Sohei Yamamoto

Loss of ARID1A protein expression occurs as an early event in ovarian clear-cell carcinoma development and frequently coexists with PIK3CA mutations

ARID1A is a recently identified tumor suppressor gene that is mutated in ∼50% of ovarian clear-cell carcinomas. This mutation is associated with loss of ARID1A protein expression as assessed by immunohistochemistry. The present study aimed at determining the timing of the loss of ARID1A protein expression during the development of ovarian clear-cell carcinoma and assessing its relevance in correlation to PIK3CA gene mutations. A total of 42 clear-cell carcinoma cases with adjacent putative precursor lesions (endometriosis-associated carcinoma cases (n=28) and (clear-cell) adenofibroma-associated carcinoma cases (n=14)) were selected and subjected to immunohistochemical analysis for ARID1A protein expression and direct genomic DNA sequencing of exons 9 and 20 of the PIK3CA gene. ARID1A immunoreactivity was deficient in 17 (61%) of the 28 endometriosis-associated carcinomas and 6 (43%) of the 14 adenofibroma-associated carcinomas. Among the precursor lesions adjacent to the 23 ARID1A-deficient carcinomas, 86% of the non-atypical endometriosis (12 of 14) and 100% of the atypical endometriosis (14 of 14), benign (3 of 3), and borderline (6 of 6) clear-cell adenofibroma components were found to be ARID1A deficient. In contrast, in the 19 patients with ARID1A-intact carcinomas, all of the adjacent precursor lesions retained ARID1A expression regardless of their types and cytological atypia. Analysis of 22 solitary endometrioses and 10 endometrioses distant from ARID1A-deficient carcinomas showed that all of these lesions were diffusely immunoreactive for ARID1A. Among the 42 clear-cell carcinomas, somatic mutations of PIK3CA were detected in 17 (40%) tumors and majority (71%) of these were ARID1A-deficient carcinomas. These results suggest that loss of ARID1A protein expression occurs as a very early event in ovarian clear-cell carcinoma development, similar to the pattern of PIK3CA mutation recently reported by our group, and frequently coexists (not mutually exclusive) with PIK3CA mutations.

PIK3CA mutation is an early event in the development of endometriosis‐associated ovarian clear cell adenocarcinoma

Clear cell adenocarcinoma (CCA), a highly lethal histological subtype of ovarian carcinoma, is a type of human cancer with a high frequency of activating mutations in the PIK3CA gene. In this study, we aimed to determine how these mutations contribute to tumour development of CCAs. Exons 9 and 20 of the PIK3CA gene were analysed by direct genomic DNA sequencing of 23 CCAs with synchronous putative precursor lesions (ie endometriosis adjacent to carcinoma, with or without cytological atypia) and their mutational statuses were compared. Somatic mutations of the PIK3CA gene were detected in 10/23 (43%) carcinomas and in all cases the type of mutation was H1047R in the kinase domain. The identical H1047R mutation was also detected in the coexisting endometriotic epithelium, adjacent to the CCAs, in nine of ten (90%) cases. Moreover, in six of the nine lesions, the H1047R mutation was identified even in the endometrioses lacking cytological atypia. These findings provide evidence that mutations of the PIK3CA gene occur in the putative precursor lesions of CCA, strongly suggesting that they are very early events in tumourigenesis, probably initiating the malignant transformation of endometriosis. A specific kinase inhibitor to mutated PIK3CA may potentially be an effective therapeutic reagent against these carcinomas. Copyright

Actinin-4 gene amplification in ovarian cancer: a candidate oncogene associated with poor patient prognosis and tumor chemoresistance

Actinin-4, an isoform of non-muscular α-actinin, enhances cell motility by bundling the actin cytoskeleton. We previously reported a prognostic implication of high immunohistochemical expression of actinin-4 protein in ovarian cancers. Chromosomal gain or amplification of the 19q12–q13 region has been reported in ovarian cancer. We hypothesized that the actinin-4 (ACTN4) gene might be a target of the 19q12–q13 amplicon and play an essential role of ovarian cancer progression. In total, 136 advanced-stage ovarian cancers were investigated for the copy number of the ACTN4 gene on chromosome 19q13, using fluorescence in situ hybridization, and the correlation of the ACTN4 copy number with actinin-4 protein immunoreactivity and major clinicopathological factors was investigated. A higher copy number (≧4 copies) of the ACTN4 gene was detected in 29 (21%) cases and was highly associated with the intensity of actinin-4 immunoreactivity (P<0.0001), a high histological tumor grade (P=0.030), a clear-cell adenocarcinoma histology (P=0.012), resistance to first-line chemotherapies (P=0.028), and poor patient outcome (P=0.0011). Univariate analyses using the Cox regression model showed that a higher ACTN4 gene copy number was able to predict patient outcome more accurately than high actinin-4 immunoreactivity (relative risk: 2.48 vs 1.55). Multivariate analysis showed that a higher copy number of the ACTN4 gene and the degree of residual disease were independent prognostic factors for overall patient survival. The actinin-4 gene may be a target of the 19q amplicon, acting as a candidate oncogene, and serve as a predictor of poor outcome and tumor chemoresistance in patients with advanced-stage ovarian cancers.

Gene amplification and protein overexpression of MET are common events in ovarian clear-cell adenocarcinoma: their roles in tumor progression and prognostication of the patient

The aim of this study was to assess protein overexpression and gene copy number alterations of MET in ovarian clear-cell adenocarcinoma, and to assess its potential as a novel therapeutic target. Ninety cases of clear-cell adenocarcinoma were analyzed for MET protein overexpression and copy number alterations of the MET gene by immunohistochemistry and brightfield double in situ hybridization, respectively. In addition, 101 cases of the non-clear-cell type ovarian carcinomas at advanced stages were also evaluated for comparison. MET overexpression was assigned when complete membrane staining with moderate or strong intensity was observed in at least 10% of the tumor cells examined. Double in situ hybridization was determined as positive when the tumor exhibited high-level polysomy (≥4 copies in ≥40% of tumor cells) or MET gene amplification. MET overexpression was detected in 20 of 90 clear-cell adenocarcinomas (22%) and none of 111 non-clear-cell type ovarian carcinomas. Double in situ hybridization was positive in 21 of 89 informative clear-cell adenocarcinomas (24%) and only 3 non-clear-cell type ovarian carcinomas (3%). In the whole population, true amplification of the MET gene was detected only in the clear-cell adenocarcinoma histology (five cases, 6%). In clear-cell adenocarcinomas, double in situ hybridization positivity was highly correlated with the presence of MET overexpression and a poorly differentiated histology of tumors (P=0.0105 and 0.00038, respectively). For the patients with clear-cell adenocarcinomas, MET overexpression, as well as advanced clinical stage and the poorly differentiated histology of tumors, was identified as an independent unfavorable prognostic factor for overall survival. In conclusion, among ovarian carcinomas, the amplification of the MET proto-oncogene is highly selective and commonly occurs in clear-cell adenocarcinoma. MET could serve as a biomarker for the prognostication of patients with clear-cell adenocarcinoma and tumor progression, and has potential as a novel therapeutic target for this carcinoma.

PIK3CA mutations and loss of ARID1A protein expression are early events in the development of cystic ovarian clear cell adenocarcinoma

Somatic mutations of PIK3CA and ARID1A are the most common genetic alterations observed in ovarian clear cell adenocarcinomas (CCA). In a previous report, we showed that PIK3CA gene mutations and loss of ARID1A expression occur early during the development of CCA. In the present study, using direct genomic DNA sequencing for exons 9 and 20 of PIK3CA and immunohistochemistry for ARID1A protein expression, we analyzed the association of these molecular alterations with various clinicopathological parameters in a total of 90 cases of primary ovarian CCA, including 42 previously examined cases. The presence of PIK3CA mutations, identified in 34 (39%) of the 88 informative cases, was significantly associated with a grossly cystic tumor, the presence of adjacent endometriosis, prominent papillary architecture of tumor growth, the presence of hyalinized and mucoid stroma, and the absence of clear cell adenofibroma components (P < 0.05, each). There was no significant association of PIK3CA mutations with other clinical variables, such as age, clinical stage, or clinical outcome of the patients. The intensity of immunoreactivity for ARID1A was assigned as negative, weakly positive, and strongly positive in 44%, 22%, and 33% of tumors, respectively. Compared to tumors immunoreactive for ARID1A, ARID1A-negative tumors were significantly associated with the presence of adjacent endometriosis (P = 0.025), but there was no statistically supported association with other examined clinicopathological parameters. Compared with CCAs strongly positive for ARID1A, CCAs negative for ARID1A more frequently harbor PIK3CA mutations (P = 0.013). PIK3CA gene mutations and ARID1A immunohistochemistry lacked prognostic significance. These data further support the idea that these molecular alterations occur as very early events during tumor development of ovarian CCA.

Clear cell adenocarcinoma associated with clear cell adenofibromatous components: a subgroup of ovarian clear cell adenocarcinoma with distinct clinicopathologic characteristics.

We occasionally encountered clear cell adenofibromatous (CCAF) components coexisting in the ovarian clear cell adenocarcinoma (CCA). To reveal the clinicopathologic significance of CCAF components in CCA, we classified 67 cases of surgically resected CCA into CCA with and without CCAF components [CCAF(+) and (−) groups], and compared clinicopathologic parameters, that is, patient age, clinical stage, the degree of optimal cytoreduction, patient outcome, histologic grade and Ki-67 labeling index of the CCA, and the presence of endometriosis, between these 2 groups. Fourteen cases (21%) and 53 cases were classified as CCAF(+) and CCAF(−) groups, respectively. Of these 14 CCAF(+) cases, the CCAF components with atypia were observed adjacent to the CCAF components without atypia in 10, and adjacent to the obvious CCAs in 13 cases. In comparison with the CCAF(−) group, the CCAF(+) group showed a higher frequency of histologically low-grade tumors [93% (13 of 14) vs. 43% (23 of 53), P=0.0027], a lower Ki-67 labeling index (mean 35.9% vs. 44.0%, P=0.0492), and better patient prognosis (5-year survival 78.8% vs. 49.3%, P=0.0277). Endometriosis was much less frequent in the CCAF(+) group than in the CCAF(−) group [14.7% (2 of 14) vs. 67.9% (36 of 53), P=0.00096]. Multivariate analysis identified only optimal cytoreduction as independent favorable prognostic factor. These results suggest that CCAF besides endometriosis is associated with the development of CCA, and that the CCAF(+) group may be a distinct subgroup of CCA with less aggressive biologic behavior.

Clear-cell adenofibroma can be a clonal precursor for clear-cell adenocarcinoma of the ovary: a possible alternative ovarian clear-cell carcinogenic pathway.

Several studies have reported that ovarian clear‐cell adenocarcinoma can be derived from endometriosis. Although the clear‐cell adenofibroma (CCAF), a major form of benign and borderline ovarian clear‐cell tumour, has been suggested as another precursor for clear‐cell adenocarcinoma (CCA), there is no supportive genetic evidence for this presumption. To examine the genetic linkage between CCAF and CCA of the ovary, we conducted allelotype analysis for both CCAF and adjacent CCA components in 14 cases of CCA associated with benign CCAF and/or borderline CCAF. DNA isolated from laser‐microdissected tissue was subjected to polymerase chain reaction and analysis for loss of heterozygosity (LOH), using 17 polymorphic markers located on 11 chromosomal arms: 1p, 5q, 8p, 9p, 9q, 10q, 11q, 13q, 18q, 19p and 22q. For all informative loci, the frequency of LOH in adenocarcinoma was 49% (54/110 loci), and was significantly higher than those in the components of benign CCAF (22%, 20/92 loci) and borderline CCAF (30%, 25/83 loci) (χ2 test; p < 0.05, respectively). The concordance rate in allelic patterns at all informative loci was 74% between benign CCAF and adenocarcinoma components, 81% between borderline CCAF and adenocarcinoma components, and 95% between benign CCAF and borderline CCAF components. Furthermore, between CCAF and adenocarcinoma components, an identical LOH pattern, involving the same alleles, was found in 13 (93%) of 14 cases at one or more chromosomal loci, and estimation of probability indicated that these events were very unlikely to have occurred by chance. Among the markers examined, LOHs on 5q, 10q and 22q were frequent in both CCAF and adenocarcinoma components, whereas LOHs on 1p and 13q were rare in CCAF components but frequent in adenocarcinoma components. These findings suggest that CCAF can be a clonal precursor for ovarian clear‐cell adenocarcinoma. Copyright

Expression of platelet-derived growth factors and their receptors in ovarian clear-cell carcinoma and its putative precursors

Recent studies have shown that platelet-derived growth factors and their receptors are frequently co-expressed in ovarian cancers. Herein, we investigated the role of the platelet-derived growth factor pathway in the development of ovarian clear-cell adenocarcinoma, a highly chemoresistant form of ovarian cancer. Immunohistochemical expression of platelet-derived growth factor receptor-α and receptor-β, platelet-derived growth factor A-chain and B-chain was examined in 31 cases of clear-cell adenocarcinoma and 56 coexisting putative precursor lesions: 17 non-atypical and 19 atypical endometrioses, and 10 non-atypical and 10 atypical clear-cell adenofibroma components. Twenty-one solitary endometrioses were also examined. Vascular endothelial cells were always positive for all the markers examined, and were used as positive controls. The frequencies of positivity for platelet-derived growth factor receptor-α and receptor-β, and platelet-derived growth factor A-chain increased in accordance with higher cytologic atypia in the putative precursors: 71, 47, and 59% in the 17 non-atypical endometrioses, 84, 73, and 84% in the 19 atypical endometrioses, 0% each in the 10 non-atypical clear-cell adenofibromas, 100, 90, and 90% in the 10 atypical clear-cell adenofibromas, and 97, 97, and 100% in the 31 clear-cell adenocarcinomas, respectively. Positivity for platelet-derived growth factor B-chain increased in accordance with increased atypia in clear-cell adenofibroma: 0% in non-atypical clear-cell adenofibromas, 30% in atypical clear-cell adenofibromas, and 60% in coexisting carcinomas. However, in contrast, positivity for platelet-derived growth factor B-chain decreased in accordance with increased atypia in endometriosis coexisting with clear-cell adenocarcinomas: 35% in non-atypical endometrioses, 11% in atypical endometrioses, and 5% in coexisting carcinomas. Platelet-derived growth factor receptor-α and receptor-β, and their ligands A-chain and B-chain were positive in 14, 29, 19, and 62% of the solitary endometrioses, respectively. These results indicate activation of the platelet-derived growth factor pathway in ovarian clear-cell adenocarcinomas and suggest biological differences between carcinomas that arise in association with clear-cell adenofibroma vs endometriosis.

HER2 testing on core needle biopsy specimens from primary breast cancers: interobserver reproducibility and concordance with surgically resected specimens

BackgroundAccurate evaluation of human epidermal growth factor receptor type-2 (HER2) status based on core needle biopsy (CNB) specimens is mandatory for identification of patients with primary breast cancer who will benefit from primary systemic therapy with trastuzumab. The aim of the present study was to validate the application of HER2 testing with CNB specimens from primary breast cancers in terms of interobserver reproducibility and comparison with surgically resected specimens.MethodsA total of 100 pairs of archival formalin-fixed paraffin-embedded CNB and surgically resected specimens of invasive breast carcinomas were cut into sections. All 100 paired sections were subjected to HER2 testing by immunohistochemistry (IHC) and 27 paired sections were subjected to that by fluorescence in situ hybridization (FISH), the results being evaluated by three and two observers, respectively. Interobserver agreement levels in terms of judgment and the concordance of consensus scores between CNB samples and the corresponding surgically resected specimens were estimated as the percentage agreement and κ statistic.ResultsIn CNB specimens, the percentage interobserver agreement of HER2 scoring by IHC was 76% (κ = 0.71) for 3 × 3 categories (0-1+ versus 2+ versus 3+) and 90% (κ = 0.80) for 2 × 2 categories (0-2+ versus 3+). These levels were close to the corresponding ones for the surgically resected specimens: 80% (κ = 0.77) for 3 × 3 categories and 92% (κ = 0.88) for 2 × 2 categories. Concordance of consensus for HER2 scores determined by IHC between CNB and the corresponding surgical specimens was 87% (κ = 0.77) for 3 × 3 categories, and 94% (κ = 0.83) for 2 × 2 categories. Among the 13 tumors showing discordance in the mean IHC scores between the CNB and surgical specimens, the results of consensus for FISH results were concordant in 11. The rate of successful FISH analysis and the FISH positivity rate in cases with a HER2 IHC score of 2+ differed among specimens processed at different institutions.ConclusionIt is mandatory to study HER2 on breast cancers, and either CNB or surgical specimen can be used.

Clinicopathological significance of WT1 expression in ovarian cancer: a possible accelerator of tumor progression in serous adenocarcinoma

Recently, oncogenic potential of the WT1 gene has been proposed in some human solid tumors and leukemias. Although previous studies have shown the frequent expression of the WT1 protein in ovarian serous adenocarcinomas (OSAs), its clinicopathologic significance is still unclear. We immunohistochemically examined the expression status of WT1 in 119 OSAs and analyzed the correlation of the intensity of WT1 immunoreactivity with the level of WT1 mRNA expression by quantitative real-time polymerase chain reaction, clinicopathologic variables, expression of p53, Bcl-2, and Ki-67 labeling index (LI). Of 119 OSAs, nuclear WT1 immunoreactivity was positive in 99 (83%), of which 44 (44%) and 55 (56%) exhibited high and low WT1 immunoreactivities, respectively. The quantitative WT1 mRNA levels were significantly correlated with the intensity of WT1 immunoreactivity (P < 0.05). In comparison with WT1-negative OSAs, the WT1-positive OSAs showed a higher grade (P = 0.007), advanced stage (P = 0.018), and higher Ki-67 LI (P < 0.001). Additionally, high WT1 immunoreactivity was correlated with a higher grade (P = 0.003), Ki-67 LI (P = 0.012), Bcl-2 expression (P = 0.003), and poorer patient outcome (5-year survival, 36.5 vs 63.8%, P = 0.008 by log-rank test). The WT1 protein may be an accelerator of the progression of OSA.