serum insulin-like growth factor-i and insulin-like growth factor binding protein-3 following...

4
ANZ J. Surg. 2003; 73 : 905–908 ORIGINAL ARTICLE Original Article SERUM INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-3 FOLLOWING CHEMOTHERAPY FOR ADVANCED BREAST CANCER IAN M. HOLDAWAY, BARBARA H. MASON, ANNE E. LETHABY, VIJAY SINGH, VERNON J. HARVEY, PAUL I. THOMPSON AND BARRIE D. EVANS Departments of Endocrinology and Oncology, Auckland Hospital, Auckland, New Zealand Background: Insulin-like growth factor-I (IGF-I) and IGF binding protein-3 (IGFBP-3) appear to influence the growth of breast cancer cells in vitro, and epidemiological studies suggest higher serum IGF-I levels increase the risk of breast cancer. IGF-I and IGFBP-3 have therefore been measured in women with advanced breast cancer to determine if changes in serum levels predict the response to treatment by chemotherapy. Methods: Serum IGF-I and IGFBP-3 levels were measured in 14 patients before and after 1 week of chemotherapy. Changes in serum levels were compared with duration of survival. Results: Mean basal serum levels of IGF-I and IGFBP-3 were not significantly different between patients with advanced breast cancer and controls or women with early breast cancer. Serum IGFBP-3 fell significantly 1 week after initiation of chemotherapy. Patient survival was not significantly related to baseline IGF-I or IGFBP-3 levels, but when the fall in serum levels 1 week after starting treatment was expressed either as absolute change or as a percentage of baseline, those individuals with a decrease in IGFBP-3 greater than the median had significantly poorer survival (median survival 5.5 months vs 18 months). These results were independent of other prognostic variables such as previous disease-free survival, and were also unaffected by the change in serum albumin with treatment. The fall in IGF-I and IGFBP-3 with chemotherapy mainly occurred in those with hepatic metastases, but prediction of survival was explained solely by the extent of the fall in IGFBP-3. Conclusions: This preliminary study has shown that serum IGFBP-3 falls significantly following initiation of chemotherapy and the extent of reduction significantly predicts the response to treatment. Key words: breast cancer, chemotherapy, insulin-like growth factor-I, insulin-like growth factor binding protein-3. Abbreviations: CMF, cyclophosphamide, methotrexate and 5-fluorouracil; IGF-I, insulin-like growth factor-I; IGFBP-3, insulin-like growth factor binding protein-3. INTRODUCTION Insulin-like growth factor-I (IGF-I) circulates bound to binding proteins, mainly insulin-like growth factor binding protein-3 (IGFBP-3), and appears to be important in the regulation of growth of normal cells and resistance to apoptosis in neoplastic tissue. There is debate as to whether breast carcinoma cells themselves elaborate mRNA for IGF-I or synthesize IGF-I peptide. 1 Insulin- like growth factor binding proteins are produced by numerous breast cancer cell lines in vitro, 2,3 and the binding capacity of the major binding protein, IGFBP-3, may be altered by specific serum proteases. 4–7 Insulin-like growth factor binding protein-3 may exert an inhibitory influence on breast cancer growth. 8 Studies in early breast cancer have shown serum IGF-I to be either increased or unchanged compared with control subjects, 9–11 and epidemiological studies have suggested increased risk of breast cancer in those with higher serum levels of IGF-I and lower concentrations of IGFBP-3. 12 Several investigators have examined the relationship between survival and IGF-I levels. High breast tumour content of IGF-I receptor has been associated with significantly longer patient survival, 13 whereas increased IGFBP-4 has indicated poorer survival. 14 However, Bhatavdekar et al. could not demonstrate a significant relationship between basal serum IGF-I concentrations and survival in patients with advanced breast cancer. 15 Endocrine therapy with tamoxifen appears to reduce serum IGF-I levels in breast cancer patients. 16 A preliminary study also shows that the extent of tamoxifen-induced suppression of serum IGF-I correlates with clinical response in patients with metastatic breast cancer. 17 Blockade of IGF-I receptors in breast cancer cells cultured in vitro enhances the response to chemotherapy, 18 and adjuvant chemotherapy in breast cancer patients caused a reduction in serum IGF-I levels. 19 The relationship between chemotherapy-induced changes in IGF-I and survival is, however, uncertain. We therefore studied the relationship between basal serum IGF-I and IGFBP-3 levels and known prognostic variables in patients with advanced breast cancer, and assessed the influence of chemotherapy on these measurements, particularly the relationship with survival. METHODS Fourteen patients with advanced breast cancer about to undergo chemotherapy were recruited for the study and informed consent I. M. Holdaway MD, FRACP; B. H. Mason BSc; A. E. Lethaby MA; V. Singh BSc; V. J. Harvey MD, FRACP; P. I. Thompson MD, FRACP; B. D. Evans MD, FRACP. Correspondence: Ian M. Holdaway, Department of Endocrinology, Auckland Hospital, Private Bag 92024, Auckland, New Zealand. Email: [email protected] Accepted for publication 19 June 2003.

Upload: ian-m-holdaway

Post on 06-Jul-2016

214 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: SERUM INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-3 FOLLOWING CHEMOTHERAPY FOR ADVANCED BREAST CANCER

ANZ J. Surg.

2003;

73

: 905–908

ORIGINAL ARTICLE

Original Article

SERUM INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-3 FOLLOWING CHEMOTHERAPY

FOR ADVANCED BREAST CANCER

I

AN

M. H

OLDAWAY

, B

ARBARA

H. M

ASON

, A

NNE

E. L

ETHABY

, V

IJAY

S

INGH

, V

ERNON

J. H

ARVEY

, P

AUL

I. T

HOMPSON

AND

B

ARRIE

D. E

VANS

Departments of Endocrinology and Oncology, Auckland Hospital, Auckland, New Zealand

Background:

Insulin-like growth factor-I (IGF-I) and IGF binding protein-3 (IGFBP-3) appear to influence the growth of breastcancer cells

in vitro

, and epidemiological studies suggest higher serum IGF-I levels increase the risk of breast cancer. IGF-I andIGFBP-3 have therefore been measured in women with advanced breast cancer to determine if changes in serum levels predict theresponse to treatment by chemotherapy.

Methods:

Serum IGF-I and IGFBP-3 levels were measured in 14 patients before and after 1 week of chemotherapy. Changes inserum levels were compared with duration of survival.

Results:

Mean basal serum levels of IGF-I and IGFBP-3 were not significantly different between patients with advanced breastcancer and controls or women with early breast cancer. Serum IGFBP-3 fell significantly 1 week after initiation of chemotherapy.Patient survival was not significantly related to baseline IGF-I or IGFBP-3 levels, but when the fall in serum levels 1 week afterstarting treatment was expressed either as absolute change or as a percentage of baseline, those individuals with a decrease in IGFBP-3greater than the median had significantly poorer survival (median survival 5.5 months

vs

18 months). These results were independentof other prognostic variables such as previous disease-free survival, and were also unaffected by the change in serum albumin withtreatment. The fall in IGF-I and IGFBP-3 with chemotherapy mainly occurred in those with hepatic metastases, but prediction ofsurvival was explained solely by the extent of the fall in IGFBP-3.

Conclusions:

This preliminary study has shown that serum IGFBP-3 falls significantly following initiation of chemotherapy andthe extent of reduction significantly predicts the response to treatment.

Key words: breast cancer, chemotherapy, insulin-like growth factor-I, insulin-like growth factor binding protein-3.

Abbreviations

: CMF, cyclophosphamide, methotrexate and 5-fluorouracil; IGF-I, insulin-like growth factor-I; IGFBP-3,insulin-like growth factor binding protein-3.

INTRODUCTION

Insulin-like growth factor-I (IGF-I) circulates bound to bindingproteins, mainly insulin-like growth factor binding protein-3(IGFBP-3), and appears to be important in the regulation of growthof normal cells and resistance to apoptosis in neoplastic tissue.There is debate as to whether breast carcinoma cells themselveselaborate mRNA for IGF-I or synthesize IGF-I peptide.

1

Insulin-like growth factor binding proteins are produced by numerousbreast cancer cell lines

in vitro

,

2,3

and the binding capacity of themajor binding protein, IGFBP-3, may be altered by specific serumproteases.

4–7

Insulin-like growth factor binding protein-3 may exertan inhibitory influence on breast cancer growth.

8

Studies in early breast cancer have shown serum IGF-I to beeither increased or unchanged compared with control subjects,

9–11

and epidemiological studies have suggested increased risk ofbreast cancer in those with higher serum levels of IGF-I and

lower concentrations of IGFBP-3.

12

Several investigators haveexamined the relationship between survival and IGF-I levels.High breast tumour content of IGF-I receptor has been associatedwith significantly longer patient survival,

13

whereas increasedIGFBP-4 has indicated poorer survival.

14

However, Bhatavdekar

et al

. could not demonstrate a significant relationship betweenbasal serum IGF-I concentrations and survival in patients withadvanced breast cancer.

15

Endocrine therapy with tamoxifen appears to reduce serum IGF-Ilevels in breast cancer patients.

16

A preliminary study also showsthat the extent of tamoxifen-induced suppression of serum IGF-Icorrelates with clinical response in patients with metastatic breastcancer.

17

Blockade of IGF-I receptors in breast cancer cells cultured

in vitro

enhances the response to chemotherapy,

18

and adjuvantchemotherapy in breast cancer patients caused a reduction in serumIGF-I levels.

19

The relationship between chemotherapy-inducedchanges in IGF-I and survival is, however, uncertain. We thereforestudied the relationship between basal serum IGF-I and IGFBP-3levels and known prognostic variables in patients with advancedbreast cancer, and assessed the influence of chemotherapy on thesemeasurements, particularly the relationship with survival.

METHODS

Fourteen patients with advanced breast cancer about to undergochemotherapy were recruited for the study and informed consent

I. M. Holdaway

MD, FRACP;

B. H. Mason

BSc;

A. E. Lethaby

MA;

V. Singh

BSc;

V. J. Harvey

MD, FRACP;

P. I. Thompson

MD, FRACP;

B. D. Evans

MD, FRACP.

Correspondence: Ian M. Holdaway, Department of Endocrinology,Auckland Hospital, Private Bag 92024, Auckland, New Zealand.Email: [email protected]

Accepted for publication 19 June 2003.

Page 2: SERUM INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-3 FOLLOWING CHEMOTHERAPY FOR ADVANCED BREAST CANCER

906 HOLDAWAY

ET AL

.

was obtained. One patient had received a brief course of mito-zantrone 2 years prior to the study, and one had received severalcycles of cyclophosphamide, methotrexate and 5-fluorouracil(CMF) chemotherapy 5 years prior to the present treatment. Theage range of the patient group was 43–76 years, with a meanage of 55 years, and the weight range was 63–94 kg, with a meanof 75 kg. Blood samples were obtained at baseline and on day 7of their first chemotherapy cycle. Twelve women receivedoral cyclophosphamide (100 mg/m

2

) daily for 14 days, intrave-nous methotrexate (40 mg/m

2

) on days 1 and 8, 5-fluorouracil(600 mg/m

2

) on days 1 and 8, and prednisone (40 mg) daily for14 days. The chemotherapy was repeated every 4 weeks. Twopatients, one with cerebral secondaries, received CMF and dexa-methasone, and another received mitozantrone on a 3-weeklycycle and subsequently intravenous CMF on a 3-weekly cycle.The mean number of CMF cycles was 4 (range 1–6). Eightpatients subsequently received further chemotherapy. Liver meta-stases were sought in all patients by high resolution transabdomi-nal ultrasound. Serum was also obtained from 31 patients withearly breast cancer (recent tumour diagnosis with no evidence ofmetastases before surgical removal) and control women, matchedfor age with the study group. Serum was frozen at –20

°

C untilassay for IGF-I, IGFBP-3 and albumin. The protocol wasapproved by the North Health Medical Ethics Committee.

Measurements

Serum IGF-I was measured by radioimmunoassay followingextraction using acid/ethanol cryoprecipitation, using a rabbitantiserum to rh-metIGF-I at a final titre of 1 : 150 000. Recom-binant hIGF-I was iodinated by a modified chloramine-Tmethod.

20

The assay sensitivity was 2

µ

g/L, the intra-assay andinterassay coefficients of variation were 5% and 7%, respec-tively, and the normal range was 90–300

µ

g/L. Serum IGFBP-3was measured using a commercial kit (Bioclone Australia Ltd,Marrickville, NSW, Australia), with an assay sensitivity of0.7 mg/L, intra-assay and interassay coefficients of variation of5% and 6%, respectively, and a normal range of 2.2–5.6 mg/L.Serum albumin was measured using a commercial kit (Boeh-ringer Mannheim, Mt Wellington, Auckland, New Zealand). Theinterassay variation was 1.5% and the detection limit was 2 g/L,with a normal range of 35–50 g/L.

Statistical analysis

Univariate survival curves were calculated using the product limitestimates method together with the log–rank test for comparingdifferences. Multivariate survival analysis was performed usingCox’s proportional hazards model to assess independent effectsof changes in serum IGF-I measurements, controlling for patho-logical and clinical prognostic factors known to affect outcome.Differences between mean group levels were assessed by

ANCOVA

with age as the covariate. Student’s

t

-test was used to test ifthe mean change between paired samples was different from zero.Repeated measures

ANCOVA

was used to show changes in serumvariables during the first cycle of chemotherapy.

Survival analyses were performed comparing survival of thosewith serum levels above or below the median basal value meas-ured before chemotherapy, or by comparing survival betweengroups where the percentage and absolute change in IGF-I andIGFBP-3 values was above or below the median at the end of thefirst week following treatment.

RESULTS

The mean basal serum IGF-I and IGFBP-3 values for womenwith advanced breast cancer (before treatment with chemo-therapy) are shown in Table 1, compared with measurementsobtained in controls or women before surgical treatment of earlystage breast cancer. There was no significant difference in theIGF-I or IGFBP-3 values between the groups.

When the patients with advanced breast cancer were sub-divided according to recognized prognostic factors (disease-freeinterval and liver metastases), there was also no significant differ-ence in IGF-I or IGFBP-3 between groups.

Serum IGF-I and IGFBP-3 concentrations before and 1 weekafter initial treatment with chemotherapy are shown in Table 2.There was a significant fall in the serum concentration of IGFBP-31 week after commencing chemotherapy.

In order to assess the influence of baseline IGF-I and IGFBP-3on survival, patients were subdivided into groups above andbelow the median value for these variables and survival curveswere constructed (Fig. 1). There was no significant difference insurvival between the different subgroups, although there was aconsistent trend for those with values above the median to havelonger survival. By comparison with results using baseline meas-urements, there was a significant difference in survival accordingto whether the percentage change in serum IGFBP-3 values overthe first week following chemotherapy was above or below themedian for the group (Fig. 2), with survival reduced in thosehaving the greater proportional fall in IGFBP-3. Those with apercentage fall in serum IGFBP-3 greater than the median had amean survival of 5.5 months compared with 18 months for thoseless than the median. In addition to the findings when measure-

Table 1.

Serum insulin-like growth factor-I (IGF-I) and insulin-likegrowth factor binding protein-3 (IGFBP-3) concentrations (mean

±

SD)in controls, women with early and advanced breast cancer, andaccording to standard prognostic variables

Patient group

n

IGF-I (

µ

g/L) IGFBP-3 (mg/L)

Normal subjects 14 115

±

39 2.78

±

0.38Early breast cancer 31 128

±

40 2.7

±

0.51Advanced breast cancer 14 135

±

66 3.15

±

1.17Liver metastases

Present 7 106

±

47 2.9

±

1.0Absent 7 164

±

73 3.4

±

1.4Disease-free interval

<1 year 5 102

±

66 2.5

±

0.9>1 year 9 153

±

62 3.5

±

1.2

Table 2.

Mean serum insulin-like growth factor-I (IGF-I) and insulin-like growth factor binding protein-3 (IGFBP-3) concentrations beforeand 1 week after chemotherapy in 14 patients with advanced breastcancer

Basal After chemotherapy

% Change

n

14 14IGF-I (

µ

g/L) 135 139 +4IGFBP-3 (mg/L) 3.2 2.8* –14

*P

0.05 compared with baseline.

Page 3: SERUM INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-3 FOLLOWING CHEMOTHERAPY FOR ADVANCED BREAST CANCER

BREAST CANCER, IGF-I AND CHEMOTHERAPY 907

ments were stratified by the median value, there was also a signif-icant inverse linear relationship between length of survival andthe extent of the fall in IGFBP-3. In a Cox’s regression analysis,neither the previous disease-free survival nor the presence orabsence of liver secondaries significantly influenced the results,and the fall in serum IGFBP-3 was the dominant factor predictingsurvival.

Liver metastases are usually associated with poor outcomein breast cancer but, surprisingly, in the present study the pres-ence of liver metastases did not significantly predict survival(

P

= 0.08). The changes in serum IGF-I and IGFBP-3 concentra-tion 1 week after receiving chemotherapy are shown according tothe presence or absence of liver metastases in Table 3, expressedeither as mean values or as the percentage change from baseline.There was a significant fall in the levels of IGF-I and IGFBP-3concentrations 1 week after receiving chemotherapy in those withliver metastases, whereas there was no significant change in thoseapparently free of liver tumour. The fall in IGFBP-3 with chemo-therapy in the overall group appeared to be predominantly inthose with hepatic metastases.

To exclude a possible non-specific effect of chemotherapy onthe synthesis or turnover of serum IGF-I or IGFBP-3 as an expla-nation for the present findings, the change in serum albumin wasassessed following chemotherapy. In contrast to the findings withIGF-I and IGFBP-3, neither basal albumin nor change in albuminat 1 week significantly related to survival using Cox’s regressionmodel or the log–rank test comparing changes above and belowthe median value (Fig. 3).

DISCUSSION

In the present study, women with advanced breast cancer had basallevels of IGF-I and IGFBP-3 that were indistinguishable from

Fig. 1.

Survival of women with advanced breast cancer from com-mencement of chemotherapy according to basal serum values ofinsulin-like growth factor-I (IGF-I) and insulin-like growth factorbinding protein-3 (IGFBP-3) before treatment (

n

= 14). Solid line,group with baseline measurement above median; dashed line, groupwith baseline measurement below median.

Fig. 2.

Survival of women with advanced breast cancer from commence-ment of chemotherapy in groups divided according to the percentagechange in serum insulin-like growth factor-I (IGF-I) or insulin-likegrowth factor binding protein-3 (IGFBP-3) after 1 week of chemo-therapy (

n

= 14). Solid line, group with baseline measurement abovemedian; dashed line, group with baseline measurement below median.

Table 3.

Mean serum insulin-like growth factor-I (IGF-I) and insulin-like growth factor binding protein-3 (IGFBP-3) concentrations before and1 week after chemotherapy for women with and without liver secondaries

Liver metastases No liver metastases Basal After

chemotherapy% Change Basal After

chemotherapy% Change

n

7 7 7 7IGF-I (

µ

g/L) 106 95* –9* 164 182 +17IGFBP-3 (mg/L) 2.9 2.2*** –30** 3.4 3.4 –2

*P

= 0.02; **

P

= 0.01; ***

P

= 0.002.

Fig. 3.

Survival of women with advanced breast cancer from com-mencement of chemotherapy in groups divided according to thepercentage change in serum albumin after 1 week of chemotherapy(

n

= 14). Solid line, group with baseline measurement above median;dashed line, group with baseline measurement below median.

Page 4: SERUM INSULIN-LIKE GROWTH FACTOR-I AND INSULIN-LIKE GROWTH FACTOR BINDING PROTEIN-3 FOLLOWING CHEMOTHERAPY FOR ADVANCED BREAST CANCER

908 HOLDAWAY

ET AL

.

control subjects. However, the serum concentration of IGFBP-3 fellsignificantly 1 week after commencement of chemotherapy. Thisfall occurred predominantly in those with hepatic metastases. Theextent of the reduction in IGFBP-3 was inversely related to sur-vival. Serum IGF-I did not change significantly with chemotherapyin the overall group, although there was a small reduction in meanIGF-I following treatment in those with hepatic metastases.

There was no significant relationship between basal serum IGF-Iand survival in the present study, confirming the results of Bhatav-detar

et al

.,

15

and basal serum IGFBP-3 concentrations also did notappear to predict survival. Survival was decreased in patientsshowing a greater percentage fall of IGFBP-3 with chemotherapy.This could reflect non-specific effects on hepatic protein synthesisin those with greater tumour burden, but the observation thatchanges in other proteins such as IGF-I and albumin did not relateto survival suggests that this finding is unlikely to be due to a non-specific result of chemotherapy or an overall decrease in proteinsynthesis with disease progression or treatment. A fall in IGF-I fol-lowing tamoxifen treatment of breast cancer has been associatedwith improved survival.

17

A fall in IGFBP-3 could increase the bio-availability of free unbound IGF-I, with resulting stimulation oftumour growth from an increase in the systemic pool of bioavail-able IGF. The effect of chemotherapy on proteolysis of IGF-I andIGFBP-3 has yet to be studied, but could also contribute to the sur-vival changes noted in this study. Serum IGFBP-3 may also exertdirect antitumour effects,

21

so that a fall in serum levels could beassociated with disease progression and reduced survival.

Women with liver metastases appeared particularly likely tohave significant falls in the concentration of IGFBP-3, and to alesser extent IGF-I, following chemotherapy. This could resultfrom enhanced hepatotoxicity with chemotherapy in women withhepatic tumour spread. However, in the present study, the fall inIGFBP-3 concentration with treatment was a better predictor ofsurvival than the presence of liver metastases alone.

It is important to note that the present results were obtained in arelatively small patient group. This should therefore be considereda preliminary investigation and further study is required to confirmthese findings. It remains possible that small changes in IGF-I withchemotherapy in the total group could have been missed becauseof the small number of individuals studied, but there was no trendapparent in the IGF-I values to support this possibility (Table 2).Conversely, the reduction in IGFBP-3 with chemotherapy, par-ticularly in those with hepatic metastases, represented a large per-centage change that was highly statistically significant and thatis unlikely to be a type 1 error. Based on the results to date, womenwith metastatic breast cancer in whom IGFBP-3 levels fall dramat-ically in the first week of standard chemotherapy treatment couldbe considered for alternative or more intensive therapy, as theirprognosis appears to be especially poor.

ACKNOWLEDGEMENTS

This study was funded by a research grant from the CancerSociety of New Zealand.

REFERENCES

1. Cullen KJ, Smith HS, Hill S, Rosen N, Lippman ME. Growthfactor messenger RNA expression by human fibroblasts frombenign and malignant lesions.

Cancer Res.

1991;

51

: 497–85.2. Shemer J, Yaron A, Werner H

et al.

Regulation of insulin-likegrowth factor (IGF) binding protein-5 in the T47D human breastcarcinoma cell line by IGF-I and retinoic acid.

J. Clin. Endocri-nol. Metab.

1993;

77

: 1246–50.

3. Sheikh MS, Shao ZM, Hussain A

et al.

Regulation of insulin-like growth factor-binding protein-1, 2, 3, 4, 5 and 6: synthesis,secretion, and gene expression in estrogen receptor-negativehuman breast carcinoma cells.

J. Cell Physiol.

1993;

155

:556–67.

4. Giudice LC, Farrell EM, Pham H, Lamson G, Rosenfeld RG.Insulin-like growth factor binding proteins in maternal serumthroughout gestation and in the puerperium: effects of a pregnancy-associated serum protease activity.

J. Clin. Endocrinol. Metab.

1990;

71

: 806–16.5. Davenport ML, Isley WL, Pucilowska JB

et al.

Insulin-likegrowth factor binding protein-3 proteolysis is induced after elec-tive surgery.

J. Clin. Endocrinol. Metab.

1992;

75

: 590–5.6. Cwyfan Hughes SC, Cotterill AM, Molloy AR

et al.

The induc-tion of specific proteases for insulin-like growth factor-bindingproteins following major heart surgery.

J. Endocrinol.

1992;

135

: 135–45.7. Suikkari A-M, Baxter RC. Insulin-like growth factor (IGF)

binding protein-3 in pregnancy serum binds native IGF-I but notiodo-IGF-I.

J. Clin. Endocrinol. Metab.

1991;

73

: 1377–9.8. Schofield PN, Lee A, Hill DJ, Cheetham JE, James D, Stewart C.

Tumour suppression associated with expression of humaninsulin-like growth factor II.

Br. J. Cancer

1991;

63

: 687–92.9. Barni S, Lissoni P, Brivio F et al. Serum levels of insulin-like

growth factor-I in operable breast cancer in relation to the mainprognostic variables and their perioperative changes in relationto those of prolactin. Tumori 1994; 80: 212–15.

10. Peyrat JP, Bonneterre J, Hecquet B et al. Plasma insulin-likegrowth factor-I (IGF-I) concentrations in human breast cancer.Eur. J. Cancer 1993; 29A: 492–7.

11. Favoni RE, de Cupis A, Perrotta A et al. Insulin-like growthfactor-I (IGF-I) and IGF-binding proteins blood serum levels inwomen with early- and late-stage breast cancer: mutual relation-ship and possible correlations with patients’ hormonal status.J. Cancer Res. Clin. Oncol. 1995; 121: 674–82.

12. Hankinson SE, Willett WC, Colditz GA et al. Circulating con-centrations of insulin-like growth factor-I and risk of breastcancer. Lancet 1998; 351: 1393–6.

13. Bonneterre JP, Peyrat R, Demaille A. Prognostic significance ofinsulin-like growth factor I receptors in human breast cancer.Cancer Res. 1990; 50: 6931–5.

14. Yee D, Sharma J, Hilsenbeck SG. Prognostic significance ofinsulin-like growth factor-binding protein expression in axillarylymph node-negative breast cancer. J. Natl Cancer Inst. 1994;86: 1785–9.

15. Bhatavdekar JM, Patel DD, Karelia NH et al. Tumor markers inpatients with advanced breast cancer as prognosticators: a pre-liminary study. Breast Cancer Res. Treat. 1994; 30: 293–7.

16. Pollak M, Constantino J, Polychronakos C et al. Effect oftamoxifen on serum insulin-like growth factor I levels in stage 1breast cancer patients. J. Natl Cancer Inst. 1990; 82: 1693–7.

17. Kiang DT, Kollander R, Kiang B, Kao PC. Role of plasma IGF-Iin endocrine therapy for breast cancer. Proc. Am. Soc. Clin.Oncol. 1992; 11: 51, [abstract 31].

18. Beech DJ, Parekh N, Pang Y. Insulin-like growth factor-I recep-tor antagonism results in increased cytotoxicity of breast cancercells to doxorubicin and taxol. Oncol. Rep. 2002; 8: 325–9.

19. Kajdaniuk D, Marek B. Influence of adjuvant chemotherapy withcyclophosphamide, methotrexate and 5-fluorouracil on plasmainsulin-like growth factor-I and chosen hormones in breast cancerpremenopausal patients. J. Clin. Pharm. Ther. 2000; 25: 67–72.

20. Gluckman PD, Johnson-Barret TJ, Butler J, Edgar BW, GunnTR. Studies of insulin-like growth factors-I and -II by specificradioligand assays in umbilical cord blood. Clin. Endocrinol.1983; 19: 405–13.

21. Holdaway IM, Mason BH, Lethaby AE et al. Serum levels ofinsulin-like growth factor binding protein-3 in benign and malig-nant breast disease. Aust. N. Z. J. Surg. 1999; 69: 495–500.