Tobacco Documents Online

f Our Re£~IMI46M ° Mr. M.L. Reynolds, ..- , Brown & Wiliiamson Tobacco Corporation, P.O. Box 35090, . . Louisville, Kentucky 40232, U.S.A. ;....°~r.. • • 15thJune, 1988 Deaf La;IC*, I was asked sometime ago :o comment on the draft paper submitted by Professor Cain to B.&W. relating to the Capri study he umdertook for you earlier taxis year. For your information, a copy of my originat views ate em¢Iosed. As a result of this, Alan Heard asked me to elaborate further oa one of the poin~ (No. 2) of this note. Essentially, my comments arose from looking at Table VII from Cain's drafter paper (whicl~ is also appended). By emamirdng the dif:f~tnces in prtxtuct types in terms of both the weights bm'nt and analytes me.astmut, a pattern should emerge m terms flint the % weight saving of Capri relative to the other products should be reflected in the % reductions of the analytcs momtored for (this is expressed in the table below). ,% REDUCTIONS BETWEEN PRODUCT TYPES Parameter "Weight Burnt CO Hydrocarbons NO, Particles (I) Particles (2) DoralLights - Ca.p.ri % Doral Liglus 34 23 39 0 11 *-3 Marlboro-Capri % Marlboro 36 26 41 14 19 14 VirginiaSlims - Capri % VirginiaSlims 34 2O 24 8 13 16 * NOTE: -re indicates Capri is higher than Dotal I../ghtsf C3 0 (.ra d~ BATCo document for Legal Services • Health Canada 20 May 1999

This spread of remits is somewhat largo and b probably du¢ in part to the malytical difficulties associated with EI'S Cain then gem on in the paper to discuss on the basis of Table VII that Cagd has the kighest ETS particulate (TSP) potential per gram of product burnt, and this is true for scnm cf the other analytes as well (This fact is supported by our own sidemtam ETS studies). I lowever, on balance within the paper, I feel that to try and put the whole subject into context, it would be advisable for someone to commettt on the variability of the data that can be derived from Table VIL point out that Capri produces the lowest absolute ETS leveLs and discu~ the ~ part/~ate (TSP) imtemia!" of Capri and tim other products. • -..:.....~. t-. X~:. 't : ~ :e~_'t?t~.!~"~i.. ~:::.t."~' : " '" I trust these vitws wiU heIp peoples thinking towards this pape~ by Cain. if you have any f'mther ~, please do not hesitate to contact me. I remain. Yours sincerely, P.D. ~IlCS. c.c. Mr. A.L Heard N.O.O. Dr. T. Hirji Dr. R. Binns Mr. ff.A. Lake O O iX.) "',4 O', ¢..n BATCo document for Legal Services • Health Canada 20 May 1999

..~ ° TO REF BAT [U.K.andExport] Limited RESEARCH & DEVELOPMENT CENTFIt smrrl~M~Cm Hr. R. P. Ferris Mr. A. L. Heard PDC/MM/46M FROM DATE P. D. Case 19th April, 1988 f" ,,..o COMMENTS ON THE PAPER BY PROFESSOR W. CAZN:- PSYCHOLOGICAL AND PHYSICAL MEASURES OF ETS FOR FOUR BRANDS OF CIGARETTES Further to my involvement with the Ca~orl exercise under-taken by R&D in 19B7 and discussions with Professor Cain in January 1988, I have been requested by R. P. Ferris to comment on the above paper. Basically, I feel some of the data Incorporated In the paper can be used but its current format does give rise to r~orne concerns. Essentially, the concerns reduce to a very simple argument i.e. do we work on a cigarette basis (in which Capri wins hands down) or should the discussion relate to a tobacco weight burnt basis, in which case, Capri dos not necessarily win! The paper currently works In both area~ The major areas of Interest are listed below:- 1. The statement, "It would aDpeaJ" then that to a first approximation amount of tobacco burned can predict the odour and irritation of ETS irrespective of whether the tobacco comes from a conventional, slim, or ultrasiim cigarette* needs to be eliminated. This statement Is almost in direct contradiction to earlier comments relating to "the departure from paraJlelism caused by Marlboro" I.e. comments made in dose response relationship terms. EssentiaJJy, I feel Professor Cain is claiming slmillar dose response curves for the 4 cigarette types (can he be so bold when only two points on the dos~ response curve exist for each product l:ype)? Examination of figures 27 to 45 does not fully support this. Additionally, in many of these figures for an equal cluantlty of tobacco burnt or an equal C~ concentration Capri often produces the highest objective score! This can of course, be counter-argued by saying one has to burn 15 Ca~ori cigarettes to generate this dose rather than 10 of the other product types, hence showing the potential advantage to Capr~. My G Lrl U1 P~J C~ G~ _.° BATCo document for Legal Services • Health Canada 20 May 1999

o. • '.":i point Is why bother getting Into an aJ'gumentJcounter arguement like this In the first place 1.¢ Keep it on a per cigarette basis in the first instencei 2. In the .objective data, the X weight saving results of Capri relative to the other products, Is not reflected in the ~; reductions of the indlviduaJ components monitored in ETS. 3= Table 6 Is of some concern. On going from smoking 10 cigarettes to 15 cigarettes, the CO concentration should increase by a factor of 1.5; apart from two of the examples, the concentration rises by more than 1.5. 4. Professor Cain in the discussion and flnaJ two pages of text Illustrates that Capri has the highest ETS Tar potential per gram of tobacco burnt. This is true, bearing In mind point 2 above, and would be extended to other analytes. This is true from our own data but does It need pointing out? AddltlonaJly, Cain states Capri Is the slowest burning product, without evidence to suppor~ this, and in SBR terms, this Is not the case. 5. Returning to the subjective attributes, figures ? to 10 show Increases in subjective Intensity for various parameters assessed. No comment Is made as to whether this has any bearing onthe order In which the paneJiists make the assessments of these parameters. 61 Finally, part of the process Involving Professor Cain was to establish if results on the same products examined in R&D translate across to his laboratory. Thus we are not helped very much by R&D examining Marlboro Box a 79ram product in detail and Cain examining an 85ram Marlboroi In summary, I do not wish to appear to being negatively disposed towards this paper, the finding on Capri having the lowest subjective and objective results are correct and need to be highlighted. However, some of the other factors currently incorporated within the paper if published, could be potentially turned agaJnst Cain (and hence us) by other parties. Consequently, I feel we need Initially to rationailse the~e concepts amongst ourselves before deciding on the next phase of the operatJon which should involve further discussions, with Professor Cain, who we clearly do not wa~t to antagonise. Before discussion with Cain, we need to have addressed s couple of questions internally: (1) Do we actively encourage Cain to change some of the format of the paper, bearing in mind we have a "hands off" policy towards third party consultants? 0 0 Q;I r~O C~ BATCo document for Legal Services • Health Canada 20 May 1999

• .o- (Z) P. D. Case • , ...~...;,~ ~ -.. Cain is starting to underst~und the weight argument towards sidestream reduction and Is elucidating such in this p&oer. To help him in this, should we be more open In our findings in relation to weight end sidestreaan yields, and hence suggest to him, there is no need for him to 9o into this ¢letaJ| in his paper? : • J . :. . . ~T' w - d -t; : . ~'~ • :., o ' ['"": c.c, Dr. R. Binns Dr. T. Hirji Hr. J.A. Luke °'" b O C) U7 --,.j O~ Co BATCo document for Legal Services Health Canada 20 May 1999

°. • ,°, ," ,. ,~:.-. ~ ~i• Table VII. Comparison of Brands Brand: Capri ..: •... • " ~. -:°-°4 . • Doral Lights i Marlboro Va Slims Lights n= 4 3 Burned (g) 4.86±0.16 7.31+0.16 4 7.55±0.O8 3 7.36+_0.64 Gases: Dec Feb Dec Feb n- 2 2 1 " 2 CO (ppm ±0.3) 4.3 4.2 5.8 5.4 HC (ppm .+. 1.0) 4.7 3.5 7.1 6.2 NOx (ppm ±0.01) 0.12 0.12 0.12 0.12 Nicotine (ug/m3) 74±7 est 89±4 Dec Feb Dec Feb 2 2 1 2 6.0 5.5 7.7 6.2 0.15 0.13 73±6 5.4 5.4 5.5 5.3 0.13 0.13 89±7 Particles: n,= 4 3 TSP (ug/m3,Dec) 872±32 976±0 (ug/m3,Feb) 1078+25 1047±114 (ug/g-m3,Dec} 184±5 133 (ug/g-m3,Feb) 217±3 155±1 4 1077±167 1250±71 143±5 166±2 : 3 1 ooo±I 1278±91 144 169±1 0 0 ~n ,43 .. .,. • .. • , • ;" _ BATCo document for Legal Services • Health Canada 20 May 1999

-DRAFT- Psychophysical and Physical Measures of ETS for Four Brands of Cigarettes Submitted by William S. Cain, Ph.D. Pierce Lab / Yale University 290 Congress Avenue New Haven, CT 06519 Submitted to Dr. William H. Deines B&W B & W Tower Louisville, KY 40202 Date April 7, 1988 i s• O O Lr~ Lrl Po BATCo document for Legal Services • Health Canada 20 May 1999

,3 Part 1: Psychophysicai Measurements INTRODUC~ON The purpose was to chart odor and irritation from ETS produced by cigarettes that varied in diameter: two of conventional diameter, one of slim diameter, and one of ultraslim diameter. METHOD .~: Thirty-five subjects, 18 males and 17 females, participated in four sessions each, one for each of four brands Average age equalled 28 (SD = 8.1). Twenty percent smoked. The majority had sewed in similar experiments previously. On any of the eight days of testing, approximately half the subjects participated. ~: Brands studied were Capri (97 mm length, 17 mm circumference), Doral Lights (85 ram, 24.5 ram), Marlboro (85 ram, 24.5 ram), and Virginia Slims Lights (98 ram, 23 ram), all nonmenthol. The cigarettes were maintained in unopened packs under normal room conditions, temperature of 21 C and relative humidity of 30 to 40 %. Packs of a given brand were opened as needed on the day of smoking (one brand per day) and the contents distributed into aluminum foil packets in the following fashion: two packets of 10 cigarettes and two of 15 cigarettes. The quantities per packet represented hourly amounts to be smoked during four-hr test sessions: 10 cig per hr for the first and second hours and 15 per hr for the third and fourth hours. The appropriate aluminum-foil packet was delivered to the smokers at the beginning of each hour. After it was smoked, a cigarette was extinguished by being placed into a scintillation vial which was then sealed. (The vial also served as ashtray.) The collection procedure permitted weighing and measurement of butts for subsequent correlation between level of sensory effect and amount of tobacco consumed. ~: Smoking took place in a 1200-cu if, aluminum environmental chamber. Relevant parameters were as follows: temperature, 22 C; relative humidity, 36% (SD=6); recirculation rate, O U'l "~1 j~ BATCo document for Legal Services • Health Canada 20 May 1999

.o 95 air changes per hr (ach); and ventilation rate, 2.5 ach. Temperature, dew point, and carbon monoxide concentration were monitored throughout a session. ~: ETS was generated by three or four smokers. These persons smoked in serial order, i.e., person A smoked a cigarette, then person B smoked a cigarette, etc. On any given day, the smokers smoked onJy one brand. Table 1 displays the order of brands smoked across days. Table 1: Schedule of Sessions Date Brand Date Brand December 28 Marlboro January 4 DoraI Lights December 29 Doral Lights January 5 Marlboro December 30 Capri January 6 Va Slims Lights December 31 Va Slims Lights January 7 Capri The smoking procedure remained the same from day to .day, except for brand smoked. During the first two hours, the smokers smoked a total of 10 cig per hour. A new cigarette was [it every 6 min and was smoked for 6 min. Hence, one cigarette was being smoked at any given time. Because the smokers had considerable prior experience with such a regimen, they followed it with ease and smoked the cigarettes comfortably, neither very fast nor very slow. During the first hour, carbon monoxide rose to a more-or-less steady level. During the second hour, level remained approximately steady and the first set of psychophysical judgmnents were collected. During the third and fourth hours, the smokers increased the smoking rate to 15 per hr. This regimen required a new cigarette to be lit every 4 rain and to remain lit for approximately 6 rain. After the first four rain, there were, therefore, alternating 2-rain periods of one cigarette being smoked and two cigarettes being smoked. The third hour was used to approach a steady level of carbon monoxide. Psychophysical judgments were collected during the fourth hour. ====,b O O U-1 U-1 p,,.) "--4 u'l BATCo document for Legal Services • Health Canada 20 May 1999

Before beginning their psychophysical judgments, subjects were grouped into threes for their visits into the chamber. A trio would enter for a 3-rain period. At t = 0, 1, 2, and 3 rain, they filled in judgment sheets of the sort shown in Fig. 1. The sheets requested judgments of both the intensity and the acceptability of eye, nose, and throat irritation, and of odor. The sheets also elicited a judgment of overall acceptability. After three minutes in the chamber, one trio would leave and be replaced by the next. The entire group would complete a round of judgments in 20 rain. There were three such rounds per hr in the second and fourth hours of the session. Hence, each individual occupied the chamber for a total of 9 min per hr. Data Analysis: Perceived intensity was derived from graphic ratings by measuring the distance in rnm from the top (zero) end of the scale. Corresponding judgments of an attribute within a round, e.g., judgments at t = 0, were treated as replicates and hence averaged across subjects. For reference in interpretation, we note that the tics marks on the graphic rating scale occurred at 13-ram intervals (see Fig. 1). Therefore, averages of 13-ram would represent 'slight,' whereas those of 26 would represent 'moderate,' and those of 39 'strong .' Acceptability scores are expressed in terms of the fraction of subjects dissatisfied. The measure offers the most direct way to compare the acceptability results with perceived intensity. RESULTS Intensity of irritation and odor increased significantly from a smoking rate of 10 to a rate of 15 cig per hr (Figs. 2 to 5). (Tables 2 to 5 display ANOVAs relevant to the data in Figs. 2 to 5.) For three of the brands, the increase occurred in more-or-less parallel fashion. Marlboro, however, exhibited a more dramatic increase than the others. Whereas its intensity fell among that of the other brancs at the smoking rate of 10 cig per hr, it climbed above the others at 15 cig per hr. The departure from paralellism caused by Marlboro presumably accounts for statistically significant or near- significant brand-by-smoking rate interactions for the three of the four sensory attributes. The brands also differed significantly in the magnitude of their sensory effects irrespective of smoking rate. The pattern 0 O "-.4 BATCo document for Legal Services • Health Canada 20 May 1999