Tumor tissue levels of Tissue Inhibitor of Metalloproteinases-1 (TIMP-1) and outcome following adjuvant chemotherapy in premenopausal lymph node-positive breast cancer patients: A retrospective study.

Abstract

We have previously demonstrated that high tumor tissue levels of TIMP-1 are associated with no or limited clinical benefit from chemotherapy with CMF and anthracyclines in metastatic breast cancer patients. Here, we extend our investigations to the adjuvant setting studying outcome after adjuvant chemotherapy in premenopausal lymph node-positive patients. We hypothesize that TIMP-1 high tumors are less sensitive to chemotherapy and accordingly that high tumor tissue levels are associated with shorter survival.
From our original retrospectively collected tumor samples we selected a group of 525 pre-menopausal lymph node-positive patients (adjuvant treatment: CMF, 324 patients; anthracycline-based, 99 patients; no adjuvant chemotherapy, 102 patients). TIMP-1 levels were measured using ELISA in cytosolic extracts of frozen primary tumors. TIMP-1 was analyzed as a continuous variable and as a dichotomized one using the median TIMP-1 concentration as a cut point between high and low TIMP-1 groups. We analyzed the benefit of adjuvant CMF and anthracyclines in univariate and multivariable survival models; endpoints were disease-free (DFS) and overall survival (OS).
In this selected cohort of high-risk patients, and in the subgroup of patients receiving no adjuvant therapy, TIMP-1 was not associated with prognosis. In the subgroup of patients treated with anthracyclines, when analyzed as a continuous variable we observed a tendency for increasing TIMP-1 levels to be associated with shorter DFS (multivariable analysis, HR 1.75, 95% CI 1.00-3.07, P = 0.05) and a significant association between increasing TIMP-1 and shorter OS in both univariate (HR 3.52, 95% CI 1.54-8.06, P = 0.003) and multivariable analyses (HR 4.19, 95% CI 1.67-10.51, P = 0.002). No statistically significant association between TIMP-1 and DFS was observed in the CMF-treated patients although high TIMP-1 was associated with shorter OS when analyzed as a dichotomized variable (HR 1.64, 95% CI 1.02-2.65, P = 0.04).
In the subgroup of patients receiving adjuvant chemotherapy we found an association between shorter survival after treatment in TIMP-1 high patients compared with TIMP-1 low patients, especially in patients receiving anthracycline-based therapy. This suggests that high tumor tissue levels of TIMP-1 might be associated with reduced benefit from classical adjuvant chemotherapy. Our findings should be validated in larger prospective studies.

Full text

Page 1

BioMed Central
Page 1 of 13
(page number not for citation purposes)
BMC Cancer
Open Access
Research article
Tumor tissue levels of Tissue Inhibitor of Metalloproteinases-1
(TIMP-1) and outcome following adjuvant chemotherapy in
premenopausal lymph node-positive breast cancer patients: A
retrospective study
Anne-Sofie Schrohl*1, Maxime P Look2, Marion E Meijer-van Gelder2,
John A Foekens2 and Nils Brünner1
Address: 1University of Copenhagen, Faculty of Life Sciences, Department of Veterinary Disease Biology, Ridebanevej 9, DK-1870 Frederiksberg C,
Denmark and 2Erasmus MC, Department of Medical Oncology, Josephine Nefkens Institute and Cancer Genomics Center, Dr. Molewaterplein 50,
3015 GE Rotterdam, the Netherlands
Email: Anne-Sofie Schrohl* - sofie@life.ku.dk; Maxime P Look - m.look@erasmusmc.nl; Marion E Meijer-van
Gelder - m.e.meijer@erasmusmc.nl; John A Foekens - j.foekens@erasmusmc.nl; Nils Brünner - nbr@life.ku.dk
* Corresponding author
Abstract
Background: We have previously demonstrated that high tumor tissue levels of TIMP-1 are associated with no or
limited clinical benefit from chemotherapy with CMF and anthracyclines in metastatic breast cancer patients. Here, we
extend our investigations to the adjuvant setting studying outcome after adjuvant chemotherapy in premenopausal lymph
node-positive patients. We hypothesize that TIMP-1 high tumors are less sensitive to chemotherapy and accordingly that
high tumor tissue levels are associated with shorter survival.
Methods: From our original retrospectively collected tumor samples we selected a group of 525 pre-menopausal lymph
node-positive patients (adjuvant treatment: CMF, 324 patients; anthracycline-based, 99 patients; no adjuvant
chemotherapy, 102 patients). TIMP-1 levels were measured using ELISA in cytosolic extracts of frozen primary tumors.
TIMP-1 was analyzed as a continuous variable and as a dichotomized one using the median TIMP-1 concentration as a
cut point between high and low TIMP-1 groups. We analyzed the benefit of adjuvant CMF and anthracyclines in univariate
and multivariable survival models; endpoints were disease-free (DFS) and overall survival (OS).
Results: In this selected cohort of high-risk patients, and in the subgroup of patients receiving no adjuvant therapy, TIMP-
1 was not associated with prognosis. In the subgroup of patients treated with anthracyclines, when analyzed as a
continuous variable we observed a tendency for increasing TIMP-1 levels to be associated with shorter DFS (multivariable
analysis, HR 1.75, 95% CI 1.00-3.07, P = 0.05) and a significant association between increasing TIMP-1 and shorter OS in
both univariate (HR 3.52, 95% CI 1.54-8.06, P = 0.003) and multivariable analyses (HR 4.19, 95% CI 1.67-10.51, P = 0.002).
No statistically significant association between TIMP-1 and DFS was observed in the CMF-treated patients although high
TIMP-1 was associated with shorter OS when analyzed as a dichotomized variable (HR 1.64, 95% CI 1.02-2.65, P = 0.04).
Conclusion: In the subgroup of patients receiving adjuvant chemotherapy we found an association between shorter
survival after treatment in TIMP-1 high patients compared with TIMP-1 low patients, especially in patients receiving
anthracycline-based therapy. This suggests that high tumor tissue levels of TIMP-1 might be associated with reduced
benefit from classical adjuvant chemotherapy. Our findings should be validated in larger prospective studies.
Published: 10 September 2009
BMC Cancer 2009, 9:322doi:10.1186/1471-2407-9-322
Received: 28 April 2009
Accepted: 10 September 2009
This article is available from: http://www.biomedcentral.com/1471-2407/9/322
© 2009 Schrohl et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 2

BMC Cancer 2009, 9:322http://www.biomedcentral.com/1471-2407/9/322
Page 2 of 13
(page number not for citation purposes)
Background
Many patients are offered adjuvant chemotherapy after
surgery for primary breast cancer. Several options exist
(reviewed in [1]) but currently, adjuvant chemotherapy
only reduces breast cancer mortality by up to 50% [2] and
hence further improvement is needed. Moreover, chemo-
therapy is most often associated with substantial side
effects. In theory, by attacking the tumor cells effectively at
an early stage unnecessary growth and potential selection
of resistant clones is avoided and a better response to
adjuvant treatment is to be expected. Choosing the most
effective adjuvant chemotherapy regimen up front should
thus lead to a further reduction in disease recurrences and,
accordingly, to increased survival of the patients. In addi-
tion, avoiding ineffective treatments would reduce
unneeded toxicity and burden for the patients and the
substantial costs associated with administration of adju-
vant chemotherapy.
Prescription of chemotherapy with no benefit to the
patient could be limited through identification and use of
predictive markers. Predictive markers serve as a tool for
tailoring therapy for individual patients, however, the
number of approved markers in primary breast cancer is
limited [3]. Thus, additional markers are needed to
improve planning of a more personalized adjuvant treat-
ment.
TIMP-1 is a naturally occurring inhibitor belonging to the
matrix metalloproteinase (MMP) system, and besides its
MMP-inhibitory functions TIMP-1 also appears to inde-
pendently influence cell growth and apoptosis (for
reviews, see [4,5]). In metastatic breast cancer patients, we
reported that a high level of TIMP-1 in the primary tumor
tissue is associated with decreased objective response to
chemotherapy [6], and a report showing a similar associ-
ation in the adjuvant setting was recently presented [7]. In
addition, it appears that TIMP-1 may be combined with
TOP2A for prediction of response to anthracycline-based
chemotherapy [8]. Similarly, high levels of plasma and
serum TIMP-1 are associated with a decreased response to
endocrine therapy in metastatic breast cancer patients
[9,10]. Besides the predictive information, which can
apparently be gained from TIMP-1, several publications
demonstrate that high levels of TIMP-1 protein in breast
cancer tissue [11-14], plasma and serum [12,15-17] are
associated with a poor prognosis. Based on these previous
studies, it may be concluded that TIMP-1 appears to carry
predictive as well as prognostic information; it has not
been clarified, though, whether TIMP-1 carries mainly
prognostic or predictive information and whether this
information may be restricted to certain patient sub-
groups.
To further study the possible use of TIMP-1 as a biomarker
in breast cancer patients, the present study investigates the
association between levels of TIMP-1 in cytosolic tumor
tissue extracts and outcome following adjuvant chemo-
therapy with cyclophosphamide-methotrexate-5-fluorou-
racil (CMF) or an anthracycline-containing regimen
(cyclophosphamide and 5-fluorouracil with adriamycin/
epirubicin (CAF/CEF); or single-agent adriamycin (A)), or
following no adjuvant treatment. We hypothesized that
high levels of TIMP-1 are associated with decreased bene-
fit from adjuvant chemotherapy and we investigated this
hypothesis in a cohort of 525 premenopausal lymph
node-positive breast cancer patients, who constitute a
subgroup of a large cohort previously described in several
studies [14,18,19]. First, we evaluated the association with
prognosis in all patients included. The included subgroup
of patients receiving no adjuvant therapy allowed for eval-
uation of association between TIMP-1 and outcome with-
out any influence from adjuvant therapy; by definition,
this is evaluation of prognosis. Secondly, in patients
treated with adjuvant chemotherapy we analyzed the pre-
dictive impact of TIMP-1. The endpoints evaluated in the
present study are DFS and OS. In particular, for the analy-
ses of predictive impact of TIMP-1, DFS is considered the
most informative as this endpoint is not influenced by
subsequent systemic therapy. The Reporting Recommen-
dations for Tumor Marker Prognostic Studies (REMARK
[20]) were adhered to where ever applicable.
Methods
Patients and study design
Tumor tissue extracts included in the present study were
collected between 1979 and 1994. Our protocol for stud-
ying molecular markers associated with disease progres-
sion was approved by the institutional Medical Ethics
Committee of the Erasmus MC Rotterdam (MEC no.
02.953). The present study, in which coded tissues were
used, was performed in accordance with the Code of Con-
duct of the Federation of Medical Scientific Societies in the
Netherlands [21].
Patients were selected for the present study from a cohort
of 2984 breast cancer patients [14] for which the original
inclusion/exclusion criteria are described in [18]. Here,
only patients who had lymph node-positive breast cancer
and who were premenopausal at the time of diagnosis
were included. This is due to the fact that very few post-
menopausal and no lymph node-negative patients
received adjuvant chemotherapy at the time when these
samples were collected. Furthermore, only patients who
received adjuvant chemotherapy with CMF, CEF, CAF or
single-agent A, or who received no adjuvant chemo- or
endocrine therapy were included. Treatment decisions
were based on standard rules at the time. Finally, patients
were included in the present study based on the availabil-

Page 3

BMC Cancer 2009, 9:322 http://www.biomedcentral.com/1471-2407/9/322
Page 3 of 13
(page number not for citation purposes)
ity of information about TIMP-1 concentration in the pri-
mary tissue, originally measured in stored cytosolic
extracts after estrogen and progesterone receptor (ER/PgR)
determination and used for a previous study [14].
A total of 525 patients fulfilled the inclusion criteria; 164
of these were ≤ 40 years, and 361 were 41-55 years.
Among included patients, 292 patients had 1-3 positive
lymph nodes and 233 had more than 3 tumor-positive
axillary lymph nodes. In 175 patients, tumors were ≤ 2 cm
(pT1), 270 patients had tumors larger than 2 cm and ≤ 5
cm (pT2), and 80 patients had tumors larger than 5 cm or
skin or chest wall involvement (pT3+4). Differentiation
grade was well or moderate in 103 tumors, poor in 291
tumors and unknown in 131 tumors. ER/PgR positive
tumors were present in 432 patients. All patients had sur-
gical removal of their tumor (313 mastectomies, 212
lumpectomies) and 391 patients received adjuvant radio-
therapy (RT; of the chest wall and/or local or regional
lymph nodes). A total of 423 patients received adjuvant
chemotherapy; 324 had CMF, 99 had an anthracycline-
containing regimen, and 102 patients had no adjuvant
systemic therapy. The characteristics of the total patient
group as well as subgroups according to adjuvant treat-
ment are summarized in Table 1. In the table, and in the
following parts, the term "Untreated patients" refers to
patients who received no adjuvant therapy; however, it
should be emphasized that these patients were treated
with surgery and some also with adjuvant radiotherapy.
Follow-up consisted of routine examinations every 3-6
months during the first 5 years and once a year thereafter.
Table 1: Patient and tumor characteristics and comparisons of patient groups
CharacteristicAll patients
N (%)
CMF-treated patients
N (%)
Anthracycline-treated
patients
N (%)
Untreated patients
N (%)
P1
Age
≤ 40 years
41-55 years
164 (31)
361 (69)
110 (34)
214 (66)
28 (28)
71 (72)
26 (25)
76 (75)
0.212
Involved lymph nodes
1-3
>3
292 (56)
233 (44)
217 (67)
107 (33)
47 (47)
52 (53)
28 (27)
74 (73)
<0.0012
Steroid hormone rec. status
Positive
Negative
432 (82)
93 (18)
269 (83)
55 (17)
76 (77)
23 (23)
87 (85)
15 (15)
0.242
Tumor size
pT1
pT2
pT3+4
175 (33)
270 (51)
80 (15)
131 (40)
147 (45)
46 (14)
21 (21)
58 (59)
20 (20)
23 (23)
65 (64)
14 (14)
<0.0012
Grade
Well/moderate
Poor
Unknown
103 (20)
291 (55)
131 (25)
65 (20)
174 (54)
85 (26)
21 (21)
61 (62)
17 (17)
17 (17)
56 (55)
29 (28)
0.332
Primary treatment
Lumpectomy
Mastectomy
212 (40)
313 (60)
169 (52)
155 (48)
4 (4)
95 (96)
39 (38)
63 (62)
<0.0012
Events3
Recurrences (disease failure)
Deaths
268
149
140
75
58
32
70
42
TIMP-1 (median, range)
12.5 (0- 113) 12.0 (0- 106)13.5 (0-51.2)13.8 (0- 113)0.204
All patients, N = 525; CMF-treated patients, N = 324; anthracycline-treated, N = 99; untreated patients, N = 102.
1Testing the hypothesis that subgroups (CMF- or anthracycline treated and untreated) are similar
2Pearson X2 test
3Events when censored at 60 months
4Kruskal-Wallis equality-of-populations rank test

Page 4

BMC Cancer 2009, 9:322http://www.biomedcentral.com/1471-2407/9/322
Page 4 of 13
(page number not for citation purposes)
The median survival time of patients alive was 99 (range,
9-255) months (CMF-treated patients, 103 (range, 12-
248) months; anthracycline-treated patients, 90 (range, 9-
213) months; untreated patients, 93 (range, 15-255)
months).
Median survival of patients alive represents median obser-
vation time.
Assays and tumor specimens
TIMP-1 concentrations were originally determined for use
in a large study of the association between TIMP-1 and
prognosis, and this study also included validation of the
assay with the present extraction buffer [14].
In brief, tumor tissue specimens were prepared following
instructions from the European Organisation for Research
and Treatment of Cancer (EORTC) regarding extraction of
tumor tissue for determination of cytosolic ER and PgR
[22]. Levels of TIMP-1 in the cytosols were measured
using an established, validated in-house enzyme-linked
immunosorbent assay (ELISA) [23]. Concentrations of
total protein were measured using the Coomassie Brilliant
Blue Method (Bio-Rad Laboratories, Hercules, CA). ER
and PgR levels were assessed previously by ligand-binding
assay or enzyme immunoassay as described in [24] and
the cut point used for classification as ER/PgR positive or
negative was 10 fmol/mg of protein.
Data analysis and statistics
Differences in TIMP-1 levels were assessed with the Mann-
Whitney U test or the Kruskal-Wallis equality-of-popula-
tions rank test when appropriate. In these tests, patient
and tumor characteristics were used as grouping variables.
Associations between continuous variables were tested
with the Spearman rank correlation (rs). Differences
between treatment groups were tested with the Pearson X2
test. Cox proportional hazard models were applied to
compute the hazard ratio (HR). Disease-free survival
(DFS) and overall survival (OS) were used as endpoints.
For DFS any relapse or secondary breast cancer was
counted as failure; non-failing patients were censored at
last date of contact. For OS, death of any cause was
counted as a failure. Surviving patients were censored at
last day of follow-up.
The protocols for adjuvant treatment (RT on the axilla
and/or systemic therapy), depending on the number of
tumor-positive lymph nodes (1-3, >3), were changed dur-
ing the study period. To accommodate these changes we
used a combination of nodal status (1-3, >3) and RT on
the axilla (no, yes) to stratify all analyses.
Proportional hazards assumptions were tested based on
Schoenfeld residuals. As to be expected, the proportional
hazards assumption was violated for hormone receptor
status and accordingly, all analyses were stratified for hor-
mone receptor status (negative, positive). Hence, none of
the variables used for stratification (nodal status, RT, hor-
mone receptor status) were included in the multivariable
analyses. Age, tumor size, and malignancy grade defined
the base model to correct for classical prognostic factors.
In multivariable analyses TIMP-1 was added to this
model. In none of these analyses the proportional hazards
assumption was violated.
To reduce skewness of the distribution of TIMP-1 levels,
these were log-transformed. No interactions were
observed between TIMP-1 and the classical prognostic fac-
tors or adjuvant therapy in the analysis for DFS. The inter-
action analysis investigates whether the contribution to
the survival model of the one variable is dependent upon
the values of the other; the estimates for the survival
model will be different for the subgroups if there is an
interaction [25].
TIMP-1 levels were dichotomized by the overall median
(12.5 ng/mg of total protein). Dichotomized levels were
used for the survival curves using the method of Kaplan
and Meier.
In this dataset, three groups relating to systemic adjuvant
treatment were available: patients treated with CMF,
patients treated with anthracyclines, and patients who
received no adjuvant chemotherapy. In the analyses of all
patients we used the group receiving no adjuvant chemo-
therapy as the reference group. For both endpoints (DFS,
OS), the hypothesis that these untreated patients are not
different, regardless of tumor tissue TIMP-1 levels (low,
high), was maintained.
For all analyses, the survival times were censored at 60
months because the numbers of patients at risk in some
treatment groups were low at 5 years; the numbers of
patients at risk in the treatment subgroups appear from
the figures (Figure 1, 2 and 3).
Computations were performed with the STATA statistical
package, version 10.1 (STATA Corp., College Station, TX).
All P-values are two-sided and P < 0.05 was considered
statistically significant.
Results
Characterization and comparison of patient subgroups
Patients were subdivided according to adjuvant chemo-
therapy regimen received (CMF, 324 patients; anthracy-
cline-containing, 99 patients; no adjuvant chemotherapy,
102 patients). Table 1 summarizes the characteristics of
the 3 subgroups together with the characteristics of the
total patient group.

Page 5

BMC Cancer 2009, 9:322http://www.biomedcentral.com/1471-2407/9/322
Page 5 of 13
(page number not for citation purposes)
The three subgroups differed significantly with respect to
the number of tumor-positive lymph nodes, tumor size,
and primary treatment (Pearson X2 test, P < 0.001, Table
1). Thus, patients who received CMF more often had only
1-3 involved lymph nodes and also had smaller tumors
(pT1) when compared with untreated patients and with
patients who received anthracycline-based chemotherapy.
Looking at primary treatments, more patients had a
lumpectomy among CMF-treated and patients receiving
no adjuvant chemotherapy than among anthracycline-
treated patients. In addition to this, fewer anthracycline-
treated patients received adjuvant RT. The differences in
number of involved lymph nodes and adjuvant RT were
partly accounted for in the stratification.
TIMP-1 levels and association of TIMP-1 with
clinicopathological variables
The median TIMP-1 concentration in the tissue extracts
was 12.5 ng/mg of total protein (range, 0-113 ng/mg pro-
tein). Median tumor tissue TIMP-1 levels in subgroups
Kaplan-Meier plot showing the DFS (A) and OS (B) of untreated TIMP-1 low and high patients
Figure 1
Kaplan-Meier plot showing the DFS (A) and OS (B) of untreated TIMP-1 low and high patients. The median
TIMP-1 concentration of the total patient group (12.5 ng/mg of total protein) was used as cut point (TIMP-1 low patients N =
57, TIMP-1 high patients N = 45). Cox univariate regression analysis; DFS: HR 0.93, 95% CI 0.58-1.49, P = 0.76; OS: HR 0.72,
95% CI 0.39-1.33, P = 0.30.
Months
0
25
50
75
100
0 1224 3648 60
TIMP-1 low TIMP-1 high
TIMP-1 high 574234 2522
17
4536 26 1814
10
TIMP-1 low
Number at risk
Cumulative percentage
A
TIMP-1 high
TIMP-1 low
Number at risk
Months
0
25
50
75
100
0 12 24 3648
60
TIMP-1 low TIMP-1 high
Cumulative percentage
57 57 50 433630
45 40 35 30 25
17
B
Kaplan-Meier plot showing the DFS (A) and OS (B) of TIMP-1 low and high anthracycline-treated patients
Figure 2
Kaplan-Meier plot showing the DFS (A) and OS (B) of TIMP-1 low and high anthracycline-treated patients. The
median TIMP-1 concentration of the total patient group (12.5 ng/mg of total protein) was used as cut point (TIMP-1 low
patients N = 46, TIMP-1 high patients N = 53). Cox univariate regression analysis; DFS: HR 1.52, 95% CI 0.88-2.63, P = 0.13;
OS: HR 2.53, 95% CI 1.19-5.39, P = 0.02.
5351
41 3329 26
TIMP-1 high
46
4444
40
3533
TIMP-1 low
Number at risk
0
25
50
75
100
01224 364860
TIMP-1 lowTIMP-1 high
Cumulative percentage
Months
53 44 322520 17TIMP1-high
4639 3228
2420
TIMP-1 low
Number at risk
Months
0
25
50
75
100
01224364860
TIMP-1 low
TIMP-1 high
Cumulative percentage
A
B