CLINICAL TRIAL

The fibroblast growth factor receptor 1 (FGFR1), a markerof response to chemoradiotherapy in breast cancer?

Carole Massabeau • Brigitte Sigal-Zafrani • Lisa Belin • Alexia Savignoni •

Marion Richardson • Youlia M. Kirova • Elizabeth Cohen-Jonathan-Moyal •

Frederique Megnin-Chanet • Janet Hall • Alain Fourquet

Received: 9 February 2012 / Accepted: 8 March 2012 / Published online: 22 March 2012

� Springer Science+Business Media, LLC. 2012

Abstract The goal of the present study was to evaluate

the role of the tyrosine kinase receptor fibroblast growth

factor-1 (FGFR1) and its ligand, the fibroblast growth

factor 2 (FGF2) in determining the response to chemora-

diotherapy of breast cancers. S14 was a phase II neoadju-

vant study carried out at the Institut Curie that recruited 59

patients between November 2001 and September 2003.

This prospective study aimed to assess the pathological

response after preoperative radiochemotherapy (5FU–Na-

velbine-radiotherapy) for large breast cancers. The

expression of FGFR1 and FGF2 in tumor cells were

assessed by immunohistochemistry. Tumors in which no

staining was seen, were considered as negative for that

protein. We used the Khi-2 test or the Fisher test to com-

pare the qualitative variables and the Student t test or the

non-parametric Wilcoxon test for the quantitative vari-

ables. We included in the present study all the 32 patients

from the S14 cohort for whom the tissue blocks from the

biopsy specimens were available with sufficient tumoral

tissue. FGFR1 and FGF2 staining were observed respec-

tively in 17 (56 %) and 22 (68 %) of the 32 tumoral

biopsies. The expression of FGFR1 was associated with the

hormone receptor positive status (p = 0.0191). Only 11 %

(1/9) of the high grade tumors failed to respond to che-

moradiotherapy compared to 68 % resistant tumors (15/22)

among the low/intermediate grade tumors (p = 0.0199).

Among the low/intermediate grade tumors, FGFR1 nega-

tive tumors did not respond to chemoradiotherapy (0/9),

compared with tumors expressing FGFR1 among which,

almost one half had a good response (6/13) (p = 0.0167).

Among the low and intermediate grade breast cancers, the

FGFR1 negative tumors were resistant to chemoradio-

therapy. The expression of FGFR1 in patients’ biopsies

may serve as a marker of response to chemoradiotherapy.

Keywords Breast radioresistance � Fibroblast growth

factor receptor 1 (FGFR1) � Fibroblast growth factor 2

(FGF2) � Neoadjuvant chemoradiotherapy �Predictive marker

Introduction

Despite substantial improvements in the treatment of breast

cancer, resistance to therapy remains a major clinical

problem [1]. Breast cancers are highly variable with respect

to histology, genetic abnormalities and gene expression

profiles [2]. This extensive biological heterogeneity could

explain why some tumors respond better than others to

treatment. Although hormone receptor status has proved to

be an effective marker of therapeutic response to hormonal

therapy and HER2 status to anti-HER2 therapy, biological

C. Massabeau � Y. M. Kirova � A. Fourquet

Department of Radiation Oncology, Institut Curie, 26 rue d’Ulm,

Paris 75005, France

C. Massabeau (&) � E. Cohen-Jonathan-Moyal

Department of Radiation Oncology, Institut Claudius Regaud,

24-26 rue du Pont St Pierre, Toulouse 31052, France

e-mail: massabeau.carole@claudiusregaud.fr

C. Massabeau � F. Megnin-Chanet � J. Hall

INSERM U612, Orsay 91405, France

B. Sigal-Zafrani � M. Richardson

Department of Tumor Biology, Institut Curie, Paris, France

L. Belin � A. Savignoni

Department of Biostatistics, Institut Curie, Paris, France

F. Megnin-Chanet � J. Hall

Institut Curie, Orsay 91405, France

123

Breast Cancer Res Treat (2012) 134:259–266

DOI 10.1007/s10549-012-2027-3

markers of response to chemo/radiotherapy have yet to be

identified. Radiotherapy for breast cancer patients

improves local control, relapse free and overall survival

[3]. However, resistance to radiotherapy, whether acquired

or intrinsic could lead to insufficient tumoral response or to

local recurrence after radiotherapy [4]. Primary radiother-

apy to tumors not amenable to wide excision has been

largely used in the Institut Curie as well as by other

investigators, and allowed breast preservation in large

subsets of patients [5–8]. It is of crucial importance to

identify these resistant tumors in order not to delay the

effective surgical treatment and to avoid ineffective and

potentially deleterious treatment. In recent years, several

published studies have attempted to combine neoadjuvant

chemotherapy with radiation for locally advanced breast

cancers with favorable results in terms of pathological

response and tolerance [9–11]. In the S14 phase II study

[11], the in-breast pathological complete response (pCR) to

neoadjuvant 5FU–Navelbine-radiotherapy was achieved in

27 % of patients while the breast conservation rate was

69 %. A histological grade 3 was the only clinicopatho-

logical factor independently associated with the pCR

(p = 0.004) [11]. The fibroblast growth factor receptor 1

(FGFR1) is one of the tyrosine kinase receptors for the pro-

angiogenic fibroblast growth factor 2 (FGF2), secreted by

endothelial and tumor cells. Recent lines of evidence

indicate that FGFR1 may play a significant role in the

biology of breast carcinomas, in particular hormonal

receptor positive and/or low grade breast carcinomas.

[12–14]. The amplification of FGFR1 (8p11.2-p12 ampli-

con) is seen in 10–30 % of breast cancers, reflecting one of

the most frequently found genetic abnormalities and is

correlated with FGFR1 protein overexpression [15, 16].

The amplification/overexpression of FGFR1 has recently

shown to be associated with a poor prognosis, early relapse

and hormone resistance in two independent series of 87 and

245 breast cancers [16]. Our hypothesis was that this

FGFR1 overexpression also reflected a more chemosensi-

tive and radiosensitive phenotype.

The main goal of the present study was to evaluate the

role of the FGF2/FGFR1 tumoral expression in determin-

ing the response to chemo/radiotherapy for breast cancers

included in the neoadjuvant phase II S14 study.

Patients and methods

Patient population and study design

S14 was a phase II prospective study that recruited breast

cancer patients from November 2001 to September 2003 at

the Institut Curie. The objective of this study was to assess

the rate of pathological complete response after a

preoperative radiochemotherapy combining breast and

regional lymph nodes radiotherapy with concurrent 5FU–

vinorelbine chemotherapy for breast cancers not amenable

for conservative surgery. Chemotherapy consisted of intra-

venous administration of 5FU and vinorelbine repeated

every 3 weeks for a total of 4 courses. Radiotherapy started

on Day 1 of the second course of chemotherapy. The whole

breast was irradiated to 50 Gy, and the internal mammary

chain and the supra- and infraclavicular areas were

irradiated to 46 Gy. Surgery was performed in all cases at

completion of chemoradiotherapy after an interval of

6-8 weeks after the end of radiotherapy. It consisted of

axillary lymph node dissection of the first two levels and of

either a tumorectomy or a modified radical mastectomy. The

decision depended on the relative volumes of the residual

tumor and of the breast when assessed both clinically and by

the same breast-imaging modalities as at inclusion. All the

surgical pieces were reviewed by the pathologist involved in

this study to assess the pathological response according to

the concept, proposed by Sataloff et al. [17]. All the 59

patients included in the S14 trial provided written informed

consent to use the biopsy samples for research purposes. We

included in the present study all the 32 patients from this

cohort for whom the tissue blocks from the biopsy speci-

mens were available with sufficient tumor tissue. We used

the reporting recommendations for tumor marker studies

(REMARK) to report the data [18].

Assessment of the radioresistant phenotype

The definition of tumoral radioresistance in patients is the

diminished amount of regression for the same dose, or the

increased dose required for an equal amount of regression

or for sterilization [4]. All the patients from the S14 study

received the same radiotherapy regimen, including the

same dose. Therefore, we considered that tumors that have

a minor (\50 %) or absence of pathological response on

surgical specimens, after neoadjuvant radiochemotherapy

would represent the radioresistant group. In contrast, the

tumors that experienced a pathological complete response

or a major response ([50 %) are considered to be radio-

sensitive. It should be noted that we focused here on in

breast tumoral radioresistance and we did not take into

account histological node response.

Specimen characteristics

The main histological findings were already described in

the first report of the S14 study [11]. Here we have com-

pleted this assessment using the tumor material from

diagnosis/pretreatment core biopsies. This consisted of 32

paraffin-embedded specimens fixed in AFA (75 % ethanol

100�/5 % acetic acid/2 % commercial formalin/18 %

260 Breast Cancer Res Treat (2012) 134:259–266

123

deionized water). For each case, the histology of the

sample analyzed was verified on a Mayer’s hematoxylin

stained preparation, and areas showing invasive carcinoma

were identified. Histological grade was assessed using the

Elston–Ellis modification of Bloom-Richardson grading

method [19]. The estrogen receptor (ER) and progesterone

receptor (PR) status were defined by immunohistochemis-

try: staining of C10 % of the tumor cells was accepted as

positivity. HER-2 overexpression was determined by IHC

when complete membranous staining was observed in more

than 60 % of cells, either of strong or moderate intensity,

which was then confirmed by gene amplification assessed

by fluorescence in situ [20]. Tumors were considered as

1/triple-negative if all the ER, PR and HER-2 were absent

2/hormonal receptor positive (HR?) if ER or PR staining

was positive (luminal-like subtype according the Perou and

Sorlie classification [21] 3/HER-2 positive (HER-2?) if

HER-2 was overexpressed and ER and PR absent.

Assay methods

Immunohistochemistry for FGF2 and FGFR1 was carried

out on pretreatment biopsies as following. 4-lm sections

were mounted and deparaffinized in toluene. Antigen

retrieval was achieved by microwaving slides in 10 mmol/

L of citrate buffer (pH 6.1; 20 min). Endogenous peroxi-

dases were inhibited by a 5-min incubation in 3 % H2O2 in

deionized water and then, the sections were blocked using

Dako Protein Block (Dako, France). The sections were then

incubated with either a mouse monoclonal antibody

directed against FGFR1 (MA1-21519, Thermo Scientific,

dilution 1/500, 4 �C overnight), and FGF2 (Millipore,

clone bFM2, dilution 1/400 4 �C overnight). Antibody

binding revealed using the Dako EnVisionTM ? Dual Link

System-HRP (DAKO France), following the manufac-

turer’s instructions. Then, slides were incubated with

3,3-diaminobenzidine for 5 min and counterstained with

Mayer’s hematoxylin. Negative controls were performed

by omitting the specific primary antibody. Positive controls

were chosen according to the protein being studied: fibro-

adenoma for FGFR1 and fibroblasts for FGF2 staining.

Immunoreactivity was scored semi quantitatively using a

light microscope by 2 independent investigators (C.M. and

B.SZ.) who were unaware of patient outcome or other

clinical findings. Tumors in which no staining was seen

were considered as negative for that protein’s expression.

When an immunostaining was present, they were consid-

ered as positive for the studied protein and the percentage

of positive cells was evaluated, from 0 to 100 % and the

intensity was recorded was evaluated, using a score ranging

from 1 to 3 (1: light, 2: moderate, 3: strong). In case of

discrepancies, a consensus was sought by a concerted

reassessment of the sample.

Statistical analysis methods

The representativeness of the 32 patients for whom bio-

logical samples were available compared to all S14 trial

population (59 patients) was checked. The data were

summarized by frequency and percentage for categorical

variables and by a median as well as a range for continuous

variables. FGFR1 and FGF2 expression was scored as

positive when immunohistochemical staining was detected

in [1 % of the cells; it was not possible to use a statisti-

cally determined cut-off for their expression due to the

small number of tissue samples available. To compare

qualitative variables the Khi-2 test or the Fisher test was

used. The quantitative variables were compared with the

Student t test or the non-parametric Wilcoxon test. To

evaluate the link between the response and clinical vari-

ables or biomarkers, univariate logistic analysis was per-

formed. A significant level was fixed at 10 % for univariate

analysis. Analyses were performed using R software

2.12.1.

Results

Patient characteristics and pathological response

The 32 patients from the S14 cohort included in the present

study do not differ from those for whom we did not have

any biological tissue, with respect to all the clinical and

pathological factors, confirming the good representative-

ness of the panel of samples available (data not shown).

The main demographic and tumor characteristics are listed

in Table 1. The median age was 49 years old (range

33–64). Most tumors (66 %) are HR? (21/32). There were

nine high grade (histological grade 3) tumors (28 %), eight

low (histological grade 1) (25 %) and 14 intermediate

grade tumors (histological grade 2) (44 %). The patho-

logical response was complete and major for 6 and 10

patients, respectively while 16 (50 %) experienced minor

or no response, these insufficient responses defining the

radioresistant group.

Description of biological markers

We performed an immunohistochemical analysis for the

detection of FGFR1 and FGF2 in tumorous cells in biop-

sies, with all the samples being processed in one batch to

avoid inter-experiment variability (Fig. 1). FGFR1 staining

observed in 17 of the 32 tumors (53 %), was predominantly

cytoplasmic (8/17) (Table 2), in agreement with a previous

report [22]. The FGF2 staining observed in 22 of the 32

tumors (68 %) was mainly localized in the nucleus (11/22)

(Table 2), supporting the hypothesis of a FGF2 nuclear

Breast Cancer Res Treat (2012) 134:259–266 261

123

translocation [23]. The intensity of staining was moderate

or strong for FGFR1 in 22 % (7/32) and for FGF2 in 38 %

(12/32) of samples. No association was found between the

expression of FGF2 and FGFR1, suggesting that the

expression of those factors are independent in tumors cells

and that FGFR1 may be expressed in absence of the FGF2

ligand and inversely.

The staining for FGF2 and FGFRl was mainly located

on the tumor cells. However, we also observed that the

stromal cells were often slightly or moderately positive for

FGFR1 (fibroblasts and smooth muscle cells of blood

vessels) and for FGF2 (capillary endothelial cells).

Relationship between clinicopathologic parameters

and FGF2/FGFR1 expression

We first investigated whether a relationship might exist

between the clinicopathologic parameters. The high grade

tumors were more frequently HR- than HR?

(p = 0.0565). Six of the nine high grade tumors (67 %)

were also HR- while only 5 of the 22 low/intermediate

grade tumors (23 %) were HR-. This observation is con-

sistent with other studies depicting the more proliferative

phenotype of HR negative tumors [24, 25].

We then examined whether FGFR1 or FGF2 expression

was associated with any clinical or pathological charac-

teristics. As previously reported, we found that FGFR1

expression was related to the molecular subtype

(p = 0.053). 67 % (14/21) of the luminal-like tumors

express FGFR1, whereas only 22 % (2/9) of the triple-

negative tumors express FGFR1. It was also noted that the

only HER-2? tumor showed FGFR1 expression. Except

for the more frequent luminal-like phenotype in the FGFR1

positive tumors, the clinical and tumor characteristics were

equally distributed between the FGF2 and FGFR1 positive

versus negative patients.

Relationship between clinical parameters

and the pathological response

We examined the association of the clinical/histological

features and the pathological response (Table 3). The rate

of resistance to chemoradiotherapy was higher for tumors

with a low mitotic index (\11 mitosis per 10 high power

field (HPF)) compared with higher mitotic index tumors

(C11 mitosis) (respectively 62 vs 30 %), but this associa-

tion was not statistically significant (p = 0.1056). No

relationship was found between the hormonal receptor or

the HER-2 expression and the pathological response. With

respect to the histologic grade (Elston-Ellis), the low/

intermediate grade tumors have an increased risk of not

responding to the chemo/radiotherapy. 15 of the 22 low/

intermediate tumors (68 %) did not respond to treatment

while only 1 tumor of the 9 high grade tumors (11 %)

failed to respond (p = 0.0139). The risk of resistance to

treatment was significantly lower for the high grade tumors

than for the low or intermediate grade tumors (HR = 0.06,

Table 1 Patient demographic and clinicopathological characteristics

N %

Age—years (median) 49 (range: 33–64)

Menopausal

No 22 69

Yes 10 31

Clinical stage

T2N0 12 38

T2N1 11 34

T3N0 3 9

T3N1 6 19

Infiltrating carcinoma

Ductal 22 69

Lobular 5 16

Medullar 0 0

Poorly differentiated 5 16

Histological grade

Low grade (grade 1) 8 25

Intermediate grade (grade 2) 14 44

High grade (grade 3) 9 28

Unknowna 1 3

Number of mitoses/10 high power field

\11 21 68

[11 10 32

HER2 overexpression

Yes 3 9

No 29 91

Hormonal receptor status (HR)

HR- 11 34

HR? 21 66

Estrogen/progesterone receptors (ER/PR)

ER-/PR- 11 34

ER-/PR? 3 9

ER?/PR- 9 28

ER?/PR? 9 28

Molecular subtypesb

HER2? 1 3

Triple-negative 10 31

Luminal-like 21 66

a Data is missing for one sample as an insufficient amount of tumor

was available to allow 10 high power fields to be counted for mitosisb Tumors were considered as 1/triple-negative if all the ER, PR and

HER-2 were absent 2/hormonal receptor positive (HR?) if ER or PR

staining was positive (luminal-like subtype according the Perou and

Sorlie classification (20)) 3/HER-2 positive (HER-2?) if HER-2 was

overexpressed and ER, PR absent

262 Breast Cancer Res Treat (2012) 134:259–266

123

95 % CI: [0-0.6]). As the histological grade was the only

factor statistically significant in the univariate analysis

carried out, multivariate analysis could not be performed.

Among the low/intermediate grade tumors, no differences

in terms of response either for the hormonal receptor

positive vs negative tumors (p = 0.1348), or for the high vs

low mitotic index tumors (p = 1) was observed.

Relationship between biologic parameters

and the pathological response

We then investigated the relationship between the FGFR1

and FGF2 expression pattern and the pathological response

to radiochemotherapy of our cohort. The rate of resistant

tumors among the FGFR1 negative tumors (9/14) (64 %)

was higher than in FGFR1 positive tumors (7/17) (41 %),

although this difference was not statistically significant.

For FGF2, the rate of response was similar (50 %) between

the FGF2? versus FGF2- tumors. As noted above, grade 3

tumors were found to respond to chemoradiotherapy (only

11 % showed an absence of response or a minor response)

and within this subset of tumors, the FGF2 and FGFR1

expression did not influence this response. However, in the

low as well as in the intermediate grade tumors, we

observed that the FGFR negative tumors did not respond to

treatment. Indeed, in these low and intermediate grade

tumors that did not express FGFR1 no response to che-

moradiotherapy was seen (9/9 i.e., 100 % resistant tumors)

compared to tumors that did express FGFR where 6/13

(46 %) were resistant (p = 0.0167) (Table 4).

Fig. 1 Immunodetection of FGFR1 and FGF2 in breast cancer

biopsies. Original magnification 9100. a No staining for FGFR1, i.e.

FGFR1 negative tumor. b FGFR1 positive tumor (95 % of stained

cells with a strong intensity). c FGF2 negative tumor. d FGF2 positive

tumor (85 % of stained cells with a moderate intensity)

Table 2 Description of the immunohistochemical staining for

FGFR1 and FGF2

FGFR1

expression*

FGF2

expression*

N (%) 17 (53 %) 22 (69 %)

Staining localization

Cytoplasmic 8 (47 %) 3 (14 %)

Nuclear 2 (12 %) 11 (50 %)

Both 7 (41 %) 8 (36 %)

Percentage of stained cells

median [min–max]

30 [0–100] 40 [0–100]

Immunostaining Intensity

Low 10 (59 %) 10 (45 %)

Moderate 5 (29 %) 8 (36 %)

High 2 (12 %) 4 (18 %)

Uninterpretable 1 (6 %) 2 (9 %)

* When more than 1 % of the tumor cells had a detectable staining,

antigen expression was considered as positive

Breast Cancer Res Treat (2012) 134:259–266 263

123

Discussion

Understanding the mechanisms of resistance to chemo-

therapy and radiation treatment, is of vital importance to

further improve the prognosis of breast cancer patients and

help to better define therapeutic strategies. To date, no

effective routinely applicable genomic or proteomic anal-

ysis or profiling has emerged for the identification of

responders/non responders. Several small neoadjuvant

studies provided proof of principle that, on pretreatment

biopsies, the gene expression profiles of chemotherapy-

sensitive cancers are different from resistant tumors

[25–28]. Most of published studies investigating predictive

factors have concentrated on chemotherapy and not on the

combination of chemo- and radiotherapy. In a recent study

of 19 triple-negative breast tumors treated with neoadju-

vant paclitaxel/radiation treatment [29], immunohisto-

chemical analysis revealed that the expression of

inflammation related proteins and immune infiltration on

pretreatment biopsies was associated with the response to

treatment. In our study, we investigated the factors asso-

ciated with the tumoral response to 5FU–Navelbine-

radiotherapy in S14 trial patients, who mostly had luminal-

like tumors. The fact that time between the completion of

the chemoradiotherapy and the surgical procedure was the

same for all the patients allow us to interprete the tumoral

regression on surgical pieces as a surrogate of chemora-

diosensitivity. As others have already shown [30–33], the

low and intermediate grade tumors in our study were at

higher risk of resistance to cytotoxic treatment than high

grade tumors (respectively 68 % vs 11 %). In addition, we

reported here that tumors which did not express FGFR1

protein on pretreatment biopsies were more resistant to

chemoradiotherapy than those with FGFR1 expression

and the difference of response between FGFR1 positive

and negative tumors was highly significant for the low and

intermediate grade tumors The FGFR1 tumoral expression

was independent from the proliferative markers (histolog-

ical grade and mitotic index), meaning that this gives us an

additional information on the tumoral phenotype. The main

Table 3 Minor or absence of

histological response according

to the clinical and conventional

pathological characteristics

(univariate analysis)

a This cut-off value was a

reference to the median ageb The low or intermediate grade

was the only factor significantly

associated with the resistance to

chemoradiotherapy (p = 0.0139)

Levels Minor or

absence of

response (%)

Odds

ratio

95 %

Confidence

interval

p value

Age (year)a \49 56 1.00 0.4773

C49 43 0.60 (0.1; 2.5)

Menopausal No 55 1.00 0.4480

Yes 40 0.56 (0.1; 2.5)

Clinical Stage T2 57 1.00 0.2455

T3 33 0.38 (0.1; 1.9)

N0 48 1.00 0.7100

N1 55 1.32 (0.3; 5.7)

Histological type Ductal 50 1.00 0.8210

Lobular 40 0.67 (0.1; 4.8)

Poorly differentiated 60 1.50 (0.2; 10.8)

Histological gradeb 1 or 2 68 1.00 0.0139b

3 11 0.06 (0; 0.6)

Hormonal receptor (HR) status HR? 52 1.00 0.7100

HR- 45 0.76 (0.2; 3.3)

Mitoses/10 high power field \11 62 1.00 0.1056

C11 30 0.26 (0.1; 1.3)

HER2 overexpression No 52 1.00 0.5515

Yes 33 0.47 (0; 5.7)

Molecular subtype HR? 52 1.00 0.9923

HER2? 0 0 (0; ?)

Triple-negative 50 0.91 (0.2; 4.1)

Table 4 Proportion of radioresistant tumors according to the FGFR1

expression in the low/intermediate grade subpopulation

Grade 1 or 2 Grade 1 or 2

FGFR 0 FGFR [ 0

Absence/minor response 9 6

Major/complete response 0 7

% of radioresistant tumors 100 % 46 %

p = 0.0167

264 Breast Cancer Res Treat (2012) 134:259–266

123

limitations of our study were first, the small number of

patients for whom biological material was available and

second, the absence of genomic analysis of FGFR1

amplification to confirm the immunohistochemical data

The validity of the formalin-fixed core biopsy in assessing

the pathologic features like the tumor histology, grade or

the expression of molecular markers by immunohisto-

chemistry has been well described by Rakha and Ellis [34].

Connor‘s study [35] shows that core needle biopsies pro-

vide reliable tissue sampling that allow the accurate and

reliable assessment of the prognostic tumor makers such as

the epidermal growth factor receptor 1, with a concordance

rate of 85–100 % when compared to surgical specimens.

Moreover, an immunohistochemical approach provides

information about the localization of the protein studied.

According to published literature, FGF2 and FGFR1 are

predominantly detected in breast tumors in association with

the stromal component; however, little has been reported

about the clinical significance of those proteins in the

breast cancer cells [36–38]. There is now evidence from

multiple cancer types to implicate FGF signaling in several

oncogenic processes, including proliferation, survival,

migration, invasion, angiogenesis and response to treat-

ment of tumor cells [39]. In the present study, we found no

relationship between the level of FGF2 expression and

resistance to therapy, in contrast to previous results of a

FGF2 radioprotective effect in lung cancer [40]. According

to our results, in the breast, the expression of FGF2 by

tumor cells was completely independent from the expres-

sion of FGFR1. This might be explained by the genetic

background of breast cancers. As reported by Turner [16],

the receptor gene amplification, which results in a supra-

physiological receptor overexpression may mediate and

activate the downstream pathways in breast tumor cells,

independently of its main ligand FGF2. Turner’s data

suggest that amplification and overexpression of FGFR1

may be a major contributor to poor prognosis in luminal-

type breast cancers, driving endocrine therapy resistance.

Our results are complementary to those data, suggesting

that those FGFR1 positive tumors may be more sensitive to

chemoradiotherapy.

Conclusion

The ability to predict tumor sensitivity to radiotherapy and/

or chemotherapy in a way analogous to making use of ER

or HER-2 status for therapeutic planning is an intriguing

prospect. The results of this preliminary study strongly

suggest that the expression of FGFR1 in patients’ biopsies

may serve as a marker of response to 5FU–Navelbine-

radiotherapy. The assessment of FGFR1 may contribute to

an optimal selection of treatment strategy especially for

patients who undergo preoperative therapy. This present

study clearly highlights the potential of pretreatment

biopsies and supports the accumulating evidence that the

analysis of the FGFR1 amplification/expression should be

integrated into the panel of other potential prognostic and

predictive markers in breast cancer, especially for the low

grade and/or luminal-like tumors. Further studies are nee-

ded : 1/to validate our data in an independent cohort 2/to

discriminate if this observation is also found in clinical

studies with neoadjuvant radiotherapy alone and chemo-

therapy alone including other regimens such as taxanes or

anthracyclins 3/to assess the association between FGFR1

expression and local relapse after radiation therapy 4/to

decipher the mechanisms of response to ionizing radiation

based on FGFR1 amplification/expression in cell lines and

in xenografts.

Acknowledgments The authors thank Sophie Dodier, Laurence

Vaslin, Vincent Pennaneach, Tomasz Zaremba and Andre Nicolas for

their substantial help with this study.

Conflict of interest The authors declare that they have no conflict

of interest.

References

1. Xia F, Powell SN (2002) The molecular basis of radiosensitivity

and chemosensitivity in the treatment of breast cancer. Semin

Radiat Oncol 12(4):296–304 Review

2. Pusztai L et al (2003) Gene expression profiles obtained from

single passage fine needle aspirations (FNA) of breast cancer

reliably identify prognostic/predictive markers such as estrogen

(ER) and HER-2 receptor status and reveal large scale molecular

differences between ER-negative and ER-positive tumors. Clin

Cancer Res 9:2406–2415

3. Early Breast Cancer Trialists’ Collaborative Group (EBCTCG),

Darby S et al (2011) Effect of radiotherapy after breast-con-

serving surgery on 10-year recurrence and 15-year breast cancer

death: meta-analysis of individual patient data for 10,801 women

in 17 randomised trials. Lancet 12(378):1707–1716

4. Conger A, Luippold H (1957) Studies on the mechanism of

acquired radioresistance in cancer. Cancer Res 17(9):897–903

5. Van Limbergen E, Van der Schueren E, Van den Bogaert W, Van

Wing J (1990) Local control of operable breast cancer after

radiotherapy alone. Eur J Cancer 26:674–679

6. Pierquin B, Owen R, Maylin C et al (1980) Radical radiation

therapy of breast cancer. Int J Radial Oncol Biol Phys 6:17–24

7. Amalric R, Santamaria F, Robert F et al (1982) Radiation therapy

with or without primary limited surgery for operable breast

cancer: a 20- year experience at the Marseille Cancer Institute.

Cancer 49:3, 34

8. Scholl SM, Fourquet A, Asselain B, Pierga JY, Vilcoq JR, Du-

rand JC, Dorval T, Palangie T, Jouve M, Beuzeboc P et al (1994)

Neoadjuvant versus adjuvant chemotherapy in premenopausal

patients with tumours considered too large for breast conserving

surgery: preliminary results of a randomised trial: S6. Eur J

Cancer 30A(5):645–652

9. Semiglazov VF, Topuzov EE, Bavli JL et al (1994) Primary

(neoadjuvant) chemotherapy and radiotherapy compared with

Breast Cancer Res Treat (2012) 134:259–266 265

123

primary radiotherapy alone in stage IIb–IIIa breast cancer. Ann

Oncol 5(7):591–597

10. Formenti SC, Volm M, Skinner KA et al (2003) Preoperative

twice-weekly paclitaxel with concurrent radiation therapy fol-

lowed by surgery and postoperative doxorubicin- based chemo-

therapy in locally advanced breast cancer: a phase I/II trial. J Clin

Oncol 21:864–870

11. Bollet MA, Fourquet A et al (2006) Pathological response to

preoperative concurrent chemoradiotherapy for breast cancer:

results of a phase II study. Eur J Cancer 42(14):2286–2295

12. Dickson C, Spencer-Dene B, Dillon C, Fantl V (2000) Tyrosine

kinase signalling in breast cancer: fibroblast growth factors and

their receptors. Breast Cancer Res 2(3):191–196

13. Turner N, Grose R (2010) Fibroblast growth factor signalling:

from development to cancer. Nat Rev Cancer 10(2):116–129

14. Reis-Filho JS, Ashworth A et al (2006) FGFR1 emerges as a

potential therapeutic target for lobular breast carcinomas. Clin

Cancer Res 15 12(22):6652–6662

15. Andre F, Job B, Dessen P et al (2009) Molecular characterization

of breast cancer with high-resolution oligonucleotide comparative

genomic hybridization array. Clin Cancer Res 15:441–451

16. Turner N et al (2010) FGFR1 amplification drives endocrine

therapy resistance and is a therapeutic target in breast cancer.

Cancer Res 70(5):2085–2094

17. Sataloff DM, Mason BA, Prestipino AJ et al (1995) Pathologic

response to induction chemotherapy in locally advanced carci-

noma of the breast: a determinant of outcome. J Am Coll Surg

180(3):297–306

18. McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark

GM (2006) Reporting recommendations for tumor marker prog-

nostic studies (REMARK). Breast Cancer Res Treat 100(2):229–235

19. Elston CW, Ellis IO (1991) Pathological prognostic factors in

breast cancer. I. The value of histological grade in breast cancer:

experience from a large study with long-term follow-up. Histo-

pathology 19(5):403–410

20. Vincent-Salomon A, Le Doussal V et al (2003) Calibration of

immunohistochemistry for assessment of HER2 in breast cancer:

results of the French multicentre GEFPICS study. Histopathology

42(4):337–347

21. Perou CM, Sorlie T, Eisen MB et al (2000) Molecular portraits of

human breast tumours. Nature 406:747–752

22. Blanckaert VD, Hebbar M, Louchez MM, Vilain MO, Schelling

ME, Peyrat JP (1998) Basic fibroblast growth factor receptors and

their prognostic value in human breast cancer. Clin Cancer Res

4(12):2939–2947

23. Hsia E, Richardson TP, Nugent MA (2003) Nuclear localization

of basic fibroblast growth factor is mediated by heparan sulfate

proteoglycans through protein kinase C signaling. J Cell Biochem

88:1214–1225

24. Genestie C, Zafrani B, Asselain B, Fourquet A, Rozan S, Validire

P, Vincent-Salomon A, Sastre-Garau X (1998) Comparison of the

prognostic value of Scarff-Bloom-Richardson and Nottingham

histological grades in a series of 825 cases of breast cancer: major

importance of the mitotic count as a component of both grading

systems. Anticancer Res 18(1B):571–576

25. Formenti SC, Dunnington G, Uzieli B et al (1997) Original p53

status predicts for pathological response in locally advanced

breast cancer patients treated preoperatively with continuous

infusion 5-fluorouracil and radiation therapy. Int J Radiat Oncol

Biol Phys 39:1059–1068

26. Ayers M, Symmans WF, Stec J, Damokosh AI, Clark E, Hess K,

Lecocke M, Metivier J, Booser D, Ibrahim N, Valero V, Royce

M, Arun B, Whitman G, Ross J, Sneige N, Hortobagyi GN,

Pusztai L (2004) Gene expression profiles predict complete

pathologic response to neoadjuvant paclitaxel and fluorouracil,

doxorubicin, and cyclophosphamide chemotherapy in breast

cancer. J Clin Oncol 22(12):2284–2293

27. Hess KR, Anderson K, Symmans WF, Valero V, Ibrahim N,

Mejia JA, Booser D, Theriault RL, Buzdar AU, Dempsey PJ,

Rouzier R, Sneige N, Ross JS, Vidaurre T, Gomez HL, Hort-

obagyi GN, Pusztai L (2006) Pharmacogenomic predictor of

sensitivity to preoperative chemotherapy with paclitaxel and

fluorouracil, doxorubicin, and cyclophosphamide in breast can-

cer. J Clin Oncol 24(26):4236–4244

28. Thuerigen O, Schneeweiss A, Toedt G, Warnat P, Hahn M,

Kramer H, Brors B, Rudlowski C, Benner A, Schuetz F, Tews B,

Eils R, Sinn HP, Sohn C, Lichter P (2006) Gene expression

signature predicting pathologic complete response with gemcit-

abine, epirubicin, and docetaxel in primary breast cancer. J Clin

Oncol 24(12):1839–1845

29. Bauer JA, Chakravarthy AB, Pietenpol JA (2010) Identification

of markers of taxane sensitivity using proteomic and genomic

analyses of breast tumors from patients receiving neoadjuvant

paclitaxel and radiation. Clin Cancer Res 16(2):681–690

30. Pinder, S.E., Murray S., Ellis I.O., Trihia H., Elston C.W., Gelber

R.D., Goldhirsch A., Lindtner J., Cortes-Funes H., Simoncini E.,

Byrne M.J., Golouh R., Rudenstam C.M., Castiglione-Gertsch M.

and Gusterson B.A. (1998) The importance of the histologic

grade of invasive breast carcinoma and response to chemother-

apy. Cancer 83(8):1529–1539

31. Von Minckwitz G, Raab G, Caputo A et al (2005) Doxorubicin with

cyclophosphamide followed by docetaxel every 21 days compared

with doxorubicin and docetaxel every 14 days as preoperative

treatment in operable breast cancer: the GEPARDUO study of the

German Breast Group. J Clin Oncol 23(12):2676–2685

32. Ring AE, Smith IE, Ashley S et al (2004) Oestrogen receptor

status, pathological complete response and prognosis in patients

receiving neoadjuvant chemotherapy for early breast cancer. Br J

Cancer 91(12):2012–2017

33. Petit T, Wilt M, Velten M et al (2004) Comparative value of

tumour grade, hormonal receptors, Ki-67, HER-2 and topoiso-

merase II alpha status as predictive markers in breast cancer

patients treated with neoadjuvant anthracycline-based chemo-

therapy. Eur J Cancer 40(2):205–211

34. Rakha EA, Ellis IO (2007) An overview of assessment of prog-

nostic and predictive factors in breast cancer needle core biopsy

specimens. J Clin Pathol 60(12):1300–1306 Review

35. Connor CS, Tawfik OW, Joyce AJ et al (2002) A comparison of

prognostic tumor markers obtained on image-guided breast

biopsies and final surgical specimens. Am J Surg 184:322–324

36. Smith K, Fox SB, Whitehouse R et al (1999) Upregulation of

basic fibroblast growth factor in breast carcinoma and its rela-

tionship to vascular density, oestrogen receptor, epidermal

growth factor receptor and survival. Ann Oncol 10:707–713

37. Linder C, Bystom P, Engel G et al (1998) Correlation between

basic fibroblast growth factor immunostaining of stromal cells

and stromelysin-3 mRNA expression in human breast carcinoma.

Br J Cancer 77:941–945

38. Penault-Llorca F, Bertucci F, Adelaıde J, Parc P, Coulier F,

Jacquemier J, Birnbaum D, deLapeyriere O (1995) Expression of

FGF and FGF receptor genes in human breast cancer. Int J Cancer

61(2):170–176

39. Turner N, Grose R (2010) Fibroblast growth factor signalling: from

development to cancer. Nat Rev Cancer 10(2):116–129 Review

40. Massabeau C, Rouquette I, Lauwers-Cances V, Mazieres J, Ba-

chaud JM, Armand JP, Delisle MB, Favre G, Toulas C, Cohen-

Jonathan-Moyal E (2009) Basic fibroblast growth factor-2/beta3

integrin expression profile: signature of local progression after

chemoradiotherapy for patients with locally advanced non-small-

cell lung cancer. Int J Radiat Oncol Biol Phys 75(3):696–702

266 Breast Cancer Res Treat (2012) 134:259–266

123