Tobacco Documents Online

289 II. INSTRUMENTAL APPROACHES TO BIOASSAY MONITORING T. M. Gayle, R. A.-Jenkins, and B. E. Gill Introduction. In order to better define the exposure in a chronic in- halation bioassay, it is necessary to obtain data more frequently than is now possible with occasional visits by ORNL personnel. While techniques developed by ORNL to define the exposures are available to all bioassay laboratories, the techniques cannot always be cost-effectively applied, even if laboratories have the facilities to perform the chemical analyses. One solution to this problem which we have been exploring is the use of monitoring instrumentation to obtain data for routine characterization of the exposures. Ideally, such instrumentation could be used easily by non- professional bioassay laboratory staff, with only a minimum or no chemical analyses. This year, we report on progress made in the application of such instrumentation, both for routine characterization of the exposures, and as a non-invasive dosimeter to be used in single animal (dog) inhalation bio- assays. Routine Monitarin~ Instrumentation. Developed under a contract with the Council for Tobacco Research USA, Inc., the optical particulate sensor has been described in detail in previous Progress Reports. Briefly, it consists of a combination light emitting diode and phototransistor. The surface area of the two units is about 10 smn2. Infrared light from the LED is backscattered from the smoke aerosol and strikes the phototransistor. The level of backscattered light is directly proportional to the aerosol concentration. The phototransistor is part of a voltage divider circuit; and the resulting voltage change is amplified to be registered as an instan- taneous smoke concentration. The integrated output of the particulate sensor

290 is proportional to the amount of particulate matter flowing past it in a given time at a constant flow. - For field testing at the bioassay laboratories, a prototype portable monitoring unit was constructed at ORNL. The particulate sensor itself is mounted inside a small tube which can be affixed to the end of the cannula of a dog exposure system. While the exposure system is operating (without the dog being present), smoke is withdrawn from the cannula past the sensor at constant flow with a small vacuum pump built into the electronics package. Particulates are collected on a filter pad mounted imnediatel,}P downstream of the sensor for purposes of calibration. Evaluation under carefully controlled conditions at ORNL indicated that integrated sensor response was directly proportional to the weight of TPM or amount of nicotine collected to within cigarette-to-cigarette variability, and was independent of cigarette type. Under field trials at the VA Hospital and Borriston Research Laboratories, it was shown that TPM or nicotine delivery could be accurately measured by the particulate sensor with only a daily four-point calibration of the sys- tem. That is, the particulate sensor could be used by bioassay personnel to document exposure system performance by attaching the system to the cannula exit and recording the integrated output of the system with only minimal calibration. This suggested that the smoke sensor could respond linearly to instantaneous smoke concentration under the less rigorously controlled con- ditions of the bioassay laboratory, and thus might be promising as part of a non-invasive particulate dosimeter (see below). For chamber exposure of several anamals to tobacco smoke, such as those used in the ORNL and Battelle'PNL rat inhalation bioassays, it has been necessary to remove small "grab" samples of the chamber atmosphere both early and late in the 30-second smoke exposure cycle in order to determine the amount of smoke particulates depleted by the animals. The difference in

291 the amount of nicotine collected on a small Cambridge.filter from a sample obtained immediately after smoke en•ters the chamber and that in a sample taken immediately prior to the time when smoke is flushed from the chamber is taken as an indication of the fraction of total available smoke inhaled by the animals. Such nicotine chamber depletion measurements require chem- ical analysis of the necessarily small amount of nicotine collected on both filter pads. A modified CTR-ORNL sensor package was tested, both at ORNL and at the Battelle PNL bioassay, to determine if an instrumental approach could be used to document the fraction of smoke in the chamber which is depleted by the animals. Briefly,the sensor is mounted on the end of a wand which fits into a small hole on the faceplate of the exposure chamber. The sensor thus moni- tors the instantaneous smoke concentration in the chamber. (A hood is placed over the exposure chamber and animal containment tubes to reduce interference from ambient light.) Figure 11-5 is taken from a typical con- tinuous trace of the system output. As smoke is generated and enters the chamber, sensor response rises above baseline. (Occasional spikes result from initially incomplete mixing of the chamber contents.) As the animals inhale the smoke, concentration of the smoke particulates in the chamber decreases. Finally, at 30 seconds into the cycle, fresh air flushes the smoke from the chamber and the system response returns to baseline. The sum of the absolute reduction in response divided by the sum of the initial responses over all of the cigarette puffs is taken as the fraction of avail- able smoke inhaled by the animals. Comparison of conventional nicotine grab samples with the depletian as measured by the particulate sensor (Table 11-8) suggests that the sensor measurement has a greater precision. However, the sensor response suggests 'that the depleted fraction is somewhat smaller than that calculated from grab sampling. This is probably because

0.81 PARTICULATE 0.52 CONCE ~TRATION (/.Lg nicotine/ml) 0.26 FIGURE 11-5 ORNL-DWG 78-9010 SMOKE CONCENTRATION IN EXPOSURE CHAMBER MADDOX-ORNL RAT EXPOSURE SYSTEM CODE 04-BNW SITE VISIT II t . 9 8 7 TIME (min) 2 1 0 8b8a,4468

293 TABLE 11-8 Comparison of Exposure Ch-amber Smoke Depletion by Rats: Nicotine Grab Samples vs Optical Particulate Sensor BNW Site Visit II Percent Depletion Cigarette Code Exposure Group No. Calculated from Nicotine A of Chamber Grab Sample Calculated from Optical nalysis Particulate Sensor s Response SEB IV (04) 1 2 3 76 69 66 61 59 56 Average 70 ± 5 59 ± 3 27 1 58 53 2 65 62 3 71 58 Average 65 ± 7 58 ± 5 13 1 42 64 2 66 58 3 40 53 Average 49 ± 14 58 ± 6 . r

294 the grab sample values are not corrected for the °chamber depletion" which is a result of the sampling itself. (The early-3-second-20 ml grab sample can potentially remove as much as 6% of the smoke in the chamber. The importance of continuous chamber monitoring is also illustrated by Figure 11-5. First, note that for the later puffs of the cigarette there is a much greater initial concentration of smoke particulates per unit volume. Secondly, it is clear that the animals deplete a much greater fraction of the later puffs of the cigarette. These findings could have important ram- ifications for assessment of the chemistry of the smoke insult to the animal, since the data suggest that the animals receive most of their smoke dose from the last few puffs of the cigarette. Particulate Sensor as a Non-Invasive Dosimeter. Presently, there are two primary methods for estimating the amount of smoke deposited during individual exposures of experimental animals. Carbon-14 tracer studies can be employed following the smoking of a radiolabeled cigarette. Dose measurement requires sacrifice of the animal. Furthermore, in experiments in which significant changes in pulmonary function may be expected, sacrificial dosimetry yields no information concerning changes in smoking behavior during the course of the experiment. The other technqiue, lung lavage, is non-sacrificial, but neither of these techniques are suited for routine use in a chronic expo- sure. Biochemical markers in physiological fluids may appear promising, but such capabilities are still in the early stages of development. The work reported on here is directed toward instrumental approaches to the determin- ation of the quantity of smoke inhaled ~and ultimately, retained) under chronic exposure conditions. 'Since response of the particulate sensor is linear with smoke particulate concentration, it was speculated that multi- plication of the concentration signal and a linear flow signal, followed by

295 integration, would result in a system output which would be proportional to the total amount of particulates pa-ssing by a fixed point, irregardless of flow rate. Thus, with a smoke sensor installed at the entrance to the tracheal cannula and the flow sensor installed at the inspiration valve at the head of the stand tube of an ADL/II exposure system, the total amount of particulates inhaled by the animal would be proportional to the integrated signal. For such a system to be feasible, the smoke sensor must respond to rapid changes in smoke concentration and the responses of both flow and sensor systems must be "i n phase°" . To determine whether the smoke sensor would respond sufficiently rapidly to abrupt changes in particulate concentrations resulting from changes in animal breathing patterns, the portable, prototype sensor system described above was employed at two of the beagle dog bioassays with the sensor mounted at the entrance to the cannula while dogs were being exposed to cigarette smoke. Since the response of the sensor is proportional to concentration, output of the system should be a function of the smoke concentration at the junction of the stand tube and the cannula. Thus, as smoke is'introduced to the head of the stand tube, animal breathing brings the smoke into the sensor area, and 'the response increases. Between breaths, if the animal has not completely cleared the smoke from the stand tube, the sensor output will re- main relatively constant and greater than zero. As the animal clears the smoke, response should return to baseline. Sensor response traces depicting smoke withdrawal from the stand tube for both cuffed and uncuffed cannulas,are compared on Figure 11-6. Clearly, the sensor does respond rapidly to abrupt changes in smoke concentration. The responses indicate that there are important differences in smoke with- drawal patterns, depending on cannula design. For the cuffed cannula, the

296 FIGURE 11-6 ORNL-DWG 78-9017 SMOKE CONCENTRATION IN STAND TUBE SENSOR LOCATED DIRECTLY UPSTREAM OF TRACHEAL CANNULA I I I SMOKE CONCENTRATION C q Lj 4 2 UNCUFFED CANNULA [CODE LN-VAH SITE VISITVII] k / I u 1 I I SMOKE CONCENTRATION 1 w LI 3 TIME (rnin) I ~-----~ 4 3 2 1 TIME (srsin) ,• CUFFED CANNULA [CODE 13-BRL SITE VISIT II] V I U

297 animal's upper respiratory tract is sealed off and thus the animal must obtain all of its breathing air through the stand tube. Thus, the animal clears the smoke quickly (a fraction of the 30 seconds between puffs). With the uncuffed cannula, the animal can inhale through the upper respiratory tract and around 'the cannula and thus, is not forced to inhale the smoke. Because of this, smoke is not quickly cleared from the stand tube. In some cases, the next puff (20-second puffing cycle) was responsible for expelling the smoke from the stand tube. Since the smoke remains in a concentrated bolus in the stand tube for quite some time, the particle size of the aerosol increases. This, in turn, suggests that lung deposition sites for smoke particulates may be different for the two types of cannulas. To determine if both smoke and flow sensor systems are adequately "°in phase" to perform analog multiplication and integration in real time, an experimental assembly of available components was constructed. Tests were performed with a large animal respirator and a manually operated syringe to simulate regular and irregular rapid breathing patterns in dogs. Figure 11-7 is a simplified block diagram of the experimental assembly. An ADL/II expo- sure system was used in the conventional manner to generate and deliver smoke. The particulate sensor was located in a small tube between the end of the stand tube and the entrance to the cannula. A pneumotachograph flowmeter with a maximum capacity of 60 liter/minute was installed in place of the inspiration valve on the front of the ADL/II such that all air inhaled by the animal (simulated) is routed through the pneumotach. The pneumotach is a flowmeter of the laminar type, which produces a small differential pressure drop which is linear with f1ow rate. The pressure drop at full rated flow rate (60 1/min) is only 10 mm H20. A differential pressure transducer with power supply-demodulator transforms the pressure drop into a linear d.c.

298 FIGURE II-7 OANN11>{ - t/.QL'r'b f,+sSE,R9eC.7" RES PlPATO R C314' SWtKW: X TsrrKMAI-P- ~ +4Kth!A ///6, i cC %'l<fi`1FR a.~~.._._.._ 4iVuT C?-EO`l '~ CdLL~~Cri~? I •~GUAi. Ta:ACC ~ A N A fNPu'r 0 "' ! C) V - - ----~ ~ E c~u ~ ~ur i ~----....-..--~- v ®____t..- VF Yy AVSK I ~c-rc~atE~~ ~ Gt1'1 T°6+t. t nYTE.r` &:IRTn ~ ~ (7 C.'r:rT M6'Gi'() r r--~ C v pt~tr,yl 6 ; VOt<METEff ~ Co --- h'bN i'l V.=1 .~7`:rw~ ExPEV+MEN r..t, ass~atc_7 G~ ~ 7) C1'! ~