interestingly,
>these genes upregulate production of both.  It's how the metastasis
>promoter-inhibitor regulation becomes uncoupled, that determines
>metastatic behavior.

When you say 'downstream', does this mean that a single defect in this
c-Ras critter can activate all of the others?

Quote:
>> I would find it very interesting to see an analysis of the probability of
>> not only the four steps identified to date, but of the additional ones
>> needed to cause metastasis and any other as yet unexplained aspects of
>> cancer, based on the assumption that these are random, independent
>> mutations.

>Thing is, they're not necessarily independent.  This study showed
>essentially that for human cancers, the change in ras, p53 (due to
>influence of the SV40 large T antigen) and telomerase expression by
>themselves are not sufficient to induce a malignant phenotype, whereas
>together they will.  They can occur independently.  However, we know from
>genetic studies of patients with Li-Fraumeni syndrome, who are born with 2
>inactive copies of p53, that these patients are already at higher risk for
>developing cancer.  It is similar with p53 knockout mice.  These mice
>develop normally, but are at extreme risk of developing tumors in early
>{*filter*}hood and later.  Together these two suggest that p53, or another
>similar situation such as inactivation of Rb (as seen in familial
>retinoblastoma and some osteogenic sarcomas), must already be present to
>initiate the process.  It also suggests that p53 inactivation allows other
>mutations to occur at a higher rate than normal.  This makes sense given
>p53's role in controlling the cell cycle.  The question becomes if p53
>mutation or inactivation must indeed occur first.  Likely not, as other
>familial cancer syndromes have mutations in different genes.  It's the
>development of genetic instability that is key--but it may not matter how
>that instability comes about.

I can see how a genetic defect which is present throughout the individual's
body could predispose to cancer.  And of course more than one of the
needed mutations would make the probability even higher.  The question is,
would chance mutations of all the other needed sites happen with a probability
that would account for what is observed?  The independence assumption of
course means that the probabilities all multiply, and since they are all
less than one, the product could be quite small.  Probability here of
course is really probability per unit time, and the question is, would
the numbers show anything like the right probability for a cancerous
genome to develop in even one cell, in a period of time comparable to a
human lifetime, or would it take more like the age of the universe for it
to occur under an independence assumption.

Of course mechanisms which couple the mutations in some way could
alter the probaiblity calculation considerably.  An 'organism' that
does carefully selected edits for its own amu{*filter*}t would be one such
mechanism, but any plausible mechanism would do.

Quote:
>Some cancers are known to be associated with specific viruses, but this is
>not the case with all cancers.  Viruses such as HPV, are known to
>inactivate p53 and/or Rb.  And even if one is infected with such a virus,
>there is no guarantee that he/she will develop its associated cancer.

Right.  Some viruses could do edits on genetic material for their own
amu{*filter*}t or benefit, but they might not be the complete set of alterations
needed to produce cancer.

Other viruses might have a different game plan in mind, and there could
conceivably be one virus that does it all.  With the right strategy, it
might even defeat the normal immune system method for detecting the presence
of a virus, the display of 'self' and 'nonself' proteins on the cell's
surface.  

Quote:
>Tell me John, how you would reconcile the "cancer organism" theory to the
>HNPCC (hereditary non-polyposis colon cancer) syndrome, where the genetic
>defect is in a specific DNA repair enzyme, coupled to the normally high
>rate of replication of colon epithelium cells? >T.

Dunno.  I never thought about it.  Is this one genetic defect sufficient
to cause cancer in this case, or are several others also required?
If it's the former, you have your answer already.  If it's the latter,
then we come back to the probability calculation.  Are the required additional
alterations wildly improbable, given an independence assumption, or not?

-John S.



Wed, 16 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"

Quote:

> Thanks for your comments.



> >c-Ras is known to be involved in the regulation of a number of genes that
> >are important for metastasis.  The important thing to remember about
> >metastasis is, that it is an excess of metastasis promoters (e.g. matrix
> >metalloproteinases) to their inhibitors that determines metastasis.  Other
> >proto-oncogenes known to be involved in regulating "metastasis genes" are
> >c-fos, c-jun, and c-ets, which are downstream of c-ras.  Interestingly,
> >these genes upregulate production of both.  It's how the metastasis
> >promoter-inhibitor regulation becomes uncoupled, that determines
> >metastatic behavior.

> When you say 'downstream', does this mean that a single defect in this
> c-Ras critter can activate all of the others?

An overactive Ras protein will indeed influence everything downstream of
it.  It is an intermediate in a signalling cascade that leads to cell
proliferation among other responses; its persistent activation decouples
the cascade from the upstream regulation.

Quote:
> I can see how a genetic defect which is present throughout the individual's
> body could predispose to cancer.  And of course more than one of the
> needed mutations would make the probability even higher.  The question is,
> would chance mutations of all the other needed sites happen with a probability
> that would account for what is observed?

Your answer is essentially yes, given the right environmental
influences.  Most cancers occur in tissues that renew themselves rapidly
and continuously, or would renew continuously given the right stimulus
(high hormone levels or chronic irritations).  It's not just that there
is a pre-malignant change (initiation).  It must be propagated
(promotion and progression).

Quote:
>  The independence assumption of
> course means that the probabilities all multiply, and since they are all
> less than one, the product could be quite small.  Probability here of
> course is really probability per unit time, and the question is, would
> the numbers show anything like the right probability for a cancerous
> genome to develop in even one cell, in a period of time comparable to a
> human lifetime, or would it take more like the age of the universe for it
> to occur under an independence assumption.

The opinion of many is that if you live long enough, eventually you will
get cancer.  There is a small, but finite risk of an error leading to a
mutation in a critical gene at each cell division.  Of course the
mutation can be lethal, harmless or promote an advantage.  But if it
promotes an advantage, at least some subsequent events will not be
independent.  The subsequent events may depend on the level or type of
stimulatory or inhibitory influences on the cell for example.  And
alterations in key genes (such as p53, a pivotal control point in the
cell cycle) lead to genetic instability, such that subsequent events
will occur at a higher rate than normal.  And there are billions of cell
divisions occurring every day in the human body.  So there is both time
and sheer number of cell targets in favor.

Quote:
> Of course mechanisms which couple the mutations in some way could
> alter the probaiblity calculation considerably.  An 'organism' that
> does carefully selected edits for its own amu{*filter*}t would be one such
> mechanism, but any plausible mechanism would do.

The "organism" theory might be one explanation.  For some animal
cancers, retroviruses containing oncogenes (or regions that can control
protooncogenes, depending on point of insertion) are often causative
agents.  But most current evidence for human cancers points to the
development of genetic instability, either through a pre-existing
mutation in a tumor suppressor gene, a random mutation in a
proto-oncogene, or its random development during a lifetime of repeated
cell replication.  Viruses play a role in only a few human cancers:
currently EBV in Burkitt's lymphoma (Africa) or nasopharyngeal carcinoma
(China), HBV and liver cancer, HPV and cervical cancer, HHV-6 in
Kaposi's sarcoma, HTLV-I and II in rare cases of human T-cell
leukemias/lymphomas.

Quote:
> >Some cancers are known to be associated with specific viruses, but this is
> >not the case with all cancers.  Viruses such as HPV, are known to
> >inactivate p53 and/or Rb.  And even if one is infected with such a virus,
> >there is no guarantee that he/she will develop its associated cancer.

> Right.  Some viruses could do edits on genetic material for their own
> amu{*filter*}t or benefit, but they might not be the complete set of alterations
> needed to produce cancer.

The effects of most of the known human culprits seem to be
multifactorial:  binding and inactivation of p53, Rb or both, or
influencing expression of various growth-promoting cytokines.  Or, in
the case of a retrovirus, inserting itself upstream of a critical
control gene thereby uncoupling its expression from its promoter, or
control, region.  (This is seen in many animal retroviruses--for example
the mouse mammary tumor virus, if it inserts upstream of the c-myc gene,
will permanently activate it.  MMTV is transmitted to offspring via
nursing.)

Quote:
> Other viruses might have a different game plan in mind, and there could
> conceivably be one virus that does it all.  With the right strategy, it
> might even defeat the normal immune system method for detecting the presence
> of a virus, the display of 'self' and 'nonself' proteins on the cell's
> surface.

So far, that one virus has not been isolated.

Quote:
> >Tell me John, how you would reconcile the "cancer organism" theory to the
> >HNPCC (hereditary non-polyposis colon cancer) syndrome, where the genetic
> >defect is in a specific DNA repair enzyme, coupled to the normally high
> >rate of replication of colon epithelium cells? >T.

> Dunno.  I never thought about it.  Is this one genetic defect sufficient
> to cause cancer in this case, or are several others also required?
> If it's the former, you have your answer already.  If it's the latter,
> then we come back to the probability calculation.  Are the required additional
> alterations wildly improbable, given an independence assumption, or not?

I'll say the latter but the required additional alterations are not
wildly improbable at all.  In HNPCC, you have a case where mutations in
DNA are not repaired efficiently, so you create a situation where future
alterations occur at a much higher rate than expected.  A similar
scenario works in xeroderma pigmentosum, where defects also occur in DNA
repair enzymes.  These patients can control (somewhat) the damage, as
long as they stay out of the sun (ultraviolet radiation).  That's an
example of the interplay between genetic and environmental factors in
cancer.

T.



Thu, 17 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"


Quote:

>The opinion of many is that if you live long enough, eventually you will
>get cancer.  There is a small, but finite risk of an error leading to a
>mutation in a critical gene at each cell division.  Of course the
>mutation can be lethal, harmless or promote an advantage.  But if it
>promotes an advantage, at least some subsequent events will not be
>independent.  The subsequent events may depend on the level or type of
>stimulatory or inhibitory influences on the cell for example.  And
>alterations in key genes (such as p53, a pivotal control point in the
>cell cycle) lead to genetic instability, such that subsequent events
>will occur at a higher rate than normal.  And there are billions of cell
>divisions occurring every day in the human body.  So there is both time
>and sheer number of cell targets in favor.

A similar thing happens in communication engineering, where sometimes
what seems like a happily low "probability of bit error" like 10^^-6
or so turns into a surprisingly frequent occurrence of bit errors,
e.g. on a very high speed link where bits go flying by at 100 million
or so per second.  In this example, of course, bad bits would go by at
about 100 per second.

However, your example needs some further analysis.  At first hearing,
it sounds as if a lot of cancer-producing mutations might occur.  However,
if I am understanding what is going on, your statement would lead to
the conclusion that SINGLE mutations would occur all over the place.
But since there are so many different cells doing the mutating, no ONE
of them would necessarily get even TWO mutations, let alone three or
four, let alone the precise three or four needed to produce cancer.
The answer still might come out as you say, that sooner or later, with
a lot of ill-considered choices of occupation, cooking.net">food additives or
contamination, passtimes, medications, and whatnot, a lot of people will
get cancer before the end of their natural lifespans.  But I'd like to
see a more careful probability calculation than I have thus far.

-John S.



Fri, 18 Jan 2002 03:00:00 GMT
 Scientists Uncover Cancer "Road Map"

Quote:



> >The opinion of many is that if you live long enough, eventually you will
> >get cancer.  There is a small, but finite risk of an error leading to a
> >mutation in a critical gene at each cell division.  Of course the
> >mutation can be lethal, harmless or promote an advantage.  But if it
> >promotes an advantage, at least some subsequent events will not be
> >independent.  The subsequent events may depend on the level or type of
> >stimulatory or inhibitory influences on the cell for example.  And
> >alterations in key genes (such as p53, a pivotal control point in the
> >cell cycle) lead to genetic instability, such that subsequent events
> >will occur at a higher rate than normal.  And there are billions of cell
> >divisions occurring every day in the human body.  So there is both time
> >and sheer number of cell targets in favor.

> A similar thing happens in communication engineering, where sometimes
> what seems like a happily low "probability of bit error" like 10^^-6
> or so turns into a surprisingly frequent occurrence of bit errors,
> e.g. on a very high speed link where bits go flying by at 100 million
> or so per second.  In this example, of course, bad bits would go by at
> about 100 per second.

> However, your example needs some further analysis.  At first hearing,
> it sounds as if a lot of cancer-producing mutations might occur.  However,
> if I am understanding what is going on, your statement would lead to
> the conclusion that SINGLE mutations would occur all over the place.

I can't remember off-hand what the actual fidelity of DNA replication
is; it's very high, but not perfect.  (An error of once in a million
bases sounds about right but I'll have to check.)  So, about 3 billion
bases in human DNA, with approximate replication error rate of 0.000001,
suggests one cell will have to deal with 3000 replication errors at each
round of DNA replication.  Most of which, fortunately, are corrected.

For the sake of argument, let's say only 10% of the DNA encodes actual
functioning genes, and that there are about 100,000 genes in a human.
That would mean, in one cell, about 300 errors (1/10 the total number of
errors) will occur in the functional coding region with each round of
DNA replication.  That's an error rate of about 0.003 for each gene, for
each round of replication in one cell (300 errors/100000 genes).  But if
we then assume about 100 billion new cells are created every day,
overall there would be 0.003 errors/gene x 100,000 genes/cell x
100,000,000,000 cells, which is 30 trillion potential replication errors
every day.  Luckily there are more limitations to this, mainly the very
high efficiency of DNA repair mechanisms and that almost all such
remaining errors will be either lethal or neutral.  I'll assume that DNA
repair systems are at least 99.99999% efficient, so that the 30 trillion
errors over the whole body per day will be reduced to 3 million
errors/body/day that go unrepaired, or 0.00001 errors/cell/day in the
coding genes.  As a baseline.  (Doesn't include the effects of
environmental mutagens.)

Let's say there are 30 or so genes that are critical "checkpoint" genes
which, if altered, will start the cell along a malignant route.  So in
all those 100 billion cells per day, with those 3 million errors, there
is a chance of 0.0003 per day that the unrepaired error will occur in a
critical checkpoint gene (30 checkpoint genes x 0.00001
errors/cell/day).  For an error, that leads to malignancy, to propagate
in one cell, it has to be a) in a critical gene, e.g. p53, that knocks
out or activates the gene function; b) not repaired before the next cell
division; and c) not lethal to the cell. Most of the errors, if not
repaired, will be lethal.  But fortunately, in a healthy body, most
errors will be repaired.  Some studies suggest that actively transcribed
genes are more likely to be proofread and caught for errors than
non-transcribed genes or intron regions.  This can cause problems if the
cell has been quiescent for a long time and is then stimulated to
divide.  If the mutation is in a gene like p53, which links DNA repair
to the cell cycle, then you have a situation where the cell becomes
*more* susceptible to mutation in succeeding generations.  Add in the
various promoters, which influence the rate of cell division (such as
estrogen or testosterone for hormone-dependent cancers), over a lifetime
of 50-70 years (the usual age at cancer diagnosis) and you begin to
think it's a wonder we don't see more cases of cancer.  Of course, it
depends whether the cell will indeed acquire more mutations or
alterations.

Quote:
> But since there are so many different cells doing the mutating, no ONE
> of them would necessarily get even TWO mutations, let alone three or
> four, let alone the precise three or four needed to produce cancer.

I suspect, and many might agree with me, that the order of mutations may
not matter, just that they occur.  And that, if a checkpoint gene gets
knocked out, it influences the rate of subsequent cell mutation in each
succeeding generation.  And it doesn't matter whether the checkpoint is
knocked out by mutation or by inactivation, e.g. by a virus, just that
it happens.  If my calculations above are correct, there is an ample
chance for enough mutations to accumulate (along with the promotional
factors), to potentially produce a cancer.

Quote:
> The answer still might come out as you say, that sooner or later, with
> a lot of ill-considered choices of occupation, cooking.net">food additives or
> contamination, passtimes, medications, and whatnot, a lot of people will
> get cancer before the end of their natural lifespans.  But I'd like to
> see a more careful probability calculation than I have thus far.

That's not including the effect of mutagens, which adds more burden to
the DNA repair system.  Many cannot be avoided (sunlight, natural
mutagens in food) but either the body can deal with them (enzymatic
detoxification, eg. by the cytochrome P450 system) or one can pay
attention and not expose him/herself necessarily.  I hope the above fits
the more careful calculation that you seek.  Basically, it suggests that
over a lifespan of 70 years, a large number of people will get cancer.
Remember too, for most of human history the average human lifespan was
something like 30 years, so most people died before they could reach an
age to get cancer.  Now the average life expectancy is over 80 years.
(Even so, cancer was 8th leading cause of death in 1900.)

Hope this makes sense.  (It's late, and I've always hated probabilities.
;-0 )

T.



Sat, 19 Jan 2002 03:00:00 GMT
 
 [ 8 post ] 

 Relevant Pages 

1. Recent News About road maps. road maps,silk road maps,oregon forest service road maps,north carolina road maps,satellite road maps

2. "Road Map".

3. "On the road to walking?"

4. Most Recent News About map quest. yahoo map quest,canada map quest,google map quest,map quest 3433 thunder road,map quest seattle

5. Most Recent News About map quest. yahoo map quest,map quest com,map quest usa,map quest 3433 thunder road,map quest us

6. Recent News About texas map. texas state map,online map of texas,texas road map,map texas,county map of texas

7. "Loved Ones"-"Terminal Cancer"- update

8. map quest : Leading News. map quest driving directions,map quest st paul alberta,quest map,map quest 3433 thunder road,map quest edmonton ab

9. MAPS - the "Mysterious Addisonian Potato Syndrome".

10. Hitlers current research "scientists"?

11. New Scientist - "The Power of Nothing"


 
Powered by phpBB® Forum Software