EWECTS OF UNDERFEEDING AND HORMONES
ON SKIN TUMOES EN MICE PRODUCED BY

9, fQ-DiMETHYL-I, E-BENZTANTHRACENE
AND CROTGN OIL

Them for Hm Degree of M. S.
MECREGAN STATE UNIVERSITY

James Frantz Long
1959

MEWWMINGMD mmonsmmmsmm
PROIIJCED BI 9510-DDEm-1.2-BENZANT}MCENE AND CROTON OIL

JAIMFRANTZLONG

AN ABSTRACT

Sublitted to the College of Science and Arts of
Michigan State University of Agriculture and Applied
Science in partial fulfillment of the requirements
for the degree of

MASTER 0? 8013103

Department of Physiology and Phameolog

1959

Approved

 

James Frantz Long
ABSTRACT

l. Berenblum and Shubik (1947) suggested a two-stage mechanism
for carcinogenesis, a short "initiation“ phase at which time normal cells
are permanently altered to become dormant tumor cells, and a longer
"promotion" phase during which time the dormant tumor cells divide and
eventually become visible tumors. The purposes of the present study
were to determine the effects on chemically-induced skin tumors of:

(a) underfeeding during the “initiation" and the “promotion" phases, and
(b) underfeeding and hormones during both phases.

The carcinogen used in these experiments was a 0.5 milliliter
solution of 0.5 percent 9,lO-dimethyl-l,2-benzanthracene (DMBA)
dissolved in benzene, and was applied over a circumscribed area in
the interscapular region over the back of mice. This was followed by
once-weekly paintings with 5 percent croton oil in liquid paraffin.

DMBA was used as an "initiating" agent and croton oil as a "promoting"
agent. The hair was clipped from the backs of albino Swiss female

mice of the CFl strain prior to application of the DMBA or croton

oil. Attempts were made to alter the "initiation" phase by underfeeding
or giving hormones for 10 days before and 10 days after DMBA application
and similar attempts were made to alter*the "promotion" phase by treat-
ments beginning 10 days after DMBA application and continuing for the
remainder of each experiment.

2. Experiment 1. When 2/3rds of the average daily food intake
of the controls was fed to mice during the ”promotion" phase. tumor
growth was inhibited significantly. Some tumors developed despite a

loss in body weight. 'When 2/3rds of the controls' food intake was fed

ii

James Frantz. Long

to mice during the "initiation" phase. tumor yield was increased as
compared with the control group fed 31%. Thyroxine (0.5 milli-
gram per kilogram diet) inhibited while thiouracil (2 percent in diet)
stimulated tumor growth when fed to mice given 2/3rds of the controls'
food intake, when compared to mice given 2/3rds feed alone. A relation-
ship was found to exist between final body weight and tumor incidence.

3. Experiment 2a. When 2/3rds of the average daily food intake
consumed by the ad 119114319 fed controls. was fed during the "promotion"
phase. tumor growth was inhibited. Feeding 2/3rds feed alone during
the “initiation" phase did not show a significant difference from the
controls‘ tumor yield. Two-thirds of the controls' food intake to—
gether with thiouracilOJ percent in diet) during the 'initiation"
phase imreased tumor yield significantly. A relationship was again
observed between tumor yield and average final body weight.

14-. Experiment 2b. The same mice were used in this eXperiment
as were used in Experiment 2a. 20 weeks after DHBA application. The
control group was continued without treatment. A group which had been
fed 51 M for 20 weeks. was put on 2/3rds feed thereafter (for
12 weeks). Another group which was restricted to 2/3rds feed for 20
weeks was fed ad mm thereafter. The group initiale fed ad
mum food intake had an average body weight of 32.0 grams and
269 tumors at the end of 20 weeks post DHBA. Four weeks later, they
had an average, body weight of 23.4 grams and only 97 tumors. The ave-
rage body weight and'tumor incidence remained practically unchanged
for the remairder of the experiment, 32 weeks after DHBA application.

‘The group fed ad, 1mm. after the previous 20 weeks on restricted

iii

James Frantz Long

food intake, gained an average of 6.1+ grams but showed an increase of
only 9 tumors during the first two weeks on eXperiment. The average
body weight remained at about 32.0 grams 10 weeks after ad 1mm food
intake was initiated. The tumor yield increased slowly and an abrupt
increase in tumor yield was noted 10—12 weeks after a_c_i_ 1mm feeding
was initiated. I

5. Experiment 3. There was some indication of an altered tumor
incidence as a result of hormone treatment or caloric restriction during
the, ”initiation" phase in the first two experiments. Attempts were rude
to confirm these findings in the mice of this experiment. For 10 days
before and 10 days after DHBA application, mice were fed? 1/2 of the
controls' average daily food intake. or 50 milligrams per kilogram diet
of hydrocortisone acetate. or 0.5 milligrams thyroxine per kilogram
diet and 2/3rds feed,'or 2 grams thiouracil per kilogram diet and 2/3rds
feed. The results showed that none of these treatments influenced
tumorogenes is Significantly.

6. A correlation analysis on the mice of the first 2 eXperi—
ments was performed to test the relationship betwaen tumor incidence
and average body‘weight. The correlation coefficient, r, equaled .85
when average tumors per tumor-bearing mouse was related to average body
weight. The r valuefor the correlation between percent animals with
tumors and final body weight was .91. .

7. It is concluded that: (a) final average body weight is a
significant factor in chemically-induced skin tumorogenesis, (b) neither
caloric intake nor hormonal treatment influences the "initiation” phase

of tumorogenesis, (c) during the "promot ion" phase of tumorogenesis,

iv

James Frantz Long

restricted food consumption inhibits tumor growth, while thiouracil
and thyroxine increase and decrease tumor development, respectively,
(d) mice with a small number of tumors after 20 weeks of underfeeding
show an increase in tumor yield when placed pn ad libitum feeding for
12 weeks; mice with a large number of tumors after 20 weeks of ad

libitum feeding, show a pronounced decrease in tumor yield when placed

on reduced food intake for 12 weeks.

EFFECTS OF UNDERFEEDING AND HORIDNES ON SKIN TUNES IN MICE
WCED BY 9,10-DIHETHYL-1.2aBENZANTHRACBNE AND CROTON OIL

Jm FRANTZ 1.0m

ATHESIS

Suhnitted to the Collegeof Science and Arts of
Michigan State University of Agriculture and Applied
Science in partial fulfillment of the requirements
for the degnee of

MASTER or SCIUCE

Department of Physiology and Pharmacology

1959

ACKWLEDGEI'ENTS

The author is grateful to Dr. J. Meites, Hofessor, Department
of Physiology and Pharmacology, for his guidance and interest through-
out this study. Thanks are also due Mr. C. S. Nicoll and Mr. P. K.
Talwalker for technical assistance. I

Indebtedness is due NIH and the Michigan Agricultural Experiment
Station for grants to Dr. J. Meites which helped to support this work.

The shydrocortisone acetate was supplied by Dr. L.‘Michaud, Merck—Sharpe

and Dohme. Rahway, N. J.

vii

TO*P%TRICIA

viii

TABLE OF CONTENTS

PAGE
ACKNOWLEDGEMENTS vii
LIST OF TABLES xi
LIST OF FIGURES xii
INTRODIBTIGN 1
LITERATURE.REVIEH’ 3
I. Chemical Carcinogenesis 3
A. History 3
B. Chemistry of Carcinogenic Polycyclic
Hydrocarbons 4
C. Histogenesis and Progression of Chemically
Induced Skin Tumors in.Mice 5
D. Cellular Components Affected by Chemical
Carcinogens 7
II. Nutrition and Carcinogenesis 12
III. Thyroid Hormone and Carcinogenesis 15
IV. Cortisone and Carcinogenesis 18
METHODS AND MATERIALS 20
I. General . 20
II. Administration of the Carcinogenic Agent 20
III. Recording Methods 21
EXPERIMENTAL.RESULTS 22
I. Experiment 1 22
A. Procedures 22
B. Results 23
II. Experiment 2a 28

III.

IV.

DISCUSSION

I.

II.

III.

IV.

V.

A. Procedures

B. Results
Experiment 2b

A. Procedures

B. Results
Experiment 3

A . Procedures

B. Results

Role of Nutrition in Tumorogenesis
Correlation Between Body Height and Tumor Incidence

Effects of Altered Caloric Intake on Previously
Developed Tumors

Role of Thyroxine and Thiouracil in Tumorogenesis

Role of the "Initiation" and "Promoting" Phase in
Tumorogenesis

SUMMARY AND CONCLUSIONS

BIBLIOGRAPHY

APPENDIXES

Appendix I. Composition and Analysis of Ration

Appendix II. Complete Data of Experiments 1, 2a, and 3

Appendix III. Statistical Fomulae

28
29
30
3o
33
35
35
#0
45
“5

47

1+9

55

57
59
63
68
69
71
75

LIST OF FIGURES

FIGURE

10.

Tumor Yield of Animals Subjected to Treatment During

"Initiation" Phase

Tumor Yield of Animals Subjected to Treatment During

"Promoting" Phase

Tumor Yield of Mice Subjected to Treatment During "Initiation

Phase (I) or "Promotion" Phase (P)

Relationship of Average Body Weight and Average Tumors Per

Tumor-Bearing Mouse in Mice Subjected to DMBA

Effects of Restricting Caloric Intake 20 Weeks After

Application ovaMBA on Tumor Growth in.Mice

Effects on Tumor Growth in Mice on Restricted Caloric

Intake Fed Ad Libitum ZO'Weeks After DMBA Application

Correlation of Average Tumors Per Tumor Mouse to Average

Body Weight with 903% Mean Confidence Belt

Correlation of Average Tumors Per Tumor Mouse to Average

Body weight with 90% Individual Confidence Belt

Correlation of Percent Mice with Tumors to Average Body

Weight with 90% Mean.Confidence Belt

Correlation of Percent Mice with Tumors to Average Body

weight with 90% Individual Confidence Belt

xi

PAGE

'1

32

38

39

51

U\
N

TABLE

LIST OF TAI‘IES

Effects of Underfeeding or Underfeeding and Hormones
During "Initiating" or "Promoting" Phase of Tumor Growth
Effects of Underfeeding or Underfeeding and Thiouracil
During the 'Initiating" Phase, and of Underfeeding During
the “Promotion" Phase

Effects of Altering Caloric Intake on Skin Tumors in
Mice Subjected to DMBA 20 Weeks Previously

Effects of Underfeeding or Underfeeding and Thiouracil
or Hydrocortisone Acetate During "Initiating" Phase of
Skin Tumorogenesis

Effects of Underfeeding or Underfeeding Plus Hormones

on Tumor erth

xii

PAGE

25

31

36

5O

INTRODUCTION

Insurance company statistics reveal that the number of cancer
fatalities is greater in persons that are overweight, and underweight
individuals are less likely to be cancer victims than persons of
normal weight (Tannenbaum, 19’40). Ancel Keys (1950) reported that
cancer fatalities dropped in Germany and Austria dtn'ing post-war
famine periods. Berenblum (1952), admitting drawbacks in the methods
of data accumulation, states that cancer is more prevalent in countries
with higher living standards than in countries with lower living standards.
If extrapolated this could mean that countries on a high plane of nutri-
tion are more susceptible to cancer.

Studies condmted on experimental animals reveal that under-
feedirg consistently depresses the tumor incidence in all types of
tumors studied. Tannenbaum (1953) reports that Moreschi in 1909 demon-
strated this by observing that a strain of mice which developed spontan-
eous mammary tumors did not develop as many tumors, and the tumors that
did deveIOp required a larger period of time to occur when subjected to
limited food intake. Tannenbaum (191w, 191+2, l9f+5) in numerous studies
involvirg chemically iniuced tumors, spontaneous mammary tumors, an}
spontaneous hepatic tumors reported similar results.

During periods of inanition it is known that the endocrine glands
are affected. Urdernutrition may result in cessation of the estrous
cycle in rodents (Heites, 1953). Kurschmann (1922) observed that under-
nutrition had a depressing action on thyroid weight. The adrenal cor-

tex decreases in activity during periods of mild inanition but hyper-

trophies during severe starvation (Selye, 19513. It is also known that
changes in endocrine function may alter nutritional requirements (Meites,et
1957). Thus hyperthyroidism increases while hypothyroidism decreases
dietary needs.

Berenblum and Shubik (1947) postulated that tumorogenesis is com-
prised of 2 separate stages: 1) an induction or ”initiating" phase, at
which time certain.cells are altered to become "dormant tumor cells", and
2) a "promoting phase”, which comprises the total time when the ”dormant
cells" are proliferating and eventually become visible tumors. The
possible relation of underfeeding and hormones to this 2-stage mechanism
of carcinogenesis will be considered in this thesis.

The specific purposes of the experiments to be described were to
determine: (a) the effects of underfeeding on development of skin tumors
induced by chemica1.cazginogens, (b) if the effects of underfeeding are
related to the ”initiating phase” or "promoting phase" of carcinogenesis,
(c) the effects of combined underfeeding and hormone administration, and
(d) to what extent the effects of underfeeding may be attributed to

altered endocrine function.

LITERATURE Rm
I. (CHEMICAL CARCINOGENESIS'
him

Percival Potts in 1775 as reported by Wolf (1952) observed a high
incidence of scrotal cancer among chimney sweeps. Thiswas the‘first .
time a chemical cause of cancer was postulated. (For almost a century very
little information concerning chemical carcinogenesis was nude available.
In the late 1800's reports indicated a relatiomhip between skin cancer
and employees in the dye and coal tar industries. The first conclusive
eXperiment demonstrating that skin cancer could be caused by a chemical
carcinogen was by Iamigawa and Ichikawa in 1918. They'produoed. camer
on the ears of rabbits by repeated application of coal tar. Tsutsui (1918)
observed the same effects of coal tar on the skin of mice. In 1933 one of
the active substances in coal tar was isolated ard identified as Lit-benz-
pyrene (Cook. eggs)“ 1933). Since then many related substances were
isolated or synthesized ani tested for carcinogenic activity.

The most widely studied carcinogens are the polycylic hydrocarbons .
Nitrogen and sulfur containing analogs of the 'same‘hydrocarbons can. pro-
duce neoplasms. Boyland and Brues (1937) found naphthylamine to'be,’
specifically carcinogenic in the bladder of dogs. Ioshidu (19310
observed that liver tumors could be produced in rats by oral adyuinistra-
tion of azo dye compounds. The first experimental evidence that cancer
may be caused by a natural occurring substance was reported by Laces-
sagne (1932). He injected estrogens into male mice and observed develop-
ment of breast tumors. 7

‘ Other chemical substances found to be carcinogenic are urethane

(Nettleship and Henshaw, 1943), carbon tetrachloride (Edwards, 1941). and
tannic acid (Karpassay ani Kovacs. 19149). Strong acids or bases have
been demonstrated to be tumorogenic as well as some other non-specific
irritants (Karsner. 1950).
WWW

Heiger (1930) . after determining the spectra of coal tar agents
with carcinogenic activity, synthesized the first chanically pure car-
cinogens. 6-isopropy1-l.2-benzanthracene and 1.2.5 .6-dibenzanthracene.
Since then may other compounds of related structure ani activity have
been'synthesized. The (potent carcinogenic hydrocarbons are 2-methyl-
cholanthrene (which may be synthesized from naturally occm'ring compounds).
9 .lo-dimethyl-l .2 . -benzantl'rracens , 5 ,6-dimethy1-1,2-benzanthracene , 1 .2 ,
5,6-dibenzanthracene, 3.1+.benzphenanthrene, and 5.6-pdimethyl-l,2-benz-
anthracene (Wolf, 1952).

. These compounds all have a characteristic fluorescence spectrum f
of 14000 to 4400' angstroms. Those which occur in coal tar are in the high
boiling. neutral. nitrogen-free fraction (Greenstein. 1954). Another
characteristic which they share is a phenanthrene nucleus (Haddcw. 191+?) .
However. phenanthrene itself is not carcinogenic. Other roots. of the
above named compounds, such as anthracene. benzanthracene, phenanthracene
or benzphenanthrene are not carcinogenic. In order for these cornpounds
to show carcinogenic activity the molecule must be activated by the
addition of methyl or benzene groups at specific sites. The addition
of methyl groups at the 9 an! 10 positions of dibenzanthracene acti-
vates the bond between the 3-h- carbon atoms. This region is known as

the up region (Pullman. 19146. 1947).

In the netabolism of these carcinogens, a hydroxy group is added
to the hydrocarbon. but not at the active site, the so-called 'K' region
(Heildelberger arr! Jones, 1948).. This iniicates that the active site of
the molecule is in some way inhibited after it enters the biological
media. weigert and Hottram (1946) obtained evidence that the metabolized
carcinogen was attached to a protein coniplex. Miller and Miller (1947)
were the first to present direct evidence that the combination between
the carcinogen and a protein existed in those cells which developed
into tumor cells. The site of combination is the activated portion of
the molecule. 1

The structural peculiarities of the active polycylic hydrocarbons
indicates a definite mode of entry into the cell. If the nathy’l group
on position 10 of the active carcinogen 9.10-dimethyl-l.2-benzantln‘acene
is replaced by a longer chain, the compound loses its carcinogenic
activity in spite of the fact that the "K" region is'sufficiently active
to bind with protein (Wolf, 1952). The carcinogenic hydrocarbons are all
fat soluble ani enter the cell via the lipid portion of the cell membrane.
The steric configuration of the molecule dictates whether the molecule

will penetrate the cell surface (Greenstein. 1951+).

H 014:“; z... -W, oneu : '13”): ,.. T..- n e
The normal mouse skin consists of two layers of undifferentiated
epithelial cells. After the application of the carcinogen the epithelial
cells differentiate into basal cells. spinous cells. and cells of the
stratum granulosum and stratum corneum. This resembles human skin ad
is thought to be due to increased mitotic activity immediately after the

carcinogen application. Three to four days after carcinogen application,

the cells in the center of the painted area become necrotic, leaving a
single layer of thin elongated epithelial cells covered by a massive
amount of keratin. The margins of the painted area are hyperplastic. The
hair follicles and sebaceous glands follow the same pattern as the epi-
dermis. The central area shows extensive and severe degeneration of the
cells lining the hair follicles. At the end of the first week regeneration
starts. Repair results inns multilayered well differentiated keratinized
epithelium. Small blood vessels and lymphatics of the dermis dilate,

and inflammatory cells appear, particularly mast cells. The damaged
sebaceous glands and hair follicles regenerate slowly; Some hair folli-
cles return to normal activity, others remain dormant, while others show
progressive changes leading to tumors. These changes are an increased
rate of mitotic activity in.the epithelial stumps of the hair follicles,
thickening of the epidermis, broad masses of epithelium in the dermal
layer. and cysts in the modified hair follicles (Stewart, 1953).

Dermal changes durirg development of carcinoma are characterized
by hyperemia and increased talangiectasis. Inflammation is usually
present with mast cells accumulating until the cancer finally develops.
Then the mast cells gradually disappear (Stewart, 1953).

Some chemical changes associated with skin tumorogenesis are
decreased calcium. iron, copper, and zinc and increased magnesium,
Cytochrome oxidase, cytochrome c and flavins are decreased while
apyrase, succinic dehydrogenase, and‘arginase are markedly increased
(Karsner. 1950). The cytoplasmic RNA increases as it does when‘cells
are subjected to radiation ®eRobertis. et a1. .1954). Shubik, et al.(l953) re-

ported that four types of tumors appeared after carcinogen application on thq

skin. The four types described are:

l. The sessile papilloma. This is an outgrowth of fibrous tissue
bulging out at the surface of the skin and covered by epithelium. These
resemble warts found in humans.

2. The pedunculated papilloma. This tumor is present on a fibrous
stalk containing blood vessels, and is covered by a thin epithelium which
is highly keratinized.

3. The conical tumor. on gross appearance this type demonstrates
a craterous depression penetrating the dermis. Microscopic observation
reveals a basement membrane impermeable'to epithelial cells. The rete
pegs are elongated and arborous.

h.. The carcinoma. This is the malignant form of skin tumor.
Grossly it appears like the conical tumor. Microscopic examination re-
veals invasiveness by epithelial cells. The mitotic rate is very high.
Cellular constituents are enlarged and cytoplasm is decreased. The cells
have a hyperchromatic nucleus. Blood sinusoids are common.

The sessile papillomas may progress to'carcinoma (Shubik. 1953).
The author subjected 100 female mice to twice weekly paintings with 9.10—
dimethyl-l,anenzanmhracene. He found that of 883 tumors arising as
sessile papillomas, 96 of these progressed to carcinomas. Only 4 car-
cinomas developed directly. The progression is not correlated to growth
of the tumor: Some small pedunculated papillomas became malignant while

some larger pedunculated papillomas do not progress to malignancy.

1). 9mg“; Cgmgggggtfi Agrggteg g2 Chggica] nggigggggg

The cellular components of the cell.affected by the carcinogenic

aromatic hydrocarbons is not known. Calcutt (1958). utilizing radio-

active 01""

, found 1,2-benzanthracene. 1.2.5.6-dibenzanthracene, and 3, 1+-
benzpyrene. all carcinogens, in the nucleus, the mitochondria, and the
microsomes. He found the non—carcinogenic hydrocarbons. anthraoene and
phenanthrene. in the microsomes. Crabtree (19479 observed that the non—
carcinogen, anthracene. couldinhibit the action of 3.Lb-benzpyrene. This
inhibition was thought to be due to a selection of active sites by the
non—carcinogen. The fact that these non-carcinogens do not enter the
nucleus of the cell and thatthey inhibit tumorogenesis in some way. led
Calcutt (1958) to postulate that the active site of the carcinogenic
action was extranuclear.

Conan (191+8) demonstrated that the appearance of the cell surface
was altered by 9.10—dimethyl-l,2-benzanthraoene. By precipitating
chromium on the surface of cells treated with the carcinogen, and then
hydrolyzing the cell he acquired a photographable replica of the cell
surface. These photographs revealed a very rough cell surface for the
carcinogen treated cell while normal cells have a smooth surface.“ He

suggested that the altered appearance may in turn affect the permeability

of the cell membrane.

WWW

There are mam theories com erning the mode of action of chemical
carcinogens. Few of them, however, have been able to withstand direct
experimentation. A few of the theories that have withstood experimental
examination are presented below.

Cramer and Stowell (19141) postulated that mitotic activity is not
directly stimulated by the carcinogen, but that the chemical carcinogens

exert a transient or suppressing effect on mitotic activity. The prolif-

eration of epithelial cells. which takes place three to four weeks after
carcinogen application and eventually leads to the tumor mass. is due
to the formation in skin of a substance that stimulates epithelial cells
to rapid division over a long period of time. The authors did not offer
any idormaticn comerning the nature of this I'stimulatory" substance.

Haddaw (19M) hypothesised that the genetic material of the somatic
cell was in some unknown way altered by the carcinogenic substance. Had-
dow states. “In response to the application of a chuical carcinogen.
some cells are destroyed while others are genetically altered to form a
new cell type which is able to reprodlme and establimi a new population
of cells. This new population of cells is not inhibited and continues
dividing. eventually yielding a tumor." This theory of carcinogenesis
is comicnly referred to as the “somatic cell mutation theory". Recent
experimental evidence (discussed previously). iuiicating that the site
of action of the carcinogens is extranuclear. suggests a mechanism diff-
erent than chromosomal alteration. Haddow further suggests that am
biological alkylating agent may be carcinogenic .

Berenblum and ambik (191+?) reported that two separate stages are
involved in carcinogenesis. In the first stage. the “initiating“ stage.
a few cells are converted into “dormant tumor cells“. This process
causes a rapid. irreversible transforntion of a few of the normal cells .
During the second stage. the I'promctixg" stage. these dormant cells are
stimulated to grow and eventually become the visible tumor. The same
authors (1959) tested their hypothesis by subjectiig female Swiss mice
to a single application of 9.lO-dimethyl-l.2-benzanthracene. followed

by weekly applications of croton oil. The croton oil applications were

 

10

commenced at different intervals after the carcinogen was applied. They
observed identical tumor yields if the applications of croton oil were
started one week. two weeks or even twenty-four weeks after the carcin-
ogen was applied. The carcinogen apparently altered a few of the normal
cells which existed as "dormant tumor cells" and the croton oil promoted
these cells to multiply. eventually yielding the visible tumor. At this
time croton oil was thought to be non-carcinogenic. but since thenIBoutifl
well, et a1. (1957) reported croton oil to be a very weak carcinogen.

Studies with ascites tumor cells (a cultured strain of cancer
cells) show that the respiratory system of the cancer cell is altered.
The energy requirements of the ascites tumor cell are met by glycolysis
rather than by respiration. These tumor cells have been shown to have
lower concentrations of the oxidative enzymes and B vitamins (Greenstein.
195h). Potter e1.al. (1950) reported that succinic dehydrogenase. cyto-
chrome c. and cytochrome oxidase activity of the rat hepatoma are only
22-30 per cent of that contained in normal liver tissue.

From this type of information and work of his own originating
in 1922, Otto Warburg (1956) postulated the anaerobic theory of carcin-
ogenesis. He states, "No one today can doubt that we understand the
origin of cancer cells if we know how their large fermentation origin-
ated, or. to express it more fully. if we know how the damaged respira-
tion and the excessive fermentation of the cancer cells originates.‘
He concludes that carcinogens cause some type of irreversible damage
to the respiratory system of the cell. For these cells to remain alive
a new source of energy is required. This energy requirement is met by

increased fermentation. The ability of the cell to supply sufficient

energy by.means of fermentation requires a long period of time and many
cell divisions. Thus the latent period associated with tumorogenesis is
merely the time required for this fermentation process to proceed to a
point where the cell can create enough energy to exist among the normal
cell population. One of the consequences of this increased fermentation
is the loss of structure of the affected cell. This loss of structure
yields undifferentiated cells which characterize the tumor mass.

Some respiratory poisons are urethane, polycyclic hydrocarbons.
arsenious acid, hydrogen sulfide and irritants which reduce blood supply.
All of these have been.shown to be carcinogenic. ‘Harburg states that any
respiratory poison.may'be carcinogenic. The controversy revolving about
‘Harburg's hypothesis is that most of his conclusions are drawn from work
done on cancer tissue cultures. which are characterized by a low aerobic
metabolism. Some tumors have a very high aerobic metabolism but in every
case the anaerobic metabolism is increased over the normal tissue (Wein-
house. 1955)t It is the consensus of critical opinion at present that
the cancer cell does not differ chemically from a normal cell. except

perhaps. in a quantitative sense.

12

II. NUTRITION AND CARCINOGENESIS

Moreschi in 1909 found that the development of mammary tumors in
a specific strain of mice was delayed due to underfeeding. Tannenbaum
(191+2) verified these results by demonstrating that mice treated with
3,h-benzpyrene and subjected to underfeeding, developed fewer tumors
and the latent period was lengthened. Tannenbaum (1942) also conducted
an experiment to show the effects of caloric restriction pg; gg. All the
experimental mice received 2 grams daily of a basic ration containing all
the essential nutrients. One group had cornstarch added to their diet.
thus increasing their food intake to 3 grams per day. The group_receiving
the additional calories developed more tumors and showed a decreased
latent period. Tannenbaum commented: WNithin the limits of our present
knowledge it is evident that caloric restriction pgr,§§ is the principal
cause of the observed inhibition of tumor formation.‘

In.l9uh. after Berenblum's "two—stage mechanism" was postulated.
Tannenbaum.subjected mice to dietary restrictions during each of the
stages of carcinogenesis. One group was fed dQ.libiEBmtf0? the duration
of the experiment. A second group‘was given full feed for 10 weeks, or
until the time the first tumors became visible. In a third group he
restricted the food intake during the time the tumors were visible and
growing. The fourth group was underfed throughout the experiment.
Repeated applications of 9.10-dimethyl-l.Z-benzanthracene served as the
tumorogenic agent. Sixty-nine per cent of the mice in the first group
contracted tumors while only 2h per cent on restricted food consumption
contracted tumors. The group subjected to underfeeding while the tumors

were visible had a tumor incidence of 34 percent and the group on

13

restricted feed before the tumors were visible had a tumor incidence
of 55 per cent. This indicated to the author that the major effect of
caloric restriction occurred during the time the tumors were visible
and growing. or during the ”promoting phase" of tumorogenesis.

Tannenbaum and Silverstone (l9lt9a)tested the effects of caloric
intake when mice were subjected to conditions that altered their dietary
requirements. One group of mice was subjected to cold room temperatures
and fed ad libitum. Two other groups had dinitrOphenolfland sodium fluo-
ride added to their diets respectively. The results demonstrated that
fewer tumors were contracted in spite of an increased food intake, in the
groups subjected to the treatment described. The data suggested to the
authors that neither food intake nor the rate of metabolism are consis-
tently related to tumor formation.

A linear relationship between caloric intake and tumor incidence
was reported in mice subjected to methycholanthrene. ard in 03H mice
develOping spontaneous hepatomas (Tannenbaum and Silverstone. 1949b) .
Varying the proportion of protein within limits that maintained normal
body weight had no effect on skin tumor incidence of spontaneously-
developing mammary tumors. However. 03H mice, which develop spontan-
eous hepatomas. had fewer tumors when given a ration containing 9 per
cent casein. This same level had no inhibitory effect on the skin or
mammary tumors (Tannenbaum and Silverstone. 1949c).

By increasing the, fat, content of the diet‘. Bauman and Rusch (1939)
observed an increased incidence and decreased latent period in chemically
induced skin tumors. Other investigators reported similar results with

increased ingestion of fat. Boutwell at al. (1949), by determining the

lfl

caloric equivalent of ingested food. postulated that the increased tumor
incidence is related to increased caloric intake.

Manipulation of the vitamin content of the diet of mice subjected
to chemical carcinogens has little or no effect. if the animals are
allowed to maintain normal body weight. If the vitamin content is
decreased to a point where it retards caloric intake. the tumor inci-
dence is decreased (Rusch. Boutwell and Brusch. 1949). These observa-
tions indicate that any dietary alteration which affects the total body
weight of the experimental animals may also affect tumor growth in a
similar fashion.

15

III . THXROID HORMNE AND CARCINOGENESIS

Hyperthyroidism. in addition to elevating the metabolic rate of an
animal. has other physiological effects which can affect carcinogenesis.
i.e.: increased heart rate, peripheral dilatation of blood vessels. '
increased movement 'of blood. perhapsan increased blood volume. increased
cellular differentiation. increased hair growth and increased skin thick-
ness (Turner, 1955). At the cellular level. increased thyroid homone
may increase the amount of oxidative enzymes. including phosphatase.
succinic and cytochrome oxidase, cytochrome c. succinoxidase and hexo-
kinase (Guyton. 1956).

The carcinogenic reSponse in animals given thyrcxine or thiouracil
have been inconsistent and not sufficiently tested. These inconsistencies
could evolve from the high dosage of carcinogen applied or the high dose
of thyroid substance often administered. As the dosage of carcinogen
increases, the carcinogenic respome increases regardless of experimental
manipulations (Tannenbaum. 191+0). An excessive dose of thyroid hormones
can affect the metabolic requirements of the experimental animal to the
extent that body weight declines and a synirone relative to inanition
appears .(Gilroy. 1930). Gilroy also observed that administration of
thyroid extract reduced the number of skin tumors without reducing the
body weight of the mice subjected to a chemical carcinogen. However. if
he gave arginase along with the thyroid hormone. a decrease in tumorogen-
esisdid not occur.

‘ Kreyburg (1938) observed an increase in tumor incidence when mice
were given thyroxine and subjected to coal tar applications on the skin of

mice. Smith et a1 . (1942) found tint hyper- or hypothyroidism ind no effect

16

on mice painted on the skin with methylcholanthrene. The author admits
that this could be due to an excessive amount of carcinogen applied to
the mice. Lerman (1947) suggested that thyroid function did not have any
influence either on experim31ta1 or human cancer. He said, however. that
thyroidectonv decreases tumor growth and also decreases body weight. A
stimulatory effect on skin tumorogenesis was reported by Silverstone and
Tannenbaum (1949) . when mice were given thyroxine only during the time
3,l&-benzpyrene was administered. If thyroxine was given during the growth
phase of tumor production. they observed a depressant action. However.
they gave a very high dose of thyroid hormone and observed that although
the mice ingested 30-40 percent more food than control animals. they
showed a weight loss of 10 percent. They concluded that the depressing
effect of thyroxine on tumor incidence was due to reduced body weight
rather than to the administered thyroxine.

Thyroxine and thiouracil have been reported to inhibit and stim-
ulate chemical induction of sarcomas. respectively (Bather and Francks,
1952). These sane effects were not observed in transplanted or well
established sarcomas. Grad (1957) observed that AKR mice rerdered hyper-
thyroid by thyroxine administration had a significantly lower incidence
of spontaneous bukemia than did AKR mice rendered hypothyroid by thiou-
racil administration. They observed that the hyperthyroid mice also had
a lower body weight. In another experiment reported by the same authors.
excess vitamins and liver extracts together with thyroxine were admin-
istered. and they observed an increase in leukemia greater than that
present in the mice given thiouracil.

Bielschowsky (1958) fourd that thyroidectomized rats given amino-

l7

fluorene orally did not develop hepatomas. If they administered growth
hormone to these rats. hepatomas did develop. The authors concluded that
the absence of hepatomas in the athyroid rat was due to an effect on the
hypOphysis which inhibited the secretion of growth hormone.

Thyroxine was shown by Meites (1958) to inhibit chemically-induced
skin tumor production in mice without a coincident loss of body weight or
decrease in food consumption. The level of thyroxine administration was
16-24 times the normal secretion rate. In the same report. it was sham
that thiouracil increased the tumor yield in spite of a decrease in body
weight. The experimental mice were of the Swiss strain ani subjected to
a single application of 9.lO—dimethyl-l.2-benzanthracene followed by once-'
weekly paintings with croton oil. This indicated a specific action of

thyroxine on tumorogenes is unaffected by body weight or caloric intake.

18

IV. OORTISONE AND CARCINOGENESIS

Some of the physiological responses to cortisone administration
which could affect tumor induction in the mouse skin are: (1) decreased
skin thickness (2) decreased hair growth (3) reduction in number of
epidermal cells and (u) atrophy of sebaceous glands. Increased corti-
sone also decreases the number of eosinophils and lymphocytes in the
peripheral blood circulation. Another effect of cortisone which could
exert an effect on growing tumors is that in large doses the animals may
lose body weight (Turner. 1955).

Reports on the tumorogenic response to cortisone administration
have been contradictory. .Sulzberger at g1. (1953) reported a decreased
skin tumor incidence in mice subjected to methylcholanthrene. Engelbreth-
Holm and Asboe-Hansen (1953) reported an increase in skin tumors in.mice
treated with 9.lO-dimethyl-l.z-benzanthracene and consequent cortisone
administration. An.inorease in skin tumor incidence in BALD/C mice was
observed by Spain £1 al, (1956) when cortisone was given 5 days before
and 10 days after the first painting with methylcholanthrene. or when
cortisone was administered throughout the experiment. The carcinogen.
methylcholanthrene. was applied twice weekly for four weeks at which
time sessile papillomas started to appear. Daserga and Shubik (1954)
observed an inhibition of skin tumorogenesis in Swiss mice subjected to
cortisone throughout the experiment. These investigators also used
multiple paintings of methylcholanthrene until tumors became visible. The
only obvious difference between the two experiments just described is
that different strains of mice were employed.

Cortisone administration was shown to have no effect on well

19

established sarcomas in Swiss mice (Baserga and Shubik, 195%. The same
investigators showed that cortisone could increase the metastatic poten-
tialities of carcinomas in Swiss mice. Tumor implants have been observed
to grow more rapidly if the host animal is pretreated with certisone (Begg,
1951). Similar results are apparent when the host is pretreated with
whole-body radiation (Fox. 1958). The authors postulated that the increased
rate of growth of the implanted tumor was due to a decreased number of
lymphocytes in the blood of the host. The lymphocytes, by antibody produc-
tion, were believed to inhibit growth of the implanted tumor for a period
of time. The duration of inhibition was dependent on the type and amount
of tumor implanted. Complete regressionof a lymphosarcoma and a masto—
cytoma. due to cortisone administration, have been reported by Heilman
and Kerdall (190:4) ani by Bloom (1952), respectively.

None of the experimmts described above consider the dietary
alterations or body weight charges in animals subjected to cortisone.
From the preceding discussion on nutrition it seems imperative that infor-

mation of this type should accompany these experimental results.

MATERIALS AND I‘ETHODS

I, Ggng]

Female Swiss albino mice of the CFl strain, purchased from Car-
worth Farms, New City, New York, were used in the ’4 experimnts to be
described. The mice were kept in wire cages in an air conditioned room
at a temperature of 70-72 degrees F. The basic ration. given in steel
feeders, was prepared in the departmental laboratories.. The ingredients

and analyses of the ration are listed in Table I, in the appendix.

II. d t f t C o A

The carcinogen used in all the experiments was a 6% solution of
9,10-dimethy1-l.2-benzanthracene (mm) dissolved in distilled benzene.
The hair was removed from the backs of the mice with electric clippers and.
2 days later .05 ml. of the carcinogenic solution was applied to the inter-
scapular region. This provided a dose of 250 micrograms of DMBA to each
mouse. To avoid evaporation of the benzene; the DMBA solution was pre—
pared for only two groups at a time and was kept in an ice bath during
application. A 1‘ cc. syringe with a blunt needle was used to administer
the DMBA solution. The blunt needle was used to insure‘that the DMBA
was applied topographically and not subcutaneously. It took an average
of about 5 hours to apply the carcinogen to 200 mice.

Starting one week after the DMBA application and continued weekly
thereafter. a 5% solution of croton oil in liquid paraffin was painted
over the backs of the mice. A glass rod served as the application

vehicle for this "promoting” agent.

21

III. Reggming Methgds

In all experiments the mice were weighed weekly as a group and the
average weight per mouse was calculated. At the time the tumors becam
visible (7 to 8 weeks after DMBA application) , the mice bearing tumors
were marked by clipping the ears. A chart was prepared for each mouse
with tumors, and the approximate area of the tumor or tumors was repro-
duced on this chart. During the time the tumors were beirg established
on the backs of the mice. charting was done weekly. After it appeared
that the number of tumors developing reached a plateau. charting was done
biweekly. Thus we had information available on the average weight of the

animals. number of tumors per mouse and approxinete size of each tumor.

EXPERDENIAL assume
I. Elcperiment 1
W235. A .

In Berenblum's (19Lt9) hypothesis the ”initiation" of tumors is a
very rapid, perhaps instantaneous process while the "promotion" of
tumors is .a much slower process. To test'this hypothesis we subjected
some of the mice to nutritiOnal and humoral alterations during the time
the DHBA was in the skin. which is presume-myths time of "initiation".
The remairxier of the mice..with the exception of 1 group which served
as controls. were subjected to similar hermonal and nutritional treat—
ment after the DEA was presumably removed from the skin, or during the
time or “promotion" of tumor growth. Darchun and Hadler (1956) ,‘ using A
c1“ labeled DMBA, observed that only 1 percent of the applied activity
remained in the skin it days after application. Because we did not know
what effect altered nutritional and hormonal balances would have on the
rate of DMBA removal, we subjected the experimental animals to altered
. nutritional and hormonal balances for either 10 days before and 10 days
after the carcinogen was applied. 'or else the treatment was begun 10
days after the DEA was applied. Heildelberger, .et a1. (1948) observed
starved animals retained 1.2.5.6—dibenzanthracene for a lorger period
of time than did his control mice. He did not comment on the signifi-
cance of his results.

Two hundred, 11-13 week-old Swiss female mice were divided into 8
groups of 25 each. The average weight per mouse in each group was 21+.1
g .1 grams. The groups were treated as follows:

1. Controls - no treatmmt

2. 2/3 of the controls' daily food consumption for 10 days before

23

and 10 days after DMZBA application.

3. "Ibid. plus 0.5 milligrams l-thyroxine per kilogram diet.

14-. Ibid. plus 0.2 milligrams growth hormone injected daily (sub—
c ut aneously) .

5. 2/3 of the controls’ food consumption startirg 10 days after
DMBA application, and continuing through remainder of experi—
ment.

6. Ibid. plus 0. 5 milligrams 1-thyroxine per kilogram diet.

7. Ibid. plus 2.0 grams thiouracil per kilogram diet.

8. Ibid. plus 0.2 milligrams growth hormone injected daily (sub-

cutaneously) .

B. Rgsglts

The first tumors appeared at 7—8 weeks after the carcinogen was
applied. The progression of the tumors wasxrecorded for 12-13 weeks
depending on the time the first tumor appeared. The duration of the
experiment was 20 weeks after DMBA application.

In Table l, the beginning and final average body weights of the 7
groups are given together with the total tumor yield. average tumors
per tumor mouse and. the percent of mice with tumors. The controls
(group 1) were fed ad 11th throughout the eXperiment and had a total
yield of 87 tumors, with 72 percent of the mice showing tumors. Group
2. which was fed 2/3 of the controls' daily food consumption for 10 days
before and 10 days after DMBA application, acquired almost 3 times as
may tumors as the control group and 92 percent of these mice had tumors.
Feeding 2/3 feed plus thyroxine (group 3) during the “initiation" phase

reduced the total tumor yield and percent of mice developing tumors. as

compared to the control group. The tumor yield is statistically less
than.the control group at the 10 percent level. Grouplu which.received
2/3 feed plus growth hormone during the “initiation" phase. did not show
a significant difference from the control group. In group 5. the mice
were fed 2/3 of the controls' daily food consumption during the "promo—
tion" phase. This treatment reduced tumor yield significantly (1 per-
cent level). Two-thirds feed plus thyroxine during the 'promotion"
phase (group 6) inhibited tumor growth completely. Group 7 which was
given 2/3 feed plus thiouracil during the "promotion" phase showed a
greater tumor yield than group 5 which was given 2/3 feed only. This
increase was significant at the 5 percent level. In relation to the
controls, group 7 had significantly fewer tumors (5 percent level).
Group 8 was removed from the experiment because of infections at the
site of growth hormone injections.

Figureiiillustrates the progression of tumor growth in the mice

subjected to treatment during the "initiation" phase. The time at which
tumors began to appear is the same for all groups. at about 8 weeks.
The greatest increases in tumors occurred between 10 and 1M weeks after
DEA application. Group 1+, which received 2/ 3 feed plus growth hormone
during the "initiation" phase, is not plotted because the line coincides
with the control group.

The mice subjected to treatment during the "promotion” phase of
tumorogenesis reacted as shown in.Figure 2. The time at which the tumors
appeared in these experimental groups was delayed. The control group
is included for comparison.

Sessile and pedunculated papillomas were the Only types of tumors

25

ounce wquwdopuaofipp mom unease u oesosuam
cease mfisoaosm n ma...
oases weasesssee a He.

eosseoaasme «use nosed eases om.

 

 

we m.: e: m.nm H.:m Amv assessesea + soon n\m mm .5
o o o e.ms o.am Amy oeaaoease + soon n\m NH .m
mm :.m NH s.m~ m.a~ sesame soon m\m mm .m
on m.m was m.wm o.:m “HV use + econ n\m mm .s
so m.s mm m.~m o.am aHv oasuosnsa + econ n\m Hm .n
mm m.m mam n.mm m.sm eoaHv econ n\~ am .m
we m.a em n.mm H.am eaoseeoo mm .H
unease ness leases-a soc easeasa .ear om sense as“: .oe
eon: u unease Hosea “any .a.m .s4 seoaseosa sedan goose

 

magma No.59 ho Ram sanBoamms mo sozuadmaHsz

Gammon amoxmom a: GHHEQRD mo wan—”ES no mean: in a

Total Tumors

I“)
O\

 

etc

220-

200‘

180-

was

120‘

100*

807

 

 

2/3 Feed (I) - /\\

      

- Controls

 

 

 

l ‘ ‘r
6 8 lo 12 1g .16 18 .20

Weeks After DMBA

Figure 1. Tumor Yield of Animals Subjected to Treatment
During IInitiation.” Phase

Total Tumors

 

   

 

 

 

 

 

2&0
220s
200-
180-
160‘
luo~
120-
80—
- Controls fed ad lib.
6o—
; / I ‘— - ‘I x
“0‘ ,"L 2/3 need a Thiouracil(P)
20s
‘A— ...... »*=756ntrols 2/3 lbed.(P)
0 ‘l - - 2f} Weed + Thyroxi no
0 lb lé Silk T6 is is
weeks After DMBA
Figure 2. Tumor Yield of Animals Subjected to Treatment

During ”Promoting” Phase

 

28

seen 20 weeks after the DEA application. All pedunculated papillomas
were charted, while only those sessile papillomas whose surface areas
were greater than 2 millimeters square were charted. Some of the sessile
papillomas regressed durirg the experiment and similar new tumors
appeared intermittently. Table II in the appendix gives the canplete
tabulated results of experiment 1.

Some of the pertinent observations of this experiment are: (l) a
decrease in food intake decreased tumor yield. (2) thiouracil increased
while thyroxine inhibited tumorogenesis. when administered in conjunc-
tion with underfeeding during the ”promotion” phase as compared to the
group given 2/3 feed_only. (3) reading 2/3 feed during the p'initiation-
phase increased tumor yield and (1+) a relationship appears to exist
between the final average body weight of the mice and the tumor yield.
The possible significance of these findings are discussed in a succeed-

ing section.
II. Experiment 2a

A. W

The purpose of this experiment was two-fold (l) to attempt to
substantiate the results observed in the first experiment and (2) to
test the effects of urethane as a tumorogenic agent on the skin of Swiss
mice. Fifty milligrams of urethane was injected per mouse intraperi-
toneally into 5 groups. This was followed by once—weekly painting with
croton oil. Fifteen weeks after the urethane was injected. no tumors
had appeared and this phase of the experiment was terminated. The

urethane dose administered did not appear to have carcinogenic effect

29

on the skin of the mice. although this agent has been shown to produce
skin tumors.
The remaining 4 groups were subjected to the same type of treat-
ment followed in.Experiment l. DMBA application.was followed by once-
weekly painting with croton oil. One hundred female Swiss mice were
divided into ’4 groups of 25 each. The average initial weight per mouse
in all the groups was 2h.0 ; .2 grams. The groups were treated as
follows:
1. Controls - no treatment
2. 2/3 of the controls' food consumption per day for 10 days
before and 10 days after'DMBA application.
Ibid. plus 2 grams thiouracil per kilogram.diet.

a. 2/3 of the controls' food consumption starting 10 days after
DHBA application and continuing through remainder of experi-
ment.

It is important to mention that this experiment was performed in
a different room.than in the let experiment, and fresh air flow, light-
ing, and other conditions were not exactly the same. This may account

for some of the differences noted in this experiment.

B. Bsfifllii

The results of this experiment are tabulated in Table 2. The
control mice (group 1) were fed §d_11bitgm throughout the experiment.
They developed 141 tumors and 71 percent of the mice had tumors 20 weeks
after DMBA application. In this eXperiment the mice fed 2/3 fbed during
the "initiation” phase (group 2) had fewer tumors than the control

group; however, the difference is not statistically significant at the

30

10 percent level. Group 3 which received 2/3 feed plus thiouracil
during the "initiation" phase had almost twice as many tumors as the
control group. The mice fed 2/ 3 feed during the "promotion” phase again
displayed significantly fewer tumors than the control group.

Figure 3 illustrates the progression of the total tumor yield.

The tine required for tumors to develop ..was delayed slightly by under-
feeding either during the "initiation" or the ”promotion" phase. The
majority of the tumors appeared between 10 and 11+ weeks after the DMBA
was applied. Table III in the appendix gives the complete results of
this experiment.

The tumors observed were sessile and pedunculated papillomas. The
same methods of recording the data were utilized as in Experiment 1. The
pertinent observations in this experiment are: (l) thimracil, in conjunc-
tion with underfeeding, increases the tumor yield in mice when given
during the "initiation” phase, (2) underfeeding retards tumor yield when
given during the "promotion" phase. (3) a relationship exists between
final body weight arr! tumoryield. These observations will be discussed

in a subsequent section.

III . EXperimen't 2b

A. W
Because a decrease in food intake decreased tumor yield. it was

of interest to test the effects of underfeeding on established. benign
tumors. We also wanted to test the effects of increasirg the food in-
take on tumor developnent in mice which received limited food intake

for 20 weeks after carcinogen application.

31

ounce wuuudonuuoesa won cheese n consanam

sedan maaaosoumsae

sedan weaasapecHeo

scapeoaamge damn seams ammo: om vacuooou «gossa-

 

 

i m4. is. 03m New .22.: coo.— Qm mm .a
CV Haoefiofie
mm 0.3 mom 9mm cram age soon Cm mm .m
mm Tl. mm; «on m.nm :33 soon 3m mm .m
fl m . m d: 0. mm o. am 38:80 am . H
chosen. 9:) menacing mom {cheese .3; cm and; 00.3- .on
one: m cheese aeooa Aemv .z.m .>4 «assesses Henry gnome

 

“mam 5590205": Hun. GZHNDQ azHQEMHQHD ho and .flmdmm sazHadnaHnH:
Hus azHMDn AHodeona nz4_¢zHQMthquD mo ozHQfifihMHnnD ho maoafihfl .N Hands

Total Tumors

32

260

 

2l+o - I
220 -
zoo . '
180- /
160 ~- ,
11m -
120 a
mo ~
80 ~

6OJ

 

20.. “7’ 22/} Food (P) .. / /
/

’ ' —-——-——_—_———

 

A
1 l l 1 l '

6 s 10 12 it 16 18 20
Weeks After MA
Figure 3. Tumor Yield of Mice Subjected to Treatment

During “Initiation“ Phase (I) or “Promotion“
Phase (P).

 

33

The mice from Experiment 2a which appeared in good health, were
used for this experiment. The control group was continued exactly
as during the previous 20 weeks after DEA application. and served as
the control group for this experiment which lasted for 12 weeks. Group
3 of kperiment 2a, which received 2/ 3 feed plus thiouracil durirg the
'initiation“ phase and was fed ad W for the duration of that
experiment, was used to test the effects of limited food intake on
already developed benign tumors. Group 14 which was on restricted food
consumption during the 'promotion" phase of Experiment 2a was used to
see whether the tumor yield would increase if the mice were fed £51
libitum starting 20 weeks after the DHBA was applied. Weekly croton
oilpaintings were continued in all the mice in this experiment.

B. stasis
Table 3 shows the effects on tumor incidence in mice subjected to
altered food consmnption 20 weeks after DEA application. At the start
of this experiment, the mice in the control group had an average body
weight of 32.0 grams and had a total of ll+1 tumors. Titelve weeks later
the average body weight was the same and 137 tumors remained. Three
mice died that did not have tumors at the beginning of the experiment.
This is the reason for the higher percentage of‘mice with tumors at the
end of the experiment. \
The mice whichrwere on full feed and then subjected to 2/3 feed
at the start of this experiment had an average body weight of 32.0
grams art! a total tumor yield of 269 tumors when the experiment began.
Two weeks later the average body weight fell by “.8 grams and the total

number of tumors was decreased by 125. The percent of mice with tumors

34

was only reduced by 2. percent . The following two weeks showed another
A decrease in average body weight of 2.8 grams and a loss of 1+7 tumors.
The number of animals with tumors was reduced by 11 percent. The average
body weight stayed about the same but the total tumors decreased by 31
during the next two week period. At this point the average body weight,
total tumors and the percent of animals with tumors leveled off and
remained fairly constant for the duration of the experiment.

The group of mice that had been on 2/3 feed. and was then begun
on full feed at the beginning of this experiment. showed an abrupt
increase in average body weight for the first 2 weeks on experiment from
23.8 to 30.2 grams. while the number oftumors increased slightly from
11+ to 23 tumors . The average body weight of the mice varied somewhat
during the next 8 weeks and appeared to level off 10 weeks after the
experiment was initiated. The total number of tumcrs increased grad-
ually throughout the experiment, with a total of 101 tumors recorded at
the conclusion of the experiment.

Figure 1+ illustrates the average weight and the average number of
tumors per tumor mouse from the time DEA was applied until Experiment
2b was terminated 32 weeks later.. From 12 weeks after the DEA was
applied until the experiment was concluded. the lines plotting the .,
average body weight and the average number of tumors per mouse paralleled
the X axis.

Figure 5 illustrates the effects of underfeeding on benign tumors.
Four weeks of restricted food intake reduced the average number of
tumors per tumor mouse from 15 to 6.5. The average body weight fell

from 32'to 24.4 grams. The slope of the tumor curve from 20 to’zlt

35

weeks after DEA application is -2.1. The slope of the body weight
curve from 20 to 214 weeks after DEA application is -1.9. After 214
weeks, both curves leveled off and reversed their direction slightly
for the remaining 8 weeks of the experiment.

The effects on tumor growth due to increased food consmnption
20 weeks afteriDEA was applied is illustrated in Figure 6. The
average body weight at the start of experiment 2b, or 20 weeks after
DEA, was 23.8 grams. Twalve weeks after experiment 2b was begun, the
average body weight was 32.0 grams. The average number of tumors per
tumor mouse was it? at the beginning of this experiment and 10.1 when
the experiment was concluded 12 weeks later. The slopes of the body
weight curve and the average tumors per tumor mouse line were .68 and
.145 respectively for the 12 week period.

The control group which was fed ad mm for 32 weeks after
the DEA was applied, had 1 mamary adenocarcinoma and l conical skin
tumor. The rest of the tumors were sessile and pedunculated papillomas.
The two groups which had a history of restricted food intake from the
time DEA was applied. displayed only sessile or pedunculated papillomas.
It appears that some of these benign tumors are not autonomous and re-

spond to altered dietary conditions.
IV.' Experiment 3

11- Emsduras
The results obtained in experiments 1 and 2 indicated that there
might be a mechanism operating during the "initiation" phase which

influences tumorogenic response. In order to further test the effects

36

63.3%. eds 455 sofas execs om use.“ new on each n\m noun «do: 990.3 .39 u coo.— Hashes.

dogmas eds 455 .338 cases om coon Qm no use owes some enema u econ Qmee

.eesoe guesthouse won cacao» no Aeneas own—word s cases a? .bds

 

 

 

 

mm H.os Hos c.0n as m.o mm «.mm mm m.w and o.mn mm
or ~.m we o.mm He H.~ Ha o.~m mm o.w was n.mn on
on m.m Ha s.~m Hm o.m mm o.mm ea m.» was o.mm mm
mm s.m Hm m.mm ma m.m em 3.:m ea a.» med m.an am
an m.m mm m.om on m are m.~m ea m.~ ems ~.Hm mm
as 2.: as s.e~ as as see e.ee as e.e ass e.ee as
u. .s s m .s m... .s M, s .s
m m s a a. m I n a m I m
m n m. . m or m. . m e m
m m u m m u m Me n cogeoaammd 4min
sosuseesoo:
Sauces H2 sauce.— n\m 39380
non ma amazes new» munmmwwmmac"muuuecmnumummmmnmawmwmmmm .n names

37

wins £908 099:9“

mm

«as o» eosooncdm son.
a.“ eases. weaueomuwoasa son a base amassed ens page: boom oweuohd no man-e033;

33.33.54 nomads—"930 .334 exec: I 03.3..

mm .zm om 0H NH

I D P b h

who
“:2-

4. 0.33m

di-

 

mm

 

«p-
‘D
up

4 a . q .

-F
d
--

\0

.. «some: seem omesoi

  

 

 

ma

we

Ironing; eSueav

38

lusts; fipog eSereav

so“: we museum Hosea
no «man no sensuoaaaeq.sose4 asses om oneseH educate wessoasseom so essence .m seamen

doaaeouamgd mowoaaoneo no»H4 exec:

mm am am on we

APSE.
I
...:

O

‘i-
O

 

.1
.4
di-
db

unl-

—
-

  

onwaox_heom owdwobd I

   

ammo: house you unease cmeuoa4.|/ \
r / \ l
A \

 

 

 

 

ma

sromnm afieaeav

Average Tumors

em

Hm

 

 

.. swan-me teem:

‘8

\

\

L\ r p

p
b

 

\ .. veg-ho acne: New 553. zone.
. -

inlay neon «0033

b. b F

um

-uo

L

 

mm

 

2. .m an

no m:

um

mm

Average Body Weight

teem. knees. owner—omen Pentagon

apnea. m. unnece- ob Halo». 9.2.3 5 Eco ob Hosanna!» 0.3.8.5 Hues—no Hen b..— 5.595
no mean. >38. 99» happenings.

 

of altered nutritional and hormonal balances durirg the "initiation“
phase. a third experiment was begun with the purpose of determining
what these effects might be. Underfeedirg during the ”initiation'
phase in the first experiment produced a definite increase in tumor
yield. but a decrease in tumor yield in the second experiment. The
degree of underfeeding may have influenced the two groups in a different
way. 1.9.. it may have been more severe for one group than for the other
group, due to the fact that these 2 experiments were performed in
different rooms ard the conditions were not the same.

In light of the above. one group of mice was fed only 1/2 as much
as the control group during the “initiation" phase. Another group was
given 2/3 as much food as the control group during the same period of
time. If an increase in tumor incidence occurred in mice which were
underfed during the "initiation" phase. would this increase occur if
the mice were kept on 2/3 feed for the duration of the experimant? (‘ne
group of mice was used to explore this possibility.

Thyroxine not only increases the metabolic rate of the animals
but also dilates peripheral blood vessels. increases blood flow and ‘
stimulates hair growth in mice. For this reason. it was suspected that
DEA may be relieved faster from the skin of mice given thyroxine for
10 days before and 10 days after application. To insure that the
effects of thyroxine were only acting while the DMBA was in the skin.
one group of mice was underfed for 10 days before and 10 days after
DEA application. and thyroxine was added to their feed during the 10
days before DEA application. Another group of mice was used to try
to repeat the effects observed when thiouracil was given in conjunc-

tion with underfeedirg for 10 days before and 10 days after DEA.

1+1

Hydrocortisone inhibits hair growth. reduces skin thickness and

increases the catabolism of protein in the body. The effects of this

hormone were therefore determined on the "initiation" phase of carcino-

genesis .

One group of mice was fed 50 milligrams of ludrocortisone

acetate per kilogram diet for 10 days before and 10 days after the car-

cinogen was applied. This groupwas otherwise fed Ad 1m.

Seven groups of 25 mice each with an average bow weight of 23.3

3 .1 grams per mouse. were used in this experiment. and they were treated

as follows:
1. Controls - no treatmnt.
2. 1/2 of the controls' average daily food intake for 10 days
before and 10 days after DMBA application.
3. 2/3 of the controls' average daily food intake for 10 days
before and 10 due after DEA application.
it. Ibid. plus 2 grams thiouracil per kilogram diet iii conjum-
tion with the unierfeeding.
5. Ibid. plus 0.5 milligrams tivroxine per kilogram diet for 10
days before DEBA application.
6. 2/3 of the controls' average daily food intake for the dura-
tion of the experiment.
7. 50 milligrams per kilogram diet of hydrccortisone acetate
for 10 days before and 10 days after DEA application.
Bo Results

In Table it. the results of this experiment are smarised. The

beg inning and final average body weights. the total number of tumors per

tumor mouse. and the percent of mice with tumors are tabulated for the

1H3
!‘_.

time 17 weeks a-fter the DHBA was applied. In the previous eXperiment it
was found that the majority of tumors appeared from 10 to 1“ weeks after
the carcinogen was applied. In view of this. it seemed legitimate to use
the data available at 1? weeks after DMBA application. Group 1. the
controls. had a total of 113 tumors and 86 percent of the mice had tumors.
Group 2, which was fed only l/2 that of the control groups average daily
food intake during the “initiation” phase had fewer tumors than the
controls. and this was statisticalr significant at the 5 percent level.

Group 3 which received 2/3rds of the controls' food intake during
the 'initiation" did not show a significant difference from the controls
in tumor yield. Group 4 which received 2/3rds of the control groups
average daily food intake plus thiouracil during the "initiation" phase
was not significantly different from the controls at the 10 percent level.
Thyroxine administration for 10 days before DMBA application together
with 2/3 of the controls' average daily food intake for 10 days’before
and 10 days after the carcinogen was applied (group 5) . increased tumor
yield. This increase is significant at the 10 percent level but is not
significant at the 5 percent level. Group 6. which received 2/3rds feed
for the duration of the experiment had significantly fewer tumors. Group
7. which received hydrocortisone acetate during the "initiation“ phase
did not show a significant difference from the control group.

The type of tumors observed were again sessile and pedunculated
papillomas. The same methods of recording data were employed as in
the first 2 experiments. The pertinent observation in this experiment
is that fairly severe alterations with hormones or underfeeding during

43

the time DMBA is in contact with the skin ("initiating" phase) does not
have a consistent effect on tumorogenesis. In fact, it may have no

effect whatsoever. ‘ This will be discussed in a subsequent section.

141+

483 mean chad S can shame.» as 0a .333 33:32:. Ions.

duodhaoana u onaaea

amxn enemas ease on not need» eenwoeeaa u use.

dogcczmmc 483 mean execs NH daemons." unease...

 

 

a: m.m mm m.nm m.mm Aaoneeuanv amen n\m mm .~
aw . H.~ nHH m.wm n.mm “av eaoeaeeeeoneam ma .0
m~ m.~ ona ~.mm n.nm eeeaHv ona + eons m\m :m .m
00H m.~ and o.~m «.mm e.ma + “Hy seen n\m Hm .:
mN m.: mm ~.~m :.nm aHv coon «\H om .m
Np n.: mm m.mm ~.nm oAHv eepn n\m mm .m
mm m.w mam m.mm n.nm .Heesmeo am .a
chosen. n3: ease: noes.“ access .3; 5 :3» no“: .on
and. a mom unease $309 Aamv .m.m $4 aqeapceua and: 93.8

 

uHmfizmoomoxDa HHMm ho Ham cczuadHaHzH a ozHMDn Refined
amomHBmooomE Mo AHoaona a: ozHgas no 232% he mag .: g

DISCUSSION
I. Role of Nutrition in Tumorogenesis

In the 4 experiments presented. a reduction in food intake during
the "promotion" phase significantly reduced tumor incidence. The
”promotion" phase is defined as the time after the carcinogen is re-
moved from the skin and the "domant tumor dens" are dividing to
produce observable tumors. That insufficient nutrients are available
to grow tumors does not seen to be the reason for the reduction in tumors.
Tumors were seen to grow in spite of body weight losses by the host
animal. It is also known that increasing specific nutrients without
increasing the caloric value of the ingested diet does not increase
tumor incidence (Tannenbaum. 1953).

Bullough (1950) postulated that the mean mitotic activity of a
tissue. including the latent tumor cell. must be high in order for a
carcinogenic response to occur. mergy in the form of carbolqdrate or
carbohydrate intermediates is required for this mitotic activity to take
place. The decreased tumor incidence observed in animals which are
underfed my be due to a reduced mitotic activity. A contradiction to
this theory is apparent in considering the effects of thyroxine on tumor
growth which is discussed in a later section.

Chronic restriction of caloric intake depresses endocrine func-
tion (Heites. 1953). The secretion of thyroxine is certainly reduced
in a mouse on restricted food intake. It has been shown however. that
3 mice given thiouracil. which inhibits the secretion of thyroxine. have

an increased rather than a decreased tumor incidence (Heites, 1958).

146

During periods of severe starvation the adrenal cortex hypertrophies
and secretes greater than normal amounts of glucocorticoids while mild
reductions in food intake reduce adrenal cortical function. Baserga and
Shubik (1951+) and Sulzberger. gt, 3],. (1953) observed a def-rinsed incidence
of skin tumors (in mice subjected to hydrocortisom. Tannenbaum's (1953)
data and the results presented here inlicate a limar relationship be-
tween caloric intake and tumor yield. Since the tumor incidence is
linear while adrenal function is decreased in response to mild inanition
and then increased in response to severe umerfeedirg. the secretion of
adrenal glucocorticoids does not seem to be the principal factor involved
in decreased tumorogenesis when mice are subjected to underfeeding.

Other effects of reduced food intake are: a decrease in metabolic
rate, decrease in metabolic pools in the tissues and organs. and alter-
ation of the levels of tissue and body fluids (Maynard. 1951). Some
of the specific effects on the skin in underfed aninmls are redmed blood
supply." thinning of the epithelium. dry scaly appearance. and reduced
hair growth (Keys, 1950). In vmw of these observations. we can assune
a reduction in turnover of metabolites in the skin of mice subjected to
underfeeding. Therefore . the energy requirements of the epithelial
cells is also reduced. In accordance with Harburg's (1956) hypothesis. I
when more energy is required by cells previously subjected to a chanical
carcinogen. the greater is the tumorogenic response. when the carcinogen
is applied. some of the energy producing components of the cell are
destroyed. If complete oxidative capacity is lost a cell will die.

If the energy requirement of a tissue is high, sane mechanism

mustoperate to supply‘ this energy. werburg's hypothesis states that

47

this energy requirement is met by increased glycolysis in cells whose
respiratory system is partially destroyed. This fermentation does not
maintain structure in the cell and an abnormal growth, the tumor.
results. Because the energy requirements of the epithelial cells in
mice subjected to underfeeding is redmed, the oxidative mohanism
remaining in the cell subjected to the carcinogen may be able to supply
sufficient energy for normal vital processes. The increased fermenta-
tion is not needed and therefore fewer tumors develop.

Caloric intake pg; .53 is not believed to be the complete explana-
tion for a tumorogenic response. Tanenbaum ard Silverstone (194%) demon-
strated by different procedures that a redmed tumor incidence may be
attained despite equal caloric intake. They subjected mice to a cold
room without increasing caloric intake and observed a decreased tumor
incidence when compared to a control group on equal caloric intake. Sim-
ilar results were obtained by the same authors (19490) in a group of
animals given dinitrOphenol. Tumorogenesis involves not only caloric
intake and metabolic turnover. but also the resultant body weight of
mice. Our experiments show a definite relationship between fiml body

weight and tumor incidence.

II. W.m MW'MMW

The only obvious difference which might explain those results
which were conflicting in the first two experiments is the final aver-
age body weight of the mice. In order to test if a definite relation-
ship existed between tumor incidence and body weight, the results were

subjected to a statistical correlation analysis. All the data from the

1&8

first two experiments were accumulated. He chose to test both the
average number of tumors per tumor mouse and the percent of animals
with tumors. Table 5 gives the data which were utilized in the statis-
tical correlation.

Figure 7 illustrates the regression line existing when the average
number of tumors per tumor mouse is plotted against average body weight.
The slope of this line is .778 and is significant at the 0.1 percent
level. The correlation which exists between average tumors per tumor
mouse and body weight is .853. This correlation is also significant at
the 0.1 percent level. If the correlation value is squared we can find
out exactly how much of the variation in average tumors is due to body
weight. This means that 72 percent of our variation in the average
number of tumors per tumor mouse is due to final body weight. A 90
percent mean confidence belt is included in Figure 7. This means that
an average of all the means of am experimental groups of Swiss mice
under the same experimental conditions. would have an average tumor
yield between the two lines 90 percent of the time for their specific
average body weight.

In Figure 8 an individual average 90 percent confidence belt is
placed around the regression line .. This confidence belt implies that
for a single group of mice under the same experimental conditions. the
average tumor yield per tumor mouse for a specific average body weight
would exist between these two lines 90 percent of the time. If the
group of mice did not fall between these lines. we could be reasonably
sure (10 percent error involved) that other conditions were operating

to alter the tumor picture.

.In order to prove that body weight not only affected the number of
tumors developed in individual mice, but also the number of mice acquir-
ing tumors. the percent of mice contracting tumors was correlated with
average body weight. The regression line is shown in Figure 9. The
slope of the regression line is 6.1411 and is statistically significant
at the 0.1 percentlevel. The statistically significant correlation is
.908. which if extended by squaring implies that over 82 percent of the
variation existing between the percent of animals with tumors can be
attributed to final body weight. A mean 90 percent confidence belt is
plotted around the regression line in Figure 9. Figure 10 illustrates
the 90 percent individual confidence belt. A complete listing of the

statistical formulae used in the thesis is given in the appendix.

111- mmmmmnmmmwm

We observed that tumor regression occurred when food intake was
reduced 20 weeks after tie carcinogen was applied. Milk and egg produc-
tion are also inhibited by underfeeding. The actual mechanism for this
response is not known. hit the protein synthesis mechanism may be in-
volved in some way. If underfeedirg can cause an inhibition of these
processes we would expect a reduction in tumor growth. Regression of
observable tumors may be due to an increased catabolism of proteins. The
effects of severely reducing caloric intake result in a decrease followed
by an increase in glucocorticoid production. The glucocorticoids have
the capacity to increase catabolism of protein and this would result in
a decreased growth of tumor tissue. Benign tumors are known not to be

as autonomous as malignant tumors (Greenstein. 1951+) and they respond

rs r-x

TABLE 5. EFFECTS OF UNDERFEEDING OR UNDERFEEDING PLUS
HORMONES 0N TUMOR GROWTH

(Data Subjected to Correlation Analysis?

 

 

2...... 52..“ “2.3%“ mm. 31:21:32..
Controls 2h 32.0 8.3 71
Controls 25 28.3 H.8 72
2/3 Feed (1)° 21+ 29.8 9.8 92
2/3 Feed (I) 22 30-2 7-7 69
2/3 Feed (r) 22 22.u 2.h 23
2/3 reed (P) 22 2h.o h.6 1h
2/3 Food + Thiouracil (I) 22 32.0 15.0 82
2/3 Food + Thiouracil (P)°° 23 ‘ 23.8 h.3 23
2/3 Feed + Thyroxine (I) 21 27.9 u.6 67
2/3 reed + Thyroxine (P) 17 18.u o 0
2/3 Feed 4 STE (1) 21 28.9 6.6 80

 

iAll data taken 20 weeks after carcinogen application
"Equal beginning weights for all groups
I° . “Initiation! phase treatment 10 days before and after DMBA application.

P°°'= “Promotion! phase treatment started 10 days after DHBA application.

:m

J anom oogdfifiuoo 30:
mom no“: unmask boom weekend on ammo: moans mom among owcuobdéoncgsaoumoo .N enema

"3mg: awarded
an on mm Lmh mm mm om , we

 

 

 

. flee eonoeCnoo one: mom #1:}---

e _ _ _ _ l . o

 

03A noun-sung 5. z.”

 

 

saomng eSsaeav

paom eonmcfimnoo Hdfidabadfim

mom new: unwaom moon omnmce< on case: moans mom enemas swamped mo nowacaouuoo .w emnmam

pawns: ewdmo><

 

 

 

 

 

rm _. mm . @n _ Mm _ wm _ Am .. mm b 9m ma
d J _ . _ fi _ _ ‘\q H - _ a n O
\JU:
0;! w

{I
\ :NH

\ fiem oodonfldoo Haemeeefi mom uuuuuuuuuuuu
.\ \ \ coma noaauoHMem .u
\\ \ \ f

 

ma

exomnm eSsaenv

53

. 3.5 322.38 .3»
o . 0
53.. mom a»? 22o: eon emote: oe Sousa an? eon. enooeom no deepest o m E

onmaem owdmobd

 

 

 

 

m m an
e a a. a". a... m a . o
.
1r
-Iom
it
to:
tow
T
flee 883.38 nee: mom 3:31;-..3- tow
scan mode-emmem
.1

 

 

OOH

Sb

seem ooneeaenoo managemenH

 

 

 

 

&om near pmmfiem 3cm emcmebd op Pecans 5.? ea anoomem no noandaeuuoo .0." 3de
unwaor cwumobd
en an on mm mm am mm om we
1 x L. .e x w t. l n n “x J. n x o
Iom
no:
\ [low
\
\ .T
tow
\ \ omen ooeeeaeeoo Heoeeeaenu mom 3. ..
\ 2H3 moanaomwem
\ ooa

 

 

 

55

to altered biological emironments.

The increased tumor yield observed when mice were fed ad mm
after a prior 20—week period of underfeeding, can be explained by the
idea that the I'dormant tumor cells" still exist in the tissue of the
mice. Thus the increased food intake has the same effect on these
cells as if full feed were administered shortly after the carcinogen
was applied. The fact that "dormant tumor cells" remain in the skin
was shown by Berenblum (1952). He found that the interval of time be-
tween carcinogen application ard commencement of croton oil paintirgs
had no effect on the tumor incidence. The time at which tin tumors
appeared after the first painting with croton oil was identical. we
also noticed the greatest increase in tumor growth 10 to 12 weeks after
full feeding was begun. This is approximately the same time the great-
est increase in tumor growth occurred when mice were started on full
feed one week after DMBA was applied.

One mammary adenocarcinoma and one conical tumor were seen in the
group fed _a_d_ M for 32 weeks. Only sessile and pedunculated papill-
omas were found in the other groups which were underfed. This suggests
that the progression of tumors is reduced when mice are subjected to

underfeeding sometime after the carcinogen is applied.

Iv. ._.'LeRo ‘aflmezcxineandmiw inlanmasmis

Our experiments show that thiouracil. increased while thyroxine
inhibited tumor growth in the underfed mice during the "promotion"
phase of tumorogenesis. We also observed that the weight of the mice

given thiouracil was increased while the mice given thyroxine had a

56

decreased body weight as compared to mice on reduced caloric intake
alone. The report by Heites (1958) indicates that the resultant body
weight of these mice was not the sole factor responsible for the results
seen. When mice were fed 31 M. Heites noted that thyroxine
decreased tumor incidence despite an increase in average body weight.
and that thiouracil increased tumor yield despite a decrease in body
weight. Berenblum (1954) suggested that the "promotion" phase of tumor—
ogenesis is essentially a delayed cell maturation process. The dormnt
tumor cells. without maturing, divide and accumulate. eventually acquir—
ing the characteristics of a tumor. Since thyroxine increases while
thiouracil decreases the rate of cell maturation. Meite‘s postulated that '
the incidence of skin tumors may have been decreased or increased accord-
ing to the rate of cell maturation.

The hypothesis of Bullough (1950) that an increased mitotic rate
is associated with tumor growth does not seem to be applicable in this
situation. Thyroxine increases while thiouracil decreases mitotic activ-
ity; yet the tumor incidence was decreased and increased reSpectively.
This indicates that other factors than mitotic activity are involved in
tumorogenesis. This also provides some evidence that the chromosome
in the nucleus of the cells are not altered by the carcinogen. ‘If a
somatic cell mutation was irximed by the carcinogen. the tumor response
should be greater in subjects with increased mitotic activity. ,Like-
wise the tumor incidence would be expected to decrease in thiouracil-
treated animals since the mitotic activity was decreased.

In mice given thyroxine. the blood supply is increased. the

peripheral vessels are dilated. skin thickness is increased and hair

DY

growth is stimulated (Turner. 1955). This indicates that the epidermal
cells have a higher energy requirement than mice not given thyroxine.

In view of‘warburg's hypothesis, we would then expect an increased tumor
yield in mice given thyroxine. However. Guyton (1956) states that the
oxidative enzymes in thyroxine treated animals are increased quantita-
tively. This increase in oxidative enzymes may increase the respiratory
capacity of the epidermal cells. In view of this possibility. the fermen-
tation process need not be increased and the cells can retain their
structure. thus tumorogenesis would be inhibited. Thiouracil. by decreas-
ing thyroxine secretion. reduces the quantity of oxidative enzymes in

the cells. Thus an increased fermentation supplies energy for the vital
processes. but structure is not retained and neoplastic growth results.

Of course. this is only an hypothetical explanation of the results
obtained by altered thyroid function. Finally. it should be mentioned
that thyroxine or thiouracil can alter adrenal. gonadal and pituitary
function. and thus some of the effects of these substances may be

mediated through these glands.

V. R_clacfiha_i_iatiun_”lnt ' "aial.___ana_i_n_"Pr t0"1_°hassinlumamaensai§

Our results suggest a specificity of action by the carcinogen in
tumorogenesis. By altering the thickness. metabolism. blood supply and
available nutrients of the skin by changes in caloric intake or hormone
levels, we were not able to show a definite change in tumor induction.
The 3rd experiment demonstrates that despite the altered conditions of
the skin. tumor yield was not changed significantly. If the average

body weight of the mice in Experiment 3 is corrected to coincide with

58

the average body weights observed in the first 2 experiments. the tumor
incidence falls within our confidence belt around the regression line

in our correlation between tumor incidence and average body weight.‘When
the mice were starved. hr given thiouracil or hydrocortisone. the number
of cells available for contact with DMBA was decreased. Thus in effect.
there was a higher concentration of carcinogen per cell and a higher
tumor incidence could be expected. ‘When thyroxine was administered the
cell number increased and therefore a lower concentration of DMBA was
present per cell. ani a reduced tumor yield could be expected. When
these treatments are restricted to the time DMBA is presumably in con-
tact with the epidermal cells. we did not have a significant difference
in tumor response. The action of the carcinogen therefore must be very
specific. and is unrelated to metabolic functions in the skin at the
time of its action.

The tumorogenic reSponse is apparently dependent on the conditions
existing in the "promotion" phase rather than onthe corditions existing
at the time the carcinogen is applied. Thus redmed caloric intake.
thyroxine or thiouracil definitely altered tumor incidence during this

phase of tumorogenes is .

59‘

SW AND CONCHBIONS

1. Berenblum and Shubik (1947) suggested a two-stage mechanism
for carcinogenesis. a short “initiation" phase at which time normal
cells are permanently altered to become dormant tumor cells. and a
longer "promotion" phase during which tine the domant tumor cells
divide and eventually become visible tumors. The purposes of the
present study were to determine the effects on chemically-induced skin
tumors of: (a) underfeeding during the "initiation” am the ”promotion"
phases. an! (b) underfeeding and homones during both phases.

The carcinogen used in fliese experiments was a 0. 5 milliliter
solution of 0.5 percent 9.lO-direthyl-l.2-benzanthracene (DEA)
dissolved inbenzene. and was applied over a circumscribed area in
the interscapular region over the back of mice. This was followed by
once-weekly paintings with 5 percent croton oil in.liquid. paraffin.
DEA was used as an 'initiating" agent and proton oil as a 'pranoting"
agent. The hair was clipped from. the backs of albino Swiss female
mice of the CPI strain prior to application of-- the DEA or croton
oil. Attempts were made to alter the "initiation" phase by underfeeding
or givirg hormones for 10 days before and 10 days after DEA application
and similar attempts were made to alter the "pmmotion" phase by treat-
ments beginning 10 days after DEA application and continuing for the
remainder of each experiment.

2. Experiment 1. when 2/3rds of the average daily food intake
of the controls was fed to mice during the "promotion" phase. tumor
growth was inhibited significantly. Some tumors developed despite a

loss in body weight. When 2/3rds of the controls' food intake was fed

to mice during the "initiation" phase, tumr yield was increased as
compared with the control group fed £51 M. Thyroxine (0.5 milli-
gram per kilgram diet) inhibited while thiouracil (2 percent in diet)
stimulated tumor growth when fed to mice given 2/3rds or the controls'
good intake. when compared to mice given 2/3rds feed alone. A relation-
shipwas found to exist between final body weight and tumor incidence.

3. Experiment 2a. When 2/3rds of the average daily food intake
consumed by the ad mm fed controls. was fed during the "promotion'
phase. tumor growth was inhibited. Feeding 2/3rds feed alone durirg
the 'initiation" phase did not show a significant difference from the
controls' tumor yield. Two-thirds of the controls' food intake to-
gether with thiouracil (0.2 percent in diet) during the "initiation"
phase increased tumor yield significantly. A relationship was again
observed between tumor yield and average final body weight.

ll». Experiment 2b. The same mice were used in this experiment
as were used in Experiment 2a. 20 weeks after DEA application. The
control group was continued without treatment. A group which had been
fed ad W for 20 weeks. was put on 2/3rds feet thereafter (for
12 weeks). Another group which was restricted to 2/3rds feed for 20
weeks ,was fed ad libitum thereafter. The group initially fed 3d
1m food intake had an average body weight of 32.0 grams art!

269 tumors at the end of 20 weeks post DMBA. Four weeks later. they
had an average body weight of 23.4 grams and only 97 tumors. The ave-‘
rage body weight and tumor incidence remained practically unchanged
for the remainder of the experiment. 32 weeks after DEA application.

The group fed ad mm. after the previous 20 weeks on restricted

61

food intake. gained an average of 6.4 grams but showed an increase of
only 9 tumors during the first two weeks on experiment. The average
body weight remained at about 32.0 grams 10 weeks after ad 11m food
intake was initiated. The tumor yield increased slowly and an abrupt
increase in tumor yield was noted 10-12 weeks after ad libitum feeding
was initiated.

5. Experiment 3. There was some indication of an altered tumor
incidence as a result of hormone treatment or caloric restriction during
the "initiation' phase in the first two experiments. Attempts were made
to confirm these firdings in the mice of this experiment. For 10 days
before and_10 days after DMBA application, mice were fed 1/2 of the
controls' average daily food intake, or 50 milligrams per kilogram diet
of hydrocortisone acetate, or 0.5 milligrams thyroxine per kilogram
diet ard 2/3rds feed. or 2 grams thiouracil per kilogram diet and 2/3rds
feed. The results showed that none of these treatments influenced
tumorogenesis significantly.

6. A correlation analysis on the mice of the first 2 eXperi-
ments was performed to test. the relationship between tumor incidence
and average body weight. The correlation coefficient. .r, equaled .85
when average tumors per tumor-bearing mouse was related to average body
weight. The r value for the correlation between percent animals with
tumors and final body weight was .91.

7. It is concluded that: (a) final average body weight is a
significant factor in chemically-induced skin tumorogenesis. (b) neither
caloric intake nor hormonal treatment influences the "initiat ion" phase

of tumorogenesis, (c) durirg the ”promotion” phase of tumorogenesis,

62

restricted food consumption inhibits tumor growth, while thiouracil
and thyroxine increase and decrease tumor development, respectively,
d) mice with a small number of tumors after 20 weeks of underfeeding,
show an increase in tumor yield when placed on ad libitum feeding
for 12 weeks; mice with a large number of tumors after 20 weeks

of ad libitum feeding, Show a pronounced decrease in tumor yield

when placed on reduced food intake for 12 weeks.

KW

Bather. R.. and Francks. w. R. Further studies on the role of thyroxim
in chemical carcinogenesi. Can Rm 12:2147-248, 1952.

Baserga, R., and Shubik, P. The action of hydrocortisone on transplanted
and irdmed tumors in mice. C 11mm}; 11+: 12-16. 1951+.

Begg, R. H. Steroid hormones aid tumor-host relations. Qanaa; M
11: 5406MB. 1951.

Berenblum.I., and Shubik. P. A new. quantitative approach to the stat
of the Jstages of chemical carcinogenesis in the mouse' 3 skin.

Exits}. .m 1: BBB-396.19”?

Berenblum, I. am Shubik. P. An experimental stuth of the initiating
stage of carcinogenesis. arfl a re-eucamination of the somatic cell
mutation theory of cancer. m. J. m 3: 3814-386. 1949.

Berenblum, I. mm m. The Johns Hopkins Press. Baltimore..
Md.: 20-21. 1952.

Berenblum. I. A speculative review: The probable nature of promoting
action and its significance in the understardirg of the mechanism
of carcinogenesis. 93m: 13m 1“: #71477, 1951+.

Bielschowsky. F. Carcinogenesis in the thyroidectomized rat. gnu. ,1.
9am 12, 2: 231-23lb, 1958.

Bloom. F. Effect of cortisone on mast cell tumors (mastocytoma) of the

dog. he m- m. 1.3.1.91. mice. 79: 651-65“. 1952-

Bollough, W. S. Mitotic activity and carcinogenesis. k111- sl- 2am
“'8 329-336. 1950-

Boutwe11;'R. K” Rusch. H. P.. and Brusch, M. K. On‘the stimulating
effect of dietary fat on carcinogenesis. WW 9:
6o7-®80 1%.

Boutwell. R. K.. Bosch. D.. and Rusch. H. P. On the role of croton oil
in tumor formation. 93mg: W 17: 71-75. 1957.

Boyland. E., and Brues. A. H. The carcinogenic action of dibenzcarbozoles.
2.229- 391 he a- E- 122: ”29.1441. 193?.

Calcutt. G. The distribution of polycyclic hydrocarbons within the cells
of some mouse and rat tissues. gut“ .Qanm 12: 1149-160. 1958.

Conan. D. R. Talk presented at Detroit Institute of Cancer. 1958.

Cook. J. Wu and Hewett, C. L. The isolation of a cancer-producing
hydrocarbon from coal tar. _J_. Chgm. .S__d.: 398—405, 1933.

Crabtree, H. G. Anti-carcinogenesis. Brit. fl, Bull. 4: 345—316, 191W.

Cramer, A., and Stowell, P. Effects of carcinogens on cell constituents.

lo Eati- Qancanlnai- 2: 397-384. 191+2.

Darchun, V.. and Hadler, H. Metabolic and carcinogenic studies with
9,1g-dimethyl-1,2-benzanthracene. Caagar Eaaaazah 16: 316-323,
195 .

DeRobertis, E.D.P. , Nowinski, W. ‘51., and Saez, F. A. General C
(2 ed). W. B. Saunders 00., Philadelphia. Pa.: 314-320, 1951+.

Edwards, J. E. Hepatomas in mice induced with carbon tetrachloride. 3.
Nat. Cancer Inst. 2:197-199, 19lel.

Engelbreth-Holm, J ,, and AsboeJ-Iansen. G. Effect of cortisone on skin

carcinogenesis in mice. Agta. m. at Mjgngigl., MM
32: 560-564. 1953.

Fox. A. 3. Talk presented in Biochemistry Seminar, Michigan State Uni-
versity, 1958.

Gilroy, E. Comparison of the effects of arginase and thyroxine upon
tumor growth rate in the mouse. W. i. 21+: 1181-1187, 1930.

Grad. B. The influence of hyper- and hypothyroidism on the induction, of
lymphatic leukemia of AKR mice. Qaaaan am 17 : 266-267, 1957.

Greenstein, J. .P. W 3;; 93119.92: 2 ed. Academic Press, Inc.,
New York, N. Y.: 32.145. 195Llr.

Guyton, A. C. W a: W1 W, W. B. Saurders 00., Phil-

Haddow, A. Transformation of cells and viruses, Ham 154: 194-2093 1941+.

Haddow, A. Mode of action of chemical carcinogens. Brit. fled. Bull. 1*:
331-342. 1947.

Heidelberger, 0., and Jones, H. B. The distribution of radioactivity in
the mouse following administration of dibenzanthracene labeled in
the 9 and 10 positions with carbon 11+. Cagggr 1: 252-260, 1948.

Heidelberger. 6., Kirk, 14.11., and Perkins. M. S. The metabolic degrada—
tion in the mouse of dibenzanthracene labeled in the 9 and 10
positions with carbon 1b,. Gama: 1: 261-275. 1918.

Heiger, I. The spectra of cancer-producing tars and oils and of related
substances. 21mm. :1. 24:505-511. 1930.

65

Heilman. F. R.. and Kendall. E. C. The influeme of ll-dehydro-17-hydroxy-
corticosterone (commund E1) on the growth of a malignant tumor in

the mouse. W 34: 14-16—920. 19M.

Karpassy. B.. and Kovacs, K. Experimental liver cirrhosis in rats pro-
duced by prolonged subcutaneous administration of solutions of

tannic acid. Brit. :1. m. 23311.30: 266-272. 1949.

Karsner. H. T. Hanan W, J.B. Lippimott Co.. Philadelphia. Pa.:
640-685. 1950.

Keys. A. 1112 11mm 9_f_ film W. University of Minnesota Hess.
Minneapolis , Minna 612-629, 1950.

Kreyburg, L. The influence of intrinsic factors on the deve10pmmt of

induced tumors in animals. data. 2am. at Mia)... W-
and. Supp. 37: 317-338. 1938.

Kurschmann, H. Effects of starvation on the thyroid gland. Agta. dad.
m. 57: 2140-246. 1922.

Lacassagne. A. Appearance of mmmary cancer in the male mouse injected
with folliculin. 9.911119}.- Bagd. flzad. d. Sc. 195: 630-632. 1932.

Lerman. J. W at 1129922119 Diseases. Oxford University Press=
329-3379 1947. ’

Levine. M., and Hugel, V. H. Disturbance of the thyroid mechanism and its
effects on tumor growth in mice. an]. ,1. gander, 10: 817-828. 1933.

Maynard. L. A. am Mon. McCraw.Hill Book Co.. Inc., New York.
N.Y.: 300-310. 1951.

Meites, J. Relation of nutrition to endocrine-reprodictive functions.
loud and 92119.1: .4.- of. Science. 28: 194m. 1953.

Meites. J.. Ferg, 1.8.1... and Hilwerth, A.M. The effects of endocrine im-
balances on vitamin requirements. An. _J_. W M 5: 381-392.
1957.

Meites, 'J. Effects of thyroxine and thiouracil on induction of skin
tumors in mice by 9.lO-dimethyl-l.2-benzanthracene and croton oil.
Cancer Research 18: 176-180. 1958.

Miller. J. A.. and Miller. E. C. The presence and significance of bound
aminoazo dyes in the livers of rats fed p-dinethylaminobenzene.

fiance: W 7: 468-“80'. 19+?-

Horeschi, C. In: Tannenbaum, A. 113.9. Wait Qanaan. Hoeber-
Harper. New York. N.Y.: 398, 1953.

Morrison. F. B. Feada and Fgading. The Morrison Publishing Co.. Ithaca.
N.Y.: 1136—11141. 1951.

66

Battleship.A .. and Henshaw, P. S. Induction of pulmonary tumors in
mice with ethyl carbamate. “Bat, Beggar Ingt. :309-319. 1943.

Potter. V. R... Price. J. M.. Miller. E. C.. and Miller. J.A. Studies
on intercellular composition of rats fed various aminoazo dyes:
IV. Effects of succincxidase ard oxalacetic acid oxidase. Bans-g;

Emmi: 10: 28-35. 1950

Pullman. A. Electronic strmture and carcinogenesis activity of condensed
aromtic hydrocarbons. Bull. Ass. France m 33: 120—130. 1946.

Pullman. A. Electronic structure and carcinogenesis activity of aromatic
molecules. Bull. Ass. France. 9mg; 34: 245—258. 1947.

Rusch. H. P.. Boutwell. R. K.. and Brusch. M. K. The influence of various
dietary factors on the induction of epithelial tumors in mice.
Scientific Proceedings. 1949. reported in gangs; 11mm 93 .607.
1949.

Selye. H. The general adaptation syndrome and the diseases of adaptation.
Am... :1. lied- 10‘ 549-555. 1951.

Shubik. P.. Baserga. R.. and Ritchie. A. C. The life and progression of
induced skin‘tumors in mice. B211. 1. Cancg: 7: 342—351. 1953.

Silverstone. H.. arm. Tannenbaum. A. Influence of thyroid hormone on the
formation of skin tumors in mice. Bangor Baggazgh 9: 684-688. 1949.

alith. D. L.. Hells. J. A.. and D'Amour. I". 'J.'The relationship of the
endocrine system to carcinogenesis. 9am W 2: 40—44.
1%20

Spain. D. M.. Molomut. N.. arr! Novikoff. A. B. Cortisone and carcinogenesis
. in mouse skin. I. Effect of cortisone during multiple painting
' with methylcholantrene. Canaan Baggargh 16:138-141. 1956.

Stewart. H. L. In: WI gm, Hoeber-Harper. New York. N.
Io: 62-779 1953’ ‘

Sulzberger. H. 3.. Hermann. F.. Piccogli. R.. and Frank. L. Incidence.
of epidermal metlwlcholantrene tumors in mice after administration
of cortisone. Prgg. ”m. Big]. and Bed. 82: 673-675. 1953-

Tannenbaum. A.‘ Relationship of body weight to cancer incidence. m.
Bath- 30: 509-517. 19140-

.Tannenbaum. A. The genesis and growth of tumors. II. Effects of caloric
restriction. per se. 93.3.19: Maggy-112: 460-467. 1942.

Tannenbaum. A. The deperdence of the genesis of induced skin tumors on
the caloric intake during different stages of carcinogenesis.

fiance: Because}: 4: 673-677. 1944.

67

Tannenbaum. A. The dependence of tumor formation on the degree of
caloric restriction. Canggz Baggamh 5: 609-615. 1945.

Tannebaum. A.. and Silverstone. H. The genesis and gnowth of tumors.
IV. Effects of varyirg the proportion of protein (casein) in the

diet. Quasar. Baccarat 9: 162.173. 1949a.

Tannenbaum. A.. ard Silverstone. H. The influence‘of _caloricvrestriction
on the formation of skin tumors and hepatomas in mice. 9am
Research 9: 724-727. 194%.

Tannenbaum. A.. and Silverstone. H.. Effects of low environmental taup—
erature. dinitrophenol. or sodium fluoride on the format ion of

tumors in mice. m m 9: 403-410. 1949c. .

Tannenbaum. A.. In: W g: Baum, Hoeber—Harper. New York,
N. Y.: 392-437. 1953.

Tsutsui. H. Artificially produced cancroid in mice. 9m 123 17-21.
1918. ‘

Turner. C. D. mm W. W. B. Saunders Co.. Philadelphia.
Pa.: 28-34; 175-178, 1955. ‘

Warburg. Oé Anaerobic theory of carcinogenesis. figigngg 123: 309-314.
195 .

Weigert. F.. and Mottram. J. C. The biochemistry of benzpyrene: II.
The course of its mtabolism and the chemical nature of the metab-

olites. mm 6: 109-120. 1946.

'Iieinhouse. S. ma 1;; gang-.9; W. III. Academic Press. New
York, N.Y.: 270-327, 1955.

Wolf. G. n W Wm: ganggz. Harvard University Press. Cam-
bridge, “3830‘ 9‘37, 19520 .

Iamakiwa. H.. and Ichikawa. K. Experimental study of the mthogemsis of
carcinoma. _J_. mm 3:1-28. 1918.

Yoshida. T. Development of experimental hepatoma by the use of o-amino.
lazotoluem. 1:. Jan. Bath. _S_Qd. 24: 523-530. 1934.

APPENDUES

68

APPENDIX I

COMPOSITION AND ANAHSIS 0F RATION

69

7O

operatesz ospaeoowan Hosea n moat.

pocupxm monk cowouuwz I name

 

 

 

 

no mm m.: ~.m m we --- doaodm co m
o.sm mn.mm m:n.s m~.m mow.~ mm.~a ooa nausea
ooo.a a snow .eoH
“we e.m 0mm. as. 0mm. ow.m as are: Hao eooosaq
m.m o.a emm. as. one. ma.a m coco» onosonm
:.mm o.m owo.a so. oem.m mm.: om noesem can: each:
Heme mood
mwm n.m can. we.a mes. nH.H w seaeeaa.eoeeaeanon
0.0m m.~a mes. mm. mes. on.n mm econ: assess
s.mm m.sm omc. oe. mom.a Ho.n mm deco oases eddone
mam new new new new . new new
..zna same has noses can daeoonm “mummd neeaeoamnH

 

.wdfleooh one epoch

Amman .eo seam
.coaflago: sown nachdmde pcofieoamch

EOH: AdBZHSHmHmNE OB mam one<m.ho mHmHA4zd.nz< onanomxoo

.H HAM<H

APPENDIX II

RECORDED DATA OF EXPERIMENTS 1. 2a. AND 3

71

a/h
h./l

 

 

odesnom seesaw - mam ascensoase u ona seasonese n ma
cane: meencomnaossa mom unease owcnmpd u omdozna\a.>4ueu compoaowm u may donneapwda n H.
m: as m: mm on on m: mm mm as om m
m.: m.mm ~.s o.mm H.: m.mm m.m s.mm m.H o,mm w.H n.am “my came + comm M\m .N
o o o o o o o o o o o o
o s.ma o o.ma o m.ma o m.mm o H.mm o m.wa “meme a some m\m .m
mm mm mm as mm NH mm as ma : z a
:.m s.mm m.m m.mm s.m m.mm m.m m.mm n.a 0.0m o.a m.am seams econ m\m .m
on man we Hos as mm as me mm mm mm N
m.m m.wm m.w m.mm m.m :.mm s.m m.mm m.m w.wm m.a H.mm “as mam + eeoa n\m .:
em mm mm me me :e mm em an mm m n
m.: m.em m.: n.wm m.: o.mm m.e w..m H.m m.am m.a m.~m fiHV ma . econ m\m .m
mm mam ooH mmm mm ems 00a mma on mm mm as
m.o m.mm :.e m.mm w.e o.mm H.m m.mm w.m :.mm m.a m.mm same eons n\t .m
me am On an me me me mm mm me am ms
w.s m.mm 1.: ~.mm m.m H.wm s.m m.~m m.m m.~m m.m m.~m accesses a
Aev Ase fies any may Apv Aev on ,Aev Ase unease can: we anessa fiestas moose
nov Adv new Adv Ase Adv Aev Aev on Adv oaso:-a\a.os<o Aosuv.>.m.os¢r
execs om eases ma esooa.ma ease: as exec: NH ease: oH

 

ooaoooaamde amen scene asses

 

mus 20 mmzozmom mqu ozHQMRMMHQZD m0 GZHQMMhmMQaD ho maommmm

AH amenauomwa no mpasnomv

«man 09 nHBomHMDm NOH: 2H masomw mOZDB ho onmmmmwomm

.HH figm<s

73

.Socacoa I can!“
.35 nouns ghee o.” coups». unclean.» .ondam escape-cams Info:
.45: note :8 2 en- oaeuoo one 3 63am oneness—Hana:

.oudoa gugpunoac» won unease no woe—Ba emanate .. undo: u\a.>4o

 

 

mm 1: m a. D e S w E m. w n

11...“. 4mm m.n Emu m.m 9n... o.m New and 1.9 m4 win our: eon.— Qm .1.
mm mom we 3m 1 08 one. r: mm mm cm a

0.3 9mm 11.? men 5: :An fim don m6 mén as one Seconds + eon.— Qm .n
mm «mm i. mus B m: 2. R u: on . om 3

To Ram mé {an we won 06 non 1m wén man new LC coon Qm .m
i. .2: 2. 9: S 9.: am m2 am mos we mom

m.» 9% n.» :ém on flmm .1 men an mém we: a. zone—co .a
E E E E E E E E 3 E 5.2 s: a. case use:

3 3 Rev G: 3 A3 3 A3 3 Adv nonexiiaopdo 338.34. announces
execs on cases a: cases an some: E ass: ma ones: 3

 

dosedofimfia an: no»: case:

 

com amusing no 3385

453. ca 2838 no:— In magma moms“... ho Hon—”mg."
a no anom mDAm” anus—Egg mu uzwnmfihmflg ho was .HHH a

7a

.485 oaomoc. .5 OH 5&3 05858 n was:
.4nxn noses once on end ensues ease as .ooend eeoaeeaoaeHe.u “Hes.
easel nag nose non chic» no EASE owsuobd .- case: .195»:

 

 

am an em on mm m m a

m. u.n~ m.m o.~m m.a o.am o.a . m.a~ “consensus coca nxu .~
nu mas we and me o~ we on

H.~ m.w~ m.~ m.wm :.: «.mm m.m m.m~ any eeeeaenoeonenm .m
me and mm was me Hod mm . ma

~.~ ~.mm u.m n.mm m.m a.m~ m.m m.mm “Hecama a earn n\m .m
8” R: 8a and me n: mm on

m.~ o.~m n.~ n.~m a.m m.m~ u.m u.mm stoma . lav econ n\m .:
me mm on me ca Hp .mm as

~.: stem ~.: m.~m o.m u.mm m.m m.mm “as corn «\H .n
- mm mm mm we em as ea

n.: «.mm o.m ~.mm m.n m.:m m.: m.mm seams earn n\m .m
we mas am mm mm mm an an

m.m u.w~ ~.m ~.mm n.: m. m «.3 «.mm caeaeeoo .a
flee “as new “as Ace “as encasa.doan.ue coo-ea assess

any Adv any “3 Adv Adv sense: .9? . obdo cameo: . pee panacea.“

exec: 2 use... 1% Eco: ma exec: OH

 

BAA Gnu—nag 3 oannD ho mag 33H no mag n: a

doaoecaadca amen nona4_eaoor

an ode-«nods: no seas-one

nan-H8839... ho Ham cannahnnne a Rug Hnoxmom

APPENDIX III

STATISTICAL FORMULAE

75

STATISTICAL FORMULAE

I. Basic Conversion Formulae
x = X - i
Exz = 3x2 - E133
n

2

EXy=EXY-@%L§Y.)_

 

II. Linear Regression Formulae

 

 

 

 

’3? . a . bx
b :‘Exg (Slope of the line)
a g ELSE
V<Ex2><Ey2)
III. :near Regression Mean Confidence Limit Formulae
sy x LSyEyZnev -L%).2
= 53’ x % I32

= Y .. sY ’thig (Upper two-tailed 100p¢ conf idenc‘e belt
for the average number of tumors in a
population of mice for X average body
weight.)

u; = Y - SY t1? (Lower two-tailed 100p3$ confidence belt
for the average number of tumors in a
population of mice for X average body
weight.)

IV. Linear Regression Individual Confidence Limit Formulae

._ Ey2- )2

n-l

 

 

n-2

51?

'T'

syox V1 1-

77

 

Y- 3113159

u;=Y-3Ytl§n

1

x2

'h' 4- E372 (Standard deviation of an irdividual

estimated value of Y for a given value
of 1.)

(Upper two-tailed 100p‘,‘ confidence belt
for the average nunber of tumors in a
group of mice for X average body
weight.)

(Later two-tailed 100p$ confidence
belt for the average number of tumors
in a group of mice for I average
body weight .

V. Formulae for Significance Tests

A.

B.

CO

Testing two sample means (small sample method):

33;.

3112-91123 ., E12 (“>2

v

 

n1
n1+n2-2

n2

tax:-

Testing slope of regression line:

.g... wZ-iggié

 

n-2

sang...

1::

3!.
Sb

(n - 2 degrees of freedom to test
whether the slope is significantly
different from zero.)

Testirg correlation:

t-

1‘

11-2 (n - 2 degrees of freedom to test

l-r

whether r is significantly different
from zero.)

2.". V“: 1““? T Q:-

‘LJ-a. L352— 13ij
' '6

AUG 30 ism 339' KM
S‘FP 9Q 10‘“ 7??? ,