Evaluation of the Tobacco Heating System 2.2. Part 2: Chemical composition, genotoxicity, cytotoxicity, and physical properties of the aerosol Academic research paper on "Earth and related environmental sciences"

Accepted Manuscript

Evaluation of the Tobacco Heating System 2.2. Part 2: Chemical composition, genotoxicity, cytotoxicity, and physical properties of the aerosol

Jean-Pierre Schaller, Daniela Keller, Laurent Poget, Pascal Pratte, Etienne Kaelin, Damian McHugh, Gianluca Cudazzo, Daniel Smart, Anthony R. Tricker, Lydia Gautier, Michel Yerly, Roger Reis Pires, Soazig Le Bouhellec, David Ghosh, Iris Hofer, Eva Garcia, Patrick Vanscheeuwijck, Serge Maeder

PII: S0273-2300(16)30290-2

DOI: 10.1016/j.yrtph.2016.10.001

Reference: YRTPH 3693

To appear in: Regulatory Toxicology and Pharmacology

Received Date: 4 July 2016 Revised Date: 4 October 2016

Accepted Date: 5 October 2016

Please cite this article as: Schaller, J.-P., Keller, D., Poget, L., Pratte, P., Kaelin, E., McHugh, D., Cudazzo, G., Smart, D., Tricker, A.R., Gautier, L., Yerly, M., Pires, R.R., Le Bouhellec, S., Ghosh, D., Hofer, I., Garcia, E., Vanscheeuwijck, P., Maeder, S., Evaluation of the Tobacco Heating System 2.2. Part 2: Chemical composition, genotoxicity, cytotoxicity, and physical properties of the aerosol, Regulatory Toxicology and Pharmacology (2016), doi: 10.1016/j.yrtph.2016.10.001.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Evaluation of the Tobacco Heating System 2.2. Part 2: Chemical composition, genotoxicity, cytotoxicity, and physical properties of the aerosol

Jean-Pierre Schaller1,2, Daniela Keller1, Laurent Poget1, Pascal Pratte1, Etienne Kaelin1, Damian McHugh1, Gianluca Cudazzo1, Daniel Smart1, Anthony R. Tricker1, Lydia Gautier1, Michel Yerly1, Roger Reis Pires1, Soazig Le Bouhellec1, David Ghosh1, Iris Hofer1, Eva Garcia1, Patrick Vanscheeuwijck1, Serge Maeder1

1Philip Morris International R&D, Philip Morris Products S.A., Quai Jeanrenaud 5, 2000 Neuchatel, Switzerland (part of Philip Morris International group of companies) Corresponding author, e-mail: jean-pierre.schaller@pmi.com; Tel.: +41 (58) 242 26 82

Keywords: Tobacco heating system, Heat-not-burn, THS2.2, Modified risk tobacco product, Harmful and potentially harmful constituents, HPHC, Aerosol Chemistry, Mutagenicity, Cytotoxicity.

Highlights:

• Mainstream aerosol characterization of the Tobacco Heating System 2.2 (THS2.2)

• Comparison of the THS2.2 aerosol with the smoke of the 3R4F reference cigarette

• The majority of the toxicant yields were reduced by more than 90% in the THS2.2

• The THS2.2 in vitro genotoxic and cytotoxic potencies were reduced by about 90%

• The particle sizes of THS2.2 aerosol and cigarette smoke were similar

Word counts: Abstract: 199, Text: 17855, References: 3048

25 The chemical composition, in vitro genotoxicity, and cytotoxicity of the mainstream aerosol from the

26 Tobacco Heating System 2.2 (THS2.2) were compared with those of the mainstream smoke from the

27 3R4F reference cigarette. In contrast to the 3R4F, the tobacco plug in the THS2.2 is not burnt. The low

28 operating temperature of THS2.2 caused distinct shifts in the aerosol composition compared with

29 3R4F. This resulted in a reduction of more than 90% for the majority of the analyzed harmful and

30 potentially harmful constituents (HPHCs), while the mass median aerodynamic diameter of the aerosol

31 remained similar. A reduction of about 90% was also observed when comparing the cytotoxicity

32 determined by the neutral red uptake assay and the mutagenic potency in the mouse lymphoma assay.

33 The THS2.2 aerosol was not mutagenic in the Ames assay. The chemical composition of the THS2.2

34 aerosol was also evaluated under extreme climatic and puffing conditions. When generating the

35 THS2.2 aerosol under "desert" or "tropical" conditions, the generation of HPHCs was not significantly

36 modified. When using puffing regimens that were more intense than the standard Health Canada

37 Intense (HCI) machine-smoking conditions, the HPHC yields remained lower than when smoking the 3 8 3R4F reference cigarette with the HCI regimen.

Abbreviations: 4-(#-nitrosomethylamino)-1-(3-pyridyl)-1-butanone (NNK), Electrically Heated Cigarette Smoking System (EHCSS), flame ionization detection (FID), gas chromatograph-mass spectrometer (GC/MS), gas chromatography (GC), gas-vapor phase (GVP), global evaluation factor (GEF), harmful and potentially harmful constituent (HPHC), Health Canada Intense (HCI), inductively coupled plasma mass spectrometry (ICP-MS), International Agency for Research on Cancer (IARC), limit of quantification (LOQ), liquid chromatograph tandem mass spectrometer (LC-MS/MS), liquid chromatography (LC), lower boundaries (LB), lowest observed genotoxic effect level (LOGEL), mass median aerodynamic diameter (MMAD), Modified Risk Tobacco Product (MRTP), mouse lymphoma assay (MLA), mutation frequency (MF), neutral red uptake (NRU), nicotine-free dry particulate matter (NFDPM), nitrogen oxides (NOx), ^V-nitrosoanabasine (NAB), ^-nitrosoanatabine (NAT), ^-nitrosonornicotine (NNN), phosphate-buffered saline (PBS), programmable dual syringe pump (PDSP), relative humidity (RH), relative total growth (RTG), Intermediate Precision (IP), Tobacco Heating System 2.2 (THS2.2), total particulate matter (TPM), upper boundaries (UB),

39 1 INTRODUCTION

40 The U.S. Family Smoking Prevention and Tobacco Control Act (FSPTCA) defines a Modified Risk

41 Tobacco Product (MRTP) as "any tobacco product that is sold or distributed for use to reduce harm or

42 the risk of tobacco related disease associated with commercially marketed tobacco products"

43 (Family Smoking Prevention and Tobacco Control Act). This publication is part of a series of nine

44 publications describing the nonclinical and part of the clinical assessment of a candidate MRTP,

45 THS2.2 regular and a mentholated version (THS2.2M). The series of publications provides part of the

46 overall scientific program to assess the potential for THS2.2 to be a reduced risk product. The first

47 publication in this series describes THS2.2 and the assessment program for MRTPs (Smith et al.,

48 submitted (this issue)). This is followed by six publications, including this one, that describe the

49 nonclinical assessment of THS2.2 regular and THS2.2M (Kogel et al., submitted (this issue); Oviedo et

50 al., submitted (this issue); Schaller et al., submitted (this issue)-a; Schaller et al., submitted (this issue)-

51 b; Sewer et al., submitted (this issue); Wong et al., submitted (this issue)). The eighth publication in the

52 series describes a clinical study to assess whether the reduced formation of Harmful and Potentially

53 Harmful Constituents (HPHCs) for THS2.2 regular also leads to reduced exposure to HPHCs when the

54 product is used in a clinical setting (Haziza et al., submitted (this issue)). A final publication utilizes

55 data gathered from the reduced exposure clinical study on THS2.2 regular to determine if a systems

56 pharmacology approach can identify exposure response markers in peripheral blood of smokers

57 switching to THS2.2 (Martin et al., submitted (this issue)). This publication is the second of the series

58 and presents the chemical analyses, the physical characterization, and the in vitro genotoxicity and

59 cytotoxicity assessments of the mainstream aerosol of the THS2.2.

61 The smoke produced by the combustion of tobacco in a combustible cigarette (CC) is a complex and

62 dynamic chemical mixture which contains more than 8000 identified chemical compounds (Rodgman

63 and Perfetti, 2013). Tobacco smoke from CC consists of an aerosol containing liquid droplets

64 ('particulate phase') suspended in a carrier gas and surrounded by its own gas-vapor phase. It is

65 generated by complex and overlapping burning, pyrolysis, pyrosynthesis, distillation, sublimation, and

66 condensation processes (Borgerding and Klus, 2005). With minor exceptions, both pyrogenesis and

67 pyrosynthesis of HPHCs result from the thermal decomposition of organic compounds present in

68 tobacco occurring at temperatures up to 900°C observed in cigarettes (Torikaiu et al., 2005; Baker,

69 2006); thus, a reduction of these toxicants may be achieved by heating rather than burning tobacco to

70 produce an aerosol. The development of heat-not-burn tobacco products is not new and earlier efforts

71 to develop such products (notably Premier and Eclipse products from The R.J. Reynolds Tobacco

72 Company and Accord from Philip Morris) have been reviewed by Baker (Baker, 2006). The

73 Electrically Heated Cigarette Smoking System (EHCSS) was the first generation of tobacco heated

74 products commercialized by Philip Morris. The EHCSS series-E has been subject to extensive

75 analytical and toxicological evaluation, demonstrating simplified smoke chemistry compared with the

76 1R4F reference cigarette of the University of Kentucky (Patskan and Reininghaus, 2003). Notably

77 there was a significant reduction in carbon monoxide (CO) and an increased yield of formaldehyde in

78 EHCSS-E mainstream smoke, compared with the 1R4F cigarette. On a per-milligram total particulate

79 matter (TPM) basis, the concentration of formaldehyde was increased approximately sevenfold

80 (Stabbert et al., 2003). In later developments of EHCSS (series-JLI and series-K), in order to reduce

81 these excessive levels of formaldehyde, ammonium magnesium phosphate (AMP) was used in the

82 cigarette paper to replace calcium carbonate. It was anticipated that ammonia released during the

83 pyrolysis of AMP would condense with formaldehyde to form hexamethylenetetramine (HMT)

84 (Schorp et al., 2012). Chemical analysis of smoke from the EHCSS-JLI and EHCSS-K cigarettes

85 containing AMP showed lower yields of formaldehyde and several HPHCs, a further decrease in CO

86 yield, and increased yields of ammonia and HMT (Roemer et al., 2008; Werley et al., 2008; Zenzen et

87 al., 2012). The THS2.2 is the latest generation of heat-not-burn products from Philip Morris

88 International. It produces an aerosol by carefully heating the tobacco with a heater blade reaching a

89 maximum temperature of 350°C. This system enables a careful control of the energy applied to the

90 tobacco plug (Smith et al., submitted (this issue)) and limits the thermal physicochemical processes

91 while producing an aerosol capable of satisfying adult smokers enabling them to switch from

92 cigarettes.

94 Although the causal relationship between smoking and several diseases is well established (Doll et al.,

95 2004), there is still very little understanding of the underlying mechanisms by which smoking causes

96 disease. Among the more than 8000 chemical compounds that have been identified in cigarette tobacco

97 smoke (Rodgman and Perfetti, 2013), public health authorities and others have proposed some 100

98 HPHCs as possible causes of smoking-related diseases such as lung cancer, heart disease, and

99 emphysema (Health Canada, 2000; World Health Organisation, 2008;

100 U.S. Food and Drug Administration, 2012). For the US Food and Drug Administration, the notion of

101 "harmful and potentially harmful constituent" includes any chemical or chemical compound in a

102 tobacco product or in tobacco smoke that is, or potentially is, inhaled, ingested, or absorbed into the

103 body, including as an aerosol (vapor) or any other emission; and causes or has the potential to cause

104 direct or indirect harm to users or non-users of tobacco products (U.S. Food and Drug Administration,

105 2016). However, there is no consensus, that lowering or eliminating any single compound (or even a

106 combination of compounds) in smoke would have a significant impact on risk. The current approach,

107 which eliminates direct tobacco combustion and limits tobacco pyrolysis by heating at significantly

108 lower temperatures than encountered in CC, has the potential to reduce a broad range of HPHCs in the

109 THS2.2 aerosol. Consequently, criteria were established to develop a list of relevant analytes, including

110 HPHCs to assess their reductions in the THS2.2 aerosol, compared to the mainstream smoke of the

111 University of Kentucky reference cigarette 3R4F, as follows:

113 • Criterion 1: Smoke constituents determined by International Organization for Standardization (ISO)

114 methods. This list includes total particulate matter (TPM)

115 (International Organisation for Standardization, 2011), water in TPM

116 (International Organisation for Standardization, 2011), nicotine

117 (International Organisation for Standardization, 2013), nicotine-free dry particulate matter (NFDPM)

118 (International Organisation for Standardization, 2011); carbon monoxide (CO)

119 (International Organisation for Standardization, 2010b) and benzo[a]pyrene

120 (International Organisation for Standardization, 2012).

122 • Criterion 2: Priority toxicants in tobacco smoke selected from the lists issued by regulatory bodies, or

123 proposed by cognizant authorities (Health Canada, 2000; World Health Organisation, 2008;

124 U.S. Food and Drug Administration, 2012). This list includes the analytes recommended by ISO

125 under Criterion 1. In addition, the following HPHCs are also included: 1-aminonaphthalene, 2126 aminonaphthalene, 3-aminobiphenyl, 4-aminobiphenyl, acetaldehyde, acetone, acrolein,

127 butyraldehyde, crotonaldehyde, formaldehyde, methyl ethyl ketone, propionaldehyde, acrylonitrile,

128 1,3-butadiene, benzene, isoprene, pyridine, quinoline, styrene, toluene, catechol, m-cresol, p-cresol, o-

129 cresol, hydroquinone, phenol, resorcinol, N-nitrosoanabasine (NAB), N-nitrosoanatabine (NAT), 4-

130 (^-nitrosomethylamino)-1-(3-pyridyl)-1-butanone (NNK), ^-nitrosonornicotine (NNN), ammonia,

131 hydrogen cyanide, nitric oxide (NO), nitrogen oxides (NOx), arsenic, cadmium, chromium, lead,

132 mercury, nickel, and selenium.

134 • Criterion 3: Toxicants with an established biomarker of exposure, i.e., for use in a clinical study to

135 determine exposure to the parent toxicant (Haziza et al., submitted (this issue)). The toxicants include

136 some analytes already listed under "Criterion 1" and "Criterion 2": CO (biomarker: blood

137 carboxyhemoglobin (COHb) (Pojer et al., 1984), nicotine (biomarker: serum cotinine (Benowitz and

138 Iii, 1984) or total nicotine equivalents in urine (Benowitz and Jacob, 1994)), 2-aminonaphthalene, 4139 aminobiphenyl and o-toluidine (biomarker: parent amines in urine (Riedel et al., 2006)), acrolein

140 (biomarker: 3-hydroxypropylmercapturic acid (3-HPMA) in urine (Mascher et al., 2001)),

141 crotonaldehyde (biomarker: 3-hydroxy-2-methylpropyl mercapturic acid (HMPMA) in urine (Scherer

142 et al., 2007)), acrylonitrile (biomarker: 2-cyanoethylmercapturic acid (CEMA) in urine (Minet et al.,

143 2011)), acrylamide (biomarker: acrylamide mercapturic acid (AAMA) in urine (Urban et al., 2006)),

144 1,3-butadiene (biomarker: 1-hydroxy-2-(N-acetyl-cysteineyl)-3-butene (MHBMA) in urine (van

145 Sittert et al., 2000)), benzene (biomarker: S-phenylmercapturic acid (S-PMA) in urine (Medeiros et

146 al., 1997)), NNK (biomarker: total 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) in urine

147 (Carmella et al., 2003)), NNN (total NNN in urine (Kavvadias et al., 2009)), benzo[a]pyrene and

148 pyrene (biomarker: total 1-hydroxypyrene (1-OHP) in urine (Strickland et al., 1996)).

150 • Criterion 4: Toxicants which are predominantly formed below 400°C, and which are not included

151 under "Criterion 2": acrylamide (possibly formed from asparagine and reducing sugars through a

152 Maillard type of reaction occurring between 120 and 200°C (Stadler et al., 2002; Blank et al., 2005;

153 Becalski et al., 2011)), acetamide (possibly formed from the pyrolysis of Amadori compounds

154 (formed by the reaction of amino acids and sugars) and from the decomposition of ammonium acetate

155 at around 250°C (Moldoveanu, 2010)), propylene oxide (possibly formed by dehydration of propylene

156 glycol which is used as a humectant in CC and for application of flavors to tobacco (Diekmann et al.,

157 2006; Laino et al., 2012)), nitrobenzene, ethylene oxide and vinyl chloride (the source of the 3 last

158 compounds is less clear but ethylene oxide and vinyl chloride are both classified by the International

159 Agency for Research on Cancer (IARC) as Group 1 (carcinogenic to humans), and nitrobenzene as

160 Group 2B (possibly carcinogenic to humans)).

162 • Criterion 5: Toxicants which are predominantly formed above 400°C, and which are not included

163 under "Criterion 1" and "Criterion 2": dibenz[a,h]anthracene and benz[a]anthracene (McGrath et al.,

164 2007).

166 In addition, glycerin (a humectant used during CC and Tobacco stick manufacturing) and menthol (for

167 mentholated products only) were also quantified. This results in a total of 59 analytes for THS2.2

168 regular products (60 for mentholated products) that were quantified to perform the chemical assessment

169 of the THS2.2 aerosol. Among them, 54 are HPHCs targets for reduction compared to 3R4F when

170 developing heat-not-burn products: carbon monoxide, benzo[a]pyrene, 1-aminonaphthalene, 2171 aminonaphthalene, 3-aminobiphenyl, 4-aminobiphenyl, acetaldehyde, acetone, acrolein, butyraldehyde,

172 crotonaldehyde, formaldehyde, methyl ethyl ketone, propionaldehyde, acrylonitrile, 1,3-butadiene,

173 benzene, isoprene, pyridine, quinoline, styrene, toluene, catechol, o-cresol, m-cresol, p-cresol,

174 hydroquinone, phenol, resorcinol, NAB, NAT, NNK, NNN, ammonia, hydrogen cyanide, nitric oxide,

175 nitrogen oxides, arsenic, cadmium, chromium, lead, mercury, nickel, selenium, pyrene, o-toluidine,

176 acetamide, acrylamide, ethylene oxide, nitrobenzene, propylene oxide, vinyl chloride,

177 benz[a]anthracene and dibenz[a,h]anthracene. As previously mentioned, this list was mainly based on

178 analytes proposed by public health authorities and it covers a large range of potential toxicants

179 identified in cigarette smoke. However, the scientific literature continues to describe new compounds

180 with a potential toxicity mainly in aerosols from new products. For instance, only recently has glycidol

181 been identified as a potential toxic compound in aerosols from electronic cigarettes (Sleiman et al.,

182 2016). The present list of HPHCs was based on knowledge available at the time of designing the

183 studies and glycidol, which was not included in the FDA list (U.S. Food and Drug Administration,

184 2012), was not identified at the time as a potential target to be included.

186 In addition to the 54 HPHCs listed above, TPM, water, nicotine, NFDPM, glycerin and menthol (for

187 mentholated products only) are used to assess product performance but are not targets for reduction

188 compared to the 3R4F. These analytes are major aerosol constituents for which the level has to be

189 maintained in order to provide satisfactory sensory properties. For instance, nicotine is addictive and

190 has toxic properties. However, at the levels nicotine is consumed from tobacco products, it is not

191 considered to be a primary cause of smoking related disease (The Royal College of Physicians, 2016).

192 As noted by the Royal College of Physicians (The Royal College of Physicians, 2016) "The ideal

193 harm-reduction device should therefore deliver nicotine in a manner as similar as possible to cigarettes,

194 while at the same time maximizing palatability and nicotine delivery to approximate the experience of

195 cigarette smoking more closely." Following this logic, the nicotine content in the delivered aerosol was

196 carefully monitored and the nicotine level was by design maintained at an appropriate level. It is

197 important to note that, although TPM and NFDPM are analytes that were originally defined in the

198 context of CC smoke, they can also be used in the context of aerosol produced by heating tobacco, as

199 they characterize the mass of aerosol delivered. However, TPM and NFDPM yields obtained from

200 heated products should not be directly compared to the values obtained for cigarettes, as the smoke

201 produced by combustion and the aerosol produced by heating tobacco have substantially different

202 chemical compositions.

204 The exposure to HPHCs may also be affected by the physical properties of the inhaled aerosol. The

205 particle/droplet size distribution determines the fraction of aerosol or smoke that is able to pass through

206 the upper respiratory tract to reach the lungs and the fraction that is retained in the respiratory system

207 (Robinson and Yu, 2001; Bernstein, 2004; Kane et al., 2010). An aerosol is considered respirable when

208 the mass median aerodynamic diameter (MMAD) calculated from the measured size distribution is

209 below 2.5 micrometers (Hinds, 1999). In the past decades, several instruments have been used to

210 determine the size distributions of CC mainstream smoke, and more recently, of e-cigarette aerosols as

211 well (Singh et al., 2006; Ingebrethsen et al., 2012; Fuoco et al., 2014; Geiss et al., 2015). These

212 analytical techniques have been shown to have both advantages and drawbacks in measuring the

213 aerosol physical parameters. Past studies revealed that different methodologies led to size distributions

214 ranging from 0.2 to 0.9 micrometers for CC smoke (Ishizu et al., 1978; Davies, 1988). This wide range

215 of droplet sizes obtained for CC was presumably due to particular aerosol sampling and dilution

216 methodologies. The methodologies enabling a real-time aerosol physical characterization used the

217 measured aerosol number size distribution to calculate the related mass concentration distribution and

218 the MMAD, assuming spherical particles or droplets with a chosen arbitrary density. Consequently, the

219 conversion from number to mass distribution could lead to the overestimation or underestimation of the

220 MMAD. To minimize these problems, the selected physical characterization of the aerosol was based

221 on multistage cascade impactor technology, allowing a determination of the MMAD. The impactor

222 technique enabled the gravimetrical classification of aerosol droplets in distinct size classes, and was

223 often associated with the related aerosol deposition behavior in lungs (Hinds, 1999). Additionally, the

224 technique allowed the determination of the MMAD assuming equivalent unit density and spherical

225 droplet shapes.

227 The development of a new product such as the THS2.2 should be accompanied not only by the analysis

228 of its chemical composition but also by in vitro and in vivo toxicological assays, because current

229 knowledge is insufficient for predicting the effect of complex mixtures based solely on the chemical

230 composition (Carchman, 1988; Institute of Medicine, 2001; Tewes et al., 2003). To further extend the

231 characterization of the THS2.2 aerosol, both the cytotoxic and mutagenic activity of the aerosol were

232 compared to the activities of mainstream smoke of the reference cigarette 3R4F. Furthermore, two 90233 day rodent inhalation studies are also reported in this issue (Sewer et al., submitted (this issue); Wong

234 et al., submitted (this issue)).

236 The current studies report the aerosol characterization from 4 different versions of the THS2.2 Tobacco

237 Sticks: two THS2.2 Regular (THS2.2 FR1 and THS2.2 D2) and two THS2.2 Menthol (THS2.2 FR1 M

238 and THS2.2 D1 M). These studies include the cytotoxic and genotoxic activities of the mainstream

239 THS2.2 aerosol compared to mainstream smoke of the reference cigarette 3R4F and a comprehensive

240 physical and chemical characterization of the produced aerosols. For both types of aerosols, the

241 cytotoxicity of the gas-vapor phase (GVP) and the particulate phase (TPM) was assessed using the

242 neutral red uptake (NRU) assay (Borenfreund and Puerner, 1985) and the genotoxicity was assessed

243 using both the Salmonella typhimurium reverse mutation assay (TPM only) (Ames (Ames et al., 1973))

244 and the mouse lymphoma assay (MLA) (Clive et al., 1972). The physicochemical characterization was

245 based on the comparison of the droplet diameter of both aerosols and the presence of 54 HPHCs in

246 each aerosol using the Health Canada Intense (HCI) machine-smoking protocol (Health Canada, 2000).

247 In addition, the THS2.2 aerosol chemical composition was assessed under different climatic conditions

248 (temperature and relative humidity) and with different machine-smoking regimens to simulate different

249 use than described in the HCI machine-smoking protocol.

251 2 MATERIALS AND METHODS

252 2.1 Reference cigarette

253 The 3R4F reference cigarette was obtained from the University of Kentucky (Lexington, KY, USA;

254 http://www.ca.uky.edu/refcig/).

256 2.2 THS2.2 tobacco stick

257 The tobacco stick was designed to be used with the THS2.2 holder (Smith et al., submitted (this issue)).

258 Its construction is shown in Figure 1.

259 A tobacco stick is constructed by the sequential assembly of the following components:

260 1. Tobacco plug wrapped in a paper over-wrap

261 2. Hollow acetate tube wrapped in a paper over-wrap

262 3. Polymer-film filter, wrapped in a paper over-wrap

263 4. Mouthpiece filter wrapped in a paper over-wrap

264 All these elements are wrapped in an outer paper, and a tipping paper is added on the mouth end

265 (Figure 1).

266 Approximate position for Figure 1

267 Unlike CC, the THS tobacco stick does not burn when used, and thus its length remains unchanged.

268 The tobacco plug is made of reconstituted cast leaf tobacco containing various tobacco types from

269 different origins, as well as binders and humectants. The humectants were added to prevent the cast

270 leaves becoming too brittle. Heating the humectants caused them to evaporate and re-condense to form

271 small droplets, generating a visible aerosol. Four different tobacco stick variants were used for aerosol

272 characterization: two versions of the THS2.2 Regular (THS2.2 FR1 and THS2.2 D2) and two versions

273 of the THS2.2 Menthol (THS2.2 FR1 M and THS2.2 D1 M). These four THS2.2 tobacco sticks contain

274 flavor ingredients. THS2.2 FR1 M and THS2.2 D1 M contain natural /-menthol applied to a cellulose

275 acetate yarn included in the polymer-film filter and to the inner liner paper included in the tobacco stick

276 pack. FR1, D2 and D1 are different tobacco blends but the tobacco sticks were of the same design.

278 2.3 Cambridge glass-fiber filter pad, cigarette, and tobacco stick conditioning

279 The reference cigarette 3R4F and THS2.2 tobacco sticks were stored at 5 ± 3°C with uncontrolled

280 humidity conditions in the original packaging, before conditioning. For the ISO/Health Canada

281 conditioning, test articles were conditioned for at least 48 h at 22 ± 1°C and 60 ± 2% relative humidity

282 (RH), according to the International Organization for Standardization (ISO) method 3402

283 (International Organisation for Standardization, 2010a). Cambridge glass-fiber filter pads were

284 conditioned under the same conditions.

285 Prior to the analyses performed in "tropical" and "desert" conditions, test articles were either

286 conditioned for at least 48 h at 30 ± 1°C and 75 ± 2% RH ("tropical" conditions), or at 30 ± 1°C and 35

287 ± 2% RH ("desert" conditions). The corresponding Cambridge glass-fiber filter pads were conditioned

288 under the same climatic conditions.

290 2.4 Generation of THS2.2 aerosol and mainstream smoke of 3R4F according to the HCI machine-

291 smoking regimen

292 The mainstream aerosol of THS2.2 and the smoke of the reference cigarette 3R4F were generated on a

293 Borgwaldt linear smoking machine type LM20X (Borgwaldt KC GmbH, Hamburg, Germany) for the

294 determination of all analytes (except elements) according to the HCI machine-smoking regimen

295 (Health Canada, 2000). For the elements, the mainstream aerosol of THS2.2 and the smoke of the

296 reference cigarette 3R4F were generated from a Burghart rotary smoking machine type RMB 20

297 (Burghart Tabaktechnik GmbH, Wedel, Germany) with the same smoking regimen.

298 For the in vitro biological assays test battery (Ames, MLA, and NRU), the 3R4F mainstream smoke

299 and THS2.2 aerosols were generated using a Burghart rotary smoking machine type RMB 20 (Burghart

300 Tabaktechnik GmbH, Wedel, Germany) according to the Health Canada Intense (HCI) smoking

301 regimen (Health Canada, 2000). The generated aerosol and smoke were trapped to analyze the aerosols.

302 After trapping, the samples were analyzed and processed (section 2.7). Processing describes the

303 conversion from a primary result, e.g., a peak area or counts to a value per cigarette or per stick, taking

304 the number of accumulations, the trapping or extraction volume or dilution into account. For in vitro

305 assessments, the aerosol was fractionated into two parts, TPM and GVP, during the same aerosol

306 collection (except for the Ames assay, where only TPM was tested). At the end of the aerosol

307 generation, the collected TPM aerosol fraction on the Cambridge glass-fiber filter pad was solubilized

308 in dimethyl sulfoxide (DMSO), and the water-soluble GVP fraction was immobilized into an impinger

309 of ice-cold Ca and Mg -free phosphate-buffered saline (PBS) solution.

310 The reference cigarette 3R4F was smoked to a butt length of 35 mm using a bell-shaped puff profile

311 and 100% blocking of ventilation holes. THS2.2 Regular and Menthol tobacco sticks were 'smoked'

312 using a bell-shaped puff profile to a defined puff count of 12 puffs. The limitation to 12 puffs is based

313 on the fixed settings of the THS2.2 system, which is programmed to finish heating after a maximum

314 period of 6 minutes, and the puff interval of the HCI regimen (30 seconds).

316 2.5 Generation of THS 2.2 aerosol under different climatic conditions

317 THS2.2 aerosol was generated under different ambient temperature and RH conditions of 22 ± 2°C and

318 60 ± 5% RH, 30 ± 2°C and 75 ± 5% RH, and 30 ± 2°C and 35 ± 5% RH to simulate "Mediterranean",

319 "Tropical" and "Desert" climates, respectively (Table 1), using a linear smoking machine prototype

320 SM405XR (Cerulean Molins PLC, Milton Keynes, UK ) and the HCI machine-smoking regimen

321 (Health Canada, 2000). The smoking machine was housed in a conditioned air cabinet (temperature

322 range: 10°C to 35°C; humidity range: 10% to 80% RH) fitted with a Delta 335 air conditioning unit

323 (Design Environmental Ltd, Ebbw Vale, UK). The atmosphere of the cabinet was constantly refreshed

324 with conditioned air. The temperature and RH in the cabinet was monitored using a TH1 dataloger

325 (ELPRO-Buchs AG, Buchs, Switzerland).

326 Approximate position for Table 1

327 2.6 Generation of THS2.2 aerosol under alternative puffing regimens

328 THS2.2 aerosol was generated according to the alternative puffing regimen presented in Table 2. A

329 Cerulean SM450RH smoking machine (Cerulean Molins PLC, Milton Keynes, UK) was used to

330 generate aerosols for the analysis of all analytes except nitric oxide (NO) and nitrogen oxides (NOx).

331 The NO and NOx measurements were performed on a Borgwaldt linear smoking machine type LM20X

332 (Borgwaldt KC GmbH, Hamburg, Germany). Since this smoking machine is limited to puffs of 100 ml,

333 NO and NOx measurement were not performed for the LR-3 regimen (Table 2).

334 Approximate position for Table 2

335 The alternative puffing regimens (SR-1, SR-4, SR-5, SR-6 and LR-3) were selected according to

336 human puffing behavior observed with THS2.2 users (Campelos et al., 2016).

338 2.7 Chemical analyses

339 All analytes were determined using 15 separate aerosol and smoke generations. The 15 separate analyte

340 groups and corresponding individual analytes are shown in Table 3.

341 Approximate position for Table 3

342 The description of the analytical methods used to quantify the analytes in the THS2.2 aerosol and in the

343 smoke of the 3R4F reference cigarette are presented in the supplementary material section.

345 2.8 Physical measurements

346 The droplet size distribution measurements were conducted using a PIXE multistage cascade impactor

347 (PIXE International Corp., Tallahassee, FL USA) using a sampling flow rate of 1 l/min. During this

348 study, the PIXE cascade impactor was composed of nine impactor stages. For both 3R4F and THS2.2

349 Tobacco Sticks, the average mass median aerodynamic diameter (MMAD) and geometric standard

350 deviation (GSD) were estimated from 10 replicate aerosol samples. The average GSD was determined

351 as the square root of the average GSD . The test items were connected to the inlet of a programmable

352 dual syringe pump (PDSP) (Burghart Messtechnik GmbH, Wedel, Germany). The outlet of the PDSP

353 was connected to a glass T-junction that allowed aerosol transfer before it entered the PIXE cascade

354 impactor. The outlet of the PIXE cascade impactor was connected to a pump (Vacuubrand GmbH +

355 CO KG, Wertheim, Germany) (Figure 2).

356 Approximate position for Figure 2

358 The calculation of MMAD and GSD was done separately for each measurement. The following steps

359 were performed:

360 1. Calculation of the net weight for each impactor stage of the cascade impactor:

361 AM; = [Weight loaded (g)]i - [Weight empty (g)]i (i = 1,..., i).

362 2. Calculation of the total mass by summing up all of the i net weights: TM = X AM;.

363 3. Calculation of the mass fraction for each stage: Mf =AM; / TM (i = 1,..., i).

364 4. Normalization of the mass fraction for each stage by the width of successive cutoff diameter

365 [AD50%]: The cutoff diameter represents the smaller size that can be captured on the related

366 stage, whereas the maximum size collected on that stage is related to the next larger cutoff

367 diameter.

368 nMfi = Mfi/[AD50%]i(i = l,...,i)

369 The values for the cutoff diameters [AD50%]i (i = 1,..., i) are given in Table 4.

370 Approximate position for Table 4

371 5. The calculation of the MMAD and GSD was performed using Igor Pro version 6.3.2.3

372 (WaveMetrics, Inc., Lake Oswego, OR, USA) using a lognormal mono-modal fitting

373 distribution (Equation 1) of the normalized mass fraction nMfi vs. the respective mid-point

374 diameter D;

Illml )-ln(MM4fl )r f(dd)=---T=e wgsd"!

375 ddln(GSDh/2^ (j)

376 where dd is the size diameter in micrometers, GSD is the geometric standard deviation, MMAD

377 is the mass median aerodynamic diameter, and f(dd) represents the normalized mass fraction

378 data.

380 The mass of each impactor stage was recorded prior to and after each aerosol collection. Subsequently,

381 the mass fraction deposited on the different impactor stages was calculated, whereas the size bins were

382 normalized by dividing the mass fractions with their respective widths. This process permitted the

383 transformation of discrete data points into density functions that could be fitted with a continuous

384 lognormal distribution function, from which both the MMAD and the GSD were calculated. The

385 negative values were not replaced by zero. The lower boundaries (LB) and upper boundaries (UB) were

386 calculated at the 95% confidence interval using the following equation:

387 LB = MMAD/GSD2; UB = MMAD x GSD2

389 2.9 In vitro Toxicology

390 All in vitro studies were performed in full accordance with the principles of Good Laboratory Practice.

392 2.9.1 Neutral red uptake (NRU) assay

393 The mouse embryonic fibroblast cell line Balb/c 3T3 (clone A31) was obtained from the European

394 Collection of Authenticated Cell Cultures (Salisbury, UK), and was used to perform the NRU assay

395 according to INVITTOX protocol 3a (INVITTOX, 1990), with some modifications (Borenfreund and

396 Puerner, 1985). Sodium dodecyl sulfate was used as the positive control.

397 In brief, 20 to 28 hours prior to aerosol fraction generation, cells were trypsinized and resuspended in

398 Dulbecco's modified Eagle's medium supplemented with heat-inactivated fetal bovine serum (10%

399 v/v) (Thermo Fisher Scientific, Waltham, MA, USA), 4 mM L-glutamine, 100 U/ml penicillin, and 0.1

400 mg/ml streptomycin. Subsequently, 4.75-5.25 x 10 viable cells were seeded in each well of a 96-well

401 plate and cultured at 37°C (5% CO2 and 70% RH). Cells were exposed to eight concentrations of each

402 test substance for 23 ± 1 hours in 96-well plates with six wells used per concentration. The exposure

403 plates were sealed with a CÜ2-permeable plastic film to prevent potential carry-over of volatile

404 substances. Following the exposure phase, the cell culture medium was replaced with cell culture

405 medium containing neutral red dye at 50 (ig/ml, and incubated at 37°C (5% CO2 and 70% RH) for an

406 additional 3 ± 0.5 hours. Subsequently, the cells were washed with PBS, and the neutral red dye taken

407 up by cells was extracted by the addition of destaining solution (ethanol, water, and acetic acid, mixed

408 in a 50:49:1 ratio). The plates were mechanically shaken using a vibrating platform shaker (Titramax

409 1000, Heidolph Instruments, Schwabach, Germany) for 10 min at approximately 450 strokes/min.

410 Neutral red absorbance was measured at 540 nm with a microplate reader (Safire 2, Tecan GmbH,

411 Grodig, Austria). The measured absorbance for each concentration was normalized against the

412 appropriate solvent control and converted to a percentage value. Cytotoxicity was expressed as 1/EC50,

413 and expressed as a function of the mass of TPM trapped on the Cambridge glass-fiber filter pads (TPM

414 basis) and on a per-mg nicotine basis. For consistency and to compare GVP fractions, data were

415 calculated and expressed on a per-mg TPM basis and on a per-mg nicotine basis (Roemer et al., 2014).

416 The EC50 endpoint measurement corresponds to the concentration of test substance for which a

417 decrease of 50% in the uptake of the neutral red dye is observed, and was determined with the SAS

418 Enterprise guide® 4.3 (SAS 9.2) software program (SAS, Cary, NC, USA).

420 The relationship between the concentration of the substance and the decrease in the uptake of neutral

421 red dye has a sigmoid shape and is described by the Hill function, which is a four parameter non-linear

422 function. Statistical analysis was performed using SAS. Unless mentioned otherwise in the text, all

423 reagents were obtained from Sigma-Aldrich (St. Louis, MO, USA).

425 2.9.2 Ames assay

426 Mutagenic activity was evaluated by using the Salmonella typhimurium tester strains TA98, TA100,

427 TA102, TA1535, and TA1537 with and without an S9 enzymatic metabolizing fraction, by following a

428 pre-incubation method (Maron and Ames, 1983) and the OECD 471 test guideline

429 (Organisation for Economic Co-operation and Development, 1997). The S9 enzymatic metabolizing

430 fraction was obtained from Aroclor 1254-induced male Sprague-Dawley rat liver (Moltox, NC, USA).

431 The TPM mainstream smoke fraction from the 3R4F reference cigarette was generated and tested in

432 parallel with the THS2.2 aerosol fraction. The strains were grown overnight in a shaking incubator at

433 37°C for approximately 10 hours in Oxoid Nutrient Broth No. 2 (Fisher Scientific, Reinach,

434 Switzerland). To determine mutagenic activity, seven different concentrations of THS2.2 TPM diluted

435 in DMSO were tested. The bacteria (approximately 1 x 109 in 100 (il) were combined with 50 (il of

436 either the test item, the solvent, or the positive control, and 500 (il of the cofactor buffer (pH 7.4)

437 supplemented with S9 as appropriate and pre-incubated at 37°C for 20 min prior to adding 2 ml of

438 histidine (50 (xM final concentration) supplemented soft top agar and plating the entire mixture onto

439 histidine-deficient 90 mm minimal glucose agar base plates for 2 days at 37°C. Revertant colonies were

440 counted using an automatic colony counter (Sorcerer, Perceptive Instruments, Bury Saint Edmunds,

441 UK). All experiments were performed in triplicate. Toxicity was detected as either a reduction in the

442 number of histidine revertants or as a thinning of the auxotrophic background lawn. The mutagens used

443 as positive controls, i.e. substances known to induce a mutagenic response to demonstrate the assay is

444 working efficiently, in experiments without the S9 mix were 4-nitrophenylenediamine (10 ^g/plate)

445 for TA98 and TA100, sodium azide (1.25 ^g/plate) for TA1535 and TA1537, and cumene

446 hydroperoxide (3 (ig/plate) for TA102. In the experiments that included the S9 fraction, benzo[a]pyrene

447 (1 (ig/plate) was used for TA98, and 2-aminoanthracene (2.5 (ig/plate) was used for TA100, TA102,

448 TA1535, and TA1537. DMSO (50 ^l/plate) served as the solvent control. All positive control

449 chemicals were obtained from either Sigma-Aldrich (St. Louis, MO, USA) or Moltox (Boone, NC,

450 USA). The biological activity after 1 mg of TPM exposure is reported as a means to permit a rapid

451 assessment of the impact of the 3R4F aerosol vs. the THS2.2 aerosol. One milligram was the maximum

452 dose tested of the 3R4F aerosol (for toxicity reasons) and thus comparison at 5mg or higher

453 concentrations was not technically possible.

455 2.9.3 Mouse lymphoma assay (MLA)

456 The L5178Y tk+/- cell line (sub-clone 3.7.2C (IVGT), Public Health England, UK) was used in the

457 MLA. Spontaneously-occurring tk~h mutants were purged from working stocks using methotrexate to

458 select against tk-deficient cells and thymidine, hypoxanthine, and guanine to ensure optimal growth of

459 tk-proficient cells as previously described (Chen and Moore, 2004). Cells were maintained in RPMI

460 1640 medium (Thermo Fisher Scientific, Waltham, MA, USA) supplemented with heat-inactivated

461 horse serum (10% v/v) (Thermo Fisher Scientific, Waltham, MA, USA), penicillin (200 U/ml),

462 streptomycin (200 (ig/ml), L-glutamine (2 mM), sodium pyruvate (1 mM), and pluronic acid F68 (0.1%

463 v/v) (Sigma-Aldrich, St. Louis, MO, USA). Short-term treatments (4 hours) were carried out in

464 reduced serum medium (3% v/v), while 10% v/v-containing medium was utilized for longer-term

465 exposures (24 hours). Cloning was carried out in 20% v/v-containing serum medium in the absence of

466 pluronic acid F68. Aroclor 1254-induced male Sprague-Dawley rat liver S9 in 0.15 M KCl (Moltox,

467 USA), in combination with a cofactor mix of glucose-6-phosphate and nicotinamide adenine

468 dinucleotide phosphate (both from Roche Applied Science, Basel, Switzerland), were used as the

469 exogenous metabolic activation system in the assay. The final concentration of S9 in cell cultures was

470 0.95 mg protein/ml (2% v/v). The controls used were methyl methanesulfonate (15 and 20 (ig /ml) for

471 the 4 hour S9- test arm, methyl methanesulfonate (5 and 7.5 (ig /ml) for the 24 hour S9- arm, and 7,12472 dimethylbenz[a]anthracene (1 and 1.5 (ig/ml) for the 4 hour S9+ arm.

474 The microwell version of the MLA was performed according to the OECD TG 490 guideline

475 (Organisation for Economic Co-operation and Development, 2015). Briefly, on two independent test

476 occasions, L5178Y cells in single replicate cultures were seeded at a density of 5 x 105 or 2 x 105

477 cells/ml (at least 6 x 106 cells), and exposed to 14 concentrations of TPM and GVP derived from

478 THS2.2 and the 3R4F reference cigarette for 4 hours in the presence (+S9) and absence (-S9) of

479 metabolic activation and 24 h in the absence of metabolic activation (-S9) treatment conditions,

480 respectively. For each treatment condition, cells were exposed for 4 or 24 hours at 37°C in the presence

481 of 5% CO2 and RH >65%. Following treatment, cells were washed and sub-cultured at a maximum

482 density of 2 x 105 cells/ml (at most 6 x 106 cells) for two further days to allow phenotypic expression

483 of the tk gene prior to mutant selection. Cells at 8 cells/ml or 1 x 104 cells/ml were then distributed into

484 96-microwell plates (200 (il per well) to determine final levels of TPM- and GVP-induced cytotoxicity

485 and mutagenicity, respectively. Cytotoxicity was determined from the relative total growth (RTG) of

486 the cell cultures following treatment and sub-culture periods in non-selective growth medium (typically

487 10-11 days). TPM- and GVP-treated cell cultures which underwent excessive cytotoxicity were

488 discarded through the assay procedure as mutagenicity data derived from these cells are difficult to

489 interpret because of their questionable biological relevance. Tk mutants were detected following culture

490 in trifluorothymidine (TFT)-containing growth medium (Sigma-Aldrich, St. Louis, MO, USA) for

491 typically 14 days. Mutant colonies were enumerated visually; colonies with a size less than a quarter of

492 the microwell's surface area were defined as small colonies, while ones covering more than a quarter of

493 the microwell's surface area were defined as large colonies. Mutation frequencies were calculated

494 according to published method (Clements, 2000). The controls used were methyl methanesulfonate (15

495 and 20 (ig /ml) for the 4 hour S9- test arm, methyl methanesulfonate (5 and 7.5 ^g /ml) for the 24 hour

496 S9- arm and finally 7,12-dimethylbenz[a]anthracene (1 and 1.5 (ig/ml) for the 4 hour S9+ arm.

497 The data generated from solvent-treated and positive controls in each treatment condition on the

498 separate test occasions were evaluated for acceptability according to OECD TG 490 guideline

499 (Organisation for Economic Co-operation and Development, 2015) and the laboratory's historical

500 control database. Furthermore, a response to an aerosol fraction was considered positive, i.e.

501 mutagenic, in the MLA if there was a concentration-related increase in mutation frequency (MF) with a

502 corresponding RTG not lower than 10%, and if an MF exceeded the sum of the microwell global

503 evaluation factor (GEF) of 126 plus the mean MF of the solvent-treated controls (Moore et al., 2006;

504 Moore et al., 2007). The microwell GEF of 126 mutants per 106 viable cells was previously defined as

505 the mean of the negative/solvent control MF distribution plus one standard deviation in a multi-

506 laboratory microwell MLA study (Moore et al., 2006). Mutagenic potencies were evaluated using the

507 lowest observed genotoxic effect level (LOGEL) (Guo et al., 2015). A LOGEL in this study was

508 defined as the lowest concentration of TPM or GVP (expressed on a per-mg nicotine basis) tested that

509 induced a mutagenic response which exceeded the GEF.

511 2.10 Statistical analyses

512 The results of the analyte quantifications are expressed in two different ways: On a per-Tobacco

513 Stick/cigarette basis, and on a per-mg nicotine basis. For all analyte, the number of values (N), the

514 arithmetic mean (M), and the confidence interval of the mean at 95% (CI95%) are given. For groups

515 including at least one measured value below the limit of quantification (LOQ) of the analytical method,

516 only the median or the LOQ was given, depending on whether the median was above or below the

517 LOQ. The mean on a per-mg nicotine basis and the respective confidence interval of the mean were

518 calculated as follows:

^ M(A)

519 M- v y

520 where A denotes the value of an analyte on a cigarette basis, and B denotes the mean value for nicotine

521 on a cigarette basis

522 and

M ( A)

M(A\ M ( B) \M ( A)2 M ( B)

rt2 rr2

SM ( A) , SM ( B) '

523 ± CI95J—= ±t(0.975, N -1)

M (B) VN

524 where A denotes the mean of an analyte on a cigarette basis, B denotes the mean for nicotine on a

525 cigarette basis, S is the standard deviation, N is the number of measurements, t(p,df) is the percentile of

526 Student's distribution, and df is degrees of freedom.

527 For the NRU assay, the cytoxicity (1/EC50) of the 3R4F reference cigarette and the THS2.2 FR1 and

528 THS2.2 FR1 M Tobacco Sticks, expressed on a per-TPM basis and on a per-mg nicotine basis, was

529 compared using one-way analysis of variance followed by Dunnett's multiple comparison procedure,

530 with the 3R4F reference cigarette as the reference group. This statistical approach was used because 2

531 THS2.2 test items were compared with 3R4F within one study. However, the cytotoxicity of the

532 THS2.2 D2 and THS2.2 D1 M tobacco sticks was compared against the 3R4F reference cigarette using

533 Student's t-test. In this case the t-test was used because 2 studies were performed, each with one test

534 item and the reference cigarette 3R4F.

536 3 RESULTS

537 3.1 Chemical composition of the THS2.2 aerosol, and comparison with the mainstream smoke

538 from the 3R4F reference cigarette

539 Two versions of the THS2.2 Regular (THS2.2 FR1 and THS2.2 D2) and two versions of the THS2.2

540 Menthol (THS2.2 FR1 M and THS2.2 D1 M) were compared with the 3R4F reference cigarette under

541 HCI machine-smoking conditions. Fifty-nine analytes (60 for menthol products) were determined,

542 covering various chemical classes present in different aerosol phases (Table 5 and Table 6). Fifteen

543 analytes in THS2.2 FR1 (benzo[a]pyrene, 3- and 4-aminobiphenyl, quinoline, NAT, pyrene,

544 nitrobenzene, vinyl chloride, dibenz[a,h]anthracene, and all the elements except mercury) and 11

545 analytes in THS2.2 D2 (2-aminonaphthalene, 3- and 4-aminobiphenyl, vinyl chloride,

546 dibenz[a,h]anthracene, and all the elements except mercury) were below the LOQ. In THS2.2 FR1 M,

547 the yields of 11 analytes (2-aminonaphthalene, 4-aminobiphenyl, quinoline, NAT, all the elements

548 except mercury and selenium, vinyl chloride, and dibenz[a,h]anthracene and 8 analytes in the THS2.2

549 D1 M (2-aminonaphthalene, 3-aminobiphenyl, arsenic, cadmium, lead, selenium, vinyl chloride, and

550 dibenz[a,h]anthracene) were below the LOQ. The 54 HPHCs which were targets for reduction (Section

551 1) were lower in the THS2.2 aerosol than in the 3R4F smoke. The yields of analytes expressed on a

552 per-mg nicotine basis are presented in Table A and in Table B of the supplementary material.

553 Approximate position for Table 5

554 Approximate position for Table 6

555 The pie charts in Figure 3 illustrates the differences in TPM composition between the 3R4F and the

556 THS2.2 FR1. While the THS2.2 FR1 delivered about the same TPM yield as the 3R4F, the THS2.2

557 aerosol composition was qualitatively and quantitatively different from that of 3R4F. The quantities of

558 water and glycerol relative to total TPM were considerably higher for the THS2.2, whereas the amount

559 of nicotine was approximately 30% lower for the THS2.2. Therefore, the relative yields of the 'other'

560 aerosol constituents were noticeably lower in the THS2.2 TPM. This is also evident when comparing

561 the color of the Cambridge glass-fiber filter pads after collection of the same amount of aerosol mass

562 and TPM from THS2.2 tobacco sticks and 3R4F cigarettes, respectively (Figure 3). Previous heat-not-

563 burn products such as Premier (deBethizy et al., 1990) and Eclipse (Borgerding et al., 1998) also

564 produce TPM compositions that contained mainly water and glycerin. Visually, the collected aerosol

565 from THS2.2 on a Cambridge glass-fiber filter pad appears similar to that seen for both Premier and

566 Eclipse.

567 Approximate position for Figure 3

569 3.2 Chemistry of the THS2.2 aerosol generated under different climatic conditions

570 The three climatic conditions, "Mediterranean", "desert", and "tropical", were selected according to

571 ICH (International Council for Harmonisation, 2003) and WHO

572 (World Health Organisation Technical Report, 2015) guidelines for stability testing. Since the

573 generation of HPHCs either through pyrosynthesis or distillation should be enhanced when increasing

574 the temperature, lower temperatures were not considered for this comparison. The impact of

575 temperature and relative humidity (RH) on the deliveries of the different analytes in the THS2.2 FR1

576 aerosol is presented in Table 7. The data expressed on a per-mg nicotine basis are presented in Table C

577 of the supplementary material.

578 Approximate position for Table 7

580 3.3 Aerosol chemistry of the THS2.2 FR1 generated with different machine-smoking regimens

581 The range of machine-smoking regimens used during this test was quite broad, and ranged from a total

582 puff volume of 210 ml for the ISO conditions to 1120 ml for the LR-3 regimen (see Table 2). Despite

583 these substantial differences in puff volumes, benzo[a]pyrene, 1-aminonaphthalene, 2584 aminonaphthalene, 3-aminobiphenyl, 4-aminobiphenyl, dibenz[a,h]anthracene, and vinyl chloride

585 remained below the LOQ for all machine-smoking regimens. For the other analytes, ISO and SR-1

586 delivered the lowest yields of HPHCs, while machine-smoking regimens SR-4, SR-6, and LR-3

587 delivered the highest HPHC yields. A summary of the obtained results is presented in Table 8 together

588 with the yields obtained for the 3R4F reference cigarette smoked in the HCI conditions. All the

589 individual results are presented in Tables D and E of the supplementary material.

590 The LR-3 smoking regimen could not be performed with the Borgwaldt linear smoking machine type

591 used to quantify NO and NOx. Therefore, this value was not reported. For TPM and carbonyls, the

592 breakthrough was specifically tested with the most intense smoking regimens to ensure that losses were

593 negligible and did not affect the accuracy of the measurements.

594 Approximate position for Table 8

596 3.4 Physical measurement of the aerosol

597 Ten series of measurement were performed on the 3R4F reference cigarette and on the THS2.2 FR1.

598 All the calculated MMAD, GSD, and boundary values are presented in Table 9. The upper boundaries

599 (UB) were below 2.5 (xm. Consequently, the smoke generated from the 3R4F and the aerosol generated

600 from the THS2.2 were respirable for all replicates with a margin of error of 5% and, according to Hinds

601 (Hinds, 1999), more than 85% of the aerosol droplets could reach the alveoli in the lung.

602 Approximate position for Table 9

604 3.5 Neutral red uptake (NRU) assay

605 Two versions of the THS2.2 Regular (THS2,2 FR1 and THS2.2 D2) and two versions of the THS2.2

606 Menthol (THS2.2 FR1 M and THS2.2 D1 M) were tested and compared with the 3R4F under HCI

607 machine-smoking conditions in the NRU assay in independent studies. The cytotoxic activity of both

608 the TPM and GVP fractions from the THS2.2 and the 3R4F reference cigarette was determined. A clear

609 concentration-dependent decrease in the number of viable cells was observed for the aerosol fractions

610 generated from the THS2.2 Regular (FR1 and D2) and THS2.2. Menthol (FR1 and D1) and the smoke

611 fractions generated from the 3R4F reference cigarette. The cytotoxicity levels induced by both products

612 covered a range spanning from no or low to high cytotoxicity. The 1/EC50 values (expressed on a per-

613 mg TPM basis and on a per-mg nicotine basis) were used to compare the relative cytotoxicity of the

614 THS2.2 and the 3R4F reference cigarette. The results showed that the relative in vitro cytotoxicity of

615 the THS2.2 Regular (FR1 and D2) and Menthol (FR1 and D1) aerosol fractions was reduced by

616 approximately 95% when expressed on a per-mg TPM basis, compared with the 3R4F reference

617 cigarette (Tables F and G of the supplementary material). The in vitro cytotoxicity of THS2.2 Regular

618 (FR1 and D2) and THS2.2 Menthol (FR1 and D1) aerosol fractions, expressed on a per-mg nicotine

619 basis, was reduced by 85-90% compared with the 3R4F reference cigarette (Table 10 and Table 11).

620 Approximate position for Table 10

621 Approximate position for Table 11

623 3.6 Ames assay

624 The TPM from the THS2.2 D2, THS2.2 D1 M, and the 3R4F was tested with the S. typhimurium

625 strains TA98, TA100, TA102, TA1535, and TA1537 in both the presence and absence of S9 (Table

626 12).

628 A positive response in the Ames test for the TPM from the 3R4F reference cigarette was detected in 3

629 of the 5 S. typhimurium tester strains in the presence of the S9 fraction, namely TA98, TA100, and

630 TA1537 (see Table 12). Nevertheless, despite testing up to 10 mg of TPM for THS2.2 D2 and 5 mg/per

631 plate for THS2.2 D1 M, no mutagenicity in any of the tester strains was detected under the conditions

632 of this assay.

633 Approximate position for Table 12

635 3.7 Mouse lymphoma assay (MLA)

636 The in vitro MLA was used to assess the mutagenicity of both TPM and GVP derived from THS2.2

637 D2, THS2.2 D1 M, and 3R4F. In both tests under the three treatment conditions, TPM and GVP

638 derived from THS2.2 D2 and 3R4F induced concentration-dependent increases in cytotoxicity and

639 mutagenicity. In the presence of S9, the mutagenic responses reproducibly surpassed the GEF threshold

640 for mutagenicity at or just above the cytotoxicity limit of the assay, i.e. RTG 10%-20%. Increases in

641 both large and small colonies were observed for TPM and GVP derived from both test articles. For

642 THS2.2 D2, LOGELs for the TPM fraction were markedly higher (on average 17-fold) than for 3R4F-

643 derived TPM (Figure 4A and Table 13). A similar mutagenicity profile was observed for the GVP

644 (Figure 4B). In both treatment conditions (4 hours and 24 hours) conducted in the absence of S9,

645 mutagenicity which exceeded the GEF threshold was also observed, however, this finding was not

646 always reproducible between the tests. When it was the case, LOGELs for THS2.2 D2 -derived TPM

647 were again markedly higher (on average, at least 14-fold) than for 3R4F TPM (exemplar responses

648 illustrated in Figure 5A and Table 13). Moreover, in these treatment conditions, the LOGELs for TPM

649 always occurred at the RTG 10%-20% cytotoxicity level. Similar mutagenicity profiles were also

650 observed for GVP under the same treatment conditions. Similar results were obtained for the

651 mentholated version of THS2.2 (exemplar responses illustrated in Figure 5B, Figure 6, and Table 13).

652 Mutagenicity data expressed on a per-mg TPM basis are presented as part of the supplementary

653 material (Supplementary Figures A-C and Supplementary Table H).

654 Approximate position for Figure 4

655 Approximate position for Figure 5

656 Approximate position for Figure 6

658 Approximate position for Table 13

660 4 DISCUSSION

661 The objective of this study was to assess the potential for reduced exposure to HPHCs from THS2.2

662 compared with 3R4F based on chemical analysis of HPHCs, in vitro genotoxicity, and cytotoxicity

663 assessments. To evaluate the robustness of the data for the products under different conditions, HPHC

664 yields were also measured when using the product under simulated real-life smoking conditions,

665 additional tests were performed under different climatic conditions and with different puffing regimens.

667 4.1 Aerosol physics and aerosol chemistry; comparison with the 3R4F reference cigarette

668 During this assessment of the THS2.2, following the feedback from taste panels and to ensure

669 sustainability of tobacco sources, the blend FR1 was replaced by blend D2 for the regular product and

670 by blend D1 for the menthol product (Smith et al., submitted (this issue)). Therefore, it was considered

671 important to present the chemical characterization of all the four products (THS2.2 FR1, THS2.2 FR1

672 M, THS2.2 D2 and THS2.2 D1 M) in this publication. When comparing the four products, it can be

673 observed that they delivered HPHC yields that were in the same range. The influence of the tobacco

674 blend composition on the HPHC yields is presented in the 3rd publication of this issue (Schaller et al.,

675 submitted (this issue)-b). When comparing menthol and regular THS2.2 products, no substantial

676 influence of menthol on the HPHC yields was detected. Schmeltz and Schlotzhauer have reported that

677 menthol pyrolysis in a closed system produced benzo[a]pyrene leading them to suggest that menthol

678 pyrolysis could act as a potential precursor to benzo[a]pyrene to the smoke of mentholated cigarette

679 products (Schmeltz and Schlotzhauer, 1968). A significant contribution of the menthol to the yield of

680 benzo[a]pyrene was not observed in the THS2.2. The FR1 Menthol product deliverd a slightly higher

681 yield of benzo[a]pyrene (1.29 ± 0.10 ng/ stick) than the FR1 Regular product (<1.00 ng/stick), but the

682 yield of menthol in the D1 Menthol product (1.08 ± 0.09 ng/stick) was on the low side compared to the

683 D2 Regular product (1.19 ± 0.08 ng/stick). In addition, in the pyrolytic conditions used by Schmeltz

684 and Scholtzhauer, benzo[a]pyrene was detected only on pyrolysis of menthol at 860°C, but not during

685 pyrolysis at 600°C. Other studies have reported that when isotopically labeled menthol was added to

686 tobacco of cigarettes, most of the menthol was transferred to the smoke unchanged and the production

687 of labeled benzo[a]pyrene was not detected (Jenkins et al., 1970; Baker and Bishop, 2004). Since (i)

688 the heater blade temperature in the THS2.2 only reaches a maximum temperature of 350°C, and (ii)

689 available literature on menthol pyrolysis to yield benzo[a]pyrene is limited, it was concluded that

690 menthol is unlikely to be a significant source of benzo[a]pyrene in the THS2.2 aerosol.

691 The mainstream aerosols produced by all the analyzed THS2.2 products were similar regarding analyte

692 yields including HPHC yields, but substantially different from the yields in mainstream smoke of the

693 3R4F reference cigarette. To quantify the exposure reduction, the yields of each HPHC for THS2.2

694 relative to those in 3R4F were calculated, and are presented on a per Tobacco Stick/cigarette (Figure

695 7). The graph presenting the results on a mg-nicotine basis is included in the supplementary material

696 (Figure D).

697 Approximate position for Figure 7

699 Figure 7 presents a general view of the ratios between THS2.2 and 3R4F. The 100% line represents the

700 yields of the 3R4F on a per-cigarette basis. When considering the HPHC yields on a Tobacco

701 Stick/cigarette basis, it can be observed that acrylamide, ammonia, butyraldehyde, butyraldehyde,

702 acetamide, and mercury presented ratios between 25 and 50%. Even if this denotes a substantial

703 reduction compared to 3R4F, it was not surprising to see ammonia, acrylamide, acetamide and mercury

704 at this level, since they could be formed or could distill out of tobacco at relatively low temperatures

705 (McDaniel et al., 2001; Stadler et al., 2002; Becalski et al., 2003; Blank et al., 2005; Moldoveanu,

706 2010; Becalski et al., 2011). The other analytes presented smaller ratios. Therefore, when compared

707 with 3R4F mainstream smoke, all HPHCs are considerably reduced in the THS2.2 aerosol of the four

708 products. This data supports the hypothesis that the controlled heating of tobacco in THS2.2 resulted in

709 a significant reduction in the pyrosynthesis of HPHCs. When compared with 3R4F, the formation of

710 polycyclic aromatic hydrocarbons, aromatic amines, phenols, and aldehydes was reduced by more than

711 75%, and for the majority of HPHCs, by more than 90% under the HCI machine-smoking conditions.

712 The analyzed HPHCs covered a broad range of chemical compounds, and several of these HPHCs have

713 been described as markers for the pyrolysis of tobacco (Moldoveanu, 2010). Since the reduction of

714 individual HPHCs was consistent across the different HPHC groups, it may be assumed that other

715 HPHCs, although not measured, were similarly reduced. In addition, it can also be observed that some

716 HPHCs that could distill out of tobacco in 3R4F were also reduced in the THS2.2 aerosol. The transfer

717 of cadmium to the aerosols of the four THS2.2 products could not be quantified (results below LOQ),

718 and the yield of TSNAs was minimal (Table 5 and Table 6). Since the nicotine yield was lower in the

719 analyzed THS2.2 products than in 3R4F, the ratios calculated on a per-mg nicotine basis were

720 somewhat higher. However, the trend remained the same, and the reductions expressed on a per-mg

721 nicotine basis were also substantial (Figure D in the supplementary material).

723 The mainstream smoke of 3R4F and the THS2.2 aerosol generated under the HCI machine-smoking

724 conditions were both shown to be respirable aerosols (Table 9). The MMAD values were similar: 0.8

725 (xm for 3R4F and 0.7 (m for THS2.2. The GSD was somewhat higher for THS2.2 (Section 3.4).

726 Therefore, THS2.2 presents respirable properties that are similar to those of 3R4F, while reducing

727 substantially the levels of the measured HPHCs.

729 4.2 Chemical composition of the THS2.2 FR1 aerosol collected under different climatic

730 conditions and extreme puffing regimens

731 Since the THS2.2 may be used by consumers using puffing regimens under climatic conditions that

732 deviate significantly from those considered in the HCI machine-smoking standard (55 ml puff every 30

733 s, 22°C, 60% RH), the aerosol of the THS2.2 FR1 was collected under different atmospheric and

734 puffing conditions described in sections 2.5 and 2.6.

736 The climatic conditions may have a significant impact on the deliveries to the mainstream smoke of CC

737 (Dymond and Hirji, 1972; Boder and Senehi, 1984). In previous studies, the filtration efficiency of the

738 tobacco rod and of the filter and the puff count were affected when CC were conditioned and smoked

739 under different climatic conditions. The effect could be observed on both gas phase and particulate

740 phase components. For instance, a temperature increase of 10°C resulted in an increased delivery of

741 nitric oxide by 18% for an American blend CC, while the phenol delivery decreased correspondingly

742 (Dymond and Hirji, 1972). It was anticipated that the careful control of the heater blade temperature up

743 to a maximum of 350°C (Smith et al., submitted (this issue)) and use of a low efficiency filter would

744 make THS2.2 less sensitive to variations in ambient conditions and alterations in the yields of the

745 different analytes. The yield of water in the THS2.2 aerosol was considered the only exception, because

746 of the high humectant content of the tobacco plug which contains glycerin in about 20% of the tobacco

747 plug weight. When the tobacco sticks were conditioned for at least 48 hours and used under "desert"

748 conditions, the delivery of water was considerably reduced when compared with the "tropical"

749 conditions (Table 7). The differences in the yield of water under both "tropical" and "desert" conditions

750 explained nearly all the variability in the TPM yields, since the NFDPM yields obtained from the

751 tobacco sticks in the three conditions ("Mediterranean", "tropical" and "desert") were similar. The

752 ranges for nicotine, formaldehyde, ethyl methyl ketone, acrylonitrile, 1,3-butadiene, benzene, styrene,

753 toluene, o-cresol, p-cresol, NNN, ammonia, and acetamide were lower than the Intermediate Precision

754 (IP) (International Council for Harmonisation, 1996; Walfish, 2006) of the respective analytical

755 methods, and the climatic conditions were considered not to have a significant impact on the yields of

756 these HPHCs. The ranges for other HPHCs were low, except for m-cresol, phenol, o-toluidine,

757 propylene oxide, and nitrobenzene, which had ranges in excess of 35%. For o-toluidine, propylene

758 oxide, and nitrobenzene, the highest yield was obtained when conditioning and machine-smoking the

759 tobacco sticks at 22°C and 60% RH. Therefore, collecting the aerosol under "tropical" or "desert"

760 conditions did not increase the yields of these HPHCs. The yield of m-cresol was increased from 0.031

761 to 0.071 (g/stick under the "tropical" condition, while phenol was increased from 1.66 to 2.49 (g/stick

762 under the "desert" condition. However, the values remained low when compared with the yields

763 obtained from 3R4F: m-cresol (3.03 (g/stick) and phenol (13.6 (g/stick) (Table 5). Therefore, the

764 variation of the climatic conditions had only a minor influence on the HPHC yields.

766 The different machine-smoking puffing regimens (Table 2) were selected to cover the puffing behavior

767 reported for users of THS2.2 (Campelos et al., 2016). They induced significant modifications of the air

768 flow and of the quantity of air used to extract aerosol from the tobacco plug in THS2.2. The minimum

769 and maximum yields obtained from the different machine-smoking regimens are presented in Table 8.

770 In general, the yields of polar HPHCs (e.g. phenol and cresol isomers) were more sensitive than apolar

771 HPHCs to the variation of the machine-smoking puffing conditions (Table 8). For each HPHC, the

772 comparison between the maximum yield and the yield obtained with the HCI smoking regimen should

773 enable the identification of the HPHC for which the standard protocol may underestimate the exposure

774 when using more extreme puffing conditions. The ratios of maximum yield to HCI yield are presented

775 in Table 8. These ratios were less than 2 for 42 of 49 analyzed compounds. Again, phenol and cresol

776 isomers were the HPHCs presenting the largest ratios. However, the HCI machine-smoking protocol

777 gave a relevant estimate of the exposure for the majority of the tested HPHCs. Interestingly, the ratio

778 obtained for nitrogen oxides was only 1.1, and the ratio for CO was only 1.2. Since NOx and CO can be

779 considered potential markers of combustion (Norman et al., 1983; Reed, 2002; Glarborg et al., 2003;

780 Im et al., 2003; Baker, 2006; Senneca et al., 2007; Cozzani et al., 2016), no evidence of tobacco

781 combustion was found even under extreme machine-smoking puffing conditions. Under extreme

782 machine-smoking puffing conditions, the yields of all toxicologically relevant compounds in the

783 THS2.2 aerosol were lower than those obtained when smoking the 3R4F reference cigarette under HCI

784 machine-smoking conditions (Table 5).

786 4.3 In vitro Toxicology

787 The in vitro toxicology results reflect the chemistry data; THS2.2 aerosol fraction-induced effects are

788 distinctly different in terms of potency from those induced by counterpart fractions from 3R4F. The

789 THS2.2 aerosol demonstrates a substantial reduction in toxicological activity compared with 3R4F

790 smoke. In the NRU assay, both the particulate phase and GVP in vitro cytotoxicity of THS2.2 Regular

791 (FR1 and D2) and THS2.2 Menthol (FR1 and D1) were reduced by 85%-95% compared with the

792 3R4F, independent of the basis used to express the activity (per-mg TPM or per-mg nicotine) (Table

793 10, Table 11 and Tables F and G of the supplementary material).

795 The Ames assay did not reveal significant mutagenicity of the TPM fraction for either THS2.2 regular

796 or THS2.2 menthol under the conditions of this test. In contrast, the TPM fraction from 3R4F was

797 mutagenic in tester strains TA98, TA100, and TA1537 in the presence of the S9 metabolizing fraction

798 from Aroclor-treated rat liver. The MLA data show that both the TPM and GVP aerosol fractions

799 derived from THS2.2 D2 and THS2.2 D1 M were mutagenic in this assay. However, the LOGELs

800 demonstrate a lower in vitro mutagenic potency of the THS2.2 aerosol fractions compared with 3R4F.

801 While a conclusion underlying the mechanism(s) of this phenomenon cannot be definitively made on

802 the basis of these data, it is reasonable to suggest that the overall reduction in the burden of toxicants

803 present in the THS2.2 aerosols may play a role in the manifestation of reduced cytotoxic and mutagenic

804 potency in vitro.

806 5 CONCLUSION

807 The low operating temperature of THS2.2 results in significantly lower concentrations of HPHCs in the

808 mainstream aerosol compared with the mainstream smoke of the 3R4F reference cigarette when

809 expressed on either a per-Tobacco Stick/cigarette or a per-mg nicotine basis, while the MMAD of both

810 aerosols remains similar. The reductions in the concentrations of most HPHCs in the THS2.2 aerosol

811 were greater than 90% when compared with 3R4F, and were not affected by machine-smoking of

812 THS2.2 under extreme climatic conditions. No evidence of tobacco combustion was found when using

813 the THS2.2 device with puffing regimens that were significantly more intense than the HCI conditions.

814 The mutagenic and cytotoxic potencies of the mainstream aerosol fractions from THS2.2, when

815 evaluated by the Ames, mouse lymphoma, and NRU assays were reduced by at least 85%-95%

816 compared with the mainstream smoke aerosol of 3R4F.

818 Conflict of Interest statement

819 The work reported in this publication involved a candidate Modified Risk Tobacco Product developed

820 by Philip Morris International (PMI) and was solely funded by PMI. All authors are (or were)

821 employees of PMI R&D or worked for PMI R&D under contractual agreements.

823 Acknowledgments

824 The authors are very grateful to the staff of Philip Morris Research Product Testing Laboratories, for

825 their excellent technical assistance.

826 6 REFERENCES

827 Ames, B. N., Lee, F. D., Durston, W. E., 1973. An Improved Bacterial Test System for the Detection

828 and Classification of Mutagens and Carcinogens. Proceedings of the National Academy of

829 Sciences. 70, 782-786. http://www.pnas.org/content/70/37782.abstract

830 Baker, R. R., 2006. Smoke generation inside a burning cigarette: Modifying combustion to develop

831 cigarettes that may be less hazardous to health. Progress in Energy and Combustion Science.

832 32, 373-385. http://www.sciencedirect.com/science/article/pii/S036012850600013X

833 Baker, R. R., Bishop, J., 2004. The pyrolysis of tobacco ingredients. Journal of Applied Pyrolysis. 71,

834 223. http://www.sciencedirect.com/science/article/pii/S0165237003000901

835 Becalski, A., Brady, B., Feng, S., Gauthier, B. R., Zhao, T., 2011. Formation of acrylamide at

836 temperatures lower than 100 degrees C: The case of prunes and a model study. Food Additives

837 & Contaminants. Part A. 28, 726-30. http://dx.doi.org/10.1080/19440049.2010.535217

838 Becalski, A., Lau, B. P. Y., Lewis, D., Seaman, S. W., 2003. Acrylamide in Foods: Occurrence,

839 Sources, and Modeling. Journal of Agricultural and Food Chemistry. 51, 802-808.

840 http://dx. doi.org/10.1021/jf020889y

841 Benowitz, N. L., Iii, P. J., 1984. Daily intake of nicotine during cigarette smoking. Clinical

842 Pharmacology & Therapeutics. 35, 499-504. http://dx.doi.org/10.1038/clpt.1984.67

843 Benowitz, N. L., Jacob, P., 1994. Metabolism of nicotine to cotinine studied by a dual stable isotope

844 method. Clinical Pharmacology & Therapeutics. 56, 483-493.

845 http://dx. doi.org/10.103 8/clpt. 1994.169

846 Bernstein, D., 2004. A review of the influence of particle size, puff volume, and inhalation pattern on

847 the deposition of cigarette smoke particles in the respiratory tract. Inhalation Toxicology. 16,

848 675-89. http://informahealthcare.com/doi/abs/10.1080/08958370490476587

849 Blank, I., Robert, F., Goldmann, T., Pollien, P., Varga, N., Devaud, S., Saucy, F., Huynh-Ba, T.,

850 Stadler, R. H., 2005. Mechanisms of acrylamide formation. in: Friedman, M., Mottram, D.,

851 (Eds.), Chemistry and Safety of Acrylamide in Food. vol. 561. Springer US, pp. 171-189.

852 Boder, J. B., Senehi, F., Influence of the atmospheric pressure and the relative humidity on smoke

853 delivery 1984. http://legacy.library.ucsf.edu/tid/obw56e00 (accessed 03.2016)

854 Borenfreund, E., Puerner, J. A., 1985. Toxicity determined in vitro by morphological alterations and

855 neutral red absorption. Toxicology Letters. 24, 119-124.

856 http://www.sciencedirect.com/science/article/pii/0378427485900463

857 Borgerding, M., Klus, H., 2005. Analysis of complex mixtures - Cigarette smoke. Experimental and

858 Toxicologic Pathology. 57, Supplement 1, 43-73.

859 http://www.sciencedirect.com/science/article/pii/S0940299305000369

860 Borgerding, M. F., Bodnar, J. A., Chung, H. L., Mangan, P. P., Morrison, C. C., Risner, C. H., Rogers,

861 J. C., Simmons, D. F., Uhrig, M. S., Wendelboe, F. N., Wingate, D. E., Winkler, L. S., 1998.

862 Chemical and biological studies of a new cigarette that primarily heats tobacco.: Part 1.

863 Chemical composition of mainstream smoke. Food and Chemical Toxicology. 36, 169-182.

864 http://www.sciencedirect.com/science/article/pii/S0278691597001087

865 Campelos, P., Jeannet, C., Maeder, S., Poget, L., 2016. Development of models for the estimation of

866 mouth level exposure to aerosol constituents from a heat-not-burn tobacco product using

867 mouthpiece analysis. Beiträge zur Tabakforschung / Contributions to Tobacco Research.

868 Accepted for publication (volume 27, issue 5).

869 Carchman, R. A., Identification of animal or organism-based short-term tests which are available for

870 the assessment of the potentially harmful properties of tobacco smoke., Reduction Report of

871 Canada's Expert Committee on Cigarette Toxicity Reduction: Conference Proceedings &

872 Invited Papers., Toronto, Canada, September 21 - 22, 1988, pp. Appendix, paper no 8.

873 Carmella, S. G., Han, S., Fristad, A., Yang, Y., Hecht, S. S., 2003. Analysis of Total 4874 (Methylnitrosamino)-1-(3-Pyridyl)-1-Butanol (NNAL) in Human Urine. Cancer Epidemiology

875 Biomarkers & Prevention. 12, 1257-1261.

876 http://cebp.aacrjournals.org/content/12/! 1/1257.abstract

877 Chen, T., Moore, M. M., 2004. Screening for Chemical Mutagens Using the Mouse Lymphoma Assay.

878 in: Humana Press, (Ed.), Optimization in Drug Discovery: In Vitro Methods., pp. 337-352.

879 Clements, J., 2000. The Mouse Lymphoma Assay. Mutation Research/Fundamental and Molecular

880 Mechanisms of Mutagenesis. 455, 97-110.

881 http://www.sciencedirect.com/science/article/pii/S002751070000066X

882 Clive, D., Flamm, W. G., Machesko, M. R., Bernheim, N. J., 1972. A mutational assay system using

883 the thymidine kinase locus in mouse lymphoma cells. Mutation Research/Fundamental and

884 Molecular Mechanisms of Mutagenesis. 16, 77-87.

885 http://www.sciencedirect.com/science/article/pii/0027510772900668

886 Cozzani, V., McGrath, T. E., Smith, M., Schaller, J. P., Zuber, G., Absence of combustion in an

887 electrically heated tobacco system - An experimental investigation 21st International

888 Symposium on Analytical and Applied Pyrolysis, Nancy, France. 9th - 12th of May, 2016.

889 https://www.pmiscience.com/library/absence-combustion-electrically-heated-tobacco-system-

890 experimental-investigation (accessed 6.2016)

891 Davies, C. N., 1988. Cigarette smoke: Generation and properties of the aerosol. Journal of Aerosol

892 Science. 19, 463-469. http://www.sciencedirect.com/science/article/pii/0021850288900213

893 deBethizy, J. D., Borgerding, M. F., Doolittle, D. J., Robinson, J. H., McManus, K. T., Rahn, C. A.,

894 Davis, R. A., Burger, G. T., Hayes, J. R., Reynolds, J. H., Hayes, A. W., 1990. Chemical and

895 biological studies of a cigarette that heats rather than burns tobacco. The Journal of Clinical

896 Pharmacology. 30, 755-/63. http://dx.doi.org/10.1002/j.1552-4604.1990.tb03639.x

897 Diekmann, J., Douda, M., Rustemeier, K., 2006. Rapid and sensitive method for the determination of

898 propylene oxide in cigarette mainstream smoke by gas chromatography-mass spectrometry.

899 Journal of Chromatographic Science. 44, 32-34.

900 http://chromsci.oxfordj ournals.org/content/44/1/32. abstract

901 Doll, R., Peto, R., Boreham, J., Sutherland, I., 2004. Mortality in relation to smoking: 50 years'

902 observations on male British doctors. BMJ. 328, 1519.

903 http://www.bmj.com/content/bmj/328/7455/1519.full.pdf

904 Dymond, H. F. D., Hirji, T., The effect of climatic conditions on the smoke delivery - Part II: Effect of

905 temperature and relative humidity. 1972. http://legacy.library.ucsf.edu/tid/wfm00f00 (accessed

906 03.2016)

907 Family Smoking Prevention and Tobacco Control Act, Public Law No. 111-131 (June 22, 2009).

908 Fuoco, F. C., Buonanno, G., Stabile, L., Vigo, P., 2014. Influential parameters on particle concentration

909 and size distribution in the mainstream of e-cigarettes. Environmental Pollution. 184, 523-529.

910 http://www.sciencedirect.com/science/article/pii/S0269749113005307

911 Geiss, O., Bianchi, I., Barahona, F., Barrero-Moreno, J., 2015. Characterisation of mainstream and

912 passive vapours emitted by selected electronic cigarettes. International Journal of Hygiene and

913 Environmental Health. 218, 169-180.

914 http://www. sciencedirect. com/science/article/pii/S 1438463914000972

915 Ghosh, D., Jeannet, C., 2014. An improved Cambridge filter pad extraction methodology to obtain

916 more accurate water and "Tar" values: In situ Cambridge filter pad extraction methodology.

917 Beiträge zur Tabakforschung / Contributions to Tobacco Research. 26, 38-49.

918 www.degruyter.com/view/j/cttr.2014.26.issue-2/cttr-2014-0008/cttr-2014-0008.xml

919 Glarborg, P., Jensen, A. D., Johnsson, J. E., 2003. Fuel nitrogen conversion in solid fuel fired systems.

920 Progress in Energy and Combustion Science. 29, 89-113.

921 http://www. sciencedirect. com/science/article/pii/S03 6012850200031X

922 Guo, X., Heflich, R. H., Dial, S. L., Richter, P. A., Moore, M. M., Mei, N., 2015. Quantitative analysis

923 of the relative mutagenicity of five chemical constituents of tobacco smoke in the mouse

924 lymphoma assay. Mutagenesis.

925 http://mutage.oxfordjournals.org/content/early/2015/05/29/mutage.gev039.abstract

926 Haziza, C., de La Bourdonnaye, G., Skiada, D., Ancerewicz, J., Baker, G., Picavet, P., Lüdicke, F.,

927 submitted (this issue). Evaluation of the Tobacco Heating System 2.2. Part 8: 5-Day

928 randomized reduced exposure clinical study in Poland.

929 Health Canada, Health Canada - Tobacco Products Information Regulations S0R/2000-273, Schedule

930 2 2000. http://laws-lois.justice.gc.ca/PDF/S0R-2000-273.pdf

931 Hinds, W. C., 1999. Aerosol Technology: Properties, Behavior, and Measurement of Airborne

932 Particles, 2nd Edition. JOHN WILEY & SONS, INC.

933 Im, H., Rasouli, F., Hajaligol, M., 2003. Formation of nitric oxide during tobacco oxidation. Journal of

934 Agricultural and Food Chemistry. 51, 7366-7372. http://dx.doi.org/10.1021/jf030393w

935 Ingebrethsen, B. J., Cole, S. K., Alderman, S. L., 2012. Electronic cigarette aerosol particle size

936 distribution measurements. Inhalation Toxicology. 24, 976-984.

93 7 http: // informahealthcare. com/doi/abs/10.3109/08958378.2012.744781

938 Institute of Medicine, 2001. Clearing the smoke: The science base for tobacco harm reduction.

939 Washington, DC: The National Academy Press.

940 International Council for Harmonisation, 1996. ICH Guideline Q2A & Q2B "Guidelines for Validation

941 of Analytical Methods". http://www.ich.org/products/guidelines/quality/quality-

942 single/article/validation-of-analytical-procedures-text-and-methodology.html

943 International Council for Harmonisation, 2003. ICH Guideline Q1AR2: "Stability Testing of New Drug

944 Substances and Products". http://www.ich.org/products/guidelines/quality/article/quality-

945 guidelines.html

946 International Organisation for Standardization, 1999. ISO 14644-1:1999,Cleanrooms and associated

947 controlled environments http://www.iso.org/iso/catalogue_detail.htm?csnumber=25052

948 International Organisation for Standardization, 2010a. ISO 3402:1999, Last rev: 2010. Tobacco and

949 tobacco products -- Atmosphere for conditioning and testing.

950 http://www.iso.org/iso/catalogue_detail.htm?csnumber=28324

951 International Organisation for Standardization, 2010b. ISO 8454:2007, Last rev: 2010. Cigarettes -952 Determination of carbon monoxide in the vapour phase of cigarette smoke -- NDIR method.

953 http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=41168

954 International Organisation for Standardization, ISO Standard 4387: 2000. Last Revision 2011.

955 Cigarettes -- determination of total and nicotine free dry particulate matter using a routine

956 analytical smoking machine., 2011.

957 http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=28323

958 International Organisation for Standardization, 2012. ISO 22634:2008, Last rev: 2012. Cigarettes -959 Determination of benzo[a]pyrene in cigarette mainstream smoke -- Method using gas

960 chromatography/mass spectrometry.

961 http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=41040

962 International Organisation for Standardization, ISO 10315:2013, Cigarettes -- Determination of

963 nicotine in smoke condensates -- Gas-chromatographic method. 2013.

964 http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=56744

965 INVITTOX, 1990. Protocol No. 3a: the FRAME modified neutral red uptake cytotoxicity test. In: The

966 ERGATT/FRAME data bank of in vitro techniques in toxicology. Nottingham, INVITTOX.

967 Ishizu, Y., Ohta, K., Okada, T., 1978. Changes in the particle size and the concentration of cigarette

968 smoke through the column of a cigarette. Journal of Aerosol Science. 9, 25-29.

969 http://www.sciencedirect.com/science/article/pii/0021850278900605

970 Jenkins, R. W., Jr., Newman, R. H., Chavis, M. K., 1970. Cigarette Smoke Formation Studies: II.

971 Smoke Distribution and Mainstream Pyrolytic Composition of Added 14C-Menthol (U).

972 Beiträge zur Tabakforschung / Contributions to Tobacco Research. 5, 299.

973 www.degruyter.com/view/j/cttr.1970.5.issue-6/cttr-2013-0254/cttr-2013-0254.xml

974 Kane, D. B., Asgharian, B., Price, O. T., Rostami, A., Oldham, M. J., 2010. Effect of smoking

975 parameters on the particle size distribution and predicted airway deposition of mainstream

976 cigarette smoke. Inhalation Toxicology. 22, 199-209.

977 http : //informahealthcare. com/doi/abs/10.3109/089583 70903161224

978 Kavvadias, D., Scherer, G., Urban, M., Cheung, F., Errington, G., Shepperd, J., McEwan, M., 2009.

979 Simultaneous determination of four tobacco-specific N-nitrosamines (TSNA) in human urine.

980 Journal of Chromatography B. 877, 1185-1192.

981 http://www. sciencedirect. com/science/article/pii/S 1570023209001925

982 Kogel, U., Titz, B., Schlage, W. K., Nury, C., Martin, F., Oviedo, A., Lebrun, S., Elamin, A., Guedj, E.,

983 Trivedi, K., Ivanov, N. V., Vanscheeuwijck, P., Peitsch, M. C., Hoeng, J., submitted (this

984 issue). Evaluation of the Tobacco Heating System 2.2. Part 7: Systems toxicological assessment

985 of a mentholated version revealed reduced cellular and molecular exposure effects compared

986 with mentholated and non-mentholated cigarette smoke.

987 Laino, T., Tuma, C., Moor, P., Martin, E., Stolz, S., Curioni, A., 2012. Mechanisms of propylene

988 glycol and triacetin pyrolysis. The Journal of Physical Chemistry A. 116, 4602-4609.

989 http://dx.doi.org/10.1021/jp300997d

990 Maron, D. M., Ames, B. N., 1983. Revised methods for the Salmonella mutagenicity test. Mutation

991 Research/Environmental Mutagenesis and Related Subjects. 113, 173-215.

992 http://www. sciencedirect. com/science/article/pii/0165116183900109

993 Martin, F., Talikka, M., Ivanov, N. V., Haziza, C., Hoeng, J., Peitsch, M. C., submitted (this issue).

994 Evaluation of the Tobacco Heating System 2.2. Part 9: Application of systems pharmacology to

995 identify exposure response markers in periferal blood of smokers switching to THS2.2.

996 Mascher, D. G., Mascher, H. J., Scherer, G., Schmid, E. R., 2001. High-performance liquid

997 chromatographic-tandem mass spectrometric determination of 3-hydroxypropylmercapturic

998 acid in human urine. Journal of Chromatography B: Biomedical Sciences and Applications.

999 750, 163-169. http://www.sciencedirect.com/science/article/pii/S0378434700003856

1000 McDaniel, R. L., Torrence, K. M., Self, D. A., Chang, M. J., 2001. Determination of Mercury in

1001 Mainstream Cigarette Smoke by Conventional and Amalgamation Cold Vapor Atomic

1002 Absorption Spectrometry. Beiträge zur Tabakforschung / Contributions to Tobacco Research.

1003 19, 267. www.degruyter.com/view/j/cttr.2001.19.issue-5/cttr-2013-0713/cttr-2013-0713.xml

1004 McGrath, T. E., Wooten, J. B., Geoffrey Chan, W., Hajaligol, M. R., 2007. Formation of polycyclic

1005 aromatic hydrocarbons from tobacco: The link between low temperature residual solid (char)

1006 and PAH formation. Food and Chemical Toxicology. 45, 1039-1050.

1007 http://www.sciencedirect.com/science/article/pii/S027869150600370X

1008 Medeiros, A. M., Bird, M. G., Witz, G., 1997. Potential biomarkers of benzene exposure. Journal of

1009 Toxicology and Environmental Health. 51, 519-539.

1010 https ://www. scopus.com/inward/record. url? eid=2-s2.0-

1011 0030849236&partnerID=40&md5=ebf89fd952f176827aad879249cf9bdc

1012 Minet, E., Cheung, F., Errington, G., Sterz, K., Scherer, G., 2011. Urinary excretion of the acrylonitrile

1013 metabolite 2-cyanoethylmercapturic acid is correlated with a variety of biomarkers of tobacco

1014 smoke exposure and consumption. Biomarkers. 16, 89-96.

1015 http://www. ncbi.nlm. nih.gov/pmc/articles/PMC3 026342/

1016 Moldoveanu, S., 2010. Pyrolysis of Organic Molecules with Applications to Health and Environmental

1017 Issues. Elsevier.

1018 Moore, M. M., Honma, M., Clements, J., Bolcsfoldi, G., Burlinson, B., Cifone, M., Clarke, J., Clay, P.,

1019 Doppalapudi, R., Fellows, M., Gollapudi, B., Hou, S., Jenkinson, P., Muster, W., Pant, K.,

1020 Kidd, D. A., Lorge, E., Lloyd, M., Myhr, B., O'Donovan, M., Riach, C., Stankowski, L. F., Jr,

1021 Thakur, A. K., Van Goethem, F., 2007. Mouse lymphoma thymidine kinase gene mutation

1022 assay: Meeting of the International Workshop on Genotoxicity Testing, San Francisco, 2005,

1023 recommendations for 24-h treatment. Mutation Research/Genetic Toxicology and

1024 Environmental Mutagenesis. 627, 36-40.

1025 http://www.sciencedirect.com/science/article/pii/S138357180600369X

1026 Moore, M. M., Honma, M., Clements, J., Bolcsfoldi, G., Burlinson, B., Cifone, M., Clarke, J.,

1027 Delongchamp, R., Durward, R., Fellows, M., Gollapudi, B., Hou, S., Jenkinson, P., Lloyd, M.,

1028 Majeska, J., Myhr, B., O'Donovan, M., Omori, T., Riach, C., San, R., Stankowski, L. F.,

1029 Thakur, A. K., Van Goethem, F., Wakuri, S., Yoshimura, I., 2006. Mouse lymphoma thymidine

1030 kinase gene mutation assay: Follow-up meeting of the international workshop on Genotoxicity

1031 testing—Aberdeen, Scotland, 2003—Assay acceptance criteria, positive controls, and data

1032 evaluation. Environmental and Molecular Mutagenesis. 47, 1-5.

1033 http://dx.doi.org/10.1002/em.20159

1034 Norman, V., Ihrig, A. M., Larson, T. M., Moss, B. L., 1983. The effect of some nitrogenous blend

1035 components on NO/NOx and HCN levels in mainstream and sidestream smoke. Beiträge zur

1036 Tabakforschung / Contributions to Tobacco Research. 12, 55-62.

1037 www.degruyter.com/view/j/cttr.1983.12.issue-2/cttr-2013-0525/cttr-2013-0525.xml

1038 Organisation for Economic Co-operation and Development, 1997. Test No. 471: Bacterial Reverse

1039 Mutation Test. OECD Publishing.

1040 Organisation for Economic Co-operation and Development, 2015. Test No. 490: In Vitro Mammalian

1041 Cell Gene Mutation Tests Using the Thymidine Kinase Gene. OECD Publishing.

1042 Oviedo, A., Lebrun, S., Kogel, U., Ho, J., Tan, W. T., Titz, B., Leroy, P., Vuillaume, G., Bera, M.,

1043 Martin, F., Rodrigo, G., Ivanov, N. V., Hoeng, J., Peitsch, M. C., Vanscheeuwijck, P.,

1044 submitted (this issue). Evaluation of the Tobacco Heating System 2.2. Part 6: 90-day OECD

1045 413 rat inhalation study with systems toxicology endpoints demonstrates reduced exposure

1046 effects of a mentholated version compared with mentholated and non-mentholated cigarette

1047 smoke.

1048 Patskan, G., Reininghaus, W., 2003. Toxicological evaluation of an electrically heated cigarette. Part 1:

1049 Overview of technical concepts and summary of findings. Journal of Applied Toxicology. 23,

1050 323-328. http://dx.doi.org/10.1002/jat.923

1051 Pojer, R., Whitfield, J. B., Poulos, V., Eckhard, I. F., Richmond, R., Hensley, W. J., 1984.

1052 Carboxyhemoglobin, cotinine, and thiocyanate assay compared for distinguishing smokers from

1053 non-smokers. Clinical Chemistry. 30, 1377-80.

1054 http://www.clinchem.org/content/30/8/1377.abstract

1055 Reed, T. B., 2002. Encyclopedia of Biomass Thermal Conversion: The Principles and Technology of

1056 Pyrolysis, Gasification & Combustion (Hydrogen Generator Gas). Golden, Co : Biomass

1057 Energy Foundation, 2002.

1058 Riedel, K., Scherer, G., Engl, J., Hagedorn, H.-W., Tricker, A. R., 2006. Determination of Three

1059 Carcinogenic Aromatic Amines in Urine of Smokers and Nonsmokers. Journal of Analytical

1060 Toxicology. 30, 187-195. http://jat.oxfordjournals.org/content/30/3/187.abstract

1061 Robinson, R. J., Yu, C. P., 2001. Deposition of cigarette smoke particles in the human respiratory tract.

1062 Aerosol Science and Technology. 34, 202-215. http://dx.doi.org/10.1080/027868201300034844

1063 Rodgman, A., Perfetti, T. A., 2013. The Chemical Components of Tobacco and Tobacco Smoke,

1064 Second Edition. CRC Press, Boca Raton.

1065 Roemer, E., Dempsey, R., Schorp, M. K., 2014. Toxicological assessment of kretek cigarettes: Part 1:

1066 Background, assessment approach, and summary of findings. Regulatory Toxicology and

1067 Pharmacology. 70, Supplement 1, S2-S14.

1068 http://www.sciencedirect.com/science/article/pii/S0273230014002906

1069 Roemer, E., Stabbert, R., Veltel, D., Müller, B. P., Meisgen, T. J., Schramke, H., Anskeit, E., Elves, R.

1070 G., Fournier, J. A., 2008. Reduced toxicological activity of cigarette smoke by the addition of

1071 ammonium magnesium phosphate to the paper of an electrically heated cigarette: Smoke

1072 chemistry and in vitro cytotoxicity and genotoxicity. Toxicology in Vitro. 22, 671-681.

1073 http://www.sciencedirect.com/science/article/pii/S088723330700361X

1074 Schaller, J. P., Keller, D., Poget, L., Pratte, P., Kaelin, E., McHugh, D., Cudazzo, G., Smart, D.,

1075 Tricker, A. R., Gautier, L., Yerly, M., Reis Pires, R., Le Bouhellec, S., Ghosh, D., Garcia, E.,

1076 Hofer, I., Vanscheewijck, P., Maeder, S., submitted (this issue)-a. Evaluation of the Tobacco

1077 Heating System 2.2. Part 2: Chemical composition, genotoxicity, cytotoxicity, and physical

1078 properties of the aerosol. Regulatory Toxicology and Pharmacology.

1079 Schaller, J. P., Pijnenburg, J. P. M., Ajithkumar, A., Tricker, A. R., submitted (this issue)-b. Evaluation

1080 of the Tobacco Heating System 2.2. Part 3: Influence of the tobacco blend on the formation of

1081 harmful and potentially harmful constituents of Tobacco Heating System 2.2 aerosol.

1082 Regulatory Toxicology and Pharmacology.

1083 Scherer, G., Urban, M., Hagedorn, H. W., Feng, S., Kinser, R. D., Sarkar, M., Liang, Q., Roethig, H. J.,

1084 2007. Determination of two mercapturic acids related to crotonaldehyde in human urine:

1085 Influence of smoking. Human and Experimental Toxicology. 26, 37-47.

1086 http://het. sagepub. com/content/26/1/3 7.abstract

1087 Schmeltz, I., Schlotzhauer, W. S., 1968. Benzo[a]pyrene, phenols and other products from pyrolysis of

1088 the cigarette additive, (d,/)-menthol. Nature. 219, 370-371.

1089 Schorp, M. K., Tricker, A. R., Dempsey, R., 2012. Reduced exposure evaluation of an Electrically

1090 Heated Cigarette Smoking System. Part 1: Non-clinical and clinical insights. Regulatory

1091 Toxicology and Pharmacology. 64, S1- S10.

1092 http://www.sciencedirect.com/science/article/pii/S0273230012001651

1093 Senneca, O., Ciaravolo, S., Nunziata, A., 2007. Composition of the gaseous products of pyrolysis of

1094 tobacco under inert and oxidative conditions. Journal of Analytical and Applied Pyrolysis. 79,

1095 234-243. http://www.sciencedirect.com/science/article/pii/S0165237006001252

1096 Sewer, A., Kogel, U., Talikka, M., Wong, E. T., Martin, F., Xiang, Y., Guedj, E., Ivanov, N., Hoeng, J.,

1097 Peitsch, M. C., submitted (this issue). Evaluation of the Tobacco Heating System 2.2. Part 5:

1098 MicroRNA expression from a 90-day rat inhalation study indicates reduced effects of the

1099 aerosol on lung tissue compared with cigarette smoke exposure.

1100 Singh, R. P., Jaiswal, V. K., Jain, V. K., 2006. Study of smoke aerosols under a controlled environment

1101 by using dynamic light scattering. Applied Optics. 45, 2217-2221.

1102 http://ao. osa. org/abstract. cfm?URI=ao-45-10-2217

1103 Sleiman, M., Logue, J. M., Montesinos, V. N., Russell, M. L., Litter, M. I., Gundel, L. A., Destaillats,

1104 H., 2016. Emissions from Electronic Cigarettes: Key Parameters Affecting the Release of

1105 Harmful Chemicals. Environmental Science & Technology. 50, 9644-9651.

1106 http://dx.doi.org/10.1021/acs.est.6b01741

1107 Smith, M. R., Clark, B., Luedicke, F., Schaller, J. P., Vanscheeuwijck, P., Hoeng, J., Peitsch, M. C.,

1108 submitted (this issue). Evaluation of the Tobacco Heating System 2.2. Part 1: Description of the

1109 System and the Scientific Assessment Program. Regulatory Toxicology and Pharmacology.

1110 Stabbert, R., Voncken, P., Rustemeier, K., Haussmann, H. J., Roemer, E., Schaffernicht, H., Patskan,

1111 G., 2003. Toxicological evaluation of an electrically heated cigarette. Part 2: Chemical

1112 composition of mainstream smoke. Journal of Applied Toxicology. 23, 329-339.

1113 http://dx.doi.org/10.1002/jat.924

1114 Stadler, R. H., Blank, I., Varga, N., Robert, F., Hau, J., Guy, P. A., Robert, M.-C., Riediker, S., 2002.

1115 Food chemistry: Acrylamide from Maillard reaction products. Nature. 419, 449-450.

1116 http://dx. doi. org/10.1038/419449a

1117 Strickland, P., Kang, D., Sithisarankul, P., 1996. Polycyclic aromatic hydrocarbon metabolites in urine

1118 as biomarkers of exposure and effect. Environmental Health Perspectives. 104, 927-932.

1119 http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1469694/

1120 Tewes, F. J., Meisgen, T. J., Veltel, D. J., Roemer, E., Patskan, G., 2003. Toxicological evaluation of

1121 an electrically heated cigarette. Part 3: Genotoxicity and cytotoxicity of mainstream smoke.

1122 Journal of Applied Toxicology. 23, 341-348. http://dx.doi.org/10.1002/jat.925

1123 The Royal College of Physicians, Nicotine without smoke: Tobacco harm reduction.

1124 The Royal College of Physicians, London, 2016.

1125 https: //www. rcplondon.ac.uk/proj ects/outputs/nicotine-without-smoke-tobacco-harm-reduction-

1126 0

1127 Torikaiu, K., Uwano, Y., Nakamori, T., Tarora, W., Takahashi, H., 2005. Study on tobacco

1128 components involved in the pyrolytic generation of selected smoke constituents. Food and

1129 Chemical Toxicology. 43, 559-568.

1130 http://www. sciencedirect.com/science/article/pii/S0278691505000104

1131 U.S. Food and Drug Administration, 2012. Harmful and Potentially Harmful Constituents in Tobacco

1132 Products and Tobacco Smoke; Established List. Federal Register Food and Drug

1133 Administration. 77, 20034.

1134 http://www.fda.gov/downloads/TobaccoProducts/GuidanceComplianceRegulatoryInformation/

1135 UCM297981.pdf

1136 U.S. Food and Drug Administration, 2016. Harmful and Potentially Harmful Constituents" in Tobacco

1137 Products as Used in Section 904(e) of the Federal Food, Drug, and Cosmetic Act (Revised)*.

1138 http://www.fda.gov/downloads/TobaccoProducts/Labeling/RulesRegulationsGuidance/UCM24

1139 1352.pdf

1140 Urban, M., Kavvadias, D., Riedel, K., Scherer, G., Tricker, A. R., 2006. Urinary Mercapturic Acids and

1141 a Hemoglobin Adduct for the Dosimetry of Acrylamide Exposure in Smokers and Nonsmokers.

1142 Inhalation Toxicology. 18, 831-839. http://dx.doi.org/10.1080/08958370600748430

1143 van Sittert, N. J., Megens, H. J. J. J., Watson, W. P., Boogaard, P. J., 2000. Biomarkers of exposure to

1144 1,3-butadiene as a basis for cancer risk assessment. Toxicological Sciences. 56, 189-202.

1145 http://toxsci.oxfordjournals.org/content/56/1/189.abstract

1146 Walfish, S., 2006. Analytical Methods: A Statistical Perspective on the ICH Q2A and Q2B Guidelines

1147 for Validation of Analytical Methods. BioPharm International. 19.

1148 http://www.biopharminternational.com/analytical-methods-statistical-perspective-ich-q2a-and-

1149 q2b-guidelines-validation-analytical-methods

1150 Werley, M. S., Freelin, S. A., Wrenn, S. E., Gerstenberg, B., Roemer, E., Schramke, H., Van Miert, E.,

1151 Vanscheeuwijck, P., Weber, S., Coggins, C. R. E., 2008. Smoke chemistry, in vitro and in vivo

1152 toxicology evaluations of the electrically heated cigarette smoking system series K. Regulatory

1153 Toxicology and Pharmacology. 52, 122-139.

1154 http://www.sciencedirect.com/science/article/pii/S0273230008001177

1155 Wong, E. T., Kogel, U., Veljkovic, E., Martin, F., Xiang, Y., Boue, S., Vuillaume, G., Leroy, P.,

1156 Guedj, E., Rodrigo, G., Ivanov, N. V., Hoeng, J., Peitsch, M. C., Vanscheeuwijck, P., submitted

1157 (this issue). Evaluation of the Tobacco Heating System 2.2. Part 4: 90-day OECD 413 rat

1158 inhalation study with systems toxicology endpoints demonstrates reduced exposure effects

1159 compared with cigarette smoke. Regulatory Toxicology and Pharmacology.

1160 World Health Organisation, The scientific basis of tobacco product regulation. Report of a WHO Study

1161 Group (TobReg). WHO Technical Report Series 945, Geneva. 2008.

1162 http://www.who.int/tobacco/global_interaction/tobreg/publications/tsr_951/en/

1163 World Health Organisation Technical Report, 2015. Annex 2 to WHO Technical Report Series, No.

1164 953: Stability testing of active pharmaceutical ingredients and finished pharmaceutical products.

1165 http://www.who.int/medicines/areas/quality_safety/quality_assurance/StabilityConditionsTable

1166 2UupdatedMarch2015.pdf? ua=1

1167 Zenzen, V., Diekmann, J., Gerstenberg, B., Weber, S., Wittke, S., Schorp, M. K., 2012. Reduced

1168 exposure evaluation of an Electrically Heated Cigarette Smoking System. Part 2: Smoke

1169 chemistry and in vitro toxicological evaluation using smoking regimens reflecting human

1170 puffing behavior. Regulatory Toxicology and Pharmacology. 64, S11-S34.

1171 http://www.sciencedirect.com/science/article/pii/S0273230012001614

1176 7 FIGURES AND TABLES

1177 Figure 1

1180 1181

Tobacco Plug Polymer-Film Filter Mouthpiece

Hollow Acetate Tube

Tipping

test item inserted into tobacco stick holder

Programmable dual syringe pump (PDSP)

High-efficiency particulate air (HEPA) filter

dilution system

Impactor

1188 1189

1190 Figure 4 A

J2 400'

Nicotine (mg/ml)

<n 500-1

V 13 ro 400-

> <£> o 300-

<n c ro 200i

D £ 100-

u. S 0-

—I—

Nicotine (mg nicotine equivalent/ml)

1197 Figure 6 A

« 500-1

« 800-1

Nicotine (mg/ml)

Nicotine (mg nicotine equivalent/ml)

1200 Figure 7

Relative Yield Ratios (THS2.2/3R4F) on per Tobacco Stick/Cigarette Basis

Isoprene

l,3-fc>utadiene

Benzene s Ethylene oxide ( p-cresol | Acrylonitrile | m-cresol | Hydrogen cyanide | Toluene ^ o-toluidine ^ o-cresol a Carbon monoxide m Resorcinol a Styrene B NNK -Nitrogen oxides a Nitric oxide a Methyl ethyl ketone a Benz[a]anthracene a Acetone a NNN B Crotonaldehyde ■ NAT a Acrolein a Phenol Hydroquinone Formaldehyde Propylene oxide Propionaldehyde Acetaldehyde Catechol Pyridine Mercury Acetamide Butyraldehyde Ammonia Acrylamide

0% 10% THS2.2 D1 M/3R4F

20% 30% THS2.2 D2/3R4F

10% 50% 60% I THS2.2 FR1 M/3R4F

I THS2.2 FR1/3R4F

1202 1203

1205 Captions for figures

1207 Figure 1: THS2.2 tobacco stick

1209 Figure 1: Experimental setup for aerosol physical measurement

1211 Figure 3: Cigarette total particulate matter (TPM) compared with THS2.2 aerosol composition generated according

1212 to the HCI machine-smoking conditions for measuring emissions. The photographs of the Cambridge glass-fiber

1213 filter pads after the collection of cigarette smoke (left) and THS2.2 aerosol (right) are also shown.

1215 Figure 4: The mutagenic responses induced by aerosol fractions derived from THS2.2 D2 and 3R4F in the 4 h +S9

1216 treatment condition in two independent tests expressed on a per-mg nicotine basis. A. TPM. MFs for the DMSO-

1217 treated controls in tests #1 and #2 were 129.77 ± 22.57 and 78.73 ± 1.55 mutants/10 viable cells, respectively.

1218 •THS2.2 D2 #1; «1182.2 02 #2; A3R4F#1; ▼3R4F#2; ♦GEF+DMSO MF #1; O GEF+DMSO MF #2. B. GVP.

1219 MFs for the PBS-treated controls in tests #1 and #2 were 107.73 ± 10.40 and 69.44 ± 3.10 mutants/106 viable cells,

1220 respectively. •THS2.2 D2 #1; BTHS2.2 D2 #2; A3R4F #1; ▼3R4F #2; ♦GEF+PBS MF #1; OGEF+PBS MF #2.

1222 Figure 5: The mutagenic responses induced by aerosol fractions derived from THS2.2 variants and 3R4F in the 4

1223 h-S9 treatment condition in two independent tests expressed on a per-mg nicotine basis. A. TPM from THS2.2 D2

1224 and 3R4F. MFs for the DMSO-treated controls in tests #1 and #2 were 99.52 ± 19.21 and 112.81 ± 23.41 mutants/106

1225 viable cells, respectively. •THS2.2 D2 #1; BTHS2.2 D2 #2; A3R4F #1; ▼3R4F #2; ♦GEF+DMSO MF #1;

1226 OGEF+DMSO MF #2. B. GVP from THS2.2 D1 M and 3R4F. MFs for the PBS-treated control(s) in tests #1 and #2

1227 were 122.33 and 94.10 ± 26.37 mutants/106 viable cells, respectively. OTHS2.2 D1 M #1; HTHS2.2 D1 M #2; A3R4F

1228 #1; ▼3R4F #2; ♦GEF+PBS MF #1; OGEF+PBS MF #2.

1230 Figure 6: The mutagenic responses induced by aerosol fractions derived from THS2.2 D1 M and 3R4F in the 4 h +S9

1231 treatment condition in two independent tests expressed on a per-mg nicotine basis. A. TPM. MFs for the DMSO-

1232 treated controls in tests #1 and #2 were 89.01 ± 6.76 and 93.90 ± 13.64 mutants/10 viable cells, respectively.

1233 •THS2.2 D1 M #1; HTHS2.2 D1 M #2; A3R4F #1; ▼3R4F #2; ♦GEF+DMSO MF #1; OGEF+DMSO MF #2. B.

1234 GVP. MFs for the PBS-treated control in tests #1 and #2 were 97.43 and 68.20 mutants/106 viable cells, respectively.

1235 OTHS2.2 D1 M #1; HTHS2.2 D1 M #2; A3R4F #1; ▼3R4F #2; ♦GEF+PBS MF #1; OGEF+PBS MF #2.

1237 Figure 7: Mainstream aerosol HPHCs from THS2.2 compared to the mainstream smoke HPHCs from the 3R4F

1238 reference cigarette (constituent levels set at 100%) on a per-unit basis under the Health Canada Intense (HCI)

1239 machine-smoking regimen. When one value or more was below the LOQ, the results were not presented in the

1240 graphs (NAT: N-nitrosoanatabine, NNK: 4-(N-nitrosomethylamino)-l-(3-pyridyl)-l-butanone, NNN: N-

1241 nitrosonornicotine).

1244 Table 1: Climatic conditions

Conditions Temperature Relative Humidity

[°C] [% RH]

Mediterranean 22 ± 2 60 ± 5%

Tropical 30 ± 2 75 ± 5%

Desert 30 ± 2 35 ± 5%

1248 Table 2: Smoking machine settings for the generation of THS2.2 aerosol generation under alternative puffing regimens

Regimen Puff volume Puff duration Puff interval Number of puffs*

[ml] [s] [s] [n]

ISO 35 2.0 60 6

SR-1 40 2.4 30 8

SR-5 80 2.4 30 8

HCI 55 2.0 30 12

SR-4 60 2.4 25 14

SR-6 80 2.4 25 >14

LR-3 110 4.5 22 14

1249 *: The number of puffs results from puff intervals of the different smoking

1250 regimens and the fixed settings of the THS2.2 system, which is

1251 programmed to finish heating after a maximum period of 6 minutes and

1252 allows up to 14 puffs to be taken during that time.

1255 Table 3: Analyte groups and corresponding individual analytes

Analyte Group Individual Analytes

ISO parameters and product-specific constituents Total particulate matter (TPM), water, nicotine, nicotine-free dry particulate matter (NFDPM), carbon monoxide (CO), glycerol

Volatiles- and semi-volatiles 1,3-butadiene, isoprene, benzene, toluene, styrene, pyridine, quinoline, acrylonitrile

Carbonyls acetaldehyde, acetone, acrolein, butyraldehyde, crotonaldehyde, formaldehyde, methyl ethyl ketone, propionaldehyde

Aromatic amines 1 -aminonaphthalene, 2-aminonaphthalene, 3-aminobiphenyl, 4-aminobiphenyl, o-toluidine

Nitrogen oxides nitric oxide (NO), oxides of nitrogen (NOx)

Hydrogen cyanide hydrogen cyanide

Ammonia Ammonia

Epoxides and vinyl chloride ethylene oxide, propylene oxide, vinyl chloride

Tobacco-specific nitrosamines ^V-nitrosoanabasine (NAB), ^V-nitrosoanatabine (NAT), ^V-nitrosonornicotine (NNN), 4-(#-nitrosomethylamino)-1-(3-pyridyl)-1-butanone (NNK)

Phenols and acid derivatives catechol, o-cresol, m-cresol, ^-cresol, hydroquinone, phenol, resorcinol, acetamide, acrylamide

Polycyclic aromatic hydrocarbons benzo[a]pyrene, benz[a]anthracene, dibenz[a,A]anthracene, pyrene

Nitrobenzene Nitrobenzene

Elements (except mercury) arsenic, cadmium, chromium, lead, nickel, selenium

Mercury Mercury

Menthol Menthol

1258 Table 4: Cutoff diameters and mid-point diameters for each stage of the PIXE cascade impactor

Cutoff diameter (|a,m) Mid-point (|0m)

Stage AD50% Di

7 16 16

6 8 12

3 1 1.5

2 0.5 0.75

1 0.25 0.375

L2 0.12 0.185

LI 0.06 0.09

1260 1261

1262 Table 5: Analyte yields from THS2.2 FR1, THS2.2 FR1 M, and 3R4F obtained under HCI machine-smoking

1263 conditions and expressed on a per-cigarette/tobacco stick basis

THS2.2 FR1 THS2.2 FR1 M 3R4F

Parameter Unit mean ± CI95% N mean ± CI95% N mean ± CI95% N

TPM mg/stick 48.2 ± 2.4 4 43.5 ± 1.5 4 49.0 ± 4.8 4

Water mg/stick 36.5 ± 3.1 4 29.7 ± 3.6 4 15.8 ± 2.9 4

Nicotine mg/stick 1.32 ± 0.16 4 1.21 ± 0.09 4 1.89 ± 0.16 4

NFDPM mg/stick 10.3 ± 0.9 4 12.6 ± 2.2 4 31.2 ± 1.8 4

Carbon monoxide mg/stick 0.531 ± 0.068 4 0.594 ± 0.110 4 32.8 ± 2.4 4

Benzo[a]pyrene ng/stick < 1.00 4 1.29 ± 0.10 4 14.2 ± 0.3 4

Puff Count /stick 12 ± 0 4 12 ± 0 4 10.6 ± 0.4 4

Menthol mg/stick n.a. 2.62 ± 0.1 4 n.a.

Glycerin mg/stick 4.63 ± 0.83 4 3.94 ± 0.87 4 2.42 ± 0.14 4

1 -aminonaphthalene ng/stick 0.077 4 0.086 4 20.8 ± 1.3 4

2-aminonaphthalene ng/stick 0.046 ± 0.008 4 < 0.035 4 11.0 ± 0.6 4

3 -aminobiphenyl ng/stick < 0.032 4 0.032 4 3.77 ± 0.47 4

4-aminobiphenyl ng/stick < 0.051 4 < 0.051 4 3.26 ± 0.12 4

Acetaldehyde ^g/stick 219 ± 31 4 205 ± 12 4 1555 ±184 4

Acetone ^g/stick 40.7 ± 6.2 4 39.4 ± 2.3 4 736±129 4

Acrolein ^g/stick 11.30 ± 2.36 4 9.15 ± 0.43 4 154 ± 20 4

Butyraldehyde ^g/stick 26.1 ± 2.3 4 26.7 ± 2 4 88.4 ± 10.7 4

Crotonaldehyde ^g/stick 4.14 ± 0.23 4 3.24 ± 0.21 4 68.8 ± 14.4 4

Formaldehyde ^g/stick 5.53 ± 0.69 4 4.55 ± 0.25 4 56.5 ± 12.1 4

Methyl ethyl ketone ^g/stick 7.18 ± 1.19 4 6.93 ± 0.64 4 187 ± 30 4

Propionaldehyde ^g/stick 14.5 ± 2.4 4 13.9 ± 0.7 4 125 ± 16 4

Acrylonitrile ^g/stick 0.258 ± 0.041 4 0.220 ± 0.014 4 31.9 ± 1.8 4

1,3-butadiene ^g/stick ¿10.294 ± 0.042 4 0.265 ± 0.024 4 63.8 ± 3.5 4

Benzene ^g/stick 0.649 ± 0.074 4 0.640 ± 0.040 4 97.6 ± 4.7 4

Isoprene ^g/stick 2.35 ± 0.39 4 2.11 ± 0.18 4 798 ± 49 4

Pyridine ^g/stick /7.54 ± 0.26 4 7.21 ± 0.25 4 36.1 ± 2.2 4

Quinoline ^g/stick < 0.012 4 < 0.012 4 0.513 ± 0.023 4

Styrene ^g/stick 0.608 ± 0.058 4 0.561 ± 0.033 4 24.5 ± 1.2 4

Toluene ^g/stick 2.59 ± 0.43 4 2.39 ± 0.16 4 188 ± 11 4

Catechol ^g/stick 16.3 ± 1.5 4 17.1 ± 1.1 4 91.4 ± 5.6 4

o-cresol ^g/stick 0.069 ± 0.008 4 0.095 ± 0.025 4 4.47 ± 0.16 4

rn-cresol ^g/stick 0.029 ± 0.004 4 0.033 ± 0.006 4 3.03 ± 0.08 4

^-cresol ^g/stick 0.072 ± 0.008 4 0.083 ± 0.010 4 9.17 ± 0.44 4

Hydroquinone ^g/stick 8.10 ± 0.48 4 8.98 ± 1.02 4 83.1 ± 5.5 4

Phenol ^g/stick 1.16 ± 0.12 4 1.60 ± 0.4 4 13.6 ± 0.9 4

Resorcinol ^g/stick 0.041 ± 0.003 4 0.048 ± 0.004 4 1.85 ± 0.08 4

NAB ng/stick < 3.15 4 < 3.15 4 33.7 ± 8.5 4

NAT ng/stick 20.5 ± 0.5 4 19.7 ± 3.6 4 318 ± 74 4

NNK ng/stick 6.7 ± 0.6 4 5.9 ± 0.4 4 266 ± 15 4

NNN ng/stick 17.2 ± 1.25 4 13.7 ± 1.21 4 309 ± 41 4

Ammonia ^g/stick 14.2 ± 1.1 4 13.8 ± 0.7 4 39.3 ± 3.2 4

Hydrogen cyanide ^g/stick 4.81 ± 0.35 4 5.14 ± 0.70 4 493 ± 78 4

Nitric oxide ^g/stick 16.8 ± 2.3 4 12.3 ± 1.7 4 491 ± 38 4

Nitrogen oxides ^g/stick 17.3 ± 2.6 4 12.6 ± 1.7 4 537 ± 43 4

Arsenic ng/stick < 1.13 4 < 1.13 4 8.51 ± 0.34 4

Cadmium ng/stick < 0.350 4 < 0.350 4 161 ± 4 4

Chromium ng/stick < 0.55 4 < 0.55 4 < 0.55 4

Lead ng/stick < 3.35 4 < 3.35 4 37.0 ± 0.7 4

Mercury ng/stick 1.17 ± 0.05 4 1.34 ± 0.18 4 4.80 ± 0.13 4

Nickel ng/stick < 0.55 4 < 0.55 4 < 0.55 4

Selenium ng/stick < 0.550 2 0.780 4 1.62 ± 0.32 4

Pyrene ng/stick < 5.00 4 9.06 ± 0.68 4 87.3 ± 2.5 4

o-toluidine ng/stick 1.260 ± 0.187 4 0.777 ± 0.287 4 85.5 ± 2.7 4

Acetamide ^g/stick 4.02 ± 0.18 4 4.30 ± 0.24 4 13.9 ± 0.5 4

Acrylamide ^g/stick 1.73 ± 0.12 4 1.91 ± 0.16 4 4.8 ± 0.3 4

Ethylene oxide ^g/stick 0.201 ± 0.014 4 0.202 ± 0.013 4 29.4 ± 2.0 4

Nitrobenzene ng/stick < 0.188 4 0.335 ± 0.164 4 8.62 ± 1.10 4

Propylene oxide ^g/stick 0.148 ± 0.018 4 0.149 ± 0.017 4 1.32 ± 0.12 4

Vinyl chloride ng/stick < 3.54 4 < 3.54 4 96.7 ± 2.0 4

Benz[a] anthracene ng/stick 1.45 ± 0.14 4 2.49 ± 0.17 4 28.0 ± 0.6 4

Dibenz[a, h] anthracene ng/stick < 0.100 4 < 0.100 4 1.70 ± 0.11 4

1264 N is the number of determinations, CI is the confidence interval of the mean, n.a.: not analyzed

1265 <: median lower than the limit of quantitation, in this case LOQ is given

1266 If at least one value is below the LOQ, the median is given and the CI is not mentioned

1267 TPM: Total particulate matter, NFDPM: Nicotine-free dry particulate matter, NAB: W-nitrosoanabasine, NAT: W-nitrosoanatabine, NNK: 4-(W-

1268 nitrosomethylamino)-1 -(3-pyridyl)-1 -butanone, NNN: W-nitrosonornicotine

1271 Table 6: Analytes yields from THS2.2 D2, THS2.2 D1 M, and 3R4F obtained under HCI machine-smoking conditions

1272 and expressed on a per-cigarette/tobacco stick basis

THS2.2 D2 THS2.2 D1 M 3R4F

Parameter Unit mean ± CI95% N mean ± CI95% N mean ± CI95% N

TPM mg/stick 54.1 ± 2.4 4 53.8 ± 3.6 4 46.3 ± 2.9 4

Water mg/stick 39.4 ± 4.6 4 39.1 ± 3.6 4 13.3 ± 1.6 4

Nicotine mg/stick 1.26 ± 0.24 4 1.32 ± 0.11 4 2.09 ± 0.14 4

NFDPM mg/stick 13.4 ± 2.8 4 13.4 ± 0.6 4 30.9 ± 1.9 4

Carbon monoxide mg/stick 0.598 ± 0.072 4 0.620 ± 0 4 30.7 ± 3.0 4

Benzo[a]pyrene ng/stick 1.19 ± 0.08 4 1.08 ± 0.09 3 13.7 ± 0.8 4

Puff Count /stick 12 ± 0 4 12 ± 0 4 10.7 ± 0.7 4

Menthol mg/stick n.a. 2.98 ± 0.21 4 n.a.

Glycerin mg/stick 4.1 ± 1.07 4 4.59 ± 0.47 4 2.39 ± 0.15 4

1 -aminonaphthalene ng/stick 0.063 ± 0.006 4 <0.061 4 19.7 ± 1.6 4

2-aminonaphthalene ng/stick <0.035 4 <0.035 4 14.8 ± 1.9 4

3 -aminobiphenyl ng/stick <0.013 4 <0.013 4 3.90 ± 0.42 4

4-aminobiphenyl ng/stick <0.021 4 n.a. 3.13 ± 0.60 4

Acetaldehyde ^g/stick 213 ± 19 4 220 ± 22 4 1589±76 4

Acetone ^g/stick 33.8 ± 6.4 4 42.6 ± 8.1 4 729 ± 36 4

Acrolein ^g/stick 9.44 ± 0.87 4 10.91 ± 2.98 4 193 ± 21 4

Butyraldehyde ^g/stick 25.3 ± 2.7 4 26.4 ± 0.9 4 103.9 ± 8.3 4

Crotonaldehyde ^g/stick 3.75 ± 0.34 4 4.15 ± 0.64 4 92.1 ± 13.2 4

Formaldehyde ^g/stick 5.22 ± 0.24 4 6.19 ± 2.00 4 68.7 ± 7.8 4

Methyl ethyl ketone ^g/stick 7.94 ± 0.75 4 10.19 ± 2.23 4 241 ± 16 4

Propionaldehyde ^g/stick 13.6 ± 1.5 4 15.9 ± 2.2 4 147 ± 8 4

Acrylonitrile ^g/stick 0.186 ± 0.028 4 0.196 ± 0.016 4 31.6 ± 2.3 4

1,3-butadiene ^g/stick 0.319 ± 0.073 4 0.411 ± 0.093 4 91.8 ± 11.0 4

Benzene ^g/stick 0.575 ± 0.072 4 0.628 ± 0.073 4 100.4 ± 2.8 4

Isoprene ^g/stick 2.44 ± 0.50 4 2.63 ± 0.60 4 869 ± 50 4

Pyridine ^g/stick 9.38 ± 0.95 4 10.08 ± 0.46 4 51.8 ± 7.5 4

Quinoline ^g/stick 0.014 ± 0.002 4 0.010 ± 0.003 4 0.390 ± 0.101 4

Styrene ^g/stick 0.672 ± 0.063 4 0.632 ± 0.079 4 28.9 ± 2.2 4

Toluene ^g/stick 1.61 ± 0.17 4 1.67 ± 0.37 4 198.8 ± 10.9 4

Catechol ^g/stick 16.4 ± 0.6 4 12.8 ± 1.3 4 88.7 ± 2.6 4

o-cresol ^g/stick 0.105 ± 0.017 4 0.059 ± 0.007 4 4.86 ± 0.50 4

rn-cresol ^g/stick 0.042 ± 0.006 4 0.032 ± 0.005 4 3.71 ± 0.34 4

^-cresol ^g/stick 0.073 ± 0.009 4 0.042 ± 0.007 4 8.50 ± 0.78 4

Hydroquinone ^g/stick 7.86 ± 0.63 4 6.21 ± 0.86 4 84.1 ± 3.3 4

Phenol ^g/stick 1.51 ± 0.23 4 1.00 ± 0.17 4 13.2 ± 0.9 4

Resorcinol ^g/stick 0.055 ± 0.013 4 0.036 ± 0.005 4 1.95 ± 0.55 4

NAB ng/stick 3.52 ± 0.48 4 3.27 ± 0.15 4 34.1 ± 3.0 4

NAT ng/stick 22.3 ± 1.6 4 61 18.6 ± 2.9 4 300 ± 53 4

NNK ng/stick 10.1 ± 0.4 4 7.9 ± 1.1 4 257 ± 39 4

NNN ng/stick 10.3 ± 0.4 4 7.7 ± 1.0 4 268 ± 50 4

Ammonia ^g/stick 15.6 ± 1.1 4 13.9 ± 1.1 4 39.2 ± 4.1 4

Hydrogen cyanide ^g/stick 3.78 ± 0.44 4 5.57 ± 0.35 4 451 ± 47 4

Nitric oxide ^g/stick 21.0 ± 8.1 3 18.4 ± 3.6 4 501 ± 33 3

Nitrogen oxides ^g/stick 22.6 ± 8.8 3 19.4 ± 4.0 4 541 ± 74 3

Arsenic ng/stick <1.13 4 <1.13 4 6.56 ± 0.46 i 4

Cadmium ng/stick <0.350 4 <0.350 4 122 ± 12 4

Chromium ng/stick <0.17 4 0.44 4 2.70* 2

Lead ng/stick <3.35 4 <3.35 4 25.1 ± 2.1 4

Mercury ng/stick 1.02 ± 0.05 4 1.12 ± 0.19 4 4.17 ± 0.74 4

Nickel ng/stick <0.55 4 0.88 4 1.30* 2

Selenium ng/stick <0.550 4 <0.550 4 1.43 ± 0.15 4

Pyrene ng/stick 7.93 ± 0.78 4 7.71 ± 0.63 3 87.3 ± 4.1 4

o-toluidine ng/stick 1.204 ± 0.149 4 0.868 ± 0.087 4 90.5 ± 3.1 4

Acetamide ^g/stick 4.13 ± 0.21 4 3.43 ± 0.17 4 13.7 ± 0.7 4

Acrylamide ^g/stick 2.27 ± 0.28 4 1.90 ± 0.12 4 5.3 ± 0.4 4

Ethylene oxide ^g/stick 0.314 ± 0.011 4 0.273 ± 0.036 4 34.2 ± 3.6 4

Nitrobenzene ng/stick 0.092 ± 0.008 4 0.155 ± 0.004 8 0.55 ± 0.04 4

Propylene oxide ^g/stick 0.175 ± 0.03 4 0.14 ± 0.019 4 1.72 ± 0.16 4

Vinyl chloride ng/stick <3.47 4 <3.47 4 95.3 ± 12.3 4

Benz[a] anthracene ng/stick 2.58 ± 0.17 4 S2.50 ± 0.06 3 26.6 ± 1.7 4

Dibenz[a, h] anthracene ng/stick <0.100 4 <0.100 4 1.79 ± 0.14 4

1273 N is the number of determinations, CI is the confidence interval of the mean, n.a.: not analyzed, *: CI not calculated

1274 <: median lower than the limit of quantitation, in this case LOQ is given

1275 If at least one value is below the LOQ, the median is given and the CI is not mentioned

1276 TPM: Total particulate matter, NFDPM: Nicotine-free dry particulate matter, NAB: A-nitrosoanabasine, NAT: A-nitrosoanatabine, NNK: 4-(A-

1277 nitrosomethylamino)-1 -(3-pyridyl)-1 -butanone, NNN: A-nitrosonornicotine

1280 Table 7: Yields on tobacco stick basis from THS2.2 FR1 obtained under three climatic conditions

Mediterranean Desert Tropical

Unit 22°C, 60% RH 30°C, 35% RH 30°C, 75% RH Range1

Parameter mean ± CI95% N mean ± CI95% N mean ± CI95% N

TPM mg/stick 47.1 ± 2.0 5 27.9 ± 1.4 5 65.1 ± 4.5 5 37.2

Water mg/stick 33.5 ± 1.5 5 15.1 ± 0.4 5 50.6 ± 3.7 5 35.5

Nicotine mg/stick 1.42 ± 0.05 5 1.46 ± 0.05 5 1.21 ± 0.02 5 0.25*

NFDPM mg/stick 12.2 ± 1.1 5 11.3 ± 1.4 5 12.6 ± 1.5 5 1.3

Carbon monoxide mg/stick 0.612 ± 0.020 5 0.454 ± 0.047 5 0.468 ± 0.024 5 0.158

Benzo[a]pyrene ng/stick <1.00 5 1.03 5 <1.00 5

Glycerin mg/stick 4.68 ± 0.36 5 4.25 ± 0.25 5 4.23 ± 0.09 5 0.45*

1 -aminonaphthalene ng/stick <0.069 5 <0.069 5 <0.069 5

2-aminonaphthalene ng/stick <0.035 5 <0.035 5 0.088 ± 0.173 5

3 -aminobiphenyl ng/stick <0.032 5 <0.032 5 <0.032 5

4-aminobiphenyl ng/stick <0.051 5 <0.051 5 0.073 ± 0.022 5

Acetaldehyde ^g/stick 193 ± 2 5 179 ± 11 5 229 ± 5 5 50

Acetone ^g/stick 37.7 ± 1.7 5 37.0 ± 5.5 5 43.0 ± 2.9 5 6.0

Acrolein ^g/stick 9.76 ± 0.91 5 8.87 ± 1.81 5 11.54 ± 0.81 5 2.67

Butyraldehyde ^g/stick 27.3 ± 0.7 5 26.0 ± 0.5 5 29.9 ± 2 5 3.9

Crotonaldehyde ^g/stick 4.13 ± 0.55 5 3.69 ± 0.51 5 4.64 ± 0.41 5 0.95

Formaldehyde ^g/stick 3.52 ± 0.3 5 3.65 ± 0.36 5 3.57 ± 0.3 5 0.13*

Methyl ethyl ketone ^g/stick 7.58 ± 0.71 5 6.75 ± 0.84 5 8.68 ± 0.57 5 1.93*

Propionaldehyde ^g/stick 14.4 ± 0.6 5 13.5 ± 2.1 5 18.1 ± 0.7 5 4.6

Acrylonitrile ^g/stick 0.167 ± 0.036 5 0.178 ± 0.021 5 0.189 ± 0.035 5 0.022*

1,3-butadiene ^g/stick 0.277 ± 0.035 5 0.248 ± 0.016 5 0.318 ± 0.027 5 0.070*

Benzene ^g/stick 0.603 ± 0.042 5 0.591 ± 0.031 5 0.613 ± 0.030 5 0.022*

Isoprene ^g/stick 2.19 ± 0.21 5 1.83 ± 0.17 5 2.60 ± 0.25 5 0.77

Pyridine ^g/stick 7.47 ± 0.31 5 5.76 ± 0.42 5 7.26 ± 0.29 5 1.71

Quinoline ^g/stick <0.012 5 <0.012 5 <0.012 5

Styrene ^g/stick 0.640 ± 0.035 5 0.619 ± 0.041 5 0.695 ± 0.038 5 0.076*

Toluene ^g/stick 2.11 ± 0.2 5 1.85 ± 0.15 5 2.25 ± 0.19 5 0.4*

Catechol ^g/stick 16.9 ± 1.2 5 15.0 ± 2.0 5 15.2 ± 1.4 5 1.9

o-cresol ^g/stick 0.109 ± 0.006 4 0.123 ± 0.021 5 0.135 ± 0.017 5 0.026*

rn-cresol ^g/stick 0.031 ± 0.003 4 0.061 ± 0.01 5 0.071 ± 0.029 5 0.040

^-cresol ^g/stick 0.070 ± 0.006 4 0.099 ± 0.019 5 0.099 ± 0.012 5 0.029*

Hydroquinone ^g/stick 8.51 ± 0.63 5 7.11 ± 1.17 5 8.21 ± 0.25 5 1.40

Phenol ^g/stick 1.66 ± 0.36 5 2.49 ± 0.45 5 2.23 ± 0.26 5 0.83

Resorcinol ^g/stick 0.054 ± 0.004 5 0.047 ± 0.007 5 0.044 ± 0.005 5 0.01

NAB ng/stick 3.37 ± 0.21 4 <3.15 5 <3.15 5

NAT ng/stick 21.7 ± 1.4 4 20.1 ± 1.7 5 17.0 ± 0.6 5 4.7

NNK ng/stick 9.2 ± 0.3 4 8.3 ± 1 5 7.7 ± 1.2 5 1.5

NNN ng/stick 16.4 ± 1.3 4 14.6 ± 1.6 5 13.9 ± 0.8 5 2.5*

Ammonia ^g/stick 14.6 ± 0.4 5 15.4 ± 0.2 5 15.3 ± 0.9 5 0.8*

Hydrogen cyanide ^g/stick 4.84 ± 0.29 4 4.14 ± 0.32 5 4.39 ± 0.68 5 0.70*

Nitric oxide ^g/stick 18.3 ± 1.2 5 16.6 ± 0.3 4 18.2 ± 0.2 4 1.7

Nitrogen oxides ^g/stick 19.5 ± 1.5 5 17.6 ± 0.2 4 19.0 ± 0.6 4 1.9

Pyrene ng/stick 5.66 ± 0.61 5 7.27 ± 1.63 5 5.85 ± 1.27 5 1.61

o-toluidine ng/stick 1.144 ± 0.113 5 0.649 ± 0.318 5 0.62 ± 0.59 5 0.524

Acetamide ^g/stick 4.24 ± 0.13 4 4.28 ± 0.55 5 4.19 ± 0.16 5 0.09*

Acrylamide ^g/stick 2.31 ± 0.12 5 2.30 ± 0.31 5 1.94 ± 0.05 5 0.37

Ethylene oxide ^g/stick 0.267 ± 0.017 5 0.269 ± 0.012 5 0.355 ± 0.032 5 0.088

Nitrobenzene ng/stick 0.138 ± 0.003 4 0.087 ± 0.005 5 ^0.087 ± 0.009 5 0.051

Propylene oxide ^g/stick 0.144 ± 0.01 5 0.101 ± 0.004 5 0.113 ± 0.009 5 0.043

Vinyl chloride ng/stick <3.54 5 <3.54 5 <3.54 5

Benz[a] anthracene ng/stick 1.39 ± 0.1 5 1.51 ± 0.35 5 ¿1.39 ± 0.2 5 0.12

Dibenz[a, h] anthracene ng/stick <0.100 5 <0.100 5 <0.100 5

1281 N is the number of determinations, CI is the confidence interval of the mean

1282 1 Range: largest mean obtained among the three climatic conditions minus the smallest mean obtained among the three climatic conditions

1283 * : Range smaller than the Intermediate Precision (IP) (International Council for Harmonisation, 1996) of the analytical method

1284 <: median lower than the limit of quantitation, in this case LOQ is given

1285 If at least 1 value is below LOQ, the median is given and CI is not mentioned

1286 TPM: Total particulate matter, NFDPM: Nicotine-free dry particulate matter, NAB: W-nitrosoanabasine, NAT: W-nitrosoanatabine, NNK: 4-(W-

1287 nitrosomethylamino)-1 -(3-pyridyl)-1 -butanone, NNN: W-nitrosonornicotine

1291 Table 8: Summary of the THS2.2 FR1 yields from extreme puffing regimens; comparison with the Health Canada

1292 machine-smoking regimen and the 3R4F reference cigarette

THS2.2 FR1 THS2.2 FR1 THS2.2 FR1 3R4F

Extreme Regimen Extreme Regimen HCI HCI

Parameter Unit Maximum Yields Minimum Yields Regimen Ratio** regimen

mean puf. reg. mean puf. reg. mean Max/HCI mean

TPM mg/stick 59.0 SR-4 26.9 ISO 56.8 1.0 49.0

Water mg/stick 45.1 SR-4 21.4 ISO 44.6 1.0 15.8

Nicotine mg/stick 2.19 LR-3 0.49 ISO 1.36 1.6 1.89

NFDPM mg/stick 17.4 LR-3 5.1 ISO 10.8 1.6 31.2

Carbon monoxide mg/stick 0.660 SR-6 0.238 ISO 0.532 1.2 32.8

Benzo[a]pyrene ng/stick <1.00 <1.00 <1.00 14.2

Glycerin mg/stick 5.66 LR-3 1.91 ISO 4.59 1.2 2.42

1 -aminonaphthalene ng/stick <0.069 <0.069 <0.069 20.8

2-aminonaphthalene ng/stick <0.035 <0.035 <0.035 11.0

3 -aminobiphenyl ng/stick <0.032 <0.032 <0.032 3.77

4-aminobiphenyl ng/stick <0.051 <0.051 K0.051 3.26

Acetaldehyde ^g/stick 205 SR-4 145 SR-1 196 1.0 1555

Acetone ^g/stick 40.7 SR-4 22.0 SR-1 37.9 1.1 736

Acrolein ^g/stick 12.9 LR-3 4.89 ISO 8.83 1.5 154

Butyraldehyde ^g/stick 26.7 SR-6 16.7 ISO 22.0 1.2 88.4

Crotonaldehyde ^g/stick 4.90 LR-3 1.88 ISO 3.04 1.6 68.8

Formaldehyde ^g/stick 7.73 LR-3 1.85 ISO 3.77 2.1 56.5

Methyl ethyl ketone ^g/stick 7.39 SR-4 3.78 SR-1 7.28 1.0 187

Propionaldehyde ^g/stick 14.4 SR-4 8.5 SR-1 13.5 1.1 125

Acrylonitrile ^g/stick 0.228 LR-3 <0.111 SR-1* 0.163 1.4 31.9

1,3-butadiene ^g/stick 0.357 SR-4 <0.236 ISO* 0.295 1.2 63.8

Benzene ^g/stick 0.708 SR-4 0.298 SR-1 0.597 1.2 97.6

Isoprene ^g/stick 2.82 SR-5 1.37 SR-1 2.56 1.1 798

Pyridine ^g/stick 8.00 SR-4 4.21 ISO 7.36 1.1 36.1

Quinoline ^g/stick 0.051 LR-3 <0.011 ISO* 0.016 3.2 0.513

Styrene ^g/stick 0.686 SR-6 0.314 SR-1 0.619 1.1 24.5

Toluene ^g/stick 2.15 LR-3 0.97 ISO 2.02 1.1 188

Catechol ^g/stick 17.9 SR-4 5.9 ISO 16.8 1.1 91.4

o-cresol ^g/stick 0.438 LR-3 0.021 ISO 0.108 4.1 4.47

rn-cresol ^g/stick 0.212 LR-3 <0.010 ISO* 0.046 4.6 3.03

^-cresol ^g/stick 0.399 LR-3 <0.010 SR-1 0.100 4.0 9.17

Hydroquinone ^g/stick 9.99 SR-6 3.66 SR-1 8.76 1.1 83.1

Phenol ^g/stick 10.87 LR-3 0.06 ISO 1.93 5.6 13.6

Resorcinol ^g/stick 0.056 SR-6 0.020 ISO 0.047 1.2 1.85

NAB ng/stick 4.31 LR-3 <3.15 ISO* 3.46 1.2 33.7

NAT ng/stick 26.8 SR-6 8.5 ISO 22.4 1.2 318

NNK ng/stick 10.2 SR-6 4.1 ISO 8.7 1.2 266

NNN ng/stick 19.1 LR-3 6.5 ISO 16.1 1.2 309

Ammonia ^g/stick 31 LR-3 4.1 ISO 15 2.1 39.3

Nitric oxide ^g/stick 19.4 SR-4 11 ISO 18 1.1 491

Nitrogen oxides ^g/stick 20.3 SR-4 11.2 ISO 19 1.1 537

Pyrene ng/stick 6.50 SR-4 <5.00 ISO* <5.00 87.3

o-toluidine ng/stick 2.146 SR-6 0.489 SR-1 1.195 1.8 85.5

Acetamide ^g/stick 6.62 LR-3 1.32 ISO 4.18 1.6 13.9

Acrylamide ^g/stick 4.23 LR-3 0.69 ISO 2.33 1.8 4.8

Ethylene oxide ^g/stick 0.323 LR-3 0.157 SR-1 0.242 1.3 29.4

Vinyl chloride ng/stick <3.54 <3.54 <3.54 96.7

Benz[a] anthracene ng/stick 1.61 SR-4 <1.00 ISO* <1.00 28.0

Dibenz[a, h] anthracene ng/stick <0.100 <0.100 <0.100 1.7

1293 <: median lower than the limit of quantitation, in this case LOQ is given

1294 *: At least one other smoking regimen was also below LOQ; see Tables D a in the supplementary material

1295 * *: When the, minimum value was inferior to the LOQ, the LOQ value was used to calculate the ratios

1296 TPM: Total particulate matter, NFDPM: Nicotine-free dry particulate matter, NAB: W-nitrosoanabasine, NAT: W-nitrosoanatabine, NNK: 4-(W-

1297 nitrosomethylamino)-1 -(3-pyridyl)-1 -butanone, NNN: W-nitrosonornicotine

1300 Table 9: MMAD and GSD results from 3R4F and THS2.2 FR1

3R4F THS2.2 FR1

Repetition MMAD GSD2 LB UB MMAD GSD2 LB UB

[^m] [^m] [^m] [^m] [^m] [^m]

1 0.9 2.1 0.4 1.8 0.8 2.6 0.3 2.1

2 0.8 1.9 0.4 1.5 0.7 2.3 0.3 1.7

3 0.8 1.9 0.4 1.5 0.7 2.1 0.3 1.4

4 0.9 1.7 0.5 1.5 0.7 1.9 0.4 1.3

5 0.9 1.6 0.5 1.4 0.7 2.0 0.3 1.3

6 0.8 1.9 0.4 1.5 0.7 2.2 0.3 1.6

7 0.9 1.9 0.5 1.6 0.8 2.5 0.3 1.9

8 0.6 1.4 0.5 0.9 0.6 2.3 0.3 1.3

9 0.8 1.8 0.5 1.5 0.6 2.3 0.3 1.5

10 0.7 1.8 0.4 1.3 0.7 3.2 00.2 2.3

Mean 0.8 1.8 0.7 2.3

Mean GSD 1.3 1.5

1301 MMAD: Mass median aerodynamic diameter; GSD: Geometric standard deviation

1302 LB: Lower boundaries with a 95% confidence interval; UB: Upper boundaries with a 95% confidence interval

1305 Table 10: Cytotoxicity of TPM and GVP, expressed as 1/EC50 (ml/mg nicotine)

THS2.2 D2 3R4F analyzed during the THS2.2 D2 study THS2.2 D1 M 3R4F analyzed during the THS2.2 D1 M study

TPM GVP TPM GVP TPM GVP TPM GVP

Mean 17.34 28.40 208.55 289.06 19.73 26.07 239.51 276.21

SEM 0.52 1.20 6.92 22.38 0.87 1.78 6.07 22.51

N 3 3 3 3 3 3 3 3

Relative cytotoxicity (%) 8.3 9.8 100 100 8.2 9.4 100 100

1306 SEM: Standard error of the mean

1307 TPM: Total particulate matter, GVP: Gas-vapor phase

1308 Relative cytotoxicity (%) = (cytotoxicity of THS2.2/3R4F) x 100

1312 Table 11: Cytotoxicity of TPM and GVP, expressed as 1/EC50 (ml/mg nicotine)

THS2.2 FR1 THS2.2 FR1 M 3R4F

TPM GVP TPM GVP TPM GVP

Mean 21.33 30.86 27.43 28.61 186.78 242.01

SEM 1.57 2.87 1.6 2.53 7.58 14.06

N 3 3 3 3 3 3

Relative cytotoxicity (%) 11.4 12.8 14.7 11.8 100 100

1313 SEM: Standard error of the mean

1314 TPM: Total particulate matter, GVP: Gas-vapor phase

1315 Relative cytotoxicity (%) = (cytotoxicity of THS2.2 / 3R4F) x 100

Table 12: Revertant colonies obtained following exposure to the TPM (1 mg per plate) from THS2.2 D2, THS2.2 D1 M, or 3R4F

Salmonella typhimurium Strain THS2.2 D2 3R4F A Solvent Control PositiveB Control THS2.2 D1 M 3R4F A Solvent Control PositiveB Control

Mean* SD Mean* SD Mean* SD Mean* SD Mean* SD Mean* SD Mean* SD Mean* SD

TA98 22 4 658 89 21 1 109 7 21 2 636 24 25 2 97 17

+S9 TA100 94 21 428 25 87 2 481 22 93 12 440 20 90 6 471 67

TA102 358 12 409 15 272 22 1005 29 290 15 399 16 265 23 968 25

TA1535 9 3 17 6 6 1 70 8 15 6 15 6 10 3 113 11

TA1537 8 5 98 9 6 2 50 5 15 3 94 9 7 2 35 5

TA98 16 4 17 5 23 6 81 6.2 22 3 10C 6 26 3 93 10

-S9 TA100 61 8 87 13 66 3 195 25 81 11 96 21 62 8 187 17

TA102 291 15 282 12 267 21 709 5 230 56 264 4 258 26 620 8

TA1535 12 3 7 3 6 2 37 8 9 4 16 6 12 2 51 6

TA1537 6 5 3 2 6 3 84 5 15 5 17 5 7 2 83 6

A: These samples were generated and tested concurrently with the respective THS variant B: Details of dose and substance are provided in the Ames methods section C: Toxicity was detected at this dose

*: Each mean and SD value was derived from 3 plates and the values were rounded

1326 Table 13: The LOGELs (expressed on a nicotine basis) achieved following treatment with TPM (^g/ml) and GVP (^g

1327 nicotine equivalent/ml) derived from THS2.2 variants and 3R4F values are shown when GEF threshold was exceeded.

4 h -S9 4 h +S9 24 h -S9

THS2.2 D2 3R4F TPM 18.66 34.48 41.80 17.31

GVP TPM GVP 71.90 83.17 249.87 300.99 43.14

1.31 A 2.47 69.07 1.91 50.79 1.14 1.05 8.64 12.61

TPM 12.55 27.82 21.71 A

THS2.2 D1 M

GVP 83.30 74.53 240.34 267.92 A

TPM 0.96 1.11 3.05 1.99 A

GVP 16.27 10.311 53.74 6.14

1328 A: Concentration-dependent increase in MF observed but below the GEF threshold