Clinical Cancer Research Vol. 6, 4653-4657, December 2000
© 2000 American Association for Cancer Research
Expression of Metastases-associated Genes in Cervical Cancers Resected in the Proliferative and Secretory Phases of the Menstrual Cycle
Silvia Formenti1,
Juan Felix,
Dennis Salonga,
Kathy Danenberg,
Malcolm C. Pike and
Peter Danenberg
New York University School of Medicine, New York, New York 10016 [S. F.], and University of Southern California Keck School of Medicine, Los Angeles, California 90033 [J. F., D. S., K. D., M. C. P., P. D.]
 |
ABSTRACT
|
|---|
Previous retrospective studies suggest that the phase of the menstrual
cycle at surgery (proliferative versus secretory) for
breast cancer may significantly affect patient survival. Fluctuations
during the menstrual cycle of the expression of genes involved in
metastases in breast cancer tissue have also been reported. We
hypothesized that the menstrual phase may also affect similar
changes in gene expression of other cancers. We focused our attention
on cancer of the uterine cervix because the hysterectomy specimen
obtained at original surgery for the cancer can be used retrospectively
to determine cycle phase. We analyzed tumor specimens from 36
premenopausal cervical cancer patients who had undergone hysterectomy
as their primary treatment. We used reverse transcription-PCR to
quantify gene expression during the different phases of the menstrual
cycle as determined from the endometrial specimen. We explored a panel
of genes that may affect metastatic propensity, namely,
metalloproteinase-9 (MMP-9), tissue inhibitor of metalloproteinase-2
(TIMP-2), cyclooxygenase 1 and 2 (COX-1 and COX-2), and vascular
endothelial growth factor (VEGF). A significantly higher level of
TIMP-2 and COX-2 gene expression (P = 0.007 and
0.030, respectively) was detected during the proliferative phase
compared to the secretory phase of the cycle. The expression of the
other genes was not significantly affected by the stage of the
menstrual cycle. The finding that TIMP-2 and COX-2 expression in
cervical cancer may be affected by the stage of the menstrual cycle
supports the hypothesis that ovarian hormones may affect the expression
of genes involved in metastasis. These findings need to be replicated,
and their implications for tumor angiogenesis, invasion, and metastatic
propensity need to be explored both in human studies and in
experimental models.
|
INTRODUCTION
|
|---|
Numerous studies have suggested that survival among premenopausal
breast cancer patients without distant metastases may be affected by
the hormonal milieu at the time of surgery, especially in women found
to have axillary lymph node involvement (1, 2, 3, 4, 5, 6, 7, 8)
. Women
undergoing their surgery during the proliferative phase tend to have
worse survival than those operated on during the secretory phase.
Several biological mechanisms to account for this have been suggested
(9, 10, 11)
.
These include cyclical patterns of immune function, as well as aspects
of cancer cell division and apoptosis, that may be affected by the
hormonal fluctuations of the menstrual cycle. If genes involved in
various aspects of carcinogenesis were hormonally regulated, their
up-regulation at specific times in the menstrual cycle might predispose
seeding cancer cells toward better survival, faster growth rate, or
increased metastatic potential. To our knowledge, two molecular studies
have been published to date on this subject. The first study, reported
in the Lancet by Saad et al. (9)
,
used Northern blot analysis of RNA to measure the expressions of a
number of metastases-related genes, namely, cathepsin-L,
MMP-9,2
MMP-2, TIMP-1, TIMP-2, and TP53, in breast cancer specimens from 27
premenopausal women. They classified the specimens in terms of
menstrual cycle phase by serum levels of estradiol and progesterone on
the day of operation. They found that cathepsin L, MMP-9, and wild-type
p53 expressions were significantly elevated in breast tumors resected
during the phases of the menstrual cycle in which estradiol was 
20
nmol/liter and progesterone was low, i.e., on days in which
a proliferative endometrium should be present. More recently, in a
study of 198 breast cancer patients, Balsari et al.
(11)
identified a statistically significant increase of
HER2neu protein expression in tumors from women operated on during the
follicular phase of the menstrual cycle, suggesting that overexpression
of the HER2 gene may explain the worse outcome for patients who are
operated on during such a phase.
We hypothesized that a similar phenomenon could be occurring in other
tumors, and we selected invasive cervical cancer for study because in
premenopausal patients initially treated by hysterectomy, the phase of
the cycle (proliferative versus secretory versus
menstrual) can be identified by direct microscopic inspection of the
endometrium. We have investigated fluctuations in the expression of two
of the genes studied by Saad et al. (9)
,
namely, MMP-9 and TIMP-2, in premenopausal cervical cancer specimens.
We have also investigated fluctuations in COX-2 and VEGF because of
their emerging central role as regulatory molecules for tumor
angiogenesis, metastatic propensity, and cell adhesion
(12, 13, 14, 15)
and because of the reported fluctuations in COX-2
with estrus in mice and sheep (16
, 17)
. We also
investigated COX-1 as a "control" molecule (18)
.
|
MATERIALS AND METHODS
|
|---|
We identified 38 frozen cervical cancer specimens from
premenopausal patients treated by hysterectomy at Los Angeles
County/University of Southern California Medical Center. All women had
undergone hysterectomy as the primary treatment for stage IB-IIA
cervical cancer. In 36 of 38 cases, the original
pathological specimen of the uterus was available for review to
describe the phase of the menstrual cycle. Their endometrial phase was
classified as proliferative or secretory by an experienced pathologist
(J. F.).
Total mRNA was isolated from 36 cervical cancer specimens. mRNA was
isolated from human biopsy tissues using the guanidinium isothiocyanate
method with the Quickprep Micro mRNA Purification Kit (Amersham
Pharmacia Biotech). Briefly, biopsy tissues were snap-frozen in liquid
nitrogen and pulverized with a mortar and pestle. The pulverized tissue
sample was then homogenized using a guanidinium isothiocyanate
solution. mRNA isolation and purification were carried out according to
instructions provided in the kit. The isolated mRNA was then dissolved
in 60 µl of diethyl pyrocarbonate-treated water. The mRNA was
converted to cDNA, and quantitative RT-PCR (19)
was
performed to determine the gene expressions of MMP-9, TIMP-2, VEGF,
COX-2, and COX-1. RT-PCR was performed using a fluorescence detection
method with the TaqMan (ABI PRISM 7700 Sequence Detection System;
Perkin-Elmer Applied Biosystems). This method uses a dual-labeled
fluorogenic probe that anneals specifically between the forward and
reverse primers. The fluorogenic probes are labeled with a quencher dye
(TAMRA) at the 3' end and a reporter dye (6FAM) at the 5'
end. Laser stimulation within the capped wells containing the reaction
mixture causes emission of the quencher dye until the probe is cleaved
by the 5'
3' exonuclease activity of DNA polymerase during the
extension phase of PCR. An amplicon is produced simultaneously when
cleavage of the quencher dye causes release of the reporter dye, which
results in the emission of a fluorescent signal that is detected by the
TaqMan detection camera. The amount of signal produced at a threshold
cycle within the purely exponential phase of the PCR reaction reflects
the starting copy number of the sequence of interest. The gene
expression data are expressed as a ratio between the PCR products
formed by the gene of interest relative to that formed by ß-actin,
the internal reference gene (18)
. For each tumor specimen,
each gene expression value was assigned to the phase of the cycle
determined by endometrial dating (proliferative versus
secretory versus menstrual).
For each marker, gene expression measurements at RT-PCR were repeated
at least two times.
For VEGF expression, sufficient mRNA was available in only 18 of 36
specimens. The other markers were tested in all 36 patients. The
variables of interest (i.e., the ratios of expression of the
genes of interest to that of ß-actin) were not normally distributed.
Logarithmic values were therefore taken for statistical analysis.
Differences between the proliferative and secretory specimens were
tested for by Student’s t test, and the statistical
significance levels were also computed using the Mann-Whitney
nonparametric two-sample approach. The means calculated (and quoted in
Table 1
, see "Results") are geometric means with associated 95% CIs
obtained by exponentiating the means and 95% CIs of the
logarithmically transformed values. All statistical significance levels
(Ps) quoted are two-sided. There was little difference
between any of the Ps calculated from the t tests
and from the nonparametric approach, and the values quoted are from the
t tests.
View this table:
[in this window]
[in a new window]
|
Table 1 Comparison of gene
expressiona in cervical cancers
removed during the proliferative (P) and secretory (S) phases of
the menstrual cycle
|
|
|
RESULTS
|
|---|
The phase of the menstrual cycle at the time of surgery was based
on endometrial dating: (a) in 21 of 36 patients, the tumor
specimens were removed during the proliferative phase; (b)
in 12 of 36 patients, the tumor specimens were removed during the
secretory phase; and (c) in 3 of 36 patients, the tumor
specimens were removed during the menstrual phase. Table 1
summarizes
the results of the study. The geometric means of TIMP-2 and COX-2 gene
expression are approximately 3-to 3.5-fold greater in the cervical
cancer tissues removed during the proliferative phase of the cycle
compared with cancer tissues removed during the secretory phase
(P = 0.007 and 0.030, respectively). The expression of
MMP-9, VEGF, and COX-1 was similar in the two groups. Figs. 1
2
3
4
5
show the individual data point for the results for each gene measured,
plotted against the timing of surgery as determined by pathological
dating of the matched endometrial tissue.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 1. The individual data points for the ratio of
COX-2:ß-actin gene expression values measured by RT-PCR from 36
cervical cancer specimens are displayed with the timing of surgery
during the menstrual cycle, as classified by endometrial dating.
|
|
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 2. The individual data points for the ratio
of TIMP-2:ß-actin gene expression values measured by RT-PCR from 36
cervical cancer specimens are displayed with the timing of surgery
during the menstrual cycle, as classified by endometrial dating.
|
|
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 3. The individual data points for the ratio of
MMP-9:ß-actin gene expression values measured by RT-PCR from 36
cervical cancer specimens are displayed with the timing of surgery
during the menstrual cycle, as classified by endometrial dating.
|
|
View larger version (11K):
[in this window]
[in a new window]
|
Fig. 4. The individual data points for the ratio
of COX-1:ß-actin gene expression values measured by RT-PCR from 36
cervical cancer specimens are displayed with the timing of surgery
during the menstrual cycle, as classified by endometrial
dating.
|
|
View larger version (8K):
[in this window]
[in a new window]
|
Fig. 5. The individual data points for the ratio of
VEGF:ß-actin gene expression values measured by RT-PCR from 18
cervical cancer specimens are displayed with the timing of surgery
during the menstrual cycle, as classified by endometrial
dating.
|
|
Noticeably, a large range of measured gene expression values were
observed for each of the markers (gene of interest:ß-actin). In the
case of COX-2, the values varied from 47 to 91,269, a 1,936-fold range.
However, within the secretory phase, only a 260-fold range was observed
(47–9,790). For TIMP-2, the range of gene expression was 1.29–1,252,
a 960-fold range. For MMP-9, the range of gene expression was
0.07–119, a 1,700-fold range. For COX-1, the range of gene expression
was 1–1,897, a 1,897 fold range. For the 18 patients whose
specimens had sufficient mRNA to conducting and repeating the
experiments twice, VEGF gene expression ranged from 0.09–25, a
2,700-fold range. Tumors removed during the proliferative phase
consistently had a trend for the highest gene expression.
|
DISCUSSION
|
|---|
Our data demonstrate that in cervical cancer, intratumoral gene
expression of COX-2 and TIMP-2 fluctuates significantly during the
menstrual cycle, with highest values in the proliferative phase
of the endometrium. This is consistent with the data reported by Saad
(9)
and Balsari (11)
on breast cancer, who
found that other markers for tumor aggressiveness and invasion such as
cathepsin D and HER2 tend to be overexpressed during the follicular
phase (i.e., proliferative phase of the endometrium).
It is intriguing to notice that we found the same trend for COX-2 and
TIMP-2 in cervical cancer. Although the geometric means of the values
were significantly different for TIMP-2 and COX-2 gene expression
during the different phases of the cycle, considerable overlap of
values during the proliferative and secretory phase was noticed
for all markers studied. This is to be expected if gene expression
reflects the physiological hormonal fluctuations during the menstrual
cycle because circulating estradiol concentrations peak during both the
secretory and proliferative phase of the endometrium. In the
absence of blood specimens to document the exact hormonal content at
the time of surgery, it is impossible to correlate gene expression with
hormonal measurements and precise dating. Future studies with a much
larger sample of patients and concurrent blood hormonal levels are
likely to better identify which window of time during the proliferative
phase is associated with increased intratumoral markers of metastatic
potential.
This is the first report of menstrual cycle-associated COX-2
fluctuations in humans, although fluctuation has been observed
previously in animals. COX-2 determination during the different phases
of the estrus cycle in a rat model showed COX-2 overexpression at the
peak of the serum estradiol level that precedes ovulation
(17)
. COX-2 was found to be highly and transiently
expressed in sheep 12–15 days after the estrous cycle and declined
thereafter to undetectable levels (16)
. In contrast, COX-1
expression did not change significantly during the menstrual cycle in
either the animal studies or the present study, which is consistent
with previous studies showing its expression to be constitutive and
noninducible (18)
.
COX-2 is often found to be elevated in cancers of major concern
and has received much attention in recent years because of the
possibility that tumors might be prevented by its inhibition.
Investigations into the role of COX-2 in tumorigenesis have disclosed
that it promotes angiogenesis and cell adhesion and inhibits apoptosis
(13, 14, 15)
. These findings suggest several mechanisms by
which higher expression of COX-2 could be a risk factor at the time of
surgery for cervical cancer. For example, cervical cancer surgery
carried out during periods of high COX-2 expression might enhance the
adhesion properties and thus enhance the probability of survival
of cancer cells that detach from the tumor during the surgery.
Similarly, a decreased propensity for apoptosis would help the
survival of cells that detach from the tumor.
If it is shown that survival of cervical cancers patients is related to
COX-2 expression, it may also be possible to eliminate this risk by
treating patients with COX-2 inhibitors before surgery.
Similar to the findings reported by Balsari et al.
(11)
and Saad et al. (9)
,
our results should only be considered exploratory, requiring
confirmation and extensive studies to elucidate possible mechanisms and
the importance of the findings. However, it is intriguing to think that
in premenopausal women, cyclic hormonal fluctuation might potentially
affect the behavior of cancer cells when tumor surgery is performed. We
have begun a tumor registry-based retrospective study of survival of
premenopausal cervical cancer patients to investigate whether the phase
of the menstrual cycle at operation has affected patient survival. Much
more valuable would be a prospective study with determination of COX-2
and other gene expressions as well as precise hormonal determinations
at the time of operation to be correlated with subsequent survival.
Finally, it could also be speculated that hormonal changes during the
follicular phase of the menstrual cycle may affect the biology of
tumors other than breast or cervical cancer in premenopausal women:
if such differences at the time of surgery prove to reflect on patient
outcome, it could become crucial to correctly schedule any tumor
surgery in premenopausal women.
|
FOOTNOTES
|
|---|
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.
1 To whom requests for reprints should be
addressed. Present address: Department of Radiation Oncology, New York
University School of Medicine, 566 First Avenue, New York, NY 10016.
Phone: (212) 263-2601; E-mail: silvia.formenti{at}med.nyu.edu 
2 The abbreviations used are: MMP, matrix
metalloproteinase; RT-PCR, reverse transcription-PCR; TIMP, tissue
inhibitor of metalloproteinase; COX, cyclooxygenase; VEGF, vascular
endothelial growth factor; CI, confidence interval.
Received 2/20/00;
revised 10/11/00;
accepted 10/11/00.
|
REFERENCES
|
|---|
-
Hrushesky W. J., Bluming A. Z., Gruber S. A., Sothern R. B. Menstrual influence on surgical cure of breast cancer. Lancet, 2: 949-952, 1989.[CrossRef][Medline]
-
Badwe R. A., Gregory W. M., Chaudary M. A., Richards M. A., Bentley A. E., Rubens R. D., Fentiman I. S. Timing of surgery during menstrual cycle and survival of premenopausal women with operable breast cancer. Lancet, 337: 1261-1264, 1991.[CrossRef][Medline]
-
Gnant M. F., Seifert M., Jakesz R., Adler A., Mittlboeck M., Sevelda P. Breast cancer and timing of surgery during menstrual cycle. A 5-year analysis of 385 pre-menopausal women. Int. J. Cancer, 52: 707-712, 1992.
-
Veronesi U., Luini A., Mariani L., Del Vecchio M., Alvez D., Andreoli C., Giacobone A., Merson M., Pacetti G., Raselli R. Effect of menstrual phase on surgical treatment of breast cancer. Lancet, 343: 1545-1547, 1994.[CrossRef][Medline]
-
Kroman N., Hojgaard A., Andersen K. W., Graversen H. P., Afzelius P., Lokdam A., Juul C., Hoffmann J., Bentzon N., Mouridsen H. T. Timing of surgery in relation to menstrual cycle does not predict the prognosis in primary breast cancer. Eur. J. Surg. Oncol., 20: 430-435, 1994.[Medline]
-
Saad Z., Vincent M., Bramwell V., Stitt L., Duff J., Girotti M., Jory T., Heathcote G., Turnbull I., Garcia B. Timing of surgery influences survival in receptor-negative as well as receptor-positive breast cancer. Eur. J. Cancer, 30: 1348-1352, 1994.[CrossRef]
-
Senie R. T., Tenser S. M. The timing of breast cancer surgery during the menstrual cycle. Oncology (Basel), 11: 1509-1522, 1997.[Medline]
-
Jager W., Sauerbrei W. Effect of timing of surgery during the menstrual cycle of premenopausal breast cancer patients. Breast Cancer Res. Treat., 34: 279-287, 1995.[CrossRef][Medline]
-
Saad Z., Bramwell V. H., Wilson S. M., O’Malley F. P., Jeacock J., Chambers A. F. Expression of genes that contribute to proliferative and metastatic ability in breast cancer resected during various menstrual phases. Lancet, 5: 1170-1173, 1998.
-
Hrushesky W. J., Gruber S. A., Southern R. B., Hoffman R. A., Lakatua D., Carlson A., Cerra F., Simmons R. L. Natural killer cell activity: age, estrous- and circadian-stage dependence and inverse correlation with metastatic potential. J. Natl. Cancer Inst., 80: 1232-1237, 1988.[Abstract/Free Full Text]
-
Balsari A., Casalini P., Tagliabue E., Greco M., Pilotti S., Agresti R., Giovanazzi R., Alasio L., Rumio C., Cascinelli N., Colnaghi M. I., Menard S. Fluctuation of HER2 expression in breast carcinomas during the menstrual cycle. Am. J. Pathol., 155: 1543-1547, 1999.[Abstract/Free Full Text]
-
Ferrara N. Vascular endothelial growth factor. Eur. J. Cancer, 32: 2413-2422, 1996.
-
Tsujii M., DuBois R. N. Alterations in cellular adhesion and apoptosis in epithelial cells overexpressing prostaglandin endoperoxide synthase 2. Cell, 83: 493-501, 1995.[CrossRef][Medline]
-
Boolbol S. K., Dannenberg A. J., Chadburn A., Martucci C., Guo X. J., Ramonetti J. T., Abreu-Goris M., Newmark H. L., Lipkin M. L., DeCosse J. J., Bertagnolli M. M. Cyclooxygenase-2 overexpression and tumor formation are blocked by sulindac in a murine model of familial adenomatous polyposis. Cancer Res., 56: 2556-2560, 1996.[Abstract/Free Full Text]
-
Tsujii M., Kawano S., Tsuji S., Sawaoka H., Hori M., DuBois R. N. Cyclooxygenase regulates angiogenesis induced by colon cancer cells. Cell, 93: 705-716, 1998.[CrossRef][Medline]
-
Charpigny G., Reinaud P., Tamby J. P., Creminon C., Martal J., Maclouf J., Guillomot M. Expression of cyclooxygenase-1 and -2 in ovine endometrium during the estrous cycle and early pregnancy. Endocrinology, 138: 2163-2171, 1997.[Abstract/Free Full Text]
-
Shoda T., Hatanaka K., Saito M., Majima M., Ogino M., Harada Y., Nishijima M., Katori M., Yamamoto S. Induction of cyclooxygenase type-2 (COX-2) in rat endometrium at the peak of serum estradiol during the estrous cycle. Jpn. J. Pharmacol., 69: 289-291, 1995.[Medline]
-
Jun S. S., Chen Z., Pace M. C., Shaul P. Estrogen up regulates cyclooxygenase-1 gene expression in ovine fetal artery endothelium. J. Clin. Invest., 102: 176-183, 1998.[Medline]
-
Horikoshi T., Danenberg K. D., Stadlbauer T. H., Volkenandt M., Shea L. C., Aigner K., Gustavsson B., Leichman L., Frosing R., Ray M. Quantitation of thymidylate synthase, dihydrofolate reductase, and DT-diaphorase gene expression in human tumors using the polymerase chain reaction. Cancer Res., 52: 108-116, 1992.[Abstract/Free Full Text]
Top
ABSTRACT
INTRODUCTION
