WDEUBTtyl/46D I November 1989
CARCINOGENS IN TOBACCO
This review has been based on the summary of
information provided in the 1989 US
Surgeon General's report, "Reducing the health
consequences of smoking - 25 years
of progress". Where necessary reference has been
made to the probable sources of
the data in view Of Sortie apparent errors in the
levels quoted. In addition, it should also
be understood Iha( the use of the term
"carcinogen" in this context should be qualified.
It is not surprising that in view of the 2500 +
compounds so far identified in tobacco, a
few should be found to be carcinogenic when
administered in isolation in experimental
animals. Many substances have been designated as
carcinogens under such
conditions, but for only a very small proportion
has this finding been generally agreed
to be relevant to human experience.
Although a substantial number of carcinogenic
substances has been identified in
tobacco smoke. relatively few have been found in
tobacco. It follows that the majority
of tobacco smoke carcinogens are formed by
combustion/pyrolysis inechanisms when
tobacco is smoked as cigarettes or other smoking
articles.
A recent summary of carcinogens found in tobacco
and Mainstream cigarette smoke
is given in Table 7 of the attached sectionof a
prepublication version of the U.S. Surgeon
General's 1989 report. Table 7 illustrates the
point made in the previous paragraph,
for example the listed polycyclic aromatic
hydrocarbons (PAPI) do not occur naturally
in tobacco or other plant materials, but are
formed when these are burned. The source
references for the data in Table 7 are not given,
but it is likely that tile quoted 0.1 - 90
ng per gram range for tobacco benzo (a) pyrene
(13P) refers to results for U.S. snuff
brands, see Table I of the allached paper by
Hoffmann et al (JNCl, 79. 1281-1286.
1987). The authors comment that during the
processing of snuff, especially during
curing, the tobacco can be contaminated with
traces of PAH. In another paper by these
authors (JNCI, 76, 435-437. 1986) this
contamination is attributed to fire-curing or
flue-curing. The former seems more likely, becausp
during fire-curing tobacco comes
into direct contact with wood smoke. Titus for
tobacco ptepared by the more common
air and flue-curing methods, BP levels are likely
to be very low, subject only to air
pollution.
N-nitrosamines are found in cured tobacco at
levels which vary widely depending on
tobacco type, curing process and subsequent
tobacco treatments. The Surgeon
General's data again seerns to reflect snuff
values as indicated by results ill the attached
paper "N-nitrosaniines in tobacco caicinogeilesis"
(Banbury Report 12, 1982).
Generally, lower levels are found in cigarette
tobaccos as shown by the attached table,
"Table 1 N-nitrosamines in commercial tobacco
products" taken from page 748 of CD
C)
"N-nitroso compounds: occurrence, biological
effects and relevance to human cancer" 110
co
CYN
00
C711
BATCO document for Province of BritiSh Columbia 28
October 1999


published by (lie IARC, 1984. Note also that both
tile Baribury and IARC references
givedata for anotherN-nitrosamirie,
N-nitrosoarialabine (NAT) riot listed bythe
Surgeon
General. The Banbury paper (page 219) also
contains the statement:-
*Our major interest in the tobacco-specific
nitrosamines is based on the fact
that these compounds are the only known organic
carcinogens in processed
tobacco . ......
However, the later Surgeon General's report also
records three aldellydes,
1,1,-dimethylhydrazine and ethyl carbaniate in
Table 7.
The aidehyde data are once again derived from
snuff samples and [fie JNCI 1907
reference discussed earlier. From that reference
it is clear that tile acetaldellyde results
are in units of pg/gram, not mg/gram as given by
the Surgeon General's report.
Generally the literature has paid little attention
to the levels of these substances in
tobacco probably because of [tie higher levels in
tobacco SITloke.
The 1, 1 -dinietllylhydrazine data in the Surgeon
General's report are taken from a paper
bySchmeltzetal. (Cancer Letters, 2,125-132,1977).
However, based on the original
reference the units should be ng/graf7i, not
pg/gram. A U.S. commercial blend and a
flue-cured Virginia tobacco gave levels of 60 and
147 ng/gram respectively. The
Surgeon General's table does not include the
result for a Burley tobacco - none was
detected.
The ethyl carbarnate results in the Surgeon
General's Table 7 also originate with
Schmeltz et at (Journal of Analytical Toxicology,
2. 265-268, 1978). Two samples of
Burley leaf, with and without maleic hydrazide
treatment, gave results of 310 and 375
ng/grain respectively.
Table 7 of the Surgeon General's report also cites
seven inorganic "compounds".
Hydrazine results have probably been taken from a
paper by Hecht el at (DHEW
Publication No. (NIH) 76-1221, pages 191-202,
1976). Maleic hydrazide treated Burley
tobacco gave the highest level (51 ng/cigarette)
compared to 22 ng/cigarette for
untreated, hand-suckered Burley tobacco. (Maleic
hydrazide is an anti-suckering
agent). Treated and untreated flue-cured tobaccos
showed no difference in hydrazine
content (12 and 14 ng respectively). A U.S.
Commercial cigarette contained 30 ng.
Table IV of the same reference quotes tobacco
level ranges for arsenic, nickel,
chromium, cadmium, lead and polonium-21 0 which in
most cases are of tile same order
of magnitude as those cited by the Surgeon
General. The latter's polonium-21 0 range
is probably taken from snuff samples analysed by
Hoffmann et al, see Table I of the
1987 JNCI paper cited earlier. C:)
%10
c0
0 I.
00
C11
4-PI
BATCo document for Province of BritiSh Columbia 28
October 1999


Finally, it should be noted that with over 2,550
known compounds in unadulterated
tobacco (see the extract from the Surgeon
General's report) and with variable criteria
for demonstration of a substance's
carcinogenicity, it is not possible to produce a
definitive list on this subject. The attached
Table 7 of the Surgeon General's report is,
however, a very recent attempt to do this and has
been used as the central basis of
this note.
,~, 1) 1 )--
W.D.E. Irwin
C~p
1.0
Co
01%
cc
C7%
Cr%
BATCO doCuMent for Province of BritiSh COIUMbia 28
October 1999