King's Research Portal

King’s Research Portal

Link to publication record in King's Research Portal

Citation for published version (APA): McNeill, A., Fix, B. V., Leigh, N., Palumbo, M., Hitchman, S., Nahhas, G., King, B., Hammond, D., Michael Cummings, K., Borland, R., Fong, G. T., O’Connor, R. J., & Goniewicz, M. L. (Accepted/In press). Characterization of vaping liquids used in vaping devices across four countries: Results from an analysis of selected vaping liquids reported by users in the 2016 ITC Four Country Smoking and Vaping Survey. Tobacco Control.

Citing this paper Please note that where the full-text provided on King's Research Portal is the Author Accepted Manuscript or Post-Print version this may differ from the final Published version. If citing, it is advised that you check and use the publisher's definitive version for pagination, volume/issue, and date of publication details. And where the final published version is provided on the Research Portal, if citing you are again advised to check the publisher's website for any subsequent corrections. General rights Copyright and moral rights for the publications made accessible in the Research Portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognize and abide by the legal requirements associated with these rights.

•Users may download and print one copy of any publication from the Research Portal for the purpose of private study or research. •You may not further distribute the material or use it for any profit-making activity or commercial gain •You may freely distribute the URL identifying the publication in the Research Portal Take down policy If you believe that this document breaches copyright please contact [email protected] providing details, and we will remove access to the work immediately and investigate your claim.

Download date: 27. Sep. 2021 Tobacco Control

Confidential: For Review Only Characterization of vaping liquids used in vaping devices across four countries: Results from an analysis of selected vaping liquids reported by users in the 2016 ITC Four Country Smoking and Vaping Survey Word Count: 3846

Journal: Tobacco Control

Manuscript ID tobaccocontrol-2020-056338.R2

Article Type: Original research

Date Submitted by the n/a Author:

Complete List of Authors: Fix, Brian; Roswell Park Comprehensive Cancer Center, Health Behavior OConnor, Richard; Roswell Park Comprehensive Cancer Center, Dept. of Health Behavior Goniewicz, Maciej; Roswell Park Comprehensive Cancer Center, Department of Health Behavior Leigh, Noel; Roswell Park Comprehensive Cancer Center, Dept. of Health Behavior Cummings, Michael; Medical University of South Carolina, Psychiatry & Behavioral Sciences hitchman, sara; King’s College London, Department of Addictions Fong, Geoffrey; University of Waterloo, Psychology el Nahas, Georges; Medical University of South Carolina, Hollings Cancer Center Hammond, David; University of Waterloo, School of Public Health and Health Systems McNeill, Ann; King’s College London, UK Centre for Tobacco Control Studies, National Addiction Centre, Institute of Psychiatry Borland, Ron; University of Melbourne, School of Psychological Sciences King, Bill; cancer council victoria, vichealth centre for tobacco control Palumbo, Mary; Roswell Park Comprehensive Cancer Center, Dept. of Health Behavior

Keywords: Electronic nicotine delivery devices, Nicotine, Public policy

https://mc.manuscriptcentral.com/tobaccocontrol Page 1 of 47 Tobacco Control

1 2 3 4 5 6 7 8 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 9 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 10 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 11 with the terms applicable for US Federal Government officers or employees acting as part of their official 12 Confidential: For Review Only duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 13 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author unless you are acting as an employee on behalf of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

https://mc.manuscriptcentral.com/tobaccocontrol Tobacco Control Page 2 of 47

1 2 3 Characterization of vaping liquids used in vaping devices across four countries: Results from an analysis of selected vaping 4 5 6 liquids reported by users in the 2016 ITC Four Country Smoking and Vaping Survey 7 8 Authors: Brian Confidential:V. Fix1, Noel J. Leigh1, Mary Palumbo1, Sara For C. Hitchman Review2,3, Georges J. Nahhas4 , OnlyBill King5, David Hammond7, Ann 9 10 McNeill2,3, K. Michael Cummings4, Ron Borland,5,6 Geoffrey T. Fong7,8, Richard J. O’Connor1, Maciej L. Goniewicz1 11 12 Affiliations: 1Roswell Park Comprehensive Cancer Center, Buffalo, New York, USA; 2King’s College London, London, 13 14 15 UK;3Shaping Public hEalth poliCies To Reduce IneqUalities and harm (SPECTRUM) UK; 4Medical University of South Carolina, 16 17 Charleston, South Carolina, USA; 5University of Melbourne, Melbourne, Australia; 7University of Waterloo, Waterloo, Ontario, 18 19 Canada; 8Ontario Institute for Cancer Research, Toronto, Ontario, Canada 20 21 22 23 Funding: This study was supported by grants from the US National Cancer Institute (P01 CA200512), the Canadian Institutes of 24 25 Health Research (FDN‐148477), and by the National Health and Medical Research Council of Australia (APP 1106451). G.T.F. was 26 27 supported by a Senior Investigator Award from the Ontario Institute for Cancer Research. NicoTAR shared resource is supported by 28 29 grants from the National Cancer Institute (NCI) (P30CA016056), and the NCI and the Center for Tobacco Products of the Food and 30 31 32 Drug Administration (U54CA228110). 33 34 Reference Count: 29 35 36 37 Word Count: 3,846 38 39 40 Keywords: Electronic Nicotine Delivery Devices, Nicotine, Public Policy 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 3 of 47 Tobacco Control

1 2 3 Author Contributions: All authors contributed to the analysis of the data presented in this manuscript and have also drafted and 4 5 6 critically revised it. All authors have granted final approval for the work published in this manuscript. 7 8 Confidential: For Review Only 9 10 11 12 What this study adds: 13 14 15  A descriptive comparison of the content of e-liquids selected initially on the basis of popularity with vapers purchased in 4 16 17 countries with different regulations. 18 19 20  Composition, particularly with respect to nicotine and flavoring content, varies by country, indicating both marketplace 21 22 preferences and potential regulatory influence. 23 24  Samples purchased in Canada (n=10) and Australia (n=15) were labeled with a nicotine value of “0” and had no detectable 25 26 27 nicotine present, consistent with the regulations in those two countries. Similarly, among 48 samples purchased in the US, 46 28 29 (95.8%) had detectable mean concentrations within 20% of the labeled value. Among 164 samples purchased in England, only 30 31 94 (57.3%) had detectable mean values within 20% of the labeled value. 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 4 of 47

1 2 3 ABSTRACT 4 5 6 OBJECTIVES: This study presents an analysis of vaping products (VPs) purchased in the United States, Canada, England, and 7 8 Confidential: For Review Only 9 Australia and assesses whether differences in regulations were associated with differences in the chemical composition of the VPs. 10 11 12 METHODS: April-September 2017, a total of 234 VP refill liquids and pre-filled cartridges were purchased in convenience samples 13 14 of retail locations in each country. Products were chosen from brands and styles most commonly reported by current VP users in the 15 16 2016 ITC Four Country Smoking and Vaping Survey. All products were tested for nicotine, tobacco-derived nitrosamines (TSNAs), 17 18 19 minor tobacco alkaloids, organic acids, and flavouring chemicals. 20 21 22 RESULTS: Consistent with the laws in Canada and Australia at the time of product purchase, nicotine was not detected in any of the 23 24 VPs (n=10 in Canada; n=15 in Australia). US liquids (n=54) had a mean nicotine concentration of 16.2 mg/ml, (range=0.0-58.6), and 25 26 English liquids (n=166) had a mean concentration of 11.9 mg/ml [ (range=0.0-31.2) F(3,244)=12.32, p<0.001]. About 5% of English 27 28 29 samples exceeded the United Kingdom’s 20 mg/mL nicotine limit. Substantial country differences were observed in levels of 30 31 tobacco-specific nitrosamines (TSNAs), with the US being higher than elsewhere. Of all products tested, 18.8% contained at least one 32 33 organic acid. Liquids purchased in England contained far more identifiable flavoring chemicals than those in the other countries. 34 35 36 37 CONCLUSIONS: VP composition, particularly with respect to nicotine and flavoring, varies by country, likely reflecting both 38 39 marketplace preferences and country-specific regulations. There are differences between nicotine levels claimed on the package and 40 41 actual levels, particularly in England. 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 5 of 47 Tobacco Control

1 2 3 INTRODUCTION 4 5 6 Use of vaping products (VPs) has increased significantly over past few years in many countries. Use of VPs varies by age, and 7 8 Confidential: For Review Only 9 in the United States and Canada, vaping has become more popular among teens than cigarette smoking [1]. Given patterns of use, 10 11 countries have taken different approaches to nicotine VPs. In Australia, nicotine VPs are considered unauthorized medicines and 12 13 cannot be legally sold, although vaping products without nicotine can be sold in some states. Canada also banned nicotine VPs but 14 15 16 then legalized them in May 2018. In the United Kingdom, a comprehensive regulatory regime was implemented in 2016 with a 17 18 transition period until May 2017, in line with the broader European Union Tobacco Products Directive (TPD), which requires 19 20 manufacturer notification prior to introduction of nicotine-containing products, standards for health warnings and packaging, including 21 22 23 a maximum nicotine concentration (20 mg/mL), tank capacity (2 mL) and refill bottle size (10 mL). Additionally, UK law, the TPD 24 25 and other EU regulations prohibit certain flavoring chemicals and additives, including vitamins, caffeine, taurine, diacetyl, 2,3- 26 27 pentanedione, and ethylene glycol [2]. Products not containing nicotine are not subject to these regulations in the UK. In the US, VPs 28 29 were initially regulated by FDA as unapproved drug delivery devices, but this was blocked by legal action. Nicotine VPs were 30 31 32 incorporated into the tobacco regulatory system via the ‘deeming’ regulation in August 2016. This included mandated warnings about 33 34 nicotine content but did not remove any products from the market or set manufacturing and product performance standards [3]. 35 36 37 Most VPs consumed contain nicotine and are used by cigarette smokers to cut back or stop smoking [4]. Nicotine is a 38 39 pharmacologically active compound which can create and sustain dependence in those using tobacco products. In most nicotine 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 6 of 47

1 2 3 liquids used in VPs, the nicotine is derived from tobacco, providing an exposure route for tobacco-specific toxicants (e.g., tobacco- 4 5 6 specific nitrosamines (TSNAs) -- N’-nitrosonornicotine (NNN), 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), N’- 7 8 nitrosoanabasineConfidential: (NAB), and N’-nitrosoanatabine (NAT)), asFor well as minor Review tobacco alkaloids (nornicotine, Only anatabine, anabasine). 9 10 Tobacco-flavored liquids may also contain extracts of cured tobacco leaves, commonly called Natural Extract of Tobacco (NET) 11 12 liquids [5], and these may also be contaminated with TSNAs. At the same time, evidence suggests that carcinogenic chemicals are 13 14 15 found in lower concentrations in VP aerosols compared to the smoke from cigarettes, and use of nicotine containing VPs is 16 17 substantially less harmful compared to smoking cigarettes [6,7]. [8]. Recently, VPs that contain organic acids have gained popularity 18 19 among users, particularly in the United States, where JUUL dominates the market. The addition of organic acids to vaping liquids 20 21 22 creates nicotine salts, which decreases the harshness of inhaling nicotine into the airways [9], thereby facilitating more rapid uptake of 23 24 unprotonated nicotine in the body [10]. Less harsh aerosol also enables the use of higher nicotine concentrations when paired with 25 26 lower heating temperatures. 27 28 29 Health risks from inhaling aerosols from VPs are likely to depend on the type of products used and patterns of use. The acute 30 31 32 and chronic inhalation effects of propylene glycol (PG) and vegetable glycerin (VG), two primary solvents of nicotine used in VPs, 33 34 have not been well-studied, although it is known that prolonged exposure can result in chronic irritation of the airways.[1] Flavoring 35 36 agents found in vaping products are commonly used in foods and indoor fragrances, but little is known about the effects of chronic 37 38 39 inhalation at the levels that can be produced by vaping products.[1,4] Many companies use ingredients that are commonly used as 40 41 food additives and are classified as Generally Regarded as Safe (GRAS) by regulatory agencies, although that designation only applies 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 7 of 47 Tobacco Control

1 2 3 when ingested, not when inhaled. England has banned certain flavorings such as diacetyl, where concerns about harms have been 4 5 6 raised, and at least some manufacturers have voluntarily stopped using it in other jurisdictions. Other flavoring chemicals are less well 7 8 understood in theConfidential: inhalation context. Some flavorings, such asFor benzaldehyde Review [11] and cinnamaldehyde,[12] Only appear to have negative 9 10 respiratory health effects at the concentrations used in VPs, although their presence in e-liquids is less common than other flavors 11 12 [13,14]. Glycols [15] and acetals [16] formed as reaction products of flavorings with other components of the mixture have been 13 14 15 identified in some samples of e-liquids and their associated aerosols. Although some harmful chemicals (particularly aldehydes) may 16 17 be generated during a product’s use,[17-19] others could be present because of poor quality control over the manufacturing process 18 19 (e.g., use of poorly purified nicotine, use [20] of flavoring chemicals of unknown quality, and stabilizers)[21]. 20 21 22 23 While a number of previous studies have reported on VP contents and emissions, few studies have compared popular VPs 24 25 across countries having different regulatory environments. The study was undertaken to determine if differences in regulatory regimes 26 27 for VP across countries were reflected in the characteristics of VPs themselves ( nicotine, minor tobacco alkaloids, tobacco-specific 28 29 nitrosamines (TSNAs), selected organic acids and flavorings) in Australia, Canada, England, and United States. These analytes were 30 31 32 selected to explore the complexity of liquid composition and highlight areas where impurities might be introduced, and focuses on the 33 34 contents of the unheated liquids used in the products, whether contained in cartridges/pods or as refill liquid. We hypothesized that 35 36 products from countries with stricter regulation towards VPs would have laboratory-measured nicotine concentrations that more 37 38 39 closely match the labelled concentration, and have lower levels of contaminants (minor alkaloids, TSNAs), and flavorings that have 40 41 been shown to have significant inhalation toxicity. 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 8 of 47

1 2 3 METHODS 4 5 6 Acquisition of VPs. Between April and September of 2017, VP liquids and refills were purchased in a convenience sample of retail 7 8 Confidential: For Review Only 9 locations in and around Charleston, SC, USA; Waterloo, ON, Canada; Melbourne, VIC, Australia; and London, England, UK (see 10 11 supplemental material). Products were chosen from brands and styles most commonly reported by current VP users in each country in 12 13 wave 1 of the ITC Four Country Smoking and Vaping Survey (ITC4CV) [22]. Methodological details for each country are available 14 15 16 in Thompson et al. [23] and via the ITC website (http://www.itcproject.org/methods). 17 18 19 We characterized a selection of brands (for closed systems) or types (for open systems) reported by 2,143 participants who 20 21 indicated that they were current daily, weekly, or monthly VP users at the time of the survey. After this initial level of product 22 23 identification, we selected brands in the market that were available in multiple flavors and nicotine levels to allow for intra-brand 24 25 26 comparisons. Tobacco, menthol, and cherry (as a representative fruit) flavors were selected as these were reported to be the most 27 28 commonly reported flavors and were offered by many manufacturers [24]. Labelled nicotine concentrations of 0, 5-20 mg/mL, and (in 29 30 US) >20 mg/mL were targeted to cover a range of commonly reported values, which also allowed us to evaluate the level of 31 32 33 compliance in England with the EU TPD limit of 20 mg/mL nicotine. In total, 245 liquids were purchased across the four sites (166 in 34 35 England, 54 in the US, 15 in Australia, and 10 in Canada). While the number of products purchased in Australia and Canada was 36 37 small, the variety of products available for purchase in Australia and Canada were reflective of the marketplace at the time of 38 39 purchase. 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 9 of 47 Tobacco Control

1 2 3 4 5 6 Measurement of Nicotine Concentration. Samples were analyzed at the Nicotine and Tobacco Product Assessment Resource 7 8 Confidential: For Review Only 9 (NicoTAR) in Roswell Park Comprehensive Cancer Center, Buffalo, NY, USA. After arrival at the laboratory, each product was 10 11 catalogued and assigned a unique sample number. All samples were stored in their original containers in a dark space at 4°C prior to 12 13 analysis, in order to minimize the risk of compound degradation. Aliquots of 10μL of each product were collected from each original 14 15 16 container using the reverse pipetting technique and transferred to chromatography vials pre-filled with 1mL of HPLC-grade methanol 17 18 (Fisher Scientific, Hampton, NH, USA) and 100 µL internal standard solution (1 mg/mL of quinoline [Acros Organics, Hampton, NH, 19 20 USA] in methanol). Analyses were performed using gas chromatography with nitrogen-phosphorous detector (GC-NPD, Agilent, 21 22 23 Santa Clara, CA, USA) using a modified NIOSH 2551 method for determination of nicotine in air (1998). The limit of quantitation 24 25 (LOQ) was 1.0 mg/mL. Recovery, STDEV, SEM and % RSD were calculated for all calibrators and QC’s. Any calibrator or QC that 26 27 was outside 20% recovery was excluded from the calibration curve for the samples. The recovery range was between 80.0-114.9% for 28 29 all runs, the STDEV range was between 0.1-4.9 mg/ml, the SEM Range was between 0.0-1.6 mg/ml and the %RSD was between 1.0- 30 31 32 20.5% for interday runs. 33 34 35 Analysis of TSNAs (NNN, NNK, NAT, NAB). Aliquots of 25 µL of each product were transferred to vials containing 2.5mL of 36 37 100mM ammonium acetate (Fisher Scientific) and 50 µL of mixed internal standard solution (100 ng/mL of deuterated NNN, NNK, 38 39 NAT, and NAB [Toronto Research Chemical, Toronto, ON, Canada] in methanol). Analysis was performed by UPLC-MS/MS on a 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 10 of 47

1 2 3 Waters Acquity UPLC-Xevo TQ-S (BEH C18 column) operating in positive ESI mode with 10mM ammonium formate in water 4 5 6 (Fisher Scientific) and methanol in 1% acetic acid (Fisher Scientific) serving as the mobile phases. LOQs were 1.0 ng/mL for NNK 7 8 and 0.5 ng/mL forConfidential: NNN, NAT, and NAB. Recovery, STDEV, For SEM and Review% RSD were calculated forOnly all calibrators and QC’s. Any 9 10 calibrator or QC that was outside 20% recovery was excluded from the calibration curve for the samples. The recovery range was 11 12 between 81.0-120.0% for all runs, the STDEV range was between 0.0-15.1 ng/ml, the SEM Range was between 0.0-7.6 ng/ml and the 13 14 15 %RSD was between 0.6-31.7% for interday runs. 16 17 18 Identification of flavorings, selected organic acids, and tobacco minor alkaloids. Aliquots of 10μL of each product were collected 19 20 from each original container using the reverse pipetting technique and transferred to chromatography vials pre-filled with 1mL of 21 22 23 HPLC grade dichloromethane (Fisher Scientific). Analyses were performed using gas chromatography mass spectrometry (GC-MS, 24 25 Agilent) using a modification of the NIOSH 2551 method for determination of nicotine in air (1998). Scan data were then matched 26 27 against both the National Institute of Standards and Technology (NIST) and Flavor and Fragrance (FFSCN3) databases to identify the 28 29 total number of chemicals, as well as known flavoring or fragrance chemicals. Match factors >70% on 2/3 replicate runs were 30 31 32 retained and reported. This same process was used to identify the presence of organic acids (benzoic acid, levulinic acid, salicylic 33 34 acid) and minor tobacco alkaloids (myosmine, nicotine oxide, anatabine, anabasine, nicotyrine). 35 36 37 Measurement of liquid pH. A 500µl aliquot of each refill liquid was collected, diluted with 5ml Milli-Q water, vortexed for 5min 38 39 using a VWR VX-2500 multi-vortex, sonicated for 5min using a VWR 75D ultrasonic cleaner, allowed to return to room temp over 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 11 of 47 Tobacco Control

1 2 3 15min, then pH and temperature was measured using a Mettler Toledo SevenCompact pH meter. Measurements were read in triplicate 4 5 6 and the average pH value was reported. 7 8 Confidential: For Review Only 9 Data analysis. All analyses were performed using SPSS 21.0. ANOVA was used to compare mean compound concentrations 10 11 between countries in the VPs purchased and by labeled nicotine value (presence or absence). Because none of the samples purchased 12 13 in Australia or Canada contained nicotine, comparisons were limited to samples purchased in England and the United States. Values 14 15 16 below the limits of quantification were replaced with LOQ / sqrt(2). Flavorings are reported descriptively as their concentrations were 17 18 not calculated. The number of flavoring chemicals identified was used as the dependent variable in this case. 19 20 21 RESULTS 22 23 24 Nicotine Concentration in VPs. Comparing the mean concentration of nicotine in liquid solutions purchased at retail locations in 25 26 these 4 countries, we found the liquids purchased in Canada (n=10) and Australia (n=15) did not contain any measurable nicotine, 27 28 29 whereas liquids purchased in the US (n=54) had a mean nicotine concentration of 16.2 mg/ml (95%CI: 12.2-20.3) and in England the 30 31 liquids purchased (n=166) had a mean concentration of 11.9 mg/ml (95%CI: 10.9-12.9) [F(3,244)=19.29, p<0.001]. In England, 7.8% 32 33 (n=13) and in the US, 16.7% (n=9) contained no detectable nicotine (Table 1). 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 12 of 47

1 2 3 Table 1. Mean Nicotine Concentration in Sampled Products by Country and Nicotine Labeling 4 5 6 7 N Mean Std. Deviation Minimum Maximum 8 AUSTRALIA Confidential:15 ND For Review Only 9 10 CANADA 10 ND 11 ENGLAND 166 11.9 6.6 0.0 31.2 12 13 US 54 16.2 14.9 0.0 58.6 14 Total 245 12.0 11.9 0.0 118.0 15 16 17 Product labeled as containing nicotine 179 13.0 6.2 0.0 37.7 18 19 Product not labeled as containing nicotine 58 3.2 7.5 1.6 29.6 20 Total 237* 10.6 7.8 0.0 37.7 21 22 *8 samples had no values indicated anywhere on the packaging 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 13 of 47 Tobacco Control

1 2 3 Nicotine labelling consistency and compliance. Overall, the correspondence between indicated and measured nicotine was 0.80 4 5 6 (Spearman Correlation, p=0.01, Figure 1). Of 245 analyzable samples, 8 (6 from the US; 2 from England) had no nicotine value listed 7 8 anywhere on theConfidential: packaging (mean=42.23; range=22.1-58.6). For Overall, 11 Reviewsamples, all of which were Only purchased in England, had a 9 10 labeled nicotine value of “0” but had nicotine values greater than 0 (mean=16.95, range=5.5-29.6). Among US samples, nicotine was 11 12 not detected in any of the products with no outward indication of the presence of nicotine, while those products indicating the presence 13 14 15 of nicotine had a mean nicotine concentration of 15.27 mg/ml [F (1,46) = 27.03, p<.001]. Among English samples, the products that 16 17 did not indicate the presence of nicotine had mean concentration of 7.78 mg/ml, while samples that did indicate the presence of 18 19 nicotine had a mean nicotine concentration of 12.34 mg/ml [F (1,162) = 11.51, p<0.001]. Of the 54 liquids tested in the US (where 20 21 22 there is no legal limit), 14 (25.9%) had a measured nicotine level greater than 20 mg/ml. Of the 166 liquids tested in England, 9 23 24 (5.4%) had a measured nicotine level greater than 20 mg/ml, above the legal limit allowed. 25 26 27 28 29 Among the 237 samples labelled with a nicotine value, we examined the relative differences between the labelled nicotine 30 31 concentration and the laboratory analyzed mean nicotine concentration, using a threshold of 20% less or greater than the labeled value 32 33 for each sample, a common threshold in pharmaceuticals to distinguish trivial deviations from potentially impactful mislabeling. 34 35 Overall, 165 of the samples (69.6%), were within 20% of the labeled value; 52 samples (21.2%), had detectable mean nicotine 36 37 38 concentrations that were more than 20% less than the labeled value; 20 samples (8.4%), had detectable mean nicotine concentrations 39 40 that were greater than 20% more than the labeled value. In terms of differences across countries, all samples purchased in Canada 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 14 of 47

1 2 3 (n=10) and Australia (n=15) were labeled with a nicotine value of “0” and had no detectable nicotine present. Among 48 samples 4 5 6 purchased in the US, 46 (95.8%) had detectable mean concentrations within 20% of the labeled value, while 1 sample was more than 7 8 20% below and Confidential:1 sample was greater than 20% above the labeled For value. AmongReview 164 samples purchased Only in England, 94 (57.3%) had 9 10 detectable mean values within 20% of the labeled value, while 51 (31.1%) were below and 19 (11.6%) were above the labeled value 11 12 by >20%, respectively. Because the samples purchased in Canada and Australia did not contain nicotine, chi-square tests were used to 13 14 15 compare US and English samples (chi-square=24.6, p<0.001). In an ANOVA analyzing the labeled nicotine concentration in samples 16 17 characterized as within 20%, >20% below, and >20% above the labeled value, the mean labeled value was higher in samples where 18 19 the tested nicotine value >20% less than the labeled value (14.37; 95%CI: 12.88-15.85) when compared to samples where the tested 20 21 22 value was within 20% of the labeled value (10.42; 90%CI: 9.11-11.74) and or > 20% above the labeled value (4.5; 95%CI: 1.50-7.50) 23 24 [F(2,236)=12.12, p<0.001]. 25 26 Figure 1. Mean measured nicotine versus labeled nicotine level 27 NOTE: The solid red line is a reference line through 0 (perfect correspondence). The black line is the regression fit with 95% confidence band (grey lines). 28 29 30 31 32 33 34 Levels of TSNAs. There were substantial differences across countries in the presence of nitrosamines in refill e-liquids and cartridges 35 36 37 (Table 2). NNN and NNK were most commonly detected in US products. However, despite containing no nicotine, all 15 samples 38 39 from Australia and 10 from Canada contained quantifiable levels of NNK, NAT, and NAB. We found that the concentrations of 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 15 of 47 Tobacco Control

1 2 3 NNN, NAT, and NAB were statistically significantly higher in the US e-liquids tested when compared across countries. As shown in 4 5 6 table 3, there was a trend, although not statistically significant) for products without detectable nicotine to have lower mean levels of 7 8 NNN, NNK, NAT,Confidential: and NAB. When compared to other flavors For (Table 3), Review tobacco-flavored products Only had significantly higher levels of 9 10 NAT (11.8 vs 2.4, p<0.001) and higher, yet not statistically significant values of NNN, NNK, and NAB. A few products were 11 12 identified as significant outliers. Vuse menthol (purchased in US) had NNN and NNK levels an order of magnitude higher than other 13 14 15 products. 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 16 of 47

1 2 3 4 5 6 Table 2. Mean Comparisons of NNN, NNK, NAT, and NAB Levels in 227 VP Samples by Country 7 8 Confidential: For Review Only

9 10 Std. 11 Country N Mean Dev Minimum Maximum df F Sig. 12 NNN US 51 17.1 21.8 0.4 125.8 13 Australia 15 0.4 0.0 0.4 0.4 14 15 Canada 9 0.6 0.5 0.4 1.4 16 England 151 0.9 2.4 0.4 22.0 17 Total 18 226 4.5 12.5 0.4 125.8 3 31.66 P<0.001 19 NNK US 51 6.6 18.0 10.4 120.8 20 21 Australia 15 0.7 0.1 0.7 1.1 22 Canada 10 0.8 0.4 0.7 1.9 23 24 England 151 1.7 4.6 0.5 35.9 25 Total 227 2.7 9.5 0.4 120.8 26 3 3.95 P=0.009 27 NAT US 51 16.6 23.9 0.4 100.4 28 Australia 15 0.4 0.2 0.4 1.2 29 30 Canada 10 0.4 0.1 0.4 0.5 31 England 151 3.0 8.2 0.3 60.0 32 Total 33 227 5.8 14.3 0.3 100.4 3 15.10 P<0.001 34 NAB US 51 4.4 4.4 0.4 45.2 35 36 Australia 15 0.4 0.0 0.4 0.4 37 Canada 10 0.4 0.0 0.4 0.4 38 39 England 151 0.7 1.1 0.4 10.2 40 Total 227 1.5 4.9 0.4 45.2 41 3 8.32 P<0.001 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 17 of 47 Tobacco Control

1 2 3 4 LOQ for each nitrosamine: NNN:0.5ng/ml, NNK: 1.0ng/ml, NAT:0.5ng/ml, NAB:0.5ng/ml 5 6 7 8 Confidential: For Review Only 9 10 11 12 13 14 15 16 17 18 19 20 21 Table 3. Comparison of Mean NNN, NNK, NAT, and NAB Levels in 226 VP Samples by Nicotine Concentration and by 22 23 Flavoring Type (Tobacco vs. Others) 24 25 26 27 28 Std. 29 N Mean Deviation Minimum Maximum df F Sig. 30 NNN No Nicotine 46 1.1 2.3 0.4 10.8 31 Detected 32 Nicotine 33 180 5.3 13.7 0.4 125.8 Detected 34 35 Total 226 4.4 12.4 0.4 125.8 1 4.62 0.03 36 Tobacco 37 81 5.0 9.7 0.4 53.1 Flavoring 38 All Other 145 3.2 14.0 0.4 125.8 39 Flavors 40 Total 226 4.4 12.4 0.4 125.8 1 1.07 0.30 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 18 of 47

1 2 3 NNK No Nicotine 47 0.9 0.8 0.7 4.6 4 Detected 5 Nicotine 6 180 3.1 10.5 0.7 120.8 7 Detected 8 Confidential: For Review1 2.11 Only0.15 Total 227 2.7 9.5 0.7 120.8 9 Tobacco 10 82 3.5 7.3 0.7 44.5 Flavoring 11 All Other 145 1.3 10.8 0.7 120.8 12 Flavors 13 Total 227 2.7 9.5 0.7 120.8 1 2.21 0.14 14 NAT No Nicotine 15 47 0.7 1.3 0.4 7.5 16 Detected 17 Nicotine 180 3.8 9.0 0.4 60.0 18 Detected 19 20 Total 227 3.1 8.2 0.4 60.0 1 3.77 0.05 21 Tobacco 82 4.3 8.0 0.4 44.4 22 Flavoring All Other 23 145 2.0 8.2 0.4 60.0 24 Flavors 25 Total 227 3.1 8.2 0.4 60.0 1 24.65 <0.001 26 NAB No Nicotine 47 2.0 7.8 0.4 40.6 27 Detected 28 Nicotine 29 180 4.8 14.0 0.4 100.4 Detected 30 31 Total 227 4.2 13.1 0.4 100.4 1 2.27 0.13 32 Tobacco 82 8.6 19.3 0.4 100.4 33 Flavoring 34 All Other 145 0.9 3.6 0.4 34.8 35 Flavors 36 Total 227 4.2 13.1 0.4 100.4 1 5.51 0.02 37 38 39 LOQ for each nitrosamine: NNN:0.5ng/ml, NNK: 1.0ng/ml, NAT:0.5ng/ml, NAB:0.5ng/ml 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 19 of 47 Tobacco Control

1 2 3 Flavorings. With respect to liquid contents, because a quantitative analysis was not performed, we compared countries and product 4 5 6 types on how many chemicals were identified as present. This was found to vary substantially across products and across countries. 7 8 Liquids and refillsConfidential: purchased in England contained far more Foridentifiable chemicalsReview than those in other Only countries (Figure 2). Liquids 9 10 purchased in the US contained the fewest number of flavorings (Figure 2). Table 4 presents mean chemicals and flavorings in liquids 11 12 by presence of nicotine and tobacco flavor, while table 5 (see supplemental material) provides examples of the most commonly found 13 14 15 chemicals identified across brands in each country. 16 17 18 Figure 2. Average Number of Chemicals and Flavoring Chemicals Detected in Sampled Products by Country. 19 20 21 Figure notes: 22 Average number of chemicals: F(3,241)=61.37, p<0.001 23 Average number of flavoring chemicals: F(3,241)=8.27, p<0.001 24 Significant differences (p<0.001) were observed between the average number of chemicals in samples purchased the United States and England and between the average number 25 of chemicals in samples purchased in Australia and England (Bonferroni Comparison) 26 Significant differences (p<0.001) were observed between the average number of flavoring chemicals in samples purchased in the United States and England (Bonferroni 27 Comparison) 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 20 of 47

1 2 3 Table 4. Mean Chemicals and Flavorings Detected by Nicotine and Tobacco Flavor Presence 4 5 6 7 8 Confidential:Std. For Review Only N Mean Dev Minimum Maximum df F Sig. 9 Chemicals Detected No Nicotine 47 25.19 18.85 4 89 10 Detected 11 Nicotine 198 34.49 20.31 2 99 12 Detected 13 Total 245 32.71 20.33 2 99 1 8.18 0.005 14 Chemicals Detected Tobacco 15 89 26.94 12.60 2 63 Flavoring 16 All Other 156 35.99 23.04 4 99 17 Flavors 18 Total 245 32.71 20.33 2 99 1 19 11.72 0.001 Flavorings Detected No Nicotine 20 47 14.32 11.58 2 59 21 Detected Nicotine 22 198 11.70 9.52 1 49 Detected 23 24 Total 245 12.20 9.98 1 59 1 2.63 0.106 Flavorings Detected Tobacco 25 89 9.96 6.54 1 27 26 Flavoring All Other 27 156 13.49 11.31 1 59 28 Flavors 29 Total 245 12.20 9.98 1 59 1 7.29 0.007 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 21 of 47 Tobacco Control

1 2 3 Tobacco Minor Alkaloids. Approximately half the samples contained minor tobacco alkaloids nicotyrine (55.9%) and/or myosmine 4 5 6 (44.5%). Anatabine, Anabasine, and nicotine oxide were not identified in any samples across all countries. 7 8 Organic acids andConfidential: pH. Finally, we examined the pH of the Forrefill liquids Review and cartridge contents, and Only the presence of any organic acids 9 10 that might be employed to change the pH (i.e., use of nicotine salts). Because pH is particularly relevant for nicotine delivery and 11 12 none of the samples purchased in Australia and Canada contained nicotine, analyses were restricted to samples purchased in England 13 14 15 and the United States. In samples purchased in both countries, mean pH levels (overall and within each country) were higher in 16 17 samples where nicotine was detected, compared to those samples where no nicotine was detected. Overall, no significant differences 18 19 between mean pH levels in samples where organic acids were present, compared to those without organic acids, though mean pH was 20 21 22 higher in products with no organic acids present. In products purchased in England, those with no organic acids detected had a mean 23 24 pH of 8.5, compared to a mean of 7.6 in samples with organic acids present (p<0.001). The inverse of this relationship was observed 25 26 in samples purchased in the United States (p=0.57). 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 22 of 47

1 2 3 Table 5. Mean pH in products purchased in England and the United States 4 Lower 5 F p-value 6 Overall N Mean CI Upper CI Minimum Maximum 7 no nicotine detected 47 4.7 4.4 5.1 3.4 9.4 784.9 <0.001 8 Confidential: For Review Only nicotine detected 187 8.6 8.5 8.7 4.1 9.6 9 10 US Only 11 no nicotine detected 9 4.4 4.2 4.6 4.1 4.8 297.9 <0.001 12 13 nicotine detected 45 8.5 8.3 8.8 5.9 9.4 14 England Only 15 no nicotine detected 13 5.2 4.1 6.2 3.7 9.4 185.9 <0.001 16 17 nicotine detected 132 8.6 8.5 8.7 4.1 9.6 18 Overall 19 20 no organic acids detected 189 7.9 7.6 8.1 3.4 9.6 0.63 0.43 21 organic acids detected 45 7.6 7.2 8.1 3.8 9.1 22 US Only 23 24 no organic acids detected 52 7.8 7.4 8.3 4.1 9.4 0.33 0.57 25 organic acids detected 2 8.5 3.1 13.5 8.1 8.9 26 27 England Only 28 no organic acids detected 114 8.5 8.3 8.8 3.7 9.6 11.45 <0.001 29 organic acids detected 41 7.6 7.3 8.2 3.8 9.1 30 31 32 33 34 Of all products tested, 45 (18.4%) samples contained at least one organic acid, with benzoic acid being most common (11.8%). 35 36 Only 6% contained levulinic acid, and no salicylic acid was detected in any of the samples analyzed. Significant differences across 37 38 39 country with 25.0% of England samples containing at least one organic acid, compared with 10.0% in Canada, 3.7% in the US, and 40 41 6.7% in Australia (chi square=14.63, p=0.002). 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 23 of 47 Tobacco Control

1 2 3 DISCUSSION 4 5 6 The current study provides a cross-country characterization of vaping liquids matched on nominal nicotine and flavour. 7 8 Confidential: For Review Only 9 Consistent with the laws in Canada and Australia at the time of sample collection, none of the e-liquids purchased in Canada (n=10) or 10 11 Australia (n=15) contained nicotine, whereas US liquids (n=54) had a mean nicotine concentration of 16.2 mg/ml and England liquids 12 13 (n=166) had a mean concentration of 11.9 mg/ml [F(3,244)=19.21, p<0.001]. By contrast, in the US and England the levels of nicotine 14 15 16 in e-liquids tested were substantially higher and were generally labelled accurately. 17 18 19 A qualitative analysis of flavoring chemicals as a potential indicator of product complexity and response to regulatory 20 21 demands found that e-liquids and refills purchased in England contained far more identifiable additives than those in the other 22 23 countries. US liquids contained the fewest number of chemicals. This is in spite of the fact that the range of product flavour labels 24 25 26 (i.e., tobacco, menthol, cherry) was largely overlapping. It is unclear to us why the e-liquids in England contained more chemicals. It 27 28 is possible that it derives from the mandate in EN that nicotine-containing VP additives adhere to food standards, where constituents at 29 30 levels above 0.1% are reportable on the label, whereas those less than 0.1% are considered confidential or trade secret. Manufacturers 31 32 33 of e-liquids may use lower concentrations of a greater number of components in order to keep their formulations secret in the UK. In 34 35 the US, there is no such threshold, and thus manufacturers may simplify their formulations. The observation of numerous flavor 36 37 chemicals in e-liquids also speaks to the importance of distinguishing a product’s overall ‘flavour’ as a label descriptor from the 38 39 presence or absence of actual flavoring chemicals. Multiple chemicals or combinations of chemicals may be used to achieve the same 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 24 of 47

1 2 3 experienced flavor. Although this study examines the number of flavourings present as a potential indicator of the complexity of a 4 5 6 product, there is some evidence suggesting the number of flavourings present in VPs can be harmful at a cellular level. In a study 7 8 comparing the cytotoxicityConfidential: of VP refill liquids using embryonic For and adult Review cells, cytotoxicity was found Only to be correlated with the 9 10 number of and concentration of flavouring chemicals [25].Another study, using gas chromatography-mass spectrometry, found that 11 12 the more flavouring chemicals in a VP liquid, the more it was likely the liquid was to be toxic [26] 13 14 15 16 17 18 19 20 21 22 We found detectable levels of TSNAs in VPs in all four countries, with levels substantially higher and more varied in the US. 23 24 Substantial differences were observed in levels of NNN, NNK, and NAT, with the US being higher than England. Though not 25 26 significant, there was a trend toward tobacco-flavored products containing higher levels of nitrosamines, and also for nicotine-free VP 27 28 29 to contain lower levels of nitrosamines. One would expect that nicotine-free VPs should not contain nitrosamines, and yet some did, 30 31 which may indicate poor quality control and constitutes a potential target for better regulation and enforcement of good manufacturing 32 33 product standards. For many of these chemicals, the levels produced under normal vaping are often well below levels seen in cigarette 34 35 36 smoke, so that an immediate focus should be on those e-liquids where high levels can be achieved within the range of normal vaping 37 38 patterns. Combined, these findings may point to the use of lower-purity sources of nicotine, and/or contamination of tobacco-derived 39 40 flavourings by nitrosamines during extraction or concentration. 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 25 of 47 Tobacco Control

1 2 3 Interestingly, in this convenience sample of purchased e-liquids in 2017, we found about 19% of products contained organic 4 5 6 salts, with benzoic acid being the most common (11% of products). Salt-based formulations with higher pH became much more 7 8 widespread in 2018,Confidential: suggesting that our purchases may have Forpicked up on Review the early emergence of aOnly novel formulation. This 9 10 demonstrates the value of ongoing market surveillance since it is important to have an understanding of how products may be 11 12 changing over time in order to assess changes in patterns of product use. 13 14 15 16 This study sought to examine only a subset of products selected based upon self-reports of preference among ITC survey 17 18 participants. As such, the results from this study should not be taken as comprehensive assessment of the marketplace of e-liquids. 19 20 While the products were stored under standardized conditions once acquired by our investigators, we do not know the provenance of 21 22 23 the products prior to purchase, such as how long they were on the shelf or what environmental conditions they had experienced prior 24 25 to our purchasing them. Without a manufacture date or known gap to “best before” date, and of how it is kept prior to receipt at our 26 27 lab, the level of breakdown of the components present at the time of manufacture cannot be estimated. However, the storage of 28 29 products used by vapers would be similar. 30 31 32 33 The regulatory environment in the US has evolved since the time of data collection. Youth use of VPs, including JUUL, grew 34 35 dramatically from 2017-19. In 2019, pursuant to a court order, the compliance dates for VPs to meet FDA’s testing, reporting, and 36 37 premarket authorization requirements [27] were moved forward from 2022 to 2020 [28]. Finally, we had limited samples from 38 39 Canada and Australia. The sample of brands purchased in Canada (and in Australia) reflected only the stores from which products 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 26 of 47

1 2 3 were purchased and not the overall market, in which products are obtained from the internet [29]. In addition, since 2017, the market 4 5 6 has changed in Canada with formal legalization of retail sales of e-liquids, while in Australia restrictions on the sale of vaping devices 7 8 and e-liquids haveConfidential: remained and in two states (Queensland andFor Western Australia)Review have become evenOnly more stringent with complete 9 10 bans on the sale of VPs products. Despite these changes, we have continued to acquire the most popular vaping products reportedly 11 12 used in each country so that we can continue to track the evolution of VPs. 13 14 15 16 CONCLUSIONS 17 18 The current study provides a descriptive comparison of the content of e-liquids selected initially on the basis of popularity with 19 20 vapers purchased in countries with different regulations. Not surprisingly in the two countries where the sale of nicotine was not 21 22 23 allowed at the time of purchase, none of the products tested had nicotine in them. VP composition, particularly with respect to 24 25 nicotine and flavoring content, varies by country, indicating both marketplace preferences and potential regulatory influence. Future 26 27 research should continue to elucidate cross-country differences in VP composition and performance as the market continues to evolve. 28 29 The current study points to a range of possibilities for the future regulation of VPs to minimize harm to users. The available 30 31 32 evidence is consistent with TSNA being extremely potent carcinogens. As this study has shown, two contamination pathways are 33 34 leading to vapers being exposed to TSNA. More stringent regulation of nicotine solutions used in vaping fluids may be the most 35 36 effective means of reducing vapers’ exposures to TSNA. Regulation of permissible levels of contaminants in tobacco extracts or even 37 38 39 prohibition of the use of tobacco extracts in vaping fluids may also be effective in reducing vapers’ exposures to TSNA. The findings 40 41 also suggest the need for regularized product surveillance activities to track changes in the marketplace over time, and to identify 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 27 of 47 Tobacco Control

1 2 3 emergent trends such as the use of organic acids, in their nascent stages. This could provide early warning as to products that could 4 5 6 emerge as potentially problematic. 7 8 Confidential: For Review Only 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 28 of 47

1 2 3 Disclaimers 4 5 6 The sponsors had no role in the design and conduct of the study; collection, management, analysis and interpretation of the data; 7 8 preparation, reviewConfidential: or approval of the manuscript; and decision For to submit Review for publication. The views Only and opinions expressed in this 9 10 letter are those of the authors and do not necessarily represent the official position of the US National Institutes of Health or the US 11 12 Food and Drug Administration. Ann McNeill is a National Institute for Health Research (NIHR) Senior Investigator. The views 13 14 15 expressed in this article are those of the authors and not necessarily those of the NIHR, or the Department of Health and Social Care. 16 17 18 19 Competing Interests 20 21 22 K. Michael Cummings and David Hammond have each served as paid expert witnesses in legal challenges against cigarette 23 24 companies. 25 26 Geoffrey T. Fong has served as a paid expert witness on behalf of governments in litigation involving the tobacco industry. 27 28 29 Maciej L. Goniewicz reports research grant from Pfizer and membership in scientific advisory board to Johnson & Johnson, outside 30 31 the scope of this paper. 32 33 34 35 Author Contributions 36 37 38 39 All authors contributed to the analysis of the data presented in this manuscript and have also drafted and critically revised it. All 40 41 authors have granted final approval for the work published in this manuscript. 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 29 of 47 Tobacco Control

1 2 3 4 5 6 Acknowledgements 7 8 Confidential: For Review Only 9 10 We would like to thank Elle Wadsworth at KCL for buying products in England. 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 30 of 47

1 2 3 4 5 6 7 8 Confidential: For Review Only 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 31 of 47 Tobacco Control

1 2 3 References 4 5 6 7 8 1. HammondConfidential: D, Reid J, Rynard V, Fong G, Cummings ForK, McNeill AReview et al. Prevalence of vaping Only and smoking among adolescents 9 10 in Canada, England, and the United States: repeat national cross sectional surveys. BMJ. 2019;:l2219. 11 12 13 2. E-cigarettes: regulations for consumer products [Internet]. GOV.UK. [cited 11 February 2020]. Available from: 14 https://www.gov.uk/guidance/e-cigarettes-regulations-for-consumer-products. 15 16 17 3. Deeming Tobacco Products To Be Subject to the Federal Food, Drug, and Cosmetic Act, as Amended by the Family Smoking 18 19 Prevention and Tobacco Control Act; Restrictions on the Sale and Distribution of Tobacco Products and Required Warning 20 Statements for Tobacco Products [Internet]. Federal Register. 2020 [cited 12 March 2020]. Available from: 21 22 https://www.federalregister.gov/documents/2016/05/10/2016-10685/deeming-tobacco-products-to-be-subject-to-the-federal- 23 24 food-drug-and-cosmetic-act-as-amended-by-the 25 26 4. Yong H, Borland R, Cummings K, Gravely S, Thrasher J, McNeill A et al. Reasons for regular vaping and for its 27 28 discontinuation among smokers and recent ex‐smokers: findings from the 2016 ITC Four Country Smoking and Vaping 29 30 Survey. Addiction. 2019;114(S1):35-48. 31 32 5. Belushkin M, Tafin Djoko D, Esposito M, Korneliou A, Jeannet C, Lazzerini M et al. Selected Harmful and Potentially 33 34 Harmful Constituents Levels in Commercial e-Cigarettes. Chemical Research in Toxicology. 2019;33(2):657-668. 35 36 37 38 39 6. Daynard R. Public health consequences of e-cigarettes: a consensus study report of the National Academies of Sciences, 40 41 Engineering, and Medicine. Journal of Public Health Policy. 2018;39(3):379-381. 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 32 of 47

1 2 3 7. Nicotine without smoke: Tobacco harm reduction [Internet]. RCP London. 2019 [cited 2020Mar12]. Available from: 4 5 https://www.rcplondon.ac.uk/projects/outputs/nicotine-without-smoke-tobacco-harm-reduction. 6 7 8. Farsalinos K, Gillman I, Melvin M, Paolantonio A, Gardow W, Humphries K et al. Nicotine Levels and Presence of Selected 8 Confidential: For Review Only 9 Tobacco-Derived Toxins in Tobacco Flavoured Electronic Cigarette Refill Liquids. International Journal of Environmental 10 11 Research and Public Health. 2015;12(4):3439-3452. 12 13 14 9. Keithly L, Ferris Wayne G, Cullen D, Connolly G. Industry research on the use and effects of levulinic acid: A case study in 15 cigarette additives. Nicotine & Tobacco Research. 2005;7(5):761-771. 16 17 18 10. Pankow J. A Consideration of the Role of Gas/Particle Partitioning in the Deposition of Nicotine and Other Tobacco Smoke 19 20 Compounds in the Respiratory Tract. Chemical Research in Toxicology. 2001;14(11):1465-1481. 21 22 11. Kosmider L, Sobczak A, Prokopowicz A, Kurek J, Zaciera M, Knysak J et al. Cherry-flavoured electronic cigarettes expose 23 24 users to the inhalation irritant, benzaldehyde. Thorax. 2016;71(4):376-377. 25 26 12. Clapp P, Lavrich K, van Heusden C, Lazarowski E, Carson J, Jaspers I. Cinnamaldehyde in flavored e-cigarette liquids 27 28 temporarily suppresses bronchial epithelial cell ciliary motility by dysregulation of mitochondrial function. American Journal 29 30 of Physiology-Lung Cellular and Molecular Physiology. 2019;316(3):L470-L486. 31 32 13. Czoli CD, Goniewicz ML, Palumbo M, Leigh N, White CM, Hammond D. Identification of flavouring chemicals and potential 33 34 toxicants in e-cigarette products in Ontario, Canada. Can J Public Health. 2019;110(5):542-550. doi:10.17269/s41997-019- 35 36 00208-1. 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 33 of 47 Tobacco Control

1 2 3 14. Havermans A, Krüsemann EJZ, Pennings J, et alNearly 20 000 e-liquids and 250 unique flavour descriptions: an overview of 4 5 the Dutch market based on information from manufacturersTobacco Control Published Online First: 04 November 2019. doi: 6 7 10.1136/tobaccocontrol-2019-055303. 8 Confidential: For Review Only 9 15. Erythropel H, Jabba S, DeWinter T, Mendizabal M, Anastas P, Jordt S et al. Formation of flavorant–propylene Glycol Adducts 10 11 With Novel Toxicological Properties in Chemically Unstable E-Cigarette Liquids. Nicotine & Tobacco Research. 12 13 2018;21(9):1248-1258. 14 15 16. Duell A, McWhirter K, Korzun T, Strongin R, Peyton D. Sucralose-Enhanced Degradation of Electronic Cigarette Liquids 16 17 during Vaping. Chemical Research in Toxicology. 2019;32(6):1241-1249. 18 19 20 17. Talih S, Salman R, Karaoghlanian N, El-Hellani A, Saliba N, Eissenberg T et al. “Juice Monsters”: Sub-Ohm Vaping and 21 Toxic Volatile Aldehyde Emissions. Chemical Research in Toxicology. 2017;30(10):1791-1793. 22 23 24 18. EL-Hellani A, Salman R, El-Hage R, Talih S, Malek N, Baalbaki R et al. Nicotine and Carbonyl Emissions From Popular 25 26 Electronic Cigarette Products: Correlation to Liquid Composition and Design Characteristics. Nicotine & Tobacco Research. 27 2016;:ntw280. 28 29 30 19. Conklin D, Ogunwale M, Chen Y, Theis W, Nantz M, Fu X et al. Electronic cigarette-generated aldehydes: The contribution 31 32 of e-liquid components to their formation and the use of urinary aldehyde metabolites as biomarkers of exposure. Aerosol 33 Science and Technology. 2018;52(11):1219-1232. 34 35 36 20. McNeill A. Levels of toxins in oral tobacco products in the UK. Tobacco Control. 2006;15(1):64-67. 37 38 21. Vreeke S, Peyton D, Strongin R. Triacetin Enhances Levels of Acrolein, Formaldehyde Hemiacetals, and Acetaldehyde in 39 40 Electronic Cigarette Aerosols. ACS Omega. 2018;3(7):7165-7170. 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 34 of 47

1 2 3 4 5 6 22. O’Connor R, Fix B, McNeill A, Goniewicz M, Bansal‐Travers M, Heckman B et al. Characteristics of nicotine vaping 7 products used by participants in the 2016 ITC Four Country Smoking and Vaping Survey. Addiction. 2019;114(S1):15-23. 8 Confidential: For Review Only 9 10 23. Thompson M. Methods of the International Tobacco Control (ITC) Four Country Survey. Tobacco Control. 11 12 2006;15(suppl_3):iii12-iii18. 13 14 24. O’Connor R, Fix B, McNeill A, Goniewicz M, Bansal‐Travers M, Heckman B et al. Characteristics of nicotine vaping 15 16 products used by participants in the 2016 ITC Four Country Smoking and Vaping Survey. Addiction. 2019;114(S1):15-23. 17 18 19 25. Bahl, Vasundhra et al. “Comparison of electronic cigarette refill fluid cytotoxicity using embryonic and adult 20 models.” Reproductive toxicology (Elmsford, N.Y.) vol. 34,4 (2012): 529-37. doi:10.1016/j.reprotox.2012.08.001 21 22 23 26. Sassano, M Flori et al. “Evaluation of e-liquid toxicity using an open-source high-throughput screening assay.” PLoS 24 biology vol. 16,3 e2003904. 27 Mar. 2018, doi:10.1371/journal.pbio.2003904 25 26 27. Premarket Tobacco Product Applications (PMTA) for ENDS [Internet]. U.S. Food and Drug Administration. 2020 [cited 11 27 28 February 2020]. Available from: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/premarket- 29 30 tobacco-product-applications-electronic-nicotine-delivery-systems-ends 31 32 28. Statement on the agency’s actions to tackle the epidemic of youth vaping and court ruling on application submission deadlines 33 for certain tobacco products, including e-cigarettes [Internet]. U.S. Food and Drug Administration. 2020 [cited 11 February 34 35 2020]. Available from: https://www.fda.gov/news-events/press-announcements/statement-agencys-actions-tackle-epidemic- 36 youth-vaping-and-court-ruling-application-submission 37 38 39 29. Czoli C, Goniewicz M, Palumbo M, White C, Hammond D. E-cigarette nicotine content and labelling practices in a restricted 40 market: Findings from Ontario, Canada. International Journal of Drug Policy. 2018;58:9-12. 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 35 of 47 Tobacco Control

1 2 3 Reviewer(s)' Comments to Author: 4 5 6 Most comments have been adequately addressed. As one of the reviewers pointed out, they need to be 7 clear what are they contributing to the literature. From their responses, it is data from countries with 8 different regulatory environments. The introduction could be more focused on the regulatory 9 environment in each country and if the dates policies have been implemented coincide with when the 10 samples Confidential:were purchased. It might be good for For Table 1 toReview include a column on Only the regulatory 11 environment in each country and dates. On the tables, they might want to consider removing the df and 12 F values. On page 25, line 6, authors might want to clarify what "within and across countries" refers to. 13 It might be understood as if samples could have been purchased from several regions within the same 14 country. 15 16 17 Response: The authors appreciate the suggestions and have made the following revisions: 18 19 In order to better focus the introduction on what the manuscript adds to the literature, we moved the 20 paragraph describing the regulatory environments in each of the 4 countries from the end of the 21 introduction section to the first paragraph of the introduction. 22 23 “Use of vaping products (VPs) has increased significantly over past few years in many countries. 24 25 Use of VPs varies by age, and in the United States and Canada, vaping has become more popular 26 among teens than cigarette smoking [1]. Given patterns of use, countries have taken different 27 approaches to nicotine VPs. In Australia, nicotine VPs are considered unauthorized medicines 28 and cannot be legally sold, although vaping products without nicotine can be sold in some states. 29 Canada also banned nicotine VPs but then legalized them in May 2018. In the United Kingdom, 30 a comprehensive regulatory regime was implemented in 2016 with a transition period until May 31 2017, in line with the broader European Union Tobacco Products Directive (TPD), which 32 33 requires manufacturer notification prior to introduction of nicotine-containing products, 34 standards for health warnings and packaging, including a maximum nicotine concentration (20 35 mg/mL), tank capacity (2 mL) and refill bottle size (10 mL). Additionally, UK law, the TPD and 36 other EU regulations prohibit certain flavoring chemicals and additives, including vitamins, 37 caffeine, taurine, diacetyl, 2,3-pentanedione, and ethylene glycol [2]. Products not containing 38 nicotine are not subject to these regulations in the UK. In the US, VPs were initially regulated 39 40 by FDA as unapproved drug delivery devices, but this was blocked by legal action. Nicotine VPs 41 were incorporated into the tobacco regulatory system via the ‘deeming’ regulation in August 42 2016. This included mandated warnings about nicotine content but did not remove any products 43 from the market or set manufacturing and product performance standards [3].” 44 45 The authors also added some detail to the objectives, at the end of the introduction. 46 47 “While a number of previous studies have reported on VP contents and emissions, few studies 48 have compared popular VPs across countries having different regulatory environments. The 49 study was undertaken to determine if differences in regulatory regimes for VP across countries 50 were reflected in the characteristics of VPs themselves ( nicotine, minor tobacco alkaloids, 51 52 tobacco-specific nitrosamines (TSNAs), selected organic acids and flavorings) in Australia, 53 Canada, England, and United States. These analytes were selected to explore the complexity of 54 liquid composition and highlight areas where impurities might be introduced and focuses on the 55 contents of the unheated liquids used in the products, whether contained in cartridges/pods or as 56 refill liquid. We hypothesized that products from countries with stricter regulation towards VPs 57 58 59 60 https://mc.manuscriptcentral.com/tobaccocontrol Page 37 of 47 Tobacco Control

1 2 3 would have laboratory-measured nicotine concentrations that more closely match the labelled 4 concentration, and have lower levels of contaminants (minor alkaloids, TSNAs), and flavorings 5 6 that have been shown to have significant inhalation toxicity.” 7 8 9 10 11 The authorsConfidential: considered adding the policy details For and implementation Review dates Onlyto table 1 as suggested, 12 but ultimately chose to describe these in the manuscript text. Given the results reported in the 13 manuscript are a point-in-time observation in each of the countries rather than a compliance 14 evaluation, we decided to leave the descriptions as-is. We also elected to keep the df’s and F values in 15 the tables to provide more context and information for the reader. 16 17 18 19 20 21 The authors also added clarification to the sentence on page 25, as suggested. 22 23 “With respect to liquid contents, because a quantitative analysis was not performed, we 24 compared countries and product types on how many chemicals were identified as present. This 25 26 was found to vary substantially across products and across countries.” 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/tobaccocontrol Tobacco Control Page 38 of 47

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Figure 1. Mean measured nicotine versus labeled nicotine level 27 NOTE: The solid red line is a reference line through 0 (perfect correspondence). The black line is the regression fit with 95% confidence band (grey lines). 28 29 30 225x132mm (96 x 96 DPI) 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/tobaccocontrol Page 39 of 47 Tobacco Control

1 2 3 4 5 6 7 8 9 10 11 Confidential: For Review Only 12 13 14 15 16 17 18 19 20 21 163x65mm (150 x 150 DPI) 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/tobaccocontrol Tobacco Control Page 40 of 47

1 2 3 SUPPLEMENTAL TABLES 4 5 6 Brands Purchased In Each Country 7 8 Confidential: For Review Only 9 Nominal Mean Country 10 Nicotine Nicotine 11 Brand Flavor (mg/ml) (mg/ml) 12 United States Vuse Menthol None 54.5 13 United States Vuse Original None 58.6 14 United States Vuse Original None 52.5 15 United States Vuse Original None 22.1 16 United States Vuse Original None 45.6 17 United States Vuse Mint None 47.8 18 United States Mark Ten XL Classic 24 37.7 19 United States Mark Ten XL Menthol 40 36 20 United States Mark Ten XL Classic 40 35.8 21 United States Blu Classic Tobacco 24 19.3 22 United States Blu Classic Tobacco 24 24.6 23 24 United States Blu Classic Tobacco 24 25.8 25 United States Blu Classic Tobacco 24 24.6 26 United States Blu Glacier Mint 24 24.5 27 United States Blu Cherry Crush 24 25 28 United States V2 Red Tobacco Flavor 18 12.3 29 United States V2 Menthol 18 15.9 30 United States V2 Red Tobacco Flavor 18 14.9 31 United States My Start E-Liquid Menthol 0 ND 32 United States My Start E-Liquid Menthol 0 ND 33 United States My Start E-Liquid Menthol 0 ND 34 United States My Start E-Liquid Menthol 6 6 35 United States My Start E-Liquid Menthol 6 5.5 36 United States My Start E-Liquid Menthol 6 5.4 37 United States My Start E-Liquid Menthol 11 10.2 38 United States My Start E-Liquid Menthol 11 10.5 39 United States My Start E-Liquid Menthol 11 10.2 40 United States My Start E-Liquid Menthol 18 16.7 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 41 of 47 Tobacco Control

1 2 3 United States My Start E-Liquid Menthol 18 17 4 United States My Start E-Liquid Menthol 18 17.2 5 United States My Start E-Liquid USA Mix, Full Flavor 0 ND 6 United States My Start E-Liquid USA Mix, Full Flavor 0 ND 7 United States My Start E-Liquid USA Mix, Full Flavor 0 ND 8 Confidential: For Review Only United States My Start E-Liquid USA Mix, Full Flavor 6 6.1 9 United States My Start E-Liquid USA Mix, Full Flavor 6 6 10 United States My Start E-Liquid USA Mix, Full Flavor 6 6.3 11 12 United States My Start E-Liquid USA Mix, Full Flavor 11 11.3 13 United States My Start E-Liquid USA Mix, Full Flavor 11 11.2 14 United States My Start E-Liquid USA Mix, Full Flavor 11 11.2 15 United States My Start E-Liquid USA Mix, Full Flavor 18 17.9 16 United States My Start E-Liquid USA Mix, Full Flavor 18 18.1 17 United States My Start E-Liquid USA Mix, Full Flavor 18 18.1 18 United States My Start E-Liquid Cherry Crush 0 ND 19 United States My Start E-Liquid Cherry Crush 0 ND 20 United States My Start E-Liquid Cherry Crush 0 ND 21 United States My Start E-Liquid Cherry Crush 6 5.8 22 United States My Start E-Liquid Cherry Crush 6 5.2 23 United States My Start E-Liquid Cherry Crush 6 5.4 24 United States My Start E-Liquid Cherry Crush 11 9.8 25 United States My Start E-Liquid Cherry Crush 11 10 26 United States My Start E-Liquid Cherry Crush 11 10.2 27 United States My Start E-Liquid Cherry Crush 18 16.2 28 United States My Start E-Liquid Cherry Crush 18 15.8 29 United States My Start E-Liquid Cherry Crush 18 15.9 30 Australia Bake It Vapor Nana Pie' (Banana) 0 ND 31 Australia Frooty E-Liquid Fresh Mango 0 ND 32 Australia Liqua Premium Vape Liquid Apple 0 ND 33 Australia Liquia Premium Vape Liquid Two Mints 0 ND 34 Australia Liquia Premium Vape Liquid American Blend 0 ND 35 Australia Royale E-Juice Apple None ND 36 37 Australia Royale E-Juice Menthol None ND 38 Australia Royale E-Juice B&H Flavor None ND 39 Australia str8 Premium ejuice Menthol 0 ND 40 Australia Byron Bay Cloud Co. Peaches & Pineapple 0 ND 41 Australia MT Baker Vapor Extreme Ice None ND 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 42 of 47

1 2 3 Australia Clouds Down Under 0 ND 4 Australia Cool Culture Royale 0 ND 5 Australia Mount Baker Vapor East Coast Tobacco None ND 6 Australia Mount Baker Vapor Thug Juice None ND 7 Canada Kloud-E-Juice Smooth Smoke 0 ND 8 Confidential: For Review Only Canada PurELiquid Cooling Menthol 0 ND 9 Canada Mad Dog Juice Cherry Garcia 0 ND 10 Canada Naked 100 Brain Freeze Menthol 0 ND 11 12 Canada Naked 100 Tobacco 0 ND 13 Canada Flavour Crafters Black Cherry Tobacco 0 ND 14 Canada Prizm Menthol 0 ND 15 Canada Blackwood Duke Tobacco 0 ND 16 Canada Darts Canadian Tobacco 0 ND 17 Canada Darts Canadian Menthol 0 ND 18 England Blu Tobacco 18 18.8 19 England Blu Tobacco 18 18.1 20 England Blu Tobacco 18 18.5 21 England E-Lites Regular/TruTaste 24 14.9 22 England E-Lites Regular/TruTaste 24 15.5 23 England E-Lites Regular/TruTaste 24 14.7 24 England Ten Motives 16 13.4 25 England Ten Motives 16 8.3 26 England Ten Motives 16 13.6 27 England Vype 12 12.5 28 England Vype 6 7.6 29 England Vype 6 6.2 30 England Vype 6 6 31 England Vype 18 14 32 England Vype 18 13.9 33 England Vype 18 13.1 34 England Vype None 23.8 35 England Vype 12 12.8 36 37 England Vype 12 12.5 38 England Vype None 21.9 39 England Vype None 25.6 40 England Vype None 31.1 41 England Vype None 21 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 43 of 47 Tobacco Control

1 2 3 England Vype None 24.3 4 England Vype None 29.6 5 England Liberty Flights Tobacco 0 ND 6 England Liberty Flights Tobacco 6 5.3 7 England Liberty Flights Tobacco 12 7.2 8 Confidential: For Review Only England Liberty Flights Tobacco 18 11.1 9 England Liberty Flights Cherry 0 ND 10 England Liberty Flights Cherry 6 3.3 11 12 England Liberty Flights Cherry 12 7.3 13 England Liberty Flights Cherry 18 11.7 14 England Liberty Flights Menthol 0 ND 15 England Liberty Flights Menthol 0 ND 16 England Liberty Flights Menthol 6 3.8 17 England Liberty Flights Menthol 12 10.2 18 England Liberty Flights Menthol 18 11.8 19 England Pure E-Liquid Tobacco (Mab) 0 ND 20 England Pure E-Liquid Tobacco (Mab) 5 3.4 21 England Pure E-Liquid Tobacco (Mab) 10 6.9 22 England Pure E-Liquid Tobacco (Mab) 15 10.8 23 England Pure E-Liquid Cherry 0 ND 24 England Pure E-Liquid Cherry 0 ND 25 England Pure E-Liquid Cherry 5 3 26 England Pure E-Liquid Cherry 10 6.2 27 England Pure E-Liquid Cherry 15 14 28 England Pure E-Liquid Menthol 0 ND 29 England Pure E-Liquid Menthol 5 3.8 30 England Pure E-Liquid Menthol 10 10.6 31 England Pure E-Liquid Menthol 10 6.9 32 England Pure E-Liquid Menthol 15 11.7 33 England Nicolites Tobacco 16 12.3 34 England Nicolites Tobacco 16 8.7 35 England Nicolites Tobacco 16 1.6 36 37 England Nicolites Tobacco 16 8.6 38 England Nicocig Tobacco 16 8.5 39 England Vapourdays Cherry 18 11.2 40 England Vapourdays Menthol 18 118 41 England Vapourdays Menthol 0 ND 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 44 of 47

1 2 3 England Vapourdays Tobacco Blend 18 18 4 England Vapourdays Tobacco Blend 6 6.1 5 England Vapourdays Tobacco Blend 12 12 6 England Vapourdays Tobacco Blend 0 ND 7 England Vapourdays Cherry 12 12.8 8 Confidential: For Review Only England Vapourdays Cherry 0 ND 9 England Vapourdays Cherry 6 5.9 10 England Vapourdays Menthol 6 6 11 12 England Vapourdays Menthol 12 11.2 13 England Ten Motives Menthol 11 9.1 14 England Ten Motives Cherry 16 15.2 15 England Ten Motives Regular 11 9.5 16 England Ten Motives Regular 18 14.7 17 England Blu Tobacco 18 18.8 18 England Ten Motives Menthol 16 13.5 19 England Ten Motives Cherry 16 14.3 20 England Ten Motives Regular 16 13.3 21 England Ten Motives Menthol 16 13.6 22 England Vype Dark Cherry 18 18.8 23 England Vype Crisp Mint 18 19.3 24 England Vype Crisp Mint 18 18.8 25 England Vype Golden Tobacco 6 5.6 26 England Ten Motives Regular 18 14.4 27 England Ten Motives Regular 18 14.1 28 England Ten Motives Menthol 16 12 29 England Vype Dark Cherry 18 16.9 30 England Blu Tobacco 8 8.1 31 England Blu Tobacco 16 16.7 32 England Blu Tobacco 17 16.4 33 England Ten Motives Regular 18 14.9 34 England Blu Cherry 0 ND 35 England Blu Cherry 8 8.1 36 37 England Blu Menthol 8 8.1 38 England Blu Menthol 16 16.7 39 England Blu Menthol 18 17.2 40 England Blu Cherry 18 17.1 41 England Blu Tobacco 18 17.1 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 45 of 47 Tobacco Control

1 2 3 England Blu Menthol 18 16.8 4 England Blu Menthol 9 8.3 5 England Blu Cherry 9 8.8 6 England Blu Cherry 18 18 7 England Blu Tobacco 9 8.8 8 Confidential: For Review Only England Vype Crisp Mint 18 19.7 9 England Blu Tobacco 18 17.8 10 England Blu Menthol 18 18.3 11 12 England Blu Tobacco 9 8.3 13 England Blu Menthol 9 8.2 14 England Logic Light 12 8.2 15 England E-Lites Light 16 12.2 16 England Logic Lights 12 12.8 17 England E-Lites Regular 24 18.2 18 England Ten Motives Regular 16 12.7 19 England E-Lites Menthol 16 11.4 20 England Elites Menthol 16 12.1 21 22 England Elites Menthol 16 11.8 23 England Logic Tobacco 6 4.7 24 England Logic Cherry 6 4.6 25 England Logic Cherry 12 9.3 26 England Logic Tobacco 18 13.9 27 England Logic Cherry 6 4.3 28 England Logic Tobacco 6 4.7 29 England E-Lites Regular 24 18.6 30 England Vype Crisp Mint 36 31.2 31 England Logic Menthol 12 9.1 32 England Logic Tobacco 18 13.4 33 England Logic Cherry 12 8.9 34 England Logic Menthol 12 9.1 35 England Vype Dark Cherry 18 17.9 36 England Vype Dark Cherry 18 17.5 37 England Vype Dark Cherry 12 28.4 38 England Vype Golden Tobacco 18 18.36 39 40 England Vype Dark Cherry 6 5.9 41 England Vype Crisp Mint 6 5.9 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 46 of 47

1 2 3 England Vype Dark Cherry 6 6.1 4 England Vype Crisp Mint 6 5.7 5 England Vype Crisp Mint 0 5.5 6 England Vype Crisp Mint 6 12 7 England Vype Crisp Mint 12 7.3 8 Confidential: For Review Only England Vype Golden Tobacco 6 11.8 9 England Vype Golden Tobacco 12 5.9 10 England Vype Dark Cherry 6 11.3 11 12 England Vype Dark Cherry 12 11.8 13 England Vype Golden Tobacco 0 11.3 14 England Vype Golden Tobacco 12 11.6 15 England Vype Crisp Mint 12 11.9 16 England Vype Crisp Mint 0 13.3 17 England Vype Crisp Mint 12 17.4 18 England Vype Dark Cherry 0 5.7 19 England Ten Motives Regular 16 17.5 20 England Logic Tobacco 18 13.3 21 England Vype Dark Cherry 18 17.4 22 England Vype Golden Tobacco 6 5.7 23 England Vype Golden Tobacco 0 ND 24 England Vype Golden Tobacco 18 17.5 25 England Vype Dark Cherry 18 17 26 England Vype Dark Cherry 6 17.6 27 England Vype Dark Cherry 12 18.2 28 England Vype Golden Tobacco 0 18.3 29 England Vype Crisp Mint 18 17.4 30 England Vype 18 19 31 England Vype 18 19 32 England Vype 18 19 33 England Vype 18 19.3 34 England Vype 18 18.9 35 England Vype 18 19.1 36 37 England Logic Tobacco 12 9.3 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 47 of 47 Tobacco Control

1 2 3 Five most commonly identified flavoring chemicals by country 4 5 6 Potential 7 Country Chemical CAS # Components Occurrence Respiratory 8 Confidential: For Review Only Irritant 9 10 Levomenthol 2216-51-5 cooling, camphoreous, minty, clean, spicy 40.0% X 11 Menthol 1490-04-6 cooling, mentholic, minty 40.0% X Australia 12 Maltol 118-71-8 sweet, cotton candy, caramellic with jammy fruity and berry 40.0% X 13 (n=15) 14 Vanillin 121-33-5 vanilla, vanillin, sweet, creamy, spicy, phenolic, milky 40.0% 15 Ethyl Vanillin 121-32-4 sweet, creamy, vanilla, smooth, caramellic 26.7% X 16 17 Vanillin 121-33-5 vanilla, vanillin, sweet, creamy, spicy, phenolic, milky 42.9% 18 Hydroxyacetone 116-09-6 sweet, slightly green, burnt 42.9% Canada 19 Maple Lactone 765-70-8 sweet, maple, bready, caramellic with nutty nuances 42.9% 20 (n=10) 21 2-methyl-Butanoic acid 116-53-0 fruity, dirty, acidic with a dairy buttery and cheesey nuance 35.7% 22 Ethyl Acetate 141-78-6 etherial, fruity, sweet with a grape and cherry nuance 28.6% 23 Propyl Methyl Ketone 107-87-9 sweet, fruity and banana-like with a fermented nuance 98.4% 24 25 Isopropyl Methyl Ketone 563-80-4 Camphor 99.0% England 26 Acetone Alcohol 116-09-6 sweet, slightly green, burnt 88.0% 27 (n=166) Menthol 1490-04-6 cooling, mentholic, minty 65.1% X 28 29 Ethyl Acetate 141-78-6 etherial, fruity, sweet with a grape and cherry nuance 80.7% 30 Vanillin 121-33-5 vanilla, vanillin, sweet, creamy, spicy, phenolic, milky 55.6% 31 Maple Lactone 765-70-8 sweet, maple, bready, caramellic with nutty nuances 35.2% 32 US 33 Levomenthol 2216-51-5 cooling, camphoreous, minty, clean, spicy 31.5% X (n=54) 34 Potato Pyrazine 14667-55-1 raw, musty, nutty, potato 29.6% X 35 36 Benzyl Alcohol 100-51-6 chemical, fruity, cherry, almond, balsamic, bitter 21.1% X 37 38 39 40 41 42 43 44 45 https://mc.manuscriptcentral.com/tobaccocontrol 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Tobacco Control Page 48 of 47