more reeenily. kaljisclmigg kuiojvan studies thai unestigatci ......3.2.4 nicotine and cotinine...
TRANSCRIPT
More reeenily. KalJisclmigg at ul. (2008) conducted a detailed review ui' U.S. and
Kuiojvan studies thai unestigatci; the relation hetween S'l use and non-neoplasiic oral diseases.
HascJ on i i U.S. .-todies, leukoplakia was deleeied in 50 - 70% of moist snuff users, but at
much lower Ireijucncv among users ofchewing tobacco (bngival recession was associated in
several studies with ST use. as noted by Johnson and Slack <2(i(_l); and a possible increase in
rusk of tooth loss was observed lor tobacco chewers, based on a recent large U.S. prospective
study (l)isiTkheLol.,10()j). An increase in dental caries was found for chewing tobacco, but
no! moist snuff (Kalljschrugg; el at., 2008).
hi summary, although there is evidence that ST use causes gingival recession and white
mucosal lesions, cigarette smoking causes far greater damage to periodontal tissues (Johnson &
Slack_2.00i).
2.7.5 Cytologieul Changes Associated with ST Use
Both cigarette smoking and ST may cause cytological changes, termed "leukoplakia"
(literally "white plaque") within the oral cavity (Grady e( uL 1990). According 1o Ri)dji&
(jodshail, 2006, oral leukoplakias occur in up to 60% of ST users, primarily at the site ol'S f
placement. The frequency of leukoplakias depends on the type of ST used, with moist snuff
more often leading to leukoplakias than chewing tobacco, most likely because moist snuff is
morealkaline (RoiIiL&JLilld^ Oral leukoplakiasoccur in around 20% of smokers,
primarily on the undcrsmface of the tongue and in the throat area (Rodu ><•: (mdsh.-ill. 2006).
Dysplasia, a prec-uicerous change, occurs in less than V'-'n of ST leukoplakias out in around .:U%
of leukoplakias due to smoking (Rodtufe GodshalL.2006 and citations therein).
S I leukoplakias rarely progress to cancer. In one prospective study, 1,550 users with
leukoplakias were followed for ten years, and no case of cancer was found (Rodu &. Jan.sson.
2004), and ;i second study reported no case of oral cancer among 500 regular S f users followed
for six years (Christen }')Q}). In contrast, Silverman etai. (1.984) reported that 17% of smoking
leukoplakias transformed into cancer within seven years.
Jn summary, oral leukoplakia occurs commonly in ST users, but it primarily represents
irritation, and very rarely progresses to cancer (Rodu & Godjshal], 2006).
3, Bhtmurker, Preclinical, and Chrmistn Studies
3.1 Introduction
Although the health outcomes from actual use of smokeless tobacco products reviewed in
the previoussection constitute the most direct evidence of disease risk, other "lines pi evidence
are frequently considered in assessing potential risk from use of tobacco products. 1he areas
other than health outcomes proposed by a pane! of tobacco control and public-policy experts lor
use in assessing risk from tobacco products are:
(1) biomarkers of exposure and effect in humans
(2) preclinical cytotoxicity and genotoxicity in cell cultures and animal models
for toxicity and disease
(3) toxic constituents in tobacco products and smoke emissions
(Zdkreta/.. 2009).
The following sections of this appendix provide Information in each of these areas
regarding ST products, with comparison to cigarette smoke or smoking.
Given the diversity of ST products, published data on the chemistry of specific ST
product subcategories arc limited. Hven more limited are in vilro. animal, and human hiomarker
studies across smokeless subcategories. To the extent available, such data for individual product
subcategories are reviewed here. Otherwise the term *'ST" will refer simply to the category of
non-combustible oral tobacco products generally available in the U.S. today.
3.2 Biomarkers Studies in Human Smokers and Smokeless Tobacco Users
3.2.1 Bioimirkers of Exposure
Biomarkersofexposure are generally considered short-term measures of exposure to a
chemical. A biotnarker of exposure is a constituent or metabolite that is measured in a biological
fluid or tissue, or that is measured after it has interacted with critical subcellular, cellular, or
target tissues. Biomarkers of exposure offerthe advantage of providing an aggregate measure of
exposure over hours, days, or weeks, depending on the specific compound's clearance rate
(Ashlev elai. 2010).
for evaluating tobacco exposures, biomarkers may he considered direct iMich .is
cadmium levels in blood or serum) or indirect (such as levels of the tobucco-speeitic nitrosamine
("TSN'A") iuctabolitc. total 4gmelhvlniirosaiuinuH-(3-nyridylH-butanoI (".VNAL") m urine)
measures ofa tolxicco-clcrivxd constituent or metabolite, which can provide some quantitative
and comparative estimate of tobacco exposure (Haisukami t±al. .2006). r?ecause biomuiker
studies examine actual human users of tobacco products, these studies account for differences in
routes of administration and other factors that chemical and in vitro studies cannot.
3.2.2 Biomarkers of Exposure in Human Smokers
Awide range of biomarkers have been assessed in smokers, obtained from a variety of
biological matrices, for example, in exhaled breath, levels of carbon monoxide ([ARC. 2004,
page 1060) and volatile organic compounds such as benzene, 1.3-butadiene. and 2.5-
dinieih) Ifuran are higher in smokers compared with nonsmokers (Gordon ct al. 2002). 1evcls
of cofinine (a nicotine metabolite) in saliva can be used to distinguish smokers from nonsmokers
(Li\RG.-..?flQi- pg- I066). Evaluation of scrum thiecyanate can also be used to distinguish
smokers from smokeless tobacco users (Holiday et al. 1995). A number of urinary markers
have likewise been used to evaluate exposures to cigarette smoke (reviewed in llceht,_2_002:
1JSD1IHS. 2010. p. 230). Urinary levels of nicotine and its metabolites, NNAL and its
glucuronides (metabolites of NNK), correlate generally with the number of cigarettes smoked
per day (insephj.-Aj://., ;_0()5). although saturation may occur ai high levels of smoking. Uiinary
levels of I-HOP (hydroxypyrcne. a pyretic metabolite considered a surrogate of PAH exposure)
may be used as a parameter to detect PAH exposure from cigarette smoking (Heudorf &
AngcTcr_.2Q01). 1,'rinary biomarkers are the most widely applied biomarkers of carcinogen
exposure in smokers (Hecht. 2002). Biomarkers detected in blood or serum include
curboxyhemoglobin (a measure of carbon monoxide exposure), nicotine, and NNAL, all of
which are elevated in smokers (1 ARC. 2004. pp. 1060-1068). A recent study suggests that
urinary levels of total NNAL may also be statistically associated with lung cancer risk in a dose-
dependent manner (Yuan et al. 2009). However. Watanabe el al, (200*)} in a statistictical
evaluation of the relationship between NNK in cigarette smoke and incremental lifetime cancer
risks, reported that NNK would only account for a small proportion of the lung cancer risk
derived from epidemiological data. In addition, Watanabe el at. (2009) offered a view that
complete removal of NNK. NNN. and B(a)P from the smoke of cigarettes would bring little to
78
no reduction in cancer risks due to smoking (Watanabe rt ,ii. 200'f). A number of other potential
bioniiirkers of cigarette smoke exposure have been described, but most are difficult !o Interpret.
since contributions lioni sources other than tobacco exposure {est. diet, other environmental
exposures) and wide differences in individual metabolism may confound results, for example,
oxidative damage to UNA in smokers has been estimated by measurements of R-
oxodooxyguanosine levels tu urine fLofgci Poulscn.J_W8). I be levels of this analv te tire
generally higher in smokers than in nonsmokers, and fall after smoking cessation. Nonetheless,
there is no epidemiological evidence that high levels of oxidative DN.A modification in tissue, or
high urinary excretion of oxidative!}' modilied nucleic acid products, are predictive for cancer
development tit humans (Paulsen, 200:.). Comparisons of smokers and nonsmokers have shown
that total NNAL (NNA1 plus its glueuronide metabolites) is the most useful diseriminatorv
carcinogen hiomarker. since the parent compound. NNK. is found exclusively in tobacco
(USDIIUS, 20in, g. 234).
Anothercommonly assessed hiomarker in smokers js the measurement of carcinogen
DNA .ultlucts. DNA adducts ;<re useful markers of carcinogen exposure, providing an integrated
measurement of carcinogen intake, metabolic activation, and delivcrv to the target
macroinolecule m target tissues (Phlijij;_s._200_5) as well as markers of tobacco-induced DNA
damage (Hecht._2003). According to the 2010 report of the ITS. Surgeon General.
overwhelming evidence indicates that DNA adducts are higher in most tissuesof smokers
compared with nonsmokers (USDH1SS. 2010, p. 245). Because of the difficulties of obtaining
target tissue samples from smokers for carcinogen-DNA adduct analysis, some have considered
.-arejnogen-hernoglobm adducts as a surrogate (reviewed in t.SDHHS. 20)0, p. 245). In spite of
the attractiveness of adducts as biomarkers, limitations exist. While the presence of adducts are
consistent with a carcinogenic hazard, there is of yet no direct relationship between adduct levels
and the level of disease risk (Phillips. 2005).
Urinary mutagenicity in smokers has been documented in a number o\ studies (reviewed
in PeMaririk 2004). Urinary mutagenicity generally correlates with the number ofcigarettes
smoked, but is similar regardless of the level of'tar* derived from combusted tobacco in
cigarettes. Indirect evidence suggests that the chemicals responsible for urine mutagenicity arearomatic or heterocjclie amines.
c
79
1he 2010 report of the U.S. Surgeon (ieneral summarized the overall data regurdiim
biomarkers of exposure irons .smoking, stating that ""several biomarkers piovide an accurate
assessment (4 exposure to toxic chemicals in cigarette smoke." But that it remained to he
determined as to "how accurately they can characterize differences in exposure between tobacco
burning cigarettes and the variety of potentially reduced-exposure products unreduced into Ihe
market during the last few years.'' (USDHHS, 2010. p. 55)
i.23 Biomarkers of Exposure in Smokeless Tobacco Users
Biomarkerstudies of ST users have focused primarily on traditional srnoking-rciated
endpoints. including measurements of nicotine, cotinine. 1SNAs. other tobacco toxicants, and
their metabolites in serum, urine, and saliva. Biomarker-of-exposure studies showed that,
despite different routesof exposure, differentdoses per ir-e, and different pharmacokinetic
patterns, smokers and S I users maintain similar daily plasma nicotine levels (foulds et al. 2003:
1.SRO. 2()n,S, p. 55). S f users generally have higher plasma cotinine levels, but this result mav
reflect rapid metabolism of orally ingested nicotine that occurs in ST u.scrs. albeit to a lesser
extent in smokers (Lbbcrt et «_/_____2004).
Compared with smokers. ST users have higher median levels of NNAL and its
glucuronides per milliliter of urine (Hecht el al, 2007). and possibly higher levels of the
hemoglobin adductof the TSNA 4-hydroxy-U G-pyridyL-l-buianone (IIPR) (Schalller e/ ui..
J22221)- Despite these findings, the urine of ST users is less mutagenic than that of smokers
(Curva.ljg/ ai.. ]987: Bcnowit/. of al. 1989; Sarkar et ai.. 2OJ0). Compared with smokers. S'f
users also have lower levels of serum thiocyanatc, a bioiriarkerof exposure to hydrogen cvanide
(Holiday el al.. 1905). A more detailed discussion of biomarkers of exposure in ST users is
presented below.
3.2.4 Nicotine and Cotinine
Typical U.S. ciyarettes contain about 15-18 mg of nicotine per gram of unburned
tobacco, or an average of 10-12 rng of nicotine in the unburned tobacco per cigarette. The
amount of nicotine absorbed from a cigarette depends on individual smoking behaviors, such as
puff frequency, puJT volume, and inhalation behavior. The amount of smoke (and. therefore,
nicotine) taken in may also be affected by whether ventilation holes are blocked by fingers or
lips. Typically, between I rng and 2 mg of nicotine enter the systemic circulation from a
SO
cigarette, but the deliveries from different styles and as affected by individual smoking behavaus
mav vary between 0.5 mg and over 3 nig of nicotine per cigarette (RCP..20.07. p. 92).
A small study by Benovvitz et al (1988) investigated the systemic dose of nicotine
delivered by moist snuff chewing tobacco, nicotine gum, and cigarette smoke. ST users were
asked to hold 2.5 g of moist snuff or 7.9 g (average) of chewing tobacco in their mouths for 50
minutes; cigarette smokers were asked to take one puff on their usual brand of cigarette every 4s
seconds for 9 minutes (12 puffs); other participants were asked to use two pieces of nicotine
gum. Blood nicotine concentration was then measured at regular intervals. Blood nicotme levels
indicated that smokers rapidly absorbed nicotine through pulmonary circulation and reached
peak blood nicotine concentration in about 10 minutes: the levels subsequently dropped off
quickly as nicotine was distributed from vascular space to tissues. Initial absorption of nicotine
from ST across oral mucous membranes also took place rapidly, and then slowed until peak
blood nicotine concentrations were readied in about 30 minutes, and dropped off gradually.
Although peak nicotine levels seen with cigarette smoking and ST use were similar. S i users
absorbed more total nicotine dm: to their prolonged period of exposure, (figure below; adapted
from Benowitz et ai. 1988)
Moodnicotine concentrations with cigarette smoking and (he use of smokelesstohueeo in single doses. Error bars represent standard error of the mean for 10subjects.
81
In spite of the very different absorption kinetics, regular use of cigarettes or ST gunlucts
throughout the day results in similar blood levels of nicotme. As a consequence. Benow.it/
concluded that any adverse health effects from smoking attributed to nicotine would be similar
for ST users as well.
A 1997 review by Benowitzsummarized data indicating that, although the nicotine
content of ST and cigarettes differ, the systemic absorption and levels of nicotine are similar in
users of smokeless tobacco and cigarette smokers. Benowitz,also concluded that both the levels
and patterns of nicotineexposure during day-long usage by cigarette smokers and users of moist
snuff, as determined by blood analysis, were very similar (Benowitz. I9.S9: table below; adapted
from Benowitz, 1997).
0900 1200 1600 2000 2400 04OQ 0800
TIME
Blood nicotine concentrations with daily cigarette smoking and use ofsmokeless tobacco.Error bars represent standard error of the mean of eight subjects.
Similarly. Curvail et al (1987) examined mean urinary nicotine and cotinine levels in
smokers and Swedish users of snuff. The authors found no significant differences in urinary
nicotine or cotinine levels between moist snuff users and smokers.
Cotinine. a biotnarker for nicotine exposure, is easier to assay for subjects in a clinical
setting since it has a longer half-life than nicotine (15- 24 h vs. 2-3 hours, respectively), and is
present in higher serum concentrations. Higher levels of serum (Naufal et al. 201 1) aud urinaiv
(1 Iccht et al. 2007) cotinine have been reported for ST users, compared with smokers. A 2011
X2
study bv Natilal et ai. used NHAM'.S data to show that levels of serum cotinine were higher in
S'f users than in smokers (table below; adapted from Naomi e.i al, 2011).
Serum Cotinine
Geometric Mean
(95% CI)
Cigarette Smokers
(n=47S0)
1?/.7
(11.9.1,135.6)
Smokeless Tobacco
Consumers
0i=359)138.7
(157.9,735.1)
Non-Consumers of
Tobacco/NRT
_(n=15532["o.oso"
(0.04 7,0.054)
3.2.5 Tobacco-Specific .Sitrosamuies
Some researchers consider TSNAs the most important carcinogens in ST and also in
cigarette smoke (Heeht el al. 2008a). Particular attention has been paid to 4-
(meth\lnitrosamino)-l-(3-pyridyl)-l-biitanone (NNK) and N-nitrosonormcotine (NNN) because
of their levels in tobacco products and (heir potency in rodent carcinogenicity studies. NNK and
NNN have also been listed by 1ARC as carcinogens (ITechi._et_aL _2_008a). After absorption by
tissues, NNK is metabolized to NNAL and glucuronide conjugates of this metabolite. NNAL
and its metabolites may be detected in urine, and have been used to estimate NNK uptake in S J
users.
Users of ST products have been reported to have levels of NNAL analyzed as total
urinary NNAL (NNAL - glucuronides) (jlechl etaf^iHY?') or as NNAL higher than those found
in smokers (Nautal et aj,.._2})])_: Ilatsukand et al •2007), The biological relevance of this finding
is unclear, however, because cigarette smoking is associated with a substantially greater
incidence of cancer than is S'f use. Some researchers have suggested that major conversion ot
TSNAs may occur in the gastrointestinal tract of ST users, (f that suggestion is correct, it is
difficult to conclude that high I SNA levels are primarily responsible for oral cancels in ST users
(LSKO. 2008, p. 58). Similarly, compared with smokers. ST users display higher levels of
hemoglobin adducts of the TSNA 4 hydroxy-1 (Upyridyl)-l-butanonc (flPB) in blood samples
(Scfaaffler el ui, 1993). Although TSNAs in tobacco products receive a great deal of scientific
interest, whether TSNAs from cigarette products are causally related to induction of lung cancer
or ether cancers in humans is not known with certainty. Watanabe el al. conducted a statistical
review of the relationship of NNK in cigarette smoke and incremental lifetime risk for lung
cancer. Conclusions from that work were that NNK could only account for a small proportion
(approximately less than 2%) of the overall risk, and that if NNK and other potential carcinogens
such as NNN and B(a)P were completely removed hum cigarette smoke it likely would bringlittle to no reduction in cancer risks due to smoking (Watanabe <•/ ,7;. ;^)09).
Hecht et ui. recently detcrtiiiued the conversion rate ofNNK to total NNAL in SI" users(14-17%) aud used ibis conversion rate to estimate the total dose ofNNK to an ST user over a
penod of20 years (Hecht et ai_^2u0Ha). fhey estimated that the NNK dose was -60-fokl lower
than the lowest total dose that induced asignificant incidence oflung aud pancreatic tumors inrats when NNK. was given chronically in drinking water.
3.2.6 Other Exposure Biomarkers in Smokeless Tobacco Lsers
A 1995 study Investigating possible biomarkers that could distinguish among nonusers.
ST users, and smokers found that serum levels of thiocyanatc (a biomarker for hydrogen cyanideuptake) were 4.5-fold higher in smokers than in STusers (Holiday et ai. 1995).
"Nautili et at. (20,1 ]_) analyzed NHANES data on 35 blood and urine biomarkers in
cigarette smokers. ST (combined moist snuff and chewing tobacco) users, and nonusers of
tobacco. The cigarette smokers had significantly higher levels of biomarkers for 18 tobacco
constituents (e.g. benzene, styrcne. naphthalene) compared with S I" users. The biomarker levels
in SI users generally resembled those observed in nonusers of tobacco - no significant
differences in 21 ofthe 33 analytcs tested. Smokers, by contrast, differed significantly from
nonusers in levels of 28 biomarkers. Data as described by Naufa! el ai. (20]..!.) are included in
the two following fables.
84
levels of urinary metabolites by tobacco use category (creatinine-corrected)
Level in
nonusef!
of tobacco
AnalyteUnit of
measure
Level inleuftl in ! ~n change in
Snickers smokeless vs smokersusers
1 "iy^ro«ypyrf-r:f
ft OH4:yTl_7 hyrirnxyftuorpnc
[l OH-Hucr!
-Ohydroxytluorere
; >OH ;f,oj
•j hycroxyfiUor'Te
jO-OHTIuor)
T
t-ydroxyohcncrthrfinr
(1 OH P".en)
2
hydroxy ph ens r•t tire r eL'-OH-Pbeni
r'gA 47 K | 6 74 127N
rg/L 196.4 il)! U 767.9
rg/f ""1.5 ; 135 (S 555.6
"S/l :tii<> >.S"; 6 UI.6
r>g/ l 'J 0 !4S4 !<>H6
j frydf cxypherwthrene! j j-OH-Phen)
I 4-j hydrc*yphenar,threne! (4-0H-Phenj: i-• rydroxynspnthc'ler:^
(l-OM-WapM
I 2-ryrtroxyaphtha!pnf»52 CH fysah) j
nj'/L
i-R/L
r-8/L
ns/L
"g/L
17 0
17
1S0S
Arson ic (Total)
(As)
i
pg/L '7.58
Cadmium Mg/L 0.24
; Cobalt
(Co) tig/L j 0.34!
Lead 4g/L 0.5')
iVercury (Total) ^g/L 0.54
4-
(i^ethyinttrosarrn'no)-
1 (4 •pyndyl.-l- iig/l- 0.00 It
butanol
(NNAL)
61) 7
:t».9
17.
1579
18X1
6.1 '"
0,16
0.76
(J.5S "0.76
0.99
•S? 6
:9 6
"!S"'
S9^
8.08
"0J4"
0.33
0.72
"0.42"
0.21
;\';ilne i>siumfitamly lower in Miiykdessnser^compared with ^sicken.Wiluc r=. uot sumfiCiiiily ..fitiereiit benreeu smokele^ u-ers .md ^mokeisVrihie rs Hgiiiilcmtiy lipjier m;.mokeics=; users; compared withsmokers
o i
OSyV
-7^.6'
77.1
\S.9'
-5I.H1
-81.4'
•7';.(;'
-25C?
-52/3£
-21.::
-14.3*
•371.4'
85
levels of blood/serum metabolites by tobacco use category
L&urJ in ' l^yal inLr' *• -f , , .evei in
(ton:,-s&ts >rnokefeiS'Measure
of ?ab.3Cco . ui«i
3.ii:er- ! -i'mL •
roLib-ne" ' Vjrn:
Wrf-P ""7/nL j<-/>x»^>e , nzj'.nl
rtlyfbf'^.-pre ' , '""[;t:r,':h?-;!
'v.,-,1 |
Ac-, .wrie (A/,i :,,-!./ '
i BaJJu-.i 5'fb
WPfjcTi';*? .^Aj j pt^ol/j
(Hs.-'.-rcfu.-ci r* Hb .
i.J.3,.*,5,5,?,5-Otta- jcMs-oti bewo-p- j ^.yi; liu d
iji:j*ir. (O-r.DD. >
l.J,*,-J.5,J,3-Ht'i):3 |C lu.'.jO:rjtr,-^-p.
£.-«.*« rrOstPO:
1.J.3 S.j'.S-Hei.!
diox.n ili>CDDi
cl'br:'d:bf.r/ofi.T,r
Pri/s :'3'J
rc/'g lipiiJ
j =f/l? i'P'd
l.'l.'iA'f.i-ieti- !
rhlrTOfiibenro^j^r i . p/g raid
2M.?.3-Penia-
rg/g l;aid
I
li I:.-
-I. •'!
.1.
t IS
v cr.arge tr
SnnoV.tirs t ^fT^o^ijIew vs ;-r".<,"*
'Value is significantly lower insmokeless us-eis compared with smokers•Vahit: is not sisiitkniiflvdirTeient between smokeless iters ;«ir! smokets
Value is jiaiuficaiiliy tiglit'r m .^iiiokeles.s users compared with sinokeis>
* Note that values of ihcse anaiytes are tower In smokers than m nonusers oftobacco. Whea compared to levels in nonusers ot tobacco, the level of each ofthese anaiyies is increased less than 10% except for HpCDF. which is elevated43.4%.
86
Trine mutagenicity testing hasbeen used as a biomarker lor exposure to environmental
auen'.s. including cigarette smoke. Data reported by Benowitz r-7 al (i%9) indicated that users
ofchewing tobacco and moist snuff as well as nonusers bad significantly lower urine
rnuiatienicifv than smokers. Moist snuff users did not have increased urine mutagenicity
eoiiipared to nonusers; urinary mutagenicity in users of chewing tobacco "tended to be
increased" when compared to that of nonusers. but litis increase was nut statistically significant
at the p '• 0.05 coniidenee level (Bmowhiel al. 17389). Benowitz laterpresented a graphic
illustration td' the low level of urine mutagenicity among users of American chewing tobacco and
moist snuff, compared with smokers, (figure below: adapted from Benowitz (,|W7)
>00-
p too
Sfsl CH SM A SN CH SPJi A
Urinary sodium and mutagenic activity with di/ify cigarette smoking and smokeless use. SN —snuff;CII " chewing tobacco; SM —cigarette smoking: A ~ tobacco abstinence. * —p<0. t)5, compared withabstinence; f: -• p<lhlt), compared with abstinence.
Similarly, Cnrvail et al. !'J_MZ) examined urine mutagenicity in a small sample of
Swedish moist snuff users, smokers, and ncver-users. The study reported no significant
difference in the urinary levels of nicotine and cotinine between moist snuff users and smokers
(table below; from .CuryalK',/ al, J.987).
MFAN I RtKARY LEVELS OF NICOTINE AND COTININE IN SMOfCKRS, SNUFF USEKS AND KON TOBACCO USERS
Subjects
Smukeis in - $}Snuff t)«-rs (n « 8)Abstinent snuff us«! (n - 6)
Nan tobacco users (/i —t)
* M-cm .standard 4cv}ario«).
IJHnarv nicotine
ine/1
1.67 (1.04) *
1.39 (UI)O.iXXl (0.006)3.005 (0.002)
mg/24 tl
2.04 (1.17}
1.00 a. 77)
aois twio)O.tXW (0.002)
Urinary coboioc
ir.g/t
144 (!.;S)
3.46 (U3)0.014(0,007)
af»ft(0.OG5)
mgy?4 h
3.15 (l."6)i\2 fl.39)0.020(0.010)0.0OS «0 Wi5>
87
However urine samples front smokers were far more mutagenic than samples from cither
current miu IT users, snuff users who had abstained from snuff use for one week, or nonusers of
tobacco, fhere were no significant dilferences irt samples from snuff users compared with
never-smokers (table below; from Curva!! et aL 1987).
MUTAGKNtC ACTIVITY OF URINE CONCE^RATESFROM 5MOKF.RS. SNUFF USf.R.S AMD NON TOBACCO LSI KSTOWARDS SALMONELLA STRAIN* TA'JHWITH THE ADDITION OF S9
Sub.ja.Is Revert ants j>c-r ml of unite Ros cfLanis pi?r 11 a (10 J
Mean S.D.
- -
K.ingr M.ean s.n,
iT--
R-m£.e
Smoker* ( n =* M) &.S 4.7-12.8 H.6 i 2-17.6
Snuff b%crs(n»8) 0,« 0.4 0.3- 1.5 1.1 0.8 0.3- 15Abstinent snutf uvs* (n 6) 0.* 0.3 0.4- 1,1 13 0? 0.5- 14Non io£*cco us«s (n —6) 0.5 0.2 0.2- 0.9 0.9 0.7 0.4- 2.1
The study authors concluded that in the case of smoking, normal consumer levels of
nicotine are accompanied by elevated urinary levels of mutagens, while no such increase is
observed for the same nicotine levels among users of Swedish snuff.
hi summary, biomarker of exposure studies shows that ST users, in general, are generally
exposed to lower levels of many tobacco toxicants than are cigarette smokers. These differences
between exposures are consistent with the epidemiological studies showing reduced chronic
disease risk in S'f users compared to smokers.
3.2.7 Hiomarker.? ot'Kt'fcct
Biomarkers of effect are measuresof early biological alterations due to exposure
(Instituteof Medicine. 2001). As such, the 201 0 report of the U.S. Surgeon General has
collectively referred to these alterations us "'biomarkersof biologic events with the potential to
lead to harm" (USDHH^_201_0, p. 54). l-xamples of biomarkers of effect have been provided by
Stratton el al (Institute of Medicine. 2001). some of which are discussed below.
3.2.8 Biomarkers of Kffcct in Smokers
A large number of biomarkers of effect have been described for cigarette smokers. 1hey
encompass gross markers such as cough, osteoporosis, hyperplasia or dysplasia in bronchial or
other tissues, abnormalities in blood, and periodontal disease. Molecular biomarkers of effect
have also been described, including changes in the expression patterns of rnRMA or proteins.
DNA damage such as the occurrence of promoter hypennethykition. and mutations in certain
K8
genes. CiearcUe smokiim has been associated w-itb all of theseconditions, but most are lint
spcciilc to cigarette smoke exposure (TjSDliilS. 2010: p. 5b).
3.2.9 Biomarkers of Kffcet in Smokeless Tobacco Users
Investigations of effects from ST products for potential barm haveconcentrated on one
cateuurv of endpoints relevant to conditionssuspected of being associated with SI use. The
occurrenceof leukoplakia in the oral cavity is considered a biomarkerof damage to oral tissues.
but the verv low- rate at which lesions associated with S 1 use progress to more serious disease,
and the completeness of regression following cessation of ST use suggest that its occurrence in
ST users predicts a different outcome than when found in smokers {see discussion on p.76.
ait]ra). InTmarkers for cardiovascular effects for ST users are thought to be mediated primarily
by nicotine in S I products, and are discussed below.
3.2.10 ( aniiovnseuiar Biomarkers
Cardiovascular biomarkers of effect include blood pressure levels, serum lipid levels,
fibrinolvtie variables, and anti-oxidatit vitamin levels. A recent review' has considered the effect
of S'f use on these and related endpoints (Asplund. 2003). The review cited early studies that
indicated increased blood pressure among oral snuff S'f users during the period of lime that the
smokeless tobacco product was held in the month with a suggestion that an elevated blood
pressure was still present for at least a few minutes after removal of Ihe product from the mouth
(Squires el al. 1.984). In addition, compared to nonusers of tobacco, an increase of diastolic
blood pressure was reported in consumers of smokeless tobacco products, although no details
were provided in the letter to the editor regarding the period between most recent use ami blood
pressure observation (Schroeder & Chen. 1985"). Systolic blood pressure and integrated heart
rate ''tended to be greater" for users of moist snuff and chewing tobacco compared with smokers.
Smoking and ST use resulted in similar maximal increases in heart rate (Benowitz et al.. 1988).
However, more recent studies {e.g. firnstcr et ai.. 1990; Sie.g.eKy' al.. 1.992} have failed to
confirm this finding. Asplund (2003) speculated that changes in the sodium content of snuff
could at least partly explain the discrepant observations in early and later studies, i he absence
of an effect on blood pressure during nonexposure to tobacco is consistent with the reported
absence of sustained hypertension in smokers (Green el al. 1986). Transient increases in both
89
blood pressure and heart rate have been reported after ST use (Benoujty. cial..] 9stS); but. as with
resting blood pressure, resting heart rate appears largely unaffected by S'f use (Asplund. 2003).
S1does not cause elevated levels of hemoglobin or hematocrit, an increase in leukocvte
counts orhigh .sensitivity C-reaeihe protein (two important markers ofsystemic inilammalion
thai are elevated in smokers), impairment ofthe fibrinolytic system, or reduction in circulating
antioxidant vitamins (Aspjuna^2003). Smokers often have a less favorable lipid profile than
nonusers of tobacco. In part due to diets that tend to be high in saturated fats {Aspjumi. .7003).
'ihe lipid profiles of snuff users resemble those of nonusers of tobacco rather than those of
smokers (Asplund. 2003: Siegel. 1992).
In the weight- of-evidence approach taken in the LSRO's evaluation of risks associated
with ditterent categories of tobacco products, changes in lipids, biomarkers of Inilammalion. and
measures of atherosclerosis were weighted more heavily than were changes in blood pressure or
heart rate. The LSRO concluded that ST users appear to have a lower degree ofCVD risk than
smokers (LSRO, 2008 p. 6).
In summary, studies of biomarkers of effect indicate that ST users have far fewer
cardiovascular effects than do cigarette smokers.
3.3 Preclinical Analysis of Cigarette Smoke and Smokeless Tobacco Products
3.3.1 hi Vitro Assays of Cigarette Smoke
fhe potential mechanisms by which cigarettesmokingcause cancerhave been intensely
studied over the last two decades. And a variety of in vilro assays based on the understanding of
those mechanisms have been developed, which provide short-term measurements of biological
changes that serve as indicators of potential harm. Assay methods generally measure
cytotoxicity, cell proliferation, cell cycle control, apoptosis, and genotoxicity (reviewed in
Johnson et al.. 2009). One potential mechanism of cancer induction involves the genotoxicity of
cigarette smoke (DeMarini. 2004). De.Marini reviewed the scientific literature on the
genotoxicity of mainstream cigarette smoke and cigarette smoke condensate, and concluded that
cigarette smoke condensate is genotoxic in nearlyall systems in which it has been tested (e.g.
Ames mutagenicity, sister chromatid exchanges, mieronuclei in bone marrow and lung cells.
DNA adduct, and other genetic abnormalities). DeMarini reported that smoking-associated
90
genotosio effects ba\e been found in ail of the organ sites examined to date at which smoking
causes cancer in humans: ontf'nasai: esophagus: pharynx/larynx; lung: pancreas; myeloid
otgans: bladder/ureter: and, uterine cervix. According to DeMarini. these data support a model of
cigarette smoke carcinogenesis in which the components of smoke induce mutations that
accumulate in fields of tissues that drive the carcinogenic process (DcManni, 200-1) A number
of recent reviews have considered the extensive body of literature regarding the in vitri)
toxicology of cigarette smoke ({.ARC. 2004: USDHllS. 2004: Jo.hnson ct al. 2009: \jSDiUlS._
2(')_[_0). All of these publications document the fact that that cigarette smoke is both cvtotoxic and
genotoxie. and promotes other effects consistent with the known harms associated with smoking.
3.3.2 In Vitro Assays of S'F products
Genotoxicity and cytotoxicity studies of ST products have also been conducted, but S'f
products have been much less thoroughly investigated than cigarette smoke. In addition, the
wide variation in methodologies across in vitro studies of S'f makes it difficult to make
quantitative comparisons between S f products and cigarettes. As pointed out in a
comprehensive review of in vitro toxicity assays of tobacco products (Jofijispn. ej.al.._20Q9).
although the available in vilro assays are reliable as screening tools for qualitative assessments,
they have been poorly validated for quantitative comparisons of different types of tobacco
products. As a consequence, it is especially difficult to interpret older studies that have used a
variety of methodologies for ST extraction and testing. Moreover, the composition of ST
products have changed significantly during the last two decades (eg. significant reduction in
TSNAs: Djordievict7/t//., 199T Hatsukami et al. 2007't and ongoing changes make the older
studies less applicable to the current ST products.
Perhaps the most carefully considered studies relevant to obtaining a datasetfor
contemporary ST products using tests designed and widely used to demonstrate genotoxicity and
cytotoxicity of cigarette smoke condensates are thosereported in Rickert et al. (2007) and
Rickert et al (2009). No moist snttff products were mutagenic in the Ames assay, using strain
f.A98. a strain in which cigarette smoke condensates show a strong dose-response and hiuh
levels ofmutagenic activity (DeMarini 2004). Some positive mutagenic response was observed
using strain TA100, and Rickert et al (2009) made a comparison ofa variety oftobacco
products. As seen in the figure below (from BM.^MmLZQIE, P- 328), chewing tobacco and
moist siiutf products exhibit only a fraction of the mutagenic activity ofcigarette smokecondensate.
TA10CUS9 Revertants per mg nicotine
7."ifi;.. rincv.my icnqcco nfoi••"5 r.iI?p]|Cf'avWiin:nKi jmifl gg 1P9S.»¥ rniSiStrjw ,vin;e:y*->r srn.-ii "pzsg ifiti.'J
r.ipipiso iu'.v-ntoi',iijo snuff
:i2E3:e>s?
Flue ctiied cigaretts bssnEBS355G?i>iPipe t!>tiH'X:0 i t-SSJSIESSS'E&j
Hli.ii 1
Plpo lotwcco 2
Ciqa/ 1 bssEcCicjanilo i p^23&
C^a-ilic 2 CssiESSJSSi
1 " ICtrjar £
KUJ.'I
ispyy |F,|f?8 1
Vtim+MWiMim •niilwii'iiiii^ili Tffi' 1 1\ 13250
:;»:K! 100O3 ILC'OO .'DCvO
TA100*S9 Revertants per rng nicotine2M:-K>
Comparison ofTAIOfH-.S') Revertants per miHigram nieuime for various types of tobacco products.
Other in vitro toxicity tests have been applied to a selection of STproducts, including
moist snuff, chewing tobacco, and snus (Rickert el al. 2009). The results all indicated that, as to
S'f products, mutagenicity determined by Ames testing, clastugenicity determined by the
micronucleus assay, and cytotoxicity determined by the Neutral Red assay, were all less than
10% of those observedas to extracts of mainstream cigarette smoke condensate. These results
are consistent with the finding that differentiates ST" users from cigarette smokers in that the
urine of S'f users is not greater in mutagenic activity than that of non-tobacco users iCurvail et
<>LJl9$1)- One study (Benowitz. 1989) did Ilnd a slight elevation inchewing tobacco users, but
thisfinding was not statistically significant at the p < 0,05 confidence level. (See additional
discussion of urine mutagenicity at p. 87-88.supra.)
In summary, even taking into account the limitations of interpreting in vitro tests of ST
products, it is clear that, in a wide variety of in vitro tests, cigarette smoke is significantly more
genotoxic and cvtotoxic than ST.
92
3.3.3 Animal Bioussavs of Cigarettes and ( igarctte Smoke
the use of experimental animal studies ;o investigate the carcinogenicity of cigarette
smoke and smoke extracts was initiated decades ago. and is still being refined, i he application
of cigarette smoke "tar" to mouse skin with subsequent induction of tumors was first
accomplished in the 1950s (e.g Wvnderc/ al. 1957), Since that time, various skm-painling
protocols have repeatedly demonstrated that cigarette smoke condensate induces cytotoxicity,
cellular proliferation, generation of damaging free radicals, and inflammation as contributors to
tutnor promotion (e.g. Curtin et al... 2004a). Studies have also shown that, in addition to tumor
promotion, cigarette smoke condensate may act as a tumor initiator, tumor accelerator, and co-
carcinogen(Hecht,.2005). Mouse skin painting studies arc currently used to examine the effects
of different tobacco types, processing techniques, or ingredients. The results of such studies are
typically reported in comparison to results from a standardized reference cigarette (l SD1II IS.
20) 0. p. 62).
Historically, animals have not been good models for the types of lung tumors that are
found in human smokers. A wale range of animal species have been utilized in cigarette smoke
inhalation experiments - with only mixed resuhs. Monkeys, dogs, hamsters, and other rodents
have all been used in such experiments: and in some cases, exposures have resulted in tumor
formation. In recent years, a variety of mouse strains, following cigarette smoke exposure by
inhalation, have exhibited some responses with significant increases in lung tumors (Witschi et
aLKHH: Curtin c/,^L,20.04h: IJtitt. e/.///.= 2905). A number of different rat strains have been
utilized in similar inhalation studies, with some reported increases in tumors of the respiratory
tract reported following cigarette smoke exposure (Djijbey_t>/[al, 1980: Mauderly etai.. 2004).
1he2004 fARC Monograph provides a comprehensive .survey of,animal testing of
cigarette smoke and smoke condensates (1ARC. 2004, p. 973). The 201 0 U.S. Surgeon
General's report provides a summary and more recent review ofsuch tests (IJSDHHS. 2OJ0. p.
61). As noted in the 2001 JOM Report, the use ofexperimental animal studies is more
qualitative than quantitative due to both physiological and methodological limitations; and
extrapolation ofresults in such studies to humans is far from concordant (Institute ofMejjkine,
.2001). Nonetheless, decades ofanimal studies support the overwhelming evidence that cigarettesmoke is carcinogenic.
I
i
j 3.3.4 Animal Bioussays of Smokeless Tobacco
Anumber ot animal models have also been used in tests for evaluating carcinogenicity ofsmokeless tobacco, barly studies focused on hamsters, whose cheek pouches seemed a logical
j choice for simulating snuff use in humans. Multiple studies using this model failed to produce1 either tumors ordyspiastic lesions in the cheek pouch ororal cavity (reviewed in Grasso &
Mann. 1998).
In 1981, 1lirsch and Thilander reported a technique whereby an artificial -canal"' was
surgically created in the lower lip ofrats, into which S'for other substances could be placed for
\ extended periods (.Hjjrsch_& Thi lander. 198.1), Althouuh evidence ofepithelial hvpernlasia and| ...dysplasia were observed in their experiment, no oral tumors were produced from long-term
administration of a high dose of powdered snuff.
A 1986 study (Hecht etal. 1986) used a similar "lip canal'' technique in rats to
investigate effects of snuff, and also used an oral swabbing technique to investigate the effects of
snuffextracts, pure preparations of the nitrosamines NNN and NNK. and a mixture of snuff
extract plus nitrosamines. Oral tumors were observed in the 'lip canal" of snuff-treated rats,
although the incidence rate wars not statistically significant compared to controls. No tumors
were observed in rats treated with oral swabbing of a snuffextract. The rats treated with oral
; swabbing of high levels of NNN and NNK exhibited a statistically significant increase in benign
tumors of the oral mucosa, but a lower number of tumors when the nitrosamine solution was
mixed with snuff extract. The authors speculated that snuff extract may actually contain
inhibitors of NNN and NNK activation (Hecht er a/., 1986).
Johansson and colleagues (Johansson el al. 1989). also using the "lip canal" technique,
subsequently reported a low {4 oral squamous cell tumors among29 snuff-treated rats) but
significant incidence of benign and malignant tumors in rat oral tissues. A laterstudy (Johansson
et al. 1991) using rats with a surgically-created 'lip canal", which investigated the promotion
potential of snuff (using a reference moist snuff sample) following chemical initiation, found no
significant promotion of epithelial tumors derived from the lip or oral squamous epithelium. Tne
authors reported a significant increase in lip sarcomas in rats treated with 4-NQO
(tiitroquinolinc-N-oxidc) followed by snuff but not in a group treated with 4-NQO alone.
However. DMBA initiation followed by snuff did not demonstrate a promotional effect of snuff.
74
DMBA alone did not yickl a significant effect on the frequency of lip sarcomas, the authors
reported that .snuff alone increased the number of lip sarcomas although a significant response
was not observed in the previous study. 1he authors speculated that the 4-NQO served as a more
potent tumor initiator than DMBA. Even when the results of studies were somewhat disparate,
the authors concluded however, that the presence of squamous lesions in some experimental
groups suggests that snuff has a weak carcinogenic potential with regard to squamous lesions.
A 1998 comprehensive review of the evidence from animal models regarding S'f and oral
cancer considered that the sum total of experimental work in animal systems suggests that
"snuff (types not specified, but both rnotst and dry snuff described) is not carcinogenic to the
oral mucosa of hamsters or rats (Gra£s<2ife774illl?TJ2^Tsi)- Ifie review1 points to significant Daws
in both the experimental technique (the "lip canal"), .statistical approaches, and experimental
designs in which adequate controls were lacking or where pretreatrnent with excessive amounts
of 4-NQO likely caused localized tissue trauma. Ihe surgical creation of the lip canal results in a
substantial inflammatory response, which leaves a hyperplastic epithelium with scar tissue
formation: such tissues are prone to cancer induction, even when no carcinogens are applied.
And rather than treating control animals with some biologically inert material insert into lip
canals, the investigators generally did not subject the control groups to any further treatment
(Grasso_& Mana 1998). These reviewers considered that the reactive lesionscreated by the lip
canal surgery were sufficient to account for the low levelsof oral tumors that were reported.
Overall, the limited data from animal studies would indicate that the harmful effects of
ST. if any. are significantly lower than those of cigarette smoke or cigarette smoke extracts.
3.4 Chemical Composition of Cigarette Smoke and Smokeless Tobacco Products
Cigarettes and cigarette smokechemistry have been analyzed in considerable detail over
decades (e.g. Dube & Green. 1982; Hoffmann & Hoffmann. 1997). It has been estimated that
the number of chemical constituents in cigarette smoke may be 8700 or more(Rodeman &
Pcrfctti. 2009). Clearly, the burning of a cigarette during smoking notonly transfers materials
that tire present in the cigarette, but also may create many additional new constituents that are
part of the"'vapor phase"' and -paniculate phase" of cigarette smoke (liuk^r, 1999: in Da\is&
Nielsen. 1999; Hojimannc/ ^7,2001).
95
Some reviews ofcigareUc smoke constituents have been conducted with respect topotential health impact (I.owJcs,4J?^n^2Jj03.). for example. Hcehl (2006) reported thepresence of62 carcinogenic compounds In cigarette smoke. 15 of which are considered
carcinogenic in humans. On the basis ofthat list of62 compounds. Hecht estimated that
smoking delivered approximately 1.4 2.2 nig ofcarcinogens per cigarette. Included in the list
01 62 carcmogcns were products ofcombustion such as polycyclic aromatic hydrocarbons(PARs). aromatic amines, heterocyclic aromatic amines, and the volatile hydrocarbons, 1.3-butadiene and benzene.
Studies ofthe chemistry of ST and ST products have generally focused on the relativelysmall number ofconstituents found in ST that are also present in cigarette smoke, and are
thought to play a role in the health effects ofcigarette smoking (e.g. Sjepanov et al.. 2098:
PjiQm.c±al. 2008; Richter.c./.a/., 2008). According to the National Cancer Institute (NCI),smokeless tobacco contains at. least 28 carcinogens m varying concentrations (NCI, 2010, china
\MLL2iU)7- see also lininnenTan.ri..&. Hoffmann. '992). Widely considered to be the most
harmful are the. fSNAs. which are formed during the growing, curing, fermenting, and aging o\~
tobacco. Other carcinogens include polycyclic aromatic hydrocarbons, such as benzo(a)p>Tene:
volatile aldehydes, such as formaldehyde, acetaldehyde, and crotonaiclehyde; hydrazine: heavymetals such as arsenic, nickel, and cadmium; andpokmium-210. Aswith smoked tobacco, ST
contains nicotine.
Clearly, determining the presence ofconstituents in smokeless tobacco and cigarette
smoke provides some insight into potential exposures: but it is also clear that, even though both
product categories contain many of the same toxic constituents, use ofST, compared with
cigarette smoking, results in vastly lower risk of virtually all tobacco-related diseases. A 2008
review by ESRO noted that the obvious major chemical differences between cigarettes and ST
products relate to the burning of cigarettes (LS.RO. 2008 pM 53), Because S'f products are not
combusted, they do not produce cither mainstream orsidestream smoke, and thus expose users to
no combustion products or to very much lower levels of combustion products, many of which
have been strongly implicated in the healthrisks from smoking (ESRO,...2008.p.J 53). As
indicated in the review by ESRO, "ST undoubtedly confers Lower risk of lung disease than
cigarettes and does not expose bystanders to environmental tobacco smoke (LSRO. 2008 p.
153)."
96