TOXICOLOGICAL ASSESSMENT OF TOBACCO INGREDIENTS

1

• TOXICOLOGICAL ASSESSMENT OF TOBACCO INGREDIENTS
• Richard R. Baker
• British American Tobacco
• Southampton
• UK
• LSRO Meeting, Denver, CO, USA
• 8/9 June 2004

2
PLAN OF PRESENTATION

• General aspects, definitions etc.
• Briefly review past work
• Overview of BAT work
• Bioassays
• Pyrolysis
• Smoke chemistry

3

• SOME DEFINITIONS (a)
• Tobacco constituent
• A substance naturally present in tobacco
• Tobacco ingredient
• A substance, generally a flavor material, added
  to tobacco during the cigarette manufacturing
  process

4

• SOME DEFINITIONS (b)
• FLAVORS
• Impart a specific taste, flavor or aroma
• Casings - applied to pre-cut tobacco (few )
• - often recognised foodstuffs
• Flavorings (top flavors)
• - applied to cut and processed
  tobacco (ppm levels, several flavors
  in mixture)

5

• SOME DEFINITIONS (c)
• ADDITIVES
• Used for a specific technological purpose, e.g.
• Humectants increase tobacco moisture-holding
  capacity
• Preservatives protect product deteriation from
  microorganisms
• Binders and strengtheners maintain physical
  state of product
• Fillers contribute to volume without
  contributing to odor, taste or flavor

6
TYPICAL CIGARETTE TOBACCO BLENDS
COMPONENT USA () UK ()
Virginia lamina (flue-cured) 35 75
Burley lamina (air-cured) 26
Oriental lamina (sun-cured) 11
Stem 22
Reconstituted tobacco 25 3
Casings, humectants 2.5
Flavorings 0.5
TOTAL 100 100
7

• Potentially, ingredients can
• Distil into smoke
• Decompose/oxidise and products enter smoke
• Reaction products react with smoke constituents
  and affect their yields and generate other smoke
  products

8
GENERAL ASSUMPTIONS BY HEALTH AUTHORITIES

• Flavor ingredients increase the toxicity of smoke
• Low tar cigarettes have higher levels of flavor
  ingredients than higher yield cigarettes

9
US Surgeon Generals Report, 1979

• In a section discussing technical achievements
  to develop low tar cigarettes, stated
• All of these developments have led to increased
  use of flavor additives, especially for low-tar,
  low-nicotine cigarettes. In fact, these new
  cigarettes require flavor corrections by
  additives in order to be acceptable to the
  consumer.

10
Wrong assumption

• Within British American Tobacco, flavor
  ingredients are not used any more on low tar
  cigarettes than on higher yield cigarettes
• - menthol is an exception
• - its use increases as tar yield decreases

11
STUDIES ON INGREDIENTS SINCE 1950s

• Pyrolysis
• Effects on smoke chemistry
• Mouse skin painting
• Inhalation toxicity
• In vitro bioassays - genotoxicity
• - cytotoxicity

12
PUBLISHED REVIEWS ON TOBACCO INGREDIENTS

• Paschke, Scherer and Heller, 2002
• Rodgman, 2002, two reviews, including much
  previously unpublished RJRT work
• Dixon et al., 2000, effects of ammonia
  ingredients on nicotine transfer and
  bioavailability

13
RECENT MAJOR STUDIES ON INGREDIENTS

• Carmines et al., 2002, four papers chemistry
  and biology
• Gaworski et al., 1997-2002, four papers biology
• Baker et al., 2004, four papers pyrolysis,
  chemistry and biology

14
Paschke, Scherer and Heller

• 198 papers/patents from 1952-2002 on ingredients
  reviewed
• Over 300 ingredients
• Smoke chemistry 150 single ingredients 61
  combinations
• Pyrolysis (161 papers)
• Smoke biological activity (37 papers)

15
Paschke, Scherer and Heller - Conclusions

• Tobacco ingredients used commercially do not
  increase the biological activity of cigarette
  smoke
• Many gaps in knowledge on pyrolysis and transfer
  to smoke
• Standard analytical methods needed for influence
  of ingredients on smoke chemistry

16
Rodgman Reviews - (1) Flavorings -
(2) Casings

• Includes previously unpublished RJRT studies
• Includes work aimed at identifying precursors of
  smoke toxins
• -predicted that relatively volatile flavors
  would distil out of cigarette burning zone
• -studies on ingredients that could potentially
  generate smoke toxins

17
Rodgman - Conclusions

• Neither flavorings nor casing and humectant
  ingredients added to tobacco during commercial
  cigarette manufacture in the USA increase the
  toxicity of cigarette smoke

18
Carmines and co-workers, 2002

• Study of 333 ingredients added to tobacco in 3
  mixtures at normal and 1.5 3 x normal use
• Effects on 51 Hoffmann analytes in smoke
• Effects on Ames and neutral red uptake bioassays
• Effects on sub-chronic inhalation toxicity
  (90-day rat inhalation)

19
Carmines and co-workers - conclusions

• The addition of the 333 ingredients had not
  affected the toxicity of smoke, even in the
  exaggerated high level mixtures.

20
Gaworski et al., 1997 - 2002

• Effects on biological activity of 175 ingredients
  singly and in combinations
• Sub-chronic smoke toxicity (90-day inhalation
  using rats)
• Mouse-skin painting

21
Gaworski et al., conclusions

• Ingredients had no discernible effect on
  inhalation toxicity or tumor-promoting activity
  of smoke

22

• BAT STUDIES
• Pyrolyse in isolation look at products
• 2. Add to cigarette and see what happens to
• smoke chemistry Hoffmann analytes
• 3. In vitro bioassays
• 4. Inhalation toxicity

23

• ADD TO CIGARETTES
• 482 ingredients
• 460 flavors
• 1 flavor/solvent
• 1 solvent
• 7 preservatives
• 5 binders
• 5 humectants
• 1 filler
• 2 process aids (one is water)
• Mixtures added to US blended tobaccos
• 19 Test cigarettes in 3 series made
• 44 Hoffmann analytes determined
• Bioassays and inhalation

24

• CIGARETTE SERIES
• Series A Flavorings
• Series B Flavorings and casings
• Sheet ingredients
• Series C Casings

25

• Inhalation toxicity
• 90-day inhalation with rats
• Series A, B and C cigarettes no
  statistically-significant differences in the
  animals subjected to smoke from the test and
  control cigarettes

26
Cigarette series A Ames test (TA98 S9)
27

• In vitro bioassays on smoke particulate matter
• 1. Genotoxic endpoints
• - Ames
• - Micronucleus bioassay
• 2. Non-genotoxic endpoint
• - Neutral red uptake for cytotoxicity
• None of the test cigarette particulate matters
  produced changes different from their controls

28

• Three approaches to assess chemical
• effects of ingredients
• 1. Add to cigarette and see what happens to
• smoke chemistry
• 2. Pyrolyse in isolation
• 3. Add labelled substance and measure
• labelled products

29

• Approaches in present pyrolysis study
• Develop pyrolysis to simulate conditions during
  smoking
• Use pyrolysis to measure amount of decomposition
  during smoking

30

• SOME DEFINITIONS (1)
• Pyrolysis Decomposition due to heat
• Pyrosynthesis Thermal decomposition of substance
  followed by reaction of their decomposition
  products to form new, larger molecules

31

• SOME DEFINITIONS (2)
• Pyrolysis in inert atmosphere
• Thermal decomposition, pyrosynthetic reactions
  can occur
• Pyrolysis in atmosphere containing oxygen
• Combustion reactions can also occur
• Sometimes called oxygen-sensitised or
  combustion-sensitised pyrolysis

32
Distillation-Pyrolysis Zone
Combustion Zone
Air
33
(No Transcript)
34
PYROLYSIS

• Pyrolysis techniques used in many studies over
  many years to establish component-smoke product
  relationships
• Many false relationships published
• Laboratory pyrolysis conditions must match
  combustion conditions inside cigarette

35
Example (1) of a False Pyrolysis Relationship

• Schmeltz Schlotzhauer (1968) pyrolysed menthol
  at 600C 860C
• They found 22 84 pyrolysed respectively
• The pyrolysis products included phenol
  benzoapyrene
• BUT smoking of cigarettes containing
  radiolabelled menthol, shows that 99 of the
  menthol transfers to the mainstream intact. No
  phenol or benzoapyrene is detected.

36
Example (2) of a False Pyrolysis Relationship

• Schmeltz et al. (1979) pyrolysed labelled
  nicotine added to tobacco in combustion tubes at
  600 - 900C
• The nicotine underwent simple degradation to
  pyridines, and extensive degradation and
  re-arrangement to quinolines, arylnitriles,
  aromatic hydrocarbons.
• They also smoked the cigarettes.
• They found much of the nicotine distilled
  unchanged to MS and SS smoke, small amount of
  simple degradation to pyridines, and no extensive
  degradation.

37
TOBACCO PYROLYSIS - DEVELOPMENT OF AUTHENTIC
CONDITIONS - 1

• Mapped out cigarette combustion conditions
• (Baker, 1970s/1980s)

38
(No Transcript)
39
TOBACCO PYROLYSIS - DEVELOPMENT OF AUTHENTIC
CONDITIONS - 2

• Effect of pyrolysis conditions
• temperature, heating rate, atmosphere
• (Tiller Gentry, 1977 Muramatsu et al., 1979
  Baker, 1980s Stotesbury, 1990s)

40
TOBACCO PYROLYSIS - DEVELOPMENT OF AUTHENTIC
CONDITIONS - 3

• Transfer of labelled substances from cigarette to
  smoke
• (Larson Harlow, 1958 Jenkins et al., 1970s
  Houseman,1973 Schmeltz el al., 1979 Best,1987
  Eble, 1987 J. D. Green et al., 1989 Stevens
  and Borgerding, 1999, Stotesbury et al., 2000)

41
TOBACCO PYROLYSIS CONDITIONS (BAT STUDIES)

• Atmosphere of 9 O2 in N2
• Gas flow of 5 ml/s
• Hold at 300oC for 5 s
• Heat from 300C to 900oC at 30 oC/s
• Hold at 900oC for 5 s

42
Schematic of Pyroprobe interface with GC
Heated interface
Pyrolysis gas in
Probe
Septum purge
Injection port
Split vent
To MS
GC column
43
(No Transcript)
44
Results of pyrolysis versus unchanged labelled
transfer to mainstream smoke
45
USE OF PYROLYSIS IN ASSESSING INGREDIENTS

• Pyrolysis system developed gives good predictions
  of smoke transfer/pyrolytic behaviour of
  relatively volatile tobacco ingredients added to
  cigarette in small amounts
• For involatile substances, the pyrolysis system
  tends to overestimate the amount of decomposition
  that occurs during smoking
• Useful screening tool to indicate which
  ingredients undergo significant decomposition
  during smoking

46
PYROLYSIS OF SINGLE-SUBSTANCE, SEMI-VOLATILE
INGREDIENTS (CUMULATIVE)

• 291 flavour ingredients pyrolysed
• 92 (32) transfer to smoke with lt1 decomposition
• 184 (63) transfer to smoke with lt5
  decomposition
• 248 (85) transfer to smoke with lt20
  decomposition

47
FOR INGREDIENTS THAT DO UNDERGO PYROLYSIS, CAN
CALCULATE MAXIMUM LEVEL OF EACH PYROLYSIS PRODUCT
IN MAINSTREAM SMOKE FOR UNFILTERED CIGARETTE

• Productmax (µg)
• Weight of ingredient in cigarette (µg) max.
  appication level
• x Proportion of product in pyrolysate
• x Proportion of tobacco burnt in puffing 0.5
• x Proportion of transfer of ingredient/product
  to MS smoke 100

48
Examples of maximum pyrolysis yields from
semi-volatile ingredients and cigarette smoke
yields (µg/cigarette)
Ingredient Product Max. level from ingredient Typical smoke level (non-filter cigarette)
Anisyl phenylacetate Phenol 0.03 80 - 160
Benzyl cinnamate Styrene 0.2 10 - 20
Cinnamyl cinnamate Phenol 0.2 80 - 160
a-Methylbenzyl acaetate Styrene 0.1 10 - 20
Phenylacaetc acid Toluene 0.07 100 - 200
p-Tolyl acatate Cresol 0.09 11 - 37
49
FOR SINGLE-SUBSTANCE, SEMI-VOLATILE INGREDIENTS
THAT DO UNDERGO PYROLYSIS

• Hoffmann analytes detected amongst pyrolysis
  products generally low/insignificant compared to
  smoke yields (lt5)

50
Pyrolysis of non-volatile tobacco ingredients

• 159 non-volatile and complex ingredients
• Most ingredients decomposed in the pyrolyser
• many products in small amounts
• significant levels of some Hoffmann analytes
  predicted
• Pyrolysis products with toxicological concern
• - checked by adding ingredient to cigarette
• - smoked by machine
• - comparing smoke yields to control (no
  ingredient) cigarette

51
Comparison of 2-furfural predicted by pyrolysis
and measured by smoking
Ingredient Max. level predicted by pyrolysis (µg/cig) Smoke Analysis Smoke Analysis Smoke Analysis
Ingredient Max. level predicted by pyrolysis (µg/cig) added to cigarette Test cig. yield (µg/cig) Control cig. yield (µg/cig)
Cellulose 410 2.4 7.7 11.0
Sorbitol 5,500 3.6 7.8 8.4
Sugar, brown 6,900 6.2 9.4 11.0
Sugar, invert 11,000 7.0 6.5 4.4
Sugar, white 10,000 10.5 12.8 11.0
Corn syrup 14,000 6.2 5.2 4.4
Honey 2,100 4.5 5.4 4.4
52
For 2-furfural generated from non-volatile
saccharides, pyrolysis experiments have grossly
overestimated the amount formed during
smoking.Pyrolysis also predicts generation of
formaldehyde from saccharide ingredients.
(Formaldehyde not detected by MS system so used
FTIR system.)
53
Generation of formaldehyde during pyrolysis
54
SMOKE CHEMISTRY

• Compare smoke yields of Hoffmann analytes in
  test cigarette (with ingredients) with yields in
  control cigarette (without the ingredients)

55

• CIGARETTE SERIES
• Series A Flavorings
• Series B Flavorings and casings
• Sheet ingredients
• Series C Casings

56
(No Transcript)
57
(No Transcript)
58
(No Transcript)
59
(No Transcript)
60
SMOKE CHEMISTRY RESULTS FLAVORINGS - OVERALL
SUMMARY

• Flavorings have either no significant effect on
  mainstream yields of Hoffmann analytes relative
  to control, or produce occasional changes in
  individual analyte levels ( and -)
• The significance of most of these occasional
  changes were not present when the long-term
  variability of the methodology was taken into
  account
• Conclude that flavorings have no effect on smoke
  chemistry

61
(No Transcript)
62
(No Transcript)
63
SMOKE CHEMISTRY RESULTS CASINGS, SHEET
ADDITIVES - OVERALL SUMMARY

• Usually no significant effect on TPM, nicotine
  and CO
• Some Hoffmann analyte levels affected,
  generally by up to /- 15 , not significant
  within long-term variability
• Significant decreases in nitrosamines (up to
  30), phenols (up to 44), and aromatic amines
  (up to 26) with some mixtures
• Carbonyls significantly increased with some
  mixtures
• - HCHO increased by up to 73 with mixtures
  containing high levels of sugars
• - HCHO increased by 68, possibly due to
  cellulosic and polysccharide materials

64
HCHO YIELDS DIFFERENT STUDIES
INGREDIENT STUDY INCREASE (µg) INCREASE ()
Cellulose BAT 16.1 68
Cellulose NCI (1980) 44 38

Sugar BAT 26.0 73
Sugar - low Carmines et al. (2002) 10.7 65
Sugar - high Carmines et al. (2002) 9.9 60
65
FORMALDEHYDE YIELDS FOR CIGARETTES WITH TAR
YIELD OF ca. 13 mg

• Experimental cigarette 62 µg
• UK benchmark study 30 - 56 µg
• World study 30 - 90 µg

66
BAT STUDY - CONCLUSIONS - 1

• 2/3 of volatile flavorings transfer to smoke with
  lt5 decomposition
• Where decomposition does occur, Hoffmann
  analytes detected amongst products generally
  low/insignificant compared to smoke yields (lt5)
• Non-volatile ingredients generally decompose in
  pyrolyser and pyrolysis experiments overestimate
  amount of compounds formed during smoking

67
BAT STUDY - CONCLUSIONS - 2

• Flavorings have no significant effect on levels
  of Hoffmann analytes in mainstream smoke
• The vast majority of casings and sheet
  ingredients have little effect on level of
  Hoffmann analytes in smoke. Several are
  decreased and one is increased.

68
BAT STUDY - CONCLUSIONS - 3

• The inhalation toxicity of the smoke from all the
  test cigarettes was the same as that from their
  respective control cigarettes
• Within the sensitivity and specificity of three
  in vitro bioassays, the specific activity of
  smoke condensate was not changed by the addition
  of ingredients to the cigarette
• -Ames test
• -Mammalian cell micronucleus assay
• -Neutral red uptake cytotoxicity assay

69
OVERALL CONCLUSIONS

• THERE IS BROAD AGREEMENT BETWEEN
• Chemical and biological studies published over 50
  years (Paschke et al., 2002, Rodgman, 2002)
• Chemical and biological work undertaken by R.J.
  Reynolds (included in the Rodgman reviews)
• Philip Morris chemical and biological studies
  (Carmines et al., 2002)
• Lorillard biological studies (Gaworski et al.,
  1997 2002)
• BAT pyrolysis, smoke chemistry and biological
  studies (Baker et al., 2004)

70
BROAD CONCLUSIONS

• Tobacco ingredients used commercially do not
  increase the biological activity of cigarette
  smoke
• Most ingredients do not affect the smoke levels
  of Hoffmann analytes

71

• BAT PAPERS ON INGREDIENTS
• R.R. Baker and G. Smith, Toxicological aspects of
  tobacco flavour ingredients, Recent Advances in
  Tobacco Science, 2003, 29, 47-76.
• R.R. Baker and L.J. Bishop, The pyrolysis of
  tobacco ingredients, J.Anal.Appl. Pyrolysis,
  2004, 71(1), 223-311.
• R.R. Baker, J.R. da Silva and G.Smith, The effect
  of tobacco ingredients on smoke chemistry. Part
  I Flavourings and additives, Food Chem. Toxicol,
  2004, 42 Supplement, 3-37.
• R.R. Baker, J.R. da Silva and G.Smith, The effect
  of tobacco ingredients on smoke chemistry. Part
  II Casing ingredients, Food Chem. Toxicol, 2004,
  42 Supplement, 39-52.
• R.R. Baker, E.D. Massey and G.Smith, An overview
  of the effects of tobacco ingredients on smoke
  chemistry and toxicity, Food Chem. Toxicol, 2004,
  42 Supplement, 53-83.
• R.R. Baker and L.J. Bishop, The pyrolysis of
  non-volatile tobacco ingredients using a system
  that simulates cigarette combustion conditions,
  Paper presented at 16th International Symposium
  on Analytical and Applied Pyrolysi, Alicante,
  Spain, May 2004.
• 7. R.R. Baker, S. Coburn, C. Liu and J. Tetteh,
  Pyrolysis of eleven polysaccharide tobacco
  ingredients a TGA-FTIR investigation, Paper
  presented at 16th International Symposium on
  Analytical and Applied Pyrolysi, Alicante, Spain,
  May 2004.