Oral cancer is one of the most common cancers in India with rates among the highest in the world.1 In many regions of India, oral cancer incidence rates exceed 6 per 100,000 males and in some parts they are as high as 10.8 per 100,000.1 Smokeless tobacco products (products in which there is no combustion or pyrolysis at the time of use) account for over one-third of all tobacco consumed in India. There are approximately 100 million users of smokeless tobacco products in India and Pakistan. Traditional forms of smokeless tobacco include betel quid containing tobacco, tobacco with lime and tobacco tooth powder but there are also new products with increasing popularity.2 Chewing of betel quid containing tobacco is a well-established cause of oral cancer in India.2, 3, 4 Oral leukoplakia and submucous fibrosis, likely precursor lesions to oral cancer, are also strongly linked to smokeless tobacco use. In India and other parts of southern Asia, smokeless tobacco use is a major public health problem.
Tobacco-specific nitrosamines are the most prevalent strong carcinogens in smokeless tobacco products and are widely believed to play a significant role as causes of oral cancer in people who use these products.5, 6, 7, 8, 9, 10, 11 These carcinogens are formed from tobacco alkaloids during the curing and processing of tobacco. Vast amounts of data convincingly demonstrate their presence in various forms of smokeless tobacco, but products available in India have been examined in only scattered studies and there have been no reports since 1989.6, 12, 13, 14, 15, 16, 17 In view of the variety of new smokeless tobacco products now available in India and the widespread use of these products, it is important to obtain current data on levels of tobacco-specific nitrosamines. Such data are critical in approaches to the control and regulation of smokeless tobacco products in India, and ultimately to prevention of oral cancer. Therefore, we analyzed a variety of products for N′-nitrosonornicotine (NNN), 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), N′-nitrosoanabasine (NAB) and N′-nitrosoanatabine (NAT).
C5-NNK, 5-(methylnitrosamino)-1-(3-pyridyl)-1-pentanone; GC-TEA, gas chromatography with nitrosamine selective detection; 5-MeNNN, 5-methyl-N′-nitrosonornicotine; MS, mass spectrometry; NAB, N′-nitrosoanabasine; NAT, N′-nitrosoanatabine; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; NNN, N′-nitrosonornicotine.
Material and methods
Indian smokeless tobacco products were purchased from retail stores in Gujarat, Karnataka, and Mumbai, India in October–November 2003. The date and place of purchase and batch number of each purchase was recorded. The 32 brands collected for analysis represent products commonly used in India. Most of them (22 brands), such as zarda, gutka, khaini and mishri, are chewing tobacco products that have become especially popular among teenagers and young adults in many states of India. Other tobacco-containing products were creamy snuff, a toothpaste, and moist Swedish snuff that is being marketed in India under the brand name Click. Three brands of tooth powder were of unknown tobacco content, but suspected to contain tobacco on the basis of previous analyses carried out in India. Five popular brands of chewing mixtures that do not contain tobacco (supari) were also included. University of Kentucky moist smokeless research tobacco 1S3 was analyzed for comparison. For 24 hr before analysis, the tobacco was conditioned in a chamber at a relative humidity of 60%.
Tobacco-specific nitrosamines were analyzed by gas chromatography with nitrosamine selective detection (GC-TEA) using a model 5890 gas chromatograph (Hewlett Packard, Palo Alto, CA) interfaced with a model 610 Thermal Energy Analyzer (Orion Research, Beverly, MA). The GC was equipped with a DB-1301 capillary column (30 m × 0.32 mm × 0.25 μm) [6% (cyanopropylphenyl)methylpolysiloxane; J&W Scientific, Folsom, CA] and a 2 m × 0.53 mm deactivated fused silica precolumn. The flow rate was 2.6 mL/min He; splitless injection port temperature was 225°C. The following oven temperature program was used: 80°C for 2 min, then 12°C /min to 150°C, then 7 min at 150°C, then 12°C /min to 200°C, then 10 min at 200°C.
GC-mass spectrometry (MS)-selected-ion monitoring analysis for nicotine was carried out with a model 6890 GC equipped with an autosampler and interfaced with a model 5973 mass-selective detector (Agilent Technologies, Palo Alto, CA). The GC was equipped with a DB-5MS fused silica capillary column (15 m × 0.25 mm × 0.25 μm). The splitless injection port temperature was 250°C; the oven temperature was 70°C for 0.5 min, then increased to 180°C at 10°C/min, then held for 3 min, then 50°C/min to 300°C, and returned to initial conditions. The flow rate was 1 mL/min He.
Nitrate and nitrite content were determined by ion chromatography using a Dionex ICS-2000 Ion Chromatograph.
Reference NNN, NNK, NAB, 5-methyl-N′-nitrosonornicotine (5-MeNNN), and 5-(methylnitrosamino)-1-(3-pyridyl)-1-pentanone (C5-NNK) were synthesized as previously described.18, 19, 20 NAT was purchased from Toronto Research Chemicals Inc. (Toronto, Ontario, Canada). [CD3]Nicotine was obtained from Sigma Chemical Co. (St. Louis, MO).
Tobacco-specific nitrosamines analyses was carried out by a slight modification of a method described previously by Stepanov et al.20 Five-hundred milligram of humidity-conditioned tobacco and 10 mL of citrate-phosphate buffer (pH = 4.5) containing ascorbic acid were added to a 30 mL Nalgene centrifuge tube (Nalge Nunc International, Rochester, NY). Two-hundred nanograms of 5-MeNNN (internal standard for NNN, NAT, and NAB) and C5-NNK (internal standard for NNK) were added. The samples were homogenized for 30 min with a Polytron tissue homogenizer (Brinkmann Instruments, Westbury, NY) and sonicated for 1 hr. The buffer extracts were separated from the particles of tobacco by high-speed centrifugation (15,000g, 10 min). The extracts were filtered into 50 mL glass screw-top centrifuge tubes (Kimble, Vineland, NJ), and the pH was adjusted to 7 by adding 100 μL of 10 N NaOH. Each sample was applied to a 20 mL ChemElut cartridge (Varian, Harbor City, CA), eluted with 3 × 20 mL CH2Cl2, and the eluants were combined and concentrated to dryness with a model SVT200H Speedvac concentrator (Thermo Savant, Farmingdale, NY). Residues were dissolved in 0.5 mL of CH2Cl2 and further purified by solid-phase extraction using Sep-Pak Plus silica cartridges (Waters Corp., Milford, MA), pre-equilibrated with CH2Cl2. The cartridges were washed with 5 mL CH2Cl2/ethyl acetate: 50/50, and the tobacco-specific nitrosamines were eluted with 10 mL of ethyl acetate. The ethyl acetate eluants were concentrated to dryness (Speedvac). The dry residues were transferred into GC-micro vials with 3 × 50 μL methanol, dried, and re-dissolved in 100 μL of acetonitrile. Three microliters of the prepared sample were injected into GC-TEA.
Nicotine analysis was carried out as described previously.20 Fifty milligrams of humidity-conditioned tobacco and 20 mL of methanol containing 50 mg of KOH were added to 30 mL Nalgene centrifuge tubes. The samples were homogenized (Polytron) and then sonicated for 3 hr. The methanol extracts were separated from the tobacco by high-speed centrifugation. Methanol extracts (200 μL) were transferred into a silanized 4 mL vial and 20 μL of [CD3]nicotine internal standard was added. The samples were transferred to GC-micro insert vials and analyzed by GC-MS-SIM.
For nitrate and nitrite analysis, 100 mg of humidity-conditioned tobacco and 10 mL of reagent grade water (Milli-Q, Millipore Corp.) were added to a 50 mL glass screw-top centrifuge tube (Kimble) pre-washed with water. Two water negative controls and three control solutions containing known concentrations of nitrate and nitrite were included in the sample set. Tobacco was homogenized (Polytron), and the tubes were sonicated for 30 min. The suspensions were centrifuged and the aqueous tobacco extract was applied to a C-18 SPE cartridge (Waters Corp., Milford, MA) conditioned with 2 mL of methanol. The first 5 mL of eluant was discarded. The next 2 mL of eluant was collected in a prewashed plastic tube and stored at −20°C until analysis. The samples were diluted 10-fold before analysis by ion chromatography. Conditions were as follows: an AS14 anion exchange column and guard column were eluted with carbonate/bicarbonate using a 50 mL sample loop and a flow rate of 1.0 mL/min. These analyses were carried out at the University of Minnesota Geochemical Analysis Facility.
Pearson correlations were determined using Sigma Plot 2001, v. 7.101 (SPSS, Inc., Chicago, IL).
A typical GC-TEA trace of tobacco-specific nitrosamines in one of the smokeless tobacco products analyzed here is presented in Figure 1.
Levels of tobacco-specific nitrosamines, nitrate, nitrite and nicotine in the products are summarized in Table I. Each value is the mean of 2 analyses for tobacco-specific nitrosamines; the results agreed on average within <10%. Recoveries of internal standards averaged 42.8%. For nitrate and nitrite, each value represents the mean of duplicate injections of the same sample, with the average difference between the 2 values being <3%. For nicotine, a single sample of each brand was prepared, and each value is the result of a single injection.
Table I. Tobacco-Specific Nitrosamines, Nitrate, Nitrite, and Nicotine in Indian Smokeless Tobacco and Related Products1
All data per gram wet weight. ND, not detected; detection limit 50 pmol/g tobacco; NA, not analyzed.
Mean of duplicate analyses of product from one package.
Mean of duplicate injections of a single sample.
Creamy snuff/ toothpaste
On the basis of the tobacco-specific nitrosamine analyses, the products can be divided into 3 groups. The first is products with high levels (Raja and Hans Chhap khaini, Shimla zarda and Gai Chhap tobacco). In these products, levels of NNN and NNK were 38.9 ± 27.0 (SD) μg/g, range = 19.2–76.9 μg/g and 8.99 ± 13.0 μg/g, range = 2.34–28.4 μg/g, respectively. The second group comprises products with medium to low levels of tobacco-specific nitrosamines. NNN and NNK in these products amounted to 2.24 ± 2.63 (SD) μg/g, range = 0.09–8.36 μg/g (n = 20) and 0.71 ± 0.86 μg/g, range 0.04-3.09 μg/g (n = 20), respectively. In the third group (tooth powders and supari), tobacco-specific nitrosamines were rarely detected.
The highest levels of NNN and NNK, 76.9 μg/g and 28.4 μg/g, respectively, were observed in Raja khaini. The second highest NNN level, 39.4 μg/g, was observed in Hans Chhap khaini. Among the products in which tobacco-specific nitrosamines were commonly detected, the lowest levels were observed frequently in different gutka brands.
Nitrite varied from non-detectable (<0.02 μg/g wet weight tobacco) to 1,020 μg/g and 1,410 μg/g in Raja khaini and Hans Chhap khaini, respectively. The average level of nitrate was 720 ± 870 (SD) μg/g, range = 7.5–2950 μg/g (n = 32). Levels of total tobacco-specific nitrosamines did not correlate with nicotine or nitrate. A correlation was observed between total tobacco-specific nitrosamines and nitrite (r = 0.78, p < 0.0001).
We analyzed 32 Indian tobacco products, including smokeless tobacco products, tobacco-free chewing products, creamy snuff, tobacco toothpaste and tooth powder. These products were purchased in 2003 in India and are used commonly in different parts of the country.
Our study shows that the levels of tobacco-specific nitrosamines in these products vary widely. Different brands of the same type of product usually contain similar levels of tobacco-specific nitrosamines, nitrate, nitrite and nicotine. This observation can be explained by similarities in tobacco processing and is in agreement with the general principle that yields of tobacco-specific nitrosamines are influenced greatly by the processes involved in the manufacturing of smokeless tobaccos.12, 21, 22, 23 The highest levels of tobacco-specific nitrosamines were observed in 2 different brands of the same variety, khaini. Khaini is a mixture of tobacco, lime and menthol or aromatic spices. The mode of tobacco processing that likely favors the reduction of nitrate to nitrite and nitrosating agents could be responsible for the high tobacco-specific nitrosamine concentrations in these 2 brands. This seems reasonable because the levels of nitrite in these 2 brands are the highest in our study and, arguably, among the highest reported in smokeless tobacco products.13 It should be mentioned that khaini is usually placed in the mouth and kept there. An extraordinarily high amount of nitrite will then be released into saliva and swallowed. As a result, additional amounts of N-nitroso compounds could be formed endogenously.
Another tobacco product with relatively high tobacco-specific nitrosamine levels is zarda, which is usually chewed or kept in the mouth. To produce zarda, tobacco leaf is boiled in water with lime and spices until evaporation. The residual particles are then dried and colored with vegetable dyes. Four brands of this product (Goa 1000, Moolchand Super, Sanket 999, Baba 120) contain an average of 6.85 ± 1.55 (SD) μg/g NNN. The fifth brand (Shimla) is relatively high in NNN content (19.9 μg/g), even though the levels of nitrite and nicotine are similar to the other zarda brands.
The “other tobacco products” (Table I) that are used for chewing may be processed or unprocessed. It is interesting to note that the brand Gai Chhap, which is made from unprocessed tobacco, contains the highest tobacco-specific nitrosamine levels of this group. Clearly, factors other than processing can influence nitrosamine levels in these products.
Mishri is a powdered form of tobacco that is used primarily for cleaning teeth. It is prepared by baking tobacco on a hot metal plate until it becomes uniformly black. The brand Shahin mishri was found to contain 4.21 μg/g NNN and 0.87 μg/g NNK. As with some of the other products studied here, nitrosamine uptake from mishri may increase when it is used habitually (i.e., being placed and retained in the mouth several times a day).
Gutka usually contains powdered tobacco, betel nut, catechu, lime and flavors. It has been commercialized since 1975, having originally been available custom mixed from pan-vendors. The use of these products is strongly associated with oral cancer.2, 3, 4 The levels of tobacco-specific nitrosamines in gutka were lower than those in many of the other products examined here, but were still considerably higher than nitrosamine levels in food and other common products, which are typically in the low ppb range.11, 24 Supari, which is similar to gutka but does not contain tobacco, did not have detectable levels of tobacco-specific nitrosamines.
Tobacco is not mentioned as an ingredient of red tooth powders. Small amounts of nicotine and trace amounts of tobacco-specific nitrosamines were observed in these products, however, raising concerns about their safety. Considerable levels of tobacco-specific nitrosamines and nicotine were found in Dentobac, a tobacco-containing toothpaste. It is remarkable that a product containing relatively high levels of carcinogens and an addictive agent is marketed for the purpose of dental hygiene.
The levels of tobacco-specific nitrosamines in 3 of the products analyzed (Raja khaini, Hans Chhap khaini, Gai Chhap tobacco) are considerably higher than those found in most smokeless tobacco products marketed in Europe and North America, where the total amounts of these compounds are usually <10 μg/g.12, 13, 25 Levels of these carcinogens in Kentucky reference smokeless tobacco 1S3 are <8 μg/g (Table I). Our results serve to emphasize the potential hazards of these products marketed in an area of high oral cancer incidence.
To our knowledge, there have been no published reports on tobacco-specific nitrosamines in Indian smokeless tobacco products in the past 15 years. Brunnemann et al.15 reported levels of tobacco-specific nitrosamines in tobacco used in betel quid. The amounts were similar to those reported here in gutka. Nair et al.14 found high levels of tobacco-specific nitrosamines in Indian chewing tobacco and creamy snuff. Tricker and Preussmann16, 17 reported levels of tobacco-specific nitrosamines in zarda tobacco similar to those found here and also observed relatively high levels in Kiwam tobacco. It seems that little has changed in the past 15 years with respect to levels of these carcinogens in tobacco products marketed in India. High exposure to tobacco-specific nitrosamines in smokeless tobacco products is likely a major factor in the continuing epidemic of oral cancer in India. Immediate public health measures are urgently needed to decrease morbidity and mortality associated with the use of these products.
We thank S.G. Carmella for his advice and J. Hodge for technical assistance.