(CANCER RESEARCH 51, 3807-3813. July 15, 1991]
Centromeric Copy Number of Chromosome 7 Is Strongly Correlated with Tumor
Grade and Labeling Index in Human Bladder Cancer1
Frederic M. Waldman,2 Peter R. Carroll, Russell Kerschmann, Michael B. Cohen,3 Frederick G. Field,
and Brian H. Mayall
Departments of Laboratory Medicine [F. M. W., F. G. F., B. H. M.], Pathology [R. K., M. B. C.¡,and Urology ¡P.R. C.J, University of California San Francisco,
San Francisco, California 94143-0808
ABSTRACT
The relationship between interphase cytogenetics and tumor grade,
stage, and proliferative activity was investigated in 27 transitional cell
carcinomas of the urinary bladder. Using fluorescence in situ hybridiza
tion with chromosome-specific DNA probes, the copy number of pericentromeric sequences on chromosomes 7, 9, and 11 was detected within
interphase nuclei in touch preparations from tumor biopsies. Monosomy
of chromosome 9 was detected in 9 of 22 cases (41%), while tetrasomy
for chromosomes 7 and 11 was detected in 10 of 26 (38%) and 6 of 23
(26%) cases, respectively. Copy number of chromosome 7 was the most
highly correlated with increasing tumor grade (r = 0.616, P < 0.001,
Spearman rank correlation) or increasing pathological stage (r = 0.356,
P < 0.002). Copy number for chromosome 9 did not correlate with either
grade or stage (P > 0.05). Tumor labeling index (LI) was determined
after in vitro 5-bromodeoxyuridine incorporation, while proliferating cell
nuclear antigen LI was determined immunohistochemically. Increasing
LI by either method correlated with increasing copy number for all three
chromosomes tested (r2 = 0.473, P < 0.002 for 7; r2 = 0.384, P < 0.01
for 11; and r1 = 0.316, P < 0.05 for 9). Since high tumor grade, stage,
and LI are all indicative of more aggressive tumor behavior and worse
prognosis, these findings suggest that polysomy, especially for chromo
some 7, may be highly predictive for bladder tumor aggressiveness.
INTRODUCTION
Numerous specific genetic aberrations have been implicated
in bladder cancer. Cytogenetic studies have defined numerical
and structural changes involving chromosomes 1, 3, 5, 7,9, 11,
and 17 (1-8). Flow cytometric analysis of DNA content has
shown a correlation between DNA ploidy, S-phase fraction,
and tumor grade (9-14). Analyses of allelic loss rates for sites
on 9q, 11p, and 17p have shown a high fraction of tumors with
such losses (15). In particular, it has been shown (16) that losses
of 9q occur independently of tumor grade, whereas 17p allelic
loss is more specifically associated with higher grade tumors.
While the prognostic significance of these genetic and cytogenetic changes has yet to be demonstrated in prospective studies,
there is increasing evidence that the specific genes involved may
modulate tumor proliferation (17-20).
Tumor labeling index or S-phase fraction has been correlated
with prognosis in bladder tumors (10,12). S-phase fraction can
be measured directly by incorporation of [3H]thymidine or
Received 12/7/90; accepted 5/7/91.
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.
' This work was supported by NIH Grant CA47537, as part of the Bladder
Marker Network of the National Cancer Institute.
2To whom requests for reprints should be addressed, at Laboratory for Cell
Analysis, MCB 232, Dept. of Laboratory Medicine, University of California San
Francisco, San Francisco, CA 94143-0808.
3 Present Address: Department of Pathology, University of Iowa, Iowa City,
IA.
BrdUrd4 into tumor DNA5 or by immunohistochemical staining
of cell cycle-specific antigens such as Ki67 (21) or PCNA (22)
expressed only in proliferating cells. The present study was
undertaken in order to study the relationships in bladder cancer
between genotypic and phenotypic abnormalities. Tumor grade,
tumor stage, and tumor labeling index, as measured by BrdUrd
incorporation or PCNA labeling, were used to characterize
tumor phenotype, while tumor genotype was defined by cytogenetic abnormalities, as detected by FISH with DNA probes
specific for chromosomes 7, 9, and 11.
MATERIALS AND METHODS
Clinical Material. All tumor samples were obtained fresh from the
Medical Center and affiliated hospitals. University of California, San
Francisco. All patient protocols were reviewed by the Institutional
Review Board. Tumors were analyzed for pathological stage (Ta, pap
illary noninvasive; Tl, invading into lamina propria; T2, superficial
invasion of muscularis; T3, deep invasion of muscularis; T4, metastatic)
and pathological grade (23), without knowledge of interphase cytoge
netics or labeling index results.
Fresh unfixed tumor samples were collected during cystoscopy and
stored in Hanks' balanced salt solution prior to sample processing.
Adequate biopsy material was always first collected for histopathological diagnosis. Touch preparations were generated by gently touching
the urothelial surface of the biopsy to a dry microscope slide. This brief
contact allowed an adequate number of single tumor cells to adhere to
the slide surface. Five to 10 such touch preparations were made from
each biopsy specimen. Slides were then air dried at room temperature
overnight and stored under nitrogen at -20°C. Whenever possible, 5mm x 5-mm x 1-mm slices were also cut from cystectomy specimens,
and touch preparations were made as described.
After touch preparations were made, tumor slices were incubated
with BrdUrd (100 ¿iM
BrdUrd, 10 nM fluorodeoxyuridine, hyperbaric
oxygen in RPMI) for 2 h at 37°C,to label cells undergoing DNA
synthesis. Tissue was then fixed in 70% ethanol, embedded in paraffin,
and sectioned for immunocytochemical staining using monoclonal an
tibodies to BrdUrd (IU4, 1:10,000; Caltag) and cyclin/PCNA (19A2,
1:200; Boehringer Mannheim). Standard indirect immunoperoxidase
staining was used (Vectastain Elite ABC kit; Vector Laboratories). The
proportion of tumor cells staining positively for BrdUrd and PCNA
was determined from 2000 cell counts, by selecting 4-6 high power
(x40) fields showing the greatest concentration of stained nuclei.
Controls used for in situ hybridization were peripheral blood lym
phocytes from healthy human volunteers. Lymphocytes in metaphase
were obtained after short term culture by standard procedures. Control
slides were stored under nitrogen at -20°Cuntil use.
Flow Cytometry. When adequate material was available, single-cell
suspensions were generated by finely mincing tissue, and cells were
prepared according to the method of Vindelov et al. (24). Alternatively,
a vigorous bladder wash specimen taken at the time of biopsy was used
for flow analysis. The FACScan flow cytometer (Becton-Dickinson,
4 The abbreviations used are: BrdUrd, bromodeoxyuridine; PCNA, proliferat
ing cell nuclear antigen: FISH, fluorescence in situ hybridization; SSC, standard
saline citrate: LI, labeling index; ADB, antibody-diluting buffer; FITC, fluorescein
isothiocyanate.
5 Waldman, F. M., Chew, K., Ljung, B. M., Goodson, W., Horn, J., Duarte,
I... Smith, H., and Mayall, B. A comparison between bromodeoxyuridine and
[3H]thymidine labeling in human breast tumors. Modern Pathol., in press, 1991.
3807
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
BLADDER INTERPHASE
San Jose, CA) was used for analysis, with doublet discrimination gating
and CellFit (Becton-Dickinson) software. A clearly defined second peak
comprising more than 20% of total cells was necessary to define
aneuploidy.
Chromosome-specific Probes. The probes used were specific for pericentromeric sequences on chromosome 7 (p7«tet;H. Willard, Stanford
University), chromosome 9 (pHUR98; B. Moyzis, Los Alamos Na
tional Laboratory), and chromosome 11 (pLCl l A; H. Willard). Probes
were labeled by nick translation using either biotin-11-dUTP (Bethesda
Research Laboratories) or digoxigenin-11-dUTP (Boehringer Mann
heim). Approximately 5-35% of thymidine residues were substituted
during the reaction.
In Situ Hybridization. The hybridization was carried out as previously
described (25), with modifications (26). Slides were first fixed for 5 min
each in three changes of freshly prepared Carnoy's solution (3/1 methanol/acetic acid). Hybridizations containing 5-10 ng labeled probe, 2.5
/¿g
carrier DNA, 55% formamide, and 10% dextran sulfate, in 2x SSC,
were incubated overnight at 37°C.Slides were washed 3 times for 10
min each in hybridization wash buffer (2x SSC, 50% formamide for
chromosome 7, 55% formamide for chromosomes 9 and 11) at 45°C
and then 2 times in 2x SSC, first at 45°Cand then at room temperature.
To minimize nonspecific staining, slides were then incubated for 5 min
in ADB (5% Carnation milk powder, 4x SSC, pH 7, 0.02% sodium
azide, 0.1% Triton X-100). Slides on which cells had been hybridized
to biotinylated probes were incubated for 25 min at room temperature
in FITC-avidin (Vector Laboratories) diluted 1:200 in ADB. Slides on
which cells had been hybridized to digoxigenated probes were incubated
for 30 min in FITC-conjugated antidigoxigenin antibody (Boehringer
Mannheim) in ADB. Slides were then rinsed 10 min each in 4x SSC,
4x SSC/0.1% Triton X-100, and 4x SSC. Slides were then rinsed in
distilled water, air dried, and counterstained with either 0.25 Mg/ml
propidium iodide (Sigma) or 0.2 Mg/ml 4',6-diamidino-2-phenylindole
hydrochloride (Molecular Probes) in anti-fade solution (27). If neces
sary, biotinylated probe signals were amplified by preblocking in ADB
for 5 min, incubating 30 min in 1:100 goat antiavidin/ADB. washing
in 4x SSC, blocking again in ADB, and incubating 25 min in FITCavidin diluted 1:200 in ADB.
Scoring of Interphase Nuclei. A touch preparation from each biopsy
was stained with Giemsa and analyzed to determine the proportion of
malignant cells present. Those cases with <50% tumor cells in the
touch preparations (by conservative cytological estimate) were censored
from the study. Additionally, six cases showed no tumor in hematoxylin- and eosin-stained sections of the biopsy and were excluded. When
the majority of cells were tumor cells by cytological criteria, benign
fibroblasts and inflammatory cells were readily recognized and were
excluded from further cytogenetic analysis.
After hybridization, slides were scored for the number of hybridiza
tion signals in each nucleus, using an epifluorescence microscope
equipped with a 63x na 1.3 oil immersion objective. If possible, 200500 nuclei were scored for each hybridization. Nuclei in which the
nuclear boundary was broken or torn were considered damaged and
were censored from analysis. Similarly, squashed, smeared, clumped,
or overlapped nuclei were ignored. Each hybridization was accompanied
by a control hybridization using normal lymphocytes. If the control
displayed increased monosomy (>10%), parallel tumor preparations
displaying weak signals were repeated, because the likelihood of missing
signals during the counting procedure was high.
The copy number category (monosomic, disomic, trisomie, tetrasomic, pentasomic, and above) reflects the major (>50%) or dominant
(20-50%) population present. If no aneusomic (i.e., other than two
copies/nucleus) subpopulation was present at >50% of the total, then
the largest aneusomic subpopulation having >20% of total cells was
used. Thus, a tumor was considered disomic if no other subpopulations
had >20% of total cells.
RESULTS
CYTOGENETICS
age, sex, labeling indices, and interphase cytogenetics for the
three chromosomes tested. The worst pattern present for nu
clear grade and pathological stage was recorded for each case.
The mean age was 68 years, and 90% of the tumors were from
men. Four of 27 patients were treated with various intravesical
chemotherapeutic reagents prior to biopsy.
Interphase Cytogenetics. The copy number distributions of
chromosomes 7, 9, and 11 in control interphase lymphocytes
are shown in Fig. 1. The sensitivity of the reaction was apparent
from the high proportion of nuclei having two signals. Because
only 3.3, 4.2, and 5.0% of control nuclei had one target copy
Table 1 Summary of patient data and interphase cytogenetics for 27 bladder
tumors
num
ber17111222222224444543
LISex"
LI(%)0.73.61.24.86.35.430.915.73.746.922.71
(%)FMMMMMMMMFMMMMMMMMMMMMMMMMM1.33.60
Case123456789101112131415161718192021222324252627GradeIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
(yr)537885455659778158707660647170677986808764556363636767BrdUr
T1.8
T17.321.920.138.948.126.521.91.8
TCopy
" M, male; F. female.
* B, bladder wash; T. disaggregated tissue.
' Centramene copy number of major or dominant population.
d Previously treated with intravesical chemotherapy.
100 H
0)
73
60
3
40-
m
20
01234
Centromeric
Copy
Fig. 1. Centromeric copy numbers within control
Clinical. Twenty-seven cases of transitional cell carcinoma of
for chromosomes 7 (B, n = 7), 9 (ü,n = 3), and
nuclei were scored for each hybridization.
the bladder were analyzed. Table 1 shows tumor grade, stage,
3808
Number
interphase lymphocyte nuclei
11 (D, n —4). Two hundred
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
BLADDER INTERPHASE CYTOGENETICS
for chromosomes 7, 9, and 11, respectively, it was possible to
define, with high confidence, true tumor monosomy when the
proportion was above 20%. Conversely, the specificity of hy
bridization reactions was demonstrated in metaphase chromo
somes, in which only two specific targets were present.
Photomicrographs of four different hybridizations are shown
in Fig. 2. Case 20 (Fig. 2A) was easily characterized, showing
three small well defined signals in each nucleus with the probe
for chromosome 7. Case 25 (Fig. 2B) also is the result of
hybridization with probe specific for chromosome 7. Most
nuclei showed four signals, although in this case the signals
were larger than in case 20 and slightly spread. Two smaller
nuclei having only two signals were also present. Some of the
signals were split into couplets, which were scored as single
signals. Fig. 2C shows nuclei from case 11 after hybridization
to chromosome 9 probe, showing one copy in each cell. Cytoplasmic shadows could be seen staining nonspecifically with
propidium iodide. Nonspecific hybridization to nuclear DNA
was also apparent as much smaller and fainter fluorescein
signals. Monosomy for chromosome 11 in case 5 is shown in
Fig. 2D. A minor population of nuclei having two signals was
also present. In this photomicrograph, cytoplasmic fluorescence
appears green.
A summary of the interphase FISH results for two repre
sentative tumors is shown in Fig. 3. In case 11, two patterns of
copy number distribution were seen. Chromosome 9 was pres
ent predominantly as one copy/nucleus. Chromosomes 7 and
11 were detected predominantly as two copies/nucleus. This
tumor was scored as monosomic for 9 and disomic for 7 and
11. Note that there were small populations of nuclei which had
Fig. 2. Photomicrographs of four different tumors after fluorescence in situ hybridization. Probes were biotinylated and then stained with FITC-avidin (green).
Nuclei were counterstained with propidium iodide (red) or 4'.6-diamidino-2-phenylindole
hydrochloride (blue). Green cytoplasmic fluorescence is the result of
autofluorescence and nonspecific staining. Multiple exposures were used to combine colors. Original magnification, x63. /. case 20, showing trisomy 7. B, case 25,
showing tetrasomy 7. C, case 11, showing monosomy 9. D. case 5, showing monosomy 11.
3809
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
BLADDER INTERPHASE CYTOGENETICS
Case 11
tumors had a negative imbalance for chromosome 7.
Chromosome 9 was probed in 22 cases. Nine of the 22 cases
(41%) were monosomic for chromosome 9. Five of these monosomic cases (cases 9-11, 18, and 22) and one disomic case
(case 24) had a negative imbalance and two cases (cases 3 and
8) had a positive imbalance for chromosome 9.
Chromosome 11 was probed in 23 tumors. Eight cases (35%)
were disomic and six (26%) tetrasomic. There was a positive
imbalance in one disomic (case 2) and one tetrasomic (case 19)
case and a negative imbalance in one disomic (case 27) and one
trisomie (case 21) case.
100 n
80
«
u
60
40
o
•s
20
O
1
Centromeric
100
2
3
Copy
Table 2 Incidence of centromeric copy number in 27 bladder tumors
Number
Incidence (%)
Case 27
n
I copy
2 copies
3 copies
4 copies
5 copies
Chromosome
7Chromosome
9Chromosome
80
11262223124117311835g2317381826115
60
40-
g
I
20-
o
E
20-
0123456
Centromeric
Copy
Number
10 -
Fig. 3. Actual distributions of Centromeric copy number for two bladder
tumors. Chromosomes probed were 7 (•).9 (0), and 11 (D).
one copy of chromosomes 7 and 11, most likely due either to
overlap causing apparent fusion of two signals into one or to
weaker signals which were missed during scoring. These subpopulations, always <10%, were also present in control prepa
rations. The small populations (~5%) of nuclei having three
copies of chromosome 7 was less likely due to artifact, however.
These trisomy 7 nuclei most likely represented true heteroge
neity of copy number distribution, because nuclei with such
increased copy number are not present in control hybridizations
in the absence of larger populations with four or more signals.
Case 27 in Fig. 3 showed a more heterogeneous distribution
of centromeric copy numbers. Large subpopulations (30-50%)
of nuclei having either two or four copies of chromosomes 7
and 9 were present. Chromosome 11 was present predominantly
at two copies/nucleus. This tumor was scored as tetrasomic for
7 and 9 and disomic for 11.
Table 1 shows the clinical, proliferative, and cytogenetic
information for all of the cases analyzed, ordered with increas
ing pathological grade. Although sufficient tissue for flow cytometric analysis was available for only nine cases, a good
correlation was seen between DNA index and total centromeric
copy numbers. In case 9, there was a small (10% of total cells)
peak seen by flow cytometry with a DNA index of 1.8. Inter
estingly, there was a similar minor population of nuclei ana
lyzed by FISH with four copies of chromosome 7 (20%) and
11 (16%) and <4% tetrasomy for chromosome 9 (data not
shown).
The copy number distributions, for the 27 tumors, for each
of the three chromosomes tested are summarized as percentages
in Table 2. Chromosome 7 was probed in 26 cases. Ten of these
(38%) showed tetrasomy for chromosome 7, while eight cases
(31 %) showed disomy. Note from Table 1 that five cases (cases
4, 5, 14, 16, and 24) showed a positive imbalance of chromo
some 7 (i.e., 9 and 11 both had fewer copies than did 7), but no
l
Centromeric
Copy Number
2345
Centromeric
Copy Number
30 -,
20 -
10 -
30 -i
20 -
l
Cenîromeric Copy
Number
Fig. 4. Relationship between tumor grade and centromeric copy number for
chromosomes 7, 9, and 11. B, grade 1; O. grade II; D, grade III.
3810
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
BLADDER INTERPHASE CYTOGENETICS
as strongly as grade (Table 1 and Fig. 5). Stage Ta tumors were
more likely to be monosomic or disomic, and stage T2-T3
tumors were more likely to be polysomic. The correlation
between stage and chromosome 7 copy number (Spearman rank
correlation, r2 = 0.356, P < 0.002) was stronger than that
between stage and chromosome 11 copy number (r2 = 0.284, P
< 0.01); there was no significant correlation between chromo
some 9 copy number and tumor stage (r2 = 0.025, P > 0.25).
so n
I
20
10-
12345
Centromeric
Copy
Number
0.001). BrdUrd LI was compared to the chromosome copy
number, as determined by FISH. Fig. 6 shows the strong
correlation which existed for chromosomes 7 (Spearman rank
correlation, r2 = 0.473, P < 0.002), 9 (r2 = 0.316, P < 0.05),
30 -¡
o
BrdUrd Labeling Index. Tumors were labeled in vitro with
BrdUrd, to detect the fraction of cells in the S (synthesis)-phase
of the cell cycle (the tumor labeling index). The LI showed a
strong correlation with pathological grade (Spearman rank
correlation, r2 = 0.814, P < 0.001) and stage (r2 = 0.537, P <
20
-10
40 -LUI'•-1
30H
1iPm
20 -
12345
Centromeric
Copy Number
m
10 -
12345
30 -,
o
Centromeric
Copy
Number
Copy
Number
Copy
Number
20 -
a
t'o
40 10 -
CID
30H
12345
Centromeric
Copy Number
Fig. 5. Relationship between tumor stage and Centromeric copy number for
chromosomes 7, 9. and 11. B. stage Ta; m, TI; D, T2; B, T3.
T>
m
10 -
Centromeric
The copy numbers for chromosomes 7, 9, and 11 were
generally correlated with each other. The correlation between
chromosomes 7 and 11 was the strongest (r2 = 0.619, P =
0.0001, simple regression), followed by 7 and 9 (r2 = 0.209, P
= 0.037) and 9 and 11 (r2 = 0.127, P = 0.123).
Tumor Grade and Interphase Cytogenetics. Tumor grade was
strongly associated with chromosome copy number for chro
mosomes 7 and 11 (Spearman rank correlation, r2 = 0.616, P
< 0.001 and r2 = 0.524, P < 0.001, respectively) and was not
quite significant for chromosome 9 (r2 = 0.140, P < 0.10). As
C
30-
20 -
shown in Fig. 4, all grade I tumors were either monosomic or
disomic for chromosome 7, while all grade III tumors were
polysomic. Grade II tumors were more evenly distributed.
These relationships were also apparent for chromosome 11 but
012345
less strongly so for chromosome 9.
Centromeric
Tumor Stage and Interphase Cytogenetics. Tumor stage was
Fig. 6. Relationship between BrdUrd
also correlated with chromosome copy number, although not
number for chromosomes 7, 9, and 11.
3811
labeling index and Centromeric copy
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
BLADDER INTERPHASE
50 —
40 -
-j
30-
Z
O
o.
20 -
DISCUSSION
(Z)
Numerical cytogenetic aberrations in 27 carcinomas of the
urinary bladder have been characterized in interphase cells,
using fluorescence in situ hybridization with DNA probes spe
cific for pericentromeric regions on chromosomes 7, 9, and 11.
These tumors showed a high incidence of chromosomal aneusomy; only one of the 27 tumors analyzed was disomic for all
three chromosomes tested. There was a high correlation seen
between the BrdUrd and PCNA labeling indices and tumor
grade, tumor stage, and chromosomal pericentromeric copy
number. This correlation was strongest for chromosome 7 copy
number and weakest for chromosome 9 copy number.
Chromosomal copy number, as defined by FISH with peri
centromeric probes, is only an estimate of the copy number of
sequences carried on that chromosome. It should be emphasized
that each nuclear signal which was counted represented the
hybridization target only. Other regions on these chromosomes
coding for other gene sequences may have been present at
greater or fewer copy number. It should be noted that centro
meric copy number determined by FISH does not increase with
DNA replication. G2 cells show only two hybridization spots
for each centromeric signal, because sister chromâtids have not
yet separated. This is evident from the virtual absence of nuclei
with four signals in preparations of proliferating lymphocytes.
Analyses of primary bladder cancers using banded cytogenetics have shown numerous nonrandom aberrations, with increas
ing numbers of numerical and structural aberrations present in
more advanced disease. Both monosomy 9 and trisomy 7 have
been proposed as primary (initiating) changes in bladder cancer,
since they are sometimes the sole cytogenetic lesion detected
(4, 8). Structural alterations of chromosome 11 have also been
noted, particularly deletions on lip (1, 29). Although flow
cytometric analysis of DNA content of tumor cells was only
possible in a fraction of cases in this study, the strong correla
tion between FISH and flow analysis seen in other studies (30,
31) was substantiated. Flow cytometric analysis does not dis
criminate between chromosomes, however. Aneusomies, as de
tectable by FISH, which affect just one or a few chromosomes,
are frequently missed in tumors with a diploid DNA index.6
10 -
1
2
Centromeric
Copy
Number
so —
«O
—
30 20 -
U
CL
10-
1
2
3
Centromeric
Copy
Number
Centromeric
Copy
Number
so —
40-1
-j
30-
CYTOCENETICS
20 U
a.
Fig. 7. Relationship between PCNA labeling index and centromeric copy
number for chromosomes 7, 9, and 11.
and 11 (r = 0.384, P < 0.01). For chromosome 7, all monosomic and disomic tumors had a BrdUrd LI less than 5.1%,
while only one tumor with greater copy number had a labeling
index of <10% (this was a stage Ta tumor and had a LI of
4.8%). A similar relationship was present for chromosome 11
but was not present for chromosome 9.
PCNA Labeling Index. PCNA is an auxiliary protein to DNA
polymerase 6 (28). Like BrdUrd, it can be detected immunohistochemically and is a reflection of the fraction of cycling tumor
cells. The PCNA LI showed a good correlation with patholog
ical grade (Spearman rank correlation, r2 = 0.444, P < 0.005)
and stage (r2 = 0.444, P < 0.005). PCNA Lis were compared
to chromosome copy number as determined by FISH. Fig. 7
shows the distribution of PCNA LI with chromosome copy
number for chromosomes 7 (Spearman rank correlation, r2 =
0.519, P < 0.001 ), 9 (r2 = 0.040, P > 0.50), and 11 (r2 = 0.569,
The correlation between chromosomal copy number and
tumor proliferation or grade may be direct and specific or
indirect and nonspecific. Results reported here may specify a
direct connection between chromosome 7 copy number and
tumor behavior. There may be genes on chromosome 7 (epider
mal growth factor receptor, for example) which, when present
in extra copy number, lead to increased tumor proliferation.
Alternatively, the strong correlation between chromosome 7
copy number and tumor proliferation may be a general phenom
enon, in which genetic instability (and chromosomal aneusomy)
leads to any of a number of specific gene mutations or deletions
on other chromosomes, and this, in turn, may cause increased
tumor proliferation. The tighter correlation between chromo
some 7 copy number and tumor proliferation, as compared to
chromosomes 11 and 9, might be due to an increased likelihood
for random loss of 9 or 11, rather than an amplification of
specific gene(s) on chromosome 7.
A high incidence of al Mie loss in chromosomes 9, 11, and
17 (and low levels of loss in the other chromosomes tested) has
been reported in bladder cancer (15, 16, 33). The pattern of
allelic loss, when detected, is usually consistent with >90% of
tumor cells having that loss (32), strongly suggesting that a
' Submitted for publication.
P < 0.002).
3812
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
BLADDER INTERPHASE CYTOGENETICS
genetic change associated with allelic loss will lead to an in
creased proliferative advantage for that tumor clone. Thus,
finding a high S-phase fraction in tumors carrying specific
genetic aberrations is not surprising.
A strong correlation was seen in this study between tumor
grade and both BrdUrd and PCNA labeling indices. This as
sociation was also seen in a larger series of bladder tumors. The
correlation between LI and grade is consistent with tumor grade
being a descriptor of nuclear activity. Proliferating cells gener
ally have larger, less condensed nuclei and more mitoses than
nonproliferating cells, characteristics used to define higher tu
mor grade. Pathological stage is less clearly a marker of prolif
eration than of tumor invasiveness and is less strongly associ
ated with either labeling index or chromosome copy number in
our study.
In summary, the results of this study demonstrate a strong
correlation in human bladder cancer between genotypic abnor
malities and tumor phenotype. Changes in centromeric copy
number reflect genotypic modification. An increase in copy
number, particularly of chromosome 7, is associated with a less
favorable phenotype, including high nuclear grade, high path
ological stage, and high BrdUrd and PCNA labeling indices.
Conversely, a disomic or monosomic genotype, particularly for
chromosome 7, is associated with a more favorable phenotype.
The causal mechanism for this relationship is unclear, although
it may possibly involve expression or suppression of specific
genes on chromosome 7. Prospective follow-up of these patients
will establish the relationship between the observed genotype
and phenotype and actual clinical outcome.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
Note Added in Proof
I [opinan el al. (Hopman, A. H. N., Moesker, O., Smeets, W. G. B., Pauwels,
R. P. E., Vooijs, G. P., and Ramaekers, C. S. Numerical chromosome 1, 7, 9,
and 11 aberrations in bladder cancer detected by in situ hybridization. Cancer
Res., 5/: 644-651, 1991) have recently reported that FISH analysis of 40 bladder
tumors showed prevalence of numerical aberrations for chromosomes 7, 9, and
11 similar to those reported in this paper.
REFERENCES
22.
23.
24.
25.
1. Atkin. N. B., and Baker, M. C. Cytogenetic study often carcinomas of the
bladder: involvement of chromosomes 1 and 11. Cancer Genet. Cytogenet.,
75:253-268, 1985.
2. Babu, V. R., Lutz, M. D., Miles, B. J., Farah, R. N., Weiss, L., and Van, D.
D. Tumor behavior in transitional cell carcinoma of the bladder in relation
to chromosomal markers and histopathology. Cancer Res., 47: 6800-6805,
1987.
3. Berger, C. S., Sandberg, A. A., Todd, I. A., Pennington, R. D. Haddad, F.
S., Hecht, B. K., and Hecht, F. Chromosomes in kidney, ureter, and bladder
cancer. Cancer Genet. Cytogenet., 23: 1-24, 1986.
4. Gibas, Z., Prout, G. R., Pontes, J. E., Connolly. J. G., and Sandberg, A. A
possible specific chromosome change in transitional cell carcinoma of the
bladder. Cancer Genet. Cytogenet., 19: 229-238, 1986.
5. Pauwels, R. P., Smeets, A. W., Schapers, R. F., Geraedts, J. P., and Debruyne,
F. M. Grading in superficial bladder cancer. 2. Cytogenetic classification. Br.
J. Urol., 61: 135-139, 1988.
6. Sandberg, A. A. Chromosome changes in bladder cancer: clinical and other
correlations. Cancer Genet. Cytogenet., 19: 163-175, 1986.
7. Smeets, W., Pauwels, R., Laarakkers, L., Debruyne, F., and Geraedts, J.
Chromosomal analysis of bladder cancer. III. Nonrandom alterations. Cancer
Genet. Cytogenet., 29: 29-41, 1987.
8. Vanni, R., Scarpa. R. M., Nieddu. M.. and Usai, E. Cytogenetic investigation
on 30 bladder carcinomas. Cancer Genet. Cytogenet., 30: 35-42, 1988.
9. Fossa, S. D., Thorud, E., Pettersen, E. O., Shoaib, M. C., Scott-Knudsen,
26.
27.
28.
29.
30.
31.
32.
O., and Ous, S. DNA flow cytometry in human bladder carcinoma. Pathol.
Res. Pract., 181: 291-295, 1986.
Helpap, B., Schwabe, H. W., and Adolphs. H. D. Proliferative pattern of
urothelial bladder cancer and urothelial atypias. J. Cancer Res. Clin. Oncol.,
109: 46-54, 1985.
Jakobsen, A., Mommsen. S., and Olsen, S. Characterization of ploidy level
in bladder tumors and selected site specimens by flow cytometry. Cytometry,
4: 170-173, 1983.
Kirkhus, B., Clausen, O. P., Fjordvang, H., Helander, K., Iversen. O. H.,
Reitan, J. B., and Vaage, S. Characterization of bladder tumours by multiparameter flow cytometry with special reference to grade II tumours. Apmis,
96: 783-792, 1988.
Smeets, A. W., Pauwels, R. P., Beck, J. L., Geraedts, J. P., Debruyne, F. M.,
Laarakkers. L., Feitz, W. F., Vooijs, G. P., and Ramaekers, F. C. Tissuespecific markers in flow cytometry of urological cancers. III. Comparing
chromosomal and flow cytometric DNA analysis of bladder tumors. Int. J.
Cancer, _?9:304-310, 1987.
Wijkstrom, H.. Granberg-Ohman, I., and Tribukait, B. Chromosomal and
DNA patterns in transitional cell bladder carcinoma. Cancer (Phila.), 5.?:
1718-1723, 1984.
Tsai, Y. C., Nichols, P. W., Hiti, A. L., Williams, Z., Skinner, D. G., and
Jones, P. A. Allelic losses of chromosomes 9, 11, and 17 in human bladder
cancer. Cancer Res., 50: 44-47, 1990.
Olumni, A. F., Tsai, Y. C, Nichols, P. W., Skinner, D. G., Cain, D. R.,
Bender, L. I., and Jones, P. A. Allelic loss of chromosome 17p distinguishes
high grade from low grade transitional cell carcinomas of the bladder. Cancer
Res., 50:7081-7083, 1990.
Cooper, J. A., and Whyte, P. RB and the cell cycle: entrance or exit? Cell,
5*: 1009-1110, 1989.
Fearon, E. R.. Cho, K. R., Nigro. J. M., Kern, S. E., Simons, J. W., Ruppert.
J. M., Hamilton, S. R., Preisinger, A. C., Thomas, G., Kinzler, K. W., et al.
Identification of a chromosome 18q gene that is altered in colorectal cancers.
Science (Washington DC). 247:49-56, 1990.
Liotta, L. A., and Steeg, P. S. Clues to the function of Nm23 and Awd
proteins in development, signal transduction, and tumor metastasis provided
by studies of Dictyostelium discoideum. J. Nati. Cancer Inst., 82:1170-1172,
1990.
Michalovitz, D., Halevy, O., and Oren, M. Conditional inhibition of trans
formation and of cell proliferation by a temperature-sensitive mutant of p53.
Cell, 62:671-680, 1990.
Matsumura, K., Tsuji, T., Shinozaki, F., Sasaki, K., and Takahashi, M.
Immunohistochemical determination of growth fraction in human tumors.
Pathol. Res. Pract., 184:609-613, 1989.
Celis, J. E., and Celis, A. Cell cycle dependent variations in the distribution
of the nuclear protein cyclin proliferating cell nuclear antigen in cultured
cells: subdivision of S phase. Proc. Nati. Acad. Sci. USA, 82: 3262-3266,
1985.
Spiessl, B., Beahrs, O. H., et al., ed. UICC TNM Atlas, Ed. 3, pp. 246-247.
New York: Springer-Verlag, 1989.
Vindelov, L., Christensen, I. J., and Nissen, N. I. A detergent trypsin method
for the preparation of nuclei for flow cytometric DNA analysis. Cytometry,
4:323-327, 1983.
Pinkel, D., Straume, T., and Gray, J. W. Cytogenetic analysis using quanti
tative, high-sensitivity, fluorescence hybridization. Proc. Nati. Acad. Sci.
USA, 83: 2934-2938, 1986.
Balazs, M., Mayall, B., and Waldman, F. Simultaneous analysis of chromo
somal aneusomy and 5-bromodeoxyuridine incorporation in the MCF-7
breast tumor cell line. Cancer Genet. Cytogenet., in press, 1990.
Johnson, G., Davidson, R., McNamee, K., Russell, G., Goodwin, D., and
Holborrow, E. Fading of ¡mmunofluorescence during microscopy: a study of
its phenomenon and its remedy. J. Immunol. Methods, 55: 231-242, 1982.
Bravo, R., Frank, R., Blundell, P. A., and Mactonald-Bravo, H. Cyclin/
PCNA is the auxiliary protein of DNA polymerase alpha. Nature (Lond.),
326: 515-517, 1987.
Vanni, R., Peretti, D., Scarpa, R. M., and Usai, E. Cytogenetics of bladder
cancer: rearrangements of the short arm of chromosome 11. Cancer Detect.
Prev.,/0:401-403,
1987.
Hopman, A. H., Poddighe, P. J., Smeets, A. W., Moesker, O., Beck, J. L.,
Vooijs, G. P., and Ramaekers, F. C. Detection of numerical chromosome
aberrations in bladder cancer by in situ hybridization. Am. J. Pathol., 135:
1105-1117,1989.
Nederlof, P. M., van der Flier, S., Raap, A. K., Tanke, H. J., van der Ploeg,
M., Kornips, F., and Geraedts, J. P. Detection of chromosome aberrations
in interphase tumor nuclei by nonradioactive in situ hybridization. Cancer
Genet. Cytogenet., 42: 87-98, 1989.
Fearon, E. R., Feinberg, A. P., Hamilton. S. H., and Vogelstein, B. Loss of
genes on the short arm of chromosome 11 in bladder cancer. Nature (Lond.),
318: 377-380, 1985.
3813
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 1991 American Association for Cancer Research.
Centromeric Copy Number of Chromosome 7 Is Strongly
Correlated with Tumor Grade and Labeling Index in Human
Bladder Cancer
Frederic M. Waldman, Peter R. Carroll, Russell Kerschmann, et al.
Cancer Res 1991;51:3807-3813.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/51/14/3807
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 15, 2017. © 1991 American Association for Cancer Research.