Response To The FDA Summary

Technical Memorandum

July 30,20090900923.000 D0T0 JH083

Review and Analysis of FDA’s Report

Study Design

The FDA tested for the presence of key chemicals in the liquid contained in the whole cartridges, and in small aliquots of the vapor generated from the cartridges by two separate techniques. Analyte detection techniques included gas chromatography followed by mass spectrometry (GC/MS)1,liquid chromatography followed by mass spectrometry (LC/MS),and high performance liquid chromatography with ultraviolet analyte detection (HPLC/UV). With respect to NJOY’s products, four types of cartridges were tested: (1) menthol high, (2)menthol medium, (3) regular medium, and (4) regular low. Not all cartridges were tested in each analysis. Nicotine and its major metabolite, cotinine, were the only chemicals able to be quantifiably measured by any of the experimental techniques used in the analysis.

In the whole cartridge liquid,the FDA assayed for nicotine, four major TSNAs [N-nitrosonicotine (NNN), N-nitrosoanabasine (NAB), N-nitrosoanatabine (NAT) and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK)]; tobacco-specific impurities [(cotinine, nicotine-N-oxide, nornicotine, anatabine, anabasine, pseudooxynicotine, myosmine, β-nicotyrine, and 1-methyl-3-nicotinoylpyrrolidine (MNP)]; and diethylene glycol (DEG). In the vapor phase,the FDAassayed for nicotine and tobacco-specific impurities using both thesparging apparatusand a heated head space vapor method that was meant to simulate the temperatures achieved by the electronic cigarette devices. The FDA-approved Nicotrol®inhaler was used as a “control” for some test methods.1GC/MS -Amethod that combines gas-liquid chromatography (the separation and analysis of compounds that can be vaporized) and mass spectrometry (a technique to determine the elemental composition of a molecule) to identify and estimate the concentration of substances within a test sample.

Technical Memorandum

July 30,20090900923.000 D0T0 JH084ResultsWhole Cartridge: Tobacco-Specific Nitrosamines by LC-MS/MS All four TSNAs were detected in all four of the NJOY cartridges tested by whole cartridge liquid analysis. As indicated by Table 1 legend “A” the detection was below the limit of quantification(LOQ)for each of the TSNAs:NAB (LOQ = 21 ppb); NAT (LOQ = 21 ppb); NNK (LOQ = 75 ppb); NNN (LOQ = 24 ppb). The limit of detection (LOD) for these chemicals using their techniques was not given. All values for the NJOY cartridges were found to be below the specified limit of quantification given above. The Nicotrol® inhaler “control” was not tested for the presence of these compounds. Whole Cartridge: Tobacco-Specific Impurities by GC-MS and GC-MS/MS2Liquid in whole cartridges was screened for possible tobacco-specific impurities: cotinine, nicotine-N-oxide, nornicotine, anatabine, anabasine, pseudooxynicotine, myosmine, β-nicotyrine, and 1-methyl-3-nicotinoylpyrrolidine (MNP). Nicotine-N-oxide, nornicotine, anatabine, pseudooxynicotine, and MNP were not detectedin any of the NJOY samples. Cotinine, anabasine, myosmine, and β-nicotyrine were detected in whole cartridge liquid analysis of all four NJOY cartridges tested. However, none of these detection levels were quantified. The level of quantification was not provided for these chemicals in the table legend. The limit of detection (LOD)for cotinine, anabasine, myosmine, and β-nicotyrine was 20 ppb, 10 ppb, 69ppb, and 170 ppb, respectively. Additionally, the Table1legend “B” indicated that the levels of cotinine, anabasine, myosmine, and β-nicotyrine found in the liquid of the NJOY cartridges were “present, but at less than the level of the Nicotrol® inhaler specification.”The specification was listed as not more than (NMT) 0.5%for three of these compounds: cotinine, myosmine, and β-nicotyrine. Based upon reference to these manufacturing specifications, it is not clear whether or not the Nicotrol® 2GC-MS/MS –This technique is performed by adding a second phase of mass fragmentation. Tandem mass spectrometry (MS/MS) is a more powerful technique that offers a higher sensitivity and selectivity for target compound analysis found in very low levels.

Technical Memorandum

July 30,20090900923.000 D0T0 JH085

inhaler device was actually tested for the presence of these compounds or whether the manufacturer’s specifications were relied upon as a valid source of quantification. Whole Cartridge: Nicotine content by HPLC-UV Nicotine levels were measured in the liquid of all four NJOY cartridges by two different extraction techniques. Values were similar using each technique. The highest amount of nicotine detected from the NJOY menthol high cartridge was 6.76 mg. This is the amount available for extraction from an 18mg NJOY nicotine cartridge using the FDA techniques as stated in the report. The amount of nicotine extractable from the Nicotrol® device was apparently not tested. Manufacturer’s inserts for this product claim that 4 mg of nicotine is extractable from a 10 mg inhalant cartridge. Simulated Use: Nicotine and Tobacco-Specific Impurities by Head Space GC-MS (HSGC-MS) and Sparging Apparatus β-nicotyrine and cotinine were the only tobacco impurities(other than nicotine) detected by both of the vapor analyses. Two NJOY cartridges were tested in the sparging vapor analyses; menthol high and menthol medium. “Trace” amounts of cotinine were found in the menthol high cartridge. This designation “trace” is not defined. Cotinine was not detected in the vapor of the menthol medium NJOY cartridge. β-nicotyrine was detected, but not quantified in the vapor of three NJOY cartridges tested using the head space technique; menthol high, menthol low, and regular high. Nicotine levels were found to vary between three different menthol high cartridges tested by the sparging technique. Quantified nicotine levels of the three menthol high cartridges were43.2, 34.9, and 26.8 mcg/100mLpuff.Also, the menthol medium cartridge was found to have 10.6 mcg nicotine/100mLpuff.The Nicotrol® inhaler device was described to have 15.2 mcg/100mLpuff;however,there was no mention in the methods section of the report of the Nicotrol® inhaler being tested with the same apparatus as the e-cigarette. The LOQs and LODs were not specified for either of the vapor phase techniquesor for any of the impurities. Use of standards to establish LODs and LOQs was also not discussed. Additionally, there was no mention of testing for any of the TSNAs in the vapor phase analyses of these same products.

Technical Memorandum July 30, 20090900923.000 D0T0 JH086Whole Cartridge: Diethylene Glycol and Ethylene Glycolby GC/MSAll four NJOY cartridges were found negative for the presence of diethylene glycol. Discussion Tobacco-Specific Nitrosamines TSNAs are common constituents of tobacco extracts and have been reported in scientific studies to be present in FDA-approved nicotine replacement therapy (NRT) devices. TSNAs are also normal constituents of conventional cigarette smoke. The FDA reported finding non-quantifiable but detectable levels of these compounds in the liquid of NJOY’s cartridges but did not test for the presence of these compounds in the vapor phase. The levels of these compounds in the NJOY cartridges were below the LOQs listed for these compounds (21–75 parts per billion).However, in the methods section of the FDA report, the sensitivity of this assay is described as being sensitive to 40 parts per trillion (Wu et al. 2008). Further, the report states the results of the testing were quantifiable in this methods section. “The assumption was made that recovery of TSNAs from the E-cigarette cartridge assembly was as good as that published by Wu, et al., and quantitative”. Elsewhere in the report (Table 1),it states that the results were not quantifiable and were below the limit of quantitation. It is difficult to know, due to the conflicting statements, but it may be that 40 parts per trillion is actually the limit of detection in this assay. If this is the case, then the values of the TSNAs obtained by the FDA may be as low as parts per trillion and as high as just below the limits of quantitation listed in the report; ppb. For comparison purposes, the levels of these compounds in mainstream and side stream smoke from conventional cigarettes is on the order of parts per million (Hoffman et al. 1978).In Hoffman’s study, the tobacco of 5 different cigarettes contained between 0.22 and 7.0 ppm (μg/g) of the carcinogenic NNN, 0.13 and 0.74 ppm of the carcinogenic NNK and 0.44–3.2ppm of NAB. The levels of the carcinogen NNN in Nicorette® 4 mg gum are approximately 2.0 ppb (ng/g) dry gum weight. The carcinogen NNK was found in the NicoDerm CQ® 4 mg patch at levels

Technical Memorandum  July 30,20090900923.000 D0T0 JH087of 8 ppb (ng/g) dry patch weight (Stepanov et al. 2006). Osterdahl et al. found up to 380 ppb (ng/g)of TSNAs per gram of Nicorette® chewing gum(Osterdahl et al. 1990).The researchers also detected the presence of these compounds in the saliva of Nicorette users, confirming that these substances were being passed into the saliva of users of this product. All the saliva samples obtained during Nicorette® chewing contained NNN and NAT/NAB at levels ranging from 0.42 to 19 ppb (ng/g) and from 1.3 to 46 ppb (ng/g),respectively. No TSNA could be detected in saliva samples collected just before Nicorette® chewing.The only NRT found not to contain TSNAs that was tested was the Commit® 2mg lozenge (Stepanov et al. 2006). The Nicotrol® inhaler was not tested by either group above, or by the FDA for the presence of TSNAs in the liquid or the vapor (aerosol). TSNAs were also not tested for in the vapor of NJOY’s products. It is the vapor that the users of these products are exposed to, not the liquid. Nitrosamines are known to volatilize poorly into the vapor phase (Hoffman et al. 1979). Therefore, it is speculative as to whether they are actually are bioavailable to the user of e-cigarette through normal use of the device. In contrast, FDA-approved NRT products, Nicorette and NicoDerm CQ, do not volatilize their products for delivery so it is quite possible, and indeed shown in the case of Nicorette gum, that users may be receiving measurable amounts of these compounds through normal use of these products(Osterdahl et al. 1990). NJOY is not aware of any testing that indicates that TSNAs are detected in the vapor phase of their products. Based on the results of the FDA analysis of the whole cartridge liquid, if there is any TSNA in the vapor phase of NJOY products, the levels would be predicted to be on the order of less than the LOQ for each substance. Tobacco-Specific Impurities Cotinine, anabasine, myosmine, and β-nicotyrine were each detected in whole cartridge liquid analysis of all four NJOY cartridges tested, but at levels lower than those found in the Nicotrol® inhaler device and below the manufacturer specifications for the devices. Table1 legend “B” indicated that the levels of cotinine, anabasine, myosmine, and β-nicotyrine when the GC-MS/MS technique was applied were “present but at less than the level of the Nicotrol® inhaler specification.” Furthermore, in the Nicotrol® inhaler control sample, cotinine, myosmine, and

Technical Memorandum

July 30,20090900923.000 D0T0 JH088β-nicotyrine were detected individually at not more than (NMT) 0.5%. The limit of quantification was not provided for these compounds. It is not clear whether or not the liquid in the Nicotrol® device was tested for the presence of these compounds or whether the manufacturer’s specifications were relied upon as a valid source of quantification. Due to the lack of numerical valuesreportedit seems that the manufacturing specificationsmay have been used as a possiblesource of the quantitation of levels of impurities in theNicotrol®inhalerdevice. β-nicotyrine and cotinine werethe only tobacco impurities(other than nicotine) detected in NJOY cartridges by both of the vapor analyses. Information in the available scientific literature regarding these two compounds indicates that neither have been found to be carcinogenicor particularly toxic by standard [Ames] tests. Cotinine is a weak metabolite of nicotineand has been shown to lack mutagenic or carcinogenic potential through several published Ames Toxicity assays. Other studies have shown that the activity of cotinineis several hundred-fold less than nicotine in behavioral studies using rats (HSDB 2005). One in-vitro study of β-nicotyrine has shown that it can inhibit the class of enzymes responsible for the metabolism of nicotine to cotinine, thus slowing the metabolism of nicotine (Denton et al. 2004), but it currently has no established toxicity profile.NicotineNicotine levels were measured in the liquid of all four NJOY cartridges by two different extraction techniques. Values were similar using each technique. The highest amount of nicotine detected from the NJOY “menthol high” cartridge was 6.76 mg. This isthe amount available for extraction from an 18 mg NJOY nicotine cartridgeaccording to the FDA report, Table 1. The amount of nicotine extractable from the Nicotrol®device using these techniques was not testedin this report. Manufacturer’s package inserts for this product claim that4 mg of nicotine is deliverablefrom a 10 mg inhalant cartridge.Nicotine levels were found to vary between three different menthol high cartridges tested by the sparging technique. Quantified nicotine levels of the three menthol high cartridges were 43.2,

Technical Memorandum

July 30,20090900923.000 D0T0 JH08934.9, and 26.8 mcg/100mLpuff.The menthol medium cartridge was found to have 10.6 mcg nicotine/100mLpuff.The Nicotrol® inhaler device was described to deliver15.2 mcg/100mLpuff. There was no mention in the methods section of the report, however, of the Nicotrol® inhaler being tested with the same apparatus as the e-cigarette. For comparison of nicotine delivered to the user by an FDA approved device, the Nicotrol® inhaler inhalation package comes to the user as 168 10 mg cartridges. Recommended dosage is 6–16 cartridges daily. The manufacturer recommends an intensive inhalation regimen(80 deep inhalations over 20 minutes),which releases on the average 4 mg of the nicotine content of each cartridge. If each breath is approximately 100 cc or 100 mLin volume as is typically assumed, then 4 mg/80 (100mL) breaths would yield 0.05 mg nicotine/puff or 50mcg per puff, which is actually more than the highest amount of nicotine delivered by the NJOY menthol high cartridge according to FDA tests. Note that the maximum recommended daily dose of the Nicotrol® inhaler is 40mg.For the 10mg inhaler, recommended dosing is 6–16 cartridges a day. Thus, if4 of the 10mg is delivered, as is stated in the packet insert, then the use of 16 cartridges (the maximum recommended dose) would yield 64mg of nicotine per day. Study Design and Presentation of Results One of the weaknesses in the FDA study was the lack of study controls. The FDA-approved Nicotrol® inhaler was presented as a “control” for their studies, but was only used for some of the experiments. The device was never tested for the presence of the same “carcinogens and toxic chemicals” that were found in some of the e-cigarette cartridges. When it was indicated that the Nicotrol® inhaler device did contain some levels of tobacco associated impurities, it was never discussed in the report or even mentioned in the results section. These tobacco-specific impurities in the Nicotrol® inhaler were also not discussed in comparison to the values obtained for NJOY’s or Smoke Everywhere’s products. With respect to the TSNAs, the report did not mention that these substances are found in nicotine gum, the patch, nasal spray, and lozenges in concentrations that are at the very least similar to, or higher than those found in the NJOY cartridges(Stepanov et al. 2006; Osterdahl et al. 2004). Specific levels were difficult to approximate from the non-quantitative “detect” and “non-detect” results shown in the report

Technical Memorandum

July 30,20090900923.000 D0T0 JH0810because LODs were given in the absence of LOQs and vice versa. Therefore, with respect to TSN As, it is possible that the values are as low as parts per trillion and as high as parts per billion. Information was not presented regarding standards and quality control parameters. Limited information was given on the number of cartridges tested for each experiment within a group and the number of times each cartridge was tested within an experiment. Furthermore, no statistical analysis was presented. Key carcinogenic and toxic chemicals that were found in the liquid phase were not examined in the vapor phase. From a human health exposure perspective, it would have been beneficial to determine if the same chemical constituents found in the liquid phase analysis were also found in the vapor phase, especially considering that the vapor phase is the pathway of exposure to the e-cigarette user. Concluding Statement The detection of trace and non-measurable levels of TSNAs and tobacco-associated impurities in the liquid, rather than the vapor phase of NJOY’s products, at levels that are many orders of magnitude below conventional cigarettes, and at or below FDA-approved nicotine containing products, should be considered as indicators of the regulatory acceptability of the NJOY products rather than reason for concern. When considering the relative potential health risks posed by these trace levels, it is worth noting that the approved NRTs, which have been shown to contain these substances, were not judged to contain levels sufficient to warrant toxicity information or reference to these substances in their own product literature.