Objective
The purpose of this study was to assess the value of the 2003 World Health Organization (WHO) and endometrial intraepithelial neoplasia (EIN) classifications, D-score, and molecular biomarkers in endometrial hyperplasia (EH) for cancer progression.
Study Design
We conducted a review of 307 endometrial hyperplasias for WHO and EIN classifications and an analysis of biomarkers, D-score, and cancer progression-free survival.
Results
The WHO, EIN, D-score, and many biomarkers were prognostic; 7.2% of the samples progressed to cancer. The WHO and EIN classifications correlated weakly with CK5/6 and p16. The D-score was strongest prognostically. When >1, it had the lowest false-negative progression rate of all features analyzed. COX2 negativity was the only other independent multivariate cancer progression predictor in endometrial hyperplasia, but only in cases with D-score <1. Eight of 13 cases (61%), with a combined D-score of <1 and negative COX2 progressed, which contrasted with 3 of 139 of all other cases (2.8%) ( P < .0001; hazard ratio, 53.0). The biomarkers did not strengthen the prognostic value of the WHO or EIN classification.
Conclusion
Combined D-score <1 and COX2 negativity strongly predict cancer progression in endometrial hyperplasias.
The estimated incidence of endometrial cancer in the United States was 40,880 for 2005 (6% of all cancers), with an estimated lifetime probability of cancer development of 1 in 38 women. Endometrial hyperplasia (EH) is an excessive proliferation of endometrial glands that leads to a higher gland/stroma ratio, compared with normal endometrium. EH is much more frequent than endometrial cancer ; 5-10% of untreated EHs progress to endometrial endometrioid adenocarcinoma.
EH can be evaluated by the 2003 World Health Organization (WHO) or the more current approach, endometrial intraepithelial neoplasia (EIN) system. The 2003 WHO classification is based on glandular crowding and cytologic atypia that resulted in 4 subgroups with increasing cancer risk. Despite its worldwide use, the inter- and intraobserver reproducibility of cytologic atypia (the most important of the 2 prognostic WHO factors) is not very good. The EIN classification distinguishes benign reactive hyperplasias (a polyclonal lesion), which is the result of excessive stimulation by estrogens, and EIN, which is a monoclonal neoplasm.
The EIN classification is based on morphometric data and molecular monoclonality. The morphometric D-score appears to be most reproducible of all existing prognostic classifications and prognostically to be the strongest. However, the computerized morphometric analysis technology that is required for D-score assessment is not available widely at the moment. Although this will most likely change with the advent of digital pathologic evaluation, it would be preferable to have strong molecular immunohistochemical outcome predictors.
Regarding molecular clonality, EIN is the result of consecutive mutations that involve genes coding for tumor suppressors, oncogenes, cell cycle regulators, apoptosis inhibitors and inducers, mismatch repair, and other genes that are involved in controlling growth and differentiation of the cells. Inactivation of the tumor suppressor gene phosphatase and tensin homolog (PTEN) is a frequent and so far the earliest known cancer precursor event. PTEN has been reported to have a prognostic value in EH. However, the value of other molecular biomarkers is uncertain because most molecular studies of the endometrium have been based on small numbers of cases and have analyzed endometrial carcinoma rather than its precursors ( Table 1 ).
| Biomarker | Year | Study | Cases, n | Expression in endometrial hyperplasia | Prognostic value (univariate) |
|---|---|---|---|---|---|
| Survivin | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; rising from EH to EC | Prognostic ( P = .049) |
| 2009 | Chen X et al | 23 | High | ||
| 2007 | Erkanli S et al | 30 | High | ||
| 2006 | Erkanli S et al | 38 | High | ||
| P21 a | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; rising from EH to EC | Not prognostic |
| 2009 | Brucka A et al | 61 | High | ||
| 2008 | Cobellis L et al | 20 | Low | ||
| 2008 | Horre N et al | 23 | High | ||
| P16 a | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | High; rising from EH to EC. | Prognostic ( P < .0001) |
| 2008 | Horre N et al | 23 (13 non-EIN, 10 EIN) | High; higher in EIN than non-EIN | ||
| P27 | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; falling in EC | Prognostic ( P = .010) |
| 2008 | Horre N et al | 23 (13 non-EIN, 10 EIN) | Lower in EIN | ||
| 2006 | Erkanli S et al | 38 | Low | ||
| 2004 | Ozkara SK et al | 24 | High in CAH; low in simple EH | ||
| 2003 | Masciullo V et al | 29 | Low | ||
| P53 a | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; rising from EH to EC | Prognostic ( P = .038) |
| 2008 | Horre N et al | 23 (13 non-EIN, 10 EIN) | Rising from inactive through EIN to EC | ||
| 2004 | Ozkara SK et al | 24 (12 with atypia) | None in NE or EH | ||
| 2003 | Maia H Jr et al | 54 (33 have used estrogen) | Few scattered positive cells, moderate intensity in 15-83%, dependent on estrogen use (lower in estrogen users) | ||
| 2002 | Cinel L et al | 19 (9 SH, 2 SAH, 4 CH, 4 CAH) | 50% in EH, rising in EC | ||
| 2002 | Sakuragi N et al | 13 (3 atypical) | None inn EH | ||
| 2001 | Elhafey AS et al | 40 (10 atypical) | None in non-atypical EH, 30% in atypical EH, and rising in EC. | ||
| 2000 | Ioachim EF et al | 34 (10 SH, 16 CH and 8 atypical hyperplasia) | None in EH | ||
| P63 | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; falling in EC | Not prognostic |
| 2004 | Hu WF et al | 3 | Higher in EC than EH | ||
| PTEN | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Lower in EC than EH | Prognostic ( P = .003) |
| 2008 | Lacey JV et al | 138 | Low in EH | Not prognostic | |
| 2008 | Tantbirojn P et al | 55 | Decreasing from PE to non-atypical to atypical to EC | ||
| 2007 | Kapucuoglu N et al | 37 | Lower in EC than EH | ||
| 2007 | Norimatsu Y et al | 38 EIN | No PTEN negativity in PE; 34% negativity in EIN | ||
| 2006 | Cirpan T et al | 37 EIN | No difference among PE, EH, or EC | ||
| 2006 | Erkanli S et al | 38 | Decreasing from PE to EH to EC | ||
| 2005 | Baak JP et al | 103 | Low | Prognostic (33% progression in EIN; prognostic in combination with D-score) | |
| 2003 | Gao QL et al | 96 | Lower in EC than EH and NE | ||
| 2003 | Orbo A et al | 68 | Decreasing from non-atypical EH (12%) to atypical EH (14%) to EC (30%) | ||
| Cyclin E | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | High | Not prognostic |
| 2009 | Brucka A et al | 61 | High | ||
| 2007 | Horre N et al | 23 | High | ||
| 2006 | Kayaselcuk F et al | 38 | High | ||
| 2003 | Kato N et al | 20 | Low; rising in EC | ||
| Her-2 a | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low in EH; higher in EC | Not prognostic |
| 2004 | Hu WF et al | 23 | No difference among PE, EH, and EC | ||
| 2001 | Da J et al | 10 atypical hyperplasia | High in EH; lower in EC | ||
| β-catenin a | 2010 | Current study | 152 (134 non-EIN, 18 EIN) | Nuclear stain vs cytoplasm/membranous stain | Prognostic ( P = .002) |
| 2009 | Liao X et al | 51 | High nuclear in atypical EH | ||
| 2007 | Norimatsu Y et al | 38 EIN/32 BRH | High nuclear in EIN | ||
| 2005 | Brachtel EF et al | 24 CAH | High nuclear in squamous morules | ||
| 2003 | Moreno-Bueno G et al | 21 CAH | High nuclear | ||
| 2002 | Ashihara K et al | 25 | High nuclear | ||
| 2001 | Saegusa M et al | 37 simple and complex/32 atypical | High nuclear in atypical EH | ||
| Bcl-2 a | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | High | Prognostic ( P = .026) |
| 2007 | Kapucuoglu N et al | 37 | 100% in NE, SH, and CH; lower in CAH (80%). but higher (97%) in EC | ||
| 2003 | Mitselou A et al | 35 adenomatous hyperplasia | High in EH, lower in EC | ||
| 2003 | Xiang DJ et al | 10 SH and 10 atypical hyperplasia | No difference among NE, EH, and EC | ||
| 2002 | Bozdogan O et al | 26 | Lower in EC than EH | ||
| 2002 | Cinel L et al | 19 (9 SH, 2 SAH, 4 CH, 4 CAH) | Lower in EH, but rising in EC | ||
| 2002 | Risberg B et al | 60 (46 SH, 9 CH, 5 atypical hyperplasia) | Bcl-2 highest in PE; Bcl-2 significantly lower in EC than EH ( P = .002) | ||
| 2002 | Sakuragi N et al | 13 (3 with atypia) | Lowering from PE through EH to EC | ||
| 2002 | Vaskivuo TE et al | 52 (17 SH,12 CH, 27 atypical hyperplasia) | Lowering from PE through degree of EH to EC | ||
| 2001 | Kokawa K et al | 16 (7 with atypia) | Lowering from nonatypical EH to EC | ||
| 2001 | Peiro G et al | 32 (10 simple and 22 complex) | Lowering from PE through EH to EC | ||
| 2000 | Morsi HM et al | 20 (12 simple and 8 complex) | Lower in EC than EH | ||
| Akt | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; rising in EC | |
| 2006 | McCampbell AS et al | 14 CAH | High; no difference between EC and CAH | ||
| mTor b | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | High | Not prognostic |
| COX-2 | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low | Prognostic ( P < .0001) |
| 2007 | Erkanli S et al | 30 SH | Rising from PE to EH and EC | ||
| 2007 | Nasir A et al | 14 (9 EH and 5 CAH) | Rising from PE to EH to EC | ||
| 2005 | Orejuela FJ et al | 19 | Rising from NE to EH; but lower in EC | ||
| 2002 | Cao QJ et al | 6 nonatypical hyperplasia | Low in atrophic endometrium, PE, EH, and low grade EC; rising in high-grade EC | ||
| Histone 3 b | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; rising in EC | Not prognostic |
| CK 5/6 b | 2010 | Current study | 152 (134 non-EIN, 8 EIN) | Low; falling in EC | Prognostic ( P = .001) |
a Articles before 2000 were also found;
We therefore have evaluated the prognostic value of 16 promising immunohistochemical molecular biomarkers on a relatively large series of consecutive EHs with long follow-up times.
Materials and Methods
Approval by the Regional Ethics Committee was obtained before the initiation of this study. We started the study by selecting all 931 consecutive cases that had been diagnosed routinely as EH between January 1, 1980, and December 31, 2004, from the archives of the Department of Pathology, Stavanger University Hospital, Stavanger, Norway. Curettage samples were fixed in 4% buffered formaldehyde, were paraffin-embedded (4-μm thick histologic sections), and were stained with hematoxylin-eosin. The 307 cases that were selected for our analysis had (1) at least 1 endometrial sample that had been obtained as part of the follow-up evaluation after the initial diagnosis of EH and (2) a follow-up period (ie, the interval between the diagnostic index and follow-up biopsy) of >12 months. The 307 cases were rereviewed by 2 gynecologic pathologists (E.G., O.G.A.); in 156 cases, they agreed on the diagnosis of EH and classified them according to the 2003 WHO and EIN systems (none of the 151 cases on which they disagreed progressed to cancer). Special attention was paid to exclude potential mimickers of EH. In 4 of the 156 EH agreement cases, paraffin blocks were lacking. Therefore, 152 cases were available for further analysis. Therapeutic interventions that occurred between the index hyperplasia diagnosis and end of follow-up interval were not standardized formally but were representative of the gynecology practice at the time of diagnosis. In general, patients underwent rebiopsy based on symptomatic indications.
Immunohistochemical staining
Antigen retrieval methods and antibody dilutions were optimized before the onset of the study. To ensure uniform handling of the samples, all sections were freshly cut and processed simultaneously. Sections were mounted on silanized slides (#S3002; Dako, Glostrup, Denmark) and dried overnight at 37°C followed by 1 hour at 60°C, deparaffinized in xylene, and rehydrated in decreasing concentrations of alcohol. Antigen was retrieved with a highly stabilized retrieval system (ImmunoPrep; Dako Instrumec AS, Oslo, Norway). The following antigen retrieval protocols and buffers were used: 10 mmol/L TRIS/1 mmol/L ethylenediaminetetraacetic acid (pH 9.0) sections were heated for 3 minutes at 110°C, followed by 10 minutes at 95°C, and then cooled to 20°C. Target retrieval buffer (S1699; Dako) was used for survivin, for which the slides were heated for 20 minutes at 100°C followed by cooling to 20°C. Dako HerCepTest was performed according to the manufacturer’s procedures. Endogenous peroxidase activity was blocked with a peroxidase-blocking reagent (S2001; Dako) for 10 minutes. The immune complex was visualized with the Dako REAL EnVision Detection System, Peroxidase/DAB, Rabbit/Mouse (K5007; Dako) incubated with EnVision/HRP, Rabbit/Mouse, for 30 minutes and DAB+ chromogen for 10 minutes. The sections were counterstained with hematoxylin, dehydrated, and mounted. With the exception of the overnight incubations with some of the primary antibodies, all steps were performed with Dako Autostainer and tris-buffered saline (S1968; Dako) with 0.05% Tween 20 as the wash buffer. Table 2 provides information about the antibodies that were used.
| Antibody | Dilution incubation | Antigen retrieval | Clone | Producer |
|---|---|---|---|---|
| CK-5/6 | 1/100 | Tris/EDTA | D5/16 B4 | DAKO Corp, Carpinteria, CA |
| P53 | 1/200 | Tris/EDTA | DO-7 | DAKO Corp |
| P63 | 1/75 | Tris/EDTA | 4A4 | DAKO Corp |
| P27 | 1/100 | Tris/EDTA | SX53G8 | DAKO Corp |
| Survivin | 1/75 | TRS IP | D8 | Santa Cruz Biotechnology Inc, Santa Cruz, CA |
| Phospho-histone-H3 | 1/12000 | Tris/EDTA | Rabbit polyclonal antibody #06-570 | Upstate Biotechnology, Inc, Waltham, MA |
| Phospho-Akt | 1/300 ON | Tris/EDTA | Rabbit monoclonal antibody #3787 | Cell Signaling Technology, Beverly, MA |
| Phospho mTOR | 1/800 ON | Tris/EDTA | Rabbit polyclonal antibody #2971 | Cell Signaling Technology |
| Bcl-2 | 1/40 | Tris/EDTA | Bcl-2/100/D5 | Novocastra Laboratories, Ltd, Newcastle, UK |
| P16 | Ready to use | Standard p. | CINtec histology | mtm Laboratories, Heidelberg, Germany |
| P21 | 1/25 | Tris/EDTA | 4D10 | Novocastra Laboratories, Ltd |
| CyclinE | 1/40 | Tris/EDTA | 13A3 | Novocastra Laboratories, Ltd |
| PTEN | 1/300 | Tris/EDTA | 6H2,1 | DAKO Corp |
| Hercep test | Ready to use | Standard p. | #SK001 | DAKO Corp |
| β-catenin | 1/300 | Tris/EDTA | 17C2 | Novocastra Laboratories, Ltd |
| COX-2 | 1/400 | Tris/EDTA | 4H12 | Novocastra Laboratories, Ltd |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
