Tobacco Documents Online

Topical Report NCI/S&HF/f1RPlL #46 SITE VISIT IV TO HAZLETON LABORATORY Reston, Virginia, May 21-22, 1977 8-15-77 R. A. Jenkins, R. B. Quincy, and B. E. Gill Tobacco Smoke Research Program Bio/Organic Analysis Section Analytical Chemistry Division Oak Ridge National Laboratory Oak Ridge, Tennessee 37830 Interagency Agreement (ERDA-NIH/NCI) No. 40-485-74, Part II Internal (ORNL) Contract Charge Number 3390-0224 Intended for informal communication with project management only. Confidential until published or'-released by author.

SITE VISIT IV TO HAZLETON LABORATORY R. A. Jenkins, R. B, Quincy, and B. E. Gill Summary. The data gathered on a fourth working site visit to Hazleton indicates that the beagle exposures there continue to be conducted in a consistent manner. The smoke doses which the animals are offered are similar to, but slightly lower than, those measured on the previous visit. Overall machine smoke delivery efficiency continues to be good. Detailed chemical analysis of the smoke of the Code 79 variant suggests that it is very similar in composition to that of the Code 32 variant, and that no significant chemical changes occur as the particulates pass through the exposure devices. The most important new finding on this visit was that the animals of the Code 90 exposure group may be experiencing wide fluc- tuations in gas phase dose, depending on the resistance-to-draw of the cigarettes which they are smoking. The preliminary isolation of two "types° of Code 90 cigarettes could have a serious effect on interpretation of blood carboxy-hemoglobin levels as an assessment of smoke dose. Introduction. A fourth site visit was made to F{azleton Laboratory on May 21 and 22, 1977, to observe and document the chronic exposure of beagles to tobacco smoke. As has been the case on previous visits, our primary purpose was to collect data to enable us to better estimate the dose which the ani- mals receive and to determine the extent to which certain smoking factors can contribute to the variability of that dose. Both exposure and experi- mental details have been previously discussed, and can be found in reports covering previous site visits (Topical Reports NCI/S&HP/ORNL #24, 29, 33, 36A). Experimental. In order to accomplish this exposure definition, it has been necessary to obtain information concerning both the physical operating performance and the smoke delivery performance of the ADL-II exposure devices in service. However, since the ADL-II's have proven to be very reliable, extensive machine puff parameter measurements have been eliminated in favor of a few measurements of puff'vo1ume. We are placing greater emphasis in our monitoring effort on a more detailed chemical characterization of the smoke offered the animals. On this visit, sixteen samples of smoke generated

by the Code 79 variant (including both ADL-II input and cannula exit samples) were returned and subjected to a multi-component, multi-aliquot analysis. In this procedure, several key components of the particulate phase of the smoke are analyzed by gas chromatography. This permits a much more detailed chemical description of the smoke, as several of these components correlate highly with the presence of entire classes of smoke constituents. In addition, this was our first opportunity to sample, at Hazleton, the cannula exit for carbon monoxide (CO) and carbon dioxide (C02). CO is important because of its physiological impact, and the ratio of CO to C02 content in the gas phase gives an indication of pyrolysis conditions in the burning cone of the cigarette. Briefly, the procedure and apparatus for taking a gas phase sample from the cannula exit are similar to that for ob- taining a particulate phase sample, except that the effluent from the respir- ator (ventilator) is expelled into a large, evacuated Saran gas sampling bag. Following completion of smoking, the contents of the bag are mixed, and a portion is drawn into a small glass gas sampling bulb. The bulb is returned to ORNL for analysis, which is performed by gas chromatography using Amberlyst ion-exchange resin in the Ni-form as the "stationary phase." Results and Discussion. The mechanical operation and reliability of the ADL-II exposure devices continue to be very good. On this fourth visit, puff volume measurements were made on all the machines in service. The mean puff volume was found to be 33.8 ± 1.0 ml/puff, which is virtually identical to values determined on the last two visits, and differs less than 4% from the specified value (35.0 mlJpuff). Table I compares the static burn rate of the Code 04 variant in both exposure wards at Hazleton with that determined under analytical conditions at ORNL. In contrast-to previous visits, the exposure ward static burn rates are very close to those determined at ORNL. This could be due to a7 lowered ventilation air flow rate at Nazleton, plus a higher relative humidity. On this visit, the humidity was observed to fluctuate between 51% and 63% at 23°C. This is close to that of analytical smoking conditions and con- siderably , siderably above the 30% RH measured last November.

The total particulate matter (TPM) and nicotine deliveries of the cigarette smoke generated by selected ADL-II exposure systems at Hazleton are tabulated in Table II. The data indicate that the machine-variant combinations are fairly consistent in the amount of smoke generated. The nicotine:TPM ratio of the smoke is slightly lower than that measured on previous visits. Table III compares the average cigarette deliv- eries measured on-site with those determined at ORNL from the same batch of cigarettes. A comparison of the nicotine delivered by a given variant shows that the cigarettes generate approximately 80-90% of the amount of smoke that's generated under analytical conditions. Slightly lowered puff volumes and slightly higher static burn rates could be contributing factors, but they are probably not responsible for the entire difference. In general, the amount of smoke generated by the variants on-site was somewhat lower than that found on previous visits. Table IV tabulates individually the TPM and nicotine values of the smoke reaching the cannula exit of the exposure systems. For any given code, the TPM and nicotine content of the output smoke is fairly consistent, suggesting that the dose which the animals receive will not vary appreciably as they are shifted from one machine to another. Table V compares the average cannula exit results with input TPM and nicotine values for each variant (from Table III). For reasons discussed in previous reports, nicotine is believed to be a better indicator of smoke particulate phase reaching the cannula exit. Thus, the delivery efficiency of the ADL-II exposure devices can be considered to be varying between 78% and 954ro. This is good efficiency, and suggests that little particulate phase condensation occurs within the exposure devices. Probably the most important chemical parameter relating to the exposures themselves is the dose which the animals receive. Presently, our best estimate of this is the smoke dose which the animals are offered. In the case of the Hazleton study, we consider the amount of smoke reaching the cannula to be a good estimate of the amount of smoke the animal inhales, because the cuffed cannula is designed to prevent the animal from breathing in around the cannula and thus not inhaling the full bolus of smoke. Table VI lists our estimates of the amount of smoke offerea the animals, per variant, on a weekly basis. With the exception of the animals exposed to the Code 90 variant, the animals are offered about the same amount of smoke particulates (±10%). For a given variant, the nicotine doses are similar to, but slightly lower than, those measured on the last visit.

In an efforit to more exactly define the chemical composition of the smoke offered the animals, we have been conducting more detailed chemical characterization studies of the particulate phase of smoke samples acquired on-site. Table VII compares the multi-component, multi-aliquot profiles of input and cannula exit samples of the smoke of the Code 79 variant. A com- parison of the relative concentration of constituents in the input and output samples suggests that none of the visualized components (with the exception of phenol) are significant enriched or depleted upon passage through the exposure system. The apparent depletion of phenol, one of the more volatile constituents, is probably a result of re-volatilization off the filter pad during cannula exit sampling. It is interesting to note that the observed chemical composition of the smoke of the Code 79 variant is very similar to that of the Code 32 variant, reported in Topical Report #36A. Table VIII compares individual cannula exit values for TPM, nicotine, CO, and C0, for two variants, Code 04 and Code 90. If CO and C02 can be considered to be good indicators of total gas phase constituents, then the data here suggests that the animals exposed to the Code 90 variant are receiving only about half the amount of gas phase constituents as animals exposed to the SEB-IV (Code 04) cigarettes. However, the TPM and nicotine values for these samples of Code 90 smoke are somewhat lower than those on Table V, suggesting that the CO/CO2 smoke samples may not be a completely representative sampling. For both codes, the CO:nicotine ratios are fairly constant. Since CO cannot condense inside the exposure device, this suggests that variations in smoke delivery to the cannula exit are due to cigarette variability, and not to fluctuations in machine delivery efficiency. Table IX compares the average cannula exit values of CO, C02, nicotine and TPP•1 for Codes 04 and 90 with those obtained by smoking analytically samples of the same batch of cigarettes. Analytical results for Codes 79 and 13 are included for comparison. Several things should be noted. First, the CO:CO2 ratio of the exposure smoke of both codes is significantly lower than that of smoke generated analytically. Since a lower CO:CO2 ratio is indicative of a lowered effective puff volume through the combustion zone of the cigarette, the data may indicate that there are small leaks around the butt of the cigarette where it fits into the holder on the ADL-II.

Secondly, the Code 90 variant is constructed with air dilution holes in the filter, which are designed to lower effective puff volume through the tobacco rod of the cigarette. Since there was such a marked difference between on-site samples and those from cigarettes selected at random from the Code 90 cigarettes returned from Hazleton, we investigated this unexpected finding further, and have since isolated essentially two groups of Code 90 cigarettes based on the resistance-to-draw (RTD) of the air dilution holes. One group of cigarettes had a very low RTD (ti 20mm H20), the other, a much higher RTD (-. 40 mm Hz0). As can be observed in Table IX, the effect of the cigarette RTD on the amount of smoke.generated under analytical conditions by the Code 90 variant is dramatic. The lower RTD cigarettes deliver about half of the gas phase (CO, C02) and a substantially reduced particulate phase from that of a higher RTD Code 90. The fact that there appear to be two "types" of Code 90 cigarettes can have a serious impact on the interpretation of any carboxy-hemoglobin measurements made on animals exposed to this variant. Without knowing the °type° of Code 90 cigarette to which the animal had been exposed on the day of the COHb measurement, it would be difficult to relate COHb to the overall level of smoke dose. Also, the cannula exit values for the Code 04 cigarettes are lower than the analytical results. However, based on the amount of nicotine and TPM which that variant generates on-site (from Table III), a delivery of 43.2 ml COz to the cannula would be predicted. This is in good agreement with the value of 45.2 ml which was observed. This does not account for the lowered delivery of carbon monoxide.

Table I Comparison of Code 04 Static Burn Rate Exposure Conditions at HL vs Analytical Smoking Conditions at ORNL Burn RatR, mTtImi n Burn Rat9, mg/mi n Code 04 at HL, in Ward #1 (Nearer front of building) 5.23 ± 0.21 75.1 ± 3.3 Code 04 at HL, inLJard~2 (Nearer rear of building) 5.04 ± 0.27 71.3 ± 3.1 Code 04 at ORNL 4.75 ± 0.24 67.0 ± 3.2 Analytical Smoking Conditions aBurn rate given in terms of actual length of cigarette burned. bBurn rate given in terms of actual weight of cigarette burned.

Table II Total Particulate Matter (TPM) and Nicotine Delivery of Cigarettes Smoked on Selected ADL-II Smoking Machines in Service at HL on May 21-22, 1977 Cigaret Code te Machine Number TPM, mg/cig Nicotine, mg/cig Nicotine:TPM Ratio 90 14 37.4 2.03 .0543 34 33.2 1.92 .0578 31 37.8 2.08 .0550 10 34.2 1.80 .0526 30 37.2 1.94 .0522 7 43.5 2.37 .0545 25 31.4 1.66 .0529 39 45.5 1.69 .0371 Average 37.5 ± 4.9 1.94 ± 0.23 .0521 ± .0063 79 17 46.9 2.78 .0593 8 37.1 2.31 .0623 24 46.8 2.79 .0596 6 39.6 2.37 .0598 37 41.9 2.25 .0537 36 40.3 2.14 .0531 29 41.9 2.28 .0544 16 39.7 2.23 .0562 Average 41.8 ± 3.5 2.39 ± 0.25 .0573 ± .0034 SEB IV 32 45.8 2.84 .0620 (04) 40 48.7 3.13 .0643 2 46.0 2.84 .0617 13 48.6 2.90 .0597 Average 47.3 ± 1.6 2.93 ± 0.14 .0619 ± .0019 13 22 42.8 0.22 .0051 11 32.7 0.22 .0067 15 36.1 0.22 .0061 38 41.0 0.23 .0056 Average 38.2 ± 4.6 0.22 ± 0.01 .0059 ± .0007

Table III Comparison of Cigarette Deliveries: Input to ADL-II Smoking Machines at HL vs Analytical Smoking Machine ADL-II Analytical* Cigarette Code # Puffs TPM, mg/ciq Nicotine, mg/cig # Puffs TPM, mg/cig Nicotine, mg/ciq 90 15.8 ± 1.3 37.5 ± 4.9 1.94 ± 0.23 20.4 ± 1.4 45.2 ± 9.8 2.37 ± 0.47 79 15.6 ± 1.1 41.8 ± 3.5 2.39 ± 0.25 16.8 ± 0.6 57.7 ± 3.2 2.73 ± 0.20 SEB IV (04) 16.8 ± 1.0 47.3 ± 1.6 2.93 ± 0.14 15.7 ± 1.1 56.8 ± 3.0 3.69 ± 0.41 13 , 13.3 ± 1.0 38.2 ± 4.6 0.22 ± 0.01 13.0 ± 0.3 46.7 ± 1.8 0.27 ± 0.03 *Average results for 8 cigarettes per code smoked on the Phipps and Bird smoking machine at 2 puffs/minute, 1 cigarette/pad

Table IV Total Particulate Matter (TPM) and Nicotine Output from Selected ADL-II Smoking Machines in Service at HL, May 21-22, 1977 Cigarette Code Machine Number TPM, mg/ci g Nicotine, mg/cig Nicotine:TPM Ratio 90 14 29.1 1.94 .0667 34 27.1 1.75 .0646 31 27.7 1.75 .0632 10 33.6 2.14 .0637 30 23.8 1.41 .0592 .7 29.7 1.78 .0599 25 26.7 .1.52 .0569 39 22.7 1.52 .0670 Average' 27.6 ± 3.4 1.73 ± 0.24 .0626 ± .0036 79 17 33.2 1.84 .0554 8 31.9 2.04 .0639 24 26.9 1.65 .0613 6 32.7 1.83 .0560 37 29.6 1.70 .0574 36 32.8 1.75 .0534 29 34.3 1.98 .0577 16 34.4 2.11 .0613 Average 32.0 ± 2.5 1.86 ± 0.17 .0583 ± .0036 SEB IV 32 39.1 2.73 .0698 (04) 40 33.6 2.11 .0623 2 38.4- 2.39 .0622 13 36.8 2.34 .0636 Average 37.0 ± 2.4 2.39 ± 0.26 .0646 ± .0035 13 22 31.4 0.20 .0064 11 33.7 0.23 .0068 15 28.7 0.20 .0070 38 30.3 0.20 .0066 Average 31.0 ± 2.1 0.21 ± 0.02 .0067 ± .0003