You have to try to avoid trying to think these things out ourselves.  We
all have models in our head about what is going on, but the biology
doesn't care what we are thinking.  A lot about disease, including
prostate cancer, is not well understood and even the models which
experts develop from years of study sometimes turn out to be misleading.

What you describe can happen, but I have no idea how often it happens
and whether it materially affects the statistics.

Bowel cancer doesn't usually start in the prostate.  It can metastsize
to different organs, but I thought the usual place it goes to first is
the bone.  Also, they don't treat prostate cancer, even that which has
metastzsized with chemotherapy like the kind you describe.  But of
course this could be some unusual form of prostate cancer.

Actually your argument also applies to diagnosing prostate cancer by
digital rectal examination.  No double blind or otherwise suitably
randomized study has ever been done to show that life span is
increased---your criterion---by doing such examinations.   Indeed there
are those who would argue not only against PSA testing but also against
regular DREs.   They would leave all prostate cancer to be discovered
through symptoms of advanced prostate cancer.

In any event, the statments I made are supported by lots of evidence.
For example, the Partin Table studies I referred to show that after
surgery, cancers are generally more advanced in T2 or higher stages.
Also, the significant decline in advanced prostate cancer diagnosis
after the advent of PSA testing.

As usual you conflate different issues and always come up with the same
simple minded response.

There are a substantial number of false positives.  But what that means
is that you have a biopsy and the biopsy is negative.  The biopsy
procedure is quite safe and not particularly painful for most men.  It
would be better of course if they could reduce the number of false
positives,  but it is not a disaster to have a biopsy and find it comes
out negative.

It is also possible for the PSA to be normal and the patient to have
prostate cancer.   But there isn't much we can do at present about that.
 It is the reason why you should also have a digital rectal exam.  The
two of them together will catch the overwhelmingly majority of prostate
cancers.

I am not a physician myself, but I've never seen it stated that
prostatitis increased the risk of prostate cancer.   You should find
another doctor and ask his opinion.  Don't just take my word for it.

You shouldn't be trying to get medical advice over the internet.  Find a
doctor you trust, make sure you get him to explain it all to you, and
then make sure you get regular exams.

There are plenty of good doctors out there.  You shouldn't let your
experience with one doctor put you off the whole profession.

From: oitbso@yahoo.com (ron)
c palmer wrote...snip...
it is a fact that the prostate cell is the ONLY cell in the human body
that produces psa.

Hi Curtis...This doesn't affect the points you were making, but I
thought you'd be interested in knowing that other organs do produce
small amounts of PSA. The name, "PSA", was given before tests were
available with the sensitivity necessary to detect these small amounts
of PSA. Some women produce enough PSA that it can be detetcted with the
ultrasensitive PSA test, and, of course, they don't have
prostates...Best wishes and good health, Ron
====================hi ron - what i was pointing out was in the male body only.

you are right on what you said on the female body.

here's an article with references to what you are saying.

Prostate-Specific Antigen in Female Serum, a Potential New Marker of
Androgen Excess
Dimitrios N. Melegos, He Yu, Mala Ashok, Chun Wang, Frank Stanczyk and
Eleftherios P. Diamandis
Department of Pathology and Laboratory Medicine (D.N.M., H.Y., E.P.D.),
Mount Sinai Hospital, Toronto, Ontario M5G 1X5; Department of Clinical
Biochemistry (D.N.M., H.Y., E.P.D.), University of Toronto, Toronto,
Ontario M5G 1L5, Canada; and Department of Obstetrics and Gynecology
(M.A., C.W., F.S.), Division of Reproductive Endocrinology and
Infertility, University of Southern California School of Medicine, Los
Angeles, California 90033
Address correspondence and requests for reprints to: Dr. E. P.
Diamandis, Department of Pathology and Laboratory Medicine, 600
University Avenue, Toronto, Ontario, Canada, M5G 1X5.
   Abstract

 
Prostate-specific antigen (PSA) is present at very low concentrations in
female serum, but it can now be measured with highly sensitive
immunoassays. We have found that in female tissues the PSA gene is
regulated by steroid hormones through the action of steroid hormone
receptors. Thus, we examined whether female serum PSA is associated with
hyperandrogenic states. Serum PSA levels were compared between 22
hirsute women with a Ferriman-Gallwey score higher than 8 and 50 women
without hirsutism. The results show that PSA levels were higher in
hirsute women in comparison with controls. In hirsute women, levels of
PSA and 3-androstanediol glucuronide (3-AG), a specific metabolite of
androgen action, showed a significant positive correlation, whereas PSA
and 3-AG showed a significant negative correlation with patient age.
Receiver operating characteristic (ROC) analysis revealed that 3-AG was
a slightly better marker of androgen excess than PSA. We conclude that
female serum PSA may be a new biochemical marker of androgen action in
females.

 
PROSTATE-SPECIFIC antigen (PSA) is a 33-KDa serine protease with
chymotrypsin-like enzymatic activity (1). In males, PSA is produced by
the prostate gland (2), and it is present in prostatic tissue, seminal
plasma, and male serum (3). Small amounts of PSA also can be produced by
the periurethral glands, and therefore, urine contains detectable levels
of PSA (4). PSA concentration in male serum is a valuable tumor marker
for diagnosis and management of prostate cancer (5). PSA production in
the prostate is under the control of steroid hormones. Androgens
up-regulate the expression of the PSA gene through the androgen receptor
(6, 7).
Initially, PSA was believed to be completely absent from all female
tissues and fluids. However, PSA has been detected recently in some
female tissues (including breast, ovarian, and endometrial tissues) and
body fluids (amniotic fluid, milk, and breast cyst fluid) (8). The
presence of PSA in these female tissues seems to be associated closely
with steroid hormone regulation, especially androgens, glucocorticoids,
and progestins (9). This was also demonstrated using a tissue culture
system and the breast carcinoma cell line T-47D (10). Indirect in vivo
evidence was provided by two case reports (11, 12).
Using conventional PSA assays with a detection limit of 0.1–0.01
ng/mL, PSA is detectable in less than 10% of female sera (13), but when
a more sensitive PSA assay is used (detection limit 0.001 ng/mL), more
than 50% of female sera have detectable PSA levels (14). Among women who
have high levels of androgens, relatively high levels of serum PSA
should be expected if PSA production in women is under the regulation of
androgens.
One of the manifestations of androgen excess in women is idiopathic
hirsutism. In most instances, the source of androgen excess in these
women is neither adrenal (e.g. dehydroepiandrosterone sulfate) nor
ovarian (e.g. testosterone) but peripheral. The most important
peripheral sources of androgen production are sexual and nonsexual skin
tissues where testosterone is converted to the potent androgen,
dihydrotestosterone (DHT). The latter androgen is responsible for
androgen action.
Based on the above considerations, we speculated that serum PSA levels
in hirsute females may be high and that the measurement of serum PSA in
these women may have some clinical implications. In this study, we
measured PSA levels in female serum with a highly sensitive PSA assay
and compared the levels between hirsute and apparently healthy women.
Because 3-androstanediol glucuronide (3-AG) is a major metabolite of DHT
and it can be measured easily in serum, we have conducted comparisons
between PSA and 3-AG in serum of hirsute women.

We collected, with informed consent, serum specimens from 22 hirsute
women and 16 nonhirsute women from the same geographical location in
California. The population consisted mainly of Caucasians and Hispanics.
Additionally, we collected another 34 serum samples from healthy blood
donors in Toronto, Canada. The clinical diagnosis of hirsutism was based
on the hirsutism scale of Ferriman and Gallwey, as described elsewhere
(15). A score of 8 or more indicates clinical hirsutism. Control
subjects had a score less than 8. Of the hirsute women, 20 had a score
between 8 and 24 with a mean ± SD of 13 ± 4.8. The score was not
available for two hirsute women.
Other available information pertaining to the hirsute women included: 1)
serum levels of total testosterone, which was measured in 14 patients
(range: 0.12–1.53 ng/mL; mean ± SD, 0.85 ± 0.44 ng/mL); 2)
ovulation, which was available for 21 patients (17 anovulating and 4
ovulating women); and 3) body weight, which was available for 21
patients (11 obese and 10 nonobese women).
PSA and androstanediol glucuronide were measured in serum by previously
described methods (16, 17). A sandwich-type time-resolved
immunofluorometric assay was used to quantify PSA. The lowest detection
limit of the assay is 1 pg/mL, and the coefficient of variation (for
between-run precision), at levels of 2 pg/mL or higher, is less than
16%. Each serum sample was measured in triplicate. The DHT metabolite,
3-AG, was measured by direct RIA using reagents obtained from a
commercial 125I-androstanediol glucuronide RIA kit (Diagnostic Systems
Laboratories, Webster, Texas). This kit was validated extensively in the
laboratory of Dr. F. Stanczyk, Los Angeles, California (17).
The differences in serum PSA and age between cases and controls were
compared with the Wilcoxon test and the ANOVA test. The relationships
between PSA and age or 3-AG were examined with the Spearman rank
correlation coefficient test when untransformed PSA values were used.
The Pearson correlation test was used when logarithmic PSA values were
employed. The receiver operating characteristic (ROC) curve was
constructed for both PSA and 3-AG using the cutoff levels at 20%, 50%,
70%, and 90% percentile distributions of their values among all
subjects. The start and end points of the curve were hypothetical, i.e.
assuming the curve starts at 100% specificity and 0% sensitivity and
ends at 0% specificity and 100% sensitivity.

 
Table 1 shows the distributions of age, PSA, and 3-AG among the three
patient groups. The mean ages and median PSA levels between the control
groups 1 and 2 were not significantly different (P = 0.26 for age and P
= 0.28 for PSA), and therefore, the two control groups were combined
together for comparison with the patient group. The mean ages were not
significantly different between patients (29.6 yr) and combined control
groups (30.7 yr) (P = 0.48).

  Table 1. The distributions of PSA, age, and 3-AG in the patients and
controls
 
The distribution of PSA values among all the women studied was
positively skewed, but the 3-AG values were normally distributed with an
almost identical mean and median. The average levels of PSA (median) and
3-AG (mean) were significantly higher in hirsute women than in normal
women (P = 0.001 for PSA, Wilcoxon Rank Sum test; and P = 0.002 for
3-AG, one-way ANOVA).
PSA values were positively correlated with 3-AG values (Fig. 1). PSA and
3-AG values were negatively correlated with patient's age (Table 2, Fig.
2, and data not shown). The regression equation of 3-AG levels (y) vs.
age (x) was: y = -0.22x + 11.6, r = -0.58, P < 0.001.
View larger version (16K):
[in this window]
[in a new window]
  Figure 1. Correlation between 3-AG and log (PSA) for 32 patients'
sera. The Pearson r = 0.58 (P < 0.001). The regression equation is y 0.17x - 0.20. The values of PSA before logarithmic transformation are in
pg/mL.
 

  Table 2. Correlations among PSA, 3-AG, and patient
 

  Figure 2. Correlation between log (PSA) and age for 72 patients'
sera. The Pearson r = -0.37 (P = 0.001). The regression equation is y -0.040x + 1.57. The values of PSA before logarithmic transformation are
in pg/mL.
 
The distributions of logarithmic PSA and 3-AG between patients and
controls are shown in Figs. 3 and 4. Although patients tended to have
higher PSA and 3-AG than controls, there were still many patients who
had values overlapping with those of controls. The ROC curve was
slightly better for 3-AG compared with that for PSA, but none of the
markers could provide both more than 80% sensitivity and specificity at
any single cutoff point (Fig. 5).

  Figure 3. Distribution of log (PSA) values between patients (dotted
bars) and control subjects (hatched bars). For discussion, see text.
 

  Figure 4. Distribution of 3-AG values between patients (dotted bars)
and control subjects (hatched bars). For discussion, see text.
 

  Figure 5. Receiver operating characteristic curves for PSA () and
3-AG ()at various cutoff points shown in brackets. Units are in pg/mL
for PSA and ng/mL for 3-AG.
 
Based on the limited number of patients who had testosterone data, we
found that serum testosterone levels tended to be positively correlated
with 3-AG (Pearson r of 0.62, P = 0.02) and PSA (Spearman r of 0.47, P 0.09) and were negatively correlated with age (Pearson r of -0.53, P 0.05).
Obese women tended to have a higher hirsutism score than nonobese women
(median scores 16 vs. 10, respectively, P = 0.01). However, using both
Wilcoxon rank sum testing and one-way ANOVA, we could not find any
significant associations between either obesity or ovulation and age,
3-AG, PSA, or testosterone (data not shown).

 
PSA was thought to be absent from female tissues until our recent
finding of its presence in female breast tissue. Our studies have
demonstrated that the female breast is one of the female tissues that is
capable of producing PSA (9, 11). Tissue culture experiments indicate
that the production of PSA by steroid hormone receptor-positive breast
cancer cells is under the control of steroid hormones, including
androgens and progestins (10). Oral contraceptives increase PSA
production in breast tissue (11). Amniotic fluid and milk contain
significant amounts of PSA, and the levels of PSA in these fluids change
with gestational age or postdelivery time (18, 19). Serum PSA levels are
significantly higher in pregnant women in comparison with nonpregnant
women (18). Taken together, these data suggest that PSA is expressed in
breast and possibly other female tissues and that PSA expression is
under the regulation of steroid hormones, especially androgens and
progestins. The major difference in PSA expression between men and women
is that the breasts express relatively low levels compared with the
levels of male prostate.
Because of the relationship between PSA production and androgen
regulation, we hypothesized that PSA may be a marker of androgen action
in women. Women with higher levels of androgen may have higher levels of
PSA compared with women with normal levels of androgen. Hirsutism
represents a state of androgen excess in women. In this study, we report
significantly elevated serum PSA levels in hirsute women compared with
normal women.
The source of androgen excess in patients with idiopathic hirsutism is
considered to be increased peripheral conversion of androstanediol and
testosterone to DHT via the pivotal enzyme, 5-reductase. Although DHT is
the most potent endogenous androgen, it is considered to be a poor
circulating marker of androgenicity. Instead, the conjugated metabolite
of DHT, namely 3-AG, is considered to be an excellent serum marker of
5-reductase activity and peripheral androgen action and of the clinical
manifestations of hirsutism (20). For these reasons, we chose 3-AG as a
serum marker for determining the relationship of peripheral
hyperandrogenism to circulating PSA levels.
Comparison of serum PSA with 3-AG for the diagnosis of hirsutism showed
that PSA did not provide better sensitivity and specificity than did
3-AG. The two markers correlate significantly with each other. An
inverse correlation between age and PSA or 3-AG was observed in these
women. We do not as yet know which female tissue produces and releases
PSA into the circulation in hirsute women. The most likely candidate is
the female breast because this tissue has steroid hormone receptors and
is capable of producing high levels of PSA, especially after steroid
hormone stimulation (11). Nevertheless, our data show that only a
relatively small proportion of patients with hirsutism produces more
than 10 pg/mL PSA (Fig. 3).
In summary, we found that serum PSA levels were increased significantly
in women with hirsutism. PSA levels in female serum were positively
correlated with serum 3-AG and were inversely correlated with patient
age. Therefore, PSA in serum can now be regarded as another biochemical
marker of androgen action in female peripheral tissues.
Received June 11, 1996.
Revised September 18, 1996.
Revised November 8, 1996.
Accepted November 12, 1996.

McCormack RT, Rittenhouse HG, Finlay JA, et al. 1995 Molecular forms of
prostate specific antigen and the human kallikrein gene family: a new
era. Urology. 45:729–744.[CrossRef][Medline]  
Wang MC, Valenzuela LA, Murphy GP, Chu TM. 1979 Purification of human
prostate-specific antigen. Invest Urol. 17:159–163.[Medline]  
Papsidero LD, Wang MC, Valenzuela LA, Murphy GP, Chue TM. 1980 A
prostate antigen in sera of prostatic cancer patients. Cancer Res.
40:2428–2432.[Abstract]  
Iwakiri J, Granbois K, Wehner N, Graves HC, Stamey T. 1993 An analysis
of urinary prostate specific antigen before and after radical
prostatectomy: evidence for secretion of prostate specific antigen by
the periurethral glands. J Urol. 149:783–786.[Medline]  
Catalona WJ, Smith DS, Ratliff TL, et al. 1991 Measurement of prostate
specific antigen in serum as a screening test for prostate cancer. N
Engl J Med. 324;1156–1161.  
Luke MC, Coffey DS. 1994 Human androgen receptor binding to the androgen
response element of prostate specific antigen. J Androl.
15:41–51.[Abstract/Free Full Text]  
Young CY-F, Montgomery BT, Andrews PE, Qiu S, Bilhartz DL, Tindall DJ.
1991 Hormonal regulation of prostate specific antigen messenger RNA in
human prostatic adenocarcinoma cell line LNCap. Cancer Res.
51:3748–3752.  
Diamandis EP, Yu H. 1995 New biological functions of prostate specific
antigen? J Clin Endocrinol Metab. 80:1515–1517.[Medline]  
Yu H, Diamandis EP, Sutherland DJA. 1994 Immunoreactive prostate
specific antigen levels in female and male breast tumors and its
association with steroid hormone receptors and patient age. Clin
Biochem. 27:75–79.[CrossRef][Medline]  
Yu H, Diamandis EP, Zarghami N, Grass L. 1994 Induction of prostate
specific antigen production by steroids and tamoxifen in breast cancer
cell lines. Breast Cancer Res Treat. 32:291–300.[Medline]  
Yu H, Diamandis EP, Monne M, Croce CM. 1995 Oral contraceptive-induced
expression of prostate specific antigen in the female breast. J Biol
Chem. 270:6615–6618.[Abstract/Free Full Text]  
Yu H, Diamandis EP, Levesque M, Asa S, Monne M, Croce CM. 1995
Expression of the prostate-specific antigen gene by a primary ovarian
carcinoma. Cancer Res. 55:1603–1606.[Abstract]  
Yu H, Diamandis EP. 1995 Measurement of serum prostate specific antigen
levels in females and in prostatectomized males with an ultrasensitive
immunoassay technique. J Urol. 153:1004–1008.[Medline]  
Diamandis EP, Yu H, Melegos DN. 1996 Ultrasensitive prostate specific
antigen assays and their clinical application. Clin Chem.
42:853–857.[Free Full Text]  
Hatch R, Rosenfield RL, Kiru MH, Tredway D. 1981 Hirsutism:
implications, etiology, and management. Am J Obstet Gynecol.
140:815–830.[Medline]  
Ferguson RA, Yu H, Kalyvas M, Zammit S, Diamandis EP. 1996
Ultrasensitive detection of prostate specific antigen by a new time
resolved immunofluorometric assay and the Immulite
immunochemiluminescent third generation assay: potential applications in
prostate and breast cancers. Clin Chem. 42:675–684.[Abstract/Free
Full Text]  
Norang R, Tran T, Savjani G. 1994 Solid phase radioimmunoassay kit for
the quantitative measurement of androstanediol glucuronide in
unextracted serum or plasma. Clin Chem. 40:1132 (Abstract).  
Melegos DN, Yu H, Allen LC, Diamandis EP. 1996 Prostate specific antigen
in amniotic fluid of normal and abnormal pregnancies. Clin Biochem.
29:555–562.[CrossRef][Medline]  
Yu H, Diamandis EP. 1995 Prostate specific antigen in milk of lactating
women. Clin Chem. 41:54–58.[Abstract/Free Full Text]  
Horton R, Lobo R. 1986 Peripheral androgen metabolism and the role of
androstanediol glucuronide. In: Horton R, Lobo RA, eds. Clinics in
endocrinology and metabolism, Vol 15. Philadelphia: WB Saunders;
293–306.
This article has been cited by other articles:
H. Aksoy, F. Akcay, Z. Umudum, A. K. Yildirim, and R. Memisogullari
Changes of PSA Concentrations in Serum and Saliva of Healthy Women
during the Menstrual Cycle
Ann. Clin. Lab. Sci., January 1, 2002; 32(1): 31 - 36.
[Abstract] [Full Text] [PDF]

A. SULLI, C. PIZZORNI, R. SCOTTO-BUSATO, S. ACCARDO, and M. CUTOLO
Androgenizing Effects of Cyclosporin A in Rheumatoid Arthritis
Ann. N.Y. Acad. Sci., June 22, 1999; 876(1): 391 - 396.
[Full Text]

C. V. Obiezu, A. Scorilas, A. Magklara, M. H. Thornton, C. Y. Wang, F.
Z. Stanczyk, and E. P. Diamandis
Prostate-Specific Antigen and Human Glandular Kallikrein 2 Are Markedly
Elevated in Urine of Patients with Polycystic Ovary Syndrome
J. Clin. Endocrinol. Metab., April 1, 2001; 86(4): 1558 - 1561.
[Abstract] [Full Text]

H. F. Escobar-Morreale, S. Ávila, and J. Sancho
Serum Prostate-Specific Antigen Concentrations Are Not Useful for
Monitoring the Treatment of Hirsutism with Oral Contraceptive Pills
J. Clin. Endocrinol. Metab., July 1, 2000; 85(7): 2488 - 2492.
[Abstract] [Full Text]

C. Negri, F. Tosi, R. Dorizzi, A. Fortunato, G. G. Spiazzi, M. Muggeo,
R. Castello, and P. Moghetti
Antiandrogen Drugs Lower Serum Prostate-Specific Antigen (PSA) Levels in
Hirsute Subjects: Evidence That Serum PSA Is a Marker of Androgen Action
in Women
J. Clin. Endocrinol. Metab., January 1, 2000; 85(1): 81 - 84.
[Abstract] [Full Text]

V. H. H. Goh
Breast Tissues in Transsexual Women-A Nonprostatic Source of Androgen
Up-Regulated Production of Prostate-Specific Antigen
J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3313 - 3315.
[Abstract] [Full Text]

H. F. Escobar-Morreale, J. Serrano-Gotarredona, S. Avila, J.
Villar-Palasí, C. Varela, and J. Sancho
The Increased Circulating Prostate-Specific Antigen Concentrations in
Women with Hirsutism Do Not Respond to Acute Changes in Adrenal or
Ovarian Function
J. Clin. Endocrinol. Metab., July 1, 1998; 83(7): 2580 - 2584.
[Abstract] [Full Text]

C. López-Otín and E. P. Diamandis
Breast and Prostate Cancer: An Analysis of Common Epidemiological,
Genetic, and Biochemical Features
Endocr. Rev., August 1, 1998; 19(4): 365 - 396.
[Abstract] [Full Text]

C. V. Obiezu, E. J. Giltay, A. Magklara, A. Scorilas, L. J.G. Gooren, H.
Yu, D. J.C. Howarth, and E. P. Diamandis
Serum and Urinary Prostate-specific Antigen and Urinary Human Glandular
Kallikrein Concentrations Are Significantly Increased after Testosterone
Administration in Female-to-Male Transsexuals
Clin. Chem., June 1, 2000; 46(6): 859 - 862.
[Abstract] [Full Text] [PDF]

D. N. Melegos and E. P. Diamandis
Is Prostate-Specific Antigen Present in Female Serum?
Clin. Chem., March 1, 1998; 44(3): 691 - 692.
[Full Text] [PDF]

G. M. Lambert-Messerlian, J. A. Canick, D. N. Melegos, and E. P.
Diamandis
Increased concentrations of prostate-specific antigen in maternal serum
from pregnancies affected by fetal Down syndrome
Clin. Chem., February 1, 1998; 44(2): 205 - 208.
[Abstract] [Full Text] [PDF]

knowledge is power - growing old is mandatory - growing wise is optional    
"Many more men die with prostate cancer than of it. Growing old is
invariably fatal. Prostate cancer is only sometimes so."
http://community.webtv.net/PALMER_ENT/doc

Nope. So far, so good in the hair department.

How do I determine which readings were affected?
My PSA after Lupron+Casodex was 0.3, on 4/29/02. On 1/17/03 it was 5.2
and climbed to a high of 9.9 by 4/12/04. Then it dropped down to 5.17
on 7/6/04 and has stayed between 8.3 and 9.2 (8.8 most recently) since
9/2/04.

No. They said no prostate was visible, but the "prostate bed" is there.

I was 62 two months ago. When I had the surgery, almost exactly nine
years ago, my PSA was 12 and a Gleason 7.

I haven't seen any recent data about Celebrex relative to cancer. Have
you? Here's the intro of the article that got me interested:

"April 24, 2003
Researchers at the University of Texas M.D. Anderson Cancer Center
have, for the first time, identified the molecular pathway by which a
commonly prescribed arthritis medication inhibits the growth of cancer.

Before this study, scientists had linked use of celecoxib capsules
(commonly known as Celebrex) to prevention of cancer, but the way in
which the medication acted in cancer cells was unknown.

Now, investigators have found that celecoxib capsules stop a key
transcription factor known as Sp1 from turning on multiple genes in
cancer cells known to be associated with cancer growth. One of those
genes triggers production of vascular endothelial growth factor (VEGF),
the predominant angiogenic factor that leads blood vessels to grow to
feed tumors.

The findings were published in the Proceedings for the 2003 Annual
Meeting of the American Association for Cancer Research.

"Our results provide a novel molecular mechanism for the antitumor
activity of celecoxib," said Keping Xie, MD, PhD, assistant professor
in the department of gastrointestinal medical oncology.

Xie added that although the study was conducted in models of pancreatic
cancer, the results likely describe how celecoxib interferes with
development of a number of common cancers that all involve the Sp1/VEGF
pathway."