Thank you.
Yeah,
thank you.
DR sessions and members of
the symposium.
When clinicians
extrapolate from the characteristics
of a disease presented in patients,
you know,
seen by them to a larger universe.
Their their vision may be clouded by
ethnocentrism.
The tendency to regard ones
experience as the standard of reference
epidemiology may provide a
corrective for ethnocentrism
and we will try to sketch some of
the epidemiological highlights of
colon cancer in the hope that
will provide a perspective in which
the clinicians can place their
own contributions
based on patient observation.
For this purpose,
we will draw on information assembled
from international comparisons,
time trends,
migrant populations and the
precursor lesions of polyps,
uh which have just been so
uh ably discussed
then,
uh the probable reason for failure to make
greater progress in the ideology of large
bowel cancer is that virtually all the
studies have been conducted in the high risk
populations of north America and western
europe.
For a firmer grip on the
epidemiology that might lead to
ideological hypotheses.
one Ought to begin with
international comparisons.
The initial evidence for striking inter
population variations and risk
came from comparative studies of
mortality.
These have been uh elaborated more
recently by incidents data from
tumor registries made possible by
extension of the network of tumor
registries to cover some
populations and eastern
europe asia africa
and south America.
Uh some of the results first
slide please.
And I guess if you can Yeah,
that's fine.
This is a selected
uh group of comparisons of
populations.
The purpose is to show the
magnitude of range of differences
between high and low risk populations
on the order of six or seven fold.
This particular comparison is
for colon cancer incidents,
that is a sigmoid.
And above.
The next slide
gives the
that was for males.
They results for females
show the same range
of inter population difference.
The
results no
fall into roughly three
groups that is high risk
represented by Connecticut,
new Zealand.
Then we have a medium
intermediate risks in most of western
europe,
United Kingdom Norway and
then lower risks
in eastern europe
Finland at
japan and south America.
The comparable data
on carcinoma direct.
Um The next slide
for males chose
the the same type
of of variation
although a little narrower range.
And the next slide on females
rectum females
substantially the same
picture the same.
This similar
ordering for colon and
rectum can be shown in the next
slide.
Uhh this slide We have one
axis incidents rights for rectum.
The other axis giving the incidence
rates for colon.
Uh
the
uh rights
for
for males,
we can look up the
and for females.
No,
the uh the white the yellow dot
is the uh the females and the red
triangle as a team.
And you can see that uh fairly
regular ordering.
Now some of the
enter registry variation in
column rectum ratios.
We can bring this point out better in the next slide.
This gives the
directly ratios of the colon
cancer incidence to the rectum
incidents.
Uhh and for
virtually all of the the
populations,
this ratio does exceed
unity.
There are certain differences which
you can see for females ratio
generally being higher for
that sex.
Uh part of this variation that you
see in this slide might be uh due to
different practices and assignment of tumors
close to the sigmoid fletcher.
But classification artifacts seem
unlikely to be the sole
explanations.
And this set of results raises the
possibility of population differences
in anatomical localization of uh
bowel tumours.
The localization of tumors within the
colon has been reported to very in
the literature,
High risk populations,
high risk.
Most of the tumors appear to be on
the left side.
A typical result being the
63% of such
a colon tumors in England.
While in uh low risk
uh countries,
the tumors seem to be concentrated on the
right side.
Uh typical result being uh
65% right sided uh
localizations in Cali
Colombia.
In fact,
when we array the registries in ascending
order of overall cancer incidence,
there is a concentration of tumors
in the sigmoid segment
uh in those registries with the
highest overall incidents,
uh I don't have a slide to show this,
but the data from Connecticut
show this concentration.
And if we took really low risk populations such
as the Bombay and Cali,
we would have very
little concentration in this
part of the,
of the colon.
Uh The next
figure,
uhh describes the sex
ratios for colon cancer incidents.
The point to be made here is that in high
risk populations,
the male female ratio tends to
be close to unity.
There's an obvious exception here.
The Hawaiian japanese,
which we will discuss a little later.
While in the lower risk populations,
there is a tendency for female
predominance.
The picture for rectum is quite
different.
Higher rates for males prevail at all levels
of risk.
Next slide,
you can see the there's only
one exception
for rectum.
Uh The sex
ratios,
however,
based on summary age adjusted rates
obscure some rather interesting sex
age interactions.
Uh When we look at the populations at
two extremes of uh colon
cancer risk,
we see some different sex
behavior in the
age specific rates.
Next slide
Connecticut and colleague Connecticut being
a high risk population.
We can see that the
rates for the two
sexes are very close
together and with a slight
male predominance at the older age is
in Cali on the other hand,
we can see Quite a definite
female predominance after age
65.
The pattern of
uh of uh just depicted
for Cali and other low risk
communities at corresponds
closely to that observed for the subset
Of seeking a sending tumors
in higher risk
populations uh for both the female
predominance after age 65 has
established this next slide,
please.
We had the result for colleagues and
in Norway,
which is one of the intermediate risk there
for sake um and ascending.
We do get this
crossover with higher female
risks and the older age
range.
The in
contrast the
we get a similarity in contours of the age
specific rates for colon and
rectum cancers.
Using the rectal sigmoid lectures the
dividing line Uh
in very high risk populations where the
male rights after 65 are clearly
higher.
In both.
Next slide
we have this male
predominance,
but it's a more marked
for your rectum.
From these and related observations,
we make injector that the entity colon
cancer may have at least two separate components seek
him ascending and as using as a
label and sigmoid rectum.
It appears to be the latter which accounts for the
enhanced risks observed in the north
America and western europe.
If environmental factors are
responsible,
the intensity of exposures and resultant
incidents may have been changing over time
Connecticut is one of the registries in which we
have information on this point.
Next slide,
please.
Uhh in
in Connecticut for the sigmoid,
I'm looking at the two
time period Ends of the 15 Year Time
Period.
The male predominance appears to
be a little more firmly established.
Uhh next slide
These for transfers and
descending in the later time
period.
This friend to male dominance
seems to be going on.
Next slide
for Sikkim and descending.
Males have not become dominant yet.
But you can see that the gap is
narrowing.
So uh
one might speculate if if we can turn
the incidents in the United
States as an
epidemic.
one can speculate if this
epidemic will reach a climax with
the presence of a male
predominance in the
lesions in the Sikh um and
a sending no colon.
Uh In addition to these inter population
comparisons,
it's useful to discuss
changes which have appeared in
certain migrant populations
when the colon cancer
experience of the Hawaiian japanese is
contrasted with that for
japan and US white,
a displacement and risks as can be observed.
Next slide,
it's this is the baseline for
japan.
The line for US whites,
the host populations.
And you can see the uh
positioning of the mortality rates
for the the migrants
generation themselves,
the esa and their native born descendants among
males has been pushed up towards the
U.
S.
Right?
This tendency is
less marked for the females
and it is this
phenomenon I believe,
which produces this unusual
sex ratio for the Hawaiian japanese,
which occurred in an earlier slide where there
was a excess of
risk of colon cancer among
the male population.
There's uh suggests that there is
a greater male volatility and
biological uh response to
changed conditions
in the
a latent period between exposure
and development of neo plas.
Ums one of course tries to look
for precursor lesions and to use
such information and attempt to transform
the epidemiology of colon cancer
into the epidemiology of a
precursor lesion.
The Early landmark in this direction
was a paper by Helwig in the 1940s,
which looked at adenomas polyps
as a candidate precursor.
His line of attack was too search for
congruence is in the anatomical distribution
of cancers and polyps.
Uh time doesn't allow to
consider the findings of
people,
you know,
following Hellwig and the different
interpretations of these uh findings.
It started,
I'm going to make a general observation
as I say,
that they were unable to resolve the
the polyps cancer congruence
problem and related issues with
observations solely from
high risk populations.
Uh This is probably not surprising
since when you have little background variation in
the distribution and biologic behaviors of
lesions,
it becomes very difficult to
detect patterns and relationships.
A single population might have
permitted test of a very simple palla
cancer models such as the
constant probability of transition
independent of anatomical
localization.
But when you have more complicated
models,
uh,
I think that you need greater
background variation and experience
for this reason,
studies of
precursor lesions needed to
be carrying out carried out in populations
with widely different colon cancer
incidents.
Uh,
since hell Wigs data were on record.
Uh Professor Korea and
Colombia and I decided that
uh work of this type should be
done in a low risk population represented
in Cali Colombia and Korea has
published on the study of
Of 1500
bottle specimens from
necropsies there.
The results that I'm going to give
are are
still in the infancy
and they have to be treated as
exploratory and illustrative and certainly
not as uh as definitive.
However,
when we compare these results with
those from Helwig,
from results more recently obtained
in New Orleans by Professor
Strong and his group and
results from
the japanese migrants in Hawaii
and from populations in
japan.
Certain uhh
patterns emerge next slide.
Uh This compares the
distribution of aveeno
metis and hyper plastic polyps in the
autopsy specimens in
college Korea when he did the
work because
uh concentrated
on adenomas polyps,
he felt that uh this
distribution would turn out to be
quite different than in the United States is
reported by Helwig and indeed it is uh
you uhh first one
explanation these location of
these lesions are plotted with
respect to distance from the anus.
And uh but they on this
particular scale uh they're plotted in
terms of relative distance.
Uh that is converting the total length
of the of the ball
from the illegal cycle valve to the
To the and this is 100%.
So that most of the of
the tumors up to the sigmoid
fletcher would be in this the
1st 10
From 10 to 90% on this
note.
For then there is a similar
distribution for hyper plastic
polyps.
There is a concentration in
the in the
rectum as one would expect.
But there are some surprising features about
this which I'll return to
when we compare
Cali with Hellwig and with new
Orleans in the
the high risk colon population than the U.
S.
Have a higher prevalence of
adenomas polyps at all ages in both
sexes.
And this difference is
accentuated particularly for
females when you limit the contrast
to multiple at anomalies.
That is multiplicity is is not a
feature in a uh this low
risk uh population in new
Orleans.
The adenomas were also larger and they
had a much higher
proportion of adenomas
with a tapia's
also a much higher proportion of
the new Orleans.
Adenomas were concentrated in
the in the
sigmoid recto
sigmoid area.
Uh This uh
is replicated now in
the observations on the japanese
migrants.
That is the Hawaiian japanese.
Haven't adenomas polyp distribution
closely resembling that.
And all these features exhibited by
blacks and whites in New Orleans.
The observations from
Japan taken in Miyagi
prefecture have a
polyp distribution for adenomas which
resembles that of the low risk
college population.
That is uh we now have uh to
a high low risk contrast which show
essentially the same results.
Now when Korea did the work on
hyper plastic polyps in Colombia,
I think that he was doing it primarily for
sake of completeness recording the observations and
uh those plus you know,
juvenile polyps.
He like other pathologists.
I didn't think that
uh there was going to be any no
relationship uh demonstrated
with with the tumors and
the location of the
these hyper plastic pileups.
Got uh
And I think the
feeling was based on on
the Marfa logic evidence,
but big,
big surprise.
It turns out that in the
in the Hawaiian japanese and the
native japanese,
that uh there is a great
difference in the distribution
of hyper plastic polyps,
that that is while hyper plastic polyps
are certainly concentrated in the
in the area between the flexion er
and the anus.
There's just a much
greater frequency of such polyps
in the Hawaiian japanese
popular as opposed to
japan.
And this same finding is
emerging from the
comparisons in new Orleans with
the colleagues.
So new Orleans with
the colleagues,
so that while the
pathologists have been slow
to,
I have not have thought that there
was anything in the morphological studies that
would really uh yield
suggestive leads these
epidemiologic comparisons now point to
great differences in these
parents of these hyper plastic.
Just how this will
relate in the long run to the
distribution of
of all right
of of rectal
cancers remains to be seen.
My feeling is that uh
that these type of observations are
likely to uh stimulate uh
work in the area of,
you know,
cell kinetics which uh your doctor
lukin will be discussing
this afternoon and that
uh this
will be one
area in which the say the colon
cancer task force and the colon cancer,
you know,
segment of the cancer institute will be
spending much more time on
so to uhh
summarized uh
we have might have proposed
the following as a working model
in low risk populations where the disease
is endemic Colon cancers are
concentrated in the Sikh,
um and ascending colon female cases
are preponderant and most of
the rise to the maximum incidence level
occurs by age 50 55.
If,
when a new ideological factor
is introduced,
A transition from an endemic to an
epidemic phases first expressed as a
rise in sigmoid cancers among older
men over 55.
A rise in female sigmoid cancer
follows later and the time lag is
reinforced by a tendency for these female
cases to appear at somewhat older ages.
As exposures to a
suspect undefined ideological
factor becomes more intense and prolonged.
A later phase is characteristically
arise and seek them and ascending colon
cancers more marked for males and for
females so that the female excess
and seek um cancer prevailing under endemic
conditions is as dis
minister diminished.
The present
observations on the prevalence of polyps
suggests.
Some link with the level of colon cancer
incidents in the given communities,
not only with respect to magnitude of prevalence,
but also with respect to distribution by
age,
sex,
anatomical localization and
multiplicity.
I think these clues lead to
uh some ideas on
ideology which I think time
will prevent uh me from
discussing this
morning,
but
there is the possibility I think
of of of linking
up studies of
ideology and testing them with respect
to this uh difference in
anatomical distribution for both.
Um
polyps and cancers.
one thing which is occurring
and this japanese
population which is very suitable for
epidemiologic studies of
dietary hypotheses.
And I won't go into it any further than
to say this,
that the associations with
uh some of the differences between the case and
control series appear to be
much more marked for those cancers
which are uh
in the rectal sigmoid,
you know,
fletcher and above that is such associations
for tumors in the in the lower rectum
is uh is virtually
absent so that we may be
able to uh bring
about a synthesis of uh
of studies and the
traditional epidemiology logic
type represented by case control
and prospective cohort
studies with findings
uhh made by uh no
geographical a
pathologist.
Okay,
okay.