ANTICANCER RESEARCH 29: 4833-4838 (2009)

Immunohistochemistry of DNA Mismatch Repair Enzyme MSH2 Is Not Correlated with Prognostic Data from Endometrial Carcinomas

ANDREAS SCHRÖER1, FRANK KÖSTER1, DOROTHEA FISCHER1, RALPH MARTIN DUBITSCHER2, ASTRID WOLL-HERMANN2, KLAUS DIEDRICH1, MICHAEL FRIEDRICH3 and DARIUS SALEHIN3

1Department of Gynecology and Obstetrics, University of Lübeck, D-23538 Lübeck; 2Department of Gynecology and Obstetrics, Saarland University Hospital, D-66421 Homburg/Saar; 3Department of Gynecology and Obstetrics, HELIOS-Hospital of Krefeld, D-47805 Krefeld, Germany

Abstract. Background: The human Mut-S-homolog-2 (MSH2) total loss of an MMR gene due to mutation, the expression is part of the DNA mismatch repair system (MMR). Mutations values of the MMR proteins in spontaneous cancer might be in genes of the MMR are a predisposition to hereditary non- important for the prognosis of the disease. The human Mut-S- polyposis colorectal cancer (HNPCC). In women, MMR gene homolog-2 (MSH2) is one of the key enzymes of the MMR. mutations may lead to primary endometrial cancer (EC). The The MMR is responsible for the detection and the repair of important function of the MMR for the integrity of the DNA DNA base-base mismatches or insertion/deletion-loop during replication makes it probable that the MMR might also mismatches during replication and consists of the two be involved in the development and the course of sporadic heterodimers Mut-S (MSH2 with MSH3 or MSH6) and Mut-L carcinomas. Insufficient MMR activity or expression levels (Mut-L-homolog-1 [MLH1] with post-meiotic-segregation- could be prognostic markers of the disease. Patients and increased-2 [PMS2]). Mutations in one of the genes coding for Methods: Immunohistochemical analysis of MSH2 was the MMR enzymes MSH2, MLH1 or PMS2 result in an performed in 86 tumor samples from patients with EC. Results: insufficient DNA repair function during replication that can lead Compared to known tumor markers, namely estrogen and to HNPCC associated with the appearance of microsatellite progesterone receptors, histopathological grading, TNM stage instabilities (MSI), (1-3). In women mutations in MMR genes and FIGO classification, no significant correlation between are also often responsible for primary EC (4-6). Recently, Garg MSH2 immunoreactivity and EC was found. Conclusion: et al. have suggested screening criteria for women with a MSH2 immunohistochemical analysis is not of prognostic value diagnosed EC that should be routinely screened for MMR for endometrial carcinoma. deficiencies and hence the risk of HNPCC (7). It is a matter of discussion whether the intratumor Mutations in genes of the DNA mismatch repair system (MMR) expression levels of MMR enzymes in sporadically derived are known to be associated with the inheritance of hereditary carcinomas can have any prognostic value (8-16). Because non-polyposis colorectal cancer (HNPCC) also known as Lynch of its important role for the integrity of the DNA during the syndrome. In women, the most important extra-colonic tumors replication of the chromosome, insufficient MMR activity in Lynch syndrome are endometrial carcinomas (EC), indicating can lead to further genetic aberrations and increasing tumor that primary carcinomas of the endometrium are associated with malignancy. Furthermore, the MMR system is probably defects in the MMR. EC is still one of the most common involved in the sensing of DNA damage induced by therapy malignancies in women, with an estimated 150,000 new cases with alkylating cytotoxic agents. Therefore, the expression per year worldwide. In addition to malignancies derived from a levels of MMR proteins are also of interest for their role in the transmission of cytotoxic activity resulting from DNA damage induced by cytostatic agents (17-23). Prognostic and predictive values of MSH2 expression have been explored in Correspondence to: Dr. Frank Köster, Department of Gynecology various tumor tissues. Previous studies focused on and Obstetrics, University of Lübeck, Ratzeburger Allee 160, D- endometrial tumors have reported increased MSH2 23538 Lübeck, Germany. Tel: +49 4515002744, Fax: +49 4515004905, e-mail: [email protected] expression in malignant endometrial tissue, both at protein and mRNA levels (24, 25). Others have found an association Key Words: Endometrial cancer, DNA mismatch repair, MSH2, between the expression of MMR proteins and the survival of immunohistochemistry. patients with EC (26).

0250-7005/2009 $2.00+.40 4833 ANTICANCER RESEARCH 29: 4833-4838 (2009)

Here, the expression levels of MSH2 were correlated with Situ Cell Death Detection Kit, with alkaline phosphatase (AP; the tumor suppressor p53, the proliferation marker Ki-67, the Roche, Mannheim, Germany). The slides were deparaffinated and appearance of apoptosis and with clinical data from 86 rehydrated. Proteinase-K digestion was performed for 25 min at room temperature. The TUNEL reaction was undertaken at 37˚C for patients with EC. 1 h under a coverslip. For the detection of fluorescein-dUTP- labelled cells, we used an AP-linked anti-fluorescein antibody. The Patients and Methods immunoreaction was visualized with the New Fuchsin Substrate Kit (Dako, Hamburg, Germany). Tumor tissue. Tumor specimens from 86 carcinoma of the corpus uteri or the endometrium were collected between 1985 and 1990 Semi-quantitative analysis of immunoreactivity. Microscopic from patients during inter-surgery diagnosis at the female hospital of analysis was performed by two independent observers (R.M.D. and the University of Homburg/Saar, Germany. The patients were A.W.-H.). The immunoassayed specimens were viewed under a informed about and gave their written consent to their participation Leica DM4000 B microscope (Leica Microsystems, Wetzlar, in this study. The study was approved by the Ethics Committee of Germany). the University of Homburg/Saar. Residues of tumor samples were The interpretation of immunoreactivity in the tumor sections was collected after histological examination by pathologists. Grading performed using the immunoreactive score (IRS) ranging from 0 to and staging of the tumors as well as the status of steroid receptors 12 (negative=0-1; weak=2-3; moderate=4-8; strong=8-10) described were taken from the pathologists’ reports. The tumor specimens by Remmele and Stegner (27). The IRS results from the were fixed in 4% neutral buffered formalin and embedded in multiplication of a staining intensity (SI) score (negative=0; weak=1; paraffin for storage. moderate=2; strong=3) and the values of percentage of positive tumor cells (PP-score: 0-10%=1; 11-50%=2; 51-80%=3; 81-100%=4). The Slide preparation. Glass slides were silane-coated, pretreated with immunoreactivity for Ki-67 was evaluated and compared with other 0.5% ovarian albumin and dried at 60˚C overnight. Serial sections immunoreactivities with a scoring after Friedrich et al. (24) that of the tumor specimens of 6-7 μm thickness were cut on a rotation creates a ranking of the percentages of immunopositive cells (PP) microtome (Leitz 1516, Wetzlar, Germany). After floating on 37˚C (rank 0=0%; 1=1–10%; 2=11–25%; 3=>25%). distilled water the sectiones were mounted on the pretreated microscopic slides. Statistical analysis of correlation. The statistical correlations between For the staining procedure the slides were deparaffinized in a MSH2, p53 and apoptosis immunoreactivity and with patient data xylene-bath and rehydrated in decreasing concentrations of ethanol. were analyzed by the Spearman’s rank correlation coefficient in Finally they were washed in Tris-buffer. Endogenous peroxidase SPSS version 10.0 (SPSS GmbH Software, München, Germany). was blocked by incubation in a freshly prepared solution of 3% peroxidase in methanol for 10 min at room temperature. The Results sections were pretreated in a solution of 10 mM citrate-buffer, pH 6.0 by heating in a microwave-oven at 500 watts for 10 min Expression of MSH2, p53 and apoptosis. The antibody followed by two washes in distilled water. against MSH2 reacted positively in 72.1% of the tumor Immunohistochemistry. Nonspecific antibody reactions were blocked specimens: 21% with strong IRS, 48.8% with moderate, in a blocking-solution of Tris-buffer with 0.2% Triton-X-100 and 2.3% with weak and 27.9% were negative. The average IRS 1% fish gelatine. The tumor specimens were incubated with score was 4.78±3.48. Among the 86 EC specimens, 47.7% monoclonal MSH2 antibody clone FE-11 (Calbiochem, Merck did not react with the p53 antibody, 23.3% were strongly Biosciences, Schwalbach, Germany) diluted 1:50 in blocking positive, 25.5% were moderately and 3.5% were weakly solution. The sections were covered with a coverslip and the stained. The average IRS score was 3.59±3.83. The reaction took place at 4˚C overnight in a humid chamber. Against p53, the monoclonal mouse antibody clone DO-7 at 1:150 dilution immunoreaction after the TUNEL assay was positive in and for Ki-67 the polyclonal rabbit antibody A-047 at 1:50 dilution 80.2% of the specimens, with a mean IRS of 3.57±2.25. were used (both Dako, Hamburg, Germany). Strong reactivity was observed in 3.5% of the tumor For detection of the primary antibody, the avidin-biotin-peroxidase samples, moderate in 62.6% and weak in 13.9%; 19.8% of complex (ABC) method (Dako) was used. The secondary biotinylated the specimens did not show staining for apoptosis. canine anti-mouse antibody was applied at 1:200 dilution in fish- gelatine for 20 min at room temperature followed by incubation with Correlation of MSH2 expression with p53, Ki-67 and the streptavidin-biotin complex for 30 min at 37˚C. After each incubation, two washes were performed. The peroxidase activity was apoptosis. No correlation of MSH2 expression was observed visualized with 3,3’-diaminobenzidine (DAB, Dako). The sections with p53, apoptosis or the proliferation marker Ki-67 (Table were counterstained in a hematoxylin solution for 10 min followed I). Furthermore, no correlations were found of p53 by two washes in distilled water. The slides were dehydrated in expression with apoptosis, p53 expression with Ki-67 or increasing ethanol concentrations and a final xylene bath before they between apoptosis and Ki-67. were covered with Entellan® (Merck, Darmstadt, Germany). Correlation of MSH2 immunoreactivity with clinical Immunoreactive detection of apoptosis after TdT-mediated dUTP nick-end labelling (TUNEL) assay. For the detection of apoptotic parameters. Regarding the histopathological differentiation activities in tissue sections of the tumor samples, we used the In grade of the tumor cells, the highest expression of MSH2

4834 Schröer et al: MSH2 in Endometrial Cancer

Table I. Correlation between MSH2 p53 and Ki-67 expression and Table II. MSH2 immunoreactivities correlated with the grade of apoptosis following TUNEL assay. histopathological differentiation.

IRS [rs] PP-score [rs] Grading N MSH2 IRS±SEM rs: 0.108; p=0.322 MSH2 – p53 0.091 (p=0.497) 0.085 (p=0.435) MSH2 – apoptosis –0.025 (p=0.817) 0.064 (p=0.559) G1 30 (37.5%) 4.72±3.43 MSH2 – Ki-67 0.036 (p=0.739) G2 29 (33.7%) 4.28±3.30 p53 – apoptosis 0.133 (p=0.224) 0.142 (p=0.194) G3 27 (31.4%) 5.48±3.46 p53 – Ki-67 0.027 (p=0.807) Apoptosis – Ki-67 0.119 (p=0.277) rs, Spearman’s rank correlation coefficient; IRS, immunoreactive score; SEM, standard error of the means. IRS, Immunoreactive score; PP, percentage immunopositivity; rs, Spearman’s rank correlation coefficient; statistical significance: p<0.05.

Table III. MSH2 immunoreactivities correlated with clinical data.

ER PR T stage N stage M stage Histopath. grading FIGO stage

MSH2 IRS [rs] 0.074 0.117 0.048 –0.330 0.030 0.108 0.027 p-Value 0.497 0.284 0.668 0.767 0.790 0.322 0.810

ER, Estrogen receptor; PR, progesterone receptor; rs, Spearman’s rank correlation coefficient; statistical significance: p<0.05; FIGO, Fédération Internationale de Gynécologie et d’Obstétrique.

was found in the group of tumors with grading G3 and an EC (28). By contrast, Cohn et al. reported 29% of endometrial IRS score of 5.48 (Table II). Nevertheless, Spearman’s cancer patients with MMR deficiency, where either MSH2 or correlation rank test over all the samples did not show a MLH1 expression was lacking (26). A total loss of MSH2 statistically significant correlation between the grade of immunoreactivity could be caused by a mutation in one of the differentiation and the expression of MSH2. genes of the Mut-S heterodimer, namely MSH2 or MSH6. The MSH2 immunoreactivities were tested for correlation When one of the binding partners for the MMR heterodimers with further clinical parameters (Table III). The IRS of the is missing, the other will be undetectable as well. For the immunostainings for estrogen and progesterone receptors (ER interpretation of our results, we used the IRS values for and PR) did not correlate with the IRS for MSH2. With regard correlation statistics instead of differentiating strictly between to the TNM classification neither the tumor size (T stage), nor MSH2 expressing and non-expressing patients. lymph node infiltration (N stage), nor the appearance of No statistically relevant correlation was found between the metastases (M stage) correlated with the immunoreactivity for MSH2 IRS and that for tumor suppressor p53, the proliferation MSH2. Furthermore, the FIGO (Fédération Internationale de marker Ki-67 and the TUNEL assays. In previous studies using Gynécologie et d’Obstétrique) stages of the tumors did not samples from breast cancer or cervical cancer, we have correlate with the MSH2 immunostaining. observed correlations between MSH2 immuno-reactivity with that scored for p53 and apoptosis (12, 13). However, Giarnieri Discussion et al. working with samples from the uterine cervix found a correlation of high p53 immunoreactivity and the loss of MSH2 As expected, a high percentage of the tumor samples were reactivity (29). The interpretation of p53 immunoreactivity is found to express MSH2 protein and most of them exhibited difficult because of the fast degradation of the protein. P53 moderate to strong immunoreactivities. The mean IRS of 4.78 mutations frequently occur in tumors and are associated with was relatively lower than that analyzed in previous studies in unfavorable prognosis (30, 31). Of note, a marked p53 EC (mean IRS of about 9.00) (24, 25). These expression and activity may be associated with MSH2 immunohistochemical studies used frozen tissue instead of expression since a binding site for p53 has been detected in the paraffinized tumor material. The finding of 27.9% of MSH2 promoter region of the MSH2 gene (32, 33). The interpretation negative probes was in a sharp contrast to other studies e.g. of apoptosis after the TUNEL assay in paraffinized material can by Peiro et al. who found only one case in 89 patients with be associated with unspecific labeling reactions. In accordance the total loss of MSH2 expression in paraffinized material of with the results from Labat-Moleur et al., we used formalin

4835 ANTICANCER RESEARCH 29: 4833-4838 (2009) fixation and a proteinase-K pretreatment because this 5 Lynch HT, Casey MJ, Shaw TG and Lynch JF: Hereditary methodical combination achieves the highest specificity (34). factors in gynecologic cancer. Oncologist 3: 319-338, 1998. No correlation was found in this study between the 6 Watson P, Riley B: The tumor spectrum in the lynch syndrome. immunoreactivity of MSH2 and the proliferation marker Ki-67. Fam Cancer 4: 245-248, 2005. 7 Garg K, Leitao MM Jr, Kauff ND, Hansen J, Kosarin K, Shia Because the MMR is highly active in the posttranscriptional J and Soslow RA: Selection of endometrial carcinomas for phase of proliferating cells, one could speculate that a strong DNA mismatch repair protein immunohistochemistry using expression of MSH2 is exclusively associated with the patient age and tumor morphology enhances detection of proliferative activity of these tumor cells. Due to the non- mismatch repair abnormalities. Am J Surg Pathol 33: 925-933, existence of correlation between MSH2 and Ki-67 an unknown 2009. mechanism in the neoplastic tissue must be suggested for the 8 Adem C, Soderberg CL, Cunningham JM, Reynolds C, Sebo TJ, regulation of MSH2 expression. Thibodeau SN, Hartmann LC and Jenkins RB: Microsatellite instability in hereditary and sporadic breast cancers. Int J Cancer The correlations of MSH2 immunoreactivities with the 107: 580-582, 2003. clinical data of the patients did not show any significant 9 Ciavattini A, Piccioni M, Tranquilli AL, Filosa A, Pieramici T result. The highest IRS value of MSH2 was found in the and Goteri G: Immunohistochemical expression of DNA patients with a histopathological grading of G3 according to mismatch repair (MMR) system proteins (hMLH1, hMSH2) in the worst prognosis, but the correlation in patients overall did cervical preinvasive and invasive lesions. Pathol Res Pract 201: not reach statistical significance. This result was in agreement 21-25, 2005. with the study of Peiro et al. who were also unable to find a 10 Friedrich M, Meyberg R, Villena-Heinsen C, Woll-Hermann A, Reitnauer K, Schmidt W, Tilgen W and Reichrath J: correlation between MSH2 immunohistochemical analysis Immunohistochemical analysis of DNA mismatch-repair enzyme and histopathological grading (28). hMSH-2 and Ki-67 in breast carcinomas. Anticancer Res 19: With regard to other classical prognostic and predictive 3349-3353, 1999. tumor markers, namely ER and PR, TNM stage and FIGO 11 Friedrich M, Villena-Heinsen C, Meyberg R, Woll-Hermann A, classification, no significant correlations were detected with Reitnauer K, Schmidt W, Tilgen W and Reichrath J: the immunohistochemical analysis of MSH2 expression. Cohn Immunohistochemical analysis of DNA ‘mismatch-repair’ et al. analyzed the MSH2 and MLH1 expression in 336 enzyme human mut-S-homologon-2 in ovarian carcinomas. 31 patients with EC on a tissue microarray. They found a Histochem J : 717-722, 1999. 12 Koster F, Schroer A, Fischer D, Greweldinger T, Diedrich K and correlation between the lack of MSH2 or MLH1 and negative Friedrich M: Correlation of DNA mismatch repair protein prognostic factors, as well as a worse progression-free survival hMSH2 immunohistochemistry with p53 and apoptosis in (26). Garg et al., searching for HNPCC-screening criteria for cervical carcinoma. Anticancer Res 27: 63-68, 2007. patients with EC, found that the age of patients and a family or 13 Koster F, Schroer A, Fischer D, Horn AK, Diedrich K and personal history of HNPCC-associated carcinomas will Friedrich M: Immunohistochemistry of proteins for DNA underestimate the risk for HNPCC. Therefore, they suggest mismatch repair in correlation to prognostic factors of that tumor morphology and the analysis of MMR protein mammarian cancer. Oncol Rep 17: 1223-1227, 2007. 14 Mylona E, Zarogiannos A, Nomikos A, Giannopoulou I, defects by immunhistochemistry for should be included (7). Nikolaou I, Zervas A and Nakopoulou L: Prognostic value of These results indicate that the immunoreactive score of microsatellite instability determined by immunohistochemical MSH2 is not a useful prognostic test for endometrial carcinoma. staining of HMSH2 and HMSH6 in urothelial carcinoma of the bladder. APMIS 116: 59-65, 2008. References 15 Cooper WA, Kohonen-Corish MR, Chan C, Kwun SY, McCaughan B, Kennedy C, Sutherland RL and Lee CS: 1 Aaltonen LA, Peltomaki P, Leach FS, Sistonen P, Pylkkanen L, Prognostic significance of DNA repair proteins MLH1, MSH2 Mecklin JP, Jarvinen H, Powell SM, Jen J, Hamilton SR et al: and MGMT expression in non-small cell lung cancer and Clues to the pathogenesis of familial colorectal cancer. Science precursor lesions. Histopathology 52: 613-622, 2008. 260: 812-816, 1993. 16 Erdamar S, Ucaryilmaz E, Demir G, Karahasanoglu T, Dogusoy 2 Fishel R, Lescoe MK, Rao MR, Copeland NG, Jenkins NA, G, Dirican A and Goksel S: Importance of mutL homologue Garber J, Kane M and Kolodner R: The human mutator gene MLH1 and mutS homologue MSH2 expression in Turkish homolog MSH2 and its association with hereditary nonpolyposis patients with sporadic colorectal cancer. World J Gastroenterol colon cancer. Cell 75: 1027-1038, 1993. 13: 4437-4444, 2007. 3 Leach FS, Nicolaides NC, Papadopoulos N, Liu B, Jen J, 17 Chintamani, Jha BP, Bhandari V, Bansal A, Saxena S and Parsons R, Peltomaki P, Sistonen P, Aaltonen LA, Nystrom-Lahti Bhatnagar D: The expression of mismatched repair genes and M et al: Mutations of a mutS homolog in hereditary their correlation with clinicopathological parameters and nonpolyposis colorectal cancer. Cell 75: 1215-1225, 1993. response to neo-adjuvant chemotherapy in breast cancer. Int 4 Aarnio M, Sankila R, Pukkala E, Salovaara R, Aaltonen LA, de Semin Surg Oncol 4: 5-12, 2007. la Chapelle A, Peltomaki P, Mecklin JP and Jarvinen HJ: Cancer 18 Christmann M and Kaina B: Nuclear translocation of mismatch risk in mutation carriers of DNA-mismatch-repair genes. Int J repair proteins MSH2 and MSH6 as a response of cells to Cancer 81: 214-218, 1999. alkylating agents. J Biol Chem 275: 36256-36262, 2000.

4836 Schröer et al: MSH2 in Endometrial Cancer

19 Claij N and te Riele H: Msh2 deficiency does not contribute to 28 Peiro G, Diebold J, Mayr D, Baretton GB, Kimmig R, Schmidt cisplatin resistance in mouse embryonic stem cells. Oncogene M and Lohrs U: Prognostic relevance of hMLH1, hMSH2, and 23: 260-266, 2004. BAX protein expression in endometrial carcinoma. Mod Pathol 20 Clodfelter JE, M BG and Drotschmann K: MSH2 missense 14: 777-783, 2001. mutations alter cisplatin cytotoxicity and promote cisplatin- 29 Giarnieri E, Mancini R, Pisani T, Alderisio M and Vecchione A: induced genome instability. Nucleic Acids Res 33: 3323-3330, Msh2, Mlh1, Fhit, p53, Bcl-2, and Bax expression in invasive 2005. and in situ squamous cell carcinoma of the uterine cervix. Clin 21 Drummond JT, Anthoney A, Brown R and Modrich P: Cisplatin Cancer Res 6: 3600-3606, 2000. and adriamycin resistance are associated with mutLalpha and 30 Athanassiadou P, Athanassiades P, Grapsa D, Gonidi M, mismatch repair deficiency in an ovarian tumor cell line. J Biol Athanassiadou AM, Stamati PN and Patsouris E: The prognostic Chem 271: 19645-19648, 1996. value of PTEN, p53, and beta-catenin in endometrial carcinoma: 22 Fedier A, Schwarz VA, Walt H, Carpini RD, Haller U and Fink A prospective immunocytochemical study. Int J Gynecol Cancer D: Resistance to topoisomerase poisons due to loss of DNA 17: 697-704, 2007. mismatch repair. Int J Cancer 93: 571-576, 2001. 31 Kalogiannidis I, Bobos M, Papanikolaou A, Makedos A, 23 Runger TM, Emmert S, Schadendorf D, Diem C, Epe B and Amplianitis I, Vergote I, Nenopoulou E and Makedos G: Hellfritsch D: Alterations of DNA repair in melanoma cell lines Immunohistochemical bcl-2 expression, p53 overexpression, PR resistant to cisplatin, fotemustine, or etoposide. J Invest and ER status in endometrial carcinoma and survival outcomes. Dermatol 114: 34-39, 2000. Eur J Gynaecol Oncol 29: 19-25, 2008. 24 Friedrich M, Villena-Heinsen C, Reitnauer K, Schmidt W, Tilgen 32 Scherer SJ, Seib T, Seitz G, Dooley S and Welter C: Isolation W and Reichrath J: Malignancies of the uterine corpus and and characterization of the human mismatch repair gene hMSH2 immunoreactivity score of the DNA “mismatch-repair” enzyme promoter region. Hum Genet 97: 114-116, 1996. human mut-S-homologo-2. J Histochem Cytochem 47: 113-118, 33 Scherer SJ, Welter C, Zang KD and Dooley S: Specific in vitro 1999. binding of p53 to the promoter region of the human mismatch 25 Friedrich M: Analysis of the DNA “mismatch-repair” enzyme repair gene hMSH2. Biochem Biophys Res Commun 221: 722- human mut-S-homologon-2 in endometrial cancer on protein 728, 1996. and RNA level. Eur J Gynaecol Oncol 21: 273-277, 2000. 34 Labat-Moleur F, Guillermet C, Lorimier P, Robert C, Lantuejoul 26 Cohn DE, Frankel WL, Resnick KE, Zanagnolo VL, Copeland S, Brambilla E and Negoescu A: TUNEL apoptotic cell detection LJ, Hampel H, Kelbick N, Morrison CD and Fowler JM: in tissue sections: Critical evaluation and improvement. Improved survival with an intact DNA mismatch repair system J Histochem Cytochem 46: 327-334, 1998. in endometrial cancer. Obstet Gynecol 108: 1208-1215, 2006. 27 Remmele W and Stegner HE: Recommendation for uniform definition of an immunoreactive score (IRS) for immunohisto- Received June 3, 2009 chemical estrogen receptor detection (ER-ICA) in breast cancer Revised October 5, 2009 tissue. Pathologe 8: 138-140, 1987. Accepted October 7, 2009

4837