Download Preclinical Screening and Evaluation of Agents

Preclinical Screening and Evaluation of Agents for
the Chemotherapy of Cancer- A Review*
ABRAHAM GoH)m, JOHN M. VENDITTI, AND NATHAN MANTEL
(Laboratory of Chemical Pharmacology, and Biometry Branch, National Cancer Institute, Bethesda, Md.)
The current extensive programs on antitumor
chemotherapeutic screening and drug evaluation
are conducted mainly in rodents with transplantable tumors (7, l l , 13, 14, 19, ~4, ~6, 43, 45, 46,
48, 56, 57, 60-65, 67, 7~, 75, 78, 84-91, 94, 99).
Additional programs are being conducted with
spontaneous and carcinogen-induced tumors (1~,
48, 69, 79, 80, 96) and with tissue culture (8, 17,
18, ~1, ~ , 47, 59, 88, 95), microbiological (1, 5, 6,
16, 20, el, e3, 50-53, 55, 77, 8~, 88, 9~), and biochemical (9, 15, 49, 66, 71, 8~) systems, and systems involving development and differentiation
(3, 58, 81), etc. Numbers of compounds are passing
the screening and evaluation tests and are being
introduced into preclinical and clinical pharmacology and into clinical trial.
There is considerable interest in the question of
how well the screening and evaluation tests may
predict clinical usefulness (~tS, 73, 75, 83). Will activity in specific animal tumor systems correlate
well with activity in specific types of human tumor
systems? Can the animal systems not only detect
antitumor activity, but also rate drugs with respect to such activity ? Will the rating of the drugs
in the animal systems correlate with the comparative therapeutic evaluation of antitumor effectiveness in man (10~)?
Basic to any attempt to determine correlation
of animal testing and chemotherapeutic effectiveness in humans is the validity of the animal screen
or evaluation test itself. What information does
the animal screen or test provide? Is it actually
evaluating activity in the animal system? Is it
appropriately ordering the compounds in terms of
relative effectiveness? Are the criteria of measurement of therapeutic effect appropriate? ks the assay system paying sufficient attention to the hosttumor-drug relationship? This basic question is
* The authors are indebted to Dr. E. K. Marshall, Jr., for his
stimulation, interest, and many helpful suggestions in the undertaking of this review.
Received for publication March ~0, 1961.
prompted by the work of E. K. Marshall (68), who
pointed out that any evaluation of agents in infection chemotherapy should take into account the
triad of parasite, host, and drug. The current review was undertaken to determine what information the common animal screening and evaluation
procedures may provide. The parameters of drug
action examined are those employed in common
practice. In the analysis no attempt is made to be
exhaustive, but rather to review the types of coinman screening and evaluation procedures and their
meaning. I t is hoped that this may contribute to
the re-examination of ways and means of improving animal and related screens, so t h a t any attempt at correlation with results in humans may
be more meaningful.
T U M O R I N H I B I T I O N ASSAYS
In the study of the antitumor activity of drugs,
one may determine the effect of the drug on the
tumor without any regard to its effect on the host.
This procedure conceivably could be useful, if the
antitumor agents were nontoxic for the host. In
such case, the evaluation of drug effectiveness
could be limited to the determination of the
amount of drug required to produce a desired degree of inhibition or cure of the tumor. Or the evaluation could be based on the time or number of injections required to produce a specified antitmnor
effect, or the rapidity of occurrence of a refractory
state, or the degree of activity for a spectrum of
tumors, etc. Assay systems could, in the above
cases, be limited primarily to simple dose-response
studies with respect to the tumor, in the main
ignoring effects of the drugs on the host. Although
the drugs might differ in physiological disposition,
including transport, excretion, or metabolic alteration, etc., host toxicity would be removed as a
factor in the assay. The methodology would be
similar to t h a t employed in the in vitro tissue culture and bacterial systems, and the extent of correlation of results with these systems could be
determined.
1334
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN el al.~Preclinical Screening and Evaluation
From the point of view of therapeutic applicability, assay systems of drug effectiveness based
on antitumor effect alone pertain to too simple a
situation, since they presuppose that antitumor
agents are nontoxic for the host. All of the known
antitumor agents are toxic for the host as well as
for the tumor, and the toxicity for the host severely limits their usefulness.
If the tumor system employed for assay is extremely sensitive to drug treatment this could, in
effect, remove host-toxicity as a factor. In such
ease, since relatively nontoxic doses would exert
antitumor effect, it might still be possible to limit
the assay to evaluation of antitumor action of the
drug. Such a system may be provided by using a
tumor system empirically found to be sensitive, or
selection could be practiced for sensitivity. A tumor system could also be made more sensitive by
diminishing the size of the inoeulum (41, 44) or by
initiating treatment at an early time following
tumor implantation (~7, 135, 40, 78, 98). Such a
system could then be employed to compare the
relative effectiveness of drugs in inhibiting tumor
growth (57, 78, 98). In some instances, it may be
desirable to employ a relatively sensitive tumor
system for determining marginal antitumor effects of drugs. This may have a practical value-for example, in screening for drugs to be employed
as surgieat adjuvants.
IIowever, care must be exercised in the utilization of such assay systems. If two drugs completely inhibit early tumor growth at nontoxic doses,
which is the more effective drug? I t could, presumably, be the one, for example, which accomplishes complete inhibition at a lower dose. However, this might indeed be misleading if neither
drug were capable of inhibiting tumor growth in
mice with more advanced tumor.
Since the known antitumor agents are limited
in their usefulness because of their toxicity for the
host, an assay system in which drug toxicity for
the host is limiting may more closely reflect the
clinical situation. A standard procedure employed
in screening programs has been to measure the
extent of inhibition of growth of the local transplanted tumor after a specified number of treatments. Usually the animal is sacrificed and the
tumor is enueleated and weighed. Alternatively,
the local tumor may be measured in one or more
dimensions. Comparison is then made with the
size or weight of the tumor for untreated control
mice. The per cent inhibition of tumor growth is
thus determined and must usually exceed a specified figure in order for a drug to pass a primary
screen. The ratio t/c (fractional growth of treated
tumor as compared with control) is commonly
1335
employed, where t is the average weight or size of
treated tumors and c is the control average. In
order to take into account nonspecific toxicity for
the host it is usually required that t/c must not
only be less than a specified fraction, but that this
must be achieved with a limited animal weight
loss and/or a limited host mortality response. The
above type of procedure has been employed in primary screening programs to determine whether
drugs are of sufficient interest for further study
(7, 12, 19, ~4, ~6, 61-65, 7~, 84, 85, 87, 89, 90, 94).
By such a test system, the compounds may be
ordered with respect to the relative degree of effectiveness; i.e., comparisons may be made of t/c
at a maximum tolerated dose.
Where both host effect and tumor effect are to
be taken into consideration, the question arises as
to the choice of parameters for each. Toxicity for
the host is usually measured in terms of drug
lethality, or as body weight loss. The measure of
antitumor effect may be based on tmnor size or
weight, or on some other characteristic of the
tumor, such as protein content, etc. Cytotoxie effects or biochemical effects may be measured. The
growth curve of the tumor may provide a useful
measure of antitumor effect. Survival time of all
mice or, alternatively, only of mice not succumbing to drug toxicity, have been used as measures
of drug effectiveness. Where drug effects are extensive, or the tmnor challenge weak, the percentage "cures" elicited may provide an index of drug
effect.
SURVIVAL T I M E ASSAYS
Whereas routine screening procedures most
commonly employ tumor size or tumor weight as
the basis for evaluating antitumor effect, there are
some notable exceptions. For example, in chemotherapeutic screening with the LI~10 tumor, the
Cancer Chemotherapy National Service Center
employs an index of effect based on the prolongation in survival time of tumor-bearing mice (6165). Survival time as a measure of chemotherapeutic effect has been used by various investigators in
laboratory research with the tumorous host (11,
13, 14, 24, 48, 56, 57, 67, 69, 78, 80, 86, 96). Since
increasing survival time, rather than reducing tumor size, more properly reflects the purpose of
treatment, such extension would, in many circumstances, more appropriately describe the therapeutic efficacy of treatment. Accordingly, in this laboratory, assay systems have been developed in
which alternative ehemotherapics are compared
on the basis of the survival time of the tumorous
host (~27, 81, 40, 43-46, 99). Since, for each drug
on a specified treatment schedule, survival time
varies with dose level, an appropriate measure for
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
Cancer Research
1836
T U M O R SIZE VS. S U R V I V A L T I M E
If extension in the survival time of tumor-bearing animals is the motivation of experimental
chemotherapy, justification for using inhibition in
tumor size as a criterion, aside from perhaps its
simpler ascertainment, would be dependent upon the
existence of an underlying correlation between the
two measures of effectiveness. The question, then,
is whether or how tumor inhibition can be used to
select agents capable of achieving important increases in survival time.
a drug on a schedule is the maximum survival time
it elicits. At dosages beyond some optimal level,
toxic effects for the host outweigh antitumor effects so t h a t further increases in dosage lead to
reductions rather than extensions in survival time.
Use of the maximum survival time elicited with
a drug on a schedule as the measure of therapeutic
effect leads to certain modifications in laboratory
procedure. To ensure bracketing the optimal dose
level, a relatively wide dosage range should be
employed. Also, the interval between successive
<~
OC
MT___X
N
T
9
_~ 3 0
0
>
L
"
:
9
Z
!
T'
--
0 [
6-MP_
__6-THIOOUANINE
AZASERINE
"
<_2o
'~
Vol. ~1, N o v e m b e r 1961
9
|
:
,
:
,
-
-
:
.
.
.
.
.
t
9
"
:
/.
J
I
l
I
I
0.4 0.9 2_.I
I
]
I
I
9
-
:
1210
~
/
=
9
[,/:
I
LEUKEMIA
,
. :/
I
I
I
I
I
I
.
.
I
i
.
I
.
i
.
!
1.4 3.1 7.0 16
35 60 160
DAILY DOSE, M S / K S
.
1
.
l
]
24
I
I
I
I
53 120
CttART 1.--Illustration of use of survival time assay employing advanced leukemia LI~10 in mice (see reference 43). In
this experiment treatment was initiated 9 days following tumor
inoculation, this being only 3 days before the median day of
death (day 1~) of untreated controls. The optimal dose of
methotrexate (MTX), 0.9 mg/kg, yielded a median survival
time of ~6 days, 17 days beyond the initiation of treatment.
With 6-thioguanine, the optimal dose was 7.0 mg/kg, the median survival time being 19.5 days. 6-Mercaptopurine (6-MP)
at a dose level of 180 m g / k g yielded a median survival time of
~0 days, whereas 53 m g / k g daily of azaserine yielded a median
survival time of 17 days. The respective maximum increases in
doses must be small enough so t h a t the observed
optimal dose and maximal survival time may be
taken as near approximations to their unknown
true values. In employing this methodology, deseribed in detail in other publications (40, 43-46,
99), it has been the praetiee of the authors to use
the median survival time of a group of mice as the
summary measure of survival time. Individual
survival times may, nevertheless, be reported (45,
46, 99), protection thus being afforded against instances where the median survival time may be
misleading. Chart 1 provides an illustration of the
use of this procedure for determining the relative
effectiveness of several antileukemie agents.
There are two aspects to this question. First,
where an agent can, in fact, elicit important increases in survival time, or possibly even cures,
how likely is this agent to be detected in a tumorinhibition assay? Second, to what extent m a y
agents, capable of eliciting only limited, nonspecific increases in survival time, give rise to reductions in tumor size? These aspects will be discussed below.
Where the use of tumor inhibition is premised
on its being an easily ascertainable basis for selecting effective agents, the question of how much
more efficient tumor-inhibiton assays m a y be also
arises. Goldin et al. (44) have suggested that,
median survival time over untreated controls are: methotrexate, 14 days; 6-thioguanine, 7.5 days; 6-MP, 8 days;
azaserine, 5 days.
Note that, although some control mice died within a day or
two after the scheduled time of initiation of treatment, such
early deaths did not occur among mice receiving effective levels
of methotrexate. In this experiment, treatment was discontinued for the last two survivors in each group, to provide the
possibility for identifying cures resulting from previous treatmerit.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
Got, DISC et al.~Preclinical Screening and Evaluation
under certain circumstances, survival time assays
can be performed with efficient use of time and
animals. In such case, tumor-inhibition assays
may be unnecessary.
Although the emphasis in survival time assays
is on the longevity of treated animals, it is not intended that in such assays toxic manifestations
should be overlooked. One would obviously be
dubious about accepting as an effective agent one
which, while increasing survival time, resulted in
permanent injury to a vital organ.
Assays to determine effectiveness of agents.--It is
perhaps naive to expect too high a correlation between the tumor-inhibitory effects of agents and
their potential for extending survival time. Measurement of inhibition of tumor growth does not
permit accurate extrapolation of survival time.
Factors which may influence the relationship of
inhibition of tumor growth to survival time include the drug dosage, schedule of therapy, and
the nutritional status of the host. The relationship
may vary in accordance with the time at which the
observations are made. Agents effective in increasing survival time nmy achieve such prolongation
in a number of ways. For an effective agent with a
true tumorieidal effect a correlation might be
readily obtained between tumor inhibition and
survival time, provided the chemical agent is relatively nontoxic for the host. The correlation
would be less readily demonstrable if the agent is
toxic for the host. Alternatively, an agent may behave as a eareinostat, perhaps even with stasis not
being evident until after the time selected for sacrifice of animals in a tumor-inhibition assay. While
eareinostasis might lead to increased longevity, it
would not necessarily produce effects detectable
in a tumor-inhibition assay. There may be agents
capable only of inhibiting invasion and metastasis
of the neoplastic process without affecting the primary tumor; here the infltrating tumor would
appear to be sensitive and the local tumor resistant. In such instance, an agent could be overlooked readily in a tumor-inhibition assay. Conversely, there may be agents which inhibit the primary tumor but fail to retard metastatic infiltration. Increase in survival time of mice with leukemia L1210 has been observed to occur with either
progressive growth of local tumor at the subcutaneous site of inoculation of leukemic cells or a decrease in size or actual disappearance of the local
tumor (131). If agents producing the kinds of effects
described above were encountered, restriction of
measurement to observation of local tumor size
should be particularly hazardous as a measure of
therapeutic effectiveness.
1337
When one considers results obtained with a
range of dose levels, some difficulties can be visualized. For an effective drug, as the dose is increased, the rate of tumor growth may be diminished and the survival time increased. It is clear,
however, that, owing to toxicity for the host, the
very highest dose levels may give the highest degree of tumor inhibition, yet show the shortest
survival times. This difficulty can be met in part
by requiring that the dosage level employed be
one that yields little lethality (or toxicity), at
least as of the time of sacrifice. This may only
yield fresh complications, due to differences among
agents in the time at which they cause death or
other toxic manifestations. With some agents, animals which have received lethal treatment may
still be alive and in apparently good condition as
of the scheduled date of sacrifice. Tumor inhibition would here be ascertained at what is, in fact,
a nontolerated dose.
Further difficulties arise when one considers the
possibility of cumulative toxic effects for the host.
Where cumulative effects are minimal, one may
continue to treat successfully animals which have
not yet shown important tumor inhibition. Conversely, the greater inhibitory effect for another
agent may be accompanied by such cumulative
toxic effects for the host as to make impossible
further successful treatment.
It is important to emphasize that, in the employment of survival time assay systems, one is
guarded against the above types of situations. A
variety of relationships encountered between tumor inhibition results and survival time extension
is illustrated in Charts ~a, b, 3a, b.
Nonspecific effects of treatment.--Nonspecific
drug toxicity for the host, including reduced caloric intake and body weight loss, may alter tumor
growth and survival time. There need be no precise correlation between the effect of nonspecific
toxicity on tumor size and survival time. Inhibition of tumor growth by caloric restriction may or
may not result in an increase in survival time, or
it may, if the restriction is too extensive, actually
result in a decrease in survival time. With increasing weight loss there may be progressive inhibition
of tumor growth, but survival time may increase
with moderate weight loss and decrease with excessive weight loss. Various investigators (r 4, ~8,
54, 76, 80, 93,101) have produced tumor inhibition
by caloric restriction. In one series of experiments
it was shown that the degree of tumor inhibition
of Sarcoma 180 elicited with a variety of agents
corresponded to that obtained with levels of caloric restriction giving the same weight loss (~8). This
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
I
I
I
I
I
1
I
I
I
I
(~ARONOM^ 755
- CHEMOTHERAPY,DAYS 4 THRU8
(}-DAILY TREATMENTLEVEL IN MG/KG.
3 0 - I-MEDIAN SURVIVALTIME
25 35 30
~E
"~ 2 5
uZ
CHEMOTHERAPY4---
CARCINOMA 755
_i
_]
~:
I - MEDIAN SUR-~-i (c)
-~
~. I 0
6 -
MP
,~/,.--I (c}
-I
(~]
(sg)
20
-
-
N 20
//
tD
_z15
_>
.
4
>_
C:YTO
~:t0
c~
"o
5
/
,,,0
,
,
,
,
,
,
,
'
'
CYTOXAN
_J
~30
I
(leO) NO
TUMORS
DEFINITE
O
:~ 2 5 -
,,, 2 0
-t
,,,15
J,
I0
I (50)
0'
0
I
I
I0
I
I
I
20
5
l
30
i
l
40
i
I .J
50
DAYS POST TUMOR INOCULATION
O0
DEFINITE TUMORS
~
(8.3) uP TO
DAYlfl
~
t
I0
',
20
i
i
30
i
i
i
40
I
J
50
DAYS POST TUMOR INOCULATION
C~ART 2a
C}~A~m 2b
~2a AND ~b.--Examples of experimental antitumor
activity employing Carcinoma 755. These charts permit following the progressive antitumor effects of treatment and
show also its therapeutic efficacy as evidenced by a prolongation in survival time. Each curve shows tile average tumor
diameter, as measured by palpation, of a group of treated mice
on specific days following tumor implantation. The curves are
discontinued on the day of median survival of the group.
For the experiment represented by Chart 2a, daily treatment with graded dosages of compound was initiated 4 days
following tumor inoculation and was continued until death of
the animals. It can be seen that both 6-mercaptopurine (6MP) and Cytoxan were effective at the optimal dosages of 8.1
and 6.~ m g / k g daily, respectively, in increasing the median
survival from ~4 days for untreated controls (C) to 50 days.
With Cytoxan this increase in survival time was accompanied
by only a moderate tumor-inhibitory effect, contrasting with
the marked inhibition due to 6-MP. Itad tumor inhibition at
some early sacrifice date been the only criterion for evaluating
an agent, standards by which 6-MP would have been accepted
could well have led to the rejection of Cytoxan, an equally effective drug by survival time standards. It m a y be noted, especially for the Cytoxan data, that doses beyond the optimal,
while showing increased tumor inhibition, lead to decreased
survival times.
For the experiment represented by Chart el), daily drug
treatment was initiated on day 4 and continued for only 5
days. With this curtailed treatment the optimal daily dosage
was greatly increased, to 65 m g / k g for 6-MP and to 108 m g / k g
for Cytoxan. Again it is evident that the efficacy of 6-MP
(greater this time than that of Cytoxan) is accompanied by
marked tumor-inhibitory effects, these being so great that tumors are not evident for an extended period following the discontinuance of treatment. Overdosage with 6-MP to a level of
180 m g / k g daily resulted in a decrease in survival time.
~,5-Bis(1-aziridinyl)-g,6-bis(~2-methoxyethoxy)-p-beI~zoquinone
(A-159) which elicited clear tumor inhibition at a daily dosage
of 5 m g / k g did not produce any important extension in survival
time.
In other experiments with Carcinoma 755, not shown here,
treatment with 6-MP or with Cytoxan was not initiated until
as late as day 13, at which time mice already had clear palpab]e
tumors. Drug treatment at this time, which was effective in
extending survival time, caused the tumors to regress, though
ultimately there was regrowth.
CIIART$
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN et al.--Preclinical Screening and Evaluation
is not to say that the antitumor effect of the agents
was necessarily mediated through a food restriction mechanism. However, it does suggest that the
effects observed were nonspecific, reflecting only
damage occurring to the host.
Toxic effects for the host are generally evident
from the over-all weight loss which follows. Where
the nonspecific effect on a tumor is due to a toxic
effect not reflected in over-all host weight loss, the
effect may mistakenly be taken as real. Alternatively, real antitumor effccts which happen to be accompanied by weight loss can be mistaken as nonspecific. When caloric restriction is cmployed as
t h e n o n s p e c i f i c a g e n t , t u m o r - i n h i b i t o r y effects exI
I
I
i
I
I
I
I
1339
tensive enough to meet most screening requirements can be achieved. However, these effects
will, in general, be accompanied by substantial
weight loss.
The relationship of tumor inhibition and survival time of the animals m a y be dependent upon
t h e e x t e n t of u n d e r f e e d i n g . T h e r e l a t i o n s h i p m a y
change with the time at which the measurements
of t u m o r size are t a k e n . T h e s c h e d u l e of u n d e r feeding, including the time at which underfeeding
is t e r m i n a t e d , m a y also i n f l u e n c e t h e r e l a t i o n s h i p ,
etc.
A l t e r a t i o n of d i e t m a y also a l t e r t u m o r g r o w t h s u r v i v a l t i m e r e l a t i o n s h i p s , etc. W i t h f e e d i n g of
2~
]
I
--
[
T~-
!
- ~ - -
;
j
J
(o.2~) '(c)
NITROGEN
( f o . : , (o.~o) MUSTARD.
)
9---~(z.o)
o
s-
N
0
z
I
I
-
SARCOMA
37
CHEMOTHERAPYFROM DAY3( ) - DALLY TREATMENT LEVEL
IN MG/KG.
I ,,' MEDIAN SURVIVAL TIME
]
I
(9.g)
m 20-
{
I
t
.,~r~
I0
--
SARCOMA 37
CHEMOT~
~3 THRU 7~
()=DAILY TREATMENT LEVEL
i
IN MG/KG
I - MEDIAN SURVIVAL TIME
{///
r
|
5
z_
I
oc
HYDROCORTISONE
(c)
L
:~:~
0
6-THIOGUANINE
2
t
I
I
(r
o I5_
,.,.
!
7
--I (5 o)
ta..i
J
~0
<
e-~ 5 o
:~ 0
~
~,
I
I
I
I
I
I
I
(39l
I
i
t
(c)
k-
e---
5 t--
(
4 - APP
9
,., 20
,,, 2 0
<
-4 (c)
~9
//
t~
~
I0
"~ 10
--I (333)
5
N-METHYL FORMAMIDE
0
4
I
!
'
'
I
',
!
!
8
12 16 20 24 28 32 36
DAYS POST TUMOR INOCULATION
CHART
-
0(
40
33.
Caxars 3a AND 3b.--Examples of experimental antitumor
activity employing Sarcoma 37. For explanation of how results
are shown in these charts see legend for Charts 2a and ~b, first
paragraph.
For the experiment represented by Chart 3a, daily treatment with graded dosages of compound were initiated 3 days
following tumor inoculation and were continued until death of
the animals. No increase in survival time was elicited with
either nitrogen mustard or hydrocortisonc. Ovcrdosages with
these compounds which reduced survival time did not generally produce tumor-inhibitory effects. At an optmml daily
dose of 1~20mg/kg, N-methyl formamide yielded a median survival time of ~5 days, compared with 17.5 days for untreated
controls (C). This was not accompanied by important tumor
inhibition. The overdose of 333 mg/kg daily yielded much
+(3 9)
5t
o
4
:
!
I
!
~
1
I
8
12 16 20 24 28 32 36
DAYS POST TUMOR INOCULATION
C~AR'r
40
3b
more extensive tumor inhibition, but survival time was reduced.
For the experiment summarized in Chart 3b, daily treatment with 8-azaguanine, 6-thioguanine, or 4-aminopyrazolo(3,4-d)pyrimidine (4-APP) was initiated on day 3 and continued
for only 5 days. None of the three compounds was effective in
increasing survival time, and only one, 6-thioguanine, gave evidence of tumor inhibition. For the other two compounds, antitumor effects were not clearly evident even at overdosages.
In Charts ~a, ~b, 3a, and 3b, a variety of relationships between drug effects may be seen. Drug efficacy in extending survival time may be accompanied by a greater or less degree of
tumor inhibition. Drug toxicity for the host may or may not be
accompanied by tumor-inhibitory effects.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
Cancer Research
1340
Vol. ~1, N o v e m b e r 1961
desiccated tumor, the Walker tumor continued to
grow, and survival time was increased (70).
The age and weight of the mice may also alter
tumor growth-survival time relationships (30, 36).
In general, where survival time is the end-point
employed, the problem of nonspecific effects is
likely to be less critical. Work in this laboratory
has shown that, while caloric restriction may produce important reductions in tumor size, the increases in survival time to which it gives rise may
tend to be less dramatic (54).
In view of the drastic effects that diminished
food intake may have on the growth of transplantable tumors, and the possibility of underfeeding
actually increasing survival time, the importance
of characterizing experimental tumor systems with
respect to the influence of underfeeding cannot be
overemphasized. If the experimental tumor system
is sensitive to underfeeding, and/or weight loss,
what would it behoove us to screen thousands of
drugs, only to select as positive those drugs that
depress appetite! Further testing would be required, in such cases, to determine to what extent
eo~
diminished
food intake could account for the ob.._1
0
served
inhibition
of tumor growth. Experiments
rr"
ISO-CALORIC
involving paired feeding or isocaloric controls
I-COMPOUND EXP. CONTROL
Z
would be required, and they are time-consuming.
0
9
2 (1.2 MG/KG) 0
(J
It may be profitable in setting up chemothera9
2~
2 (0.6 MG/KG) 1:3
peutic assay systems to establish regression lines
22
I-for the tumors employed, relating degree of under133
17
feeding
to antitumor effect or degree of weight loss
[]
9
.J
3
to
antitumor
effect. The antitumor effect could be
5
measured as tumor weight, survival time, etc. The
s
E
I
LI_
extent of correspondence of the drug-induced anti0
(1
I
15
tumor effect to the regression lines may then be
~-I00
qdetcrInincd. A close correspondence would be ins
dicative of nonspecific antitumor action of the
80
j
drug.
rY
The complication of nonspecific toxicity in
o 60
2~
evaluation of tile antitumor effectiveness of a drag
:D
may frequently be overcome in large measure by
~- 40
appropriate adjustment of the assay procedure. If
7
it
is found that the test system employed is too
"k9' 20
REGRESSION
sensitive to nonspeeifie calorie restriction or host
Ct2
LINE
toxicity, this may sometimes be remedied by inI
I
I
I
I
1
I
a_ 0
creasing the inoculum level or withholding therapy
+10
0 -I0 -20 -30 -40
until
the tumor is more advanced. Thus, food rePERCENT CHANGE IN ANIMAL WEIGHT
striction may inhibit local tumor growth and inCltART 4.--Illustration of a test for nonspecific inhibitory
crease survival time of mice inoculated with leukeeffects of treatment with nitrogen mustards on Sarcoma 180.
mia LI~10 when the restriction is initiated within
TMs chart is taken from Goldin et al. (~28). The tumor-inhibiseveral days following leukemic inoculation. IIowtory effect for each treatment group (open tigure) and its isoever, when the disease is suftieiently advanced,
caloric control (closed figure) is plotted against the percentage
calorie restriction produces minimal inhibition of
animal weight loss occasioned by treatment. The line shown on
the chart is based on the average relationship observed to occur
the local tumor growth and does not increase surbetween tumor weight and animal weight loss when such effects vival time (54). This obviates the need for further
wcrc brought out in food restriction studies 028). In the prestesting of compounds (with isoealorie controls,
ent chart it can be seen that tumor weights for drug-treated
etc.) that are found active against advanced leumice are substantially similar to those for their isocaloric controls and that body weight loss for the host can account for
kemia LI~10. I t removes the need to be concerned
such tumor inhibition as has occurred.
about any possible failure of correlation of effect
Treatmeut was initiated one day after tumor inoculation
of drng-imposed calorie restriction on tumor
and continued daily for 5 days. The compounds and daily doses
growth and survival time in attempts to interpret
are coded as follows:
specific drug effects. Charts 4 and 5 illustrate some
(1.~ and 0.6 mg/kg)
Mcthylbis (~-chloroethyl) amine
of the points made in the foregoing, relative to the
~
N,N,N~,N'-tetrakis (~-chloro(4.4 Ing/kg)
ethyl) cthylcncdiamine
effects of food restriction.
E:~
W
i
m
i
17
8
5
1
15
Furfuryl-bis (/3-chloroethyl) amine
Ethyl-bis (~-chlorocthyl) amine
Isopropyl-bis (~-chloroethyl) amine
Bis ($-chloroethyl) amine
_O-Methoxyethyl-bis (fl-chloroethyl) amine
('2.4 m g/kg)
(0.4 mg/kg)
(0.S mg/kg)
(~20.'2 mg/kg)
(0.4 mg/kg)
TIIE THERAPEUTIC INDEX
A concept which has been useful in pharmacological research and which has been applied in ex-
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN et al.--Preclinical Screening and Evaluation
p e r i m e n t a l cancer c h e m o t h e r a p y is t h a t of the
" t h e r a p e u t i c index." T h i s index is one which takes
account of the fact t h a t the action of an a g e n t has
a d u a l nature. T h e r e are effects b o t h on t h e disease
process for which the a g e n t is being applied a n d on
the diseased host. W h e n a p p r o p r i a t e l y e m p l o y e d ,
the t h e r a p e u t i c index is a p a r a m e t e r , v a l u a b l e in
characterizing the action of a t h e r a p e u t i c agent.
I n general, the t h e r a p e u t i c index m a y be used to
rate a n t i t u m o r agents w i t h respect to the m a r g i n
of safety in their use. T h e higher the t h e r a p e u t i c
index, the greater the m a r g i n of safety in the
a c h i e v e m e n t of a specified a n t i t u m o r effect.
Specifically, the t h e r a p e u t i c index is the ratio
of the dosage level of a n agent causing a defined
degree of d a m a g e in the host to the dosage causing
a defined effect on the disease process. D a m a g e to
the host m a y be in t e r m s of lethality, weight loss,
or other toxic effects of the agent, the defined degree of such effect o r d i n a r i l y being a t some minim a l level. R e l a t i v e to the disease process, the defined effect m a y be some specified reduction in
p o p u l a t i o n of the disease-causing e n t i t y below
either its initial value or the final value for corresponding u n t r e a t e d controls; a l t e r n a t i v e l y , the
effect on the disease process m a y be in t e r m s of a n
objectively m e a s u r e d degree of i m p r o v e m e n t in
the host, or a certain percentage of cures, or a certain survival time increase. W i t h e x p e r i m e n t a l
t u m o r systems, the t h e r a p e u t i c index m a y be
t a k e n as the ratio of a specified host toxicity to a
specified a n t i t u m o r effect (56, 57, 7~, 84). F o r purposes of the current discussion, it m a y be defined
as LD~0/EDg0, where the LD10 is the dosage lethal
to 10 per cent of the test a n i m a l s a n d the EDg0 is
the dosage yielding a 90 per cent reduction in tumor weight below u n t r e a t e d controls.
T h e m e d i c a l a d v a n t a g e s of a h i g h t h e r a p e u t i c
index in some situations can readily be seen. Suppose in a clinical situation t h e index were defined
in t e r m s of the ratio of dose j u s t b a r e l y p r o d u c i n g
toxic m a n i f e s t a t i o n s in the p a t i e n t to the dose giving some large percentage of cures. W h e r e t h e
ratio is high, a t r e a t m e n t dosage l e v d can be
selected somewhere between the effective dose a n d
the tolerated dose so as to provide a wide m a r g i n
of safety. T h e r e is e n o u g h l a t i t u d e to ensure little
chance of a p a t i e n t ' s being either overtreated or
ineffectively treated. An a g e n t with a high index
is thus ideal for routine medical use with only
fairly general instructions to clinicians as t o app r o p r i a t e t r e a t m e n t levels. T h e lower m a r g i n of
safety which would o b t a i n when the t h e r a p e u t i c
index is low could restrict the use of a n a g e n t to
eases where conditions can be carefully controlled
w i t h respect to the characteristics b o t h of the pat i e n t a n d the drug. I t m i g h t be too optimistic to
1841
expect a clear-cut t h e r a p e u t i c a d v a n t a g e in m a n
w i t h a d r u g which has only a m o d e r a t e increase in
t h e r a p e u t i c index in a n a n i m a l t u m o r system.
R e t u r n now to the t h e r a p e u t i c index defined
i
--i--"
--
i
35-
'1"
[
1
i
I
t
CARCINOMA 755
EXP. I-FOOD RESTRICTION FROM DAY 4
{)-6M& FOOD OFFERED ,~
25PER DAY PER A /~_.z(r
MOUSE
/~%z~
....
uS
//~N.//\'~--,,._._._._._._.~.(2.2)
2030
tD
Z
-I,
l
{
I
15
I0
g
~o(.m
(C)=AD LIBITUM CONTROLS
iV;~'~(o.-.~)
' I=MEDIAN SURVIVAL TIME
5L
3 ~ -
I
-+--,
I
F
i
r
t
t
I
EXP. 2-FOOD RESTRICTION FROM DAY I0
w
30-
~
25
20
<
rY
uJ
'
i
~
15
I0-
/~
---~(2.a)
-40
3)
q
(o.8)
iJ
50
0
-I,
8
16
24
32
i
40
DAYS POST TUMOR INOCULATION
Cm~T &--Effects of food restriction oi1 Carcinoma 755.Results are shown in the same way here for food restriction effects
as they are for drug treatment effects in Charts ~a, ~b, 3a, and
8b. For explanation of what is shown, see legend for Charts ~a
and ~b, first paragraph.
The top panel shows how food restriction can mimic the effects of a compound which, though inhibitory to the tumor, is
ineffective in eliciting important survival time increases. Extreme food restriction, like overdosage with some drugs, leads
to more pronounced tumor inhibition but survival time is reduced.
For the experiment of the bottom panel, the initiation of
food restriction was delayed until day 10. In this experiment,
food restriction does lead to an extension in survival time (untreated controls died sooner in this experiment than in the
first). As with overdosages of an effective agent, the more extreme tmnor inhibitions with still further food restriction
caused earlier death of the animals. That food restriction can
produce some degree of extension in survival time does suggest
that agents to be judged effective should be required to yield
still greater extensions when working with this experimental
tumor system.
The phenomenon of food restriction being more effective:
when initiated late than early parallels effects observed with
some chemical agents. If treatment with a tumor-inhlbitory
agent having cumulative toxicity for the host is initiated early,
it will kill the host before the tumor does; with delayed initiation of treatment, drug lethality occurs after the normal time of
death from tumor.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
134s
Cancer Research
above as the LD~0/EDg0 in an experimental tumor
system. Clearly, it is not sufficient to inhibit a
tumor only 90 per cent; it may still be growing or
m a y resume growth and destroy the host. Suppose
one were willing to tolerate 10 per cent drug lethality for the host. Then, where the index is high, a
large relative dosage increase over the EDg0 can
be effected without encountering undue lethality,
while for a low index only a small increase can be
tolerated. The increased effect on the tumor depends, however, not only on the dosage increase
but also on the nature of the dosage-response
curve for the tumor. Where there is parallelism of
these curves for the two agents, ordinarily the
larger dosage increase permitted with the high
therapeutic index would result in greater antitumor effect. Such parallelism may not, however,
exist. I t is possible that the agent with the lower
index may nevertheless display a so much steeper
antitumor dosage response curve as to prove more
efficacious at the tolerated level. The agent with
the greater index may have so shallow a curve as
to be little more effective against the tumor at the
LD10 than at the EDg0.
The therapeutic index may not be too discriminating if it is employed with weak or highly sensitive tumor systems. If the tumor system is extremely sensitive, any number of drugs having
different therapeutic indices could effect complete
inhibition of tumor growth. In such circumstances,
the assay system would, in essence, determine the
degree of safety with which any number of drugs
may effect total inhibition of tumor growth.
A common practice in determining the therapeutic indexes of compounds is to obtain the host
toxicity data in normal mice, with the tumor mice
exclusively used to determine degree of a n t i m m o r
inhibition. An additional element of error is involved in this procedure, since there may be a lack
of correspondence of the dose-mortality curve for
the normal host and the tumorous host. In general,
the tumorous host may tolerate less of the drug.
Additional studies would be required to compare
drug toxicity in the normal and tumorous host.
Even if the drugs are so inhibitory to tumor
growth t h a t the dose range for antitumor effect is
much below the dose range for toxicity, the most
efficient practice could be to determine the entire
dose-response curves for tumor and host in the
same population of tumorous animals. If it is considered t h a t this dose range is too great, it might
be useful in order to bring the antihost and antitumor dose ranges closer, to alter the test system
so t h a t the antitumor effect is more difficult to
obtain.
Vol. ~1, N o v e m b e r 1961
THE SPECIFICITY INDEX
While intended primarily as a measure of drug
safety the therapeutic index may, under appropriate circumstances, be employed to compare the
therapeutic potencies of various drugs. This is true
when the tumor-dose response curves and also the
host-dose response curves for the various agents
are parallel. This m a y occur if the various agents
being compared have related structures and thereby display similar activities. Care, however,
should be taken in comparing therapeutic indices,
particularly for unrelated structures. In general,
when the agents do not have similar activities, the
therapeutic index fails to provide an answer to the
question, " F o r equivalent cost in toxicity or
equivalent risk for the host, which drug is the
more effective antitumor agent?"
An alternative measure, the specificity index,
has been employed (~27, 33-39) as a measure of
therapeutic efficacy. In broadest terms, it is (lcfined as the antitumor effect at a specified toxicity
risk. The question is asked, " F o r equal host toxicity, which drug gives greater inhibition of tumor
growth, or greater increase in survival time, or
greater percentage cure, etc.?" Other paralneters
of host effect and tumor effect could be employed.
The host effect could be body weight change, food
intake, hematopoietic alteration, or a cytological
or biochemical response, etc. The tumor effect
could be change in cell number, level of enzyme,
cytologic alteration including mitotic aberration,
etc.
In principle, there is not only a specificity index
for an agent but also a specificity curve. Suppose
an agent has been administered to the tumorous
host under conditions which permit ascertaining
separate dose response curves for both tumor and
host. A specificity curve can, in turn, be determined from these curves by relating the antitumor
effect of drug treatment to the corresponding effect on the host. Drugs which have distinct response curves for tumor and host may yet have
identical specificity curves.
Chart 6 illustrates this concept. For each of two
drugs, A and B, the panels of the figure illustrate
the tumor response curve, the host response curve,
and the specificity curve. In the illustration, drug
A, which has the lower response curve both with
respect to the tumor and the host, has, nevertheless, the higher specificity curve. For simplicity, in
the illustration, the response curves are shown as
parallel straight lines; this is not a general requirement for obtaining or comparing specificity curves.
The methodology may be employed to rate
drugs with respect to specificity, or to evaluate
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN et al.--Preclinical Screening and Evaluation
regimens of therapy. This type of evaluation has
been made with individual drugs (38, 39) and combinations of drugs (~7, 3~-34, 37, 39) used in treatment. The combinations of drugs have been employed in studies of synergism (~7, 3~, 39) and
antagonism (~7, 3~-34, 37).
In the above types of assay the question may
arise as to how to determine dose-response curves
for host and for tumor in the same host. This has
been accomplished in one type of experiment in
which host. mortality was used as the toxicity index and survival time of iniee not succumbing to
host toxicity as the tumor index, by designing the
experimental test system to provide a temporal
separation between the two parameters. In this
system, animals succumbed to drug dose toxicity
prior to the time that untreated controls suecumbed to tumor.
R E L A T I O N S H I P B E T W E E N T t t E TIIERAP E U T I C I N D E X AND TIIE
SPECIFICITY INDEX
If the therapeutic specificity of an antitumor
agent is defined in terms of its antineoplastic effect at a defined level of tolerance for the host
(specificity index), it is seen that there is no necessary relationship between a drug's therapeutic
specificity and its therapeutic index. 1 The therapeutic index has to do not so much with the efficacy of an agent, but rather with the care needed
to be exercised in its employment. The importance of both the efficacy of an agent and the degree of safety in its use is evident. Since, in clinical
cancer situations the utmost of care is exercised in
the use of a drug, additional attention should be
given to therapeutic specificity as an important
basis for ranking candidate agents.
The concepts of the therapeutic index and therapeutic specificity have certain parallels in the
assay procedures involving maximum survival
time increases. As described above, the assay is
directed toward ascertaining the maximum survival time which can be obtained with a drug on a
particular schedule. This maximum can be thought
of as corresponding to the drug's therapeutic specificity. Use of dosage levels either above or below
the optimal will result in shorter survival times.
Assume there is some survival time which, though
below the maximum, is nevertheless desirably high.
It should be borne in mind that, as with the therapeutic
index, nonparallelism of dose-response curves m a y also posc a
problem for the estimation of the specificity index. Alteration
of the specificity index at different toxicity levels with nonparallel dose response curves for t u m o r and host m a y create
additional difficulties in attempts at correlating therapeutic
index and specificity index.
1343
There should then be two dosage levels yielding
this survival time; one will be above, the other
below, the optimal dosage. The ratio of these two
dosages, which corresponds to the therapeutic
index, indicates how much latitude there is for
varying the dosage without encountering undesirably low survival times. This index has been employed by Creech et al. (13) and can provide an
important ancillary index to determination of the
maximum survival time in the rating of chemotherapeutic agents.
The various concepts discussed in this section
are illustrated graphically in Charts 7 and 8.
TIIE TUMOR CIIALLENGE
Any experimental measure of the effect of an
antineoplastic agent is necessarily specific to the
circumstances surrounding its determination. Effectiveness will vary with the particular neoplasm
8,00]
-f-
i001
~
oI,
O
so
l0
20
410
I~ Olll0
DOSAGE SCALE
r
I00
o
~
5O
I,I
Lt-
210
410
DOSAGE SCALE
I
so
%
~o
,oo
EFFECT ON HOST
C~IART & - - S c h e m a t i c representation of the use of the specificity index to determine tile relative effectiveness of two
drugs, A and B. The upper panels illustrate dose response
curves for the t u m o r and the host. T h e vertical scaling is not in
defined units and should be interpreted to be the appropriate
scale for a n y appropriate parameters of t u m o r and host response employed (increasing numbers indicate increased response). The dosage scale was chosen arbitrarily. It should be
noted that, for equal dosage, the a n t i t u m o r effect of drug B is
greater than that of drug A. The deleterious effect of drug B for
the host is also greater than t h a t of drug A.
The lower panel illustrates, for each drug, the antitumor effect as a function of the effect on the host. From these specificity curves, it can be seen t h a t for a n y given effect on tile
host, drug A produces a greater a n t i t u m o r effect than drug B.
For simplicity, the dose response curves for drugs A and B
are showu as being parallel. Nonparallelism between the dose
response curves for the t u m o r or between the dose response
curves for the host could result in nonparallelism between the
specificity curves for the two drugs. This would complicate
their evaluation to the extent t h a t the relative efficacy of the
drugs would vary with the level of effect on the host chosen for
their comparison. Examples of the use of the specificity index
for determining the relative merits of various therapies m a y be
seen in Charts 10 and 11.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
Z
HIGH SPECIFICITY
A ANTI-TUMOREFFECTI00 AT DESIGNATED TOXICITYFOR HOST B
I00
<
F--~.)
ANTI-TUMOR CURVE//J
r,
u..
"'
80
~,-
60 -THERAPEUTI
-
HIGH
4~
o~
LOW THERAPEUTIC/~
INDEX
'b9-o
t6 "2 / I
/
OR TU
OR
TUMO
MO~,,R,-"
/
0
C
ts.l V')
..aO
>-
60
a:
40
<
20
LOW. SPECIFICITY
D/
ANTI-TUMOR EFFECT40 AT DESIGNATED TOXICITY FOR H O S T / /
HIGH THERAPEUTIC INDEX
LOW THERAPEUTIC INDEX//
<=80
Q)
I--rfl
n.-"
l
"~'DESIGNATED,
/'
9a~ O
ta_ I - -
r'r
<
.HOST
I TOXtCtTy
7
12)
3 2 0 . nO
,/
~
_
3~o
2
/
L/
~/
/
I0
20 40 80
I
160 320 640 1280
20 40 80
DOSAGE SCALE
~
~
,
SURVIVAL /
'
I
I
i
,
6O -
\
60 -
,
'
/
/
l
I
I
I
-i
/I
j
/I NARROW
-
\ (DOUBLED
\
I
A
I00 DAYS
WIDE EFFECTIVE
I
SURVIVAL
CONTROLS
a: 80 -
I
B
/
X, a',ME)
,'--, 40 - ./~nnt JpJ ~n~l ~WvAIN,, 32o = ,6
.....
. . . . . .
J
%
0
,
MAXIMUM SURVIVAL TIME
CrIME
___2O
F-
160 320 6401280
HIGH EFFICACY
I00
o'3
I
]
Panel A illustrates an agent with a high therapeutic index
(thc ratio of the dose yielding a designated level of toxicity to
the dose yielding a designated antitumor effect is large) and a
high st)eciti(tity (there is a strong antitumor effect at a dosage
level yielding a designated toxic effect for the host). In panel B
it can be seen that the same high specificity can be obtained
even when the therapeutic index is low, whereas panel C illustrates how tlle high therapeutic index can be coupled with a
low st)ecificity. Finally, panel D shows an association ()f low
theral)eutic index with low specificity.
r
A
80
I~ /
/ /
CIIART 7.--Schematic representation of various possible relationships between the antitumor speciticity of an agent and
its therapeutic index. Each of tile four panels of tim chart
shows a hypothetical autitumor dose response curve and a
hypothetical host toxicity curve for an agent. For simplicity,
these are shown as straight lines. The vertical scaling is not in
defined units and should be interpreted to be tlm appropriate
scale for antitumor effect and host effect as the case m a y be.
Antitumor effects and effects on the host are generally in noneomparal)le units making the concept of parallelism between
their dose response curves inappropriate.
,
]
I
I ~ G E
l ~^
I
-!
:,
_~
i ,~O=z
80
\
~
_
~
L
C
WIDE EFFECTIVE
I
I
L
I
I
I
I
I
I
J
LOW EFFICACY
D
MAXIMUM SURVIVAL TIME
-I
50 DAYS
NARROW EFFECTIVE
DOSE RANGE, z-~bg-o
~m
I
_
DOSE RANGE,
40
20
0
I0
CON TROL S
"~
I
;
(
20
40
80
i
I
I
:I
160 320 640 1280
r
20
:
'
40 80
!
I
I
i
160 320 640 1280
DOSAGE SCALE
CHART 8.--Schematic representation of various possible relationships between the therapeuti(' efficacy of an agent and the
width of its effective dose range. These relationships are intended to parallel those between antitumor spccificity and
therapeutic index as shown in Chart 7. Each of thc four panels
of the chart shows a hypothctical survival time dose-response
curve. Characteristically, the survival time of treated tumorbearing animals increases with dosagc up to a I)oint; increased
dosage beyond some optimal lcvel results in reductions rather
than increases in survival time. The therapeutic ctficacy of an
agent is defined as the survival time elicited by its optimal dos-
age. Arbitrarily here, the width of tile effective close range is
defined as the ratio of the highest dose yiehting a twofold or
better increase in survival time over untreated controls to the
lowest dose doing so.
I t can be seen from the various panels that agcnts with high
efficacy (panels A and B) m a y be effective over either a wide or
narrow ctt'cctive dose range; also agents having a wide effective
dose range (panels .4 and C) m a y or m a y not be especially efficacious. These parallel the possible situations existing between the antitumor specificity of a drug and its therapeutic
index as shown in Chart 7.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN el al.---Preclinical Screening and Evaluation
11345
The work of this laboratory has been characemployed and its source, and with the particular
host employed and its age, sex, or weight. Varying terized by the use of increasing levels of t u m o r
the route of t u m o r inoculation or drug t r e a t m e n t
challenge as more effective agents have been idenm a y result in changed measure of effect (45, 100). tiffed (44). Thus, when it became apparent t h a t
Altered schedules of drug therapy m a y increase or complete inhibition of mouse leukemia LI~10
diminish effectiveness (85, 87, 40, 97). Changes in could readily be achieved with early drug treatdefinition also can change one's measure of effec- ment, provided the t u m o r inoculum was barely
tiveness. Defining the effective a n t i t u m o r dose as large enough to ensure 100 per cent takes, the chalthe one yielding 50 per cent t u m o r weight loss in lenge was made more difficult. T h e inoculum was
5 days will give different measures of the therapeu- raised to one about 100 times t h a t giving 100 per
tic index than defining it as the dose yielding 90 cent takes in the t u m o r titration. Effectiveness in
per cent inhibition in 7 days. Alternative measures such a system now became the criterion for rating
can flow from defining the effective dose in terms agents.
An increasing measure of success against this
of effect on t u m o r diameter or on survival time;
similar changes can occur from changes in defini- massive inoculum in turn led to the employment
tion of the tolerated dose, as, for example, the of still more extensive t u m o r challenges. I t was
dose yielding not more than o0 per cent weight recognized that, while appropriate therapy did
loss or the dose yielding not more than 10 per cent lead to "cure," even though the inoculmn emlethality.
ployed was massive, nevertheless, t r e a t m e n t had
Changes in measure of drug effectiveness such been initiated while the t m n o r was still not y e t
as the last, resulting from altered definitions, are palpable. This did not approximate very well the
only of academic interest. Others m a y be of impor- clinical situation. T h e further increase ill the tut a n t theoretical interest. I t m a y be worth while to mor challenge made at this point was one in which
learn why a particular drug is effective with a cer- test animals were not subjected to drug therapy
tain host, or more effective when the host is older, until tumors were palpable at the site of inoculaetc. Other causes of altered drug effectiveness m a y tion and the disease was systemic.
With this challenge amethopterin was more efbe of important practical interest. If variation in
method of use of an agent, as by altered schedules fective than the various known antileukemic
of therapy, can make the drug appear to be more agents. I t produced several weeks' extension in
or less effective, how most effectively to use the survival time, even though t r e a t m e n t was withagent is worth learning. Although somewhat rea- h e l d until only 5~ or :3 clays before death would
sonable a t t e m p t s can be made at determining es- otherwise have occurred. This provided a yardpecially effective schedules of therapy in the lab- stick for comparing the effectiveness of still other
oratory, in the clinic this remains one of the less agents. :Finally, use of the assay procedure employwell resolved problems.
ing delayed initiation of t r e a t m e n t led to the findAn important factor affecting measures of drug ing t h a t certain halogenated analogs of folic acid
effect which has been referred to in the foregoing could be used routinely to obtain a high percentis t h a t of the t u m o r challenge employed. This fac- a g e of cures. No little p a r t in the cure of experiwhich only a few years
tor is germane to the design of a program for iso- mental leukemia LIs
lating especially effective agents, whether con- ago was thought to be among the more virulent of
ducted on a mass screening or on an individual the laboratory neoplasms, w a s played by the use
of a criterion of efficacy which emphasized the
laboratory basis.
Ordinarily, the more extreme the neoplastic therapeutic results achieved rather t h a n the (Rose
challenge employed, the more difficult it will be to
level employed. On an equimolar basis, amethopfind efficacious agents. The use of rather weak
terin exerts greater antileukemic effect than 3',5'challenges has some justification in the early
dichloroamethopterin. However, the ha]ogenated
stages of a chemotherapeutic research program. I t
analog is considerably less toxic for the host. This
m a y serve at least to demonstrate t h a t the t u m o r permitted the use of relatively massive doses of the
is susceptible to t r e a t m e n t and m a y provide a halogenated analog, resulting in extensive antistarting point for further improvement. Con- leukemic effect unaccompanied by host toxicity
tinued use of too low a t u m o r challenge can, how- (51, 4~, 99).
ever, produce a misleading number of laboratory
I t m a y be noted t h a t use of a lesser t u m o r chalcancer "cures." With weak challenges it is difficult lenge may, under certain circumstances, provide a
to differentiate between agents having specific or reasonable basis for rating candidate agents. When
nonspeeifie a n t i t u m o r effects, or between agents the challenge is limited, one can limit the extent
having varying degrees of specific effects.
to which the test compound is used. Thus, when
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
1346
drugs were t e s t e d in this l a b o r a t o r y a g a i n s t t h e
more m a s s i v e t u m o r i n o c u l u m , b u t w i t h t r e a t m e n t
given early, such t r e a t m e n t was given on o n l y a
single or on o n l y a l i m i t e d n u m b e r of days. As ind i c a t e d above, this device p e r m i t t e d a s e p a r a t i o n
of t h e l e t h a l effects a n d t h e a n t i t u m o r effects of
such t r e a t m e n t . T h e agents could t h e n be r a t e d on
the basis of t h e specificity indices (38).
T h e v a r i o u s systems e m p l o y e d in this laborat o r y h a v e been useful n o t o n l y for r a t i n g c a n d i d a t e
l
I
l
I
I
1
i
I
TREATED (LOW INOC~LUM
_J I00
<~
>
!
)~ 80-
~jL____~CONTROL
1~
/
.~
I~(HIGH INOCULUM)
~o
uJ
II~CONTROL
!-~_~WLNOCULUM)
40
!
~X_
a_ 2o
o
Vol. s
Cancer Research
I
o
I0
I
20
I
40
I\
50
I
60
I
I
70
I
80
1961
T h e value of t u m o r i n h i b i t i o n , per se, as a measure of effectiveness is limited, e v e n w h e n t h e restriction is m a d e t h a t such i n h i b i t i o n be n o t acc o m p a n i e d b y u n t o w a r d toxic effects for t h e host.
I n concept, t h e r e are a v a r i e t y of m o d e s of a c t i o n
b y w h i c h an a g e n t , n a y p r o v i d e a n effective t h e r a p y w i t h o u t its being reflected in a h i g h degree of
t u m o r i n h i b i t i o n a t some specified time. T h e r e is
no necessary c o r r e l a t i o n b e t w e e n t u m o r size a n d
s u r v i v a l time. E x a m p l e s h a v e been given where
i n h i b i t i o n of t u m o r g r o w t h was a c c o m p a n i e d b y
e i t h e r a n increase, or no increase, or a decrease in
s u r v i v a l time.
T h e e x t e n t to which d r u g t r e a t m e n t increases
t h e s u r v i v a l t i m e of t u m o r - b e a r i n g a n i m a l s is suggested as a m e a s u r e of d r u g effectiveness which
gives weight to b o t h t h e a n t i t u m o r effects of t r e a t m e n t a n d its toxic effects for t h e host, a n d which
should be a p p r o p r i a [e w h a t e v e r t h e m o d e of t h e r a p e u t i c action. T h e s u r v i v a l t i m e m e a s u r e m e n t re-
1 TREATED
~[-~H-FGH
iNOCULUM)
'
30
November
i
I
I
I
1
I
28
90
DAYS
CHART 9.--Illustration of how Leukemia LI~10 is less responsive to treatment with amethopterin when the tumor
inoculum level is high than when it is low. This chart is taken
from Goldin et al. (44).
The chart shows for each treatment group and its corresponding control the percentage of ten mice surviving to any
specified day following tumor implantation. Both the low tumor inoculum (14,000 cells per mouse) and the high tumor
inoculum (500,000 cells per mouse) resulted in 100 per cent
tumor takes and subsequent death, with only a moderate displacement between the two control survival curves. With a
high tumor inoculum, treatment with amethopterin at a level
of ~4 mg/kg on days 8 and 7 post-tumor inoculation produced a
substantial extension in survival time over untreated controls,
but ultimately all mice succumbed to leukemic growth. With
the low tumor inoculum 80 per cent of the treated mice survived for 90 days.
agents, b u t also for e l u c i d a t i n g t h e effects of exp e r i m e n t a l modifications. Successive kinds of assay p e r f o r m e d in this l a b o r a t o r y e m p l o y i n g progressively m o r e severe challenges are i l l u s t r a t e d in
C h a r t s 9-lB.
SUMMARY AND CONCLUSIONS
Various aspects of t h e p h i l o s o p h y a n d c o n d u c t
of a n t i t u m o r drug screening a n d e v a l u a t i o n programs are reviewed a n d i l l u s t r a t e d in t h e p r e s e n t
report.
T h e yield of screening a n d e v a l u a t i o n p r o g r a m s
is influenced b y t h e e x p e r i m e n t a l procedures employed, t h e e n d - p o i n t s observed in m e a s u r i n g d r u g
effect, a n d t h e r a t i n g procedure b y w h i c h drug
effectiveness is e v a l u a t e d .
AMETHOPTERIN
26
i:
24
o _~
I- ~.J
~
/
A
M
NOPTERIN
i
2220
IE
','
18-
CONTROLS
16-
14
I
I
I
I
I
I
I
2
3,
4
5
6
7
8
CALCULATED PROBIT OF TOXIC MORTALITY
CtIART 10.--Illustration of the survival time vs. probit of
toxic mortality procedure for comparing the antileukemic
(LI~10) effectiveness of agents. The chart is from Goldin et al.
(38), and the basic data are available there.
This chart shows only the lines fitted to the basic data. The
line for amethopterin being higher than that for aminopterin is
interpreted as meaning that, for a given cost in lethality for the
host, amethopterin yields the greater antitumor effect. The
antitumor effect is evidenced by the average survival time of
mice not dying early of drug toxicity, but succumbing later to
their tumors. On this basis, amethopterin is the superior agent
of the two.
As explained in the reference (38), employment of this assay
procedure is based on a temporal separation of drug deaths
from tumor deaths. Probit procedures (10) are used to ascertain the lethality of any particular level of drug treatment. In
turn, least-squares regression procedures are used to relate
the data on times of tumor deaths to the calculated probits of
toxic mortality. Drugs were administered in a single dose 3 days
post-tumor inoculation.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
TREATMENT OF ADVANCED LEUKEMIA(LI210)
I
I"
'
'
I
'
I
'
1
'
I
'
I
'
I
r
30
I
I
I
I
I
I
I
-
t21
I
26_
.,...
DAY OF AMINOPTERIN TREATMENT
20 -
Dv
*-
o~
Dichloro-0mmopterln
\3'5'
_
223
i-~ J 20--
~g ,s-
/
dl w
f6
ut ~
,>
~"
P2, t
DAY
|
r~Treotment
5F
/ / ( N O NO APPARENT
APPARENT R
REGRESSIONAGAINST
~-"'J
----CONTRoL:AY
- 9
0/I
14
~z
0
/AMINOPTERIN
0.06
I
t
I
I
L
!
!
from
,
t
~
I
0.16
t
i
[
I
I
,
:3.35
9,:30
DOSE
,
I
25.9
,
1
72 0
(Mg./Kg.)
CHART 12.--Illustration of the maximum survival time procedure for comparing the antileukemic (L1210) efflcacies of
agents. This experiment is reported in Goldin et al. (4~).
These results are based on an experiment in which tumorbearing mice were treated daily beginning on day 8 with
graded amounts of drug. While untreated controls died shortly
thereafter, all the agents shown yielded clear increases in the
median survival time at their optimal doses. The maximum survival time elicited with amethopterin exceeded that for
aminopterin, but a higher dosage level was required. This was
in confirmation of the results for this drug comparison as
shown in Chart 10. With 3',5'-diehloroaminopterin, the optimal dosage was not clearly established, but at its lowest dose
this drug did give results comparable to those for amethopterin.
T R E A T M E N T OF ADVANCED L E U K E M I A
t
~
9
CHART l l.--Illustration of the lesser susceptibility of the
advanced tumor to single treatment with drug. The chart is
from Goldin et aI. (35), and the basic data are available there.
This chart illustrates how the procedure exemplified in
Chart 10 was employed to ascertain the susceptibility of L1210
to treatment with aminopterin when administered as a single
dose at various times following tumor implantation. The leukemia was found to be more susceptible on day 3 than on either
day 1 or day 6. When treatment was delayed until day 9, no
important increases in survival time were noted and there was
no evidence of progressive increase in survival time with increasing levels of aminopterin.
[-1
I
1.21
DAILY
I
2
5
4
5
6
7
8
CALCULATED PROBIT OF TOXIC MORTALITY
L
0.4,4
day 8
i
I
I
t
L
I
I
I
i
( L 1210)
(8)
t
~
i
1
I
'4)
(5) I (6)
(4)
(4)
V--- 7o<~
>
rr
60-
5~i
40
)
MTX
=
t ~
20
W
~0
0|1
CONTROLS ( o )
TREATMENT FROM DAYS7,T0 90 r
I
i
0.48 0.75 118
I
184
I
i
I'~1
!
I
I
t
i
171 26.7 41.8 65.5 102 160
DALLY D O S E - M G . / K G .
26.7 4L8 65.5 102
CHART 13.--Illustration of especially efficacious antileukemic (L1~10) compouuds detected using the maximum survival
time procedure. The chart is from Goldin et al. (31).
In this experiment, amethopterin (MTX) shows its usual
high capacity (cf. Chart 12) for extending the survival time of
mice bearing advanced leukemia Llgl0. Its dihalogenated
forms, 3',5'-dichloroamethopterin (DCM) and 3'-bromo-5'chloroamethopterin (BCM), however, show antileukemic activity of a nature not previously observed. These agents elicit
rather extreme increases in median survival time and also yield
substantial numbers of 160-day survivors (29) (100-day survivors are shown in parentheses in the figure). With survivors
160
of advanced leukemia established as the reference effect, it
m a y become desirable to alter assay procedures to permit detection of agents having the capacity to cure advanced L1210
rather than only to increase survival time.
It may be noted that the effective dose levels of the halogenated derivatives are m a n y times that of amethopterin. The
increased antileukemic efficacy stems primarily from a sharply
reduced toxicity, on a molar basis, for the host, without as extensive a reduction in antileukemic effect. This permits the
safe employment of high dose levels with resulting increased
therapeutic effects.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
1348
Cancer
fleets the net effect of the drug on the host-tumor
relationship. In general, a drug which is capable
of eliciting important increases in survival time is
likely to inhibit tumor growth. However, a therapeutic effect may be achieved, even when a drug
has little or no direct effect on a tumor, i.e. (a)
in this laboratory, increase in survival time of
mice with leukemia LI~IO has been observed with
either progressive growth of the local tumor at the
site of inoculation of the leukemic cells, or a decrease in size or disappearance of the local tumor;
(b) the drug may have a salutary effect on the host
and thereby prolong life. There are other situations in which measurement of local tumor size
alone would not adequately dcscribc the therapeutic value of a drug. (a) The drug effect on the
tumor may require the mediation of the host. I t
may, for example, be necessary for the host to
alter the drug to an active form. Such effects, and
the toxic limitation of the drug for the host, could
bc masked if measurement is limited to local tumor
growth. (b) ~Yhere a tumor mctastasizes, measuremcnt of local tumor growth would fail to take into
account the infiltration of the disease into vital
organs, etc. (c) If tumor size or weight is measured
on a specified day and animal weight change is
taken into account, this may still not accnrately
reflect therapeutic response as adjudged by sm'vival time. For example, the tumor-inhibitory effect
or toxicity for the host could occur subsequent to
the time of measurement, etc.
Thus, a screening system or evaluation program
limited to measurement of local tunmr growth m a y
miss compounds which produce therapeutic effects
without extensive inhibition of tumor growth. In
addition, such a screen may give too many false
positives. Screening systems employing survival
time as the major response with measurement of
local tumor growth response as an ancillary index
of effect would appear to be highly desirable.
Both the tumor response and the survival time
of the tumorous animals should be characterized
with respect to the influence of nonspecific effects.
Regression lines may be established relating host
response, tumor response, and survival time response to underfeeding and weight loss of the host.
If the system is too sensitive to nonspecific toxicity, steps m a y bc taken to provide a more difficult
challenge. This may be accomI)lished by increasing
the inoculum level, delaying treatment until the
tumor is well established, etc.
Survival time measurement lends itself well to
precise quantitative evaluation of drug effectiveness. It may be employed in assay systems for the
determination of the specificity of action of drugs.
Assay systems with survival time used as the primary criterion of drug effectiveness permit the
Research
Vol. ~1, N o v e m b e r 1961
ordering of compounds in terms of their maximum
effectiveness, or in terms of their effectiveness at
a given cost in toxicity for the host. The therapeutic index, as a measure of the safety with which a
drug may be employed, provides an important index of the action of the drug. Taken in conjunction
with the assay of specificity of action it provides
a broad characterization of the action of the drug.
Preferably, the therapeutic index employed should
incorporate survival time as well as tumor size as
criteria of drug response. Such assay systems are
applicable also to study of combinations of drugs
in synergism and antagonism and to determination
of the influence of schedules of therapy. The assay system may be employed in pharmacological
and bi()chemical studies involving inhibition
analysis.
The assay systems should remain flexible. The
tumor challenge should be increased as more effective drugs arc found. Whereas moderate increases
in survival time may at first be the requirement for
drug effectiveness, extensive survival time and cure
with advanced systemic disease is the ultimate
objective. The assay systems with transplantable
tumor systems may, in turn, be extended to carcinogen-in(tuced an(t spontaneous tumor systems.
The development of wholly quantitative assay
procedures for the determination, rating, and
study of drug effectiveness, and the use of screening procedures relaled to these assay procedures
should provMe a firm foundation for study of the
relationship of the influence of drugs in animal
tumor systems and the clinic.
REFERENCES
1. AsiiEStlov, I. N.; HALL, E. A.; and FLON, H. Action of
Test Compounds on Bacterial Viruses. In: A. GELLIIOtlN
and E. III~SCImErm (cds.), Investigation of 1)iversc Systems for Cancer Chemotherapy Screening. Cancer Research (Suppl. 3), pp. 57-6~, 1955.
2. BELKIN, 5[. Effect of Podophyllin on Transplanted
Mouse Tumors. J. Pharmaco[. & Exp. Therap., 93:18-25,
1948.
3. B~EBEJ~,S., and HIT(:m_nGS, G. II. The I)cvcloping Frog
Embryo and Growth Inhibitors. In: A. (~ELLII()RNand
E. HmSCHBERG(cds.), Investigation of Diverse Systems
for Cancer Chemot]mrapy Screening. Cancer Research
(Suppl. 3), pp. 80-82, 1955.
4. BISC:IIOFF,F., and I,OXG,M. L. The Influence of Calories
per se upon the Growth of Sarcoma 180. Am. J. Cancer,
32:~18-~1, 1938.
5. BRADNER, W. W. Special Microbiological Systems. I.
Anaerobic Organisms. Ann. N.Y. Acad. Sci., 76:469-74,
1958.
6. BUYAN,W. R. Problems in Viral-Tumor Therapy. A Review. Nat. Cancer Inst. Monograph No. 4, pp. 363-76,
1960.
7. CLARKE, I). A. Mouse Sarcoma 180. In: A. GELLHORN
and E. HIItSCnBERG(eds.), Investigation of Diverse Systems for Cancer Chemotherapy Screening. Cancer Research (Suppl. 3), pp. 14-17, 1955.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN et al.--Preclinical Screening and Evaluation
8. CoB~, J. P. The Comparative Cytological Effects of Several Alkylating Agents on H u m a n Normal and Neoplastic
Cells in Tissue Culture. Ann. N.Y. Acad. Sci., 84: 513-42,
1960.
9. COLES, N. W.; JotINSTONE, R. M.; and QUASTEL, J. H.
The Effects of Alkylating Agents on the Incorporation of
Glycine-l-C TMinto Tissue Proteins in Vitro. Cancer Research, 20:15o-3-29, 1960.
10. CORNFIELD,J., and MANTEL, N. Some New Aspects of the
Application of Maximum Likelihood to the Calculation of
the ])osage Response Curve. J. Am. Stat. Assoc., 45:181o10, 1950.
11. COSTA, F. M.; BLUMENTHAL, G. H.; and GREENBERG,
1). M. Evaluation of the Carcinostatie Activity of Various
Agents on Mouse Aseites Tumors. In: J. LF.ITEI~ (ed.),
Cancer Chcnmtherapy Screening Data. VIII. Cancer Research, 20:685-90, 1960.
12. CRANSTON,E. M., and BIETER, R. N. Chemotherapeutic
Studies on Recently Isolated Transplants of Mouse Mainmary Tumor. In: J. LEITErl (cd.), Cancer Chenmthcrapy
Screening Data. II. Cancer Research, 19: l-1~, 1959.
1~. CREEC:H,H. J.; BHEUNINGEtl, E.; HANKWITZ, R. F., JR.;
POLS~V, G.; and WILSON, M. L. Quantitative Studies of
the Effects of Nitrogen Mustard Analogs and Other
Alkylating Agents on Ascitcs Tumors in Mice. In: J.
I,EITEn (cd.), Cancer Chemotherapy Screening Data. VII.
Cancer Research, 20: 471-94, 1960.
14. CI~EECII,S J.; ]IAUSCHKA, W. S.; HANKI~'ITZ,I{. F., JR.,
Ln"rLETO~', B. J. ; and ANDItE, J. Studies with Four Distinct Lines of the Ehrlich Ascites Tumor for the Evaluation of Twenty-seven Potential Inhibitors. In: A. GELLHOICN and E. HIRSCIIBERG (eds.), Investigation of I)iverse Systems for Cancer Chemotherapy Screening. Cancer Research (Suppl. 3), pp. 47-51, 1955.
15. ])AVIDSON, J. D., and FttEEMAN, B. B. The Effects of
Compounds upon p~9 Incorporation into Nucleic Acids as
a Technic for Selecting Antitumor Drugs. Cancer Research (Suppl. 3), pp. 97-101, 1955.
16. DIP.~oLo, J. A., and ROSENHELD, t{. Inhibition of Yeast
Respiratory Mutants as a Technic for Selecting Compounds for Cancer Chemotherapy Study. In: J. LEITEtt
(ed.), Cancer Chemotherapy Screening Data. IV. Cancer
Research, 19: 467-87, 1959.
17. E*GLE, II., and FOLEY, G. E. The Cytotoxie Action of
Carcinolytic Agents in Tissue Culture. Am. J. Med., 21:
739-49, 1956.
18.
. Susceptibility of Cultured Human Cells to Antitumor Agents. Ann. N.Y. Aead. Set., 76:584-41, 1958.
19. Fn.:LD, J. B.; BAscoY, L. T.; FH~LEa, D. A.; BORVCZKA,
A. ; COSTA,F. ; and KLOETZEL,M. C. Relationship of Structure to Tumor-Inhibitory Activity in the Phenanthrene
Series. ln: J. LEITER (ed.), Cancer Chemotherapy Screening Data. IV. Cancer Research (Suppl.), 19: 409-28, 1959.
20. FogEY, G. E. Observations on the Use of Microbiological
Systems in Screening Potential Antitumor Agents. Current Research in Cancer Chemotherapy (report no. 6),
pp. 3-20, 1956.
21. FOLEY, G. E.; EAGLE, H.; SNFLL, H. E.; KIDDER, G. ~'.;
and THAVE~, P. S. Studies on the Use of in Vitro Procedures for the Screening of Potential Antitumor Agents:
Comparison of Activity in Mammalian Cell Cultures and
Microbiological Assays Alone and in Combination with
Experimental Antitumor Activity. Ann. N.Y. Acad. Set.,
7t]:952-60, 1958.
2~. FOLF~]', G. E., and HANDLEIt, A. H. Tumorigenic Activity
of Tissue Cell Cultures. Ann. N.Y. Aead. Set., 76;506-12,
1958.
25. FOLEY, G. E.; McC.,tRTHr, R. E.; and BIxxs, V. M. A
1349
Comparative Study of the Use of Microorganisms in
the Screening of Potential Antitumor Agents. Ann. N.Y.
Acad. Sci., 76:418-41, 1958.
24. FRENCH, F. A., and FREEDLANDEII,B. L. Chemotherapy
Studies on Transplanted Mouse "rumors. In: J. LEITER
(ed.), Cancer Chemotherapy Screening Data. VII. Cancer Research (Suppl.), 20: 505-38, 1960.
25. GELLHORN,A. Invited Remarks oll the Current Status of
Research in Clinical Cancer Chemotherapy. Cancer
Chemotherapy Rep., 5:1-12, 1959.
26. GELLHORN, A.; KELLS, A.; and GOLINO, ~. M a m m a r y
Adenocarcinoma 755, Glioma 26, and Brown-Pearce Carcinoma. In: A. GELLItOItN and E. IImscHSERG (eds.),
Investigation of l)iverse Systems for Cancer Chemotherapy Screening. Cancer Research (Suppl. 3), pp. 58-43,
1955.
~27. GonmN, A. The E m p l o y m e n t of Methods of Inhit)ition
Analysis in the Normal and Tumor-Bearing Mammalian
Organism. In: J. GREENSTEIN and A. ttADDOW (eds.),
Advances in Cancer Research, 4:113-48. New York,
N.Y.: Academic Press, 1956.
08. GOLDIN, A.; (~OLDBERG, B.; ORTEGA, L. G.; FL'~.~I*NN,
R.; FAIMAN, F.; and SCIIOENiBACIt,E. S. Effectiveness of
Nitrogen Mustards in Retarding the Growth of Sarcoma
I80 with Reference to Route of Administration, Age of
Tumors, Food Intake, and Weight Loss of the Mice.
Cancer, 2: 865-76, 1949.
~29. GOLDIN, A., and HITMPIIREYS,S. R. Studies of I m m u n i t y
in Mice Surviving Systemic Leukemia LI~10. J. Nat.
Cancer Inst., 24: 288-300, 1960.
30. GOLDIN, A.; IIu~IPHREYS, S. R.; VENDITTI, J. M.; DENms, D.; and MANTEL, N. Growth Studies on Sarcoma 37
in Mice. J. Nat. Cancer Inst., 16:1207-20, 1956.
S1. GOLDIN, A.; IIuMPHUEYS, S. R.; VENDITTI, J. M.; and
M*~TEL, N. Prolongation of the Lifespan of Mice with
Advanced Deukemia (LI~10) by T r e a t m e n t with IIalogenated Derivatives of Amethopterin. J. Nat. Cancer
Inst., 22: 811-23, 1959.
3~. GOLDIN, A., and MXNTEL, N. The Employment of Combinations of Drugs in the Chemotherapy of Neoplasia: A
Re~qeu,. Cancer Research, 17: 635-54, 1957.
33. GOLDIN, A.; MANTEL, N.; GItEEN~OI:SF, S. W.; VENmTTI, J. M. ; and HrLivm~ms, S. R. Estimation of the Antileukemic Potency of the Antimetalmlite Aminopterin,
Administered Alone and in Combination with Citrovorum
Factor or Folio Acid. Cancer Research, 13: 843-50, 1955.
84. - - - - - - . Effect of Delayed Administration of Citrovorum
Factor on the Antileukemic Effectiveness of Aminopterin
in Mice. Ibid., 14:48-48, 1954.
35. - - - - .
Factors Influencing the Specificity of Action of
an Antileukemie Agent (Aminopterin). Time of Treatmen|. and Dosage Schedule. Ibid., pp. 311-14.
86. GOLDIN, A.; VENDITTI, J. ~I.; HUMPHRFYS, S. R.; DENNIS, D.; and MANTEL, N. Factors Influencing the Specificity of Action of an Antileukemic Agent (Aminopterin). Host Age and Weight. J. Nat. Cancer Inst., 16:
709- c21, 1955.
87. GOLDIN, A.; VFNDITTI, J. M.; HITMPHREYS, S. R.; DENNIS, D.; MANTEL, N.; and GJ~EENHOUSE, S. W. Factors
Influencing the Speciticity of Action of an Antileukemie
Agent (Aminopterin). Multiple Treatment Schedules
Plus Delayed Administration of Citrovorum Factor. Cancer Research, 15: 57-61, 1955.
88. - - - - --. A Quantitative Conlparison of the Antilcukemic
Effectiveness of Two Folic Acid Antagonists in Mice. J.
Nat. Cancer Inst., 15:1657-64, 1955.
989. - - - - --. An Investigation into the Synergistic Antileukcmic Action of Amethopterin and 6-Mercaptopurine in
Mice. Ibid., 16:129-38, 1955,
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
1350
Cancer Research
40. GoLmN, A.; VENDITTI, J. M.; HUMPIIREYS, S. R.; and
MANTEL, N. Modification of Treatment Schedules in the
Management of Advanced Leukemia with Amethopterin.
J. Nat. Cancer Inst., 17:~08-12, 1956.
4 1 . - - - - - - . Influence of the Concentration of Leukemic
Inoculum on the Effectiveness of Treatment. Science,
123:840, 1956.
4~. - - - - .
Comparison of the Relative Effectiveness of Folic
Acid Congeners against Advanced Leukemia in Mice. J.
Nat. Cancer Inst., 19:1133-35, 1957.
43. ~ .
Quantitative Evaluation of Chemotherapeutic
Agents against Advanced Leukemia in Mice. Ibid., 21:
495-511, 1958.
44. GOLDIN, A.; VENDITTI, J. M.; HUMPHREYS, S. R.;
SHUSTER, L., D.~RROW,R. A.; and MANTEL,N. Advanced
Leukemia As a Tool for Chemotherapeutic Studies. Acta
Internat. Unio contra cancrum, 16:64~2-50, 1960.
45. GOLDIN,A.; VENDITTI,J. M.; KLINE, I.; and MANTEL,N.
Evaluation of Antileukcmic Agents Employing Advanced
Leukemia Ll~10 in Mice. In: J. LEITER (ed.), Cancer
Chemothcrat)y Screening Data. IV. Cancer Research
(Suppl.), 19:4~29-66, 1959.
46. - - - - - - . Evaluation of Antileukcmic Agents Employing
Advanced Leukemia Ll~210 in Mice II. In: J. LEITER
(ed.), Cancer Chemotherapy Screening Data. VI. Ibid.,
20: 88~-448, 1960.
~7. GR*Y, J. II., and ELY, J. O. Inhibition Studies on Tissue
Cultures of Walker Carcinonm 256 and Chick F m b r y o
Ileart Cells. In: J. LEITER (ed.), Cancer Chemotherapy
Screening Data. I. Cancer Research (Suppl.), 18: 891-96,
1958.
48. HAVAS, It. F., and DONNELLY, A. J. Mixed Bacterial
Toxins in the Treatment of Tumors. IV. Response of
Methylcholanthrene-induccd, Spontaneous, and Transplanted Tumors in Mice. Cancer Research, 21:17-25,
1961.
49. IIEIDELBEI~-GER,C., and KELLY:R, R. A. The Effects of
Twenty-nine Compounds on Nuclei(; Acid and Protein
Biosynthesis in Slices of Flcxner-Jobling Carcinoma an(l
Rat Spleen. In: A. GELLIIORXand E. IIiRscnnE1m (eds.),
Investigation of Diverse Systems for Cancer Chemotherapy Screening. Cancer Research (Suppl. 3), pp. 106-1~2,
1955.
50. HEMMERLY,J., and DEMEREC, M. Tests of Chemicals for
Mutagenicity. Cancer Research (Suppl. 3), pp. 69-75,
1955.
51. HIaSCIInERG, E., and MEI~SON, G. Effect of Test Compounds on the Aggregation and Culmination of the Slime
Mold Dictyostdium discoideum. Cancer Research (Suppl.
3), pp. 76-79, 1955.
52. IIITCmNGS, G. H. Special Microbiological Systems. III.
The Use of Lactic Acid Bacteria for the Evaluation of
Purine and Pyrimidine Antagonists. Ann. N.Y. Aead.
Sei., 76:490-95, 1958.
53. HITCHINGS, G. II., and ELION, G. B. The Lactobacillus
casei Screening Test. In: A. GELLItOttN and E. HIItS(.'HnERG (cds.), Investigation of Diverse Systems for Cancer
Chemotherapy Screening. Cancer Research (Suppl. 3),
pp. 66-68, 1955.
5.~. HUMPIIREYS, S. R.; GOLI)IN, A.; VENDITTI, J. M.; and
MANTEL, N. The Influence of Amcthopterin on the Survival Time of Leukemic Mice with Respect to Food Intake, Body Weight Changes, and T u m o r Growth. J. Nat.
Cancer Inst., 20:211-18, 1958.
55. ~IUTNER, S. II.; NATHAN, H. A.; AARONSON, S.; BAKER,
H.; and SCnER, S. General Considerations in the Use of
Microorganisms in Screening Antitumor Agents. Ann.
N.Y. Acad. Sci., 76:457-68, 1958.
Vol. ~1, November 1961
56. JONES, R., JR.; KESSLER, W. B.; LESSNER, H. E.; and
RANE, L. New Developments in Cancer Chemotherapeutic Agents. Cancer Chemotherapy Rep., 10:99-108,
1960.
57. JONES, R., JR.; MCKENZIE,'D.; STEVENS, M. L., DUNraNG, W. F. ; and Ct'RTIS, M. R. Usefulness of Dunning
Leukemia I R C 741 for Quantitative Pharmacological
Studies of Cancer Chemotherapeutic Agents. Ann. N.Y.
Acad. Sci., 76:659-72, 1958.
58. KARNOFSKY,D. A. Assay of Chemotherapeutic Agents on
the Developing Chick Embryo. In: A. GELLHORN and
E. HIRSCHBERG (eds.), Investigation of Diverse Systems
for Cancer Chemotherapy Screening. Cancer Research
(Suppl. 8), pp. 83-85, 1955.
59. KLINE, I., and ACKER, R. F. The Effects of Chemical
Agents on Cells Propagated in Tissue Culture Systems.
Cancer Chemotherapy Rep., 9:73-77, 1960.
60. LAmONOV, L. F. Peptides, Amides, and Esters of Chloroethylamino Derivatives of Amino Acids and of Phenylalkancarboxylic Acids. A New Class of Antitmnor Coinpounds. Cancer Research, 21: 99-104, 1961.
61. LEITER, Z.; BOURKE, A. R.; SCIIEPARTZ, S. A.; and
WODINSKY, I. Screening 1)ata from the Cancer Chenlotherapy National Service Center Screening Laboratories
V. In: J. LEITER (ed.), Cancer Chemotherapy Screening
Data. VIII. Cancer Research (Suppl.), 20: 734-919, 1960.
60~. LEITER, J., and SCIINEII)ERMAN, M. A. Screening Data
from the Cancer Chemotherat)y National Service Center
Screening Laboratories. In: J. LETTER (cd.), Cancer
Chemotherapy Screening Data. II. Cancer Research
(Suppl.), 19:31-193, 1959.
63. LEITEI~, J.; WODINSKY,I. ; and BOURKE, A. R. Screening
Data from the Cancer Chemotherapy Nati(mal Service
Center Screening Laboratories II. In: J. LmT~.R (ed.),
Cancer Chemotherapy Screening Data. I II. Cancer Research (Suppl.), 19:309-96, 1959.
6r LmTEtr J.; WODINSKY, I.; BOURSE, A. R.; and SCHNEIDraYMAN, M. A. Screening Data from the Cancer CheinothcraI)y National Service Center Screening Laboratories
III. In: J. LEITER (ed.), Cancer Chemotherapy Screening
Data. IV. Cancer Research (Supt)l.), 19 : 488-575, 1959.
65. - - - - .
Screening Data from the Cancer Chemotherapy
National Service Center Screening I.aboratories IV. In:
J. I,EITER (cd.), Cancer Chernothcrapy Screening Data.
VII. Ibid., 20:539-670, 1960.
66. LEP.~GE, G. A., and GREENLEES, J. L. ]ncorl)oration of
Glycine-~-C x4 into Ascitcs Tumor-Cell Purines as a Biological Test System. In: A. GELLIIORN and E. IIn~SCH~EIm (eds.), Investigation of Diverse Systems for Cancer
ChcInothcrapy Screening. Cancer Research (Suppl. 3),
pp. 10~-5, 1955.
67. LEWIS, M. R.; HUGIIES, B.; BATES, A. L.; and BAIINEii,
C. T. Antitumor Activity of Certain Styrylquinolincs Administered Orally to Tumor-Bearing Rats. In: J. LEITER
(ed.), Cancer Chemotherapy Screening l)ata. VIII. Cancer Research (Suppl.), 20:691-97, 1960.
68. MARSIIALL,E. K., JR. The Dosage Schedule of Chemotherapeutic Agents. Pharmacol. Rev., 4:85-105, 1952.
69. MERKER, P. C.; BABA, T.; and SINGER, K. The Effect of
Steroids on Methylcholanthrene-induced Sul)cntaneous
Tumors in CaR and B A L B / c Mice. Cancer Research,
20:146i-68, 1960.
70. MILLAR, F. K.; WHITE, J.; BROOKS, R. II.; and MIDER,
G. B. Walker Carcinosarcoma 256 Tissue as a Dietary
Constituent, I. Stimulation of Appetite and Growth in the
Tumor-Bearing Rat. J. Nat. Cancer Inst., 19:957-67,
1957.
71. MILNER, A. N. Changes in the Enzymatic Activity of
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
GOLDIN et al.--Preclinical Screening and Evaluation
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
Malignant Cells after Chemotherapy. Cancer Chemotherapy Rep., 10:115-19, 1960.
MONTGOMERY,J. A. The Relation of Anticancer Activity
to Chemical Structure: A Review. Cancer Research, 19:
447-63, 1959.
RAVDIN, I. S. The Cancer Chemotherapy Program. Cancer Chemotherapy Rep., 5:13-17, 1959.
REILLY, H. C. An Evaluation of the Use of Antimicrobial
Activity as a Screening Procedure for Tumor-inhibiting
Agents. In: A. GELLHORNand E. HIRSCHBERG (eds.), Investigation of Diverse Systems for Cancer Chemotherapy
Screening. Cancer Research (Suppl. 3), pp. 63-65, 1955.
Review of the Cancer Chemotherapy National Service
Center Program. Cancer Chemotherapy Rep., 7: ~25-46,
1960.
Rous, P. The Influence of Diet on Transplanted and
Spontaneous Mouse Tumors. J. Exp. Med., 20:433-51,
1914.
SCHABEL,F. M., JR., and SKIPPER, H. E. Effects of a Series of Tumor-inhibiting Agents and Related Compounds
Against a Spectrum of Mouse-adapted Viruses. In: A.
GELLHORN and E. HIRSC~RERG (eds.), Investigation of
Diverse Systems for Cancer Chemotherapy Screening.
Cancer Research (Suppl. 3), pp. 52-56, 1955.
SCIIMIDT,L. H. Observations on the Therapeutic Activity
and Toxicity of 2-Chloroethylmethanesulfonate--CB1506 (NSC-18016). Cancer Chemotherapy Rep., 9: 56-68,
1960.
SCHOLLER,J. Practical and Theoretical Considerations in
the Use of Induced and Spontaneous M a m m a r y Tumors
in Cancer Chemotherapy. Ann. N.Y. Acad. Sci., 76:85560, 1958.
SCHOLLER,J.; PHILIPS, F. S.; and BITTNER, J. J. Assays
with First- or Second-Generation Transplants of Spontaneous M a m m a r y Adcnocarcinomas of Mice. In: A.
GELLHORN and E. HIRSCRBERG (eds.), Investigation of
Diverse Systems for Cancer Chemotherapy Screening.
Cancer Research (Suppl. 3), pp. 32-37, 1955.
SCHULTZ, J.; ROTHMAN, N.; and ARONSON, M. M. The
Growth and Morphogenesis of Drosophila melanogaster as
Criteria for Screening Tests. Cancer Research (Suppl. 3),
pp. 86-96, 1955.
SCOTT, D. B. M.; ULBRICHT, T. L. V.; ROGERS, M. L.;
CHV, E.; and RosE, C. Effect of Substituted Pyrimidines
on Growth and Biosynthesis of Microorganisms. In:
J. LEITER (ed.), Cancer Chemotherapy Screening Data.
II. Cancer Research (Suppl.), 19:15-30, 1959.
SKIPPER, H. E. Observations of the Retiring Chairman of
the Cancer Chemotherapy Review Board. Cancer Chemotherapy Rep., 8:141-52, 1960.
SKIPPER, H. E.; MONTGOMERY, J. A.; THOMSON, J. R.;
and SCHABEL, F. M., JR. Structure-Activity Relationships and Cross-Resistance Observed on Evaluation of a
Series of Purine Analogs Against Experimental Neoplasms. Cancer Research, 19:425-37, 1959.
- - - - - - . Structure-Activity Relationships and Cross Resistance Observed on Evaluation of a Series of Purine
Analogs Against Experimental Neoplasms. In: J. LEITER
(ed.), Cancer Chemotherapy Screening Data. III. Cancer
Research (Suppl.), 19: 287-308, 1959.
SKIPPER, H. E., and THOMSON, J. R. Effects of a Series of
Tumor-inhibiting Agents and Related Compounds on
L1210 Leukemia and Drug-resistant Lines Thereof. In:
A. GELIJ~ORN and E. HIRSCHBERG (eds.), Investigation
of Diverse Systems for Cancer Chemotherapy Screening.
Cancer Research (Suppl. 3), pp. 44-46, 1955.
SLOBODA,A. E. A Method for the Quantitative Comparison of Anticancer Agents with Special Reference to a
88.
89.
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
102.
1351
Series of Folic Acid Antagonists. J. Pharmacol. & Exper.
Therap., 128: 419-27, 1960.
STOCK, C. C. Experimental Cancer Chemotherapy. In:
J. GREENSTEIN and A. HADDOW (eds.), Advances in Cancer Research, 2:425-92. New York: Academic Press,
1954.
STOCK, C. C. ; CLARKE, D. A.; PHILIPS, F. S. ; and BARCLAY, R. K. Sarcoma 180 Screening Data. In: J. LEITEH
(ed.), Cancer Chemotherapy Screening Data. V. Cancer
Research, 20:1-172, 1960.
SUGIURA, K. Studies in a Spectrum of Mouse and R a t
Tumors. In: A. GELLHORN and E. HIaSCHB~RG (eds.),
Investigation of Diverse Systems for Cancer Chemotherapy Screening. Cancer Research (Suppl. 3), pp. 18-27,
1955.
. Relative Sensitivity of the Solid and Ascites
Forms of Sarcoma 180 and Ehrlich Carcinoma to Inhibitory Compounds. Ann. N.Y. Acad. Sci., 76: 575-85, 1958.
SZYBALSKI,W. Special Microbiological Systems. II. Observations on Chemical Mutagenesis in Microorganisms.
Ann. N.Y. Acad. Sci., 76:475-89, 1958.
TANNENBAUM,A. The Role of Nutrition in the Origin and
Growth of Tumors. In: F. R. MOULTON (ed.), Approaches
to T u m o r Chemotherapy, pp. 96-127. Washington, D.C. :
Am. Assoc. Adv. Sc., 1947.
TARNOWSKI,G. S. T u m o r Inhibition Studies on RC Mammary Adenocarcinoma. In: A. GELLHORN and E. HIRSCH~ERG (eds.), Investigation of Diverse Systems for Cancer
Chemotherapy Screening. Cancer Research (Suppl. 3),
pp. 28-31, 1955.
TOPLIN, I. Preliminary Report on a Tissue Culture Cytotoxicity Test for Large-Scale Cancer Chemotherapy
Screening. Cancer Chemotherapy Rep., 3:12-15, 1959.
TURNER, C. F. Experimental T r e a t m e n t of Tumors in
Mice. In: J. LEITER (ed.), Cancer Chemotherapy Screening Data. I. Cancer Research (Suppl.), 18: 543-606, 1958.
VENDITTI, J. M.; GOLDIN, A.; and KLINE, I. The Influence of T r e a t m e n t Schedule on the Chemotherapy of Advanced Leukemia L1210 in Mice. Cancer Chemotherapy
Rep., 6:55-57, 1960.
VENDITTI,J. M.; HUMPHREYS,S. n.; and GOLDIN,A. Investigation of the Activity of Cytoxan against Leukemia
L1210 in Mice. Cancer Research, 19:986-95, 1959.
VENDITTI, J. M.; HUMPHREYS, S. R.; MANTEL, N.;
KLINE, I. ; and GOLDIN, A. Evaluation of Antileukemic
Agents Employing Advanced Leukemia L1210 in Mice
III. Congeners of Folic Acid. I n : J. LEITER (ed.), Cancer
Chemotherapy Screening Data. VIII. Cancer Research
(Suppl.), 20: 698-733, 1960.
VENDITTI, J. ~ . ; SCItRECKER, A. W.; MEAD, J. A. R.;
]~LINE, I. ; and GOLDIN, A. Influence of the Route of Administration on the Relative Effectiveness of 3',5'-Dichloroamethopterin and Amethopterin against Advanced
Leukemia (L1210) in Mice. Cancer Research, 20:145156, 1960.
WHITE, F. R. The Relationship between Underfeeding
and T u m o r Formation, Transplantation, and Growth in
Rats and Mice. Cancer Research, 21:281-90, 1961.
ZUBROD, C. G.; SCHNEIDERMAN, M.; FREI, E., I I I ;
BRINDLEY, C.; GOLD, G. L.; SHNIDER, B.; OVIEDO, R.;
GORMAN, J. ; JONES, R., JR. ; JONSSON, U.; COLSKY, J.;
CHALMERS, T.; FERGVSON,B. ; DEDERICK, M. ; HOLLAND,
J. ; SELAWRY,O. ; REGELSON,W. ; LASAGNA,L.; and OWENS,
A. H. Appraisal of Methods for the Study of Chemotherapy of Cancer in Man: Comparative Therapeutic Trial of
Nitrogen Mustard and Triethylene Thiophosphoramide.
J. Chronic Diseases, 11:7-33, 1960.
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.
Preclinical Screening and Evaluation of Agents for the
Chemotherapy of Cancer: A Review
Abraham Goldin, John M. Venditti and Nathan Mantel
Cancer Res 1961;21:1334-1351.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/21/10/1334.citation
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 17, 2017. © 1961 American Association for Cancer
Research.