research article the association between treatment for

Research ArticleThe Association between Treatment forMetabolic Disorders and Breast Cancer Characteristics

Hadar Goldvaser1 Shulamith Rizel12 Daniel Hendler1

Victoria Neiman1 Daniel Shepshelovich23 Tzippy Shochat4

Aaron Sulkes12 Baruch Brenner12 and Rinat Yerushalmi12

1 Institute of Oncology Davidoff Cancer Center Beilinson Hospital Rabin Medical Center 39 Jabotinski St Petah Tikva Israel2Sackler Faculty of Medicine Tel Aviv University PO Box 39040 Tel Aviv Israel3Department of Medicine A Beilinson Hospital Rabin Medical Center 39 Jabotinski St Petah Tikva Israel4Statistical Consulting Unit Beilinson Hospital Rabin Medical Center 39 Jabotinski St Petah Tikva Israel

Correspondence should be addressed to Hadar Goldvaser hadar7ggmailcom

Received 17 March 2016 Revised 16 July 2016 Accepted 9 August 2016

Academic Editor Andrea Tura

Copyright copy 2016 Hadar Goldvaser et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Purpose To evaluate the associations between metformin insulin statins and levothyroxine and breast cancer characteristics andoutcomeMethods Retrospective chart review of patients treated in our institute for early estrogen receptor (ER) positive humanepidermal growth factor receptor 2 negative breast cancer whose tumors were sent to Oncotype DX (ODX) analysis Patientswere grouped according to medications usage during the time of breast cancer diagnosis Each group was compared to the rest ofthe study population Results The study cohort included 671 patients Sixty (91) patients were treated with metformin 9 (14)with insulin 208 (317) with statins and 62 (94) with levothyroxine Patients treated with metformin had more intense ERstain (119901 = 0032) and a lower ODX recurrence score (RS) (119901 = 0035) Diagnosis of diabetes mellitus was also associated withlower ODX RS (119901 = 0014) Insulin usage was associated with a higher rate of angiolymphatic invasion (119901 = 0041) but lowerKi67 (119901 = 0017) Levothyroxine usage was associated with different histological subtype distribution (119901 = 002) Extendedlevothyroxine usage was associated with lower ODXRS (119901 = 0005) Statin usage had no impact on tumor characteristics Outcomewas comparable in the studied subgroups Conclusions Common medications for metabolic disorders might be associated withbreast cancer characteristics

1 Introduction

Breast cancer is a heterogonous disease Treatment andoutcome are influenced by various parameters includingtumor size nodal involvement grade estrogen receptor (ER)progesterone receptor (PR) and human epidermal growthfactor receptor 2 (HER2) statuses as well as Ki67 percentageand the presence of angiolymphatic invasion [1ndash8] All ofthese parameters are well established prognostic factors fordisease recurrence Modern oncology has added to theseparameters the multigene assays such as Oncotype DX(ODX) which provides additional prognostic and predictiveinformation based on a real time polymerase chain reaction(RT-PCR) of 21 genes A mathematical formula which uses

the quantified results generates a score that predicts distantrelapse and provides information regarding the potentialbenefit of adjuvant chemotherapy in early ER positive andHER2 negative breast cancer [9 10]

The associations between different metabolic disordersespecially obesity and diabetes mellitus (DM) and breastcancer are well established [11ndash16] However data regardingthe impact of drugs commonly used to treat these conditionson breast cancer are scarce

Metformin is a commonly used oral antidiabetic agentthat decreases hyperinsulinemia As hyperinsulinemia andinsulin-like growth factors were reported to have mitogeniceffect on breast cells [17 18] metformin may favorablyaffect patients with breast cancer Many in vitro and in vivo

Hindawi Publishing CorporationInternational Journal of EndocrinologyVolume 2016 Article ID 4658469 9 pageshttpdxdoiorg10115520164658469

2 International Journal of Endocrinology

studies found that metformin inhibits cancer cell growthreduces the risk of developing solid tumors and improvescancer outcome including breast cancer [11 19ndash23] Thereduced risk of breast cancer associated with metformin usewas clearly demonstrated in a recent large epidemiologicalstudy [24] Furthermore metformin usage was associatedwith a threefold greater pathologic complete response indiabetic patients with breast cancer receiving neoadjuvantchemotherapy compared to patients with DM not treatedwith metformin [25] In addition a meta-analysis by Xu etal found that breast cancer patients with DM who weretreated with metformin had significantly decreased risk ofall-cause and cancer specific mortality compared with theircounterparts who did not receive metformin [26] In light ofthese data the drug is currently being investigated for adju-vant breast cancer treatment of nondiabetic patients (httpsclinicaltrialsgovct2showNCT01101438 NCIC CTGMA32study Goodwin PJ et al) On the other hand several reportsdid not support the positive impact of metformin on breastcancer [21 27ndash30]

The activity of insulin as a growth factor has raisedconcern regarding the stimulation of neoplastic growth byinsulin analogues Data regarding the influence of insulinanalogues on cancer incidence are inconsistent One meta-analysis found that diabetic patients treated with insulin hadan increased cancer risk compared to non-insulin-treateddiabetics [31] while another study reported an inconsistentimpact of insulin on cancer incidence [32] More specificallya meta-analysis by Colmers et al found an increased breastcancer risk in glargine users compared to those who did notuse glargine [33]Thismay be attributed to themitogenic andproliferative activity of glargine due to its increased bindingaffinity to insulin-like growth factor-1 compared to humaninsulin [34] However a pooled analysis of 13 epidemiologicalstudies did not support the association between treatmentwith glargine and increased incidence of breast cancer [34]

There are preclinical data supporting the antineoplas-tic effect of statins Possible mechanisms include reducedexpression of the antiapoptotic protein bcl-xL and increasedtranscription of phosphatase and tensin homolog whichinhibit the oncogenic phosphatidylinositol-3-kinase pathway[35 36] Several in vitro studies demonstrated that statinsdecrease breast cancer cell proliferation [35 37ndash39] A recentmeta-analysis by Zhong et al found that statins usage wasassociated with improved cancer survival [40] Despite thesereports current data do not support lower incidence of breastcancer in patients treated with statins [41ndash43] and it is notclear if they have any effect on outcome [44 45]

Thyroid function is an additional metabolic factor thatmay influence breast cancer Data regarding a correlationbetween higher levels of free T4 (FT4) and triiodothyronineas well as lower levels of thyroid peroxidase autoantibodies(TPO-Ab) and higher breast cancer incidence are accumu-lating [46ndash49] Some data suggest that FT4 and TPO-Ablevels are associated with breast cancer characteristics [48]however the impact of levothyroxine treatment is unclear

The aim of this study was to evaluate the associationof metformin insulin statins and levothyroxine and breastcancer characteristics and outcome With the advent of

ODX as a prognostic and predictive tool for daily practiceimportant molecular information on tumor in addition totraditional pathology report is obtained [9 10] Thereforewe chose to focus on patients with early breast ER positiveHER2 negative disease for whom their physician added theODX test prior to making a decision on systemic treatment

2 Methods

This was a retrospective single center cohort study Patientswith early ER positive HER2 negative breast cancer whowere diagnosed between 42005 and 32012 were includedThe cohort included all patients treated in our institute duringthis period whose tumors were sent for ODX analysis

The medical records of all patients were retrospectivelyreviewed up to 82015 Data on patient demographicsclinical-pathological parameters treatment and outcomewere retrieved Data regarding usage of metformin insulinstatins and levothyroxine at the time of breast cancer diag-nosis were also collected Duration of treatment until breastcancer diagnosis and related comorbidities including DMdyslipidemia and hypothyroidism were retrieved from theClalit Health Service database We evaluated the influence ofthe studied medications on tumor size nodal involvementstage ODX recurrence score (RS) and histological charac-teristics Histological characteristics evaluated were ER PRHER2 grade angiolymphatic and perineural invasion Ki67P53 and histological subtype As angiotensin convertingenzyme inhibitors (ACEI) and angiotensin receptor blockers(ARB) are commonly prescribed in patients with DM andseveral studies suggested a possible relationship betweenthese medications and breast cancer [50 51] we evaluatedtheir usage in this cohort The study protocol was approvedby the institutional ethics committee

Staining for ER PR p53 Ki-67 andHER2was performedusing the Ventana Benchmark XT automated immunostainer(Ventana Tucson AZ USA) with the standard cell condi-tioner (CC1) protocol for 30min Following deparaffinizationand the CC1 protocol ready-to-use ER rabbit monoclonalantibody [anti-ER (6F11) primary antibody Ventana] wasapplied for 40min incubation at 37∘C PR rabbit mono-clonal antibody (clone 16 Novocastra Newcastle UK) wasemployed at a 1 100 dilution with 40min incubation at37∘C Ki 67 rabbit monoclonal antibody (clone SP6 ThermoFisher Scientific) was used at a 1 100 dilution for 40minat 37∘C and ready-to-use PATHWAY HER2 anti-HER2neurabbit monoclonal antibody (4B5) (Ventana) was utilizedwith 32min incubation at 37∘C Slides were counterstainedwith 410158406-diamidino-2-phenylindole (Sigma-Aldrich) andthe stainedmaterial was visualized under a BX51 fluorescencemicroscope (Olympus) The signals were analyzed manually

ER and PR immunohistochemical (IHC) staining usedthemodified version of the119867-scoremethod 1timespercentage ofweakly staining nuclei + 2 times percentage of moderately stain-ing nuclei + 3 times percentage of intensely staining nuclei100yielding a range of 0 to 3 Ki67 staining was reportedby percentage of positively stained nuclei (0 to 100)HER2 negativity was defined as an immunohistochemistrytest score of 0 or 1 If the IHC score equaled 2 HER2

International Journal of Endocrinology 3

Table 1 Tumor burden and Oncotype DX recurrence score

Population (119873) 119879Macroscopicnode positive1 Stage () Oncotype Dx RS

Mean size cm (SD) 1199012 () 1199012 I II III 1199012 Mean (SD) 1199012

All (671) 167 (079) mdash 9 mdash 712 283 05 mdash 1906 (1027) mdashMetformin3 (60) 172 (079) 0621 133 0217 65 35 0 089 164 (852) 0035Insulin3 (9) 203 (06) 0171 111 0574 444 556 0 0727 1767 (1364) 0682Statins3 (208) 177 (095) 0067 92 0994 67 325 05 0226 1811 (103) 0108Levothyroxine3 (62) 161 (068) 0509 65 0464 694 306 0 0949 171 (835) 00621

Macroscopic nodes-lymph node metastases gt 2 millimeter2119901 value refers to comparison of each variable between patients who used the specified medication to the rest of the cohortMedication usage was not available for 15 patientsRS recurrence score3Median duration of medications usage (months range) metformin 505 (1ndash805) insulin 56 (10ndash126) statins 72 (1ndash168) and levothyroxine 113 (8ndash168)

negativity was determined according to fluorescence in situhybridization test per American Society of Clinical Oncologyguidelines at the time of the test

3 Statistical Analysis

Statistical analysis was performed using SAS Software ver-sion 94 119905-test was used to compare the value of continuousvariables between study groups Chi-square (for more thantwo groups) or Fisherrsquos exact test (for two groups) were usedto compare the value of categorical variables between studygroups Pearson correlation was used to assess the associ-ations between continuous variables Continuous variableswere presented by mean plusmn standard deviation (SD) ANOVAtest was used to assess associations between categoricalvariables Categorical variables were presented by (119873 ) Inorder to evaluate the association between medications usageduration and the characteristics that were found significantwe used the duration of usage as a continuous variable

Overall survival (OS) was assessed by Kaplan-Meiersurvival analysis with the log-rank test Breast cancer spe-cific survival (BCSS) and disease-free survival (DFS) wereassessed by the Cox proportional hazards model with theFine andGray correction for noncancer death as a competingrisk Two-sided119901 values less than 005 were considered statis-tically significant To evaluate interaction betweenmetforminand insulin we performed multivariate analysis for tumorcharacteristics which were found statistically significant onunivariate analyses We added interaction of metformin andinsulin to the model 119901 value less than 015 was considered tobe significant for interaction

4 Results

41 Patients andTumorCharacteristics A total of 671 patientswere included in the study Median age was 61 years (range34ndash85) There were 662 women and 9 men Among thewomen 446 (674) were postmenopausal History of inva-sive and noninvasive breast cancer was documented in 112and 19 of the patients respectively Nine percent of thepatients had a history of malignancy of nonbreast origin

Personal history of benign breast disease was noted in 91First and second degree family history of breast cancer orany other cancer were noted in 397 and 58 respectivelyBreast cancer diagnosis was established by screening in themajority of cohort population (826) Screening rates weresimilar among the studied groups

Data regarding patient comorbidities were missing for160 (238) patients History of DM was documented in88 (172) dyslipidemia in 333 (652) hypothyroidism in74 (145) and hyperthyroidism is 6 (12) patients For177 (264) patients data regarding duration of medicationsusage were not available Data regarding medication usagewere missing for 15 patients

Tumor characteristics for the entire population cohortand for each medication group are detailed in Tables 1 and 2As expected there were correlations between the intensity ofER and PR IHC staining and theOncotype ER PR andHER2(according to RT-PCR results 119901 lt 00001 for all variables)Median ODX RS was 17 (range 0ndash88) and 504 384 and112 of the patients had low (0ndash17) intermediate (18ndash30)and high (ge31) RS respectively

42 Impact of Diabetes Mellitus and Related Medications onTumor Characteristics Patients with DM had lower ODXRS (mean 167 plusmn 10 versus 197 plusmn 207 119901 = 0014)Tumor size nodal involvement and other histological char-acteristics were not influenced by diagnosis of DM Atbreast cancer diagnosis 60 (91) patients were treatedwith metformin Median duration usage was 505 (1ndash805)months and interquartile range (IQR) was 22ndash805 monthsMetformin usage was associated with lower ODX RS (mean164 plusmn 85 versus 193 plusmn 104 119901 = 0035) (Figure 1) andmore intense ER staining (mean 259 plusmn 04 versus 246 plusmn058 119901 = 0032) These differences were not associatedwith the duration of metformin treatment Other histologicalcharacteristics as well as tumor size nodal involvement andstage were comparable Nine (14) patients were treatedwith insulin at breast cancer diagnosis Median duration ofinsulin usage was 56 (10ndash126) months and IQR was 15ndash94months Angiolymphatic invasion wasmore common amongthese patients (222 versus 58 119901 = 0041) while the Ki67percentage was lower (mean 95 plusmn 464 versus 16 plusmn 1364


Table2Histologicalcharacteris

tics

Popu

latio

n(119873)

Histolog

yGrade

ERPR

Ki67

()

P53(

)Ang

iolymph

atic

invasio

nPN

I

IDC(

)ILC(

)Other2(

)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Yes()1199011

Yes()1199011

All(671)

809

122

69

mdash202

(059)

mdash247

(057)

mdash14

7(119

)mdash

1591(1358)

mdash645

(1696)

mdash61

mdash45

mdashMetform

in3(60)

767

1583

0683

2(063)

0846

259

(04)

0032

129(105)

0217

137(111)

0204

696

(147)

0828

34

0368

51

0742

Insulin3(9)

778

111

111

0878

186(038)

0474

249

(067)

0948

112(094)

0371

95(464)

0017

85(203)

0765

222

0041

00512

Statins3(208)

798

111

910266

207

(05)

0193

248

(055)

0916

141(10

2)03461678(14

53)0318

645

(1561)0999

780212

49

0768

Levothyroxine3

(62)

694

161

145

002

202

(055)

094

247

(052

)0999

137(127)

04881528(1264

)0727592

(1656)

084

33

0339

160257

1119901valuer

eferstothec

omparis

onof

each

varia

bleb

etweenpatie

ntsw

housed

thes

pecifiedmedicationto

ther

esto

fthe

coho

rt

2Ano

ther

subtypeh

istolog

yinclu

desm

edullarym

ucinou

spapillaryand

tubu

larc

arcino

mas

Medicationusagew

asno

tavailablefor

15patie

nts

ERestr

ogen

receptorP

NIperin

euralinvasion

andPR

progeste

rone

receptor

3Mediandu

ratio

nof

medications

usage(mon

thsrange)m

etform

in505(1ndash

805)insulin

56(10ndash

126)statin

s72(1ndash

168)and

levothyroxine113

(8ndash168)


1932190819491926

164

17671811

1711

14515

15516

16517

17518

18519

19520

MetforminInsulinStatinsLevothyroxine

NoYes

= 0062 = 0108 = 0035= 0682

Figure 1 Oncotype DX recurrence score according to medicationusage

119901 = 0017) Duration of insulin treatment did not influencethese findings Insulin usage was not associated with ODXRS tumor size nodal involvement and other histologicalcharacteristics For ODX RS ER staining and angiolym-phatic invasion rate there were no significant interactionsbetween metformin and insulin in the multivariate analysis(119901 = 0199 119901 = 0413 and 119901 = 0989 resp) Treatment witheither ACEI or ARB was documented in 38 (633) patientstreatedwithmetformin and in 6 (667) patients treatedwithinsulin Neither ARB nor ACEI were associated with ODXRS intensity of ER stain or angiolymphatic invasion rate

43 Impact of Dyslipidemia and Statins on Tumor Characteris-tics Treatment with statins was noted in 208 (317) patientswith median duration usage of 72 (1ndash168) and IQR of 39ndash108months Neither dyslipidemia nor treatment with statins wasassociated with the tumor characteristics examined

44 Impact of Thyroid Dysfunction and Levothyroxineon Tumor Characteristics Hyperthyroidism and hypothy-roidism did not influence all the evaluated characteristicsSixty-two (94) patients were treated with levothyroxineMedian duration usage was 113 (8ndash168) months and IQRwas 85ndash130 months These patients had a trend toward lowerODX RS (mean 171 plusmn 835 and 1926 plusmn 1043 119901 = 0062)(Figure 1) Prolonged levothyroxine treatment was associatedwith lower ODX RS (119901 = 0005) Histological subtypedistribution differed IDC ILC and other histologies werefound in 694 161 and 145 respectively for patientstreated with levothyroxine compared to 821 118 and61 respectively for patients not treated with levothyroxine(119901 = 002) Duration of treatment had no significant impacton this histological difference Levothyroxine treatment wasnot associated with other tumor characteristics

45 Treatment Adjuvant hormonal treatment was pre-scribed to 977 of the patients Most patients (744) didnot receive adjuvant chemotherapy Patients treated withlevothyroxine were less likely to receive adjuvant chemother-apy (64 versus 272 119901 = 00001) For patients treated

0 20 40 60 80 100 12000

02

04

06

08

10

+ CensoredMetformin usage

YesNo

(Months)

OS

= 0833

Figure 2 Overall survival according to metformin usage

with metformin there was a trend toward receiving adjuvantchemotherapy less often (155 versus 262 119901 = 0082)Other evaluated medications have not been associated withchemotherapy and hormonal therapy usage rates

46 Outcome Median follow-up was 618 months (range 17ndash1146) During this period 13 (19) patients died of breastcancer 644 (959) remained alive and 14 (21) died ofother causes The estimated 5-year DFS rates for all patientswere 957with rates of 967 957 and 919 for patientswith low intermediate and high ODX RS respectively Five-year DFS rates were significantly different among patientswith low to high RS (HR = 04 95 CI 016ndash099 119901 =0047) Five-year OS rate for the whole population was 985The Cox proportional hazards model was not applicable forpatients treated with insulin due to the small number ofpatients Five-year DFS BCSS andOS rates were comparablein the studied medications groups and related comorbiditiesFigures 2 and 3 depict OS according to metformin andlevothyroxine usage respectively

5 Discussion

This study evaluated the association between four commonlyused drugs metformin insulin statins and levothyroxineand breast cancer The first three are frequently used inpatients with metabolic syndrome In an attempt to bet-ter understand the direct effects of these medications theinfluence of the related comorbidities was also evaluatedData analysis revealed several significant findings Patientswith diabetes as well as patients treated with metforminhad a significantly lower ODX RS In addition metformintreatment was associated with more intense ER staining Tothe best of our knowledge these findings are novel Theyconcur with previous reports describing improved survivalin breast cancer patients treated with metformin [26 29]Nonetheless as the majority of the patients with DM weretreated with metformin it is not clear whether the difference


Levothyroxine usage(Months)

OS

0 20 40 60 80 100 120

00

02

04

06

08

10

+ CensoredYesNo

= 054

Figure 3 Overall survival according to levothyroxine usage

in ODX RS relates to direct effect of metformin or theexistence of DM

The significantly lower ODX RS in patients treated withmetformin did not translate into improved outcome in ourcohort Given the excellent outcome for all the subgroups alarger population and longer follow-up are probably requiredto identify such differences The tendency of patients treatedwith metformin to receive less adjuvant chemotherapy treat-ment is consistent with lower ODX RS

Patients treated with insulin had significantly higherangiolymphatic invasion in their tumors however the Ki67percentage was significantly lower While evidence of higherangiolymphatic invasion could be consistent with previousreports implying exogenous insulin as a growth factor forcancer cells [34] lower Ki67 is inconsistent with this pos-tulation Of note since only nine patients in the cohortpopulationwere treatedwith insulin the conclusions that canbe drawn from this specific analysis are limited Nonethelessas large scaled epidemiological studies imply existing associ-ation between insulin usage and breast cancer incidence andmortality [52 53] further research to evaluate the potentialeffect of insulin on breast cancer is needed

As opposed to the antidiabetic drugs which might sug-gest some impact on breast cancer biology we did not detectany association between breast cancer and statins treatmentThis is in contrast to the favorable effect described by someinvestigators [35 36 40] The findings are confined to ERpositive HER2 negative early breast cancer patients andit is possible that statins could affect other breast cancersubgroups

The fourth drug investigated was levothyroxine Whileneither diagnosis of hypothyroidism nor diagnosis of hyper-thyroidism had influence on tumor characteristics patientstreated with levothyroxine were less likely to have IDCsubtype and had a tendency to have a lower ODX RS andextended levothyroxine treatment was associated with lowerODX RSThese results suggest that levothyroxine might havea role in breast cancer pathophysiology Despite the asso-ciation with extended levothyroxine usage and lower ODXRS levothyroxine usage was not associated with improved

outcome As all patients had excellent survival longer follow-up is probably needed to evaluated differences in outcome Tothe best of our knowledge all of the aforementioned findingshave not been described previously

Our study has several limitations It was a retrospectivestudy which may cause bias due to unknown or unrecordedconfounders Patientsrsquo comorbidities might cause potentialconfounders and add to the difficulty of results interpretationas was previously described in another study [16] As thiswas a single center study it is more vulnerable to unknownbias Data regarding comorbidities and duration of medi-cations usage were retrieved from the largest health serviceorganization in Israel As not all patients are affiliated tothis organization these data were missing for a quarter ofthe patients Information regarding other antidiabetic drugswhich might have some influence on breast cancer was notcollectedMoreover the efficacy of the evaluatedmedicationsincluding hemoglobin A1C cholesterol level and thyroidfunction were not documented These data might help us tobetter understand the influence of the evaluated medicationson breast cancer The study included all patients with ERpositive HER2 negative early breast cancer from bothgendersMale breast cancermight represent a distinct clinicalentity therefore their inclusionmight have caused additionalbias Nonetheless as only 9 (13) men were included in thestudy cohort they were not likely to significantly influenceour results

Strengths of this study include the large patient cohortFurthermore the chart review included detailed patient datawhich is lacking in registry-based studies The correlationbetween IHC staining results and theODXER PR andHER2(based on RT-PCR) results as well as the correlation betweenODX RS and the DFS of the study population add to thereliability and validity of the findings

6 Conclusion

Patients with ER positive HER2 negative early breast cancerwho were treated with metformin had lower ODX RS andmore intense ER staining which are both associated withfavorable outcome This supports recent research efforts toincorporate metformin in anticancer adjuvant treatmentPatients treated with levothyroxine also had distinct tumorcharacteristics Our findings might suggest that these medi-cations have a role in the pathophysiology and developmentof breast cancer As they are commonly prescribed additionalstudies are required to elucidate the possible associationsbetween these medications and breast cancer and to explorepossible clinical implications

Competing Interests

All authors declare that there is no conflict of interestsregarding the publication of this paper

Authorsrsquo Contributions

Hadar Goldvaser was responsible for substantial contribu-tions to study conception and design acquisition of data


analysis and interpretation of data drafting of the article andrevising it critically for important intellectual content andfinal approval of the version to be published and agrees to beaccountable for all aspects of the work Shulamith Rizel wasresponsible for substantial contributions to study conceptionand design analysis and interpretation of data revising thearticle critically for important intellectual content and finalapproval of the version to be published and agrees to beaccountable for all aspects of the work Daniel Hendler wasresponsible for substantial contributions to acquisition ofdata revising the article critically for important intellectualcontent and final approval of the version to be published andagrees to be accountable for all aspects of the work VictoriaNeiman was responsible for substantial contributions toacquisition of data revising the article critically for impor-tant intellectual content and final approval of the versionto be published Daniel Shepshelovich was responsible forsubstantial contributions to study conception and designanalysis and interpretation of data drafting of the article andrevising it critically for important intellectual content andfinal approval of the version to be published and agrees tobe accountable for all aspects of the work Tzippy Shochatwas responsible for substantial contributions to analysis andinterpretation of data revising the article critically for impor-tant intellectual content and final approval of the versionto be published and agrees to be accountable for all aspectsof the work Aaron Sulkes was responsible for substantialcontributions to analysis and interpretation of data revisingthe article critically for important intellectual content andfinal approval of the version to be published and agrees tobe accountable for all aspects of the work Baruch Brennerwas responsible for substantial contributions to analysisand interpretation of data revising the article critically forimportant intellectual content and final approval of theversion to be published and agrees to be accountable forall aspects of the work Rinat Yerushalmi was responsiblefor substantial contributions to study conception and designanalysis and interpretation of data drafting of the article andrevising it critically for important intellectual content andfinal approval of the version to be published and agrees to beaccountable for all aspects of the work

References

[1] M Cianfrocca and L J Goldstein ldquoPrognostic and predictivefactors in early-stage breast cancerrdquoOncologist vol 9 no 6 pp606ndash616 2004

[2] C W Elston and I O Ellis ldquoPathological prognostic factorsin breast cancer I The value of histological grade in breastcancer experience from a large study with long-term follow-uprdquo Histopathology vol 19 no 5 pp 403ndash410 1991

[3] M Schumacher C Schmoor W Sauerbrei et al ldquoThe prognos-tic effect of histological tumor grade in node-negative breastcancer patientsrdquo Breast Cancer Research and Treatment vol 25no 3 pp 235ndash245 1993

[4] P L Nguyen A G TaghianM S Katz et al ldquoBreast cancer sub-type approximated by estrogen receptor progesterone receptorand HER-2 is associated with local and distant recurrence afterbreast-conserving therapyrdquo Journal of Clinical Oncology vol 26no 14 pp 2373ndash2378 2008

[5] E K A Millar P H Graham S A OrsquoToole et al ldquoPredictionof local recurrence distant metastases and death after breast-conserving therapy in early-stage invasive breast cancer using afive-biomarker panelrdquo Journal of Clinical Oncology vol 27 no28 pp 4701ndash4708 2009

[6] K D Voduc M C U Cheang S Tyldesley K Gelmon T ONielsen and H Kennecke ldquoBreast cancer subtypes and the riskof local and regional relapserdquo Journal of Clinical Oncology vol28 no 10 pp 1684ndash1691 2010

[7] S F Schoppmann G Bayer K Aumayr et al ldquoPrognostic valueof lymphangiogenesis and lymphovascular invasion in invasivebreast cancerrdquo Annals of Surgery vol 240 no 2 pp 306ndash3122004

[8] S E Pinder I O EllisM Galea S OrsquoRourke RW Blamey andC W Elston ldquoPathological prognostic factors in breast cancerIII Vascular invasion relationship with recurrence and survivalin a large study with long-term follow-uprdquo Histopathology vol24 no 1 pp 41ndash47 1994

[9] S Paik S Shak G Tang et al ldquoA multigene assay to predictrecurrence of tamoxifen-treated node-negative breast cancerrdquoThe New England Journal of Medicine vol 351 no 27 pp 2817ndash2826 2004

[10] S Paik G Tang S Shak et al ldquoGene expression and bene-fit of chemotherapy in women with node-negative estrogenreceptor-positive breast cancerrdquo Journal of Clinical Oncologyvol 24 no 23 pp 3726ndash3734 2006

[11] S K Garg H Maurer K Reed and R Selagamsetty ldquoDiabetesand cancer two diseases with obesity as a common risk factorrdquoDiabetes Obesity andMetabolism vol 16 no 2 pp 97ndash110 2014

[12] C M Kitahara A Berrington de Gonzalez N D Freedman etal ldquoTotal cholesterol and cancer risk in a large prospective studyin Koreardquo Journal of Clinical Oncology vol 29 no 12 pp 1592ndash1598 2011

[13] E R Nelson S E Wardell J S Jasper et al ldquo27-Hydroxycholesterol links hypercholesterolemia and breastcancer pathophysiologyrdquo Science vol 342 no 6162 pp1094ndash1098 2013

[14] S S Coughlin E E Calle L R Teras J Petrelli andM JThunldquoDiabetes mellitus as a predictor of cancer mortality in a largecohort of US adultsrdquoAmerican Journal of Epidemiology vol 159no 12 pp 1160ndash1167 2004

[15] C-H Tseng C-K Chong and T-Y Tai ldquoSecular trend formortality from breast cancer and the association betweendiabetes and breast cancer in Taiwan between 1995 and 2006rdquoDiabetologia vol 52 no 2 pp 240ndash246 2009

[16] C-H Tseng ldquoDiabetes and breast cancer in Taiwanese womena detection biasrdquo European Journal of Clinical Investigation vol44 no 10 pp 910ndash917 2014

[17] N Vrachnis C Iavazzo Z Iliodromiti et al ldquoDiabetes mellitusand gynecologic cancer molecular mechanisms epidemiologi-cal clinical and prognostic perspectivesrdquoArchives of Gynecologyand Obstetrics vol 293 no 2 pp 239ndash246 2016

[18] S S Coughlin and S A Smith ldquoThe Insulin-like growth factoraxis adipokines physical activity and obesity in relation tobreast cancer incidence and recurrencerdquo Cancer and ClinicalOncology vol 4 no 2 pp 24ndash31 2015

[19] I Ben Sahra K Laurent A Loubat et al ldquoThe antidiabeticdrugmetformin exerts an antitumoral effect in vitro and in vivothrough a decrease of cyclin D1 levelrdquo Oncogene vol 27 no 25pp 3576ndash3586 2008


[20] I N Alimova B Liu Z Fan et al ldquoMetformin inhibits breastcancer cell growth colony formation and induces cell cyclearrest in vitrordquo Cell Cycle vol 8 no 6 pp 909ndash915 2009

[21] EGarcıa-Esquinas EGuino G Castano-Vinyals et al ldquoAssoci-ation of diabetes and diabetes treatmentwith incidence of breastcancerrdquo Acta Diabetologica vol 53 no 1 pp 99ndash107 2016

[22] D Margel D R Urbach L L Lipscombe et al ldquoMetformin useand all-cause and prostate cancer-specificmortality amongmenwith diabetesrdquo Journal of Clinical Oncology vol 31 no 25 pp3069ndash3075 2013

[23] M Yin J Zhou E J Gorak and F Quddus ldquoMetforminis associated with survival benefit in cancer patients withconcurrent type 2 diabetes a systematic review and meta-analysisrdquoThe Oncologist vol 18 no 12 pp 1248ndash1255 2013

[24] C-H Tseng ldquoMetformin may reduce breast cancer risk inTaiwanese women with type 2 diabetesrdquo Breast Cancer Researchand Treatment vol 145 no 3 pp 785ndash790 2014

[25] S Jiralerspong S L Palla S H Giordano et al ldquoMetformin andpathologic responses to neoadjuvant chemotherapy in diabeticpatients with breast cancerrdquo Journal of Clinical Oncology vol 27pp 3297ndash3302 2009

[26] H Xu K Chen X Jia et al ldquoMetformin use is associated withbetter survival of breast cancer patients with diabetes a meta-analysisrdquo Oncologist vol 20 no 11 pp 1236ndash1244 2015

[27] J Kim W Lim E-K Kim et al ldquoPhase II randomized trialof neoadjuvant metformin plus letrozole versus placebo plusletrozole for estrogen receptor positive postmenopausal breastcancer (METEOR)rdquo BMCCancer vol 14 no 1 article 170 2014

[28] B A Oppong L A Pharmer S Oskar et al ldquoThe effect ofmetformin on breast cancer outcomes in patients with type 2diabetesrdquo Cancer Medicine vol 3 no 4 pp 1025ndash1034 2014

[29] T Yang Y Yang and S Liu ldquoAssociation between metformintherapy and breast cancer incidence and mortality evidencefrom a meta-analysisrdquo Journal of Breast Cancer vol 18 no 3pp 264ndash270 2015

[30] B Kowall A Stang W Rathmann and K Kostev ldquoNo reducedrisk of overall colorectal lung breast and prostate cancer withmetformin therapy in diabetic patients database analyses fromGermany and the UKrdquo Pharmacoepidemiology and Drug Safetyvol 24 no 8 pp 865ndash874 2015

[31] M Janghorbani M Dehghani and M Salehi-MarzijaranildquoSystematic review and meta-analysis of insulin therapy andrisk of cancerrdquo Hormones and Cancer vol 3 no 4 pp 137ndash1462012

[32] Oslash Karlstad J Starup-Linde P Vestergaard et al ldquoUse of insulinand insulin analogs and risk of cancermdashsystematic review andmeta-analysis of observational studiesrdquo Current Drug Safetyvol 8 no 5 pp 333ndash348 2013

[33] I N Colmers S L Bowker L A Tjosvold and J A JohnsonldquoInsulin use and cancer risk in patients with type 2 diabetes asystematic review and meta-analysis of observational studiesrdquoDiabetes amp Metabolism vol 38 no 6 pp 485ndash506 2012

[34] H K Bronsveld B ter Braak Oslash Karlstad et al ldquoTreatmentwith insulin (analogues) and breast cancer risk in diabeticsa systematic review and meta-analysis of in vitro animal andhuman evidencerdquo Breast Cancer Research vol 17 article 1002015

[35] N Ghosh-Choudhury C C Mandal N Ghosh-ChoudhuryandGGhoshChoudhury ldquoSimvastatin induces derepression ofPTEN expression viaNF120581B to inhibit breast cancer cell growthrdquoCellular Signalling vol 22 no 5 pp 749ndash758 2010

[36] K K W Chan A M Oza and L L Siu ldquoThe statins asanticancer agentsrdquo Clinical Cancer Research vol 9 no 1 I pp10ndash19 2003

[37] M J Campbell L J Esserman Y Zhou et al ldquoBreast cancergrowth prevention by statinsrdquo Cancer Research vol 66 no 17pp 8707ndash8714 2006

[38] Y H Park H H Jung J S Ahn and Y-H Im ldquoStatininduces inhibition of triple negative breast cancer (TNBC)cells via PI3K pathwayrdquo Biochemical and Biophysical ResearchCommunications vol 439 no 2 pp 275ndash279 2013

[39] A Gopalan W Yu B G Sanders and K Kline ldquoSimvastatininhibition of mevalonate pathway induces apoptosis in humanbreast cancer cells via activation of JNKCHOPDR5 signalingpathwayrdquo Cancer Letters vol 329 no 1 pp 9ndash16 2013

[40] S Zhong X Zhang L Chen T Ma J Tang and J ZhaoldquoStatin use and mortality in cancer patients systematic reviewand meta-analysis of observational studiesrdquo Cancer TreatmentReviews vol 41 no 6 pp 554ndash567 2015

[41] S Bonovas K Filioussi N Tsavaris and N M Sitaras ldquoUse ofstatins and breast cancer a meta-analysis of seven randomizedclinical trials and nine observational studiesrdquo Journal of ClinicalOncology vol 23 no 34 pp 8606ndash8612 2005

[42] K Undela V Srikanth and D Bansal ldquoStatin use and risk ofbreast cancer a meta-analysis of observational studiesrdquo BreastCancer Research and Treatment vol 135 no 1 pp 261ndash269 2012

[43] J A McDougall K E Malone J R Daling K L Cushing-Haugen P L Porter and C I Li ldquoLong-term statin use and riskof ductal and lobular breast cancer among women 55 to 74 yearsof agerdquoCancer Epidemiology Biomarkers and Prevention vol 22no 9 pp 1529ndash1537 2013

[44] P Desai A Lehman R T Chlebowski et al ldquoStatins and breastcancer stage and mortality in the Womenrsquos Health InitiativerdquoCancer Causes and Control vol 26 no 4 pp 529ndash539 2015

[45] C R Cardwell BM Hicks C Hughes and L J Murray ldquoStatinuse after diagnosis of breast cancer and survival a population-based cohort studyrdquo Epidemiology vol 26 no 1 pp 68ndash78 2015

[46] A Tosovic C Becker A-G Bondeson et al ldquoProspectivelymeasured thyroid hormones and thyroid peroxidase antibodiesin relation to breast cancer riskrdquo International Journal of Cancervol 131 no 9 pp 2126ndash2133 2012

[47] A Tosovic A-G Bondeson L Bondeson U-B Ericsson JMalm and JManjer ldquoProspectivelymeasured triiodothyroninelevels are positively associated with breast cancer risk inpostmenopausal womenrdquo Breast Cancer Research vol 12 no 3article R33 2010

[48] J Brandt S Borgquist and J Manjer ldquoProspectively measuredthyroid hormones and thyroid peroxidase antibodies in relationto risk of different breast cancer subgroups a Malmo Diet andCancer Studyrdquo Cancer Causes and Control vol 26 no 8 pp1093ndash1104 2015

[49] M Soslashgaard D K Farkas V Ehrenstein J O JoslashrgensenO M Dekkers and H T Soslashrensen ldquoHypothyroidism andhyperthyroidism and breast cancer risk a nationwide cohortstudyrdquo European Journal of Endocrinology vol 174 no 4 pp409ndash414 2016

[50] Y K Chae M E Valsecchi J Kim et al ldquoReduced risk of breastcancer recurrence in patients using ACE inhibitors ARBsandor statinsrdquo Cancer Investigation vol 29 no 9 pp 585ndash5932011

[51] S Bangalore S Kumar S E Kjeldsen et al ldquoAntihypertensivedrugs and risk of cancer network meta-analyses and trial


sequential analyses of 324 168 participants from randomisedtrialsrdquoThe Lancet Oncology vol 12 no 1 pp 65ndash82 2011

[52] C-H Tseng ldquoProlonged use of human insulin increases breastcancer risk in Taiwanese women with type 2 diabetesrdquo BMCCancer vol 15 article 846 2015

[53] C-H Tseng ldquoUse of insulin and mortality from breast canceramong Taiwanese women with diabetesrdquo Journal of DiabetesResearch vol 2015 Article ID 678756 8 pages 2015

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom


studies found that metformin inhibits cancer cell growthreduces the risk of developing solid tumors and improvescancer outcome including breast cancer [11 19ndash23] Thereduced risk of breast cancer associated with metformin usewas clearly demonstrated in a recent large epidemiologicalstudy [24] Furthermore metformin usage was associatedwith a threefold greater pathologic complete response indiabetic patients with breast cancer receiving neoadjuvantchemotherapy compared to patients with DM not treatedwith metformin [25] In addition a meta-analysis by Xu etal found that breast cancer patients with DM who weretreated with metformin had significantly decreased risk ofall-cause and cancer specific mortality compared with theircounterparts who did not receive metformin [26] In light ofthese data the drug is currently being investigated for adju-vant breast cancer treatment of nondiabetic patients (httpsclinicaltrialsgovct2showNCT01101438 NCIC CTGMA32study Goodwin PJ et al) On the other hand several reportsdid not support the positive impact of metformin on breastcancer [21 27ndash30]

The activity of insulin as a growth factor has raisedconcern regarding the stimulation of neoplastic growth byinsulin analogues Data regarding the influence of insulinanalogues on cancer incidence are inconsistent One meta-analysis found that diabetic patients treated with insulin hadan increased cancer risk compared to non-insulin-treateddiabetics [31] while another study reported an inconsistentimpact of insulin on cancer incidence [32] More specificallya meta-analysis by Colmers et al found an increased breastcancer risk in glargine users compared to those who did notuse glargine [33]Thismay be attributed to themitogenic andproliferative activity of glargine due to its increased bindingaffinity to insulin-like growth factor-1 compared to humaninsulin [34] However a pooled analysis of 13 epidemiologicalstudies did not support the association between treatmentwith glargine and increased incidence of breast cancer [34]

There are preclinical data supporting the antineoplas-tic effect of statins Possible mechanisms include reducedexpression of the antiapoptotic protein bcl-xL and increasedtranscription of phosphatase and tensin homolog whichinhibit the oncogenic phosphatidylinositol-3-kinase pathway[35 36] Several in vitro studies demonstrated that statinsdecrease breast cancer cell proliferation [35 37ndash39] A recentmeta-analysis by Zhong et al found that statins usage wasassociated with improved cancer survival [40] Despite thesereports current data do not support lower incidence of breastcancer in patients treated with statins [41ndash43] and it is notclear if they have any effect on outcome [44 45]

Thyroid function is an additional metabolic factor thatmay influence breast cancer Data regarding a correlationbetween higher levels of free T4 (FT4) and triiodothyronineas well as lower levels of thyroid peroxidase autoantibodies(TPO-Ab) and higher breast cancer incidence are accumu-lating [46ndash49] Some data suggest that FT4 and TPO-Ablevels are associated with breast cancer characteristics [48]however the impact of levothyroxine treatment is unclear

The aim of this study was to evaluate the associationof metformin insulin statins and levothyroxine and breastcancer characteristics and outcome With the advent of

ODX as a prognostic and predictive tool for daily practiceimportant molecular information on tumor in addition totraditional pathology report is obtained [9 10] Thereforewe chose to focus on patients with early breast ER positiveHER2 negative disease for whom their physician added theODX test prior to making a decision on systemic treatment

2 Methods

This was a retrospective single center cohort study Patientswith early ER positive HER2 negative breast cancer whowere diagnosed between 42005 and 32012 were includedThe cohort included all patients treated in our institute duringthis period whose tumors were sent for ODX analysis

The medical records of all patients were retrospectivelyreviewed up to 82015 Data on patient demographicsclinical-pathological parameters treatment and outcomewere retrieved Data regarding usage of metformin insulinstatins and levothyroxine at the time of breast cancer diag-nosis were also collected Duration of treatment until breastcancer diagnosis and related comorbidities including DMdyslipidemia and hypothyroidism were retrieved from theClalit Health Service database We evaluated the influence ofthe studied medications on tumor size nodal involvementstage ODX recurrence score (RS) and histological charac-teristics Histological characteristics evaluated were ER PRHER2 grade angiolymphatic and perineural invasion Ki67P53 and histological subtype As angiotensin convertingenzyme inhibitors (ACEI) and angiotensin receptor blockers(ARB) are commonly prescribed in patients with DM andseveral studies suggested a possible relationship betweenthese medications and breast cancer [50 51] we evaluatedtheir usage in this cohort The study protocol was approvedby the institutional ethics committee

Staining for ER PR p53 Ki-67 andHER2was performedusing the Ventana Benchmark XT automated immunostainer(Ventana Tucson AZ USA) with the standard cell condi-tioner (CC1) protocol for 30min Following deparaffinizationand the CC1 protocol ready-to-use ER rabbit monoclonalantibody [anti-ER (6F11) primary antibody Ventana] wasapplied for 40min incubation at 37∘C PR rabbit mono-clonal antibody (clone 16 Novocastra Newcastle UK) wasemployed at a 1 100 dilution with 40min incubation at37∘C Ki 67 rabbit monoclonal antibody (clone SP6 ThermoFisher Scientific) was used at a 1 100 dilution for 40minat 37∘C and ready-to-use PATHWAY HER2 anti-HER2neurabbit monoclonal antibody (4B5) (Ventana) was utilizedwith 32min incubation at 37∘C Slides were counterstainedwith 410158406-diamidino-2-phenylindole (Sigma-Aldrich) andthe stainedmaterial was visualized under a BX51 fluorescencemicroscope (Olympus) The signals were analyzed manually

ER and PR immunohistochemical (IHC) staining usedthemodified version of the119867-scoremethod 1timespercentage ofweakly staining nuclei + 2 times percentage of moderately stain-ing nuclei + 3 times percentage of intensely staining nuclei100yielding a range of 0 to 3 Ki67 staining was reportedby percentage of positively stained nuclei (0 to 100)HER2 negativity was defined as an immunohistochemistrytest score of 0 or 1 If the IHC score equaled 2 HER2











4 Results








tics

Popu

latio

n(119873)

Histolog

yGrade

ERPR

Ki67

()

P53(

)Ang

iolymph

atic

invasio

nPN

I

IDC(

)ILC(

)Other2(

)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Yes()1199011

Yes()1199011

All(671)

809

122

69

mdash202

(059)

mdash247

(057)

mdash14

7(119

)mdash

1591(1358)

mdash645

(1696)

mdash61

mdash45

mdashMetform

in3(60)

767

1583

0683

2(063)

0846

259

(04)

0032

129(105)

0217

137(111)

0204

696

(147)

0828

34

0368

51

0742

Insulin3(9)

778

111

111

0878

186(038)

0474

249

(067)

0948

112(094)

0371

95(464)

0017

85(203)

0765

222

0041

00512

Statins3(208)

798

111

910266

207

(05)

0193

248

(055)

0916

141(10

2)03461678(14

53)0318

645

(1561)0999

780212

49

0768

Levothyroxine3

(62)

694

161

145

002

202

(055)

094

247

(052

)0999

137(127)

04881528(1264

)0727592

(1656)

084

33

0339

160257

1119901valuer

eferstothec

omparis

onof

each

varia

bleb

etweenpatie

ntsw

housed

thes


ther

esto

fthe

coho

rt

2Ano

ther

subtypeh

istolog

yinclu

desm

edullarym

ucinou

spapillaryand

tubu

larc

arcino

mas

Medicationusagew

asno

tavailablefor

15patie

nts

ERestr

ogen

receptorP

NIperin

euralinvasion

andPR

progeste

rone

receptor

3Mediandu

ratio

nof

medications

usage(mon

thsrange)m

etform

in505(1ndash

805)insulin

56(10ndash

126)statin

s72(1ndash

168)and

levothyroxine113

(8ndash168)


1932190819491926

164

17671811

1711

14515

15516

16517

17518

18519

19520


NoYes

= 0062 = 0108 = 0035= 0682






0 20 40 60 80 100 12000

02

04

06

08

10


YesNo

(Months)

OS

= 0833




5 Discussion




OS

0 20 40 60 80 100 120

00

02

04

06

08

10

+ CensoredYesNo

= 054










6 Conclusion


Competing Interests






References






























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


























4 Results








tics

Popu

latio

n(119873)

Histolog

yGrade

ERPR

Ki67

()

P53(

)Ang

iolymph

atic

invasio

nPN

I

IDC(

)ILC(

)Other2(

)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Yes()1199011

Yes()1199011

All(671)

809

122

69

mdash202

(059)

mdash247

(057)

mdash14

7(119

)mdash

1591(1358)

mdash645

(1696)

mdash61

mdash45

mdashMetform

in3(60)

767

1583

0683

2(063)

0846

259

(04)

0032

129(105)

0217

137(111)

0204

696

(147)

0828

34

0368

51

0742

Insulin3(9)

778

111

111

0878

186(038)

0474

249

(067)

0948

112(094)

0371

95(464)

0017

85(203)

0765

222

0041

00512

Statins3(208)

798

111

910266

207

(05)

0193

248

(055)

0916

141(10

2)03461678(14

53)0318

645

(1561)0999

780212

49

0768

Levothyroxine3

(62)

694

161

145

002

202

(055)

094

247

(052

)0999

137(127)

04881528(1264

)0727592

(1656)

084

33

0339

160257

1119901valuer

eferstothec

omparis

onof

each

varia

bleb

etweenpatie

ntsw

housed

thes


ther

esto

fthe

coho

rt

2Ano

ther

subtypeh

istolog

yinclu

desm

edullarym

ucinou

spapillaryand

tubu

larc

arcino

mas

Medicationusagew

asno

tavailablefor

15patie

nts

ERestr

ogen

receptorP

NIperin

euralinvasion

andPR

progeste

rone

receptor

3Mediandu

ratio

nof

medications

usage(mon

thsrange)m

etform

in505(1ndash

805)insulin

56(10ndash

126)statin

s72(1ndash

168)and

levothyroxine113

(8ndash168)


1932190819491926

164

17671811

1711

14515

15516

16517

17518

18519

19520


NoYes

= 0062 = 0108 = 0035= 0682






0 20 40 60 80 100 12000

02

04

06

08

10


YesNo

(Months)

OS

= 0833




5 Discussion




OS

0 20 40 60 80 100 120

00

02

04

06

08

10

+ CensoredYesNo

= 054










6 Conclusion


Competing Interests






References






























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


















tics

Popu

latio

n(119873)

Histolog

yGrade

ERPR

Ki67

()

P53(

)Ang

iolymph

atic

invasio

nPN

I

IDC(

)ILC(

)Other2(

)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Mean(SD)1199011

Yes()1199011

Yes()1199011

All(671)

809

122

69

mdash202

(059)

mdash247

(057)

mdash14

7(119

)mdash

1591(1358)

mdash645

(1696)

mdash61

mdash45

mdashMetform

in3(60)

767

1583

0683

2(063)

0846

259

(04)

0032

129(105)

0217

137(111)

0204

696

(147)

0828

34

0368

51

0742

Insulin3(9)

778

111

111

0878

186(038)

0474

249

(067)

0948

112(094)

0371

95(464)

0017

85(203)

0765

222

0041

00512

Statins3(208)

798

111

910266

207

(05)

0193

248

(055)

0916

141(10

2)03461678(14

53)0318

645

(1561)0999

780212

49

0768

Levothyroxine3

(62)

694

161

145

002

202

(055)

094

247

(052

)0999

137(127)

04881528(1264

)0727592

(1656)

084

33

0339

160257

1119901valuer

eferstothec

omparis

onof

each

varia

bleb

etweenpatie

ntsw

housed

thes


ther

esto

fthe

coho

rt

2Ano

ther

subtypeh

istolog

yinclu

desm

edullarym

ucinou

spapillaryand

tubu

larc

arcino

mas

Medicationusagew

asno

tavailablefor

15patie

nts

ERestr

ogen

receptorP

NIperin

euralinvasion

andPR

progeste

rone

receptor

3Mediandu

ratio

nof

medications

usage(mon

thsrange)m

etform

in505(1ndash

805)insulin

56(10ndash

126)statin

s72(1ndash

168)and

levothyroxine113

(8ndash168)


1932190819491926

164

17671811

1711

14515

15516

16517

17518

18519

19520


NoYes

= 0062 = 0108 = 0035= 0682






0 20 40 60 80 100 12000

02

04

06

08

10


YesNo

(Months)

OS

= 0833




5 Discussion




OS

0 20 40 60 80 100 120

00

02

04

06

08

10

+ CensoredYesNo

= 054










6 Conclusion


Competing Interests






References






























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















1932190819491926

164

17671811

1711

14515

15516

16517

17518

18519

19520


NoYes

= 0062 = 0108 = 0035= 0682






0 20 40 60 80 100 12000

02

04

06

08

10


YesNo

(Months)

OS

= 0833




5 Discussion




OS

0 20 40 60 80 100 120

00

02

04

06

08

10

+ CensoredYesNo

= 054










6 Conclusion


Competing Interests






References






























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


















OS

0 20 40 60 80 100 120

00

02

04

06

08

10

+ CensoredYesNo

= 054










6 Conclusion


Competing Interests






References






























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


















References






























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


























































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
















research article the association between treatment for

Documents