E&
DISCUSSION DOCUMENT 0
SMOKING AND HFkLTH: FVTURE STRATEGIES FOR
RFI;FARCH
1. EPIDFXIOLOGICAr. A.PPROACHES
W Lung Cancer
Research clainizin that smoking is associated with
lunx cancer
- Further research on this topic is unlikely to be
a high
priority for the scientific community except to
study the
associations between Io- tar products and disease.
In order
ta be of use, such studies would have to study a
generation
who had smoked only low tar products. It would
also be
necessary in drawing comparisons with previous
generations
of smokers, to control for additional factors that
may have
changed in this generation e.g. dietary habits.
It is arguable that the tobacco industry should be
carrying
cut such research. However, this is a very
long-term
project (answers could not realistically be
expected for 25
years) and it is likely that, by then, the product
may have
changed significantIv. The results may not
therefore be
directly relevant to the product of the future. In
any
case. a number of such studies are known to be
already in
progress e.g. funded by the ISCSH in the UK.
Other factors associated with lunx cancer
- Such studies have limited value. The list of
factors
associated with lung cancer is no. extremely long.
Ho-ever,
it does not directly address the association
between smoking
and lung cancer unless it could be shown that one
or more of
these factors are confounding factors e.g. all
smokers who
C)
C11
Ic
BATCo document for Province of British Columbia 28
October 1999


- 2 -
develop lung cancer are pet bird keepers or are
exposed to
diesel exhaust. Such research does, of course,
serve to
combat suggestions that smoking is responsible for
90% or
more of all lung cancer deaths and resulting
publicity may
draw attention to the fact that smoking is not the
only, nor
even the major factor claimed to be associated
with certain
diseases.
Trends in disease and in cixarette sookinx
- It has been noted by a number of scientists
(e.g. Peter Lee
using cohort analvsis) that there is frequently a
poor
correlation between the trends in a disease such
as lung
cancer in the population, and patterns of
cigarette
smoking. As above the value of such studies is
that they
can point to the relative importance of other
factors in the
disease e.g. environmental factors. Unfortunately,
the
studies are unlikel, to achieve more than to point
out that
anomalies exist and suggest they are due to other
factors;
the identification of which other factors is
unlikely to be
feasible. It is also impossible, because of
inadequate
death certificate information, to know whether
changes in
disease rates are different in smokers and
non-smokers.
Trends in lunx cancer subtypes
A number of studies have now been published
suggesting that
lung cancer rates are increasing in non-smokers
relative to
smokers. This is frequently linked to a selective
increase
in lung cancer subtypes that have been found to
occur less
frequently in smokers e.g. adenocarcinoma. It
would be
useful to investigate this trend further.
CI,
%0
ON
114
aN
LM
BATCo document for Province of BritiSh Columbia 28
October 1999


(ii) Cardiovascular disease
Research claiminx that smokinx is
associated with cardiovascular disease
- Studies are still appearing on this subject,
investigating
the various subtypes of cardiovascular disease. In
general,
risk ratios e.g. (averaging 2 - 3 for coronary
heart
disease) are low enough for many such results to
fall into
the category of 'low risk epidemiology' and
therefore to be
subject to the associated problems.
- Atherosclerosis underlies many of the
cardiovascular
diseases that are claimed to be associated with
smoking, and
is also relevant to peripheral vascular disease.
Particular
attention should therefore be paid to literature
relating to
atherosclerosis hen evaluating the claimed
association
between smoking and heart disease. It is
noteworthy that
the SRG is currently supporting work in the area
of
mechanisms of atherosclerosis, with Professor
Kak-k-ar at
King's College London.
- Intervention studies (e.g. MRFIT) have shown
that if
individuals stop or reduce smoking there is no
benefit in
terms of a reduction in heart disease risk.
Other factors associated with heart disease
- Approximately 245 different factors have been
claimed to be
associated with heart disease. It would therefore
be
inappropriate to have such studies high on a list
of
priorities.
C~
Ic
BATCo document for Province of British Columbia 28
October 1999


- Diet has long been suspected to be a major risk
factor
associated with coronary heart disease; however,
recently
Professor Michael 01 i%-er of Edinburgh University
suggested
that it could be a confoundinx factor in studies
claiming an
association between smoking and heart disease. He
suggested
that the claimed increased risk in smokers -as not
a direct
effect of smoking, but an indirect result of
smokers having
a poor diet. Such work is certainly worth further
investigation.
(iii) Respiratory disease
Research claiming that smokinx is associated
with increased risk of bronchitis and euphysena
- It is frequently stated that only smokers
develop severe
emphysema; this claim is worthy of further
investigation.
It is certainly not the case for bronchitis.
Other factors associated with respiratory disease
- A large amount of researc'-1 has suggested other
factors that
have been associated w:th both bronchitis and
emphysema.
Again, there is Probably I;ttle direct value in
giving high
prioritv to this kind cf research.
(iv) Other diseases
- A number of other cancers and diseases have been
claimed to
be associated with smoking. For most of these,
authorities
such as the US Surgeon General accept that the
evidence is
insufficient to determine that a causal
relationship
exists. For other cancers, however, a causal
relationship
BATCo document for Province of BritiSh Columbia 28
October 1999


has also been suggested e.g. laryngeal,
oesophageal and
oral cancers. The majority of industry-funded work
has
tended to concentrate on lung cancer and therefore
relativelv little attention has been paid to
anomalies in
the literature relating to other cancers. It would
be
useful to pav a little more attention to these
diseases in
future.
EPIDFXIOLOGY: IKPLICATIONS FOR RESFkRCH
Hinh Priority research
- Trends in lung cancer subtypes. It would be
useful to commission a
review considering whether the dissociation
between adenocarcinoma and
squamous and small cell carcinomas (in terms of
their association with
smoking) is a distinct one. From this, for
example, it may be
suggested that survevs should be carried out in
kev countries to
provide more information about the relative
incidence of lung cancer
in smokers and non-smokers, and associated
subtypes.
- Diet and heart disease, in particular relation
to diet as a
confounding factor in the association between
smoking and heart
disease. Michael Oliver could be approached as an
initial step.
- Is it the case that only smokers develop set-ere
emph%-sema' - A
consultant's opinion could be obtained as an
initial step.
cc
BATCo document for Province of BritiSh Columbia 28
October 1999


Low Priority research
- Trends in disease in relation to cigarette
smoking habits.
- Research into other factors in
smoking-associated diseases e.g. pet
birds.
- Long-term epidemiological studies into the
effects of changes in the
product.
2. MOLECULAR BIOLOGY/MECHANISMS OF DISFASE
Lunit Cancer
- The most informative study in this area would be
to identify a
genetic factor that increases the risk of lung
cancer, and to
investigate whether incidence of the particular
factor is different
in smokers and non-smokers. Idle's work on
metabolism of
procarcinogens has so far covered the first part
of this question,
but has not yet adequately covered the second.
However, before
additional funding were to be committed, some
additicnal questions
need to be answered e.g. do the enzymes that Idle
is interested in
metabolise chemicals that mav be relevant to the
development of
lung cancer'
- An alternative approach comes from Dr Helmut
Bartsch f,-om the 1ARC
in Lyon. Dr Bartsch is investigating the relevance
of P-450
systems in animals, which mav shed some light on
Professor Edle's
research. They have also carried out work in
smokers in,estigating
the activity of such enzyme systems, which app~-ar
to be induced
during smoking. Further research of this kind
would complement
Professor Idle's work, and possibilities are
currently being
discussed with Dr Bartsch.
C7 \
\Z
BATCo document for Province of British Columbia 28
October 1999


- A large amount of attention (both scientific and
media) has been
paid to the formation of DNA adducts for chemicals
that occur in
cigarette smoke. It i s frequently assumed that
this might hake
some bearing on carcinagenesis. However, this is
by no means
certain, and a recent review by Ashby (1988)
suggests that care
must be taken in application of tests based on
such measures. It
may not be appropriate to actually carry out
research in this area
if the relevance is in total doubt, however, Dr.
Peter Farmer.
Dr. Bartsch and Professor Bruce Ames are potential
contacts in this
area.
- Immunology is another area that is increasing in
importance. For
smoking, it has so far been relevant in two
different areas: W
respiratory disease and (ii) cervical cancer.
G) A number of studies have claimed that smoking
affects immune
components of the respiratory S"Stem e.g. alveolar
macrophage activity. The relevance of such studies
is not
totally clear but it has been suggested that a
reduction in
immune capacity would render an individual
susceptible to
disease and possibly even to the development of
cancer. It
would be extremely useful to have a review of this
area to
assess the potential need for further research in
the field
of immunology.
(ii) A study recently suggested that Langerhans
cells. a
component of the immune system, were reduced in
the cer,.ix
of smokers. It was claimed that this reduced
im.une
activity would make individuals more susceptible
to viruses
that may be associated -ith cervical cancer.
lfo-e'er'
Langerhans cells are just one component of the
immune
system; others -ere not studied. Again, a review
-ould be
useful leading to further research if necessary.
cz:
Cl-
~Ic
CY\
BATCo document for Province of BritiSh Columbia 28
October 1999


8 -
Heart Disease
- It is frequently claimed in the scientific
literature that smoking
raises blood lipid levels and affects clotting. It
would be useful to
have someone review this area in order to analyse
the evidence and
determine whether more research would be useful.
Pioloxical testi-- for carcinoxenesis
- An extremely large number of recent reviews on
the subject of
carcinogenicity testing have cautioned that
extreme care must be taken
when extrapolating from in vitro and in vivo
genotoxicity tests to
human carcinogenesis. The relevance of such tests
is not as yet
established. Similarly, the appropriateness of
other non-genotoxic
annual models of carcinosenesis to the human
situation has been
questioned. Bearing this in mind, there would seem
to be no
particular value in spending a great deal of
effort in developing ne-
tests until such time as their relevance and the
appropriateness of
the model is more clear. The exception to this, of
course, is that
where regulatory authorities recommend the use of
specific tests for
the purposes of testing novel or modified
products, when we should
ensure that we have the capacity to carry them
out.
Hixh oriority research
- Genetic differences that may influence the risk
of lung cancer.
Already being carried out with Professor Idle. and
approaches made to
Dr Bartsch at IARC.
- A review of immunology as it relates to
smoking-associated diseases
e.g. respiratory disease and cervical cancer.
- A review of the claimed effects of smoking on
factors that may be
relevant to heart disease.
ON
BATCo document for Province of BritiSh Columbia 28
October 1999


Low Priority research
- Research related to the formation of DNA
adducts. until such time as
its relevance to carcinogenesis is more clear.
- Detailed research into novel animal models of
carcinogenesis,
including both in vitro and in -ivo studies,
except where recommended
by regulatory authorities.
3. ENVIROMNFNTAT. TOBACCO SMOKE
W Evidemioloxy
- t;ith the majority of studies (17 out of 23,
with all their
flaws) sho-ins no statistically significant
effect, there is
little justification for carrying out others.
- Studies of potential confounding factors in
epidemiological
studies would be verv useful. Although Peter Lee
is still
very enthusiastic about misclassification, it is
not gaining
particularly wide acceptance as a critical factor.
Further
studies in this area are unlikely to be of value
(there are
already se,eral). It would be more useful to
concentrate on
other factors such as diet; this is currently
being
investigated in Rong Kong, e.g. by Linda Koo.
- The majority of studies have not shown a
significant
association between ETS and heart disease;
however. there
are some studies showing a significant association
between
ETS and the development of respiratory effects in
small
children. Further analysis of these studies would
be
useful, along with some research into potential
confounding
factors.
C)
CD
`0
BATCo document for Province of British Columbia 28
October 1999


- 10 -
(ii) Indoor Air Guality
- Research into sick building syndrome has
established that
ETS is not a significant factor. It is unlikely to
be
useful to carry out further research in this area
unless
required for specific purposes, e.g. in countries
where such
studies have not previously been carried out.
- We now have (thanks particularly to the
Southampton ETS
programme) a large body of data indicating that
ETS is
present in extremely low levels in a large variety
of indoor
situations. Future research in this area should
probably
now be restricted to studies that are useful for
particular
purposes e.g. to fight an airline ban on smoking;
or in
countries where ETS is just emerging as a problem,
to give a
local example of the kinds of levels of ETS that
may be
present in the air. For example, it has been
suggested that
there has been little investigation of ETS levels
in the
workplace in the USA.
- Research into indoor environments in warm
climates has been
relatively neglected, and it may be useful to
carry out a
study in such a country. This is currently being
investigated with Covington & Burling in the Far
East.
- Further investigation into the correlation
between measures
of nicotine levels in the air and measures of
nicotine
levels in saliva would be useful to address Martin
Jarvis'
work in children, which claims that they have very
high
cotinine levels in saliva if their parents are
smokers.
BATCo document for Province of British Columbia 28
October 1999


(iii) Chemistry of ETS
- Volatile organics in the air maybe of
importance,
particularly since the EEC are concentrating upon
them. It
would be useful to have more information about
levels in the
air.
- Aldehydes are major constituents of ETS: it
might therefore
be useful to know more about their concentrations
and
behaviour.
- Information about levels and behaviour of "other
noxae" in
ETS would be useful if it were possible to measure
them.
$ixh Priority research
- Further research into confounding factors in
epidemiological studies
relating to ETS e.g. diet. -
- Analysis of studies claiming to show respiratory
effects of ETS in
small children, with a view to further research
into e.g. confounding
factors.
- Research into levels of ETS components in
buildings that will serve a
particular purpose e.g. to combat a ban in a
particular country. The
ETS group in Southampton should be available to do
this.
- Research into levels of ETS in buildings in warm
climates e.g. Far
East or Central America. Currently being
investigated with Covington
& Burling. and in Costa Rica.
- Further investigation into the correlation
between measures of
nicotine levels in air and in saliva. Currently
underway by Chris
Proctor.
- Research (-here possible) into levels of "Other
Noxae" in ETS.
Z:~
C)
C'\
BATCO document for Province of BritiSh Columbia 28
October 1999


- 12 -
Low Priority research
- Studies of misclassification.
- Research into sick building syndrome.
- Non-targeted research into levels of ETS
components in air.
4. THE EFFECTS OF NICOTINE
M Addiction
- It is difficult to identify research areas that
will have a
bearing on the subject of whether or not nicotine
is
addictive. There is already some evidence
indicating
differences between nicotine and drugs that are
generally
accepted to be addicti,-e. from both human and
animal
studies. Any additional evidence in this category
would be
useful but is should be borne in mind that even
amongst the
drugs that are accepted to be addictive there are
differences.
Work on alternative theories or explanations for
why people
smoke would be more constructive. To date, Da,id
Warburton
has been something of a lone voice on the
suggestion that
smokers smoke. not because the, are addicted, but
because
smoking fulfils a useful function for them. An
analysis of
the validity of Warburton's claims and further.
more
detailed work in this area is already going on at
the
University of Leeds.
BATCo document for Province of BritiSh Columbia 28
October 1999


- it is extremelv difficult to identify research
into
biological mechanisms of addiction. since it is
not clear
what they actually are. However, BAT is alreadv
supporting
work in this area with Dr Roy Wise in 4ontreal,
looking at
effects of nicotine compared to other drugs in
brain
pathways and behavioural. systems relating to
reward and
dependence.
(ii) Other effects
- Since there are a number of scientists who agree
that
nicotine has no carcinogenic or cardiotoxic
effects,
research in this area should have low priority.
- Effects of nicotine in situations of stress,
anxiety, and on
mood and performance are all relevant to the
addiction
question and therefore worthy of support. Work in
this area
is already planned at the Universitv of Leeds.
5. PRODUCr MODIFICATION AND INNOVATION
To date. scientific e,-idence concerning specific
components of
cigarette smoke does not easily form a firm basi5
for the removal of
specific compounds from cigarette smoke. Manv of
the animal models in
which these compounds have been tested have little
relevance to the
development of cancer of the lung. However, in
spite of this, bodies
such as the IARC and regulatory authorities around
the world have
issued lists of compounds that they believe to be
particularlv
problematic in relation to cigarette smoke.
CD
BATCo document for Province of BritiSh Columbia 28
October 1999


- 14 -
From a pragmatic point of vie,. because the
scientific evidence is not
conclusive, and because these regulatory
authorities would themselves
need to approve any significant product
modification or innovation,
the most sensible approach for BAT to adopt would
be to follow their
recommendations as far as possible. The following
areas have been
considered by regulatory authorities:
1. Lo-er Tar i.e. non-specific reduction of all
smoke components.
It is generally believed that, in markets such as
the UK, tar has
no. been reduced to the limit in terms of consumer
acceptability.
Regulatory authorities such as the ISCSH have
therefore suggested
that the tar:nicotine ratio of the product be
altered such that
lower tar does not result in lo-er nicotine. This.
of course.
assumes that nicotine is the most crucial
component of cigarette
smoke. Work in this area is conflicting but the
overall consensus
has been that the nicotine level that smokers
generally take in is
0.8 mg per cigarette. The recommendation also
assumes that
nicotine is an acceptable (in "heal th" terms)
component of
cigarette smoke. Such an approach is certainly
valid and Rothmans
are also kno-n to favour it.
2. Reduction in specific components. This has been
considered in
detail by BAT in the "Other Noxae" list, and a
similar approach
-as adopted by RJ Reynolds in relation to
"Premier". In an,
product modification or innovation. reduction or
removal of these
compounds should have prior't,.
3. Nicotine. It is generall., accepted that
nicotine is not
genotoxic, carcinogenic or teratogenic. Although
Hoffman and his
group have assumed that nitrosamines formed from
nicotine may be
carcinogenic, this idea has not received
widespread scientific
acceptance. For cardiovascular disease, manv
scientists agree
that nicotine is not cardiotoxic. Such effects as
nicotine can be
C)
Q__
BATCo document for Province of British Columbia 28
October 1999


- 15 -
seen to ha% e on the cardiovascular system are
acute and not
significantly different from the effects of mild
exercise. The
major area of scientific controversy relating to
nicotine is
whether it is addictive. However, the US
Surgeon-General's Report
made surprisingly little impact except amongst
anti-smoking groups
and the criticisms that have been made of it e.g.
b~ David
Warburton are valid and acceptable to many
scientists. Except for
the problem of the "image" of nicotine (it is a
drug; without it,
is would be difficult to argue that smoking is
addicti,e) there is
presently no justification for its removal.
Nevertheless. we do
not -anta product based on nicotine alone; the
pharmaceutical
companies are covering this field quite
adequately.
Philip Morris, however. have taken a different
acoroach and
introduced a nicotine-free cigarette. It will be
interesting to
see whether or not this is acceptable to the
consumer.
Sidestream smoke. Regardless of conflicting
research results
relating to the claimed effects of ETS, regulatory
authorities
have reduction of sidestream as a priority. This
should be a
priority of BAT for aesthetic reasons, even if its
reduction for
health reasons is not justifiable. Regardless of
the health
issue, many people would like to see smoking
banned because they
find it is unpleasant and an irritant. Reduction
of sidestream
smoke is therefore a critical response. It would
be particularl,
useful to decrease visible sidestream.
Testiruz of Product modification or innovations
This area has been considered in some detail bv
the SRG in relation to
RJR's launch of 'Premier' and the projects
currently being undertaken b~
the BAT group. During the SRG's review process it
was suggested that the
comprehensive approach of RJR had perhaps set
something of a precedent,
and that it would probably be necessary for future
products to be
subjected to an equally exhaustive battery of
tests.
CD
CD
10
CU
BATCo document for Province of BritiSh Columbia 28
October 1999


- 16 -
The most critical area to consider is on what
grounds, i.e. on the basis
of what tests, would a regulatory authority be
prepared to concede that a
particular new product might provide some kind of
advantage over existing
products?
In their various reports and discussions,
regulatory authorities have
considered the aspects of tobacco products
discussed above (lower tar,
reduction of specific components, the role of
nicotine and the presence of
sidestream smoke). It would be necessary to have a
regulatory authority
concede that a company had followed its
recommendations in producing a
novel product; if it did so then by definition it
would have to concede
that such a product must constitute an improvement
over existing
products. Presumably then, for example, the UK
Independent Scientific
Committee on Smoking and Health would have to
concede by its own criteria
that a product that had lower "tar", lower
specifically identified
chemicals and lower sidestream, would represent an
improvement over
current products.
Complications. of course, arise if, in modifying
the product or
introducing a totally novel product. new
constituents are used and
different chemicals are produced. It would then be
necessary, as far as
possible, to determine that these novel
constituents did not have effects
that could be interpreted as toxic. This is the
process that was
undergone by R. J. Reynolds in relation to
glycerol as a major constituent
of 'Premier'. A list of biological tests that the
SRG considered that
regulatory authorities would wish to see carried
out is appended.
BATCo document for Province of BritiSh Columbia 28
October 1999