The present invention involves a method for early detection of
developing tamoxifen resistance in breast cancer. Tamoxifen is the
drug of choice for hormonal therapy of a first recurrence of breast
cancer, but its use is associated eventually with emergence of resistant
tumors. Whereas initial treatment is usually followed by tumor regression,
resistant tumors may actually resume growth under continued tamoxifen
treatment. Because such growth may actually be augmented by the
tamoxifen, it is essential to identify the onset of resistance as
early as possible so alternative therapy may be promptly instituted.
Monitoring increased levels of estrogenic tamoxifen isomers or estrogenic
tamoxifen metabolites signals developing resistance.
1. A method for detecting in vivo development of tamoxifen-resistant
breast tumors in a patient being subjected to a course of tamoxifen
determining ratios of cis-4-hydroxy-tamoxifen concentration to
trans-4-hydroxy-tamoxifen concentration in a fluid from said patient;
following said ratios during the course of tamoxifen treatment
to determine onset of treatment resistance as characterized by an
increase in the ratio of cis-4-hydroxy-tamoxifen concentration to
trans-4-hydroxy-tamoxifen concentration as compared to said ratio
expected for tumors sensitive to tamoxifen treatment.
2. The method according to claim 1 wherein said fluid is from breast
3. The method according to claim 1 wherein said fluid is blood
4. The method according to claim 1 wherein said fluid is blood
5. The method according to claim i wherein said fluid is blood
6. A method for detecting development of tamoxifen-resistant breast
tumors in a breast cancer patient being subjected to a course of
tamoxifen treatment comprising:
determining tamoxifen concentration in fluid or tumor samples from
said patient; and
following said concentration during the course of tamoxifen treatment
to determine onset of treatment resistance as characterized by a
decrease in said tamoxifen concentration as compared to said concentration
expected for tumors sensitive to tamoxifen treatment.
7. A method for detecting development of tamoxifen-resistant breast
tumors in a breast cancer patient being subjected to a course of
tamoxifen treatment, the method comprising:
determining tamoxifen concentration and a ratio of cis-4-hydroxy-tamoxifen
concentration to trans-4-hydroxy-tamoxifen concentration in a sample
of fluid or breast-tumor from said patient; and
following said tamoxifen concentration and ratio during the course
of tamoxifen treatment to determine onset of treatment resistance
as characterized by a decrease in tamoxifen concentration as compared
to said concentration expected for tumors sensitive to tamoxifen
treatment and an increase in the ratio of cis-4-hydroxy-tamoxifen
concentration to trans-4-hydroxy-tamoxifen concentration as compared
to said ratio expected for tumors sensitive to tamoxifen treatment.
8. A method for detecting development of resistance to treatment
by a trans triphenylethylene antiestrogen in a breast cancer patient
being subjected to a course of treatment with a trans triphenylethylene
antiestrogen, the method comprising:
determining tissue ratios of cis triphenylethylene antiestrogen
isomer to trans triphenylethylene antiestrogen isomer in fluid or
tissue from said patient; and
following said ratio during the course of treatment to determine
onset of treatment resistance as characterized by an increase in
the ratio as compared to said ratio expected for tumors sensitive
to tamoxifen treatment.
9. The method according to claim 8 wherein said tissue is breast
10. The method according to claim 8 wherein said tissue is blood
11. The method according to claim 8 wherein said tissue is blood
12. The method according to claim 8 wherein said tissue is blood
13. The method according to claims 1, 7 or 8 wherein said determining
step comprises photoactivation of a tissue sample extract.
14. The method of claim 13 wherein the photoactivation involves
irradiation with ultraviolet light.
15. The method of claims 1, 7 or 8 wherein the determining step
comprises high performance liquid chromatography.
16. A method for detecting the onset of tumor tamoxifen resistance
in a breast cancer patient being subjected to a course of tamoxifen
treatment, the method comprising, during the course of said treatment:
(a) determining a level of tamoxifen, an antiestrogenic tamoxifen
isomer or an antiestrogenic tamoxifen metabolite in a tissue or
fluid sample from said patient;
(b) determining a level of an estrogenic tamoxifen isomer or an
estrogenic tamoxifen metabolite in a tissue or fluid sample from
said patient; and
(c) detecting onset of tumor tamoxifen resistance when the level
determined in step (b) increases relative to the level determined
in step (a) as compared to the levels expected for tumors sensitive
to tamoxifen treatment.
17. The method of claim 16 where a level of antiestrogenic tamoxifen
isomer is determined in step (a) and a level of estrogenic tamoxifen
isomer is determined in step (b).
18. The method of claim 16 where a level of tamoxifen is determined
in step (a) and a level of an estrogenic tamoxifen metabolite is
determined in step (b).
19. The method of claim 18 where a level of tamoxifen is determined
in step (a) and a level of tamoxifen monophenol is determined in
20. The method of claim 16 wherein the estrogenic tamoxifen metabolite
is tamoxifen monophenol.
21. The method of claim 16 wherein the estrogenic tamoxifen metabolite
is tamoxifen bisphenol.
The present invention relates to methods for monitoring the effectiveness
of tamoxifen (2-[4-(1,2-diphenyl-1-butenyl)phenoxy]-N,N-dimethylethanamine)
therapy in the treatment of breast cancer, and more particularly
to methods of detecting the emergence of tamoxifen-resistant tumors
and the resulting treatment resistance by means of analytical assays
for antiestrogenic and other forms of tamoxifen as well as its estrogenic
Breast cancer is the most common form of malignant disease among
women of the Western world, and it is the most common cause of death
among those between 40 and 45 years of age. This disease will develop
in about six to seven percent of women in the United States, and
at the present time about one half of this group can be cured. The
treatment of breast cancer involves surgery, radiation, chemotherapeutics
and hormonal therapy, the last category including consideration
of antiestrogens for treatment of endocrine-responsive tumors.
Tumors sensitive to estrogen stimulation may regress following
competitive inhibition of estrogen receptors by tamoxifen (an antiestrogen).
Response to tamoxifen is currently predicted based on the stage
of disease and on the basis of assays for estrogen receptors (ER)
and progesterone receptors (PR) in the tissue. Many breast cancers
and all normal estrogen-responsive tissues contain these labile
cytoplasmic proteins which bind estrogen and progesterone. Patients
with positive assays for these proteins have an objective response
to hormone therapy of about 65%, while those with negative assays
have an objective response rate of <10%. For postmenopausal women
having a first recurrence of breast cancer with an ER+ or PR+ assay,
tamoxifen therapy is the treatment of choice.
Notwithstanding strong interest in the use of antiestrogens in
breast cancer treatment, however, incomplete knowledge of their
basic pharmacology persists. Substituted triphenylethylenes (including
tamoxifen) have antiestrogen effects which appear to be partially
dependent on geometric isomerism. For example, trans-tamoxifen (the
isomer used in tamoxifen therapy) is an antiestrogen, whereas the
cis isomer is a weak estrogen. The present invention makes use of
newly acquired knowledge about in vivo interconversion of tamoxifen
geometric isomers as well as the formation of tamoxifen metabolites,
including estrogenic metabolites, to improve treatment of breast
Although tamoxifen is the most widely used antiestrogen for treating
breast cancer, development of tamoxifen resistance and subsequent
tumor progression during tamoxifen therapy represents a major reason
for treatment failures. The mechanism of tamoxifen resistance has
been unknown, but an estrogenic metabolite of tamoxifen which would
promote growth in ER+ tamoxifen-resistant tumors is identified.
Tamoxifen (TAM) is a non-steroidal triphenylethylene antiestrogen
which is commonly used in the treatment of patients with breast
cancer. Tamoxifen competitively antagonizes the binding of estradiol
to estrogen receptors, which is thought to be the mechanism for
inhibiting tumor growth. However, tamoxifen also possesses both
estrogenic as well as antiestrogenic effect. In other tissues such
as human endometrial tissue, and in some species, such as dogs,
tamoxifen acts as an estrogen agonist (Furr, 1984, Ferrazzi, 1977).
The mechanisms underlying the "intrinsic" estrogenic effects
of tamoxifen remain poorly understood. Because the estrogenic agonist/antagonist
properties of tamoxifen are highly species and tissue dependent,
it has been suggested that altered metabolic pathways leading to
the production of estrogenic metabolites may be involved in the
"intrinsic" estrogenic effects noted with tamoxifen.
The systemic metabolism of tamoxifen is characterized in humans.
Pharmaceutically administered trans-tamoxifen is metabolized in
the liver to N-desmethyl-and 4-hydroxytamoxifen (Adam, 1979), for
example. Both of these metabolites are active antiestrogenic agents.
N-desmethyltamoxifen is the major metabolite found in human serum.
It is further metabolized to metabolite Z (didesmethyltamoxifen)
and to metabolite Y, (the side chain alcohol) (Kemp, 1983, Jordan,
1983). At least two estrogenic metabolites of tamoxifen have been
identified, these include metabolite E (monophenol) and bisphenol
(Lyman 1985). Neither metabolite has been identified in any significant
quantity in human serum or tissues. However, the monophenol has
been isolated from the bile of dogs, a species in which tamoxifen
is predominantly estrogenic (Fromson 1973).
It has been shown that tamoxifen can actually stimulate tumors
to grow following chronic dosing in hormone responsive human tumor-bearing
mouse models. The present inventors have shown that reduced uptake
and metabolism of tamoxifen is associated with tamoxifen resistance
and tamoxifen stimulated growth. The possibility that tumor cell
growth is dependent on the net estrogenic agonist/antagonist activity
of cellular compounds, including tamoxifen metabolites at the receptor
site, led the present inventors to further explore whether estrogenic
metabolites are present in neoplastic breast tissues. The presence
of monophenol tamoxifen (metabolite E) in human breast tumor biopsy
specimens is described herein using mass-spectrometry and HPLC identification
SUMMARY OF THE INVENTION
The present invention relates to the levels of tamoxifen isomers
and metabolites as early indicators of tamoxifen resistance (prior
to clinical treatment failure). Tamoxifen is representative of a
group of compounds called triphenylethylene antiestrogens, whose
effect is to slow or stop the growth of estrogen-dependent tumors.
While it is the most commonly used drug for treatment of breast
cancer today, tamoxifen is associated with the development of drug
resistance in virtually all patients who take it. With the onset
of resistance, tumor growth resumes or accelerates and tamoxifen
therapy should be discontinued. Renewed tumor growth will eventually
become obvious, of course, but an objective of the present invention
is to provide early warning of tamoxifen resistance so that therapy
can be changed promptly as needed.
The mechanism of tamoxifen resistance is incompletely described,
but it is known that tamoxifen exists as two geometric isomers,
the trans form (an antiestrogen) and the cis form (a weak estrogen).
The trans form, of course, is the therapeutic drug for breast cancer.
Further, however, it has been shown that there can be interconversion
from one isomer to the other, especially with the 4-OH metabolite,
in tissue culture cells and also, presumably, in patients.
A model with which the present invention was developed consists
of human breast cancer cells growing subcutaneously in athymic nude
mice. Treatment of the mice with tamoxifen results in tumor growth
inhibition for four to six months, followed by the onset of tumor
resistance and regrowth (much as in humans). Further, tumor regrowth
is actually stimulated by tamoxifen. Even when such a tumor is transplanted
into different mice, the resistant tumors will not grow unless stimulated
by tamoxifen or estrogen. Tamoxifen, in these cases, appears to
mimic the action of estrogen.
The development of tamoxifen resistance relates to conversion over
time of both tamoxifen and its metabolites from trans (antiestrogen)
forms to cis forms or conversion of tamoxifen to other metabolites
which have estrogen-like activity. Assays for both trans and cis
forms of tamoxifen as well as tamoxifen metabolites may be accomplished,
for example, with a high performance liquid chromatography (HPLC)
system and mass spectrographic analysis, in some cases. In results
described herein, it is shown that:
1) tumors from tamoxifen-resistant mice have significantly lower
concentrations of tamoxifen than sensitive tumors (whose growth
is still arrested);
2) there is an increase in the cis-4-hydroxytamoxifen/trans-4-hydroxytamoxifen
(C/T-OH-TAM) ratio in resistant tumors; and
3) there is a spectacular decrease in the ratio of tamoxifen to
tamoxifen monophenol metabolite in resistant tumors.
Values of the C/T-OH-TAM ratio are about 0.4-0.5 in sensitive tumors
and 0.8-0.9 in resistant tumors. The increased estrogen effect resulting
from an elevated C/T-OH-TAM ratio may be the reason for development
of resistance, the association appears to be consistent. Analogous
and even more clear correlations are shown between the development
of tamoxifen resistance and the appearance of certain tamoxifen
metabolites such as tamoxifen metabolite E (monophenol). The ratio
(ng/gm tissue) of total tamoxifen to metabolite E (tamoxifen monophenol),
for example, decreases spectacularly as tamoxifen resistance develops.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows HPLC tracings of 400 ng/ml stock solutions of trans
and cis 4-hydroxytamoxifen (4-OH-TAM), respectively.
FIG. 2 shows HPLC chromatographs indicating that the cis/trans
ratio of 4-OH-TAM is lower in sensitive than in resistant tumors
isolated from athymic nude mice.
FIG. 3 shows the approximately linear inverse relationship which
exists between the logarithm of total tumor tamoxifen concentration
and the cis/trans 4-OH-TAM ratio (N=32, R=0.83, linear regression
FIG. 4 shows a scatter plot of tamoxifen concentrations in sensitive
and resistant MCF-7 tumors (N=32).
FIG. 5 shows a scatter plot of 4-OH-TAM ratios (cis/trans) in sensitive
and resistant MCF-7 tumors (N=32).
FIG. 6 shows a scatter plot of tamoxifen versus cis/trans 4-OH-TAM
ratios in resistant tumors (open symbols) and sensitive tumors (solid
FIG. 7 shows HPLC chromatographs of a resistant tumor analyzed
for total tumor tamoxifen, cytosol tamoxifen, and tamoxifen found
in nuclear pellets extracted with KCl (from top to bottom, respectively).
FIG. 8 shows example HPLC chromatographs demonstrating concentrations
of cis- and trans-4-OH-TAM extracted from tumor nuclear pellets
representing patients with clinically-evident tumor resistance.
These findings of greater cis than trans levels must be contrasted
with analogous measurements for tamoxifen-sensitive tumors in the
mouse model (showing trans>cis 4-OH-TAM) because sensitive tumors
in human patients are not biopsied and the information is not available.
FIG. 9 shows example HPLC chromatographs demonstrating detection
of tamoxifen, N-des-tamoxifen, and cis and trans 4-OH-TAM following
treatment with 20 mg tamoxifen per day. These measurements demonstrate
the capability to measure tamoxifen and its metabolites in serum.
FIG. 10 shows structures of trans and cis tamoxifen isomers.
Trans-tamoxifen: R'=--OCH.sub.2 CH.sub.2 N(CH.sub.3).sub.2, R"=--H,
4-hydroxytamoxifen: R'=--OCH.sub.2 CH.sub.2 N(CH.sub.3).sub.2,
N-desmethyltamoxifen: R'=--OCH.sub.2 CH.sub.2 NH(CH.sub.3), R"=--H,
Bisphenol: R'=--OH, R"=--OH,
Monophenol: R'=--OH, R"=--H,
Metabolite Y: R'=--CH.sub.2 CH.sub.2 OH, R"=--H,
Metabolite Z: R'=--OCH.sub.2 CH.sub.2 NH.sub.2, R"=--H
Metabolite B.sub.x : R"=--OCH.sub.2 CH.sub.2 NHCH.sub.3, R"=--OH,
cIS-TAMOXIFEN: R'=--OCH.sub.2 CH.sub.2 N(CH.sub.3).sub.2, R"=--H,
4-Hydroxytamoxifen; R'=--OCH.sub.2 CH.sub.2 N(CH.sub.3).sub.2, R"=--OH,
monophenol; R'=--OH, R"=--H.
FIG. 11 shows separation of tamoxifen and metabolites in a spiked
plasma sample (FIG. 11A) and in a tamoxifen resistant human breast
tumor (FIG. 11B). Peaks in order are; bisphenol (a), monophenol
(b), cis 4-hydroxytamoxifen (c), trans 4-hydroxytamoxifen (d), metabolite
Bx (e), tamoxifen (f), metabolite Z (g) and N-desmethyltamoxifen
FIG. 12 shows electron-ionization (+) mass-spectrometric result
of a peak isolated by HPLC fractionation of tamoxifen resistant
human breast tumor (retention time=11-12 min). The peak at 300 m/z
confirms the presence of monophenol tamoxifen (MW=300).
FIG. 13 shows HPLC chromatograms showing tamoxifen and metabolites
in (FIG. 13A) MCF-7 tamoxifen resistant tumor grown in a nude mouse
following tamoxifen administration (500 ug/d.times.6 mos) and (FIG.
13B) in a plasma sample spiked with bisphenol (a), monophenol (b),
cis 4-hydroxytamoxifen (c), trans-4-hydroxytamoxifen (d), tamoxifen
(e) and N-desmethyltamoxifen (f).
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
Those skilled in the art will recognize that changes analogous
to those detected in the C/T-OH-TAM ratio or TAM/monophenol ratio
may occur for other tamoxifen metabolites (e.g. particularly estrogenic
metabolites such as bisphenol) or in other triphenylethylene antiestrogens
related to tamoxifen. Nevertheless, the best mode of implementing
the present invention now appears to be in conjunction with treatment
of estrogen-responsive breast cancer with tamoxifen and detection
of the onset of tamoxifen resistance.
Those skilled in the art will also recognize that if geometric
isomerization or metabolism to other estrogenic metabolites proves
to be the mechanism of antiestrogen resistance, synthesis of antiestrogens
which preclude isomerization or metabolization will be a fruitful
path to improved treatment. At the present time, however, tamoxifen
is the drug of choice for hormonal treatment of recurrent breast
cancer, and its invariable association with the onset of tumor resistance
and tamoxifen-enhanced tumor growth makes the present invention
a valuable aid to therapy.
Claims of the present invention are directed to indicators discovered
to be closely associated with the onset of tumor tamoxifen resistance.
Elevation of the C/T ratio for 4-OH-TAM, reduction in the TAM/monophenol
ratio in tumor blood or other body fluid, and depression of the
tamoxifen concentration in tumor are, for example, shown to be useful
predictors of recrudescence of tumor growth despite (and perhaps
in-part because of) tamoxifen treatment. The elevation of tamoxifen
metabolite E(monophenol) concentrations, particularly with respect
to tamoxifen concentrations is a preferred useful predictor. Other
relevant ratios for such determinations include levels of estrogenic
tamoxifen metabolites as compared to tamoxifen or any antiestrogenic
tamoxifen metabolite. These indicators can be used individually
For each indicator, small changes (for example, about a few percent)
in the direction indicated by this specification and claims show
early stages of tumor resistance. Larger indicator changes (up to
about 80 to 100 percent, for example) are expected when tumor resistance
is well established. Examples of in vitro and in vivo tests are
provided to demonstrate the specific utility of the claimed invention
and to support both the approximate magnitude of indicator changes
and the reasonable extrapolations suggested in this specification.
Those of skill in the art recognize that there are many methods
of assay for tamoxifen, tamoxifen isomers and tamoxifen metabolites,
particularly estrogenic ones. The development of antibodies, especially
monoclonal antibodies, having specificities for tamoxifen and tamoxifen-related
compounds will lead to ready assay methods of perhaps even greater
Examples presented herein show that acquired tamoxifen resistance
is associated with reduced tamoxifen concentrations and increasing
cis- to trans-4-hydroxytamoxifen ratios. The identification of monophenol
tamoxifen metabolite in a tamoxifen resistant human breast tumor
represents the first evidence that an estrogenic metabolite is also
present in these tissues following prolonged tamoxifen treatment.
Reduced accumulation of antiestrogenic compounds coupled with production
of estrogenic metabolites may result in a net estrogenic environment
within hormonally sensitive breast cancer cells. This observation
explains why at least some tamoxifen resistant tumors are stimulated
to grow by tamoxifen in various animal models. Further evidence
of tamoxifen stimulated growth is seen in patients who are progressing
on tamoxifen and respond to the discontinuation of tamoxifen with
no other therapy. This also suggests that tamoxifen is directly
or indirectly behaving like an estrogen agonist.
Those skilled in the art know that tamoxifen metabolites and isomers
may be detected in tumor tissue or other fluids such as blood leukocytes,
blood serum or blood plasma, for example.
These examples are presented to describe preferred embodiments
and utilities of the present invention and are not meant to limit
the present invention unless otherwise specified in the claims appended
hereto. Taken together, the examples illustrate the best mode of
implementing the invention as it is currently understood.
Mouse MCF-7 Tumor Model
In order to examine the relationship between tamoxifen metabolites
and tamoxifen resistance, athymic Balb/c nude mice were implanted
with human breast cancer tumor cells (MCF-7). Mice were treated
daily for 4-6 months with subcutaneous tamoxifen. After tamoxifen
resistance developed and tumor progression was observed, mice were
sacrificed, and tumors were collected. In addition, tamoxifen-sensitive
tumors were collected from mice still responding to tamoxifen therapy.
Quantification of tamoxifen and its metabolites in tumor specimens
was performed using photactivation and high performance liquid chromatography
(HPLC) analysis. To summarize: Each tumor sample was weighed, spiked
with an internal standard, and homogenized. Tissue samples were
then extracted with 2% n-butanol in hexane and irradiated with high
intensity ultraviolet light (254 nm). Samples were analyzed by HPLC
using a C-18 reverse phase column and eluted isocratically with
a mobile phase of water (7%) and triethylamine (0.25%) in methanol.
All solvents were of HPLC grade. Fluorescence of all compounds was
measured at a wavelength of 266 nm. The sensitivity of this assay
was approximately 2 ngs/gm for trans-tamoxifen, n-desmethyltamoxifen
and cis and trans-4-hydroxytamoxifen. Linearity was measured through
a concentration range of 25-3,000 ngs for all compounds, with a
correlation coefficient of greater than 0.993. Cis and trans-4-hydroxytamoxifen
ratios were calculated based on peak-heights. In some samples, the
tumor was first homogenized and then the nuclear pellet obtained
by centrifugation. HPLC tamoxifen measurements were made on extracts
of the entire tumor specimen or on subcellular fractions including
the cytosol or-the total nuclear pellet. In some cases, the nuclear
pellet was further extracted in 0.4M KCl to extract primary metabolites
bound to estrogen receptor.
Chromatographs (Mouse Studies)
HPLC tracings (see FIG. 1) are shown of 400 ng/ml stock solutions
of trans and cis 4-OH-TAM isomers mixed so as to yield a solution
with equal amounts, as well as a solution in which the trans isomer
predominated and one in which the cis isomer predominated. These
tracings show that specimens may be analyzed to determine relative
amounts of the geometric isomers present, thus providing a means
of easily following a patient's response to tamoxifen therapy using
a method of the present invention.
FIG. 2 shows HPLC tracings indicating relative amounts of trans
and cis 4-OH-TAM in sensitive and resistant tumors isolated from
tamoxifen-treated athymic nude mice. These figures demonstrate identification
of cis and trans 4-OH-TAM in tamoxifen-resistant and sensitive tumors.
Further, they demonstrate that the cis/trans ratio is higher in
the resistant tumor.
FIG. 3 shows the approximately linear inverse relationship of the
logarithm of total tumor tamoxifen concentration to cis/trans 4-OH-TAM
ratios (N=32, R=0.83, linear regression analysis). This graph suggests
that because total tumor tamoxifen concentration is related to the
geometric isomer ratio, only one of these quantities may be needed
to predict emergence of tumor resistance. Because determination
of the isomer ratio does not require absolute concentrations, it
may be a more convenient measure for clinical evaluation of tamoxifen-treated
FIG. 4 shows a scatter plot of tamoxifen concentrations in sensitive
and resistant MCF-7 tumors (N=32). Mean tamoxifen concentrations
in sensitive tumors is 36 ug/gm (SD=46.41), versus 4.38 ug/gm in
resistant MCF-7 tumors (SD=6.22). These differences are highly statistically
significant (P<0.0005) by the Wilcoxon Rank Sum test. See Table
1 for Data.
The median tamoxifen concentration in sensitive tumors is 12.475
ug/gm and in resistant tumors it is 2.2 ug/gm. 10 of 16 sensitive
tumors had values greater than 10 ug/gm compared to only 2 of 16
resistant tumors. This graph implies that while individual tamoxifen
concentrations might not be determinative of tamoxifen resistance,
a trend in such numbers for a particular patient could be helpful
in making therapeutic decisions.
TABLE 1 ______________________________________ Tamoxifen Concentrations
in Sensitive and Resistant MCF-7 tumors. Tamoxifen (ug/gm) (N =
16) Sensitive (N = 16) Resistant ______________________________________
1. 101.55 1.82 2. 84.47 25.12 3. 10.76 6.93 4. 159.24 3.3 5. 80.94
2.58 6. 12.69 6.08 7. 50.7 0.745 8. 15.36 0.688 9. 0.766 3.348 10.
2.78 0.45 11. 8.85 1.64 12. 8.08 0.814 13. 12.26 11.12 14. 9.45
1.66 15. 14.4 2.81 16. 3.79 0.981 Mean 36 ug/gm 4.38 ug/gm p <
0.0005 SD = 46.41 ug/gm SD = 6.22 ug/gm Range = 0.77-159.24 Range
= 0.45-25.12 10/16 are > 10 ug/gm 2/16 > 10 ug/gm median =
12.475 ug/gm median = 2.2 ug/gm ______________________________________
FIG. 5 shows a scatter plot of cis/trans 4-OH-TAM ratios in sensitive
and resistant MCF-7 tumors (N=32). The mean 4-OH-TAM cis/trans ratio
in sensitive tumors was 0.5297 (SD=0.168), whereas in resistant
MCF-7 tumors the mean ratio was 0.8141 (SD=0.196). These differences
are again highly statistically significant (P<0.0005). See Table
2 for cis/trans 4-OH-TAM data.
By reasoning similar to that for FIG. 4, this graph indicates that
while individual cis/trans ratio determinations might not be determinative
of tamoxifen resistance, a trend in such numbers for a particular
patient would be helpful in making therapeutic decisions.
TABLE 2 ______________________________________ Cis/trans 4-OH-TAM
ratios in sensitive and resistant MCF-7 tumors. Cis/Trans 40H-Tamoxifen
Sensitive (N = 16) Resistant (N = 16) ______________________________________
1. 0.575 0.86 2. 0.588 0.812 3. 0.472 0.442 4. 0.56 0.929 5. 0.645
1.019 6. 0.428 0.777 7. 0.446 1.073 8. 0.4632 0.49 9. 0.476 0.724
10. 0.506 0.868 11. 0.34 0.59 12. 0.486 0.86 13. 0.34 0.73 14. 0.38
1.16 15. 1.0 0.76 16. 0.77 0.932 Mean 0.5297 0.8141 p < 0.0005
SD = 0.168 SD = 0.196 Range = (0.34-1.0) Range = (0.442-1.16) Median
= 0.48 Median = 0.84 3/16 > 0.06 3/16 .ltoreq. 0.60 13/16 .ltoreq.
0.60 13/16 > 0.60 ______________________________________
FIG. 6 shows the scatter plot of FIG. 3 except for removal of the
regression line and addition of identification for resistant (open
symbols) and sensitive (solid symbols) tumors. This scatter plot
demonstrates the relationship among cis/trans ratios, total tumor
tamoxifen concentration, and tamoxifen resistance. Resistant tumors
have lower tamoxifen concentrations and higher cis/trans ratios.
Nevertheless, this graph like FIGS. 3, 4, and 5 further indicates
that trend data are even more useful in therapeutic decisions for
a given patient than single determinations of either tamoxifen concentration
or the cis/trans ratio.
Table 3 shows cis/trans ratios and tamoxifen concentrations in
nuclear extractions versus nuclear pellets extracted with 0.4 molar
KCl followed by 200,000 G centrifugation at 4.degree.. The material
extracted with KCl includes that bound to estrogen receptor. The
results of this experiment suggest that tamoxifen-sensitive tumors
generally have a higher percentage of the major anti-estrogen metabolite,
trans 4-OH-TAM. In addition, tamoxifen resistant tumors generally
have reduced tamoxifen concentrations, together with high levels
of the less antiestrogenic and more estrogenic metabolite cis 4-OH-TAM.
Therefore, levels of geometric isomers of the 4-OH-TAM metabolite,
together with tamoxifen levels, may indicate tumor sensitivity.
TABLE 3 ______________________________________ Study #4 Tumors
2 to 13 Cis/Trans Ratios - Tamoxifen Levels - R/S tamoxifen level
tumor # cis/trans ratio (conc/ g) S/R ______________________________________
2 Total Nuclear 0.6846 3,440.28 sensitive 2 KCl Nuclear Extract
trans only 664.26 3 Total Nuclear 1.0455 2,568.63 sensitive 3 KCl
Nuclear Extract -- 825.35 4 Total Nuclear 0.4620 -- resistant 4
KCl Nuclear Extract 0.9195 798.48 5 Total Nuclear 0.6632 4,282.17
sensitive 5 KCl Nuclear Extract trans only 412.68 6 Total Nuclear
1.0209 796.78 resistant 6 KCl Nuclear Extract 1.9248 157.96 7 Total
Nuclear 0.9401 713.43 resistant 7 KCl Nuclear Extract 1.3808 148.67
8 Total Nuclear 1.0059 1,891.92 sensitive 8 KCl Nuclear Extract
-- 454.25 9 Total Nuclear 0.3808 14,238.85 sensitive 9 KCl Nuclear
Extract 0.9351 1,330.48 10 Total Nuclear 0.8530 1,139.24 resistant
10 KCl Nuclear Extract 2.2606 228.21 11 Total Nuclear 0.4924 8,757.46
sensitive 11 KCl Nuclear Extract trans only 950.92 12 Total Nuclear
1.0904 817.98 resistant 12 KCl Nuclear Extract 2.0029 193.33 13
Total Nuclear 0.4367 8,062.12 sensitive 13 KCl Nuclear Extract trans
only 1,667.69 ______________________________________ KCl Nuclear
Extract fraction equals nuclear pellet extracted with 0.4 molar
KCl followed by 200,000 G's.
Four of seven sensitive tumors had trans only in the KCl nuclear
extract, in two the levels were below the limit of detectability
and in one the C/T ratio was 0.9. All 5 resistant tumors had high
C/T ratios. Total and KCl extracted tamoxifen was lower in resistant
than in sensitive tumors.
FIG. 7 shows an HPLC chromatograph of a resistant tumor analyzed
for total tumor tamoxifen, cytosol tamoxifen, and tamoxifen found
in nuclear pellets extracted with KCl (from top to bottom, respectively).
This figure shows that cis 4-OH-TAM/trans-4-OH-TAM ratios are larger
in the nuclear pellets extracted with KCl, which presumably represents
estrogen receptor-bound drug. This observation indicates that tamoxifen
resistance is related to the production of the cis 4-OH-TAM metabolite.
To further demonstrate that the invention is applicable in humans,
several studies were carried out which demonstrate its feasibility.
1) identification and measurement of tamoxifen and its metabolites
in tumors isolated from tamoxifen-treated patients who are known
2) measurement of tamoxifen and its metabolites in serum from patients
who are being treated with adjuvant doses of tamoxifen;
3) determination of cis/trans 4-OH-TAM ratios in treatment-resistant
Table 4 shows tamoxifen concentrations and cis/trans ratios in
14 human tumor specimens from patients who were on tamoxifen therapy.
There are two tamoxifen-sensitive tumors (L281 and G375) for comparison
because patients responding to treatment are not usually biopsied
during the responsive phase of their disease. The majority of patients
whose tumors were tested had relatively high cis/trans ratios in
the total tumor extract, as were noted with the MCF-7 mouse model.
That is, tamoxifen-resistant human tumors have significant concentrations
of the weakly antiestrogenic metabolite, cis 4-OH-TAM, a metabolite
which may also have weak estrogenic activity. The two tumors from
tamoxifen-sensitive patients, as expected, have low cis/trans ratios.
TABLE 4 ______________________________________ Tamoxifen/Metabolites
in Total Tumor, Cytosol, and Nuclear Pellets Extracted with 0.4
molar KCl 40H 40H sample cis trans patient Tamoxifen peak peak cis/trans
# weight (mg) peak height height height ratio ______________________________________
W 131 total 41 57789 4587 3394 1.35 cytosol 97 18504 1724 902 1.91
pellet 97 ni 1496 1024 1.46 N 184 total 38 1567 ni ni 1.42 cytosol
88 183 358 nd cis only U 243 total 34 44604 ni ni 1.23 cytosol 77
9067 1544 1021 1.51 pellet 77 5600 ni ni 0.97 L 281* total 44 39960
3324 12216 0.27 cytosol 80 7369 1389 3787 0.37 pellet 80 4795 1154
1194 0.97 H 312 total 33 39726 3574 7749 0.46 cytosol 87 15887 ni
ni 1.38 pellet 87 3234 ni ni 1.18 G 375* total 49 17521 1911 3373
0.57 cytosol 44 8429 613 723 0.85 pellet 44 2523 ni ni 1.29 F 462
total 45 598 ni 3968 trans only cytosol 93 301 431 1102 0.39 U 465
total 32 8823 ni 4287 trans only total 125 6672 640 1703 0.38 pellet
125 2248 508 nd cis only T 474 total 44 11987 ni 1481 trans only
total 113 3129 483 741 0.65 pellet 113 2737 311 nd cis only Y 501
total 48 4085 ni ni 1.19 cytosol 71 1288 206 nd cis only K 532 total
36 8850 ni nd -- total 85 3410 669 324 2.06 pellet 85 1131 305 nd
cis only I 559 total 65 ni ni nd -- cytosol 89 243 456 nd cis only
J 562 total 32 21965 3924 nd cis only cytosol 64 6656 1291 784 1.48
pellet 64 161 ni ni 1.45 J 578 total 36 ni ni nd -- cytosol 51 553
408 nd cis only ______________________________________ nd = not
detectable ni = no integreation (ratio obtained by direct measurement
total = total tumor cytosol = total tumor except for the nucleus
pellet = KCl nuclear extract *tamoxifen sensitive tumors
Results of these studies show that concentrations of tamoxifen
and its metabolites measured in serum, tumors, and nuclear fractions
can be used to follow the response of human tumor patients. Relative
increases in the cis/trans ratio and decreases in tamoxifen concentration,
for example, are expected to be associated with patients whose tumors
are progressing on adjuvant doses of tamoxifen. FIG. 8 shows an
example of higher cis 4-OH-TAM in the nuclear pellets (extracted
with 0.4 molar KCl, followed by ultracentrifugation, as in the mouse
Human Serum Specimens
Because directly monitoring tumor drug levels in patients requires
a biopsy for each reading, it may be impractical. Using a modified
assay method, however, detection of tamoxifen, desmethyl-tamoxifen,
and cis and trans 4-OH-TAM in serum is possible. FIG. 9 shows serum
specimens following 20 mg daily doses of tamoxifen. Each serum specimen
was spiked with cis and trans 4-OH-TAM reference standards which
marked the appropriate peak in the unspiked sample. However, the
relationship between plasma and tumor levels of the compounds of
interest remains undetermined. Analogous results were obtained when
a 160 mg tamoxifen dose was used. Metabolite E and bisphenol tamoxifen
can also be detected in serum or plasma with this method. Measurement
of the parent drug and its metabolites in peripheral blood leukocytes
may increase sensitivity of the assay.
Indicative Tamoxifen Metabolites
MATERIALS AND METHODS
The chemical structures of trans- and cis-tamoxifen and tamoxifen
metabolites are shown in FIG. 10. Tamoxifen citrate (ICI Pharma)
and its metabolites (cis- & trans-4-hydroxytamoxifen, bis and
monophenol tamoxifen) were generously provided by Professor Katzenellenbogen.
All reagents were of HPLC grade including; methanol, hexane, 1-butanol
(Fisher Scientific, Fair Lawn, N.J., and triethylamine (Sigma Chemical
Co., St. Louis, Mo.).
Human breast tumors from various patients who failed tamoxifen
therapy (20 mgs/day>one year) were obtained from the Breast Cancer
Tumor Bank. A portion of the tumor was extracted, irradiated and
analyzed by Mass Spectrometry (procedure described below). Tumor
tissue was also obtained from an in vivo nude mouse model of acquired
tamoxifen resistance. MCF-7 human breast tumors were harvested from
nude mice after the development of tamoxifen resistance (4-6 months
of tamoxifen administration, 500 ug/day). These tissues were frozen
at -20.degree. C. until analysis. Frozen tissue samples were weighed,
homogenized, and extracted. Briefly, all samples were extracted
with 6.0 ml of 2% butanol in hexane, vortexed for 1.0 min, and then
centrifuged for 10 min at 1000.times.g. The organic phase was dried
under N.sub.2 gas at 37.degree. C. and reconstituted in 200 .mu.l
methanol prior to injection. The reconstituted samples were transferred
to a Infrasil quartz cuvette (Fisher Scientific), irradiated for
one minute with a 15 W Hg vapor lamp and injected onto a HPLC column.
The HPLC system consisted of a Beckman model 320 gradient liquid
chromatograph, two model 110A pumps, and a model 420 controller.
The HPLC was equipped with a reverse phase Altex C18 ultrasphere
ODS column and a 100 .mu.l injection loop. The mobil phase consisted
of 7% H.sub.2 O and 0.18% triethylamine in methanol. The flow rate
of the mobil phase was set at 0.5 ml/min. The fluorescence of photochemically
activated compounds was detected with an Applied Biosystems 980
fluorometer with excitation wavelength set at 266 nm. Retention
times and peak heights were recorded with a Spectraphysics 4100
Non-irradiated samples were also injected onto the HPLC column
and fractions were collected every minute. Fractionations corresponding
to the retention times noted for the bisphenol and monophenol tamoxifen
were further evaluated using mass-Spectrometry. The mass spectrometric
behavior of the mono and bis phenols under electron impact ionization
mass-spectrometry (EI-MS) was utilized. HPLC separated samples,
in glass vials, were extracted with methanol, injected into glass
capillaries and dried. Samples were introduced via a direct insertion
probe. Spectra were recorded at 70 eV over the mass range of 70-500
at 10 seconds per decade with a resolution of 1000. The probe was
heated in stages to 200.degree. C.
HPLC analysis of a human breast cancer tissue specimen and a spiked
plasma sample is shown in FIG. 11. In the human breast cancer biopsy
specimen, peaks corresponding to the retention times of bisphenol,
monophenol, cis-4-hydroxy, trans 4-hydroxy, metabolite Bx, tamoxifen,
metabolite Z and N-desmethyltamoxifen were observed. Tamoxifen and
its N-desmethyl metabolite were the most abundant compounds present
(200-400 mg/gm). Peaks were also noted for metabolites Bx and Z,
however, these were not quantified. Peaks corresponding to bisphenol,
monophenol, cis 4-hydroxy, metabolite Y and trans 4-hydroxy tamoxifen
were present in low concentrations (<50 ng/gm), but due to the
close proximity of their retention times absolute identification
of these peaks required mass spectrometry analysis.
Further confirmation of the bisphenol and monophenol HPLC fractionations
were done with mass spectrometry. Mass spectrometry analysis confirmed
the presence of monophenol (FIG. 12). The HPLC peak at 11-12 minutes
corresponded to a peak at m/z 300, consistent with the molecular
weight of monophenol tamoxifen metabolite. In addition to the peak
at m/z 300 (the monophenol), a strong peak at 149 m/z was observed.
This was probably derived from phthlate ester impurities in the
sample. Although a small but clear peak co-migrates with bisphenol
on HPLC, the relative abundance of smaller weight compounds (<150)
obscured the mass spectra analysis of these peaks. Therefore, the
presence of bisphenol could not be confirmed in this study.
Results from HPLC analysis of MCF-7 tumors isolated from mice following
tamoxifen administration are shown in FIG. (13). The top chromatogram
shows results of the MCF-7 tumor analysis while the bottom shows
a 25 ng/ml plasma sample spiked with bisphenol (A), monophenol (B),
cis 4-hydroxy (C), trans 4-hydroxy (D), tamoxifen (E), and N-desmethytamoxifen
(F). Peaks with retention times corresponding to monophenol, cis
4-hydroxy, trans 4-hydroxy, tamoxifen and N-desmethyltamoxifen were
noted in the MCF-7 tamoxifen resistant mouse tumor. Metabolite Bx
was also noted on the chromatogram (RT=17-19 min) but was not quantified.
N-desmethylation and hydroxylation were the primary metabolic routes
noted. Concentrations of tamoxifen and its metabolites were as follows:
tamoxifen=3,255.9 ng/gm, N-desmethyltamoxifen=135.4 ng/gm cis 4-hydroxytamoxifen=42
ng/gm and trans 4-hydroxytamoxifen=36 ng/gm. Similar to the results
noted in the human tumor specimen, the MCF-7 tumor also had peaks
corresponding to monophenol (22 ng/gm). A very small peak was evident
at a retention time similar to bisphenol.
In Table 5, levels of tamoxifen and a variety of tamoxifen metabolites
are shown for tamoxifen-sensitive and tamoxifen-resistant tumors.
TABLE 5 ______________________________________ Sensitive and Resistant
MCF-7 Mouse Tumors: Tamoxifen and Metabolite Concentrations (.eta.g/g
tissue) Sensitive Tumors: Animal N- Cis Trans Mono- Bis- # TAM DES
4-OH 4-OH phenol phenol ______________________________________ 1
3951.20 124.70 87.67 99.91 7.69 -- 4 26926.12 331.82 139.89 337.58
109.07 44.27 7 26304.72 534.62 122.59 244.61 160.99 33.90 8 2339.52
5.95 70.83 71.01 37.72 39.80 13 20104.46 322.99 132.13 117.44 265.25
1491.44 15 15057.24 359.67 71.26 105.23 137.93 352.76 N 6 6 6 6
6 5 X 15780.54 279.95 104.06 162.63 119.78 392.43 SD 10721.90 187.05
31.19 104.87 92.31 629.19 range 2339.52- 5.95- 70.83- 71.01- 7.69-
33.90- 26926.12 534.62 139.89 337.58 265.25 1491.44 3 250.46 19.65
28.01 22.50 -- 15.12 5 709.92 68.12 89.59 79.78 24.30 18.34 6 400.13
38.44 43.33 35.04 9.63 4.26 9 3255.87 135.44 42.32 35.46 21.99 3.93
10 402.59 50.27 55.37 47.55 -- 32.75 11 710.41 65.70 51.94 53.08
20.66 9.77 12 310.19 -- 45.04 31.66 232.66 43.87 14 775.66 127.29
103.44 97.52 47.82 52.72 16 2836.88 100.43 78.46 59.73 27.13 23.38
N 10 9 10 10 8 10 X 998.40 71.28 58.36 49.91 49.35 21.33 SD 1099.60
41.17 24.03 23.47 75.00 16.84 range 250.46- 19.65- 28.01- 22.50-
9.63- 3.93- 3255.87 135.44 103.44 97.52 232.66 52.72 ______________________________________
From the data in Table 5 it should be noted that the tamoxifen/tamoxifen
monophenol ratio (TAM/monophenol) is about 132 in tamoxifen-sensitive
tumors and about 20 in tamoxifen-resistant tumors. The proportion
of this estrogenic tamoxifen metabolite level to total TAM level
shows a major increase in resistant tumors.
In patients receiving tamoxifen therapy, acquired resistance to
tamoxifen generally occurs after six months. Several mechanisms
of resistance have been suggested including hormone independence
and altered metabolic pathways (Jordan). The present example documents
that an estrogenic metabolite, monophenoltamoxifen, is present in
tumors isolated from patients who failed tamoxifen therapy. In addition,
bisphenoltamoxifen, a potent estrogenic metabolite, is also present
in these patients. These metabolites are, for example, detected
in human breast tumor tissue as well as in MCF-7 tumors isolated
The identification of estrogenic metabolites in tissues following
long-term tamoxifen administration has important clinical ramifications.
Both bisphenol and monophenol tamoxifen have documented estrogenic
activity in vitro and in vivo. Metabolite E (monophenol) is also
known to exit in isomeric form. Isomerization of transmetabolite
E to cis metabolite E has been demonstrated in vitro (Murphy, 1990).
The structure activity relationship of these isomers has also been
examined. Trans-metabolite E appears to be a weak partial agonist,
while the cis isomer is a potent estrogen agonist (Murphy, 1990).
The cis form of metabolite E has also been shown to be more potent
than bisphenol in stimulating T47D human breast cell growth (Murphy
1990). Using GC-MS, Fromson et al. identified metabolite E in the
bile fluid of dogs where it accounted for approximately 6% of the
radiolabeled compounds present in bile (Fromson 1973). Murphy et
al. tentatively identified the presence of cis-metabolite E using
GC-MS in the plasma of patients receiving tamoxifen (Murphy 1987).
They report that cis-metabolite E was present in low concentration
(0.9-2.0 ng/ml) in patients given tamoxifen for 14 days, while concentrations
of 2.8 and 7.0 ng/ml were reported in two chronically (>2 yrs)
treated patients (Murphy 1987).
Furr B. J. A. and Jordan V. C. (1984), "The Pharmacology and
Clinical Uses of Tamoxifen," Pharmacol Ther., 25:127.
Ferrazzi et al. (1977), "Oestrogen-like effect of tamoxifen
on vaginal epithelium," Br. Med. J., 1:1351.
Adam et al. (1979), "The metabolism of tamoxifen in humans,"
Biochem Pharmacol. 27:145.
Kemp et al. (1983), "Identification and biological activity
of tamoxifen metabolites in human serum," Biochem. Pharmacol.,
Jordan et al. (1983) "Determination and pharmacology of a
new hydroxylated metabolite of tamoxifen observed in patient sera
during therapy for advanced breast cancer," Cancer Res., 43:1446
Lyman et al. (1985), "Metabolism of tamoxifen and its uterotrophic
activity," Biochem Pharmacol, 34:2787.
Fromson et al. (1973), "The metabolism of tamoxifen (ICI 46,474).
I. in laboratory animals," Xenobiotica, 3:693.
Sutherland et al. (1982), "Mechanisms of oestrogen antagonism
by nonsteroidal antiestrogens," Molec. Cell Endocr., 25:5-23.
Osborne et al., "Acquired tamoxifen resistance correlates
with reduced tuor tamoxifen and trans- 4hydroxytamoxifen in human
breast cancer," JNCI, 1990 (in press).
Coezy et al. (1982), "Tamoxifen and metabolites in MCF-7 cells:
correlation between binding to estrogen receptor and inhibition
of cell growth," Cancer Res., 42:317-323.
Jordan et al. (1985), "Structure requirements for the pharmacological
activity of nonsteroidal antiestrogens in vitro," Mol Pharmacol.,
Murphy et al. (1990), "Structure-Function relationships of
hydroxylated metabolites of tamoxifen that control the proliferation
of estrogen-response T47D breast cancer cells in vitro," Molec.
Murphy et al. (1987), "Analysis of tamoxifen and its metabolites
in human plasma by gas chromatography mass spectrometry (GC-MS)
using selected ion monitoring (SLM)," J. Steroid Biochem.,
Changes may be made in the construction, operation and arrangement
of the various parts, elements, steps and procedures described herein
without departing from the concept and scope of the invention as
defined in the following claims. For example, it is understood that
many methods for measuring tamoxifen isomers and metabolites may
be used. Additionally after ratios, particularly those reflecting
relative increases in estrogenic tamoxifen isomers and metabolites
as compared to antiestrogenic analogs are equivalent. Of course
these measurements and ratios may be found in body fluids where
tamoxifen metabolites congregate, e.g., biopsy specimens, lymph
fluid, blood or even urine.