Part Three: Biomarker Changes in Cigarette Smokers Switched to Vuse Solo or Abstinence: A Randomized, Controlled Study

Biomarkers of exposure (BoE) can help evaluate exposure to combustion-related, tobacco-specic toxicants after smokers switch from cigarettes to potentially less-harmful products like electronic nicotine delivery systems (ENDS). This paper reports data for one (Vuse Solo Original) of three products evaluated in a randomized, controlled connement study of BoE in smokers switched to ENDS. Subjects smoked their usual brand cigarette ad libitum for two days, then were randomized to one of three ENDS for a 7-day ad libitum use period, or to smoking abstinence. Thirteen BoE were assessed at baseline and Day 5, and percent change in mean values for each BoE was calculated. Biomarkers of potential harm (BoPH) linked to oxidative stress, platelet activation, and inammation were also assessed. Levels decreased among subjects randomized to Vuse Solo versus Abstinence, respectively, for the following BoE: 42–96% vs. 52–97% (non-nicotine constituents); 51% vs. 55% (blood carboxyhemoglobin); and 29% vs. 96% (nicotine exposure). Signicant decreases were observed in three BoPH: leukotriene E4, 11-dehydro-thromboxane B2, and 2,3-dinor thromboxane B2 on Day 7 in the Vuse Solo and Abstinence groups. These ndings show that ENDS use results in substantially reduced exposure to toxicants compared to smoking, which may lead to reduced biological effects. the effects of from to of of ENDS products Abstinence, Day of a in-clinic switch. For the purposes this data resulting from those subjects randomized to the Vuse Solo and Abstinence groups will be reported. Data for the other Vuse ENDS investigational product (IP) groups will be a controlled study was designed to assess BoE in smokers switched to one of three closed Vuse ENDS IP groups or Abstinence during an in-clinic connement. Only the data for subjects randomized to either the Vuse Solo or Abstinence groups will be reported here. No between-group comparisons were performed.


Introduction
It is well known that cigarette smoking is a leading cause of preventable premature death and disease, including cancer, cardiovascular disease, and chronic obstructive pulmonary disease (COPD) 1,2 . In 2012, the US Food and Drug Administration (FDA) established a list of harmful and potentially harmful constituents (HPHCs) in tobacco products.
Embedded in this list, FDA categorized each HPHC as a carcinogen, respiratory toxicant, cardiovascular toxicant, reproductive or developmental toxicant, and/or addictive 3 . The type of tobacco product used, and the frequency, duration, and manner of use are associated with the level of exposure to HPHCs as well as disease risk 4 . Biomarker studies have previously con rmed that individuals who smoke fewer cigarettes and those who dramatically reduce their daily cigarette consumption (>50%) have lower toxicant exposure than continuing heavy smokers [5][6][7][8] . Furthermore, this relationship is consistent with analyses of epidemiologic data that show an increasing curvilinear relationship between cigarettes per day and the risk of certain cancers [9][10][11][12] .
Electronic nicotine delivery systems (ENDS) fall between cigarettes and nicotine replacement therapies on the continuum of risk that has been used to describe the relative risk associated with the use of different tobacco-and nicotine-containing products 4,13 . The aerosol generated during the use of ENDS has been characterized as containing lower levels of known toxicants compared to those found in cigarette smoke [14][15][16] , and signi cantly fewer toxic constituents than traditional cigarettes [17][18][19][20][21][22] . Biomarkers of exposure (BoE) are used as measures of exposure to tobacco and tobacco smoke constituents in humans. Prior research has shown signi cantly lower levels of urinary carcinogen and toxicant metabolites in ENDS users compared to cigarette smokers 20 , as well as signi cant declines in biomarkers of tobacco exposure among cigarette smokers switched exclusively to ENDS 19,23,24 .
Validated functional predictive biomarkers, also known as surrogate biomarkers of complex smoking-related diseases such as lung cancer, COPD and cardiovascular disease, do not exist 1,25 . Alternatively, tobacco-related biomarkers of potential harm (BoPH) provide an assessment of biological effects resulting from tobacco product exposure that can lead to the development of disease. Such biomarkers can serve as intermediate indicators of disease risk from tobacco product exposure and can be measured in the context of short-term switching studies.
Chronic smoking is associated with elevated oxidative stress and in ammation, which are key drivers of smoking-induced pathophysiology 26 . Smoking-induced oxidative stress and in ammatory response are indicated by an increased synthesis of arachidonic acid (AA) and its metabolites [27][28][29] . A panel of AA-derived metabolites, including prostaglandins, isoprostanes, thromboxanes, and leukotrienes, has been demonstrated as a useful set of short-term BoPH in smokers who switch from smoking to abstinence or tobacco products lower on the risk continuum for as few as ve days 30 .
This study assessed the effects of switching non-menthol smokers from combustible cigarettes (CC) to use of one of three Vuse ENDS products (Vuse Solo Generation 2 [Gen 2], Vuse Vibe or Vuse Ciro) or Abstinence, at Day 5 of a seven-day inclinic switch. For the purposes of this manuscript, data resulting from those subjects randomized to the Vuse Solo and Abstinence groups will be reported. Data for the other Vuse ENDS investigational product (IP) groups will be reported in a later publication.
The primary endpoints in this study were validated BoE to constituents recommended for evaluation in ENDS e-liquids and aerosols as listed in the draft guidance for Premarket Tobacco Product Applications for ENDS published by the US FDA Center for Tobacco Products in May 2016 31 and/or compounds identi ed by the FDA as HPHCs 3 . All of the tobacco-related BoE quanti ed in this study have been reviewed previously 32 . The secondary objectives of this study were to evaluate daily product use amounts and to assess urinary AA-derived metabolites as BoPH at Day 5 and Day 7 following an in-clinic switch from UB cigarettes to use of Vuse Solo or Abstinence. The Abstinence group was included in the study design to assess the maximum biomarker change possible when subjects experience the same environment and conditions as those who switch to an ENDS product but stop use of tobacco products completely.
Several of the AA metabolites are mediators of important physiological processes such as platelet activation and airway in ammation, and are known to be increased in chronic cigarette smokers, potentially contributing to increased cardiovascular risk and airway obstruction, respectively. Previous studies have reported rapid changes in a subset of AA metabolites upon smoking abstinence 33,34 . Results from this study build upon the existing body of research which suggests that exclusive use of ENDS products poses less individual health risk to tobacco product consumers than CC smoking. This paper is the third in a three-part series reporting on three recent studies conducted to evaluate the abuse liability 35 , nicotine pharmacokinetics 36 , and biomarkers of Vuse Solo Gen 2 (current). Adverse event (AE) information for all three studies is collectively reported in this paper.

Results
This randomized, controlled study was designed to assess BoE in smokers switched to one of three closed Vuse ENDS IP groups or Abstinence during an in-clinic con nement. Only the data for subjects randomized to either the Vuse Solo or Abstinence groups will be reported here. No between-group comparisons were performed.

Study Population
Thirty-ve subjects were randomized to the Vuse Solo group and 16 to the Abstinence group. Among these, 46 (90%) completed the study (Vuse Solo group n=35, Abstinence group n=11). Four subjects in the Abstinence group voluntarily withdrew from the study, and one was discontinued due to an AE. Subject demographics and baseline characteristics are summarized in Supplementary Table S1. The study population was similar between the two groups and was generally balanced between males (57.1%, 56.3%) and females (

Biomarkers of Exposure in Urine and Blood
Thirteen BoE (12 urinary and one blood) were evaluated as primary endpoints in this study. Levels of all evaluated BoE in both study groups statistically signi cantly (p < 0.05) decreased ve days after switching from UB cigarettes to use of Vuse Solo or Abstinence (Figure 1; Supplementary Table S2). Aromatic amines decreased 58-95% and 66-94% in the Vuse Solo and Abstinence groups, respectively. For biomarkers of semi-volatile organics, reductions ranged from 79-96% in the Vuse Solo group and 81-95% in the Abstinence group. One polycyclic aromatic hydrocarbon was measured in this study, 3-OHbenzo[a]pyrene (3-OH-B[a]P). Although statistically signi cant (p = 0.0420, step-down Bonferroni-adjusted), only a 42% reduction in the biomarker of benzo[a]pyrene (B[a]P) was observed in the Vuse Solo group due to ~50% of subjects having baseline values below the lower limit of quanti cation (100 fg/mL; Supplementary Table S4). A statistically signi cant (p = 0.0015) reduction of 72% in the biomarker of benzo[a]pyrene (B[a]P) was observed in the Abstinence group. Statistically signi cant (p < 0.05) reductions were observed for tobacco-speci c nitrosamines, with values ranging from 59-83% and 52-97% in the Vuse Solo and Abstinence groups, respectively. Additionally, a statistically signi cant (p <0.0001) reduction of 83% was observed for total NNN in the Vuse Solo group versus a non-signi cant (p = 0.1131) reduction of 97% in the Abstinence group. When one extreme value was excluded from the Abstinence group Day -1 dataset, the reduction of total NNN was 87% and statistically signi cant (p = 0.0061) Similar to reductions in urinary BoE, blood carboxyhemoglobin (COHb) levels were statistically signi cantly (p < 0.05) reduced 51% and 55% post-product switch (Day 5) for the Vuse Solo and Abstinence groups, respectively (

Biomarkers of Potential Harm
Switching from smoking to either Vuse Solo or Abstinence resulted in decreases in AA-derived metabolites (Table 1).

Daily E-Liquid Consumption
The mean amounts of Vuse Solo e-liquid consumed each day of the seven-day product switch period are presented in Table   2. These data show that Vuse Solo e-liquid consumption remained relatively constant on Days 1 and 2, increased over Days 3 through 5, with a slight decrease from maximum consumption on Days 6 and 7 (Table 2). blister, and oropharyngeal pain (one subject each). The most common AEs reported in the Abstinence group were irritability (12.5%) and back pain (12.5%). These AEs occurred after randomization; therefore, none of the AEs were related to product use. No subject in the Vuse Solo group discontinued the study as a result of an AE. However, one subject in the Abstinence group was discontinued due to an AE, which the PI determined was related to nicotine withdrawal symptoms.
Across this study and the other two reported as part of this series, 53 total AEs were reported by 68 subjects using Vuse Solo 35,36 . Of these, the study PI determined 11 AEs were possibly related and 9 AEs were related to Vuse Solo use (Supplementary Table S3). All AEs causally related to Vuse Solo use were mild, and included headache, cough, back pain, oropharyngeal pain, musculoskeletal pain, nausea, dry mouth, dry throat, and gastroesophageal re ux (Supplementary Table   S3). for about 50% of subjects in the Vuse Solo group were below the lower limit of quanti cation. Therefore, the percent changes observed were mostly dependent on the respective 24-hour urine volume collected. Comparable results of low baseline levels of 3-OH-B[a]P in smokers switched short-term to an ENDS product were also reported in a previous study by Round et al. (2019). For total NNN, when the subject with the extreme data value in the Abstinence group (n=1 of 11) was excluded from the analysis, the 97% reduction in total NNN changed from non-signi cant (p = 0.1131) to a statistically signi cant reduction of 87% (p = 0.0061).

Discussion
In general, the reductions in total NNAL were lower (59% and 52% in Vuse Solo and Abstinence groups, respectively) compared to other urinary biomarkers assessed, and the mean percent reductions were similar between both groups. The The mean percent reduction in blood COHb observed in this study's Vuse Solo group was similar to the mean percent reduction observed in the Abstinence group (51% versus 55%, respectively; Figure 1). Jay et al. 24 reported 69% -74% COHb reduction in the ENDS groups compared to 69% reduction in the Abstinence group. The results in the current study are also comparable to the earlier reported ndings by Round and colleagues 23 .
Reductions in NicEq-T in the Vuse Solo group, although statistically signi cant (p < 0.05), were smaller than the reductions observed for other urinary biomarkers in this group. NicEq-T decreased signi cantly by 29% in the Vuse Solo group versus the statistically signi cant 96% reduction in the Abstinence group. These ndings demonstrate the expected biomarker reductions with use of a nicotine-containing product, which indicates that the subjects in the Vuse Solo group remained exposed to nicotine levels closer to baseline UB smoking relative to those in the Abstinence group. These results support the concept that smokers who switch to Vuse Solo will use the product in such a way to provide them with levels of nicotine near those of smoking, while reducing their exposure to tobacco-related toxicants found in cigarette smoke. Further evidence to demonstrate that nicotine uptake occurs with Vuse Solo use, but not to the same extent as when smoking a CC, are presented in Campbell et al. 35 . In their study, Campbell et al. reported on the abuse liability of Vuse Solo, which showed that mean peak nicotine uptake from 10-minute ad libitum use of Vuse Solo was 61% less (p < 0.05) than nicotine uptake from smoking one UB cigarette.
This study also evaluated a panel of AA-metabolites, indicators of oxidative stress, in ammation, and platelet activation, as BoPH. A key nding was that switching to exclusive use of Vuse Solo or Abstinence resulted in reductions of three of these metabolites: LTE4, 11-dh-TXB2, and 2,3-d-TXB2. These ndings align with studies which demonstrate that urinary LTE4 levels return to baseline within two weeks of tobacco product abstention 39 . This similar trend toward reduction of urinary LTE4 levels was observed in smokers switched to either Vuse Solo or smoking abstinence.
Thromboxane A2 (TXA2) is a short-lived eicosanoid derived from cyclooxygenase-1 (COX-1)-dependent metabolism of arachidonic acid. TXA2 is very rapidly metabolized to Thromboxane B2 (TXB2) and further metabolized to 11-dehydro TXB2 and 2, 3-dinor TXB2. It is reported that increased levels of LTE4 are linked to airway in ammation; whereas increased levels of 11-dh-TXB2 and 2,3-d-TXB2, are re ective of platelet activation that leads to atherosclerosis and cardiovascular disease 40,41 42,43 . Thus, the decline in LTE4, 11-dh-TXB2, and 2,3-d-TXB2 levels in smokers who switch to noncombustible tobacco products suggests reduction in airway in ammation and platelet activation, which are the two key drivers of smokinginduced pathophysiology. These results of the short-term BoPH assessments further support previous ndings in which reductions in LTE4, and 2,3-d-TXB2 were observed for smokers who switched to Vuse Solo products or abstinence 30 .
While the data presented herein re ect the effects in cigarette smokers who completely switch to Vuse Solo use, there are limitations to the study design. First, to ensure that subjects used only the ENDS product or abstained from smoking, the study was conducted in a con ned setting, which reduces the variability of environmental toxicant exposure due to the controlled study environment. However, a con ned setting intentionally limits options for product use. While these data are indicative of exposure changes if cigarette smokers completely switch to Vuse Solo, in a "real-life" environment, where cigarette smokers are able to freely choose the tobacco products they use, other patterns of product use are possible. Dual use and poly use of tobacco products, which may occur in an ambulatory setting, could result in different toxicant exposures.
Secondly, this study did not include a group of subjects who continued to smoke, which would have allowed an assessment of the difference between BoE levels from continued smoking and switching to Vuse Solo or abstinence under the same study conditions. Some have reported that urinary BoE signi cantly (p < 0.001) increased in the continued smoking groups in a con nement setting 24,44 . Other studies that investigated BoE in a randomized controlled design have reported increased exposure to toxicants in the continued smoking group 8, 45,46 . Given these results, a similar continued smoking group in the current study may have impacted study results by possibly increasing or decreasing the differences from baseline. As such, the biomarker results presented here may represent the worst-case scenario for biomarker reduction given the results from prior research 24,44 .
In this series of studies, the results demonstrate that smokers who switch to Vuse Solo are exposed to much lower levels of smoke toxicants, that their nicotine exposure from a single use and from daily ad libitum use is less than smoking cigarettes, and that their nicotine exposure is similar across different avors of Vuse Solo. as well as dual users of CC and ENDS, achieved peak nicotine uptake values that were consistently lower than those for CC and higher compared to nicotine gum with nicotine uptake similar across the assessed avors. In addition, subjective measures of product liking, intent to use again, and urge to smoke were all lower after the use of Vuse Solo compared to CC but higher compared to nicotine gum.
The results from this study demonstrate that short-term switching from smoking to use of Vuse Solo leads to decreased exposure to tobacco smoke constituents. The reductions in toxicant exposure observed among subjects switched to Vuse Solo were comparable in magnitude to the reductions observed in subjects who completely abstained from smoking, suggesting that the reduction in exposure for the Vuse Solo group was similar to that of the Abstinence group for the biomarkers assessed.
The ndings presented in the present study, along with the results reported in the abuse liability study 35 35 provide support that Vuse Solo is preferred more than nicotine gum but less than CC, the results of the current study demonstrate that use of Vuse Solo has a favorable impact on reducing toxicant exposure levels relative to smoking, and this impact is comparable to abstinence.
With the well-known risks of cigarette smoking, ENDS such as Vuse Solo can serve a bene cial purpose for public health if both the abuse potential and exposure to toxicants are reduced compared to CC. Taken together, the results presented in this series of studies demonstrate that Vuse Solo has a nicotine pharmacokinetic pro le and an overall product liking pro le that may provide an acceptable alternative to CC for smokers. Additionally, switching to Vuse Solo from CC reduces exposure to smoking-related toxicants. Collectively, these results provide further support for the role of closed system ENDS products in general, and Vuse Solo speci cally, in tobacco harm reduction.

Study Design
This randomized, controlled, switching, open-label, parallel cohort con nement study (ClinicalTrials.gov registration number; Generally healthy males and females ages 21-60 years who self-reported smoking ≥ 10 combustible, ltered, non-menthol cigarettes per day for at least 6 months prior to screening, con rmed by an expired carbon monoxide (ECO) level of ≥ 12 parts per million and a positive urine cotinine test (≥ 200 ng/ml), were included in the study. The subjects were required to be willing to switch from using their UB cigarettes to either use of one of three Vuse ENDS IPs (Vuse Solo, Vuse Ciro, Vuse Vibe) or to remain abstinent from smoking through the end of the study. Other eligibility criteria were also assessed by the PI; however, women who were pregnant, breastfeeding, or intended to become pregnant during the study were excluded from participation.
Potential subjects completed a pre-screening telephone interview and also attended a screening visit at the clinical research site within 30 days before study entry. Eligible subjects were enrolled into the study on Day -2 and stayed at the clinical site for a 10-day con nement period. The subjects were allowed to smoke their UB cigarettes ad libitum, between the hours of 0700 and 2300 for the rst two days (Days -2 and -1), during which baseline measurements were taken. On Day -2, the level of dependence on cigarettes was assessed with the Fagerström Test for Nicotine Dependence, on which a score of 0-2 indicates low dependence and a score of 8-10 indicates high dependence 47 .
On Day 1, subjects were randomized to either a Vuse Solo ENDS IP group or the Abstinence group. A computer software program was used by the statistician at the CRO to generate the randomization sequences. A total of 35 and 15 subjects were randomized in the Vuse Solo and the Abstinence groups, respectively. There were 10 blocks total for Vuse Solo (7 blocks, sequence code B) and for the Abstinence (3 blocks, sequence code D) groups. The study was unblinded by necessity due to the very different visual appearances of the study products. Upon randomization, the subjects in the Vuse ENDS groups were allowed to use their assigned ENDS IP ad libitum between the hours of 0700 and 2300. The randomized subjects were supervised and monitored by the clinical site staff to ensure study product usage and protocol compliance.
Sampling for urinary BoE was performed two times by collecting 24-hour urine samples beginning on the mornings of Day -1 and Day 5 through the mornings of Days 1 and 6, respectively. Blood samples for measurement of COHb concentrations in whole blood were collected once a day (at approximately 12 hours post start of product use) on Days -1, 1, 3, and 5.
Daily ENDS IP usage was controlled by the clinical staff and restricted after 2300 hours on Days 1 through 7. The amount of IP consumed per subject per day was measured as the total mass of e-liquid used per day. This was done by weighing the cartridges before and after use and allowing the subjects to use the same cartridge until entirely depleted.
Additional urinary samples were collected on study Days -1, 5, and 7 for the evaluation of BoPH. To determine whether AA metabolism was altered following short-term switching, a panel of nine urinary AA-derived metabolites were measured in both the Vuse Solo (n=35) and Abstinence (n=11) groups after ve and seven days of switching from UB smoking.
Safety evaluations were assessed by monitoring AEs, physical examinations (including oral examinations), electrocardiograms, clinical laboratory tests (chemistry, hematology, virology, and urinalysis), pregnancy screening, and vital sign measurements. An AE was de ned as any untoward medical occurrence or condition experienced by a subject from randomization (on Day 1) until completion of the study, whether or not considered related to the use of IP by the PI. Any AEs with a start time prior to randomization was recorded as part of the subject's medical history. The severity of AEs was categorized as mild (of little concern to the subject and/or of no clinical signi cance), moderate (discomfort enough to cause interference with or change in usual activities), or severe (incapacitating or unable to work or subject in many or all usual activities; is of concern to the subject and/or poses substantial risk to the subject's health or well-being). In all cases, the PI made a determination of the relationship of the AE to the IP using a four-category system (not related, unlikely related, possibly related, or related).
The data presented in this manuscript were collected from baseline (Day -1) through the morning of Day 8. The subjects were discharged from the study on Day 8 after completing all the End of Study Assessments.

Study Product
The three Vuse ENDS products used in the study for which the methods are described above were commercially available in the USA at the time the study was conducted. The primary endpoints were assessed with 13 statistical comparison tests (including 12 urinary and one blood biomarker comparisons). For urinary BoE, raw concentrations (mass/mL) were reported by the bioanalytical laboratory. Total urinary amounts excreted per day (total mass/day) were determined by multiplying mass/mL x total urine output volume in mL over 24 hours. The statistical analysis was performed on total mass/day for all urine BoE. Total nicotine equivalents (NicEq-T) was treated as a secondary endpoint. For the blood BoE (COHb), the bioanalytical laboratory reported percent (%) carbon monoxide. Descriptive statistics (mean and standard deviation [SD]) were calculated for the 12 urinary BoE and blood COHb for baseline during cigarette smoking (Day -1) and post-product switch (Day 5). Changes in these primary BoE from Day -1 to Day 5 were assessed for statistical signi cance using a one-sided paired t-test, except for NicEq-T where statistical signi cance was assessed using a two-sided paired t-test. Percent (%) change from Baseline (Day -1) to post-product switch (Day 5) was also calculated for both the urinary and blood BoE.
To account for multiplicity and preserve the overall type I error rate of 0.05, each test on the primary endpoints was adjusted using the adaptive step-down Bonferroni method implemented in SAS version 9.4. No multiplicity adjustments were made for the secondary endpoint. Biomarker data were screened for potential extreme data values. Extreme data values were de ned as data that were beyond the mean ± 6σ for each endpoint according to internal SOP. If any extreme data values were observed, analyses were repeated with and without the extreme data values. No comparisons between the Vuse Solo and the Abstinence groups were performed.
The BoPH levels are presented as total mass excreted in 24 hours (ng/24h). Descriptive statistics were calculated for baseline (Day -1) and for Day 5 and Day 7 post-product switch to either Vuse Solo or Abstinence. To compare the differences between baseline and post-switching in each group, a paired t-test was conducted to determine statistical signi cance (p<0.05). Percent (%) change from Baseline (Day -1) to post-product switch (Days 5 and 7) was also calculated for BoPH. No adjustments were made for multiple comparisons. JMP 10 (SAS Institute, Cary, NC, USA) was used for data analysis.
Descriptive statistics of demographics, baseline characteristics, and Vuse Solo use (e-liquid consumption) are summarized by group, and time point where appropriate. All enrolled subjects were included in these summarizations, as well as the data reported for AEs.

Declarations
Data availability The applicable data generated or analyzed during this study are included in this manuscript (and its supplementary Tables).