DRITI S11-A,%11-"R1 CAN TOBACCO COMPANY LTD.
PROJECT JANUS
PROGRESS REPORT: JANUARY-APRIL 1971
lo-5-71
AUTHOR: S.R. Evelyn
DISTRIBUTION:
A.D. McCormick, Esq. Copy No. 1
Six Charles Ellis 2
Dr. S.J. Green 3. 4. 5
Dr. I.W. Hughes 6
Dr. R.A. Sanford 7
Mr. R.S. Wade 81 9
R&D Manager, B.T.Co. 10
Herrn H. Sottorf 11
Dr. D.G. Felton 12
Dr. F. Haslam 13 COPY NO: XF
00
C-0-i
BATCo document for Province of British Columbia 5
November 1999


PROJEICT JANUS
PFZ0(.;Rj-SS TU-111OUT: JANU:II?Y-AI'I?JJ, 1971
LONG-TERM EXPERTMENTS
Experiments BO-135
All these early experimonts have terminated and a
report has
been written surveying the results (RD 773-R).
Because of the
high-dose anomaly, a method was developed so that
the experiments
could be reconstructed. To do this a nimnber of
important assump-
tions had to be made and these are given in detall
in the report.
Based on a preliminary reconstruction of the data,
anote
outlining the main conclusions and giving the
derived tumorigenic
ratios was presented to the Montreal Research
Conference, in
October 1970. In this preliminary analysis, an
age-standardisation
approach was used and the curves were constricter!
to pass through
the results obtained for the 25 nig dose level. In
the revised
analysis nov reported, a Weibull distribution
function has been
used for the reconstruction which is not
constricted to pass
through the 25 mg result.
As a result, if iz is agreed that the many
assumptions are
valid, the revised tumoriaenic ratios rank these
experiments as
follows:
B3 < B2 < Do < B6
~B4 ~137
The < Indicates significance at the 9A- level.
This ranking
is identical to that given in the 1)3'Climinary
note but the improved
BATCO document for Province of BritiSh Columbia 5
November 1999


analysis indicates that the condCrIS;Lte from
samplc~ 132 is
significantly lower than that of' 130.
Experiments 1!6 and B7 (Yeast trc-zillment cuid
ep-L)
The latest results for this experiment are given
in Table 1.
There is very little chanGe in this experiment
and, with only a
few mice left alive, the -final results will be
similar to those
given in Table 1. The high dose anomaly is very
evident in all
groups but it is apparent that yeast treatment
does not lead to a
change in the tumorigenicity of the condensates.
The experiment
is continuing so that any possible differences due
to strand
width can be examined.
Experiment B8 (X-PCL-5 and 1-308)
The results, given in Table 2, show the position
at week 76.
It is apparent that for B8 samples 1, 4v 5 and 6,
the predicted
values are stabilising although there are still an
appreciable
number of healthy mice at the low dose levels.
Comparison of B8-5
(100% 1-308) with B8-1 (control blend) indicates
that the predicted
activity of the condensates is similar. It is of
interest that-the
result for B8-4, containing 60% 1-308, is also
very similar. At
this stage the activity of the condensate from
B8-6 would indicate
that the specific activity is increased by use of
a silica gel
filter.
The results for the samples manufactured from the
experimental
X-PCL-5 (B8-2 and B8-3) show that there is still a
large proportion
of healthy mice alive bUt, compared with
tho-colitrol, (BS-1), th.
activity of the condcji.-:ate is low. In adlition
a araded response
CD
from B8-3 to B8-2 and D8-1 appears to be
developing. CzD
cc
BATCo document for Province of British Columbia 5
November 1999


TA13LE 1
EXPERIMENTS 136/137 ( 128 WE-EKS)
Real thy Tumour Rate
Mice
N INO Obs Std P,.di., ted
M PSR
B6-1 Yeast treated 30 cpi
25 ing 1 57 58 6o
50 nig 0 39 82 83
75 mg 0 10 71 80
B6-2. Yeast treated 60 cp2'.
25 nig 0 56 63 64
50 mg 0 41 79 80
75 mg 0 25 74 79
B6-3 Yeast treated 120 cpi
25 mg 0 51, 55 55
50 mg 0 50 81 81
75 mg 0 39 82 82
B7-1 Control lamina 30 CPi
25 mg 0 61 63 64
50 mg 0 51 84 84
75 mg 0 38 78 81
B7-2 Control lamina 60 cpi
25 Mg 0 58 62 62
50 mg 0 .57 80 82
75 mg 0 42 82 85
B7-3 Control lamina 120 cpi
25 mg 1 1 52 56 57
50 mg 0 64 86 86
75 nig 0 52 811 811
BATCo document for Province of BritiSh Columbia 5
November 1999


TA 111, 11, 2
BXl'j;1lzjmrN'r ns (WJIEA~ 70
I
Dose flea 1 thy i Tui-our Ra to
]Level
CL
N INO
bs
td
redicted
PSR (%) j
D8-1 Cortrol Ito 37 11h Ij 6 74
50/50, lamina/CRS 60 22 6o 57 80
B8-2 X-PCL-5 6od
70 Ito 63 18 18 1 47
20/20, lamina/CRS 60 60 26 25 56
B8-3 X-PCL-5 100% Ito 80 8 8 26
6 o 59 19 20 49
118-4 1-308 6o% Ito 39 41 42 71
20/20, lamina/CRS 60 20 57 63 81
B8
-5
1-308 100%
4 o
45
33
44
73
60 19 65 67 84
-6
8
B
ilica gel filter
0
2
9
4
86
Control tobacco 6o 5 78 86 91
Experiment B9 (Etha,iol extraction, Gerlach)
In this experiment the effects of ethanol
extraction, the
Gerlach additives and the Gerlach process are
being examined. To
week 32 very few animals have died and only five
tumours have
appeared.
Experiment B10 (PCL, non-combustible fillers)
The experiment designed to compare PCL and two
Schweitzer
shect samples with the control. was cancelled
becauso of iiiexplicablY
hi& levels of nitrate, and hiGh deliveries of
oxides of nitroacii
c::)
from cigarettes made from the Schweitzer samples.
The oxporistionL Qz~
"C
100
C-14
BATCo document for Province of BritiSh Columbia 5
November 1999


is now desi&med to dotermii1c:
(a) Whether the incorporation of
iioii-cutnbu.~:tible fiilcr (N(;F) at
the 25% or 50% level will lead to an increase in
tumoricenicity
of the condensate.
(b) Whether the use of appreciable amounts of
binder (sodium
carboxymethyl cellulose) would lead to differences
in activity.
(c) The ttutiorigenicity of PCL condensate
relative to that of the
control tobacco blend manufactured from 60%
lamina, 40% stem
in a single experiment using the new strain of
mice (Carworth
Europe Ltd.).
The samples for this experiment are as follows:
B10-1 Control tobacco, lamina 60%, CRS 40%
B10-2 Canadian PCL (100%)
B10-3 NCF 25%v lamina 45%, CRS 30%
B10-4 NCF 50%, lamina 30%, CRS 20%
B10-5 CF 50%, lamina 30%, CRS 20%
(NCF contains 12% sodium carboxymethyl cellulose,
CF contains 25%.)
Condensate froin the control ciGarette will be
applied at five
dose levels from 14.1 mg to 56.6 mg, so that a
dose response curve
can be established for Carworth mice. Other
condensates will be
applied at three dose levels only. Initially, it
was intended to
use the three middle dose levels, but it has now
been decided to
use the three higher dose levels, i.e. 28.3 mg,
40.0 me and 56.6 mg.
This decision has been taken on the basis of a
preliminary toxicity
trial in which 50 mice were treated with
condenzate (1310-.',) at the
highest dose level for 3-4 weeks. Although the
condensate had a
depressant effect on the activity of the mice, no
mico died durinG
this period and the higher dose level. is being-
used to increase
BATCo document for Province of British Columbia 5
November 1999


the sensitivity of tIi( exi-ye"i'lle"t.
Animal Husbandry
Following the occurrence of chronic respiratory
disease in a
shipment of mici to Harrogate, which ne cessitated
the sacrifice
of 2t5OO mice, special checks have beon made at
Frankfurt;.
Chronic respiratory disease (CRD), which results
froin a virus
infection and/or an infection by iiii crop la sma,
causes the early
death of mice. This may occur after, say, twelve
months and coin-
pletely in validate a long-terni experiment.
Preliminary health checks have been undertaken on
recent ship-
ments of mice from CarwGrth. Samples taken froin
these small
hyperplasia shipments have proved negative in
terins of CRD lesions
despite the use of repeated stress conditions in
the forni of the
"cold sliock". A few mice with liver lesions have
been noted.
The large shipment for experiment B10 was received
on 29th
March and a small number of mice were sacrificed
for the pathological
examination. Altogether 35 mice have died, most
within 4-6 days
after the cold shock stress. Although this is only
about 1-21% of
the total, the proportion is considerably higher
than that fOujLd
previously. The dead animals were spread
throuGhout the cages and
although acute pneumonia was diagnosed in many
cases, some had lurif;
lesions similar to early CRD in appearance. A
preliminary bacterio-
logical check showed slight contamination in the
lungs of 6 mice
with moderate contamination in one further
instance. Tliree mice
had liver lesions similar in appearance to those
found in mice
with Tyzzer's disease.
Since the disease becomes more apparent with
stress, the
animals have been given a second "cold shock".
BATCo document for Province of British Columbia 5
November 1999


Painting has commenced as planned but the
a-iiiiiials %vi H. be
held in the Quarantine Rooni for up to futir
wccks. ThiS 3-00111 has
a separate air-circulatinG .),stcm and is the
ojil), otie which can
be used with safety if there is any risk of
infection.
For this period, the aninials will be held
U111IL1111bered in cages
containing 18 mice. A system of special car
markinCs or scparate
cages was devised for rcplazeinent mice.
As a further check, 40 of the 150 mice allocated
as replace-
ment animals will be used for health checks. These
will be treated
with condensate (56.6 mg dose level, B10-5) to
provide a normal
stress; ten mice will be sacrificed each week for
a thorough patho-
logical examination.
Analysis of Long-Terin Experiniciits
(a) ~.~nj~_Cojnputer System
------------
There has been sonic further delay in the delivery
of the
suite of programs, due to ininor proaranimina
difficulties and
the departure of one of the programmers. Computer
Systems
and Prograrnming Ltd. now expect to deliver the
programs and
manuals early in May.
An extensive series of tests JiE-.s been arranged
for the
acceptance test period. The data from experiment
BO has been
punched onto-IBM cards while that for experiment
BI is currently
being processed. Final checks are being made at
Battelle on
the coded data forms for experiment B2.
(b) Analysis at Battelle
In the last Progress Report it was sue:gcsted that
analysis
of pathologies other than skin tuniours inif,.jit
be utidortaken on
the basis of a proposal sent in by Dr. Mandrella
at Dattolle.
BATCO document for Province of BritiSh Columbia 5
November 1999


The proposal., howevor, di(I iiot r~ o jc~
,ivc! ;t c.1c,11. ()Ljt,jijje f L,
work which mielit I)c ")d jA is also possiblo UjaL
such dlSea.5e5 1131[~JiL reSUIL fr-om Wic natural
aGing, proccs...-es
in mice on loiig-term test. Tt was decided that it
would nat
be worthwhile attempting this type of analysis
until there
was a clear indication that the patholoCies are
related to
the type and close of condensate applied. lit the
first
instance the data will be examined on a graphical
basis.
PROMOTION TEST - FEASIBILITY STUDY
The promotion test which
study is nearing completion.
flue-cured tobacco blond) and
manufactured into Gerlach sheet),
mice pretreated with DMBA for
Table 3.
is being undertaken as a feasibility
The two condensates, 139-1 (control
B9-6 (ethatiol extracted tobacco
have been applied to Carworth
24 weeks. The results are givc1i in
To date no tumours have been detected in the
control groups
and there is a marked difference between the
numbers of tumours
produced in the two condensate groups. On the
basis of these
resultsv the promoting activity of the condensate
from B9-7
tobacco is shoAfn to be considerably more active
than that from
Gerlach sheet manufactured from extracted tobacco.
Because of the
low yield of tumours with the less active sample,
and the possible
effect on any statistical evaluation, the test is
being continued
to week 28 and possibly to week 32.
BATCo document for Province of British Columbia 5
November 1999


TAT3T,r 3
PPONIOT.FM i
(Duration Of alit) JJ.C;I -211 110(.k-S)
DMBA (pg) Condensate Ntimber n C Skin I'Limours
once thrice WOCJCJ Y an:i mal s
An:himls
TLuiiOLIrS
180 B9-1 -117 17 31
120 B9-1 1147 11 16
180 B9-6 1117 2 2
120 i3q_6 i 147 3 3
180 acetone 1117 0 0
120 acetone 147 0 0
18 o no thina 75 0 0
120 110 thin a 75 0 0
HYPERPLASIA TESTS
Treatment of Stems and Freeze-Dryinr-
Two series of cigarettes were examined to evaluate
the effect
of water treatment on CRS and the effect of
freeze-drying on a
flue-cured blend. The samples were as follows:
~2.ries_l. No. 185 Control CPS
No. 186 WTS treatment
No. 188 CRS treated with water only
(Tliese cigarettes were manufactured from 100%
CRS).
~!.ries 2. No. 174 : Control, CV48 blend
No. 175 : 100% freeze-dried tobacco (CV48).
Reports have been. issued which -iye dctaiLs of'
Ou-se or
comparable samples. Physical measurements for Llic
control CIB and
WTS are given in IUDE' report RD 750-1Z (sce TabIc
h, test 127
BATCO document for Province of BritiSh COIUMbia 6
November 1999


_10-
Control "A" and WTS 111311). R(--3ulLs %,,ere j-u
p1-cs(,,n Lott J',)L, Uir: water.
only sample (,,.ode 188); in this case the CIZS
was sprayed wiLb wator
to a target level similar L; that used fin- WTS
The evaluation of freeze-dried tobacco has been
describod 11L
R&DE report RD 715-R, AlthOUGh UM tobacco for Lhis
sample was
freeze-dried at a different time, the blend arid
processing was
identical with that us:or~ for the Mark 8 smoking
maChinC samples
examined in the report.
TABLE 11
IIYPJ',IZPLAS.IA TZESUL,rs
Code Description
No.
BO
185
186
188
174
175
Flue-cured lamina. (Control)
100% CRS (Control)
100% WTS
100% CRS - water only
Control blend (cv48)
100% freeze-dried tobacco (CV48)
Area Activity RatinG
(micron-days)
Observed Nornialised
(BO = 100)
95 100
65 68
77 81
71 75
75 79
78 82
CRS Samples
On the basis of this Ilyperplasia test, it is
apparent that
the condensate from the control (100% CRS)
cigarottes has appreciably
lower activity than that from cigarettes
nanufaLtured with WTS or,
CRS treated with wate3- 0111Y (i.e. not StCaMed).
J'JIe ictivity of
the condensate from all the 100% stein Samples is
significantly
lower than that of the 100% lamina cigarettes.
This result is in
C:)
C-D
BATCo document for Province of British Columbia 5
November 1999


agreement with the provisional rcsults of tho
long-ter-111 tcSt (nli).
Yree ze: dri e d -Se r i (, s
The ai-na activity rati.ngs indicate that Hic,
spccific ac(..Lvjty
ef the condensace is not affected by the use of
freeze-dried tohacco.
In the R&DE evaluation, some reduction in toxicity
to tetra-
hymena was noted but there was no difference in
the chorioallantoic
membrane thickening test.
After making allowance for a steni content of
about 25%, the
results for both samples are lower thaj)
anticipa[ed 1rn!!; previous
-values obtained for CN-102 blend and it appears
possible that there
Is a difference between the CNI02 and C~rl~8
blends.
Canadian PCL Samples
Four samples from Mozitrec~l were tested to
examUic the effect
of PCL manufacture on two different flue-cured
tobacco grades.
Mie non-filter cigarettes were manufactured from
the following
tobaccos:
260 Leaf grade D + CRS (ratio 3:1) Plus 3.5%
glycerol
764 Leaf grade D + CRS (ratio 3:1) made into PCL
(AJ100
process)
699 Leaf grade R + CRS (ratio 3:1) Plus 3.5%
glycerol
034 Leaf Grade R + CRS (ratio 3:1) made into PCL
(AJ100
process)
The results of this test are shown in Table 5, and
are of interest
in that they indicate that:
(a) The AJ100 process does not reduce the specific
activity of the
condensate as jud~;ed by the hyperplasia test.
(b) The activities of the condensates from both
samples based on
Leaf Grade D are appreciably lower than those of
the sample.5
manufactured from Leaf' Gi-nde R.
BATCo document for Province of British Columbia 5
November 1999


-12-
The present results for the samples ba5cd on Leaf
Grade H are
in close agroOllierLt with thOSC
ucing the A.TjOO process. In a
document 190/3), three samples
compared with regular PCL: the
with a laniina/CRS ratio of 3:1
91 for the PCL.
found For
previous
produced
activity
was 88,
earlier samples produced
hyperplasia tasL (BTC
by the AJ1OO process were
rating of both samples
::ompared with a value of
TABLE 5
CANADIAN PCL HYPERLASIA RESULTS
Area Activity Rating
(micron-days)
Code Description
No.
Observed Normalised
(DO = 100)
BO Flue-cured lamina: control 93 100
26 o Leaf D + CRS (ratio 3:1) 72 77
764 Leaf D + CRS made into PCL 76 82
699 Leaf R + CRS (ratio 3:1) 82 88
034 Leaf R + CRS made into PCL 82 88
PRT Samples
A series of cigarettes, manufactured (except for
the control.)
wholly from PRT prepared using Burley stems, was
submitted to the
hyperpla5ia test in July. The details of the sheet
materials,
which can be considercd as two sots, are as
follows:
Set A
RS63 - Burley CRS
RS30 - PRT ex hoL soaked BurLey stem (no extract
return)
RS72 - PRT ex hot soaked Durley stem (extract
returned)
CD
CD
BATCo document for Province of BritiSh Columbia 5
November 1999


-13-
~2 L-B
(RS30, PRT no extract rOtUl'n, COlItT'01.)
RS33 - Control witli Ii% calcium oxal~ tte
RS36 - Control. with 10% bismuth oxide
RS22 - Control with 13% clialk
RS25 - Control. with 21% carbon
The PRT sheets were manufactured in R&DE using the
miniature
Fourdrinier machine: see R&DE reports RD 772-1?
and RD 733-Rv
covering different aspects of these samples.
Because of the number of samplos submitted, one
saiiiple (RS33
with calcium oxalate) was examined as part of a
separate series
but the hyperplasia results are normalised using
the appropriate
value for BO.
TABi.,j,- 6
PRT jI)rpERpT,ASIA RESULTS
Area Activity Rating i
(micron-days)
Code Samp I e
Observed Normalised
(130 = 100)
BO
I 88 100
Set A
Rs63 Burley CRS 78 89
RS30 PRT - no extract return 97 110
RSIZ PRT - extract returned 62 71
Set B
PRT control (RS30) 110
RS33 PRT with calcium oxalate 86 91
v BO=95
RS36 PRT with bistmith oxide 86 98
I RS22 PRT with clialk
91
103
RS25 PRT with carbon 88 100
"0
BATCo document for Province of British Columbia 5
November 1999


Set A
The results sliow t)L;IL:
1. The Burley s Lcmj (I?S6:3) has a lower I(- t i
vi I-y tban Wic flu~.--
cured lamina but the result is not as low might
have
been postulated for a cigarette manu ['ac tit
oiiLiroly froiii
CRS.
2. There is a siCnificant reduction in the
ac'-j.vity wbeo Lite
stem is made into PUT with the extract rohirned
(RS72).
This result suggests thtt the recotistituLion
process, wfif-.11
applied to Durley tobacco stem, has a direct
al'Fect on the
activity of the condensate as mozi-stired by the
hyj)ci-plasla
3. When the extract is not returned, Lite activity
oF the canden-
sate from the PRT (IZS30) is inc-l-ea.-iod Coll
Siderab IV . 1ji
addition this method of manufacture, Jead., to a
cojisiderable
increase in the delivery of dry CO11LICJ1SDLC.
ThiS Off'OCt WiLS
demonstrated clearly in report RD 772-1?.
Set B
The results indicate that the incorporation of
calcium oxalate
and bismuth oxide into PRT with no extract return
leads to a coit-
siderable reduction in activity as jUdGed by the
hyporplasia test.
An indication of a slightly reduced activity was
found previously
when calcium acetate and bismuth laCULLe were
arlded to a fluo-cured
blend but an adequate control was not included in
that test (DTC
document 200/15), and a repeat U-St i- L)01111~
011CICrtakOll.
It is also apparent that smaller reductions in
the, specific
activity of the coodenzsatc are acMc-%,(-,j t)v
-mo ijicOrp,,,-,,tju,, ,r
chalk or carbon int-o 13t.1-1-10y St-Cill 11,10.,
with ito I.-CL111:11.
Further hyperplasia tests ij-e I)Ui11jT Co~
oxalllilte Lite CF*r(!((:F-)
C-)
~,,D
00
BATCO document for Province of BritiSh ColUmbia 5
November 1999


-15-
on the activity of tlv~ condensate From
cj.j,,-ar(,tt:cs with chalk
incorporated using a variety of mothods.
REPORTS AND OTIrEll TESTS
The following have been circulated durinU the
period under
review:
RE~port B-22
"Assessment of Acute Upper Airways Irritancy of a
Range of
Cigarettes".
In this report from Huntingdon Research Centre,
the
irritancy of some JANUS samples (B3 and BL~) is
compared with that
for some Lokstedt samples. Of the samples
examined, B4, a flue-
cured blend incorporating 50% CIZS, caused the
smallest docroase
in respiratory rate. The Gerlach sample with added
nitrate
resulted in the largest decrease in respiratory
rate. Detailed
interpretation of the results is difficult and the
meaning of the
test is not understood clearly at present.
Sebaceous Gland Test (Hamburf-) (13TC document
209/4)
In a previous report (B-21), details of the
results obtained
at Huntingdon are given. The same samples were
examined at B-AT
Hamburg using the test developed in their
laboratories.
It was found that, although the Hamburg results
did not
correlate with the Huntingdon values, there was a
very good
correlation between thp Hamburg results and (a)
the predicted
results of the JANUS long-term test5, arid (b)
JAIVUS hyporplasia
results.
It is understood that improvemciiLs li.-.ive been
madc in the
c0
BATCo document for Province of BritiSh Columbia 5
November 1999


-16-
sebaceous gland test teclLnique used at
Huntingdoii. The test
reported in B-21 is being repeated at the expense
of IffZC.
SRE/SEW/3-5
10 th May, 1971
CD
BATCo document for Province of British Columbia 5
November 1999