ANATOMIC PATHOLOGY
Original Article
Tumor Proliferative Fraction in
Solid Malignant Neoplasms
A Comparative Study ofKi-67 Immunostaining
and Flow Cytometric Determinations
AYSEGUL A. SAHIN, M.D., JAE Y. RO, M.D., ADEL K. EL-NAGGAR, M.D.,
PATRICIA L. WILSON, HT, KIM TEAGUE, B.Sc, MARK BLICK, D.O.,
AND ALBERTO G. AYALA, M.D.
Tumor proliferative fraction (TPF) has been shown to correlate
with prognosis in some malignancies. A method for its determination that is practical, accurate, and reproducible is still being
sought. In this comparative study of techniques, TPF values were
determined in mirror-image samples of 126 consecutive solid
malignant neoplasms using flow cytometry and immunostaining
with Ki-67, a monoclonal antibody that recognizes an unknown
nuclear antigen expressed during the entire cell proliferation cycle
but not in resting cells. The mean TPF values for all cases were
19.5 ± 15.6% (percentage of tumor cells stained) by Ki-67 (range,
1-86%) and 15.7 ± 9.6% (S + G2M) by flow cytometry (range,
3-60%), which correlated significantly at r = 0.53 and P = 0.005.
Tumor proliferative fraction (TPF) has been investigated
for decades to better understand tumor growth and metastatic potential and to aid in patient prognosis.1"10 It recently gained clinical significance because its measurement has been used in some types of tumors to inform
treatment selection. Although several methods have been
used to determine TPF in a research setting, a reliable
and practical method for application in a clinical practice
is still being sought."
Mitotic count, long used by surgical pathologists as a
diagnostic and prognostic criterion in malignant tumors,
The correlation was less strong in tumors with low S-phase values
(<10%, r = 0.28) than in tumors with intermediate and high Sphase values (r = 0.66). Ki-67 staining percentages did not correlate with patient age, sex, or tissue origin of the tumor. Ki-67
staining appears comparable to flow cytometry determination of
TPF in solid malignancies with intermediate and high S-phase
values. In tumors with low S-phase values, Ki-67 immunostaining
shows higher TPF values, which perhaps reflect an increase in
the proportion of G,-phase cells or dilutional effect of nonneoplastic cells in the tumors with low proliferative fraction. (Key
words: Proliferative fraction; Ki-67 antibody; Flow cytometry;
Mitotic rate) Am J Clin Pathol 1991; 96:512-519
may be subjective, may not be reproducible, and fixation
of sections can affect the number of mitoses discernible.12"14 Furthermore, the mitotic phase constitutes only
a small part of the cell proliferation cycle. The cells that
are actively synthesizing DNA can be assessed by radiolabeling of the DNA precursors (e.g., by tritiated thymidine labeling). Tumor proliferative fraction as determined
by this method has been shown to correlate with clinical
behavior in certain malignancies1516; however, the method
is laborious and time-consuming. DNA flow cytometry
offers accurate, rapid, and objective information on cell
cycle distribution,17 and TPF thus determined is increasingly used in various neoplasms as an adjunct prognostic
From the Departments of1 Pathology and 2Clinical Immunologyfactor.
and 18~20 But among the drawbacks of the technique are
Biological Therapy, The University of Texas M. D. Anderson Cancer
that it requires expensive equipment, it cannot detect cells
Center, Houston, Texas.
in Gi, and its results are prey to contamination by nonSupported in part by The University Cancer Grant Presented in Part
neoplastic cells.
at the 79th Annual Meeting of United States and Canadian Academy
Recently, immunohistochemical methods using antiof Pathology, March 1990, Boston, Massachusetts.
bodies
to cellular proteins that are involved in cell proReceived November 19, 1990; received revised manuscript and accepted for publication February 5, 1991.
liferation have been introduced to assess TPF. 21 " 24 One
Address reprint requests to: Dr. Sahin, Department of Pathology,
of these antibodies is Ki-67, a mouse monoclonal antibody
M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston,
Texas 77030.
that is commercially available. Ki-67 was first developed
512
513
SAHIN ET AL.
Tumor Prolife live Fraction
by Gerdes and co-workers21 and reacts with an as-yet unknown nuclear protein that has been shown to be present
in all phases of cell proliferation and absent in the resting
phase. Several studies have found TPF values obtained
by Ki-67 staining comparable to those obtained by radionuclide labeling or by flow cytometric determination of
S + G2M values.25"28 In most of these previous studies,
however, the various determinations were made in different portions of a tumor, i.e., without accounting for
intratumoral heterogeneity. Therefore in the present study
we compared TPF values as determined by Ki-67 immunostaining and flow cytometry in mirror-image sections of various solid malignant neoplasms, as well as by
mitotic count in immediately adjacent tissue.
MATERIALS AND METHODS
Specimen Selection
Fresh tumor specimens were obtained from 126 patients who underwent surgical resection for a solid malignant neoplasm at the M. D. Anderson Cancer Center between May 1989 and September 1989. The cases represent
consecutive surgical pathologic specimens with adequate
tumor sample. Tumor samples were processed immediately after resection. In each case, a 1-cm3 piece of tissue
was excised and divided into two equal portions. One half
was embedded in Optimal Cutting Temperature compound (Miles Scientific, Naperville, IL), and the other
half was used to prepare single-cell suspensions for DNA
flow cytometry. Tissue that was immediately adjacent to
the excised portion was fixed in formalin, processed routinely, and stained with hematoxylin and eosin for histologic evaluation and mitotic count.
Ki-67 Staining
For immunostaining, 5-/im frozen sections were cut,
air dried, and fixed in freshly prepared 4% paraformaldehyde for 10 minutes. Immunostaining was performed
by the avidin-biotin-peroxidase complex (ABC) method29
using the Vectastain Elite ABC kit (Vector Labs, Burlingame, CA). Briefly, the sections were washed with phosphate-buffered saline and incubated with normal horse
serum for 20 minutes to reduce nonspecific staining. The
sections then were incubated with Ki-67 antibody (Dakopatts, Santa Barbara, CA) for 60 minutes at a dilution
of 1:80. After washes in phosphate-buffered saline, the
sections were incubated with biotinylated secondary antibody and ABC reagent. The reaction was developed with
diaminobenzidine-hydrogen peroxide substrate. The sections were counterstained with Mayer's hematoxylin, dehydrated through graded alcohols, cleared in xylene, and
mounted with Permount.
A.J.C.P. •
Sections of lymph node with reactive germinal centers
were used as positive controls, and mouse serum applied
instead of primary antibody was used as negative control
staining. All tumor cell nuclei with any discernible staining
were designated Ki-67 positive, regardless of the staining
intensity. Percentage positivity was determined by counting the number of positively stained nuclei among the
total tumor nuclei in 10 representative high-power fields
(HPF, lOx ocular and 40x objective; minimum of 3,000
cells) in the portion of the tumor with highest staining
using a Nikon Labophat microscope (Nikon Instrument
Group, Houston, TX).
Flow Cytometry
For flow cytometric analysis, single-cell suspensions
were prepared by mincing tissue in Roswell Park Memorial Institute medium (Irvine Scientific, Santa Ana,
CA). The cells were filtered after washing with phosphatebuffered saline and the cell concentration was adjusted to
1.0 X 106 cells/mL. Simultaneous DNA/RNA staining
was performed by a two-step method using acridine orange
stain.30
Cellular DNA content was analyzed on a Coulter Profile
flow cytometer (Hialeah, FL) using an argon-ion laser operating at 15 mW. Excitation was achieved at 488 nm. A
535 band pass filter was used in conjunction with photo
multiplier tube 2 (green fluorescence), and a 610 long
pass filter was used with photo multiplier tube 4 (red fluorescence). Cell-cycle compartments were defined and S
+ G2M segments were computed using a baxogram gating
analysis program on the EPICS Profile.31
The proliferative fraction was defined as the percentage
of the total cell population in the S + G2M phase on the
histogram. Ploidy abnormality was defined by DNA index,
i.e., the ratio of the relative G 0 /Gi DNA content of tumor
cells to that of the normal G 0 /Gi cells. Diploidy was defined by a DNA index of 1.00 and aneuploidy by a DNA
index of more than 1.00 or less than 1.00.
Histologic
Evaluation
Mitoses were counted in a total of 10 contiguous HPFs
in the mitotically most active area of the tumor. Tumor
pleomorphism was graded as 1 to 3 on the basis of nuclear
size and shape (grade 1, mild; grade 2, moderate; grade
3, marked pleomorphism).
Correlations
Correlation coefficients (r) comparing Ki-67 immunohistochemistry results, proliferative fraction as determined by flow cytometry, and histologic results (mitotic
count and pleomorphism) were calculated using Spearman's rank test. In addition, Ki-67 staining percentages
ober 1991
514
ANATOMIC PATHOLOGY
Original Article
also were analyzed with respect to the variables of patient
age, sex, and tissue origin of the tumor using the chisquare test.
RESULTS
The tumors were 101 carcinomas (26 lung, 18 breast,
12 colon, 10 ovary, 9 kidney, 6 head and neck, 5 stomach,
3 endometrium, 2 thyroid gland, 2 urinary bladder, 8
miscellaneous), 17 sarcomas, and 8 malignant melano-
mas. The patients were female (66) and male (60), ranging
in age from 14 to 91 years.
Ki-67 Staining
Ki-67 decorated the nuclei of some proportion of tumor
cells in all cases (range, 1-86%; mean, 19.5 ± 15.6%). The
most common staining pattern was diffuse nuclear staining with accentuation of the nucleoli (Fig. 1). Strong Ki-
FIG. 1 (upper left). Diffuse
nuclear staining. Immunoperoxidase stain, hematoxylin counterstain, X250. (Upper right) Nuclear staining
with intensification at the
nucleoli (arrows). Immunoperoxidase stain, hematoxylin counterstain, X250.
FIG. 2. Ductal carcinoma of
breast (lower left), transitional
cell carcinoma of urinary
bladder (lower middle), and
squamous cell carcinoma of
lung (lower right), in which
Ki-67 stained 5%, 20%, and
60% of the tumor cells, respectively. Immunoperoxidase stain, hematoxylin
counterstain, X250.
Vol. 96 • No. 4
515
SAHIN ET AL.
Tumor Proliferative Fraction
67 staining also was observed on chromosomes in mitotic
cells.
Example tumors showing low, intermediate, and high
Ki-67 staining percentages are depicted in Figure 2. Only
a small minority of the tumor samples showed an even
distribution of Ki-67-positive cells; in the remaining cases,
the percentage of cells stained varied from area to area.
In some cases, Ki-67 staining was restricted to certain
architectural components of the tumor. For example, in
cases of keratinizing squamous cell carcinoma, the central
keratinizing portions of the tumor nests consistently did
not stain with Ki-67, but the peripheral, less-differentiated
cells showed immunoreactivity (Fig. 3). Ki-67 staining
percentages did not correlate with patient age or sex or
with tissue origin of the tumor.
Flow Cytometry
The proliferative fraction as determined by flow cytometry ranged from 3-60% for all cases, with a mean
value of 15.7 ± 9.6%. Fifty-seven tumors (45%) were diploid, and 69 (55%) were aneuploid. The coefficient of
variation for the diploid Go/Gi peaks for all cases ranged
from 1.4% to 3.6% (median, 2.2%). The proliferative fractions (and means) for diploid and aneuploid tumors were
4-30% (11.5 ± 6.8%) and 3-60% (19.2 ± 10.3%), respectively (Table 1). Representative DNA histograms with low,
intermediate, and high S + G 2 M values (see definitions
below) are shown in Figure 4.
Correlations Between Ki-67, S-Phase, and Histologic
Results
In Figure 5, for all 126 tumors, Ki-67 staining percentages are plotted against S + G 2 M. Some scattering is
seen, but the relationship between the two variables is
statistically significant (r = 0.53, P = 0.005).
For further analysis, we divided the tumors into those
with low, intermediate, and high proliferative activity according to the proliferative fraction determined by flow
cytometry, whereby Group I tumors had S-phase values
^ 10%; Group II, 11-20%; and Group III, ;> 21%. Figure
6 shows Ki-67 staining results according to this grouping.
There was a stronger correlation in Groups II and III
combined (r = 0.66) than in Group I (r = 0.28), in which
Ki-67 values were higher than the S + G2M values.
There was a trend for increasing mitotic counts and
increasing pleomorphism to be reflected in higher mean
Ki-67 and S + G 2 M percentages (Table 1). The increases
in either subdivision of Ki-67 values were not linear, but
there was a significant difference in Ki-67 immunostaining
according to whether the mitotic rate was less than 2 or
2 or more per 10 HPFs (P = 0.006). The increases in
pleomorphism were linear in relation to mitotic count,
- '!'
•if,.;
-d
"
. ^
•"•• -J-
. •..*;••
•
•
.*
>••.;,
•'
F*.
• V>
"W-
FIG. 3. Ki-67 mainly decorates the cells at the periphery
of tumor nests in this moderately differentiated squamous cell carcinoma. The
centers of the tumor nests
show keratinization and no
immunoreactivity with the
antibody, (left) Hematoxylin
and eosin, X250. (right) Immunoperoxidase stain, hematoxylin counterstain,
X250.
A.J.C.P. • October 1991
•£><v
•
t
\
•
.;
'
:
'•J'
A
• r >ftv
•
"1 e '
•
v
.•„
'- *- .
ap.
-.gli
.•'sss' ^ 1
'
^
' ',
' •*
*
*
•'
'•:H
516
ANATOMIC PATHOLOGY
Original Article
TABLE 1. Ki-67 AND S-PHASE VALUES BY PLOIDY, MITOSIS, AND PLEOMORPHISM
Ki-67 Staining Percentage
No. Of
Cases (%)
All cases
DNA ploidy
Diploid
Aneuploid
Mitotic rate*
<2
2-5
6-10
>10
Pleomorphismt
1
2
3
S-Phase Percentage
Mean
Range
19.5 ± 15.6
1-86
15.7+
9.6
3-60
57 (45)
69 (55)
17.8 ± 18.2
20.9 ± 13.0
1-86
2-73
11.5+ 6.8
19.2+ 10.3
4-30
3-60
16(13)
41 (33)
45 (36)
24(19)
10.8
22.1
20.4
19.0
± 7.7
±22.6
± 11.3
± 9.0
2-18
5-86
1-58
2-40
14.1 + 8.4
12.2+ 6.7
17.1 ± 9.8
20.2 + 12.2
3-22
5-39
3-48
5-60
17(13)
88 (70)
21(17)
11.2± 6.5
20.9 ± 17.5
19.9 ± 8.8
1-20
2-86
8-40
12.7+ 6.8
14.9 + 8.9
21.4+ 12.4
3-25
4-39
9-60
126
Mean
Range
t l . mild: 2 moderate: 3. marked pleomorphism.
• Per 10 high-power field.
and the tumors with very uniform nuclear features (grade
1) tended to show lower Ki-67 staining percentages than
the tumors with moderate (grade 2) and marked pleomorphism (grade 3). However, the increases in Ki-67 values were not linear in relation to increased degrees of
pleomorphism.
DISCUSSION
Previous studies have compared TPF values as determined by Ki-67 immunostaining with the results obtained
by mitotic count, thymidine labeling index, and flow cytometry, and in sum have found positive correlations.
However, the correlations have varied considerably in
strength, and a further complication is that in most of the
studies the TPF measurements were performed in different
portions of a tumor, i.e., the issue of intratumoral heterogeneity was not addressed.25"28'32
Because solid tumors may show significant intratumoral
variation in proliferative activity, our comparison of Ki67 and flow cytometric determination of TPF was between
mirror-image tumor sections. This study design yielded
results that reflected the generally good correlation between the two methods in previous reports. Among the
tumors in our series, there was a stronger immunohistochemistry-flow cytometry correlation in the tumors with
intermediate or high TPF (as defined by S + G 2 M values
greater than 10%) than in the tumors with low TPF. In
the latter group, the TPF values determined by Ki-67 immunostaining were higher than those derived by flow cytometry.
Isola and co-workers33 used Ki-67 immunostaining and
flow cytometry to evaluate TPF in 102 breast carcinomas
and found a significant correlation that was stronger in
tumors with aneuploid DNA content than in tumors with
diploid content. As is generally true, their diploid tumors
had, on average, lower S-phase values than their aneuploid
tumors. Hence, the weaker correlation between the two
methods among the diploid tumors in their series may
have reflected the low S-phase values. Schwartz and
associates34 studied 74 non-Hodgkin's lymphomas and
demonstrated an overall significant correlation between
TPF values as obtained by Ki-67 staining and by flow
cytometric S-phase measurements. In some of their cases,
however, TPF values as obtained by the two methods diverged significantly. Although the authors did not place
their cases in subgroups, it can be seen in their report that
most of the tumors with divergent results had low S-phase
values.
The weakened correlation observed in tumors with low
S + G 2 M phase values in the previous studies and our
series can be attributed to (1) inclusion of Gi phase-cells
by Ki-67 immunostaining and (2) the dilutional effect of
nonneoplastic cells, especially in diploid tumors on the
calculation of S + G 2 M phase by flow cytometry.
Although the exact nature of the nuclear antigen recognized by Ki-67 antibody is not known, detailed cellcycle analysis has shown that the antigen starts to appear
in the nuclei during the early postmitotic phase (G,),
reaches maximum expression during the postsynthetic
(post-S) phase (G2) and mitosis (M), and disappears when
the cells enter the resting phase (G0). Therefore, Ki-67
immunostaining describes the numeric relative ratio of
proliferating cells to resting cells. In DNA histograms, on
the other hand, TPF is defined as the ratio of the cells in
the S + G 2 M compartments to the total number of cells;
because the cells in Gi have DNA content equal to that
of the resting cells, they are not included in the flow cytometric determination of TPF. In theory, then, TPF
measurements by the two respective methods would reflect
the inclusion versus exclusion of tumor cells in G,. Studies
Vol. 96 • No. 4
517
SAHIN ET AL.
Tumor Proliferative Fraction
% KI-67 STAINING
100
c
o
u
N
T
20
30
40
70
SO
% S+G2M PHASE VALUES
2n
DNA
4-n
— i
FIG. 5. Comparison of tumor proliferative fraction as determined by Ki67 immunostaining (percentage of tumor cells stained) and flow cytometry (percentage of S + G2M). The overall correlation is significant (r
= 0.53, P = 0.005).
CONTENT
the d phase is relatively short—and thus proportionally
fewer cells would be in that compartment at any given
time than in the other portions of the proliferation cycle.
In such tumors, the TPF values, as determined by Ki-67
staining (G, and S + G2M) andflowcytometry (S + G2M),
would not be expected to differ significantly.
Discrepancy between flow cytometric and immunohistochemical TPF values also may result from the contamination of nonneoplastic cells (stromal and inflammatory cells). The tissue samples used forflowcytometric
analysis may contain significant number of nonneoplastic
cells leading to underestimation of TPF values because
of the dilutional effect. A chief advantage of immunohistochemical TPF measurement is that it entails simultaneous morphologic evaluation, with separation of neoplastic and nonneoplastic cells.
Artifactual underestimation of S-phase values due to
the "contamination" of nonneoplastic cells is a more
C
O
U
N
T
taunrSfr
1
I
4-n
6n
CONTENT
C
O
U
N
T
COMPARISON OF KI-67 & S+G2M PHASE (MEAN)
% KI-67 STAINING
J ^N*yf\
2n
DNA
4-n
6n
CONTENT
FlG. 4. DNA histograms of representative tumors showing S + G2M
values of 3% (upper), 12% (middle), and 22% (lower), respectively.
s10%
11-20%
> 20%
S*G2M PHASE GROUPS
of cell-cycle distribution have shown that, in terms of duration, the S, G2, and M phases in a given cell type are
fairly constant, whereas the d phase shows marked variation. It is possible that in tumors with high TPF values,
H
S'G2M PHASE
CD KI-67
FIG. 6. Ki-67 immunostaining according to the degree of tumor proliferative fraction byflowcytometry: group 1 (low, S + G2M value £ 10%),
group II (intermediate, 11%-20%), or group III (high, &21%).
A.J.C.P. • October 1991
518
ANATOMIC PATHOLOGY
Original Article
findings with morphology of non-Hodgkin's malignant lymphocommon problem in diploid tumors because of the overmas. J Natl Cancer Inst 1977; 58:499-504.
lapping peaks of tumor cells and nonneoplastic cells. Re8. Braylan RC, Diamond LW, Powell ML, Harty-Golder B. Percentage
cent advances in flow cytometric analysis using keratin
of cells in the S phase of the cell cycle in human lymphoma
determined byflowcytometry Cytometry 1980; 1:171-174.
immunostaining for gating neoplastic cells and accurate
9. Morimura T, Kitz K, Budka H. In situ analysis of cell kinetics in
calculations of cell-cycle compartments may resolve these
human brain tumors. A comparative immunocytochemical study
35
shortcomings.
of S phase cells by a new in vitro bromodeoxyuridine-labeling
technique, and of proliferating pool cells by monoclonal antibody
The relationship between mitotic count and Ki-67
Ki-67. Acta Neuropathol (Berl) 1989; 77:276-282.
staining has been studied by several investigators.26"29 In
10. Nilsson S, Nordgren H, Eklov S, Logadhl M. Expression of Ki-67
our study, the tumors with a low mitotic rate (less than
A proliferation associated antigen in prostatic cancer. Acta Oncol
2 per 10 HPFs) showed low Ki-67 values. However, in
1991;30:177-179.
11. Quinn CM, Wright NA. The clinical assessment of proliferation and
tumors with a higher mitotic rate, Ki-67 immunostaining
growth in human tumours: Evaluation of methods and applicashowed a wide range of percentage positivity. Our results
tions as prognostic variables. J Pathol 1990; 160:93-102.
in tumors with a low mitotic rate corroborate the findings 12. Donhuijsen K, Schmidt U, Hirche H, van Beuningen D, Budach
V. Changes in mitotic rate and cell cycle fractions caused by deof Barnard36 and McGurrin37 and their associates, who
layedfixation.Hum Pathol 1990; 21:709-714.
have reported significant correlations between mitotic rate
13. Ellis PSJ, Whitehead R. Mitosis counting—A need for reappraisal.
and Ki-67-derived TPF values in a series of breast carHum Pathol 1981; 12:3-4.
14. Sadler DW, Coghill SB. Histopathologists, malignancies, and uncinomas, the majority of which had mitotic rates that were
defined high-powerfields(letter). Lancet 1989; i:785-786.
less than 5 per 10 HPFs. In a series of soft-tissue sarcomas, 15. Johnson
HA, Bond VP. A method of labeling tissues with tritiated
38
in general with a high mitotic rate, Kroese and associates
thymidine in vitro and its use in comparing rates of cellular proliferation in duct epithelium, fibroadenoma, and carcinoma of
did not find a correlation with Ki-67 staining, nor did we
human breast. Cancer 1961; 14:639-643.
in tumors with a high mitotic rate.
16. Meyer JS, Coplin MD. Thymidine labeling index,flowcytometric
S-phase measurement, and DNA index in human tumors. ComOther S-phase-specific methods of determining TPF,
parisons and correlations. Am J Clin Pathol 1988; 89:586-595.
e.g., thymidine and bromodeoxyuridine labeling, also
17. Quirke P, Dyson JED. Flow cytometry: Methodology and applihave correlated significantly with Ki-67 staining. In fact,
cations in pathology. J Pathol 1986; 149:79-87.
Deshmukh and co-workers39 concluded in a recent study
18. McDivitt RW, Stone KR, Craig RB, et al. A proposed classification
of breast cancer based on kinetic information: Derived from a
that Ki-67 provided a more reliable measurement of TPF
comparison of risk factors in 168 primary operable breast cancers.
in solid tumors than did thymidine labeling.
Cancer 1986;57:269-276.
Additional research is needed to determine the clinical
19. Hedley DW. Flow cytometry using paraffin-embedded tissue: five
years on. Cytometry 1989; 10:229-241.
significance of TPF values defined by Ki-67 immunoH, Toikkanen S, Klemi PJ. DNA index and S-phase fraction
staining. In tumors with a low S-phase value, however, 20. Joensuu
and their combination as prognostic factors in operable ductal
this approach may provide a more accurate portrayal of
breast carcinoma. Cancer 1990; 66:331-340.
TPF. Furthermore, any clinical usefulness demonstrated 21. Gerdes J, Schwab U, Lemke H, Stein H. Production of a mouse
monoclonal antibody reactive with a human nuclear antigen asfor Ki-67 immunostaining in relation to tumor growth
sociated with cell proliferation. Int J Cancer 1983; 31:13-20.
fraction would be magnified by its relative ease and cost- 22. Garcia RL, Coltrera MD, Gown AM. Analysis of proliferative grade
using anti-PCNA/cyclin monoclonal antibodies infixed,embedeffectiveness.
ded tissues. Comparison withflowcytometric analysis. Am J Pathol 1989; 134:733-739.
Acknowledgment. The authors thank Bobbie L. Coleman for secretarial
23. Tsutsumi Y, Hon S, Onoda N. DNA polymerase a: An immunoassistance.
histochemical marker for proliferating cells in normal and neoplastic human tissues. Am J Clin Pathol 1990; 93:643-650.
REFERENCES
24. Clevenger CV, Epstein AL, Bauer KD. Modulation of the nuclear
antigen p 105 as a function of cell-cycle progression. J Cell Physiol
1. Meyer JS. Growth and cell kinetic measurements in human tumors.
Pathol Annu 1981; 16:53-81.
1987; 130:336-343.
2. Wright NA. Cell proliferation in health and disease. In: Antony PP,
25. Veneroni S, Costa A, Motta R, et al. Comparative analysis of [3H]MacSween RNM, eds. Recent advances in histopathology, Vol.
thymidine labelling index and monoclonal antibody Ki-67 in non12. New York: Churchill Livingstone, 1984, pp 17-33.
Hodgkin's lymphomas. Hematol Oncol 1988; 6:21-28.
3. Tubiana M, Courdi A. Cell proliferation kinetics in human solid
26. Sasaki K, Matsumura K, Tsuji T, Shinozaki F, Takahashi M. Retumors: Relation to probability of metastatic dissemination and
lationship between labeling indices of Ki-67 and BrdUrd in human
long-term survival. Radiother Oncol 1989; 15:1-18.
malignant tumors. Cancer 1988; 62:989-993.
4. Tubiana M, Koscielny S. Cell kinetics, growth rate and the natural
27. Schrape S, Jones DB, Wright DH. A comparison of three methods
history of breast cancer. The Heuson Memorial Lecture. Eur J
for the determination of the growth fraction in non-Hodgkin's
Cancer Clin Oncol 1988; 24:9-14.
lymphoma. Br J Cancer 1987; 55:283-286.
5. Meyer JS, Friedman E, McCrate M, Bauer WC. Prediction of early
28. Kamel OW, Franklin WA, Ringus JC, Meyer JS. Thymidine labeling
course of breast carcinoma by thymidine labeling. Cancer 1983;
index and Ki-67 growth fraction in lesions of the breast. Am J
51:1879-1886.
Pathol 1989; 134:107-113.
6. Meyer JS, Prioleau PG. S-phase fractions of colorectal carcinomas
29. Hsu SM, Raine L, Fanger H. A comparative study of the peroxidaserelated to pathologic and clinical features. Cancer 1981; 48:1221antiperoxidase method and an avidin-biotin complex method for
1228.
studying polypeptide hormones with radioimmunoassay anti7. Silvestrini R, Piazza R, Ricardi A, Rilke F. Correlation of kinetic
bodies. Am J Clin Pathol 1981; 75:734-738.
Vol. 96 • No. 4
SAHIN ET AL.
519
Tumor Proliferative Fraction
30. Traganos F, Darzynkiewicz Z, Sharpless T, Melamed MR. Simultaneous staining of ribonucleic and deoxyribonucleic acids in unfixed cells using acridine orange in aflowcytofluorometric system.
J Histochem Cytochem 1977; 25:46-56.
31. EPICS Profile flow cytometry manual. Coulter electronics Inc, Hialeah, Florida, 1989; part 4:107-148.
32. Swanson SA, Brooks JJ. Proliferation markers Ki-67 and pl05 in
soft tissue lesions: Correlation with DNA flow cytometric characteristics. Am J Pathol 1990; 137:1491-1500.
33. Isola JJ, Helin HJ, Helle MJ, Kallioniemi O. Evaluation of cell proliferation in breast carcinoma. Comparison of K.i-67 immunohistochemical study, DNA flow cytometric analysis, and mitotic
count. Cancer 1990; 65:1180-1184.
34. Schwartz BR, Pinkus G, Bacus S, Toder M, Weinberg DS. Cell proliferation in non-Hodgkin's lymphomas. Digital image analysis
of Ki-67 antibody staining. Am J Pathol 1989; 134:327-336.
35. Zarbo RJ, Visscher DW, Crissman JD. Two-color multiparametric
36.
37.
38.
39.
method for flow cytometric DNA analysis of carcinomas using
staining for cytokeratin and leukocyte-common antigen. Anal
Quant Cytol Histol 1989; 11:391-402.
Barnard NJ, Hall PA, Lemoine NR, Kadar N. Proliferative index
in breast carcinoma determined in situ by Ki67 immunostaining
and its relationship to clinical and pathological variables. J Pathol
1987:152; 287-295.
McGurrin JF, Doria MI Jr, Dawson PJ, et al. Assessment of tumor
cell kinetics by immunohistochemistry in carcinoma of breast.
Cancer 1987;59:1744-1750.
Kroese MCS, Rutgers DH, Wils IS, Van Unnik JAM, Roholl PJM.
The relevance of the DNA index and proliferation rate in the
grading of benign and malignant soft tissue tumors. Cancer 1990;
65:1782-1788.
Deshmukh P, Ramsey L, Garewal HS. Ki-67 labeling index is a
more reliable measure of solid tumor proliferative activity than
tritiated thymidine labeling. Am J Clin Pathol 1990; 94:192-195.
A.J.C.P. • October 1991