Smoking is one of the main preventable causes of respiratory and other chronical diseases, including COPD and lung cancers (1). The best cause of action for smokers is to quit. However, despite significant public health efforts and the availability of pharmaceutical nicotine replacement therapy products (NRTs), many smokers remain unwilling or unable to give up smoking. As a harm reduction intervention, validated low-risk tobacco products have been made available to current smokers. In the U.S., under the auspice of the FDA’s comprehensive tobacco regulation program, modified risk tobacco products (MRTPs) are an authorized category of alternatives that can be marketed to adult smokers (2).
Heated tobacco products (HTPs) are among the classes of novel tobacco products that are available in some markets. They are designed to heat tobacco through a battery-powered heating system to generate nicotine-containing aerosol for inhalation (3, 4). It has been shown that most of the toxicants found in cigarette smoke are formed as a result of incomplete tobacco combustion (5). In a typical HTP device, its heating temperature program and the aerosol formation can be accurately controlled, resulting in the elimination of tobacco combustion and therefore the aerosol formed is found to contain significantly less toxicants than cigarette smoke (6, 7). Most commercial HTPs heat the tobacco below 350 °C using a varied but controlled temperature profile. This contrasts to a burning cigarette which typically runs between 600 and 900 °C (5). However, commercial HTPs come with different heating technologies such as e.g., IQOSTM and gloTM (3, 4), where the heating energy is either transmitted from inside a tobacco rod outwardly or inwardly. More innovations are coming onto the market in HTP designs, some using resistive while others using radiative/non-contact means for heat transfer. This may lead to different degrees of tobacco heating at different locations, and therefore some differences in the chemicals formed in the aerosol. Hence it is required by regulators that each HTP should be chemically and biologically assessed under rigorous preclinical and clinical programs before it is considered for market assess and potentially reduced risk claims.
In commercial HTPs, aerosol is formed by heating reconstituted tobacco material containing aerosol-forming agents such as vegetable glycerol and propylene glycol (3). During heating, the volatile compounds and the aerosol-forming agents are vaporized from the tobacco substrate and reach supersaturation; upon cooling by the puffing air flow condense to form the aerosol. In most HTP designs, puffing airflow enters from the open end of a tobacco rod and exits through its mouth end - for comparison purpose, this design is called an open-end HTP. In this work, a novel HTP design that prevents the air flow from entering the tobacco end is described - this is called a closed-end HTP. The key design features and its working principle have been described before (8, 9), more details will be given in the experimental section.
In this work, we aimed to compare the aerosol chemistry, first using a limited number of toxicants measured on this close-ended HTP and a commercial HTP (IQOSTM). Afterwards biological properties of the aerosol produced were assessed using the standardized regulatory
Three products were tested, two commercially available HTPs and one combustible research reference cigarette 3R4F. The two HTPs were commercially available: Mr. Yeah (the closed-end test product, coded as HTP-A) and IQOSTM (the open-end comparator, coded as HTP-B). Each HTP came with its own heating device and dedicated tobacco sticks. The key features of the heating device and tobacco sticks are listed in Table 1, and further details about the products can be found in the literature (9, 10). The tobacco sticks used in HTP-A had ca. 350 mg of reconstituted Chinese flue-cured blended tobacco, 1.81 mg of nicotine, 49.5 mg of glycerol, and 12.2 mg of propylene glycol.
The heating device of HTP-A is branded as Mr. Yeah and the tobacco sticks carry the brand FARSTAR. HTP-B consists of an IQOS™ branded heating device and TEREA branded tobacco sticks.
Parameter | HTP-A | HTP-B | Combustible cigarette |
---|---|---|---|
Commercial product (name) | Mr. Yeah | IQOS™ | 3R4F |
Puffing air flow design | Closed-end | Open-end | Open |
Heater design | External heating | Central heating | Combustion |
Maximum heating temp. (°C) | 260 | 350 | / |
Puff n° under HCI regime | 8 | 12 | 8 |
Tobacco stick-length (mm) | 45 | 45 | 84 |
Tobacco stick-weight (mg) | 721.70 | 664.30 | 1035.00 |
Tobacco stick-diameter (mm) | 7.36 | 7.18 | 7.98 |
Prior to the smoking experiments, the tobacco sticks were removed from their packages and conditioned at 22 °C ± 2 °C and a relative humidity of 60% ± 3% in accordance with ISO 3402 (28). The devices were fully charged and cleaned between each smoking session according to their operating instructions.
The air flow design of HTP-A and HTP-B is schematically shown in Figure 1 (8, 9). In short, the HTP-A airflow pathway (Figure 1A) was from the perforation on the hollow acetate tube because the tobacco end of the heating device was completely sealed. The HTP-B airflow pathway (Figure 1B) was from the tobacco end (8, 9). The working mechanism behind the aerosol formation and transfer for the close-end HTP-A has been described before (8, 9). In brief: The extraction of formed aerosol in HTP-A is facilitated by a negative pressure created by the puffing air through the airflow pathway relative to the tobacco stick.
Schematic diagram comparing air flow path for the closed-end HTP-A (a: the test product) and an open-end HTB-B (b: see Ref. 8, 9 )
The mainstream aerosol/smoke was generated on a Cerulean smoking machine (Cerulean, Milton Keynes, UK) and captured by Cambridge glass-fiber filter pads (Whatman, Maidstone, UK) or cooled impingers with an appropriate solvent for volatile fractions. After the heating devices were preheated according to their respective operating instructions, 3 replicated products were smoked under the Health Canada Intense (HCI) (29) smoking regime, which yielded 8 puffs/stick for HTP-A and 3R4F cigarette, 12 puffs for HTP-B respectively. For screening purposes, we selected 9 toxicants - recommended by the WHO Study Group on Tobacco Product Regulation (TobReg) for mandated lowering - for HTP aerosol and cigarette smoke comparison: carbon monoxide, formaldehyde, acetaldehyde, 1,3-butadiene, benzene, acrolein, B[
Analyte | Brief description | Reference |
---|---|---|
Nicotine | By gas chromatography with flame ionization detection | ISO 10315 (12) |
CO | By nondispersive infrared photometry | ISO 8454 (13) |
Polycyclic aromatic hydrocarbons | Extracted by hexane and analyzed by GC-MS | CRM 91 (14) |
Extracted by acetic acid and analyzed by HPLC-MS/MS | CRM 75 (15) | |
Volatile organic compounds | By GC-MS using simultaneous trapping of adsorbent cartridges in a Cambridge filter without cryogenic impinger | WHO SOP 09 (16) |
Carbonyl compounds | By HPLC, using simultaneous trapping of adsorbent cartridges in a Cambridge filter without cryogenic impinge | WHO SOP 08 (17) |
The Ames assay was performed in accordance with the Health Canada Official Method T-501 (18). Mutagenicity of collected total particulate matter (TPM) fraction was evaluated using two bacterial strains:
Strains | Mutation | Antibiotic resistance | Positive controls | |
---|---|---|---|---|
Without S9 (μg/plate) | With S9 (μg/plate) | |||
TA98 | His D3052 | Ampicillin | 2-Nitrofluorene (4 μg/plate) | 2-Amlinofluorene (10 μg/plate) |
TA100 | His G46 | Ampicillin | 4-Nitroquinoline |
2-Amlinofluorene (10 μg/plate) |
The MN assay was performed in accordance with Health Canada Official Method T-503 (19). The Chinese Hamster Ovarian (CHO) cell line was purchased from BNCC (Beijing, China) and were maintained in Dulbecco’s modified eagle medium (Sigma Aldrich, MO, USA) supplemented with 10% fetal bovine serum (Thermo Fisher, Waltham, MA, USA) in a 5% CO2 incubator at 37 °C ± 2 °C. Cell suspension (1 × 105 cells/mL) was preincubated for 24 h before treatment.
For positive controls, cyclophosphamide A (CPA) (Dalian Meilun Biotechnology Co., Ltd., Dalian, Liaoning Province, China) for ST (short-term)+S9, mitomycin C (MMC) (GlpBio, Montclair, CA, USA) for ST-S9 and mitomycin C for LT (long-term)-S9 were used in the experiments (Table 4). For short-term exposures, the cell culture was treated with serially diluted test samples for 3 h without or with S9-mix containing Aroclor 1254-induced rat liver homogenate (Molecular Toxicology, Inc., Boone, NC, USA). After removal of the test sample, the cells were incubated for 27 h. For LT exposures, the cells were incubated with test sample for 30 h in the absence of the metabolic activation system.
Positive controls | Ultimate density |
---|---|
Cyclophosphamide A (short-term + S9) | 0.2 μg/mL |
Mitomycin C (short-term - S9) | 1.0 μg/mL |
Mitomycin C (long-term - S9) | 0.5 μg/mL |
The NRU assay was performed using CHO cell line in accordance with Health Canada Official Method T-502 (20). The CHO cell suspension (5 × 104 cells/mL) was precultured for 24 h in 96-well microtiter plates. The cells were treated with TPM fractions for 24 h. Sodium lauryl sulfate was used as the positive control. Dimethyl sulfoxide (DMSO) alone was used as solvent control. The treated cells were washed with PBS, and then incubated with a medium containing 50 μg/mL neutral red dye for 3 h. The cells were then fixed with 1% formalin solution for 1–2 min. After removal of the fixative, the neutral red dye taken up by the viable cells was extracted by adding 50% ethanol containing 1%
For aerosol and smoke chemistry, each product was analyzed in triplicates. The findings were reported as mean ± standard deviation (SD). For analytes that were systematically below the level of quantification (LOQ), not quantified (NQ) was assigned. For the biological testing data, SPSS 24.0 software (IBM, New York, NY, USA) was used for statistical analysis. All data were analyzed by ANOVA,
The TPM results and selected toxicant levels are shown in Table 5, on both per-stick and per-mg of nicotine basis. The TPM levels of HTP-A and HTP-B were slightly higher than those of 3R4F cigarettes when compared on a per-stick basis. The CO yields for both HTPs were below LOD whereas those from 3R4F were 24.69 mg/stick - the reduction of CO for the HTPs was a clear confirmation that the tobacco in each system did not experience a significant thermal breakdown or combustion. Formaldehyde, acetaldehyde and acrolein were detected in HTP-A with quantifiable levels, with HTP-A’s formaldehyde yields at 76%, acetaldehyde at 28%, and acrolein at 56% lower than those of HTP-B, respectively. These levels were also 96% (formaldehyde), 83% (acetaldehyde) and 93% (acrolein) lower than those in 3R4F cigarette smoke, respectively. In addition, 1,3-butadiene was below LOQ for both HTPs whereas in 3R4F it was at 8.13 μg/stick. The yields of benzene and benzo(
Analyte | Unit | HTP-A | HTP-B | 3R4F |
---|---|---|---|---|
Nicotine | mg/stick | 0.47 ± 0.01 | 0.98 | 1.05 |
Total particulate matter | mg/stick | 29.06 ± 1.36 | 30.15 | 24.23 |
mg/mg nicotine | 61.83 ± 2.90 | 30.77 | 23.08 | |
Carbon monoxide | mg/stick | 0.19 ± 0.01 | 0.25 | 24.69 |
mg/mg nicotine | 0.40 ± 0.02 | 0.26 | 23.52 | |
Formaldehyde | μg/stick | 1.08 ± 0.18 | 4.46 ± 0.56 | 26.60 ± 2.14 |
μg/mg nicotine | 2.30 ± 0.38 | 4.55 ± 0.57 | 25.33 ± 2.04 | |
Acetaldehyde | μg/stick | 123.5 ± 19.53 | 172.42 ± 6.10 | 722.47 ± 36.11 |
μg/mg nicotine | 262.77 ± 41.55 | 175.94 ± 6.22 | 688.07 ± 34.39 | |
Acrolein | μg/stick | 4.78 ± 0.91 | 10.90 ± 0.18 | 71.92 ± 2.84 |
μg/mg nicotine | 1.02 ± 1.94 | 11.12 ± 0.18 | 68.50 ± 2.70 | |
1,3-Butadiene | μg/stick | NQ | NQ | 8.13 ± 0.17 |
μg/mg nicotine | NQ | NQ | 7.74 ± 0.16 | |
Benzene | μg/stick | 0.17 ± 0.00 | 0.36 ± 0.05 | 49.95 ± 1.54 |
μg/mg nicotine | 0.36 ± 0.00 | 0.37± 0.05 | 47.57 ± 1.47 | |
Benzo( |
ng/stick | NQ | NQ | 24.00 ± 0.64 |
ng/mg nicotine | NQ | NQ | 22.86 ± 0.61 | |
NNK | ng/stick | 6.61 ± 0.86 | 5.32 ± 0.89 | 281.00 ± 16 |
ng/mg nicotine | 14.06 ± 0.83 | 5.43 ± 0.91 | 267.62 ± 15 | |
NNN | ng/stick | 1.79 ± 0.26 | 8.37 ± 0.55 | 163.43 ± 11.19 |
ng/mg nicotine | 3.81 ± 0.55 | 8.54 ± 0.56 | 155.65 ± 10.66 |
NQ: not quantified
Machine-smoking regime: 55 mL puff volume, 2 s puff duration, 30 s puff interval. Results are expressed as means ± standard deviation (± SD). Replicate n = 3.
To assess potential mutagenicity effect, the TPM fractions of aerosols were tested in the Ames assay. The assay was performed using two tester strains, TA98 and TA100, in either presence or absence of exogenous metabolic activation system. The results are shown in Tables 6 and 7. In the presence of metabolic activation, the 3R4F TPM gave reproducible positive responses up to 500 μg TPM per plate (Table 6). In contrast, the TPM fraction of the two HTPs aerosol did not induce significant increase in the number of revertants up to 2000 μg TPM/plate under any of the assay conditions (Table 7).
Dose (μg/plate) | TA98 | TA100 | ||
---|---|---|---|---|
−S9 | +S9 | −S9 | +S9 | |
NC | 38.0 ± 2.4 | 60.0 ± 5.9 | 115.3 ± 3.8 | 156.3 ± 2.9 |
0 | 25.7 ± 4.0 | 51.0 ± 5.1 | 121.7 ± 6.0 | 107.0 ± 6.4 |
50 | 38.3 ± 8.2 | 87.7 ± 5.2 | 125.7 ± 8.5 | 105.0 ± 3.7 |
100 | 28.3 ± 7.5 | 124.3 ± 5.4* | 127.3 ± 7.6 | 115.3 ± 7.5 |
250 | 32.7 ± 6.1 | 237.7 ± 11.8* | 143.3 ± 8.3 | 133.0 ± 4.1 |
500 | 34.7 ± 2.1 | 233.7 ± 33.0* | 137.7 ± 11.1 | 171.7 ± 13.0 |
PC | 899.0 ± 24.0 | 2437.3 ± 106.0 | 2719.7 ± 55.5 | 781.7 ± 95.7 |
NC: negative control, with no solvent (DMSO). PC: positive control, 2-aminoflourene was used for both TA98 and TA100 with metabolic activation (S9 mix), 2-nitrofluorene for TA98 without S9, 4-nitroquinoline
*: the number of revertant colonies was more than twice of NC. Results are expressed as means ± standard deviation (± SD).
Replicate n = 3.
Dose (μg/plate) | TA98 | TA100 | |||
---|---|---|---|---|---|
HTP-A | HTP-B | HTP-A | HTP-B | ||
−S9 | |||||
NC | 38.0 ± 2.4 | 38.0 ± 2.4 | 115.3 ± 3.8 | 115.3 ± 3.8 | |
0 | 25.7 ± 4.0 | 25.7 ± 4.0 | 121.7 ± 6.0 | 121.7 ± 6.0 | |
50 | 24.0 ± 3.7 | 30.3 ± 4.0 | 124.0 ± 4.1 | 108.0 ± 9.9 | |
500 | 35.0 ± 3.7 | 28.7 ± 4.1 | 120.3 ± 8.7 | 105.0 ± 8.3 | |
1000 | 30.7 ± 2.4 | 33.3 ± 5.7 | 114.3 ± 6.2 | 129.3 ± 5.3 | |
2000 | 30.7 ± 3.8 | 32.7 ± 4.5 | 120.7 ± 7.9 | 125.3 ± 3.1 | |
PC | 899.0 ± 24.0 | 899.0 ± 24.0 | 2719.7 ± 55.5 | 2719.7 ± 55.5 | |
+S9 | |||||
NC | 60.0 ± 5.9 | 60.0 ± 5.9 | 156.3 ± 2.9 | 156.3 ± 2.9 | |
0 | 51.0 ± 5.1 | 51.0 ± 5.1 | 107.0 ± 6.4 | 107.0 ± 6.4 | |
50 | 48.3 ± 7.6 | 48.3 ± 3.1 | 116.0 ± 19.0 | 100.0 ± 4.3 | |
500 | 48.7 ± 7.4 | 56.0 ± 4.3 | 112.0 ± 10.7 | 100.7 ± 5.4 | |
1000 | 49.0 ± 7.1 | 40.3 ± 5.2 | 113.7 ± 12.4 | 100.7 ± 9.2 | |
2000 | 51.0 ± 2.8 | 41.7 ± 7.0 | 94.0 ± 16.8 | 105.3 ± 1.2 | |
PC | 2437.3 ± 106.0 | 2437.3 ± 106.0 | 781.7 ± 95.7 | 781.7 ± 95.7 |
NC: negative control, with no solvent (DMSO); PC: positive control, 2-aminoflourene was used for both TA98 and TA100 with metabolic activation (S9 mix), 2-nitrofluorene for TA98 without S9, 4-nitroquinoline
To assess the genotoxicity of the TPM fractions for the HTP aerosol, the MN assay was performed under three conditions, a 3-h treatment with and without metabolic activation and a 24-h treatment without metabolic activation. The results are shown in Table 8. The 3R4F TPM displayed clear increases in MN frequency under all three conditions. For the TPM fractions of the two HTP aerosols, there were no statistically significant increases in MN frequency up to 600 μg TPM/mL treatment.
Sample | Concentration (μg/mL) | MN frequencies (%) | ||
---|---|---|---|---|
3h + 21h -S9 | 3h + 21h +S9 | 24h + 0h -S9 | ||
Vehicle | 11 | 1.49 ± 0.23 | 1.52 ± 0.54 | 1.08 ± 0.08 |
HTP-A | ||||
75 | 1.37 ± 0.33 | 1.28 ± 0.04 | 1.29 ± 0.08 | |
150 | 1.08 ± 0.37 | 1.20 ± 0.07 | 1.19 ± 0.04 | |
300 | 1.05 ± 0.07 | 1.21 ± 0.01 | 1.40 ± 0.08 | |
600 | 1.15 ± 0.17 | 1.10 ± 0.10 | 1.30 ± 0.01 | |
HTP-B | ||||
75 | 1.11 ± 0.02 | 1.20 ± 0.14 | 1.26 ± 0.16 | |
150 | 1.33 ± 0.22 | 1.07 ± 0.18 | 1.28 ± 0.11 | |
300 | 1.36 ± 0.30 | 1.28 ± 0.01 | 1.33 ± 0.13 | |
600 | 1.69 ± 0.30 | 1.29 ± 0.21 | 1.32 ± 0.11 | |
3R4F | ||||
14.3 | 2.03 ± 0.21 | 2.16 ± 0.92 | 2.53 ± 0.05 | |
28.5 | 3.21 ± 1.02 | 2.46 ± 0.98 | 2.85 ± 0.20 | |
57.0 | 3.93 ± 1.34 | 2.31 ± 1.29 | 2.65 ± 0.05 | |
114.0 | 4.49 ± 1.58 | 2.67 ± 1.49 | 3.03 ± 0.30 | |
PC | - | 5.68 ± 0.29 | 5.08 ± 0.55 | 6.09 ± 0.37 |
NC | - | 1.40 ± 0.30 | 1.32 ± 0.30 | 1.20 ± 0.11 |
NC: negative control, with no solvent (DMSO); PC: Positive Control; Mitomycin C (MMC) was used for 3h/24h treatment without S9. Cyclophosphamide A (CPA) was used for 3h treatment with S9. Results are expressed as means ± standard deviation (± SD). Replicate n = 3.
The NRU assay protocol followed the one described in the literature (21). The CHO cells showed significantly reduced viability after exposure to 3R4F smoke compared with that after exposure to the two HTP aerosols (Figure 2). The IC50 of HTPs could not be calculated. In contrast, the mean IC50 values of 3R4F was 49.18 μg/mL for TPM, and 2.21 μg/mL for nicotine.
In our previous studies, the aerosol release behaviors of the main TPM components of these two HTP systems were compared, such as glycerol, nicotine and water, on a paired closed-end
In this work, the two HTPs under investigation were not matched in their design characteristics, and therefore it was difficult to deduce the deeper mechanistic reasons behind the toxicant yields observed in Table 5. These have also been reflected by the differences in puff numbers and released nicotine levels for the two HTPs: HTP-A produced less nicotine and most of the measurable toxicants as compared to those in HTA-B on a per-stick basis, but when the yields were normalized to per-nicotine basis, the differences were marginal or reversed.
Generally speaking, the TobReg 9 toxicants cover the four main categories of releasing behavior (11, 22):
carbon monoxide being biomass-pyrolysis- and combustion-driven; formaldehyde and acetaldehyde being generated by low-temperature decomposition of sugars; 1,3-butadiene/benzene/acrolein being the products of carbohydrate pyrolysis; B[
Significant reductions across these four categories are a reasonable indication that this novel closed-end HTP (HTP-A) was maintaining the key emission performance that is typically found in the leading commercial HTP products (such as HTP-B), and therefore it was worthwhile to be investigated for its toxicological effects. The relative increases or decreases in these toxicants were most likely caused by the differences in tobacco blends, heating mechanisms, and not just by the fact that HTP-A had a reduced availability of oxygen during puffing (23).
In previous toxicological studies of differently designed HTPs (18, 24–26), standard regulatory genotoxicity and cytotoxicity assays were used as the first steps in screening in addition to toxicant emission analyses. Under the test conditions, it was impossible to determine the IC50 for the two HTPs up to 3000 μg/mL of TPM. This contrasted with the IC50 of 3R4F smoke which was determined to be 49.18 μg/mL of TPM (Figure 2). The exposure of the cells to TPMs from the two HTPs at 2000 μg TPM/plate failed to elicit cell viability responses (Table 7). These results are in agreement with the fact that the two HTPs emitted significantly lower toxicant levels as compared to those found in 3R4F smoke (Table 5). This is due to their mainly heated aerosol formation mechanism, and glycerol and water making up the main components of TPMs (6, 7). The Ames and MN results showed that the TPMs of the two HTPs were not genotoxic under these regulatory toxicological assays. These assays were not sensitive to distinguish any difference between the two HTPs, whether or not the comparisons were made on a per-unit mass of TPM and/or per-unit mass of nicotine in the aerosol. These findings are also in line with those published previously for the other HTP products (24–26). It appears that despite all the design differences in heating technology and tobacco blend compositions (U.S. blended
This work described a preliminary aerosol chemistry and regulatory