Tobacco Documents Online

Topical Report NCI/S&HP/ORNL #59 INHALATION BIOASSAY OF TOBACCO SMOKE IN PIGEONS SITE VISIT I TO BETH ISRAEL HOSPITAL (BIH) Boston, Massachusetts, November 30 - December 1, 1977 12-20-77 R. A. Jenkins 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-NIFi"/NCI) No. 40-485-74, Part II Internal (QRNL),Contract Charge No. 3390-0224 Intended for informal communication with project management only. Confidential until published or released by author.

SITE VISIT I TO BIH R. A. Jenkins and B. E. Gill Sun-nar . Data gathered on a first site visit to Beth Israel Hospital suggest that the pigeons in the new 20-bird exposure rack deplete approxi- mately 10; of the smoke offered, per bird. At the proposed exposure protocol of three per day, seven days per week, a nicotine dose level of 8.3 mg per week would be projected for the Code 23 exposure group. It has been recom- mended that the compressor pump supplying the dilution air for the exposure system be replaced with some device which prevents oil fumes from entering the air supply. Introduction. A first visit was made to the pigeon inhalation bioassay being conducted at Beth Israel Hospital's (BIH) Charlesgate Inhalation Facility on November 30 - December 1. Our primary purpose was to meet with the personnel involved, familiarize ourselves with the details of the inhalation study, discuss methods and any potential problem areas, and identify future monitoring needs. In addition, a few smoke samples (both particulate and gas phase) were taken in order to obtain some preliminary data concerning the amount of smoke which the animals retain during exposure. We found the staff, including both BIH personnel and those from the Harvard School of Public Health - who are responsible for the smoke chemistry portion of the study - to be very helpful and cooperative. We discussed at length the exposure system and its concept, in order that all parties involved could more fully understand the advantages and disadvantages of its mode of operation. Staff members aided us in making puffing performance measure- ments on the Lorillard 30-port smoking machine. General Comments. The eventual smoke exposure system to be used will be the Lorillard 30-port smoking machine. The device generates a semi-continuous stream of smoke, which is diluted with 10 volumes of air. The diluted smoking stream will then be distributed between three racks holding twenty birds each on a 20 seconds of 9% smoke`- 40 seconds of air cycle. Each rack contains five channels, each channel with four pigeons. Ideally, the flow through each channel is one-fifth of total output of the machine, or 2.31 liters per minute per channel. BIH personnel have done a very good job designing the pigeon restraint system. Briefly, each bird is positioned in a half-cyclinder and held in

place, with its chest against a foam rubber wedge, by a Velcro-stripped piece of fabric. Its beak is then placed through a hole in a piece of dental dam stretched across a cylindrical port into the smoke exposure channel. The bird's head is immobilized by slipping on a plastic "helmet" which fits over two guide pins. This.restraint method succeeds in main- taining a beak-only exposure, insuring that virtually no smoke particulates deposit in areas other than the target organs. The pressure drop across the exposure rack is very small. Consequently, anything which tends to restrict flow downstream of the point at which the 9% smoke is split five ways, including even an extra coil, in the smoke delivery tubing, would tend to alter the flow of smoke in all five channels. Even measuring the flow in any given channel becomes difficult, since almost anything placed in the smoke path to measure the flow will significantly alter the flow in that channel (and, of course, in all of the other channels in the rack). For example, the "restriction" caused by adding a length of tubing and the "weight" of one bubble in a soap bubble flow meter to the exit port of one channel resulted in a reduction of flow in that channel from 2.3 liter/min (assumed) to 0.8 liter/min. Unless exactly the same "restriction°' is added to all five channels of a rack simultaneously, there is no easy way to determine if the flow in any given channel is near the specified value. If it is assumed that the animals do not deplete a significant portion of the smoke offered to them, then minor fluctuations in flow in the exposure rack channels are not important, since the concentration of smoke to which the animals are exposed is not affected by the flow. The smoke dose which the animals receive would consist of exposure to an essentially fixed con- centration of smoke for a given time. If, however, the animals deplete a significant fraction of smoke, then the dose which they receive will be hiqhlL_qeTendent on the flow rate in the exposure channels and their indi- vidual position in the rack. That is because the concentration of smoke to which any bird downstream of the first bird will be exposed is highly dependent on the amount of smoke deplet`ed upstream. r Enerimental Methods. Puff volumes were determined by placing a soap bubble meter on the exhaust to the distribution manifold, with lighted cigarettes in the smoking ports. Particulate phase and gas phase samples were taken by attaching the inlet of a flow-controlled diaphram pump to the point to be sampled with neoprene tubing. Flow rates into the pump were monitored

by means of a calibrated mass flow meter placed directly upstream of the pump. At no time during the course of the measurements did the flow into the pump vary by more than 1%. Particulate phase samples were collected on a conventional 44-mm Cambridge filter pad placed between the exposure rack and the mass flow meter. Gas phase samples were obtained by collecting the pump effluent in a Saran gas sampling bag, and transferring an aliquot of the gas to a glass gas sampling tube, for analysis at ORNL. Exposure rack channel inlet samples were obtained by sampling a portion of the flow (%1200 riil/min) into each channel. Clearly, in such a pressure- drop-sensitive system, any sampling of a volume of this magnitude would shift the flow in all five channels away from the specified 2.31 liters/minute. Presumably, if the identical sampling rate was applied to all channels, a relative measure of the amount of smoke going into each channel could be obtained. To determine the amount of smoke depleted by the pigeons on any given channel, smoke samples were taken at the downstream end of the exposure channel both with and without birds on the channel. This "differential sampling" acts to compensate for any deposition of smoke in the exposure channel. While the absolute amount of smoke depleted by the animals deter- mined by this method should be accurate, the calculated fraction of smoke depleted will probably be significantly lower than the actual fraction, since a withdrawal of %1200 ml/min at the downstream end of the channel would probably act to increase the total smoke flowing through that particular channel. Observations and Results. The environmental conditions in the exposure room deviated significantly from those of analytical smoking. The relative humidity was determined to be 30-35~'W at 69°F. While no cigarette static burn rate measurements were made, it has been our experience at other inha- lation studies that a substantially reduced humidity acts to increase the static burn rate of the cigarettes and ultimately reduce the amount of main- stream smoke generated by the cigarettes. The BIFI staff are acting to reduce . the impact of low humidity.by storing cigarettes in humidified containers ~ until just prior to their use in exposure. Puff volumes were found to be approximately 10'1t> higher than the speci- fied value, averaging 38.4 ml/puff. The puff volume of the Lorillard smoker is easily adjustable, and efforts should be made to keep the effective puff volume near the specified value.

The el even puff srsioki ng regi men (ten pi us one "l i ghti ng" puff ) used by EIIH results in burning the Code 23 cigarettes routinely past the 23-mm butt line by approximately 12-mm. Smoking the Code 23 cigarette under analytical conditions for the Series IV study resulted in a puff number of 9.5. Thus, a puff number of 9 or 10 might be more appropriate and would help to insure better interlaboratory comparability of the bioassay. Of course, the larger puff volume and lower humidity probably contribute to a reduction in the number of puffs available from the Code 23 cigarette. One of the most potentially serious problems noted during the visit was the strong smell of oil fumes in the line supplying dilution air to the epxosure system. This air is supplied by a compressor pump located in the exposure ward. Since exposure of the experimental animals to a pump oil aerosol would complicate interpretation of the bioassay results, it is strongly recommended that measures be taken to utilize an air supply system which is oil-fume-free such as an air compressor approved for use by humans. Individual values for selected constituents in the smoke being offered the animals are tabulated on Table I. (Values given in tables are means ± standard deviations.) These data suggest that some individual differences may exist among channels in the amount of smoke offered the animals. However, no gross differences exist. (Note that the concentrations listed for the gas phase constituents are volume percentages, averaged over the total expo- sure. Thus, the actual concentration of CO and CO2 in the smoke offered the animals averages 0.36% and 0.90%, respectively. Table II compares the observed concentrations of smoke in the channel input samples with that which would have been predicted from data from PdCI's Fourth Series of Experimental Cigarettes for Code 23. The comparison suggests that the output of the Lorillard exposure system contains less M1, more nicotine, about the same CO but more C02 than that from an analytical smoking machine. The difference in the C0:C0z ratio is usually indicative of an alteration in the.combustion process and may be due to the fact that the Lorillard is a positive pressure puffing system. Table III compares the concentrati'on of several smoke constituents at the downstream end of one exposure channel, averaged over several runs, both' with and without birds being exposed. Clearly, the birds have a marked ~ effect on the concentration of smoke constituents in the exposure channel, C~ even at relatively high flow rates. Respiration adds substantial amounts ~ of C®, and water vapor to the exposure atmosphere. This is exhibited by the W

dramatic increase in C02 concentration and weight change of the filter pad (TPM). Nicotine and carbon monoxide are substantially depleted from the smoke. To a rough approximation, the average smoke depletion per bird is 10/00 (Note that this fraction depleted may be somewhat lower than that during routine exposure, since the 1.21 1/min sampling rate downstream of the exposure channel may be expected to increase the total amount of smoke available in that channel.) Assuming that each animal depletes an average of 12% of the nicotine offered per exposure, based on Series IV values for Code 23 nicotine delivery per cigarette and 21 exposures per week, the projected nicotine dose for each pigeon would be 8.3 mg nicotine per week. This is similar the nicotine dose which the rats in the high nicotine dose group in another NCI bioassay are offered on a weekly basis. The level of smoke depletion by the pigeons has some important ramifi- cations for the bioassay. The increased level of C02 in the downstream smoke will tend to stimulate respiration in the downstream animals. However, the downstream animals will be breathing smoke that is significantly depleted in both gas and particulate phase constituents. It is important that animals be rotated through the various exposure rack positions, so that an "average" exposure is obtained. BIH staff intend to do this. Furthermore, care should be taken to insure that nothing significantly alters the pressure drop in any one channel, as such could cause a marked change in the smoke dose which the animals retain.

TABLE I Smoke Constituent Concentration in Channel Input Samples Code 23 Cigarettes Channel No. TPMa, mqfpad Nicotine, mgfpad % cob %CO~b 1 26.9 2.30 0.12 0.30 2 26.5 1.75 0.09 0.26 3 32.9 2.27 0.12 0.32 4 31.3 2.06 0.12 0.30 5 32.9 2.48 0.12 0.31 aTotal Particulate Matter, by weight. bConcentration of CO/CO2 is by volume %. Note that this is the concentration averaged over the total exposure period. The con- centration in the smoke stream should be greater by a factor of three.

TABLE II Channel Input Samples Comparison of Observed and Predicted Smoke Deliveries Code 23 TPA1, m /2ad Ni coti ne, mg/pad C0, ml /sampl e C®,, ml Asampl e CO: C0, Rati o . _ Observeda 30.1 ± 3.2 2.17 ± 0.28 14.8 ± 1.7 38.9 ± 3.3 .38 Predi ctedb 36.8 ± 1.0 1.72 ± .05 15.8 ± 1.3 32.1 ± 1.5 .49 aSampling flow rate: 1210 mijmin, or 10.5% of total smoke output of Lorillard smoker. bPrediction based on Series IV per cigarette deliveries, collecting 10.5% of a ten cigarette sample.

TABLE III Depletion of Smoke Constituents in Exposure Channel Code 23 Candi ti on TPM, mralpad Ni coti ne, mg/pad % CO % CO2-_ Wi th©ut Eii rds 30.1 ± 2.3 2.73 ± 0.14 .128 ± .005 .313 ± .005 Wi th Bi rds 71.2 ± 18.4 1.38 ± 0.10 .083 ± .006 1.79 ± 0.24 Percent Change +137 -49 -35 +472