Main

Carcinomas of the endometrium are the most common malignancy of the gynecologic tract and rank fourth in incidence among invasive tumors in women, following breast, lung, and colon cancers. Endometrial carcinomas exhibit a broad morphologic spectrum thought to be the consequence of variable cellular differentiation. The most frequent subtype of endometrial carcinoma is endometrioid adenocarcinoma (>80%), followed by serous carcinoma (10%). Endometrioid adenocarcinoma that is confined to the uterus has a very favorable outcome with a 5-year-survival rate of 87.4% reported in the Annual Report (AR) of FIGO (International Federation of Gynecology and Obstetrics).1 In contrast, serous carcinoma is infrequent compared with endometrioid adenocarcinoma, and has a much higher propensity for spreading beyond the uterus early in the disease process, resulting in a much poorer clinical outcome. Therefore, it is very important for proper clinical management to distinguish serous carcinoma from endometrioid adenocarcinoma and ultimately to understand the molecular basis for the differences in biological behavior of these tumors.

Insulin-like growth factor-II mRNA-binding protein 3 (IMP3), also known as L523S or KOC (K-homologous domain-containing protein overexpressed in cancer),2, 3 is a 580 residue protein containing four K homology domains and two other RNA recognition motifs. The IMP3 gene is located on chromosome 7p11.5. IMP3 is a member of the human IGF-II mRNA-binding protein family (IMPs) whose other members are IMP1 and IMP2. These proteins are capable of strong and specific binding with mRNA transcripts including those encoding insulin-like growth factor-II (IGF-II), cell adhesion molecules and others. IMPs have been implicated in various post-transcriptional processes, such as mRNA localization,4 turnover,5 and translational control.6

IMP3 is expressed in many cells of the developing fetus. In contrast, it is not expressed in most adult tissues except in the gonads. However, IMP3 is expressed in a number of cancers. A monoclonal antibody specific for IMP3 has been developed and used successfully with routine immunohistochemical methods to study overexpression of the protein in malignant pancreatic, renal, and uterine endocervical lesions.3, 7, 8 Expression of IMP3 in endometrial carcinomas has not been previously studied. The aims of this study were to evaluate the expression of IMP3 in endometrial carcinomas especially endometrioid adenocarcinoma and serous carcinoma subtypes as well as benign endometrium using immunohistochemical techniques, to assess the diagnostic utility of IMP3 in distinguishing endometrioid adenocarcinoma from serous carcinoma, to correlate IMP3 expression with clinicopathologic features of the disease, and to understand the possible role of IMP3 in the pathogenesis of endometrial adenocarcinoma.

Materials and methods

Case Selection

One hundred and sixty-seven cases of endometrial carcinoma with adequate archive tissue for immunohistochemical study were retrieved from the files of the Department of Pathology at UMass Memorial Medical Center accessioned between August 1999 and May 2006. These included 122 cases of endometrioid adenocarcinoma and 45 cases of serous carcinoma. All studied endometrioid adenocarcinoma specimens were from hysterectomy specimens. The specimens for serous carcinoma included 6 endometrial biopsies (EMB), 2 endometrial curettings (EMC), and 37 hysterectomies. The diagnosis of endometrioid adenocarcinoma or serous carcinoma was made in accordance with published diagnostic criteria.9 Tumor grade and stage were evaluated using the FIGO criteria. Twenty control samples of benign endometrium were obtained from 20 patients who underwent hysterectomy at UMass Memorial Medical Center in 2004 for benign uterine lesions including leiomyomas, adenomyosis, and endometrial polyps.

Immunohistochemical Study

Immunohistochemical stains were performed as described previously.3, 7 Briefly, 5-μm tissue sections were cut, deparaffinized, washed, and immersed in citrate buffer (0.01 M, pH 6.0), and underwent microwave (770 W, 14 min) antigen retrieval. The slides were cooled, rinsed with PBS (three rinses, 5 min each), and stained on a DAKO Autostainer (Dako Corporation, Carpinteria, CA, USA) at room temperature. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide, and the slides were rinsed and treated with a blocking protein (ChemMate; Ventana, Tucson, AZ, USA) to prevent nonspecific staining. The sections were then incubated with L523S, a mouse monoclonal antibody specific for IMP3 (2 μg/ml; Corixa Corporation, Seattle, WA, USA) for 45 min, or with p53 mouse monoclonal antibodies (a cocktail of two clones: Ab2 at 1:300 dilution and Ab6 at 1:1000 dilution; Calbiochem, San Diego, CA, USA) for 30 min. Following brief buffer washes, the sections were treated with a cocktail of biotinylated anti-rabbit IgG and anti-mouse IgG/IgM (ChemMate) for 30 min, washed, and incubated with avidin–biotin–peroxidase complex (ChemMate) for 30 min. The sections were rinsed, developed with diaminobenzidine (ChemMate) and hydrogen peroxide (10 min), rinsed with tap water, and counterstained with hematoxylin.

Representative sections of pancreatic carcinoma were used as positive controls for IMP3 expression, and sections of high-grade ovarian serous carcinoma were used as positive control for p53. Negative controls were performed by replacing the primary antibody with nonimmune IgG. The percentage of endometrioid adenocarcinoma, serous carcinoma, or normal endometrium that was positive for IMP3 or p53 was assessed. The intensity of the staining was evaluated as weak, intermediate, or strong.

Statistical Study

Independent sample t-tests was used to compare mean values of continuous variables in different groups. P<0.05 were considered to indicate significance.

Results

Clinicopathologic Features

One hundred and twenty-two patients with endometrioid adenocarcinoma ranged in age from 34 to 89 years. The median age was 61 years. Thirty-four (28%) cases had tumor cell nuclear grade 1, 76 (62%) cases grade 2, and 12 (10%) cases grade 3. Tumor architecture grade 1 was observed in 61 (50%) cases, grade 2 in 53 (43%) cases, and grade 3 in 8 (7%) cases. FIGO stages of these 122 endometrioid adenocarcinomas were IA (n=43), IB (n=45), IIA (n=5), IIB (n=2), IIIA (n=4), and IIIC (n=6). Forty-five patients with serous carcinoma ranged in age from 35 to 85 years old with a median age of 68 years. Among these 37 cases of serous carcinoma with clinicopathologic stage assessed, 9 cases were diagnosed with stage IA disease, 6 stage IB, 4 stage IC, 1 stage IIA, 6 stage IIB, 4 stage IIIA, 6 stage IIIC, and 1 stage IVB disease. The age of the 20 control patients with hysterectomies performed for benign uterine lesions ranged from 34 to 58 years. Of these 20 cases, 6 had proliferative, 4 weakly proliferative, 8 secretory, and 2 inactive endometria. All 20 cases had leiomyomas. Additionally, 11 cases also had adenomyosis, and 3 had endometrial polyps.

Immunohistochemical Findings

IMP3 expression

All of the serous carcinoma cases showed cytoplasmic staining for IMP3 (Figure 1). As indicated in Table 1, 39 cases (86%) showed positive staining for IMP3 in more than 50% of tumor cells, and 3 cases (7%) had staining in 21–50 or 6–20% of the tumor cells. Immunohistochemical intensity for IMP3 was found to be strong in 38 (84%) and intermediate in 7 (16%) serous carcinoma cases.

Figure 1
figure 1

(a) Serous carcinoma of the endometrium (H&E); (b) cytoplasmic expression of IMP3 in serous carcinoma of the endometrium; (c) International Federation of Gynecology and Obstetrics (FIGO) grade 1 endometrioid adenocarcinoma of the endometrium (H&E); (d) negative immunohistochemical stain for IMP3 in FIGO grade 1 endometrioid adenocarcinoma of the endometrium; (e) FIGO grade 3 endometrioid adenocarcinoma of the endometrium (H&E); (f) cytoplasmic expression of IMP3 in FIGO grade 3 endometrioid adenocarcinoma of the endometrium; (g) benign proliferative endometrium (H&E); (h) negative immunohistochemical stain for IMP3 in benign proliferative endometrium.

Full size image
Table 1 Expression of IMP3
Full size table

As shown in Table 1, 54 of 122 (44%) endometrioid adenocarcinoma cases were negative for IMP3, and 68 (56%) cases showed immunohistochemical expression of the protein. Among these 68 cases, 30 cases (25%) demonstrated positive staining in 1–5% tumor cells, 20 (16%) in 6–20%, 10 (8%) in 21–50%, and 8 (7%) in >50% of the tumor cells. Strong IMP3-staining intensity was noted in 34 cases (28%), intermediate in 26 cases (21%), and weak in 8 cases (7%).

The nuclear grade, architectural grade, FIGO stage, and patient's age at the diagnosis of endometrioid adenocarcinomas were compared between endometrioid adenocarcinomas with and without IMP3 expression. Endometrioid adenocarcinoma with IMP3 expression was defined as tumor with >5% of the neoplastic cells showing positive staining for IMP3. Endometrioid adenocarcinomas with IMP3 expression had significantly higher nuclear and architectural grades than those without IMP3 expression (P=0.0000 and P=0.0002, respectively). There was no FIGO stage or patient's age differences between these two groups (P=0.5590 and P=0.8045, respectively).

All 20 samples of morphologically benign endometrium evaluated were negative for IMP3.

p53 expression

Positive immunohistochemical reaction for p53 in tumor cells was characterized by positive staining seen in the nucleus. As shown in Table 2, 35 out of 45 (78%) serous carcinoma cases demonstrated strong p53 staining in more than 50% of the tumor cell nuclei. Thirty-two of 112 (29%) endometrioid adenocarcinoma cases were negative for p53, and 80 (71%) cases showed immunohistochemical expression of the protein. Among these 80 cases, 40 cases (36%) demonstrated positive staining in 1–5% tumor cells, 23 (20%) in 6–20%, 6 (5%) in 21–50%, and 11 (10%) in >50% of the tumor cells. Strong p53-staining intensity was identified in 75 cases (67%), intermediate in 4 cases (3%), and weak in 1 case (1%).

Table 2 Expression of p53
Full size table

Discussion

Endometrial serous carcinoma is the most common aggressive form of endometrial malignancy. It has a strong predilection for extrauterine spread and poor clinical prognosis. The mechanism of its aggressive clinical behavior is unknown, although a recent focus has been done on a potential role for cellular adhesion molecules in this process.10 In this study, we have demonstrated that IMP3 is highly expressed in endometrial serous carcinoma as compared to endometrial endometrioid adenocarcinoma (P<0.0001), and expression of IMP3 in the latter form, although low in frequency, correlates with higher tumor nuclear and architectural grades, suggesting that IMP3 may be of diagnostic value in differential diagnosis, a marker of tumor aggressive clinical behavior, and important in the pathogenesis of endometrial malignancy.

These results are of particular interest because IMP3 has been found to be a biomarker associated with the progression of other unrelated cancers. Thus its expression progressively increases with advancing stage in pancreatic cancers.3 Moreover, in renal cell carcinoma, IMP3 expression identifies the majority of cases that go on to metastasize and ultimately kill patients.7 Together these findings raise the possibility that IMP3 promotes aggressive behavior in tumors and might explain why most of the more highly aggressive serous carcinomas express this molecule as well as why it is present more frequently in high- vs low-grade endometrial carcinomas.

IMPs are primarily expressed during early embryogenesis and at mid-gestation in the mouse,11 but not in most adult tissues.12, 13 IMPs are considered to be oncofetal proteins that are frequently overexpressed in various cancers. IMPs play important roles in the binding, trafficking, and stabilization of the fetal subtype of IGF-II mRNA during embryogenesis.12, 14 In a recent study to characterize the cellular functions of IMPs utilizing the RNA interference technique in cancer cells (HeLa), Vikesaa et al15 showed that IMPs are necessary for cell adhesion, cytoplasmic spreading, and invadopodia formation; implying that IMPs are involved in cellular adhesion and invasion during normal development and malignancy formation. Previous study demonstrated that antisense suppression of the IMP3 orthologue in xenopus embryos inhibited the migration of neural crest cells.16 Moreover, another study showed that transgenic overexpression of IMP3 in the pancreas resulted in increased proliferation and metaplasia of acinar cells.17 These findings might explain why IMP3 is associated with more aggressive cancers.

Further analysis of HeLa cells found that the loss of IMPs was associated with a coordinate downregulation of mRNAs encoding extracellular matrix and adhesion proteins, particularly, a 5.0 kb CD44 mRNA.15 CD44 protein is a member of Ig-superfamily cell adhesion molecules. It participates in intercellular adhesion, and promotes collagen IV degradation and tumor cell invasion by anchoring matrix metalloproteinases.18 These results suggest that CD44 mRNA regulation may be involved in IMP-mediated tumor cell invasion and metastasis.

Several studies have shown that endometrial serous carcinoma has significantly less CD44 expression than endometrial endometrioid adenocarcinoma.19, 20 The loss of CD44 expression has been correlated with a more aggressive course of endometrial carcinomas.21, 22 CD44 expression is subject to both alternative splicing and post-translational modifications,23 so the selective regulation of the CD44 transcripts by IMPs including IMP3 provides an additional level of complexity to the control of CD44 expression. It has been hypothesized that CD44 transcripts may be divided into an IMP-upregulating and an IMP-downregulating population.15 Our data imply that IMP-downregulating CD44 transcripts may play an important role in the pathogenesis of uterine serous carcinoma. Downregulation of the adhesion protein CD44 and loss of invadopodia may collectively explain the reduced cell–cell adhesiveness in serous carcinoma and early dissemination of the carcinoma cells into the peritoneal cavity through the fallopian tubes.

Mutation of the p53 tumor suppressor gene accompanied by overexpression of the mutant p53 protein is one of the major genetic alterations in endometrial serous carcinoma.24 Therefore, p53 has been used as a biomarker for endometrial serous carcinoma.25 To correlate IMP3 and p53 expressions in endometrial carcinomas, immunohistochemical expression of p53 was evaluated in all 45 cases of serous carcinoma and 112 cases of endometrioid adenocarcinoma. Our data showed in Table 2 demonstrate that IMP3 may be another useful biomarker for endometrial serous carcinoma in difficult cases.

In summary, IMP3 is highly expressed in endometrial serous carcinoma as compared to endometrial endometrioid adenocarcinoma (P<0.0001), and when expressed in endometrioid carcinomas it is more frequently found in tumors of higher grade. Therefore, in endometrial adenocarcinomas, as is in some other cancers, IMP3 expression is found in more aggressive disease.