RESEARCH ARTICLE

Clinical significance of serum insulin-like growth factor-1 (IGF-1)and insulin-like growth factor binding protein-3 (IGFBP-3)in patients with breast cancer

Faruk Tas & Senem Karabulut & Elif Bilgin &

Didem Tastekin & Derya Duranyildiz

Received: 19 April 2014 /Accepted: 10 June 2014# International Society of Oncology and BioMarkers (ISOBM) 2014

Abstract Insulin-like growth factor-1 (IGF-1) and its primarybinding protein-3 (IGFBP-3) play an important role in cellularproliferation, differentiation and apoptosis in many tumors,including breast cancer (BC). The objective of this study wasto determine the clinical significance of the serum levels ofIGF-1 and IGFBP-3 in BC patients. A total of 96 patients witha pathologically confirmed diagnosis of BCwere enrolled intothis study. Serum IGF-1 and IGFBP-3 levels were determinedby the solid-phase sandwich enzyme-linked immunosorbentassay (ELISA) methods. Age- and sex-matched 30 healthycontrols were included in the analysis. The median age ofdiagnosis was 48 years (range: 29–80). Thirty-seven (39 %)consisted of metastatic disease. No significant difference inbaseline serum was found in both IGF-1 and IGFBP-3 levelsbetween BC patients and healthy controls (p=0.92 and p=0.26, respectively). None of the prognostic parameters ana-lyzed was correlated significantly with the serum assay con-centrations. Likewise, no correlations were also found be-tween these serum concentrations and response to chemother-apy. No significant correlation was found between serum IGF-1 and IGFBP-3 levels in BC patients (rs=0.048, p=0.66).Thepatients with elevated serum IGF-1 levels had favorable insurvival than those with lower levels (p=0.05). However,serum IGFBP-3 concentrations were found no prognostic rolefor outcome (p=0.35). In conclusion, elevated serum IGF-1level is afavorable prognostic factor for overall survival in BCpatients.

Keywords Serum . IGF-1 . IGFBP-3 . Breast cancer .

Prognostic factor

Introduction

Insulin-like growth factors (IGFs) is one of the major familiesof growth factors shown to be intimately involved in theregulation of cell growth and transformation. Insulin-likegrowth factor-1 (IGF-1) is a multifunctional peptide, whichplays an important role in cellular growth, proliferation, dif-ferentiation and cellular transformation in many tumors, in-cluding breast cancer [1–5].

In vitro studies show that IGF-1increases breast cell growthand invasive potential, suggesting a role for the IGF-1 path-way in breast cancer etiology [1–5]. Although numerous trialshave confirmed that IGF-1 plays a role in breast cancer,exactly its way to play this role and the clinical outcomeassociated with circulating IGF-1 concentrations is not clear.The effect of circulating IGF-1 on disease recurrence andoutcome in breast cancer patients are studied insufficiently[4–10]. Therefore, nowadays, the clinical significance of plas-ma circulating IGF-1 levels in women with breast cancer isambiguous.

The IGFBP family of binding proteins consists of sixstructurally related proteins and all members are expressedin the normal breast tissue [10, 11]. Their role is to bind andregulate their effects. Insulin-like growth factor bindingprotein-3 (IGFBP-3) is the primary binding protein of IGF-1and regulates the mitogenic and anti-apoptotic actions ofinsulin-like growth factors (IGFs) [9]. In most instances,IGFBP-3 has been shown to correlate with circulating IGF-1levels [4, 9, 12]. In addition, notably IGFBP-3 has direct IGF-independent effects on cellular growth and apoptosis. Thestudies agree that IGFBP-3 has been correlated with theprognosis of breast cancer [3, 7, 9]. IGFBP-3 may exert aninhibitory influence on breast cancer growth [5]. Serum levelsof IGFBP-3 are significantly decreased in breast cancer [5, 9].Moreover, overexpression of IGFBP-3 by breast cancer cellswith aggressive biological features is a paradox [9].

F. Tas (*) : S. Karabulut : E. Bilgin :D. Tastekin :D. DuranyildizInstitute of Oncology, University of Istanbul, Capa 34390, Istanbul,Turkeye-mail: faruktas2002@yahoo.com

Tumor Biol.DOI 10.1007/s13277-014-2224-2

The significance of the serological levels of IGF-1 andIGFBP-3 in breast cancer patients is not known yet. Giventhe conflicting results from recent epidemiological and clinicalstudies examining the IGF axis and breast cancer, we con-ducted this study whether serum IGF-1 and IGFBP-3 concen-trations have diagnostic, predictive and/or prognostic role inbreast cancer patients.

Material and methods

Patients

A total of 96 female breast cancer patients admitted to IstanbulUniversity, Institute of Oncology were enrolled into the study.Diagnosis of breast cancer was histologically proved eitherwith tru-cut biopsy of the primary lesion or metastatic site andstaged according to sixth edition of AJCC. All the patientswere treatment-naive for at least 3 months before accrual.

For histological evaluation tissue sections (2 mm) weredeparaffinized and stained using hematoxylin and eosine.Grading of tumors was established according to modifiedBloom–Richardson grading system. Estrogen receptor (ER),progesterone receptor (PR) and HER2 status were evaluatedin the sample sections using appropriate antibodies. The im-munohistochemical staining was assessed upon visual inspec-tion with optical microscope and considered as positive if thepercentage of cells staining positive were more than 5 %. Inthe case of 2(+) staining by IHC, HER2 gene amplificationwas analyzed by fluorescent in situ hybridization (FISH).

All patients were treated with multidisciplinary approach.Overall, 66 (83 %) patients received an anthracycline-basedregimen, 14 (18 %) received hormone therapy either as aroma-tase inhibitor or tamoxifen ± LHRH analogues, two patientswere followed up without any treatment due to early stage(<1 cm). Forty patients received adjuvant treatment after mas-tectomy or lumpectomy.Other 40 patients with locally advancedand metastatic disease were treated with anthracycline-basedchemotherapy (n=37) and only hormonal therapy (n=3), re-spectively. In these patients, response to treatment was evaluatedaccording to revised RECIST criteria version 1.1.

For comparison of serum IGF-1 and IGFBP-3levels, age-and sex-matched 30 healthy controls were included in theanalysis. Informed consent was obtained from all patientsand the study was reviewed and approved by our local ethicalcommittee.

Measurement of serum IGF-1 and IGFBP-3 levels

Serum samples were obtained on first admission before anyadjuvant and metastatic treatment was given or follow-uppatients. Blood samples were obtained from breastcancer patients and healthy controls by venipuncture and

clotted at room temperature. The sera were collected follow-ing centrifugation and frozen immediately at −20 °C untilanalysis.

Serum IGF-1 (Mediagnost, Germany) and IGFBP-3(Mediagnost, Germany) levels were determined by the solid-phase sandwich ELISA method.

The IGF-1 ELISA (Mediagnost, Germany) uses a double-antibody sandwich enzyme-linked immunosorbent assay(ELISA) to determine the level of human insulin growthfactor-1 (IGF-1) in samples. Serum samples and standardswere added to the wells which are pre-coated with humanIGF-1 monoclonal antibody. IGF-1 monoclonal antibody andallowed to incubate for 1 h. Unbound material was washedaway. Biotin-conjugated anti-human IGF-1 antibody andStreptavidin–HRP was added to incubate for 30 min. Biotin-conjugated anti-human IGF-1 antibody binds to human IGF-1captured by the first antibody. Streptavidin–HRP binds to thebiotin-conjugated anti-human IGF-1 antibody. Following in-cubation (30 min), unbound Streptavidin–HRP was removedduring a wash step, and substrate solution reactive with HRPwas added to the wells and incubated for 15 min. A coloredproduct was formed in proportion to the amount of humanIGF-1 present in the sample or Standard. The reaction wasterminated by an addition of acid (stop solution) and absor-bance is measured at 450 nm using an automated ELISAreader (Rayto, RT-1904C Chemistry Analyzer, Atlanta GA,USA). The results were expressed as ng/ml.

The IGFBP-3 ELISA (Mediagnost, Germany) uses adouble-antibody sandwich ELISA to determine the level ofhuman insulin growth factor binding protein-3 (IGFBP-3) insamples. Serum samples and standards were added to thewells which are pre-coated with human IGFBP-3 monoclonalantibody. IGFBP-3 monoclonal antibody and allowed to in-cubate for 1 h. Unbound material was washed away. Biotin-conjugated anti-human IGFBP-3 antibody and Streptavidin–HRP was added to incubate for 1 h. Biotin-conjugated anti-human IGFBP-3 antibody binds to human IGBP-3 capturedby the first antibody. Streptavidin–HRP binds to the biotin-conjugated anti-human IGBP-3 antibody. Following incuba-tion (30 min) unbound Streptavidin–HRP was removed dur-ing a wash step, substrate solution reactive with HRP wasadded to the wells and incubated for 30 min. A coloredproduct was formed in proportion to the amount of humanIGBP-3 present in the sample or Standard. The reaction wasterminated by an addition of acid (stop solution) and absor-bance was measured at 450 nm using an automated ELISAreader (Rayto, RT-1904C Chemistry Analyzer). The resultswere expressed as ng/ml.

Statistical analysis

Continuous variables were categorized using mean values ascut-off point. Assessment of relationships, comparisons

Tumor Biol.

between various clinical/laboratory parameters was accom-plished using Mann–Whitney U test. Spearman’s rank ordercorrelation was used for correlation analysis. Overall survival(OS) was calculated from the date of first admission to theclinics to disease-related death or date of last contact with thepatient or any family member. Kaplan–Meier method wasused for estimation of survival distribution and differ-ences in survival were assessed by the log-rank statis-tics. A p value <0.05 was considered as significant.

Table 1 Characteristics of the patient and disease

Variables n (%)

No. of patients 96 (100)

Age, years

<50/≥50 51 (53)/45 (47)

Histological grade

I/II/III/unknown 4 (4)/26 (27)/24 (25)/42 (44)

Estrogen receptor (ER)

−/+/unknown 27 (28)/68 (71)/1(1)

Progesteron receptor (PR)

−/+/unknown 32 (33)/63 (66)/1(1)

HER2

−/+/unknown 64 (67)/31 (32)/1(1)

Classificationa

Luminal/HER2+/Triple(−)/Triple(+)/Unknown

56 (58)/18 (19)/8 (8)/13(14)1(1)

Tumor status (T)b

1/2/3/4 18 (30)/21 (36)/3 (5)/3 (5)/14(24)

Node status (N)b

Negative/Positive/Unknown 23 (39)/23 (39)/13 (22)

Metastasis status (M)

0/1 59 (61)/37 (39)

Serum LDH level (450 U/l)a

Normal/Elevated 69 (73)/26 (27)

Serum CA15-3 level (35 U/ml)a

Normal/Elevated 68 (72)/26 (28)

Response to chemotherapyc

Yes/No/Unknown 46 (78)/10 (17)/3 (5)

a Patients with unknown data concerning the variables are not included inthe analysisb In 59 nonmetastatic diseasec In 37 metastatic and 22 locally advanced disease

Table 2 The values of serum marker levels in patients with breast cancerand in healthy controls

Assay Patients (n=96) Controls (n=30) pMean (SE) Mean (SE)

IGF-1 level (ng/ml) 112.7 (12.5) 93.7 (11.0) 0.92

IGFBP-3 level (ng/ml) 4,641.7 (225.2) 4,827.2 (274.7) 0.26

Table 3 Comparisons of serum assay levels according to various clini-cal/laboratory variables

Variables IGF-1 (ng/ml) Mean(SE)

IGFBP-3 (ng/ml)Mean(SE)

Age, years (p) 0.15 0.42

<50 118.0 (15.6) 4819.4 (308.5)

≥50 106.6 (20.3) 4464.1 (329.7)

Grade (p) 0.19 0.85

I + II 132.4(24.0) 4856.6 (374.0)

III 115.0 (31.5) 4872.4 (426.5)

ER (p) 0.99 0.40

Negative 117.4 (28.2) 5046.3 (477.7)

Positive 110.9 (13.9) 4446.9 (252.6)

PR (p) 0.67 0.79

Negative 111.0 (24.5) 4807.5 (432.6)

Positive 113.6 (14.7) 4520.1 (262.6)

HER2 (p) 0.93 0.61

Negative 112.6 (14.7) 4494.3 (257.2)

Positive 112.9 (24.7) 4882.4 (453.1)

Classification (p) 0.58 0.31

Luminal 119.0 (16.5) 4323.4 (250.1)

Others 103.7 (20.0) 5038.2 (421.4)

Classification (p) 0.39 0.85

Triple-positive 72.6 (9.1) 4724.6 (820.4)

Others 119.1 (14.5) 4602.3 (230.8)

Classification (p) 0.40 0.37

Triple-negative 68.0 (9.4) 5583.6 (999.6)

Others 116.7 (13.6) 4521.5 (228.0)

Classification (p) 0.93 0.61

HER2 enriched 112.9 (24.7) 4882.4 (453.1)

Others 112.6 (14.7) 4494.3 (279.2)

Tumor status (p) 0.92 0.61

1 (small) 134.2 (37.7) 5380.3 (626.7)

2–4 (large) 128.6 (28.0) 4296.2 (261.6)

Node status (p) 0.25 0.93

Negative 120.5 (30.3) 4627.8 (436.0)

Positive 145.8 (32.4) 4546.6 (395.3)

Metastasis status (p) 0.44 0.94

Negative 119.3 (17.6) 4627.9 (276.5)

Positive 102.0 (16.7) 4662.6 (386.5)

Serum LDH level (p) 0.98 0.39

Normal 106.4 (12.2) 4755.8 (263.8)

High 132.7 (33.2) 4292.2 (452.5)

Serum CA 15-3 level (p) 0.87 0.75

Normal 114.2 (15.6) 4562.9 (229.7)

High 109.4(23.0) 4744.1 (619.6)

Response tochemotherapy (p)

0.23 0.96

Yes 103.0 (13.5) 4470.5 (347.2)

No 76.0 (23.5) 4626.6 (807.2)

Tumor Biol.

Statistical analysis was carried out using SPSS 16.0software.

Results

Between February 2010 and June 2011, 96 female breastcancer patients were enrolled into the study. The median ageof diagnosis was 48 years (range: 29–80). Thirty-seven (39%)of the patients consisted of metastatic breast cancer patientswhere the most frequent site of metastasis was bone (68 %).Majority of the patients were responsive to chemotherapy(78 %) (Table 1).

No significant differences were found in baseline bothserum IGF-1 and IGFBP-3 levels between breast cancer pa-tients and healthy controls (p=0.92 and p=0.26, respectively)(Table 2). Table 3 shows the correlation between the serumlevels of both IGF-1 and IGFBP-3 and known clinicopatho-logical factors. None of the prognostic parameters analyzedwas correlated significantly with the serum assay concentra-tions. Moreover, no correlations were also found betweenthese serum concentrations and response to chemotherapy.

Similarly, no significant correlation between serum IGF-1and IGFBP-3 levels in breast cancer patients (rs=0.048, n=96,p=0.66; Spearman’s correlation) (Fig. 1).

The mean follow-up time was 19.4±9.85 months. Themean survival for all patients was 32.7±1.2 months.

The 1-, 2-, and 3-year OS rates were 92.4 % (95 % confi-dence interval [CI], 86.9–97.9), 80.0 % (95 % CI, 70.2–99.8),and 78.2 % (95 % CI, 68.4–80.0), respectively. The negativity

of ER (p=0.007) and PR (p=0.004), elevated serum LDHconcentration (p=0.042), metastatic stage (p<0.001) and un-responsiveness to chemotherapy (p<0.001) had statisticallysignificant poor prognostic factors for OS (Table 4). Thepatients with elevated serum IGF-1 levels had favorable inOS than those with lower levels (p=0.05) (Table 4 and Fig. 2).However, serum IGFBP-3 concentrations were found noprognostic role for OS (p=0.35) (Table 4 and Fig. 3).

Discussion

In literature, only a limited number of studies have investigat-ed serum IGF-1 and IGFBP-3 concentrations in human breastcancer [4, 5, 7–10]. A small study by Holdaway et al. [5]performed an RIA for IGF-1 and IGFBP-3 in serum samplesfrom 14 patients with advanced breast cancer about to undergochemotherapy and 31 patients with early-stage breast cancer.Mean basal serum levels of IGF-1 and IGFBP-3 were notsignificantly different between patients with advanced breastcancer and controls or women with early breast cancer.Moreover, patient survival was not significantly related tobaseline IGF-1 and IGFBP-3 levels.

In the WHEL study, baseline blood samples from 510matched breast cancer cases and controls were analyzed forIGF-1, and a subset of 188 pairs were studied for otherhormones including IGFBP-3 [4]. IGF-1 was positively, butnot significantly, associated with recurrence. However,IGFBP-3 did not significantly predict recurrence of breastcancer. These results do not support an association between

Fig. 1 Correlation betweenserum IGF-1 and IGFBP-3levels in breast cancer patients(rs=0.048, n=96, p=0.66),Spearman’s correlation

Tumor Biol.

Table 4 Univariate analysis of survival

Variables Mean survival months (SE) Two-year survival rate % (SE) p

All patients 32.7 (1.2) 80 (5)

Age of patients 0.118

<50 yeas 32.4 (1.4) 87 (6)

≥50 years 30.3 (1.9) 72 (8)

Grade (p) 0.19

I–II 20.5 (1.6) 44 (10)

III 22.7 (1.9) 58 (10)

ER (p) 0.007

Negative 31.3 (1.4) 77 (9)

Positive 32.0 (2.0) 83 (6)

PR (p) 0.004

Negative 30.9 (2.1) 73 (9)

Positive 31.9 (1.3) 86 (5)

HER2 0.55

Negative 30.3 (1.5) 77 (7)

Positive 34.2 (1.6) 88 (6)

Classification 0.41

Luminal 32.6 (1.7) 82 (6)

Others 31.0 (1.5) 77 (8)

Classification 0.20

Triple positive 28.3 (2.3) 75 (16)

Others 32.1 (1.3) 80 (5)

Classification 0.073

Triple negative 25.2 (4.2) 48 (21)

Others 32.6 (1.2) 83 (5)

Classification 0.105

HER2 enriched 35.1 (1.8) 94 (6)

Others 30.4 (1.3) 75 (6)

Tumor status 0.398

T1 NR 100 (0)

T2–T4 NR 94 (5)

Node status 0.35

Negative NR 100 (0)

Positive NR 94 (6)

Metastasis status <0.001

Negative 36.1 (0.7) 97 (3)

Positive 23.0 (2.1) 51 (10)

Serum LDH level 0.042

Normal 33.2 (1.2) 83 (5)

High 27.9 (2.9) 68 (11)

Serum CA 15-3 level 0.083

Normal 33.0 (1.2) 84 (5)

High 28.9 (2.4) 69 (11)

Response to chemotherapy <0.001

Yes 32.9 (1.3) 84 (7)

No 13.4 (2.5) 18 (13)

Serum IGF-1 level 0.05

High (>mean) 34.9 (1.4) 88 (8)

Low (<mean) 30.5 (1.5) 74 (6)

Serum IGFBP-3 level 0.35

High (>mean) 32.5 (1.7) 80 (7)

Low (<mean) 30.4 (1.8) 72 (8)

Tumor Biol.

high IGF-1 levels and risk of recurrence among postmeno-pausal women.

In a retrospective study on 130minority (African Americanand Hispanic) breast cancer patients, plasma IGF-1 and serumIGFBP-3 levels were found to be correlated with tumor his-topathology, menopausal status, treatment modality, recur-rence rates, and probability of survival [9]. Plasma IGF-1and serum IGFBP-3 were measured by RIA. Patients withbreast cancer have increased plasma IGF-1 and serum IGFBP-3 levels compared to normal subjects. Plasma IGF-1 did not

correlate with age and nodal status. Both increased with tumorsize. IGF-1 did not correlate with ER status, but did increase inPR-positive patients. IGF-1 concentrations were higher inpremenopausal patients and in women with cancer recurrence.It demonstrated survival probability of breast cancer patientswith increased plasma IGF-1 levels. Additionally, a positivecorrelation between plasma IGF-1 and serum IGFBP-3 levelswas found.

Another study involved 110 women operated on for breastcancer at least a year previously, not undergoing

Fig. 3 Survival curves in breastcancer patients according toIGFBP-3levels (p=0.35)

Fig. 2 Survival curves in breastcancer patients according to IGF-1 levels (p=0.05)

Tumor Biol.

chemotherapy [8]. Serum IGF-1 was higher in the recurrencegroup, but the differences were not significant. However,high IGF-1 levels may increase the risk of breast cancerrecurrences only in the presence of high PDGF levels.This suggests that in vivo IGF-1 and PDGF may alsointeract to increase the risk of breast cancer progression.A cohort of 512 women without known diabetes, whohad early-stage breast cancer were studied for the prog-nostic associations of insulin [7]. IGF-1 and IGFBP-3were assayed using the ELISA. No significant prognos-tic effects were seen for both IGF-1 and IGFBP-3.

In a new study, IGF-1 and IGFBP-3 were measured in 582plasma samples in breast cancer patients [10]. Patients hadmostly nonmetastatic (99 %) and distributed equally as node-negative and node-positive. IGF-1 levels were higher in pa-tients not receiving hormonal therapy compared to patientsreceiving hormonal therapy. As expected, IGFBP-3 concen-trations were not different according to received hormonaltherapy. IGF-1 and IGFBP-3 were significantly related to eachother. Both of these levels were lower at higher ages atdiagnosis and postmenopausal status. No significant effectof IGF-1 levels on survival was found. However, higherIGFBP-3 levels were related to a favorable outcome in uni-variate analysis. Moreover, association of IGFBP-3 with sur-vival were seen both in patients treated with and withoutadjuvant hormonal therapy. In a multivariate Cox regressionmodel, IGFBP-3 was not independently related to OS.Additionally, high IGF-1 levels were related to worse OS inpatients treated with hormonal therapy.

In this trial, we studied serum IGF-1 and IGFBP-3 levels in96 patients with a pathologically confirmed diagnosis ofbreast cancer and age- and sex-matched 30 healthy controlsdetermined by the solid-phase sandwich ELISA methods. Nosignificant difference in baseline serum both IGF-1 andIGFBP-3 levels between breast cancer patients and healthysubjects. None of the prognostic parameters analyzed wascorrelated significantly with the serum assay concentrations.Moreover, no correlations were also found between theseserum concentrations and response to chemotherapy.Likewise, no significant correlation between serum IGF-1and IGFBP-3 levels was found in breast cancer patients. Thepatients with elevated serum IGF-1 levels had favorable in OSthan those with lower levels. However, serum IGFBP-3 con-centrations were found no prognostic role for outcome.

In conclusion, although little is known, evidence to datesuggests that the IGF family may be involved in the etiologyand progression of breast cancer. However, there are muchconflicting evidences in the literature regarding the patterns ofexpression of these gene products; therefore, the precise func-tional relevance of these alterations is not yet well understood.Similar comments were true for patterns of quantifying of the

circulating serum IGF concentrations. The small sample sizeof our study could be considered as significant limitation andmight have influenced these results. However, our study con-tributes to the literature. A standardized method remains to beestablished and validated in larger series of patients in pro-spective studies to determine the potential clinical significanceof these assays in breast cancer patients.

Conflicts of interest None

References

1. Renehan AG, ZwahlenM,Minder C, O'Dwyer ST, Shalet SM, EggerM. Insulin-like growth factor (IGF)-I, IGF binding protein-3, andcancer risk: systematic review and meta-regression analysis. Lancet.2004;363:1346–53.

2. Key TJ, Appleby PN, Reeves GK, Roddam AW. Insulin-like growthfactor 1 (IGF1), IGF binding protein 3 (IGFBP3), and breast cancerrisk: pooled individual data analysis of 17 prospective studies. LancetOncol. 2010;11:530–42.

3. Rinaldi S, Peeters PH, Berrino F, Dossus L, Biessy C, Olsen A, et al.IGF-1, IGFBP-3 and breast cancer risk in women: the EuropeanProspective Investigation into Cancer And Nutrition (EPIC).Endocrine Rel Cancer. 2006;13:593–605.

4. Al-Delaimy WK, Flatt SW, Natarajan L, Laughlin GA, Rock CL,Gold EB, et al. IGF1 and risk of additional breast cancer in theWHEL study. Endocrine Rel Cancer. 2011;18:235–44.

5. Holdaway IM, Mason BH, Lethaby AE, Singh V, Harvey VJ,Thompson PI, et al. Serum insulin-like growth factor-1 and insulin-like growth factor binding protein-3 following chemotherapy foradvanced breast cancer. ANZ J Surg. 2003;73:905–8.

6. Hankinson SE, Willett WC, Colditz GA, Hunter DJ. Circulatingconcentrations of insulin-like growth factor-1 and risk of breastcancer. Lancet. 1998;351:1393–6.

7. Goodwin PJ, Ennis M, Pritchard KI, Trudeau ME, Koo J, MadarnasY, Hartwick W, Hoffman B, Hood N. Fasting ınsulin and outcome inearly-stage breast cancer: results of a prospective cohort study. J ClinOncol. 2002;20: 42–51

8. Pasanisi P, Venturelli E, Morelli D, Fontana L, Secreto G, Berrino F.Serum insulin-like growth factor-1 and platelet-derived growth factoras biomarkers of breast cancer prognosis. Cancer EpidemiolBiomarkers Prev. 2008;17:1719–22.

9. Vadgama JV, Wu Y, Datta G, Khan H, Chillar R. Plasmainsulin-like growth factor-1 and serum IGF-binding protein 3can be associated with the progression of breast cancer, andpredict the risk of recurrence and the probability of survivalin African-American and Hispanic women. Oncology.1999;57:330–40.

10. Hartog H, Boezen HM, de Jong MM, Schaapveld M, Wesseling J,van der Graaf WTA. Prognostic value of insulin-like growth factor 1and insulin-like growth factor binding protein 3 blood levels in breastcancer. Breast. 2013;22:1155–60.

11. Oh Y. IGF-independent regulation of breast cancer growth by IGFbinding proteins. Breast Cancer Res Treat. 1998;47:283–93.

12. Figueroa JA, Jackson JG, McGuire WL, Krywicki RF, Yee D.Expression of insulin-like growth factor binding proteins in humanbreast cancer correlates with estrogen receptor status. J Cell Biochem.1993;52:196–205.

Tumor Biol.