(CANCER RESEARCH 49, 5288-5293, October 1, 1989]
In Vitro Clonogenic Growth and Metastatic Potential of Human Operable Breast
Cancer1
Yasuo Nomura,2 Hideya Tashiro, and Kazufumi Hisamatsu
Department of Breast Surgery, National Kyushu Cancer Center Hospital, 3-1-1, Notarne, Minami-ku, Fukuoka 815, Japan
ABSTRACT
In 254 patients with operable (International Union against Cancer
(1972) Stages I, II, and III| breast cancer, the relations between in vitro
clonogenic growth in soft agar of primary breast cancer tumors and their
metastatic potential expressed by the relapse-free survivals (RFS) as
well as overall survivals were studied.
Sixty-four % (163 of 254) of cancers formed distinct colonies (30 or
more colonies in a single dish, or 10 or more colonies in plural dishes).
Other breast cancers (36%, 91 of 254) were designated to be negative for
the clonogenicity. There was no correlation between the positive or
negative clonogenicity and clinicopathological characteristics in breast
cancer patients, including the age of patients, menopausa! status, tumor
size, T classification, International Union against Cancer stage, histological type (Japanese Breast Cancer Society), histologically proved axillary
lymph node metastasis, and estrogen receptor (ER).
At the time of median follow-up of 43 mo (range, 25 to 61 mo) after
mastectomy, a recurrence of malignancy occurred in 19.0% (31 of 163)
of the patients with positive clonogenic tumors, and in 8.8% (8 of 91) of
those with negative clonogenic tumors (/' = 0.03). There also was a
significant difference (P < 0.03 by log rank test, P < 0.05 by generalized
Wilcoxon test) in RFS curves between positive and negative clonogenicity
groups. These differences in RFS were also noted in Stage II patients in
favor of the negative colony formation group. In ER-negative cancer
patients, the RFS of patients with positive clonogenic cancers was shown
to be worse (P < 0.03 by log rank test, P < 0.05 by generalized Wilcoxon
test) than patients with negative clonogenic cancers. There was no differ
ence in RFS in ER-positive cancer patients. There was a trend (/' = 0.09
by log rank test) of worse overall survival rate in patients with positive
clonogenic growth. In a multivariate comparison using the principal
component analysis composed of factors including positive node, T clas
sification, histológica!type, age, ER, and colony formation, the clonogen
icity showed a significant effect on the recurrence of malignancy and also
on the survival of the patients after mastectomy.
In conclusion, in vitro clonogenic growth of the primary tumor of
breast cancer was shown to be one of the independent factors of metastatic
potential in operable breast cancer patients after mastectomy.
INTRODUCTION
Since the first reports on the utilization of a semisolid me
dium system for in vitro growth of human tumors (1, 2), most
investigators have directed their interests on the correlation of
in vitro drug effects and patients' clinical responses to chemo
therapy, as well as in vitro screening of new drugs.
On the other hand, there have been few reports on the
clonogenicity of human tumors correlated with the metastatic
potential of the tumors or a prognostic factor for recurrence of
primary cancers (3-8). The reported articles have been so far
very controversial, probably because of a relatively small num
ber of patients studied and of unselected patient populations.
By the analysis of clonogenicity of pBC3 related to the differ-
entiation and proliferation parameters, it has been reported that
cancer cell clones that grow in semisolid medium may represent
the most malignant and rapidly growing fractions of the heterogenous population of primary tumors (9-12). In experimental
animals, however, it appears to be still inconclusive whether in
vitro CF of cancer cells correlates with their metastatic potential
in vivo (13, 14).
The present study deals with the possibility that the clono
genicity of human breast cancer is related to the metastatic
potential of primary cancer after mastectomy. This will be
linked to the development of an independent prognostic factor
for recurrence and survival of patients with operable breast
cancer.
MATERIALS
AND METHODS
In Vitro Clonogenic Assay. pBC tumors were biopsied, removed from
the resected specimens at the time of mastectomy, and dissected from
surrounding normal or fat tissues. The tumors were put in sterilized
vehicles filled with Hanks' balanced salt solution and immediately sent
to the laboratory.
Tumor tissues were minced into small cubes and centrifuged with
1000 rpm for 10 min. The pellets were washed twice and placed in an
enzyme-dispersed medium [RPMI 1640 with 10% DCC-treated PCS,
insulin (0.25 lU/ml), and 0.07% collagenase] for 30 min, 37°C,being
agitated by a stirrer. The cells were then washed twice and filtered
through a stainless mesh. The filtrates were centrifuged twice, and
viable cells were counted after being stained with trypan blue solution
and then adjusted to a cell suspension of the concentration of 1 to 2 x
IO6cells/ml.
Cells were cultured in a bilayer agar system as described by Ham
burger and Salmon (1). The feeder layer consisted of Ham's F-12 (Flow
Co.) medium and Dulbecco's minimal essential medium (Gibco, Grand
Island, NY) added with 20% PCS (Flow), 20% HS, 2-mercaptoethanol
(10 MM)>insulin (0.5 lU/ml), penicillin (100 ¿¿g/ml),
kanamycin (0.1
mg/ml), and 0.5% agar.
For the upper layer, 2 to 5 x IO5 cells were suspended in enriched
RPMI 1640 medium supplemented with 20% PCS, 20% HS, and
insulin (0.5 lU/ml) and mixed with the equivalent volume of 0.6% agar.
The resultant mixtures (final concentration, 0.3% agar) were pipetted
on top of the previously prepared underlayer.
For the deprivation of endogenous steroids including estrogens, all
sera used in the experiment were pretreated as follows. The sera were
incubated twice with DCC (0.5% activated charcoal and 0.05% dextran)
and centrifuged for 30 min at 45°C.The completeness of DCC treat
ment was checked by comparing the plating efficiency of colony for
mation of MCF-7 cells on plastic dishes in the presence of 17/3-estradiol
(10 HM)and the absence of estradici (DCC-treated medium alone), as
described previously (15).
The bilayer cultures on the 35-mm Nunc plastic dishes were incu
bated at 37°Cin a 5% CO2, 100% humidified atmosphere.
Evaluation of Clonogenicity. Colony counts were made between 10
and 14 days after plating, with an inverted phase microscope at mag
nifications of x40 or xlOO. Aggregates of 30 cells or more were
considered colonies. This study has been made in a series of drug
sensitivity testing of hormonal and anticancer agents in the soft agar
culture system. A duplicate culture for nontreatment control (DCCtreated sera alone) and cultures in duplicate were made by adding E2
17/3-estradiol (10 HM) and other hormonal and chemotherapeutic
agents. The number of dishes used for soft agar culture depended on
the tumor size and cellularity of the individual tumors.
5288
Received 2/1/89; revised 5/26/89; accepted 6/13/89.
The costs of publication of this article were defrayed in part by the payment
of page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work is supported in part by a Grant-in-Aid (62S-1) from the Ministry
of Health and Welfare of Japan.
2To whom requests for reprints should be addressed.
3 The abbreviations used are: pBC, primary breast cancer; CF, colony formation
in soft agar; DCC, dextran-coated charcoal; ER, estrogen receptor; PgR, proges
terone receptor; RFS, relapse-free survival; OS, overall survival; UICC, Interna
tional Union against Cancer; PCS, fetal calf serum; HS, horse serum.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
CLONOGENIC
GROWTH AND METASTASIS OF BREAST CANCER
The clonogenicity of individual tumors was evaluated and classified
into the following 4 groups: 0, no CF in any of the dishes; 1, the
number of colonies grown was shown to be less than 30/dish in a single
dish, or less than 10 colonies in plural dishes; 2, CF of 30 colonies or
more in a single dish or 10 colonies or more in plural dishes; 3, CF of
30 or more colonies in at least 2 dishes.
Characteristics and Prognosis of pBC Patients. From December 1981
to the end of 1985, in a total of 254 patients with operable breast cancer
[UICC (1972) Stages I, II, and III], cancer tissues were provided for the
clonogenic assay. Although all breast cancer tumors were consecutively
tried for the clonogenic assay, only tumors of enough size left after
histological and hormone receptor examinations were provided for the
assay. Therefore, breast cancer patients in this study had a tendency to
include larger tumors in the total series of patients in the hospital. ER
assay, as well as the histological examination of primary tumors, was
performed in the neighboring cancer tissues. ER assay was done by the
DCC method, 3.0 fmol/mg of protein or more of the binding sites being
considered to be positive for receptor (16).
The follow-up of the patients was performed every 3 mo for 2 yr
after mastectomy, and thereafter twice a year. Recurrence of malignancy
was checked by physical examinations, biochemical examinations, Xrays of chest and bones, liver and bone scintigraphy, and computed
tomographic roentgenography of brain and/or liver, if necessary.
Table 1 In vitro clonogenic growth of human breast cancer in soft agar
Colony
formation(no.)0°1No.
cases36
formationSS
(14.2),* no colony
of
(21.7), less than 30 colonies in a single dishor
than 10 colonies in plural dishes
20 (7.9),
3
less
30 or more colonies in a single dish or 10 or
more colonies in plural dishes
143 (56.3), 30 or more colonies in plural dishes
Total
254(100)
" Nos. 0 + 1 (91; 36%); colony negative, nos. 2 + 3 (163; 64%); colony positive.
b Numbers in parentheses, percentage.
The relapse of malignancy and patients' survival were evaluated at
the end of January 1988, with a median follow-up period of 43 mo
(range, 25 to 61 mo) after mastectomy.
Statistics. The RFS and OS curves were made by the actuarial method
or Kaplan-Meier method, and their statistical differences were analyzed
by log rank and generalized Wilcoxon tests. Clinical characteristics of
the patients in relation to the in vitro clonogenicity were compared and
checked by the x2 method or by the Student t method.
Multivariate comparisons of clinicopathological factors including
the clonogenicity were done using the principal component analysis in
order to find out the degrees of effect on the recurrence of malignancy
or the survival of the patients.
RESULTS
pBC tumors from 254 female patients with operable cancer
were provided for the in vitro clonogenic assay. As shown in
Table 1, 143 of 254 or 56.3% formed distinct colonies (30 or
more colonies in plural dishes), and 20 (7.9%) showed 30 or
more colonies in a single dish or 10 or more colonies in plural
dishes. In these 143 cancers with comparable colony formation,
22% or 31 cancers showed the estrogen-dependent growth
stimulation; that is, by adding 17/3-estradiol (10~8 M) in the
medium, the mean number of colonies per dish increased to
150% or more of the DCC-treated control. These cancers were
designated positive for clonogenicity, being 64.2% (163 of 254).
In 21.7% or 55 of 254 cancers, colonies were formed in less
than 30 in a single dish or less than 10 in plural dishes, while
no colonies or less than 10 colonies in plural dishes was found
in 36 or 14.2%. These cancers were considered as a group for
negative clonogenic growth (35.8%, 91 of 254).
Table 2 indicates the correlation of positive or negative CF
with clinical factors of the patients, including the age of the
patients, menopausa! status, tumor size, T classification, UICC
stage, histological type (Japanese Breast Cancer Society), histologically proved axillary lymph node metastasis, and ER.
Table 2 Clinical factors and in vitro clonogenicity in operable breast cancer
formation+16349.4
FactorNo.
cases25413«1182
valueNS*(/>
patientsAgeMenopausa!
of
0.8°92
±
1.644 ±
0.13)NS(/>
=
(56)'
71
(44)3.2
(48)
47(52)3.0
0.27)NS(/>=0.17)NS
=
1.30(0)
±
1.02(2)
±
38
181
171637
21(13)
118(72)
12(7)
12(7)19(12)
17(19)
63 (69)
5(5)
4(4)18(20)
174
43120
115(71)
29(18)83(51)
59 (65)
14(15)37(41)
statusSize
Post-01
(cm)T
classificationStageHistological
2
34In
inn:0
type
Papillotubular carcinoma
Solid-tubular carcinoma
Scirrhous carcinoma
typesHistological
Special
Lymph node
MetastasisERCategoryPre-
88
24
22126
n: 1-3
n:S4ER+
58 (36)
11(7)
11(7)77
(47)
85
55 (34)
43143111Colony 31(19)98
(60)
65 (40)-9151.9
ER-No.of
0.20)NS(/>
(P =
0.21)P
=
30 (33)
13(14)
11(12)49
0.077NS
=
(54)
30 (33)
12(13)45
0.43)NS</>
(P =
(49)
46(51)P
= 0.14)
°Mean ±SE.
* NS, not significant.
' Numbers in parentheses, percentage.
5289
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
CLONOCENIC
GROWTH AND METASTASIS OF BREAST CANCER
There were no significant differences between positive and
negative clonogenicity patient groups in the distributions of
categories in these factors including axillary node metastasis.
There was a trend that scirrhous cancer and special types
(mucinous carcinoma, medullary carcinoma, invasive lobular
carcinoma, etc.) showed a lower but not significantly different
clonogenic growth as compared with other types of cancer.
Thus, the in vitro clonogenicity of pBC appears to be an
independent factor from the usual clinical characteristics of
breast cancer patients.
The recurrence of malignancy of these 254 patients was
studied according to in vitro CF, at the time of median followup of 43 mo after mastectomy. As shown in Table 3, a recur
rence occurred in 19.0% (31 of 163) of patients with positive
clonogenic tumors, while in 8.8% (8 of 91) of those with
negative clonogenic tumors (P = 0.03). As shown in Fig. IA,
these relations were confirmed in the actuarial RFS curves of
the two groups. There is a significant difference (P < 0.03 by
log rank test, P < 0.05 by generalized Wilcoxon test) in RFS
between positive and negative clonogenicity groups.
Fig. IB shows OS of the CF-positive and -negative patients.
78
68
STAGC 1
58
COLONY ' U=19)
48
38
COLONY •¿
ÕN=I8>
I a>
10
a
21
36
rVlNTHS
AFTER
IS
60
MASTECTOMY
"S
X
in
STAGEII
COLONY
*
COLONY
-
(N=115>
<N=$9)
Table 3 Clonogenic growth in soft agar and recurrence after mastectomy in
primary breast cancer
21
J6
MONTHS AFTER
Total+
Yes
31(19.0)°
163(100)8(8.8)
No
132(81.0)
91(100)Total
P<0.01
P<0.05
83(91.2)
254(100)Numbers
39(15.4)
(LOGRANK TEST)
(GENERALIZED WlLCOXON TfST)
108
215(84.6)
98
°ftA•<->Bai>>1=RecurrenceColony
88
percentage.X*
in parentheses,
0.03.¿988078605B48
= 4.7; /> =
78
M
Sl«0! Ill
58
^COLONY
48
30
1
28
(N=163)COLONY*
3020U0P"""-*-^^^-^___
(N=91>,0
Fig. 2. Relapse-free survival curves of operable breast cancer patients accord
ing to UICC stage and in vitro clonogenicity.
-
60COLONY
12
» VS
21
36
18
COLONY - : P<0.03
TEST)P<0.05
(LOGRANK
TEST)KIBB
(GENERALIZED WILCOXON
„¿988070685048
,
ÕN-1631COLONY
*
3828teB~*-~~-^_^COLONY
(N-91)0
12
-
60COLONY
21
36
18
«vs COLONY
- : p<:0.09 (LOGRAN«
TEST)
Fig. 1. A, relapse-free survival curves of operable breast cancer patients ac
cording to the in vitro clonogenicity. B, overall survival curves of operable breast
cancer patients according to the in vitro clonogenicity.
There is a trend (P < 0.09 by log rank test) of lower survival
rate in patients with positive clonogenic growth. However, this
difference is still marginal, probably because of the few deaths
(20 deaths in colony-positive and 14 in colony-negative groups,
respectively).
When the patients were analyzed by the UICC stage, there is
a statistically significant difference in RFS in favor of the
negative CF group in Stage II patients (Fig. 2). In Stages I and
III, there is no significant difference in RFS, probably because
of the smaller number of cases in each group. As indicated in
Fig. 3, in node-negative patients the RFS curve of patients with
negative CF tumors did not show a significant difference from
that of patients with positive CF tumors. In node-positive
patients, there was a trend of lower RFS curve in patients with
positive clonogenicity as compared with the negative counter
part.
The influences of in vitro clonogenicity on RFS in combina
tion with ER in tumors are shown in Fig. 4. In these 254
patients, the presence or absence of ER in primary tumors did
not affect RFS or OS of the patients. In 143 patients with ERpositive cancers, there was a tendency of lower RFS in the
5290
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
CLONOGENIC GROWTH AND METASTASIS OF BREAST CANCER
X
X
100
100
to
70
68
ER»
50
N-
COLONY*
(N-77)
N-
COLONY-
(N-19)
COLONY + (N=98)
40
30
21
36
18
MONTHS
AFTER
HASTECTOMY
I
COLONY - (N=1S)
a
21
60
56
18
MONTHS AFTER MASTECTOMY
60
..JJ.Ll.J.l.J..
70
ER COLONY * (N=65)
N* COLONY+
(N*86)
COLONY - <N=16i
N* COLONY- (N-12)
3
0
21
36
MONTHS
AFTER MASTECTOMY
18
60
12
21
36
MONTHS
LOGRANKX2-2.31
P'0.129
ÖENERAL1ZED WlLCOXON
0.1>p>0.05
Fig. 3. Relapse-free survival curves of operable breast cancer patients accord
ing to lymph node metastasis and in vitro clonogenicity.
AFTER
18
60
MASTECTOMY
P<0.03
(LOG-RANKTEST)
P<0.05
(GENERALIZED
WlLCOXON
TEST)
Fig. 4. Relapse-free survival curves of operable breast cancer patients accord
ing to ER and in vitro clonogenicity.
independent factor for recurrence of malignancy or survival of
patients. The evaluation of biological characteristics of clonopositive clonogenicity patients; however, there was no signifi
genic cells from human breast cancer is thought to be best and
most comprehensively done by estimating RFS and OS of the
cant difference.
On the contrary, in ER-negative patients, there was a distinct
patients after mastectomy, although it needs many patients
with pBC and longer follow-up periods.
difference (P < 0.03 by log rank test, P < 0.05 by generalized
Wilcoxon test) in RFS between colony-positive and colonyAs shown in Table 2, the clonogenicity of pBC tumors has
negative groups, in favor of the latter group.
no statistically significant correlation with already recognized
Using the principal component analysis, clinicopathological
prognostic factors, such as the age of patients, menopausa!
factors were evaluated for their degrees of contribution to the status, tumor size, UICC (1972) clinical stage, histological type
recurrence of malignancy (Fig. 5) or the survival of the patients
of cancer, axillary lymph node status, and ER. Previous workers
(Fig. 6). As shown in Fig. 5, prognostic factors influencing the have also shown no relations of clonogenicity with clinical
recurrence were revealed to be in the order of the number of parameters (4, 8). While Maestroni et al. (17) noted that CF
positive nodes > T classification > histológica! type > CF > correlated inversely to ER and PgR, Sandbach et al. (18),
age > ER. On the other hand, the number of positive nodes > Dittrich et al. (18), and Aapro and coworkers (8) found no
T classification > histological type > age > CF > ER is shown
relation between CF and hormone receptors in tumors. Thus,
to be effective in the survival of patients (Fig. 6). These multiit is likely that in vitro clonogenicity of breast cancer is desig
variate analyses suggest that clonogenicity is one of the signif
nated to be an independent characteristic of pBC from other
icant and independent factors of the recurrence of malignancy
clinical and histopathological factors.
and the survival of patients with operable breast cancer.
However, in vitro clonogenicity of human breast cancer may
As shown in Table 4, there was no significant difference in be the expression of other type(s) of tumor biology. Touzet et
al. (9) and Hug and coworkers (10) noted that well-differen
recurrence of malignancy between the patients whose tumors
showed less than 30 colonies in a single dish or less than 10 tiated pBC formed fewer colonies than poorly differentiated
colonies in plural dishes, and those without CF (4 of 36 versus tumors. According to Moezzi et al. (11), there was a significant
4 of 55). On the other hand, when distinct colonies were positive relationship between the degree of nuclear differentia
obtained according to the criteria shown previously, the number
tion and focal microscopic necrosis in the original specimen
of colonies per dish did not affect the incidence of recurrence
and its subsequent clonogenicity. Other histological parameters
after mastectomy. Thus, in vitro clonogenicity of breast cancer
did not show any relationship with clonogenicity. Flentje et al.
tumors appears to be an independent prognostic factor of pBC (12) compared in 65 various human solid tumors in vitro CF to
and an indicator for worse prognosis.
proliferation parameters simultaneously obtained by DNA flow
cytometry of the same tumor specimens. They showed that
colony growth was enhanced in aneuploid tumors in compari
DISCUSSION
son to diploid tumors, with a positive relation between the
percentage of S phase and colony growth.
The present study has been performed to investigate whether
Thus, it is possible that cancer cell clones that grow in
or not the ability of breast cancer cells to grow in vitro is an
5291
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
CLONOGENIC
GROWTH AND METASTASIS OF BREAST CANCER
astatic potential in vivo. By correlating the clonogenicity of
breast cancer in soft agar with the recurrence of malignancy or
patient's survival after mastectomy, the relationship between
OF
COEFFICIENTcmic=:[Cl~~Z]ini
NO1234567FACTORTHlSTOLOGICAL
EFFECTmm§¡•iiCOEFFICIENT4.
OF
639E-02
3. 168E-03
2
7. 509E-04
1
-9.411E-02
30SOL-TUB
1.278E-01-2.
935E-02
TYPEERAGE
2. 042E-02
Pff-JÕB
-1. 599E-K
SPECIflL
987E-BS-2.
6.
SCIRRH+41-60
1n13
742E-0¿
532E-0*7.
3.
772E-03
GROUPNo.
-1. 830E-02
-40
-1.
655E-03-7.
61->=401-3•fEFFECT
clonogenicity and metastatic potential could be elucidated.
However, there have been few reports concerning these relations
in human cancer, such as ovarian cancer (3), head and neck
squamous cell carcinoma (4, 5), multiple myeloma (6), and
breast cancer (7, 8). These studies, however, were based on the
relatively small number of patients studied and unselected
patient populations, thus being insufficient to draw any conclu
sion by analyzing the clonogenicity of human cancer as an
expression of metastatic potential in vivo, in combination with
standard prognostic factors.
Recently, Aapro and coworkers (8) reported, in 61 patients
with pBC, no correlation of CF with any of the standard clinical
parameters studied. After excluding far-advanced cancers, they
showed that there was a significant difference in RFS but not
OF
COEFFICIENT•[^]
EFFECTmm•ER
OF
NO1234567FACTORTHlSTOLOGICAL
899e-0a
OFPOSITIVE
3. 368E-0a
NODESCOLONYCONSTANTDEGREE
-9.
961E-03-1.037E-02
7. 919E-03
2
9. 703E-03
1
-1. 355E-01
IB111
247E-81-1.
1.
1.857E-B26.
426E-02
481E-B2CATEGORY4
TYPEDEGREE
1.417E-02
RECURRENCE MEAN
No
YES7.
1KS~;SS]
30SOL-TUB
PflP-TUB
-6. 80ÃŽE-0C SPECIflL
772E-02
-6. 332E-03
-1.231E-023.
1AGE
Fig. 5. A principal component analysis of clinicopathological factors including
the clonogenicity for the recurrence of malignancy in operable breast cancer
patients.
-4.
998E-036.
783E-03
GROUPNo.
semisolid medium may be the most malignant and rapidly
growing fractions of the heterogenous populations in the pri
mary tumors, or even that the colony-forming cells were se
lected to be more aneuploid during the cultivation in soft agar
(20). Therefore these colony-forming cells may be only repre
senting a small subpopulation of rapidly growing cell clones of
the original heterogenous populations.
Although these histological expressions of tumor malignancy
as well as cell proliferation kinetic parameters have been noted
to have significant prognostic significance (21), probable ag
gressiveness of the primary cancer that grows colonies in vitro
may not be necessarily directly connected to the metastatic
potential of breast cancer.
In experimental animal model systems, there have been di
verse opinions concerning the relation of cancer cells in semisolid medium with their metastatic potential in vivo. Using
various rat mammary adenocarcinoma cell clones with known
metastatic potential, Nicholson et al. (13) found that the ability
of tumor cells to grow in vivo and to spontaneously metastasize
is unrelated to their growth potential in 0.3 or 0.6% agarose.
On the contrary, according to Price (14), primary mouse mam
mary tumors differed widely in colony-forming efficiency in
vitro, and these differences were closely associated with the
ability of dissociated cells to form lung colonies after i.v. inoc
ulation. Thus she suggested that stem or clonogenic cell content
of primary tumors is one of the important determinants of the
metastatic phenotype. These animal models, however, appear
to have an essential difference from human breast cancer in
their higher colony-forming efficiency in vitro and higher met
SCIRRH+41-60
879E-03
2. 548E-03
-2.
335E-02-7.
-40
61->=4
833E-02
OFPOSITIVE
2. 701E-02
0
NODESCOLONYCONSTANT.M'COEFFICIENT1.577E-02
-4.
172E-04-8.
1-3+EFFECT
178E-03
465E-027.
1.
434E-02CATEGORY4
SURVIVAL
MEAN
389E-02
ALIVE
DIED8. -3.160E-02
Fig. 6. A principal component analysis of clinicopathological factors including
the clonogenicity for the survival of the patients with operable breast cancer.
Table 4 Clonogenic growth in soft agar and recurrence after mastectomy in
primary breast cancer
No. of
colonies/dish0
(88.9)
29»
(7.3)
51 (92.7)
55(100)
30-99'
15(18.7)
65(81.2)
80(100)
100-299
46 (80.7)
11 (19.3)
57 (100)
(80.8)215(84.6)Total36(100)
21
26(100)254(100)
300TotalRecurrenceYes4(11.1)°4
5(19.2)39(15.4)No32
" Numbers in parentheses, percentage.
'' Including cancers that showed less than 10 colonies in plural dishes.
' Including cancers that showed 10 or more colonies in plural dishes.
5292
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
CLONOGENIC GROWTH AND METASTASIS OF BREAST CANCER
OS in favor of patients without clonogenic growth of tumor
specimen as compared with those with positive clonal growth.
They suggested that the CF is an independent prognostic pa
rameter for pBC.
In the present study, by univariate analysis, there was a
statistically significant difference in RFS but not in OS of pBC
patients between those with positive and negative CF, as shown
in Fig. 1. This relationship was repeatedly conformed in Stage
II cancer patients (Fig. 2). By the principal component analysis,
however, the clonogenic growth of pBC tumors was indicated
to be a significant and independent prognostic factor for OS as
well as RFS (Figs. 5 and 6).
When the patients were analyzed by axillary node status,
there was only a trend of difference in RFS in favor of the
negative clonogenicity group. This may be partly due to the
smaller number of patients included, shorter survival followup, and particularly few patients with relapsed malignancy in
the node-negative group.
There has been a lot of controversy concerning the relation
ship between the presence or absence of ER in primary tumors
and RFS or OS of breast cancer patients after mastectomy (22).
In the present study, there was no correlation between ER and
prognosis of 254 patients. This has been confirmed by our
previous study (23). In the present study, while there was no
difference in RFS between colony-positive and colony-negative
patient groups in patients with positive ER tumors, those with
positive clonogenicity clearly showed lower RFS as compared
with those with negative clonal growth in the ER-negative group
(Fig. 4). Although these relations are still to be validated, the
clonogenic assay system may be useful in selecting negative ER
patients with worse prognosis who would be benefitted more
by the use of more aggressive adjuvant therapy.
We could not find any difference in the incidence of recur
rence of malignancy between the patients whose tumors showed
minimum CF (less than 30 colonies in a single dish or less than
10 colonies in plural dishes) and those without clonogenicity
(Table 4). The preliminary cut-off point of negative and positive
clonogenicity appears to be justified in relation to RFS of the
patients. As shown in Table 4, rather unexpectedly, if the
number of colonies formed was over 30 in a single dish or over
10 colonies in plural dishes, there was no significantly different
relationship between the number of colonies per dish or clonal
efficiency of cancers (data not shown) and the recurrence rate
of patients after mastectomy. These results coincide with those
of Nicolson et al. (13) in that, if significant clonal growth was
obtained, the metastatic potential of primary rat mammary
adenocarcinoma does not relate to the clonal efficiency. These
findings also justify the present cut-off level in the discrimina
tion of CF.
There still remain many technical problems in the human
tumor clonogenic assay system, particularly in the low clono
genicity as indicated by various investigators ( 1, 24).
Breast cancers lacking in vitro clonogenic growth may be
divided into two groups. One has the natural ability to grow in
the soft agar system but does not form colonies because of some
technical difficulties during processing. The other group origi
nally has no viability to grow in the system. This latter cancer
group may be less malignant with less potential to relapse after
mastectomy. Therefore, by lessening the former possibility with
increasing clonogenicity, the relationship between the in vitro
clonogenicity and prognosis of operable breast cancer patients
will be more clearly elucidated.
ACKNOWLEDGMENTS
The authors are indebted to T. Johma and K. Shinozuka for their
excellent technical assistance. Also thanks are due to J. B. Kawabe for
correction of the manuscript.
REFERENCES
1. Hamburger, A. W., and Salmon, S. E. Primary bioassay of human tumour
stem cells. Science (Wash. DC), 197: 846-854, 1977.
2. Courtenay, V. D., and Mills, J. An in vitro colony assay for human tumours
grown in immune-suppressed mice and treated in vivo with cytotoxic agents.
Br. J. Cancer, 37: 261-268, 1978.
3. Bertoncello, I., Bradley, T. R., Campbell, J. J., Day, A. J., McDonald, I. A.,
McLeish, G. R., Quinn, M. A., Rome, R., and Hodgson, G. S. Limitations
of the clonal agar assay for the assessment of primary human ovarian tumour
biopsies. Br. J. Cancer, 45: 803-811, 1982.
4. Mattox, D. E., and Von Hoff, D. D. In vitro stem cell assay in head and neck
squamous carcinoma. Am. J. Surg., 140: 527-530, 1980.
5. Johns. M. E., and Mills, S. E. Cloning efficiency. A possible prognostic
indicator in squamous cell carcinoma of the head and neck. Cancer (Phila.),
52: 1401-1404, 1983.
6. Durie, B. G. M., Young, L. A., and Salmon, S. E. Human myeloma in vitro
colony growth: interrelationships between drug sensitivity, cell kinetics, and
patient survival duration. Blood, 61:929-934, 1983.
7. Sutherland, C. M., Mather, F. J., Carter, R. D., Cerise, E. J., and Krementz,
E. T. Breast cancer as analyzed by the human tumor stem cell assay. Surgery,
94:370-375, 1983.
8. Aapro, M. S., Eliason, J. F., Krauer, F., and Alberto, P. Colony formation
in vitro as a prognostic indicator for primary breast cancer. J. Clin. Oncol.,
5:890-896, 1987.
9. Touzet, C. L., Ruse, F., Chassagne, J.. Ferriere. J. P.. Chollet, P., Plagne,
R., Fonck, Y.. and De Latour, M. In vitro cloning of human breast tumour
stem cells: influence of histológica! grade on the success of cultures. Br. J.
Cancer, 46:668-669, 1982.
10. Hug, V., Rashid, R., Blumenschein, G., and Spitzer, G. Clonogenic in vitro
growth and histologie grading of primary human breast tumors. Int. J. Cell
Cloning, 3: 143-148, 1985.
11. Moezzi, J., Ali-Osman, F., Nicholson, G. L., Ungerleider, J. S., and Murphy,
M. J., Jr. Relationship between histopathology and in vitro clonogenicity in
breast cancer. Breast Cancer Res. Treat., S: 147-156, 1986.
12. Flentje, D., Feichter, G., Fletje, M., Krämer,K. L., Goerttler, K., and Schlag,
P. Does in vitro colony formation and chemosensitivity relate to DNA ploidy
and S-phase fractions? J. Cancer Res. Clin. Oncol., 113: 87-90, 1987.
13. Nicolson, G. L., Lembo, T. M., and Welch, D. R. Growth of rat mammary
adenocarcinoma cells in semisolid clonogenic medium not correlated with
spontaneous metastatic behavior: heterogeneity in the metastatic, antigenic,
enzymatic, and drug sensitivity properties of cells from different sized colo
nies. Cancer Res., 48: 399-404, 1988.
14. Price, J. E. Clonogenicity and experimental metastatic potential of sponta
neous mouse mammary neoplasms. J. Nati. Cancer Inst., 77: 529-535, 1986.
15. Nomura, Y., Matsui, K., Tashiro, II.. Hamada, Y.. and Saeki. K. Response
to steroid hormones of a human breast cancer cell line (MCF-7) revealed by
the colony formation method. Jpn. J. Cancer Chemother., 10: 1335-1341,
1983 (in Japanese with English summary).
16. Nomura, Y., Tashiro, II., and Shinozuka, K. Changes of steroid hormone
receptor content by chemotherapy and/or endocrine therapy in advanced
breast cancer. Cancer (Phila.), SS: 546-551, 1985.
17. Maestroni, G. J. M., and Losa, G. A. Effect of autologous serum and steroid
hormone receptors on direct cloning of human breast cancer. Arch. Pathol.
Lab. Med., 108: 783-785, 1984.
18. Sandbach, J., Von Hoff, D. D., Clark, G., Cruz, A. B.. Jr., Obrien. M., and
The South Central Texas Human Tumor Cloning Group. Direct cloning of
human breast cancer in soft agar culture. Cancer (Phila.), 50: 1315-1321,
1982.
19. Dittrich, C., Jakesz, R., Wrba, F., Havelec, L.. Haas. O., Spona, J.. Holzner,
H., Kolb, R., and Moser, K. The human tumour cloning assay in the
management of breast cancer patients. Br. J. Cancer, 52: 197-203, 1985.
20. Tveit, K. M., Pettersen, E. O., Fossa, S. D., and Pihl, A. Selection of tumour
cell subpopulations occurs during cultivation of human tumours in soft agar.
A DNA flow cytometric study. Br. J. Cancer. 52: 701-705, 1985.
21. Meyer, J. S., McDivitt, R. W., Stone, K. R., Préy,M. U., and Bauer, W. C.
Practical breast carcinoma cell kinetics: review and update. Breast Cancer
Res. Treat., 4: 79-88, 1984.
22. Raemaekers, J. M. M., Beex, L. V. A. M., Koenders, A. J. M., Pieters, G.
F. F. M., Smals, A. G. H., Benraad, T. J., Kloppenborg, P. W. C., and The
Breast Cancer Group. Disease-free interval and estrogen receptor activity in
tumor tissue of patients with primary breast cancer: analysis after long term
follow-up. Breast Cancer Res. Treat., 6: 123-130, 1985.
23. Nomura, Y., Tashiro, H., and Hisamatsu, K. Clinical values of estrogen
receptor (ER) as a prognostic factor in breast cancer. J. Jpn. Soc. Cancer
Ther., 23: 844-858, 1988 (in Japanese with English summary).
24. Weisenthal, L. M., and Lippman, M. E. Clonogenic and nonclonogenic in
vitro chemosensitivity assays. Cancer Treat. Rep.. 69: 615-632, 1985.
5293
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1989 American Association for Cancer Research.
In Vitro Clonogenic Growth and Metastatic Potential of Human
Operable Breast Cancer
Yasuo Nomura, Hideya Tashiro and Kazufumi Hisamatsu
Cancer Res 1989;49:5288-5293.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/49/19/5288
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 18, 2017. © 1989 American Association for Cancer Research.