http://ijs.sagepub.com/International Journal of Surgical Pathology

http://ijs.sagepub.com/content/early/2008/05/15/1066896908315796The online version of this article can be found at:

 DOI: 10.1177/1066896908315796

published online 15 May 2008INT J SURG PATHOLSulkowski

Anna Malgorzata Chabowska, Mariola Sulkowska, Adrian Chabowski, Andrzej Wincewicz, Mariusz Koda and StanislawErythropoietin and Erythropoietin Receptor in Colorectal Cancer

  

- Jun 22, 2008version of this article was published on more recent A

Published by:

http://www.sagepublications.com

can be found at:International Journal of Surgical PathologyAdditional services and information for    

  http://ijs.sagepub.com/cgi/alertsEmail Alerts:

 

http://ijs.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- May 15, 2008 OnlineFirst Version of Record>>  

- Jun 22, 2008Version of Record

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

1

Erythropoietin and ErythropoietinReceptor in Colorectal Cancer

Anna M. Chabowska, MD, Mariola Sulkowska, MD, PhD,Adrian Chabowski, MD, PhD, Andrzej Wincewicz, MD,Mariusz Koda, MD, and Stanislaw Sulkowski, MD, PhD

Specifically, high expression of EpoR has been foundon the surface of the endothelial cells, neurons, liver,uterus, and on various solid tumors.5,6 This suggestsadditional considerable function for this hormone inseveral physiological and pathological processesunrelated to erythropoiesis. In particular, Epo seemsto play a crucial role in the process of neoangiogen-esis, promotion of survival of hypoxic cancer cells,and tumor growth.7 Erythropoietin was also shown tobe a factor of unfavorable prognosis in endometrialcancer.8 Erythropoietin is induced by hypoxia-inducible factor 1, which also upregulates transcrip-tion of other angiogenic molecules, such as vascularendothelial growth factor.9 Results of experimentson cultured human breast and cervical cancer celllines have also implicated Epo or EpoR signaling asa new factor influencing tumor progression. In thesecells, high messenger RNA levels of Epo and EpoRwere shown to be positively correlated with highrates of proliferation. Exposure of hybrid EA.hy926endothelial cell lines (the cells that are formed byfusion of human umbilical vein endothelial cells[HUVECs] with A549 lung carcinoma cells) to

Erythropoietin (Epo) functions via its trans-membrane erythropoietin receptor (EpoR) onhematopoietic cells and is considered as one

of the most potent stimulators of erythropoiesis.1,2

In erythroid cells, activation of Epo/EpoR signalingresults in a number of cellular events, such as stim-ulation of proliferation and differentiation of ery-throblasts and inhibition of apoptosis.3 The ability ofEpo to stimulate the erythropoiesis has led to itsclinical use in patients with chronic diseases associ-ated with anemia. Severe anemia often accompaniescolorectal tumors, and recombinant human Epo(rHuEpo) might be used for the treatment of thistype of cancer.4 However, both Epo and EpoR haverecently been found to be expressed on several normalnonhematopoietic cells and in neoplastic tissues.

Erythropoietin via erythropoietin receptor effectivelyprevents anemia, giving reasons for a clinical use oferythropoietin in patients with colorectal cancers.However, erythropoietin seems to promote survival ofthe neoplastic cells in hypoxic environment. The aimof this study was to evaluate immunohistochemicallythe expression of erythropoietin and erythropoietinreceptor in 136 primary colorectal cancers with a cor-relation to different anatomo-clinical features. Erythro-poietin correlated with erythropoietin receptor incolorectal cancers (r = 0.547, P < .00001). Erythropoi-etin and erythropoietin receptor expressions werestatistically higher in adenocarcinomas versus mucinous

carcinomas (P = .05 and P = .03, respectively) and inmoderately (G2) versus poorly differentiated (G3)tumors (P = .001 and P = .02, respectively). This invivo study is the first study that provides evidences forthe presence of erythropoietin and erythropoietinreceptor in human colorectal cancer. The expressionsof these proteins strictly depended on grading becausethe better histological differentiation probably comesfrom trophic influence of erythropoietin and erythro-poietin receptor.

Keywords: Epo; EPoR; tumor differentiation; colorec-tal cancer

From the Departments of Clinical and General Pathomorphology(AMC, MS, AW, MK, SS), and Department of Physiology (AC),Medical University of Bialystok, Bialystok, Poland.

Address correspondence to: Stanislaw Sulkowski, MD, PhD,Department of General Pathomorphology, Collegium Pathologi-cum, Medical University of Bialystok, 13 Waszyngtona St, Bialystok15-269, Poland; e-mail: sulek@zeus.amb.edu.pl.

International Journal ofSurgical Pathology

Volume XX Number XMonth XXXX xx-xx

© 2008 Sage Publications10.1177/1066896908315796

http://ijsp.sagepub.comhosted at

http://online.sagepub.com

INT J SURG PATHOL OnlineFirst, published on May 15, 2008 as doi:10.1177/1066896908315796

Copyright 2008 by SAGE Publications. at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

rHuEpo stimulated both DNA synthesis and prolifer-ation of the cells.10 However, disruption of Epo orEpoR signaling was associated with antiproliferativeeffects in malignancies of female reproductive tractand xenografts of human cancer lines in mice.11,12

Recombinant Epo did not affect tumor growth orangiogenesis in several cancer models, such as R3230rat mammary adenocarcinomas, CT-26 mouse coloncarcinomas, HCT-116 human colon carcinomas, andFaDu human head and neck tumors.13 Future studiesare warranted to fully evaluate the role of Epo orEpoR complex in carcinogenesis because recent stud-ies have reported the presence of Epo and EpoR indifferent types of human cancers.6,14

Given the essential functions of Epo or EpoR sig-naling present in cancer cells and possible wide-spread use of Epo in the treatment of anemia inpatients with colorectal cancer it is of particularinterest to examine the expressions of Epo and EpoRin these tumors. In the present study, we examinedthe expressions of Epo and EpoR in primary tumorsof colorectal cancer by immunohistochemistry. Wealso investigated the relationship between the expres-sion of Epo and EpoR in colorectal cancers and cho-sen anatomo-clinical properties of tumor texture.

Materials and Methods

Tissue Samples

The tissue samples were from 136 human colorectaltumors obtained from patients surgically treated forprimary colorectal cancer. No individuals underwenteither radiotherapy or chemotherapy before tumorresection. No individual was also given rHuEpobefore sampling of the cancer tissues. This studywas guided in concordance with the 1995Declaration of Helsinki and its latest revision in2000 and approved by the local ethical committee atthe Medical University of Bialystok.

Each neoplastic tumor was cut parallel to thelongest axis to obtain at least 1 complete cross sec-tion of 2- to 3-mm thick. It was then divided intosmall blocks of 1 to 1.5 cm in diameter. In every case,we obtained 4 to 8 segments comprising the tumorand the macroscopically unchanged colon mucosathat was taken from proximal end of the colon, atleast 10 cm from the primary tumor. On the basisof scant number of depth of tumor intramuralgrowth cancers (pT) 1, we prompted pT1 and pT2

neoplasms into 1 group (pT1 + pT2). Cancers clas-sified as pT3 and pT4 formed another group (pT3 +pT4). In this case, 12 tumors were classified as pT1or pT2 (pT1 + pT2) and 124 tumors as pT3 or pT4(pT3 + pT4). According to the guidelines of WorldHealth Organization (WHO), primary tumors of 93patients were classified as moderately differentiatedwith histological-grade G2 cancers and tumors ofother 43 patients as poorly differentiated cancers(G3). The histological type of 114 examined caseswas diagnosed to be adenocarcinoma, and other 22tumors were found to be mucinous carcinomas. Inall, 64 tumors were located in rectum and 72 tumorsin colon. Lymph nodes metastases were found in61 samples.

Immunohistochemical Tissue Staining

The samples were fixed in 10% buffered formalde-hyde for up to 24 hours and then embedded inparaffin at 56°C. Segments, 5-μm thick, were cutfrom the paraffin blocks (microtom Microm H340)and stained with hematoxylin and eosin. This stain-ing was routinely used for histological evaluation ofAmerican Joint Committee on Cancer/InternationalUnion Against Cancer Tumor Node Metastasis(AJCC/UICC TNM) stage, histological tumortype, and grade of histological differentiation (G).Immunohistochemical assays were performed onformalin-fixed, paraffin-embedded sections. Slideswere incubated with antibodies against Epo (rabbitpolyclonal, H-162 [1:200 dilution]; Santa CruzBiotechnologies, Santa Cruz, California) andEpoR (rabbit polyclonal, C-20 [1:400 dilution]).Immunohistochemical stainings for Epo and EpoRwere interpreted semiquantitatively by assessing theintensity and the extent of staining. The percentageof weakly and strongly staining cells was determinedas follows: 0, if the tumor contained less than 10%immunoreactive cells; 1, if immunoreactive cellsranged from 10% to 50% of cancer cells; and 2, ifmore than 50% of malignant cells of each tumorwere immunoreactive. The immunohistochemicalstaining was assessed by 2 pathologists in 10 differ-ent fields of each tumor at magnification of ×20using light microscope, and the mean number oftumor-positive cells was counted. Negative controlswere prepared with omission of the primary antibod-ies. Epo- and EpoR-stained specimens of breast can-cer served as positive controls.15

2 International Journal of Surgical Pathology / Vol. XX, No. X, Month XXXX

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

Statistical Analysis

The correlations between the Epo/EpoR expressionand the anatomo-pathologic features of colorectalcancer were analyzed by a Chi-square test orSpearman rank correlation test, where appropriate,with a level of significance at P < .05.

Results

Erythropoietin and EpoR expressions were predom-inantly discrete in normal colorectal mucosa that wasremote from the tumor. Both Epo and EpoR wereclearly expressed in adjacent mucosa in close relationto tumor tissue selectively in Epo-/EpoR-positivecolorectal cancers. There was a weak positive stain-ing for Epo in mucosa nearby colorectal cancer foci.Mucin of goblet cells pushed protoplasm in basalregion of cell, and the staining also is accumulatedin this area of cytoplasm (Figures 1A-1C and 1G-1I).Similar phenomenon was observed in our previousstudy with respect to immunohistochemical expres-sion pattern in breast cancers.16 No staining of thecells was observed in any of the tumor sections eitherafter omitting the first antibody or after incubationwith appropriate secondary antibody.

Expression of Epo in Colorectal Cancers

Positive Epo staining was detected in 94 (69.2%)tumors. No immunohistochemical staining for Epoin tumor tissue was observed in 42 (30.8%) sam-ples. At the cellular level, intracytoplasmic labelingpresented as confluent granules in colorectal can-cer. It was quite evenly distributed in cytoplasm ofcancer cells except the perinuclear region, wherethe staining was apparently reinforced (Figures 1Dand 1E), the immunohistochemical patterns of Epostaining comprised infranuclear and supranuclearstaining.

We noted a positive correlation between Epoexpression and the grade of differentiation of thetumor cells (P = .001). Specifically, higher Epo expres-sion was observed more frequently in moderately dif-ferentiated colorectal cancers (G2) than in poorlydifferentiated tumors (G3). We also noted a correla-tion between the expression of Epo and the type ofthe colorectal cancer. We found mucinous adeno-carcinomas expressing Epo at a lower rate than ade-nocarcinomas (P = .05). No other correlations were

found between Epo expression in the tumor cells andother examined anatomo-clinical features (Table 1).

Expression of EpoR in ColorectalCancers

We observed strong cytoplasmic or membrane EpoRimmunostaining in 61 (44.8%) and negative stain-ing in 75 (55.2%) samples. The staining pattern ofcytoplasm and cell membranes stained for EpoR incolorectal cancer was coarse granular. Beads ofEpoR-positive granules were arranged in close rowsalong the cell membranes (Figures 1J and 1K). Inaddition, in benign adjacent tissue, strong EpoRexpression was found only in vascular endothelialcells and in macrophages. We noted that strongexpression of EpoR was more frequently seen in theseverely hypoxic margin of the tumors (necroticareas; data not shown). Similar to the Epo, EpoRstaining was positively correlated with the type ofthe colorectal cancer (P = .03) with lower expressionin mucinous adenocarcinomas. Erythropoietin recep-tor was also more frequently expressed in moderatelydifferentiated tumors (G2 cancers) than in poorlydifferentiated tumors (G3 cancers; P = .02). Noother associations were found between EpoR stainingscore and primary tumor site, pT-status, pN-status,age, or sex (Table 1).

Correlation of Epo and EpoR Expressionin Colorectal Cancers

We also observed a strong positive correlation betweenimmunohistochemical expression of Epo and EpoR(r = 0.547; P = .00001) in colorectal cancer cells(Table 2). It was striking that none of the EpoR-negativetumors had any high expression of Epo, and only fewEpoR-negative tumors presented low immunoreactivityto Epo. Most tumors, which were highly positive forEpo, also expressed EpoR abundantly.

Discussion

This study was undertaken to characterize theexpression of Epo and EpoR in human primary col-orectal cancers. To the best of our knowledge, this isthe first study to report the expressions of both pro-teins Epo and EpoR in colorectal cancer cells. Thepresence of Epo and EpoR in gastric carcinoma cells

Erythropoietin and Its Receptor in Colorectal Cancer / Chabowska et al 3

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

has been shown previously.17 Acs et al8 showedincreased expression of Epo or EpoR in cancer thanin benign adjacent tissue,8 and our present results

are in concordance with these studies showingprevalence of Epo/EpoR expression in tumor cellscompared with normal colon mucosa.

4 International Journal of Surgical Pathology / Vol. XX, No. X, Month XXXX

Figure 1. Expressions of Epo and EpoR in colorectal cancer. A, Discrete immunoreactivity to Epo in normal mucosal epitheliumof colon remote from the tumor. B, C, There is a weak positive staining for Epo in mucosa in nearby colorectal cancer foci. Mucinof goblet cells pushes protoplasm in basal region of cell and the staining also is accumulated in this area of cytoplasm. D, Strongmixed cytoplasmic and membranous deposition of Epo in colorectal cancer cells. Some cells show enhanced accumulation of Epoin perinuclear area. E, Weak dispersed cytoplasm diffuse staining for Epo in colorectal cancer cells. F, G, Weak membranous andcytoplasmic expression of EpoR in normal epithelium of colon remote from the tumor. H, Anti-EpoR gives cytoplasmic pattern ofimmunoreactivity in short distance from cancer foci analogous to this one, which was observed in case of Epo in mucosa adheringto colorectal cancer fields. H, I, EpoR-positive cancer gland-like tubes. J, K, Cytoplasm and cell membranes stain for EpoR in col-orectal cancer in coarse granular pattern. Original magnification: A, B, D-F, H, I, ×100; C, G, J, K, ×200. Epo indicates erythropoi-etin; EpoR, erythropoietin receptor.

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

With respect to some discrepancies in case ofanti-EpoR C-20 antibody, it should be mentionedthat Elliott et al18 found it to label a few proteins ofa variegated molecular mass and concluded thatnone of them was EpoR. One of them was identifiedto be heat shock protein (HSP) 70, which success-fully competed with EpoR in binding with this pecu-liar antibody.18 In reaction to Elliott et al18 report,specificity of C-20 antibody was assumed to depend

on the batches of the antibody applied, so that char-acteristics of the batches ranged from high sensitiv-ity and specificity for EpoR to low or anyone.19 Onthe basis of Elliott et al18 underestimation of EpoRmolecular mass, it was concluded that anti-EpoRC-20 was not useful in immunohistochemistry forEpoR. However, Brown et al20 compared immuno-histochemical results of C-20 usage in nonsmall celllung cancers to show that immunoreactivities for

Erythropoietin and Its Receptor in Colorectal Cancer / Chabowska et al 5

Table 1. Statistical Analysis for Erythropoietin Expression and Erythropoietin Receptor in Colorectal CancerAccording to Clinicopathological Findings

Chi-square Chi-square Epo Test EpoR Test

Groups of Patients (n = 136) Negative Low High P Negative Low High P

Age≤60, n = 45 16 (35.56%) 18 (40.00%) 11 (24.44%) NS 8 (17.78%) 19 (42.22%) 18 (40.00%) NS>60, n = 91 26 (28.57%) 43 (47.25%) 22 (24.18%) 16 (17.58%) 32 (35.16%) 43 (47.25%)

SexMale, n = 69 21 (30.43%) 32 (46.38%) 16 (23.19%) NS 9 (13.04%) 29 (42.03%) 31 (44.93%) NSFemale, n = 66 21 (31.82%) 29 (43.94%) 16 (24.24%) 15 (22.73%) 21 (31.82%) 30 (45.45%)

LocalizationRectum, n = 64 20 (31.25%) 30 (46.88%) 14 (21.88%) NS 9 (14.06%) 26 (40.63%) 29 (45.31%) NSColon, n = 72 22 (30.56%) 31 (43.06%) 19 (26.39%) 15 (20.83%) 25 (34.72%) 32 (44.44%)

HP-statusAdenocarcinoma, 31 (27.19%) 52 (45.61%) 31 (27.19%) .05 17 (14.91%) 42 (36.84%) 55 (48.25%) .03

n = 114Mucinous carcinoma, 11 (50.00%) 9 (40.91%) 2 (9.09%) 7 (31.82%) 9 (40.91%) 6 (27.27%)

n = 22G-status

G2, n = 93 19 (20.43%) 44 (47.31%) 30 (32.26%) .001 12 (12.90%) 33 (35.48%) 48 (51.61%) .02G3, n = 43 23 (53.49%) 17 (39.53%) 3 (6.98%) 12 (27.91%) 18 (41.86%) 13 (30.23%)

pT-statuspT (1 + 2), n = 12 6 (50.00%) 3 (25.00%) 3 (25.00%) NS 3 (25.00%) 4 (33.33%) 5 (41.67%) NSpT (3 + 4), n=124 36 (29.03%) 58 (46.77%) 30 (24.19%) 21 (16.94%) 47 (37.90%) 56 (45.16%)

pN-statuspN (−), n = 61 14 (22.95%) 28 (45.90%) 19 (31.15%) NS 10 (16.39%) 18 (29.51%) 33 (54.10%) NSpN (+), n = 75 28 (37.33%) 33 (44.00%) 14 (18.67%) 14 (18.67%) 33 (44.00%) 28 (37.33%)

NOTE: Epo = erythropoietin; EpoR = erythropoietin receptor; HP = histopathologic; G = grading of histological differentiation; pT =depth of tumor intramural growth; pN = lymph node involvement; NS = not significant.

Table 2. Correlation of Immunohistochemical Expression of Erythropoietin With ErythropoietinReceptor in Colorectal Cancer Cells

Epo Spearman Rank Test

Groups of Patients, (n = 136) Negative(n = 42) Low (n = 61) High (n = 32) r P

EpoRNegative (n = 24) 20 (47.62%) 4 (6,56%) 0 (0.00%) 0.547 .00001Low (n = 51) 16 (38.10%) 27 (44.26%) 8 (24.24%)High (n = 61) 6 (14.29%) 30 (49.18%) 25 (75.76%)

NOTE: Epo = erythropoietin; EpoR = erythropoietin receptor.

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

both EpoR and HSP70 could overlap to producemixed membranous and cytoplasmic granular stain-ing both in neoplastic and nonneoplastic cells.20 IfC-20 antibody was preabsorbed with HSP70-5 pep-tide to avoid labeling of HSP70 in tissues, suchpreincubated antibody detected exclusively membra-nous pattern of EpoR distribution in bronchioalveo-lar cancer cells. So, the partial utility of C-20 can beaccomplished in immunohistochemical evaluationof EpoR. However, background staining that resultfrom cross-reacting with HSP70 is a certain disad-vantage.20 In our study, cytoplasm and cell mem-branes of colorectal cancer cells were stained withC-20 in granular fashion. Regarding the study ofBrown et al,20 EpoR protein is most likely showedwith membranous component of our staining withC-20. Usefulness of C-20 to detect EpoR has beenreported recently.21,22 It should not be also neglectedthat cancer cells can harbor numerous mutations,which might be responsible for cytoplasmic storageof some membranous proteins as in case of connexin26 in colorectal cancerogenesis.16,23

Overexpression of Epo and EpoR appeared as anconcomitant of moderately (G2) differentiated col-orectal cancer. Such a higher expression of Epo andEpoR could also be a strong evidence of accelerationof stimulating EpoR by Epo in acquisition of resem-blance of colorectal cancers to glandular epitheliumof large intestine in case of better differentiated can-cers. The expression of EpoR was strongly depend-ent on grading of cancer and angiogenesis in gastriccancers.17 It has also been shown that EpoR stainingwas significantly greater in tumors that showed highhistological grade in breast carcinoma.24 It is proba-bly because the better histological differentiationresults from improved supply of oxygen and trophicinfluence of hypoxia-induced Epo and EpoR signal-ing. Several studies show higher expression ofEpo/EpoR in solid tumors mainly due to hypoxiccore environment.8 This phenomenon might berelated to the role of Epo/EpoR signaling in thestimulation of neoangiogenesis in solid tumors,8,21,22

resulting in better differentiation of tumor cells.Additionally, Epo/EpoR signaling is highly conservedthroughout the species, suggesting that these pro-teins are produced by more differentiated cells.25

Furthermore, except these strong links of Epoand EpoR expression with the grade of differentiationof the tumors (G2 vs G3 grade) and histological typeof colorectal cancers, no other associations werefound between Epo/EpoR expression with regard to

depth of tumor ingrowth pT, node involvement pN,primary tumor site, or age and sex of patients withcolorectal cancers. Similar results showing no corre-lations between EpoR staining and other histopato-logical features of the squamous and neck tumorswere obtained by Hoogsteen et al.26 Those findingsmay indicate Epo/EpoR signaling as less involved intumor progression and malignancy.26

In the present study, we found strong correlationbetween Epo and EpoR expressions suggesting theirtight regulation and possible activity of the complexin human colorectal cancer cells. High expression ofboth Epo and EpoR in colorectal cancers could implytheir considerable function in this tissue. However,controversy still exists as to whether Epo/EpoRexpression underlies an essential functional signalingpathway in growth of cancer cells.27,28 Several in vitrostudies have shown that rHuEpo had no effect on theproliferation of EpoR-positive tumor cells.13,29,30 Ithas recently been suggested that EpoR plays a minorrole in tumor progression in neck squamous cancer.26

Consequently, it could be suspected that the EpoRsalthough present in malignant cells might be func-tionally inactive.29,30

However, the others have contradicted previousdata showing functional activity of Epo or EpoR sig-naling in tumor cell lines.31 It is rather widelyaccepted that Epo/EpoR signaling is affected bytumor hypoxia and may contribute to the worse sur-vival rates of patients by diminishing apoptoticpotential of these tumor cells.32 Studies using cellculture models clearly showed the proliferative andantiapoptotic effect in tumor cells after stimulationof EpoR. In this line of evidence, there are studiespresenting upregulation of Epo/EpoR expression inhuman breast cancers.32 Others have also shown apositive correlation between strong Epo and EpoRexpression but also associated poor prognosis withthis relationship of Epo and EpoR in human can-cers.8,33,34 Recent studies have supported this pointof view as they have shown a coexpression of Epo andEpoR in various solid tumors and cell lines.21,23,24,34,

In prostate cancer cells, coexpression of Epo/EpoRsuggested the potential autocrine or paracrinegrowth-promoting role. Other authors have alsoshowed rHuEpo can stimulate angiogenesis.10 Inspite of the fact that Hoogsteen et al26 reported thatEpoR had no significance for prognosis of head andneck cancer patients, the active Epo/EpoR system hasbeen also proposed to enhance cancer cell invasive-ness in human head and neck squamous cancers.33

6 International Journal of Surgical Pathology / Vol. XX, No. X, Month XXXX

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

Together, these studies suggest that Epo may influ-ence the development of certain tumor cellsalthough in colorectal cancers, definitive argumentshave not been provided so far.8,24,32,35 In our study,Epo and EpoR expressions failed to associate withnode metastases and depth of intramural growth.This is the reason why we cannot recognize theseproteins as eventual contributors to spread of col-orectal cancer.

This in vivo study provides evidences for thepresence of Epo and EpoR in human colorectal can-cer for the first time. Overexpression of Epo andEpoR appeared as an concomitant of moderately(G2) differentiated colorectal cancer. Such a higherexpression of Epo and EpoR could also be a strongevidence of acceleration of stimulating EpoR by Epoin acquisition of resemblance of colorectal cancersto glandular epithelium of large intestine in caseof better differentiated cancers. Apparently, over-expression of Epo and EpoR did not seem to be asso-ciated with progression of colorectal cancer in ourstudy because there was a lack of significant statis-tical association between Epo or EpoR expressionand nodal involvement or depth of intramuralgrowth. In light of our investigations, it is beyondany doubt that most colorectal cancers express Epoand EpoR. However, on the basis of our findings itshould be determined whether Epo supplementationto combat anemia of colorectal cancer patients doesnot exert a side effect to speed up the growth of col-orectal cancer.

Acknowledgment

This work was supported by a research grant fromthe Polish State Committee for Scientific Research(3994/P01/2007/32).

References

1. Lacombe C, Mayeux P. Biology of erythropoietin.Haematologica. 1998;83:724-732.

2. Lappin T. The cellular biology of erythropoietin recep-tors. Oncologist. 2003;8:15-18.

3. Silva M, Grillot D, Benito A, Richard C, Nunez G,Fernandez-Luna JL. Erythropoietin can promote ery-throid progenitor survival by repressing apoptosisthrough Bcl-XL and Bcl-2. Blood. 1996;88:1576-1582.

4. Stapley S, Peters TJ, Sharp D, Hamilton W. The mortal-ity of colorectal cancer in relation to the initial symptomat presentation to primary care and to the duration of

symptoms: a cohort study using medical records. Br JCancer. 2006;95:1321-1325.

5. Farrell F, Lee A. The erythropoietin receptor and itsexpression in tumor cells and other tissues. Oncologist.2004;9:18-30.

6. Acs G, Acs P, Beckwith SM, et al. Erythropoietin anderythropoietin receptor expression in human cancer.Cancer Res. 2001;61:3561-3565.

7. Hardee ME, Cao Y, Fu P, et al. Erythropoietin blockadeinhibits the induction of tumor angiogenesis and pro-gression. PLoS ONE. 2007;2:e549.

8. Acs G, Xu X, Chu C, Acs P, Verma A. Prognostic signif-icance of erythropoietin expression in human endome-trial carcinoma. Cancer. 2004;100:2376-2386.

9. Arany Z, Huang LE, Eckner R, et al. An essential rolefor p300/CBP in the cellular response to hypoxia. ProcNatl Acad Sci USA. 1996;93:12969-12973.

10. Ribatti D, Presta M, Vacca A, et al. Human erythropoi-etin induces a pro-angiogenic phenotype in culturedendothelial cells and stimulates neovascularization invivo. Blood. 1999;93:2627-2636.

11. Yasuda Y, Fujita Y, Masuda S, et al. Erythropoietin isinvolved in growth and angiogenesis in malignanttumours of female reproductive organs. Carcinogenesis.2002;23:1797-1805.

12. Yasuda Y, Fujita Y, Matsuo T, et al. Erythropoietin regulatestumour growth of human malignancies. Carcinogenesis.2003;24:1021-1029.

13. Hardee ME, Kirkpatrick JP, Shan S, et al. Humanrecombinant erythropoietin (rEpo) has no effect on tumourgrowth or angiogenesis. Br J Cancer. 2005;93:1350-1355.

14. Arcasoy MO, Amin K, Karayal AF, et al. Functional sig-nificance of erythropoietin receptor expression in breastcancer. Lab Invest. 2002;82:911-918.

15. Wincewicz A, Sulkowska M, Koda M, Lesniewicz T,Kanczuga-Koda L, Sulkowski S. STAT3, HIF-1α, EPO andEPOR-signaling proteins in human primary ductal breastcancers. Folia Histochem Cytobiol. 2007;45:81-86.

16. Kanczuga-Koda L, Sulkowski S, Lenczewski A, et al.Increased expression of connexins 26 and 43 in lymphnode metastases of breast cancer. J Clin Pathol. 2006;59:429-433.

17. Ribatti D, Marzullo A, Nico B, Crivellato E, Ria R,Vacca A. Erythropoietin as an angiogenic factor in gas-tric carcinoma. Histopathology. 2003;42:246-250.

18. Elliott S, Busse L, Bass MB, et al. Anti-Epo receptorantibodies do not predict Epo receptor expression.Blood. 2006;107:1892-1895.

19. Verdier F, Gomez S, Lacombe C, Mayeux P. Selectedanti-Epo receptor antibodies predict Epo receptorexpression [author reply]. Blood. 2006;108:1106.

20. Brown WM, Maxwell P, Graham AN, et al. Erythropoi-etin receptor expression in non-small cell lung carci-noma: a question of antibody specificity. Stem Cells.2007;25:718-722.

Erythropoietin and Its Receptor in Colorectal Cancer / Chabowska et al 7

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from

21. Ribatti D, Marzullo A, Gentile A, et al. Erythropoietin/erythropoietin-receptor system is involved in angiogene-sis in human hepatocellular carcinoma. Histopathology.2007;50:591-596.

22. Ribatti D, Poliani PL, Longo V, Mangieri D, Nico B,Vacca A. Erythropoietin/erythropoietin receptor systemis involved in angiogenesis in human neuroblastoma.Histopathology. 2007;50:636-641.

23. Kanczuga-Koda L, Sulkowski S, Koda M, Sulkowska M.Alterations in connexin 26 expression during colorectalcarcinogenesis. Oncology. 2005;68:217-222.

24. Acs G, Zhang PJ, Rebbeck TR, Acs P, Verma A.Immunohistochemical expression of erythropoietin anderythropoietin receptor in breast carcinoma. Cancer. 2002;95:969-981.

25. Wen D, Boissel JP, Tracy TE, et al. Erythropoietin struc-ture-function relationships: high degree of sequencehomology among mammals. Blood. 1993;82:1507-1516.

26. Hoogsteen IJ, Peeters WJ, Marres HA, et al. Erythropoietinreceptor is not a surrogate marker for tumor hypoxia anddoes not correlate with survival in head and neck squamouscell carcinomas. Radiother Oncol. 2005; 76:213-218.

27. Lai SY, Childs EE, Xi S, et al. Erythropoietin-mediatedactivation of JAK-STAT signaling contributes to cellularinvasion in head and neck squamous cell carcinoma.Oncogene. 2005;24:4442-4449.

28. Leo C, Horn LC, Rauscher C, et al. Expression of ery-thropoietin and erythropoietin receptor in cervical cancer

and relationship to survival, hypoxia, and apoptosis. ClinCancer Res. 2006;12:6894-6900.

29. Westphal G, Niederberger E, Blum C, et al. Erythropoi-etin and G-CSF receptors in human tumor cells: expressionand aspects regarding functionality. Tumori. 2002;88:150-159.

30. Rosti V, Pedrazzoli P, Ponchio L, et al. Effect of recom-binant human erythropoietin on hematopoietic andnon-hematopoietic malignant cell growth in vitro.Haematologica. 1993;78:208-212.

31. Westenfelder C, Baranowski RL. Erythropoietin stimu-lates proliferation of human renal carcinoma cells.Kidney Int. 2000;58:647-657.

32. Acs G, Chen M, Xu X, Acs P, Verma A, Koch CJ.Autocrine erythropoietin signaling inhibits hypoxia-inducedapoptosis in human breast carcinoma cells. Cancer Lett.2004;214:243-251.

33. Mohyeldin A, Lu H, Dalgard C, et al. Erythropoietin sig-naling promotes invasiveness of human head and necksquamous cell carcinoma. Neoplasia. 2005;7:537-543.

34. Acs G, Zhang PJ, McGrath CM, et al. Hypoxia-inducibleerythropoietin signaling in squamous dysplasia andsquamous cell carcinoma of the uterine cervix and itspotential role in cervical carcinogenesis and tumor pro-gression. Am J Pathol. 2003;162:1789-1806.

35. Sulkowska M, Wincewicz A, Chabowska A, et al. To giveor not to give recombinant Epo to anemia endangeredcancer patients. Prague Med Rep. 2006;107:281-290.

8 International Journal of Surgical Pathology / Vol. XX, No. X, Month XXXX

at Jazan University on November 25, 2013ijs.sagepub.comDownloaded from