ORIGINAL ARTICLE
Molecular Characterization of Undifferentiated Carcinoma Associated With Endometrioid Carcinoma Elisabetta Kuhn, MD,* Ayse Ayhan, MD, PhD,w Asli Bahadirli-Talbott, MSc,* Chengquan Zhao, MD,z and Ie-Ming Shih, MD, PhD*y8
Abstract: Uterine and ovarian undifferentiated carcinomas (UCs) are often associated with low-grade endometrioid carcinomas (EMCs) and are characterized by a solid growth pattern and a lack of appreciable features of differentiation. As compared with pure EMC, UC is highly malignant, and the molecular pathogenesis that leads to disease aggressiveness remains largely unknown. This study interrogates the molecular pathogenesis of UCs by comparing the molecular alterations between the UC and the EMC components. A total of 20 UCs were studied, 12 of which contained both UC and EMC components. Mutation analysis was performed for the genes commonly mutated in EMC, and immunohistochemistry was used to determine the expression pattern of b-catenin and PTEN. Sequencing analysis revealed that UCs harbored somatic mutations in PIK3CA (50%), CTNNB1 (30%), TP53 (30%), FBXW7 (20%), and PPP2R1A (20%). All somatic mutations detected in EMCs were also present in concurrent UCs. Moreover, additional somatic mutations were detected in the UC component in 5 (42%) cases with concurrent EMC and UC. Concordance of immunostaining pattern for b-catenin and PTEN was recorded in all 12 matched EMCs and UCs, except 4 cases in which nuclear accumulation of b-catenin staining was detected in one of the components but not in the other. Our findings support a clonal relationship between EMCs and their associated UCs. Additional molecular genetics alteration, including mutations of CTNNB1, PPP2R1A, and TP53, may contribute to tumor progression from EMC to UC. Key Words: undifferentiated carcinoma, endometrioid carcinoma, b-catenin, progression (Am J Surg Pathol 2014;38:660–665)
From the Departments of *Pathology; yGynecology and Obstetrics; 8Oncology, Johns Hopkins Medical Institutions, Baltimore, MD; zDepartment of Pathology, Magee-Womens Hospital, University of Pittsburgh Medical Center, Pittsburg, PA; and wDepartment of Pathology, Seirei Mikatahara Hospital, Hamamatsu, Japan. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Ie-Ming Shih, MD, PhD, Department of Pathology, Johns Hopkins University, 1550 Orleans Street, CRBII, room 305, Baltimore, MD 21231 (e-mail: [email protected]) and Elisabetta Kuhn, MD, Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD (e-mail: [email protected]). Copyright r 2014 by Lippincott Williams & Wilkins
660 | www.ajsp.com
U
ndifferentiated carcinoma (UC) represents a unique and highly aggressive form of uterine cancer. The morphologic features of UC were not well defined until Silva and colleagues proposed the defining characters including medium-sized or large-sized monotonous cells, a solid and diffuse growth pattern, and a lack of appreciable features of differentiation—that is, glandular, papillary, squamous, and neuroendocrine.1–3 On the basis of these morphologic diagnostic criteria, the investigators reported an incidence of 9% among the endometrial carcinomas in their institution.2 Therefore, UC is probably underrecognized as it is previously reported to represent only 1% to 2% of all endometrial carcinomas.1,2,4–7 Clinically, the majority of UCs presents at advanced stages displays an aggressive clinical behavior and does not respond to the conventional chemotherapy for endometrial carcinoma. These characteristics highlight the importance of a correct diagnosis of this endometrial carcinoma variant. Very interestingly, UC is often associated with endometrioid carcinoma (EMC) and is thought to arise from it.1–3,8 However, the molecular genetic evidence for its genesis is still lacking, and the pathogenesis that leads to UC remains largely unknown. In fact, apart from frequent microsatellite instability, little is known about the molecular genetic alterations in UC.3,9,10 In particular, this special type of uterine cancer was not included in The Cancer Genome Atlas (TCGA), which has detailed the molecular landscape of uterine carcinomas.11 Thus, in this study, we performed mutational and immuno histochemical analyses in UCs, many of which are associated with well-differentiated EMC, allowing direct comparison of molecular alterations between UC and EMC components from the same cases.
MATERIALS AND METHODS Case Selection We selected a total of 20 UCs from the Johns Hopkins Hospital (Baltimore, MD), Seirei Mikatahara Hospital (Hamamatsu, Japan), and Magee-Womens Hospital (Pittsburg, PA). Among them, 18 involved the uterus, and 2 involved the ovary. There were 12 cases containing both UC and EMC components. All the cases were rereviewed by 2 pathologists (E.K. and I.-M.S.) to confirm the diagnosis according to the morphologic criteria previously Am J Surg Pathol
Volume 38, Number 5, May 2014
Am J Surg Pathol
Volume 38, Number 5, May 2014
reported and recently described in the new World Health Organization classification.3,12 This study was approved by the Institutional Review Board of the Johns Hopkins Medical Institutions.
Mutation Analysis Genomic DNA was prepared from serial 5-mm-thick paraffin sections. Both UC and EMC tumor components and adjacent normal tissue were collected separately by manual microdissection as previously described.13 Genomic DNA was extracted for sequencing using a QIAamp DNA Micro Kit (Qiagen, Valencia, CA) according to the vendor’s protocol. Mutation analysis was performed in all cases for the regions that frequently harbor mutations in the following genes: CTNNB1 (b-catenin-encoding gene, exon 3), FBXW7 (exons 9 and 10), KRAS (exon 2), PIK3CA (exons 1, 9, and 20), PPP2R1A (exons 5 and 6), and TP53 (exons 4 to 9). The above genes were analyzed because they are commonly mutated in endometrial carcinomas.11 Each exon was polymerase chain reaction amplified, using primers previously reported.14–16 All polymerase chain reaction products were sequenced at Agencourt Biosciences (Beverly, MA), and all sequence variants were confirmed by at least 3 independent rounds of experiments.
Immunohistochemistry Immunohistochemical analysis was performed using antibodies against b-catenin (cat # 610154; BD Transduction Laboratories, Mississauga, ON; dilution 1:1000) and PTEN (cat # M3627; Dako, Carpinteria, CA; dilution 1:100). Formalin-fixed paraffin-embedded sections were immunostained manually or by the autostainer Bond Max Automated Immunohistochemistry Vision Biosystem (Leica Microsystems GmbH, Wetzlar, Germany) using a previously described protocol.15–17 Subcellular localization (cytoplasmic, membranous, or nuclear) was recorded. For b-catenin, a positive reaction was defined as nuclear immunoreactivity in at least 5% of the tumor cells. For PTEN, a preserved reaction was defined as positivity in at least 20% of the tumor cells. In addition, all observations were made for both UC and EMC components, when present.
Statistical Analysis Comparisons of immunohistochemistry and mutation outcomes were performed using the 2-tailed unpaired Fisher exact test. P-values of 0.05 or less were considered statistically significant. Statistical analysis was carried out using the GraphPad Prism software version 5.0 (GraphPad Software, San Diego, CA).
RESULTS Mutational Analysis We determined the sequence mutations in 6 genes that are commonly mutated in endometrial carcinomas using Sanger sequencing in 20 UCs and 12 concurrent EMCs (Table 1).11 Eighteen (90%) of 20 UCs carried mutations in at least 1 of the 6 genes tested: 10 (50%) in PIK3CA, 6 (30%) in CTNNB1, 6 (30%) in TP53, 4 (20%) r
2014 Lippincott Williams & Wilkins
Molecular Characterization of UC Associated With EMC
in FBXW7, and 4 (20%) in PPP2R1A. In contrast, 9 (75%) of 12 associated EMCs carried mutations in either 1 of the 6 genes tested: 6 (50%) in PIK3CA, 2 (17%) in CTNNB1, 3 (25%) in TP53, 2 (17%) in FBXW7, and 3 (25%) in PPP2R1A. All mutations were not detected in the adjacent normal tissue in any of the cases, confirming the somatic nature of the identified mutations. No tumors showed KRAS mutations.
Immunohistochemistry In addition to mutational analysis, we also performed immunohistochemical analysis to determine whether the staining patterns of b-catenin and PTEN were different between UC and EMC components from the same patients. The presence of somatic mutations in these genes is known to affect the expression or subcellular localization of their proteins. Specifically, PTEN deletions and mutations that occur throughout its entire coding region cause loss of protein expression.18 In contrast, mutations in exon 3 of CTNNB1, the gene encoding for b-catenin, result in its nuclear accumulation as opposed to cytoplasmic and membranous localization of wild-type b-catenin.19 We found that nuclear expression of b-catenin was present in 7 (47%) of 15 UCs and 3 (25%) of the 12 concurrent EMCs analyzed. PTEN was analyzed in 15 UCs, and the staining was preserved in 9 (60%) and lost in 6 (40%) of 15 UCs. Moreover, PTEN was preserved in 9 (75%) and lost in 3 (25%) of 12 concurrent EMCs. We observed a correlation between CTNNB1 somatic mutation and b-catenin nuclear accumulation by immunohistochemistry (Table 2). In fact, all carcinomas that harbored CTNNB1 somatic mutations presented nuclear accumulation of b-catenin. In contrast, 3 samples (1 EMC and 2 UCs) with nuclear accumulation of b-catenin did not harbor CTNNB1 mutation.
Comparison of Concurrent UCs and Low-grade EMC UCs and associated EMCs harbored identical FBXW7 and PIK3CA somatic mutations in all pairs analyzed. However, in 5 (42%) of 12 matched UC and EMC pairs at least one of those following mutations identified in UC could not be detected in the associated EMC component. Specifically, of these 5 UCs, 3 harbored additional CTNNB1 mutations (Table 1, cases 2, 4, and 8; Fig. 1), one additional PPP2R1A mutation (Table 1, case 7), and one additional TP53 mutation (Table 1, case 11) as compared with their associated EMC components. Moreover, in 2 cases with identical CTNNB1 somatic mutations in both UC and EMC components, we found a much higher percentage of nuclear b-catenin immunoreactivity in UC than the associated EMC (100% vs. 10% and 20% vs. 5%, respectively). Finally, in case 1 there was nuclear accumulation of b-catenin in the EMC component but not in the associated UC. Concordance of immunostaining pattern for PTEN was recorded in all the 12 pairs of samples containing both UC and EMC components. www.ajsp.com |
661
Kuhn et al
Am J Surg Pathol
Volume 38, Number 5, May 2014
TABLE 1. Immunohistochemical and Mutation Profiles in 20 UCs and 12 Concurrent Low-grade EMC Components Immunohistochemistry Case
PTEN
Mutation Analysis
b-catenin* (%)
CTNNB1
FBXW7
PIK3CA
PPP2R1A
TP53
KRAS
1 EMC UC
Preserved Preserved
30 Negative
WT WT
WT WT
G106_R108del G106_R108del
R183W R183W
WT WT
WT WT
EMC UC
Lost Lost
Negative 80
WT D32Y
WT WT
R38C R38C
WT WT
WT WT
WT WT
EMC UC
Preserved Preserved
Negative Negative
WT WT
WT WT
WT WT
WT WT
G266V G266V
WT WT
EMC UC
Preserved Preserved
Negative 100
WT G34E
WT WT
WT WT
WT WT
WT WT
WT WT
EMC UC
Lost Lost
Negative Negative
WT WT
WT WT
H1047R H1047R
R182W R182W
R273H R273H
WT WT
EMC UC
Preserved Preserved
10 100
D32V D32V
WT WT
WT WT
WT WT
WT WT
WT WT
EMC UC
Preserved Preserved
Negative Negative
WT WT
WT WT
WT WT
WT A252V
WT WT
WT WT
EMC UC
Preserved Preserved
Negative 70
WT T41I
WT WT
WT WT
WT WT
WT WT
WT WT
EMC UC 10 EMC UC 11 EMC UC 12 EMC UC 13 UC 14 UC 15 UC 16 UC 17 UC 18 UC 19 UC 20 UC
Preserved Preserved
Negative Negative
WT WT
R465H R465H
E545K E545K
WT WT
E287D E287D
WT WT
Preserved Preserved
Negative Negative
WT WT
R465H R465H
WT WT
WT WT
WT WT
WT WT
Lost Lost
Negative Negative
WT WT
WT WT
K111Nw K111Nw
22984C > T 22984C > T
WT S241P
WT WT
Preserved Preserved
5 20
D32N D32N
WT WT
H1047R H1047R
WT WT
WT WT
WT WT
NP
NP
WT
WT
WT
WT
R273Lz
WT
NP
NP
WT
WT
WT
WT
WT
WT
NP
NP
S33Y
WT
H1047Y
WT
WT
WT
NP
NP
WT
WT
M1043I
WT
WT
WT
NP
NP
WT
WT
T1052K
WT
WT
WT
Lost
100
WT
R465H
R38H
WT
G245D
WT
Lost
5
WT
WT
WT
WT
WT
WT
Lost
Negative
WT
R465H
WT
WT
WT
WT
2 3 4 5 6 7 8 9
*b-catenin percentage concerns the nuclear accumulation. wT1025A. zA298T, S313G. NP indicates not performed; WT, wild type.
Comparison of Molecular Genetic Changes Between UC and Uterine EMC Given that the case number for EMC was small in this study, we also retrieved the data from the TCGA
662 | www.ajsp.com
endometrial cancer database and compared the frequency of these specific molecular genetic alterations in UC from the current study with those of uterine EMCs from the TCGA database. As shown in Table 3, UC had a very r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 5, May 2014
Molecular Characterization of UC Associated With EMC
TABLE 2. Correlation Between b-catenin Immunohistochemistry (IHC) and CTNNB1 Somatic Mutations CTNNB1 Mutation
CTNNB1 Wild Type
Total
7 0 7
3 17 20
10 17 27
Positive b-catenin IHC Negative b-catenin IHC Total P < 0.0001, 2-tailed Fisher exact test.
similar frequency of somatic mutations in CTNNB1, FBXW7, PIK3CA, PPP2R1A, and TP53 as compared
with EMC. In contrast, it appeared that there was a higher frequency of KRAS mutation in EMC (26%) than in UC, in which it was not found in any case examined in this study (P < 0.001, 2-tailed Fisher exact test). In this report, UCs with CTNNB1 mutations did not have TP53 mutations in the analyzed samples, and thus there was a negative correlation between CTNNB1 and TP53 mutations (P < 0.05, 2-tailed Fisher exact test, Table 4). Similarly, by analyzing the TCGA data, we found that EMC (excluding those with hypermutation) with CTNNB1 mutations tended not to harbor TP53 mutations (P < 0.001, 2-tailed Fisher exact test). No
FIGURE 1. Representative photomicrographs from case 4. A, Hematoxylin and eosin stain shows the morphologic features characteristic of low-grade EMC and UC component (top panels). Immunostaining of b-catenin reveals diffuse nuclear accumulation of b-catenin in the UC component, whereas the staining is faint in the EMC component (bottom panels). B, Sequence chromatograms of CTNNB1 demonstrate a wild-type sequence in the EMC component and a somatic mutation in UC (G to A point mutation). r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
663
Kuhn et al
Am J Surg Pathol
TABLE 3. Correlation Between Mutation Frequency in UC Compared With EMC and Uterine Serous Carcinoma (USC) From the TCGA Data Set Mutation Rate
Gene
EMC w/o UC EMC (TCGA, Hypermutation (n = 20) n = 200) (TCGA, (%) (%) n = 121) (%)
PIK3CA CTNNB1 TP53 FBXW7 PPP2R1A KRAS
50 30 30 20 20 0
55 37 14 12 7 26
USC (TCGA, n = 53) (%)
55 46 18 14 6 20
36 0 74 26 23 2
evidence of statistically significant correlation was observed among other mutated genes.
DISCUSSION One of the most intriguing observations in studying the pathogenesis of uterine and ovarian EMCs is its occasional association with UC, which exhibits distinct morphologic features from its adjacent EMC. Those EMCs with the UC component usually behave in a more aggressive manner than those without. Thus, delineation of the molecular changes in UC may have significant biological and clinical implications. In this study, we provide new molecular evidence that UC arises from preexistent EMC because all the molecular changes detected in EMC can be found in the associated UC from the same patients, and, more importantly, additional molecular changes are recorded in UC as compared with the concurrent EMC. These findings accord to the paradigm of the tumor progression model describing that tumor progression is characterized by accumulated molecular genetic alterations.20 The finding that somatic mutations including CTNNB1, FBXW7, PIK3CA, PPP2R1A, and TP53 detected in EMC were also identified in concurrent UC is unlikely a random event and, in fact, strongly supports the clonal relationship between EMC and associated UC. This view is further supported by the immunostaining finding that PTEN loss, presumably due to an inactivating mutation/deletion, is observed in both UC and EMC from the same patients. If UC represents a tumor progression from EMC, one would expect to detect additional mutations in UC but not in EMC. By carefully analyzing the mutational profile between these 2 components from the same tumor, we demonstrated that this is TABLE 4. Correlation Between CTNNB1 and TP53 Somatic Mutations CTNNB1 Mutation
CTNNB1 Wild Type
Total
0 8 8
9 15 24
9 23 32
TP53 mutation TP53 WT Total
P < 0.05, 2-tailed Fisher exact test.
664 | www.ajsp.com
Volume 38, Number 5, May 2014
the case as additional somatic mutations were found in the UC component that could not be identified in the concurrent EMC in 42% matched EMC and UC pairs. This observation not only suggests that EMC is the precursor lesion of UC but also provides new insight into the molecular pathogenesis of how EMC progresses to UC. In particular, additional somatic mutations in CTNNB1, PPP2R1A, and TP53 were identified in UCs. It is plausible that those mutations may play a role in driving the transition from EMC to UC in some of the EMC cases, but it should not be construed that EMCs with those mutations are destined to become UC. This is because mutations in CTNNB1, PPP2R1A, and TP53 can also be detected in many pure (conventional) EMC cases without the UC component. A likely explanation is that tumor progression from EMC to UC depends on specific molecular genetic context in the EMC, so that additional mutations of those genes, and thus the dysregulation of the pathways they control, may increase the “oncogenic dosage” to EMCs and drive the progression from EMC to UC. The concept of “oncogenic dosage” describes the overall accumulated tumor-promoting effects contributed by individual oncogenic mutation and/or tumor suppressor–inactivating mutation. It is the “dosage” rather than the type of constituent molecular genetic alterations that matters. It has been shown in the engineered animal models that PTEN deletion alone is, in general, insufficient to drive ovarian tumor formation, and it requires one of the additional molecular alterations—Apc deletion, Pik3ca mutation, Kras mutation, or Arid1a—to have the sufficient “dosage” for tumor initiation.21–23 Another interesting finding is that we did not detect KRAS mutations in any of the UCs or associated EMCs, as opposed to a relatively high frequency of KRAS mutations in EMC reported by TCGA (26%). This finding raises the possibility that those EMCs that progress to UC may be molecularly different from those that are not associated with UC (eg, pure EMC). Therefore, only a minority of EMCs is prone to progress to UC, and later acquisition of mutations, like CTNNB1 and TP53 mutations, in EMC subclones may clonally expand and develop UC. On the basis of this observation, the role of CTNNB1 mutation is context dependent. We propose that additional CTNNB1 mutations may be associated with tumor progression of UC. Therefore, unlike conventional EMC, CTNNB1 mutations seem to be acquired late in the pathogenesis of EMC that progresses into UC. Future genome-wide studies are required to comprehensively catalog molecular genetic alterations in UC and associated EMC to identify the potential genes and the pathways that drive progression from EMC to UC. The highly malignant nature of UC raises a question whether UC is molecularly close to uterine serous carcinoma, the prototype of type II uterine carcinoma, and should be classified as such.24,25 By comparing the mutation profiles of CTNNB1, FBXW7, PIK3CA, PPP2R1A, and TP53 and those from the TCGA data, we show that UC has a very similar mutation profile to uterine EMC but not to uterine serous carcinoma r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 5, May 2014
(Table 3).11,14 Likewise, loss of PTEN expression, presumably because of inactivating mutations/deletions, is detected in 40% of UC, but PTEN mutations rarely occur in uterine serous carcinoma. This is expected because UC may likely represent a subsequent step of tumor progression from EMC and therefore has a mutation profile characteristic of EMC. What is surprising is that the mutation frequency of TP53 in UC is strikingly low as compared with other highly aggressive gynecologic carcinomas including uterine serous carcinoma and high-grade ovarian serous carcinoma. Finally, the fact that TP53 mutations do not co-occur with CTNNB1 mutations in UC, as well as in the TCGA endometrial cancer data, supports that mutations in both genes may be functionally redundant or may confer negative cell selection. Thus, EMC-UC tumor progression may represent another pathway that results in highly malignant uterine carcinoma besides the pathways leading to uterine serous carcinoma and carcinosarcoma. In summary, UC is a unique neoplastic disease that is frequently associated with uterine or ovarian well-differentiated EMC. Our molecular genetic analyses demonstrate that UC is clonally related to EMC, which likely represents the precursor of UC. The tumor, like EMC, is characterized by the molecular alterations in the pathways involving b-catenin and PTEN, of which molecular changes are uncommon in the type II gynecologic tumors including uterine serous carcinoma and ovarian highgrade serous carcinoma. The results from this study provide new molecular insight into this underrecognized and highly malignant gynecologic cancer. REFERENCES 1. Altrabulsi B, Malpica A, Deavers MT, et al. Undifferentiated carcinoma of the endometrium. Am J Surg Pathol. 2005;29:1316–1321. 2. Silva EG, Deavers MT, Malpica A. Undifferentiated carcinoma of the endometrium: a review. Pathology. 2007;39:134–138. 3. Tafe LJ, Garg K, Chew I, et al. Endometrial and ovarian carcinomas with undifferentiated components: clinically aggressive and frequently underrecognized neoplasms. Mod Pathol. 2010;23:781–789. 4. Abeler VM, Kjorstad KE, Nesland JM. Undifferentiated carcinoma of the endometrium. A histopathologic and clinical study of 31 cases. Cancer. 1991;68:98–105. 5. Wilson TO, Podratz KC, Gaffey TA, et al. Evaluation of unfavorable histologic subtypes in endometrial adenocarcinoma. Am J Obstet Gynecol. 1990;162:418–423. 6. Robboy SJ, Bradley R. Changing trends and prognostic features in endometrial cancer associated with exogenous estrogen therapy. Obstet Gynecol. 1979;54:269–277. 7. Benediktsdottir KR, Jonasson JG, Hallgrimsson J. Tumours in Iceland: 12. Malignant tumours of the corpus of the uterus. APMIS. 1989;97:781–786.
r
2014 Lippincott Williams & Wilkins
Molecular Characterization of UC Associated With EMC
8. Wu E, Shih IM, Diaz-Montes TP. Dedifferentiated endometrioid adenocarcinoma: an under-recognized but aggressive tumor? Gynecol Oncol Rep. 2013;5:25–27. 9. Giordano G, D’Adda T, Bottarelli L, et al. Two cases of low-grade endometrioid carcinoma associated with undifferentiated carcinoma of the uterus (dedifferentiated carcinoma): a molecular study. Pathol Oncol Res. 2012;18:523–528. 10. Romero-Pe´rez L, Lo´pez-Garcı´ a MA, Dı´ az-Martı´ n J, et al. ZEB1 overexpression associated with E-cadherin and microRNA-200 downregulation is characteristic of undifferentiated endometrial carcinoma. Mod Pathol. 2013. [Epub ahead of print]. 11. Cancer Genome Atlas Research Network. Kandoth C, Schultz N, et al. Integrated genomic characterization of endometrial carcinoma. Nature. 2013;497:67–73. 12. Kobel M, Bell D, Carcangiu ML, et al. Undifferentiated carcinoma. In: Kurman RJ, Carcangiu ML, Herrington CS, et al, eds. World Health Organization Classification of Tumors- Gynecological Malignancy. 4th ed. Lyon, France: IARC Press; 2014. (In press). 13. Kuhn E, Teller L, Piana S, et al. Different clonal origin of bilateral papillary thyroid carcinoma, with a review of the literature. Endocr Pathol. 2012;23:101–107. 14. Kuhn E, Wu RC, Guan B, et al. Identification of molecular pathway aberrations in uterine serous carcinoma by genome-wide analyses. J Natl Cancer Inst. 2012;104:1503–1513. 15. Kuhn E, Ayhan A, IeM Shih, et al. The pathogenesis of atypical proliferative Brenner tumor: an immunohistochemical and molecular genetic analysis. Mod Pathol. 2013. [Epub ahead of print]. 16. Kuhn E, Kurman RJ, Vang R, et al. TP53 mutations in serous tubal intraepithelial carcinoma and concurrent pelvic high-grade serous carcinoma-evidence supporting the clonal relationship of the two lesions. J Pathol. 2012;226:421–426. 17. Kuhn E, Ayhan A, Shih IM, et al. Ovarian Brenner tumour: a morphologic and immunohistochemical analysis suggesting an origin from fallopian tube epithelium. Eur J Cancer. 2013;49: 3839–3849. 18. Djordjevic B, Hennessy BT, Li J, et al. Clinical assessment of PTEN loss in endometrial carcinoma: immunohistochemistry outperforms gene sequencing. Mod Pathol. 2012;25:699–708. 19. Moreno-Bueno G, Hardisson D, Sa´nchez C, et al. Abnormalities of the APC/beta-catenin pathway in endometrial cancer. Oncogene. 2002;21:7981–7990. 20. Merlo LM, Pepper JW, Reid BJ, et al. Cancer as an evolutionary and ecological process. Nat Rev Cancer. 2006;6:924–935. 21. Kinross KM, Montgomery KG, Kleinschmidt M, et al. An activating Pik3ca mutation coupled with Pten loss is sufficient to initiate ovarian tumorigenesis in mice. J Clin Invest. 2012;122:553–557. 22. Dinulescu DM, Ince TA, Quade BJ, et al. Role of K-ras and Pten in the development of mouse models of endometriosis and endometrioid ovarian cancer. Nat Med. 2005;11:63–70. 23. Wu R, Hendrix-Lucas N, Kuick R, et al. Mouse model of human ovarian endometrioid adenocarcinoma based on somatic defects in the Wnt/beta-catenin and PI3K/Pten signaling pathways. Cancer Cell. 2007;11:321–333. 24. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol Oncol. 1983;15:10–17. 25. Kuhn E, Bahadirli-Talbott A, Shih IeM. Frequent CCNE1 amplification in endometrial intraepithelial carcinoma and uterine serous carcinoma. Mod Pathol. 2013. [Epub ahead of print].
www.ajsp.com |
665
Molecular Characterization of Undifferentiated Carcinoma Associated With Endometrioid Carcinoma Elisabetta Kuhn, MD,* Ayse Ayhan, MD, PhD,w Asli Bahadirli-Talbott, MSc,* Chengquan Zhao, MD,z and Ie-Ming Shih, MD, PhD*y8
Abstract: Uterine and ovarian undifferentiated carcinomas (UCs) are often associated with low-grade endometrioid carcinomas (EMCs) and are characterized by a solid growth pattern and a lack of appreciable features of differentiation. As compared with pure EMC, UC is highly malignant, and the molecular pathogenesis that leads to disease aggressiveness remains largely unknown. This study interrogates the molecular pathogenesis of UCs by comparing the molecular alterations between the UC and the EMC components. A total of 20 UCs were studied, 12 of which contained both UC and EMC components. Mutation analysis was performed for the genes commonly mutated in EMC, and immunohistochemistry was used to determine the expression pattern of b-catenin and PTEN. Sequencing analysis revealed that UCs harbored somatic mutations in PIK3CA (50%), CTNNB1 (30%), TP53 (30%), FBXW7 (20%), and PPP2R1A (20%). All somatic mutations detected in EMCs were also present in concurrent UCs. Moreover, additional somatic mutations were detected in the UC component in 5 (42%) cases with concurrent EMC and UC. Concordance of immunostaining pattern for b-catenin and PTEN was recorded in all 12 matched EMCs and UCs, except 4 cases in which nuclear accumulation of b-catenin staining was detected in one of the components but not in the other. Our findings support a clonal relationship between EMCs and their associated UCs. Additional molecular genetics alteration, including mutations of CTNNB1, PPP2R1A, and TP53, may contribute to tumor progression from EMC to UC. Key Words: undifferentiated carcinoma, endometrioid carcinoma, b-catenin, progression (Am J Surg Pathol 2014;38:660–665)
From the Departments of *Pathology; yGynecology and Obstetrics; 8Oncology, Johns Hopkins Medical Institutions, Baltimore, MD; zDepartment of Pathology, Magee-Womens Hospital, University of Pittsburgh Medical Center, Pittsburg, PA; and wDepartment of Pathology, Seirei Mikatahara Hospital, Hamamatsu, Japan. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Ie-Ming Shih, MD, PhD, Department of Pathology, Johns Hopkins University, 1550 Orleans Street, CRBII, room 305, Baltimore, MD 21231 (e-mail: [email protected]) and Elisabetta Kuhn, MD, Department of Pathology, Johns Hopkins Medical Institutions, Baltimore, MD (e-mail: [email protected]). Copyright r 2014 by Lippincott Williams & Wilkins
660 | www.ajsp.com
U
ndifferentiated carcinoma (UC) represents a unique and highly aggressive form of uterine cancer. The morphologic features of UC were not well defined until Silva and colleagues proposed the defining characters including medium-sized or large-sized monotonous cells, a solid and diffuse growth pattern, and a lack of appreciable features of differentiation—that is, glandular, papillary, squamous, and neuroendocrine.1–3 On the basis of these morphologic diagnostic criteria, the investigators reported an incidence of 9% among the endometrial carcinomas in their institution.2 Therefore, UC is probably underrecognized as it is previously reported to represent only 1% to 2% of all endometrial carcinomas.1,2,4–7 Clinically, the majority of UCs presents at advanced stages displays an aggressive clinical behavior and does not respond to the conventional chemotherapy for endometrial carcinoma. These characteristics highlight the importance of a correct diagnosis of this endometrial carcinoma variant. Very interestingly, UC is often associated with endometrioid carcinoma (EMC) and is thought to arise from it.1–3,8 However, the molecular genetic evidence for its genesis is still lacking, and the pathogenesis that leads to UC remains largely unknown. In fact, apart from frequent microsatellite instability, little is known about the molecular genetic alterations in UC.3,9,10 In particular, this special type of uterine cancer was not included in The Cancer Genome Atlas (TCGA), which has detailed the molecular landscape of uterine carcinomas.11 Thus, in this study, we performed mutational and immuno histochemical analyses in UCs, many of which are associated with well-differentiated EMC, allowing direct comparison of molecular alterations between UC and EMC components from the same cases.
MATERIALS AND METHODS Case Selection We selected a total of 20 UCs from the Johns Hopkins Hospital (Baltimore, MD), Seirei Mikatahara Hospital (Hamamatsu, Japan), and Magee-Womens Hospital (Pittsburg, PA). Among them, 18 involved the uterus, and 2 involved the ovary. There were 12 cases containing both UC and EMC components. All the cases were rereviewed by 2 pathologists (E.K. and I.-M.S.) to confirm the diagnosis according to the morphologic criteria previously Am J Surg Pathol
Volume 38, Number 5, May 2014
Am J Surg Pathol
Volume 38, Number 5, May 2014
reported and recently described in the new World Health Organization classification.3,12 This study was approved by the Institutional Review Board of the Johns Hopkins Medical Institutions.
Mutation Analysis Genomic DNA was prepared from serial 5-mm-thick paraffin sections. Both UC and EMC tumor components and adjacent normal tissue were collected separately by manual microdissection as previously described.13 Genomic DNA was extracted for sequencing using a QIAamp DNA Micro Kit (Qiagen, Valencia, CA) according to the vendor’s protocol. Mutation analysis was performed in all cases for the regions that frequently harbor mutations in the following genes: CTNNB1 (b-catenin-encoding gene, exon 3), FBXW7 (exons 9 and 10), KRAS (exon 2), PIK3CA (exons 1, 9, and 20), PPP2R1A (exons 5 and 6), and TP53 (exons 4 to 9). The above genes were analyzed because they are commonly mutated in endometrial carcinomas.11 Each exon was polymerase chain reaction amplified, using primers previously reported.14–16 All polymerase chain reaction products were sequenced at Agencourt Biosciences (Beverly, MA), and all sequence variants were confirmed by at least 3 independent rounds of experiments.
Immunohistochemistry Immunohistochemical analysis was performed using antibodies against b-catenin (cat # 610154; BD Transduction Laboratories, Mississauga, ON; dilution 1:1000) and PTEN (cat # M3627; Dako, Carpinteria, CA; dilution 1:100). Formalin-fixed paraffin-embedded sections were immunostained manually or by the autostainer Bond Max Automated Immunohistochemistry Vision Biosystem (Leica Microsystems GmbH, Wetzlar, Germany) using a previously described protocol.15–17 Subcellular localization (cytoplasmic, membranous, or nuclear) was recorded. For b-catenin, a positive reaction was defined as nuclear immunoreactivity in at least 5% of the tumor cells. For PTEN, a preserved reaction was defined as positivity in at least 20% of the tumor cells. In addition, all observations were made for both UC and EMC components, when present.
Statistical Analysis Comparisons of immunohistochemistry and mutation outcomes were performed using the 2-tailed unpaired Fisher exact test. P-values of 0.05 or less were considered statistically significant. Statistical analysis was carried out using the GraphPad Prism software version 5.0 (GraphPad Software, San Diego, CA).
RESULTS Mutational Analysis We determined the sequence mutations in 6 genes that are commonly mutated in endometrial carcinomas using Sanger sequencing in 20 UCs and 12 concurrent EMCs (Table 1).11 Eighteen (90%) of 20 UCs carried mutations in at least 1 of the 6 genes tested: 10 (50%) in PIK3CA, 6 (30%) in CTNNB1, 6 (30%) in TP53, 4 (20%) r
2014 Lippincott Williams & Wilkins
Molecular Characterization of UC Associated With EMC
in FBXW7, and 4 (20%) in PPP2R1A. In contrast, 9 (75%) of 12 associated EMCs carried mutations in either 1 of the 6 genes tested: 6 (50%) in PIK3CA, 2 (17%) in CTNNB1, 3 (25%) in TP53, 2 (17%) in FBXW7, and 3 (25%) in PPP2R1A. All mutations were not detected in the adjacent normal tissue in any of the cases, confirming the somatic nature of the identified mutations. No tumors showed KRAS mutations.
Immunohistochemistry In addition to mutational analysis, we also performed immunohistochemical analysis to determine whether the staining patterns of b-catenin and PTEN were different between UC and EMC components from the same patients. The presence of somatic mutations in these genes is known to affect the expression or subcellular localization of their proteins. Specifically, PTEN deletions and mutations that occur throughout its entire coding region cause loss of protein expression.18 In contrast, mutations in exon 3 of CTNNB1, the gene encoding for b-catenin, result in its nuclear accumulation as opposed to cytoplasmic and membranous localization of wild-type b-catenin.19 We found that nuclear expression of b-catenin was present in 7 (47%) of 15 UCs and 3 (25%) of the 12 concurrent EMCs analyzed. PTEN was analyzed in 15 UCs, and the staining was preserved in 9 (60%) and lost in 6 (40%) of 15 UCs. Moreover, PTEN was preserved in 9 (75%) and lost in 3 (25%) of 12 concurrent EMCs. We observed a correlation between CTNNB1 somatic mutation and b-catenin nuclear accumulation by immunohistochemistry (Table 2). In fact, all carcinomas that harbored CTNNB1 somatic mutations presented nuclear accumulation of b-catenin. In contrast, 3 samples (1 EMC and 2 UCs) with nuclear accumulation of b-catenin did not harbor CTNNB1 mutation.
Comparison of Concurrent UCs and Low-grade EMC UCs and associated EMCs harbored identical FBXW7 and PIK3CA somatic mutations in all pairs analyzed. However, in 5 (42%) of 12 matched UC and EMC pairs at least one of those following mutations identified in UC could not be detected in the associated EMC component. Specifically, of these 5 UCs, 3 harbored additional CTNNB1 mutations (Table 1, cases 2, 4, and 8; Fig. 1), one additional PPP2R1A mutation (Table 1, case 7), and one additional TP53 mutation (Table 1, case 11) as compared with their associated EMC components. Moreover, in 2 cases with identical CTNNB1 somatic mutations in both UC and EMC components, we found a much higher percentage of nuclear b-catenin immunoreactivity in UC than the associated EMC (100% vs. 10% and 20% vs. 5%, respectively). Finally, in case 1 there was nuclear accumulation of b-catenin in the EMC component but not in the associated UC. Concordance of immunostaining pattern for PTEN was recorded in all the 12 pairs of samples containing both UC and EMC components. www.ajsp.com |
661
Kuhn et al
Am J Surg Pathol
Volume 38, Number 5, May 2014
TABLE 1. Immunohistochemical and Mutation Profiles in 20 UCs and 12 Concurrent Low-grade EMC Components Immunohistochemistry Case
PTEN
Mutation Analysis
b-catenin* (%)
CTNNB1
FBXW7
PIK3CA
PPP2R1A
TP53
KRAS
1 EMC UC
Preserved Preserved
30 Negative
WT WT
WT WT
G106_R108del G106_R108del
R183W R183W
WT WT
WT WT
EMC UC
Lost Lost
Negative 80
WT D32Y
WT WT
R38C R38C
WT WT
WT WT
WT WT
EMC UC
Preserved Preserved
Negative Negative
WT WT
WT WT
WT WT
WT WT
G266V G266V
WT WT
EMC UC
Preserved Preserved
Negative 100
WT G34E
WT WT
WT WT
WT WT
WT WT
WT WT
EMC UC
Lost Lost
Negative Negative
WT WT
WT WT
H1047R H1047R
R182W R182W
R273H R273H
WT WT
EMC UC
Preserved Preserved
10 100
D32V D32V
WT WT
WT WT
WT WT
WT WT
WT WT
EMC UC
Preserved Preserved
Negative Negative
WT WT
WT WT
WT WT
WT A252V
WT WT
WT WT
EMC UC
Preserved Preserved
Negative 70
WT T41I
WT WT
WT WT
WT WT
WT WT
WT WT
EMC UC 10 EMC UC 11 EMC UC 12 EMC UC 13 UC 14 UC 15 UC 16 UC 17 UC 18 UC 19 UC 20 UC
Preserved Preserved
Negative Negative
WT WT
R465H R465H
E545K E545K
WT WT
E287D E287D
WT WT
Preserved Preserved
Negative Negative
WT WT
R465H R465H
WT WT
WT WT
WT WT
WT WT
Lost Lost
Negative Negative
WT WT
WT WT
K111Nw K111Nw
22984C > T 22984C > T
WT S241P
WT WT
Preserved Preserved
5 20
D32N D32N
WT WT
H1047R H1047R
WT WT
WT WT
WT WT
NP
NP
WT
WT
WT
WT
R273Lz
WT
NP
NP
WT
WT
WT
WT
WT
WT
NP
NP
S33Y
WT
H1047Y
WT
WT
WT
NP
NP
WT
WT
M1043I
WT
WT
WT
NP
NP
WT
WT
T1052K
WT
WT
WT
Lost
100
WT
R465H
R38H
WT
G245D
WT
Lost
5
WT
WT
WT
WT
WT
WT
Lost
Negative
WT
R465H
WT
WT
WT
WT
2 3 4 5 6 7 8 9
*b-catenin percentage concerns the nuclear accumulation. wT1025A. zA298T, S313G. NP indicates not performed; WT, wild type.
Comparison of Molecular Genetic Changes Between UC and Uterine EMC Given that the case number for EMC was small in this study, we also retrieved the data from the TCGA
662 | www.ajsp.com
endometrial cancer database and compared the frequency of these specific molecular genetic alterations in UC from the current study with those of uterine EMCs from the TCGA database. As shown in Table 3, UC had a very r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 5, May 2014
Molecular Characterization of UC Associated With EMC
TABLE 2. Correlation Between b-catenin Immunohistochemistry (IHC) and CTNNB1 Somatic Mutations CTNNB1 Mutation
CTNNB1 Wild Type
Total
7 0 7
3 17 20
10 17 27
Positive b-catenin IHC Negative b-catenin IHC Total P < 0.0001, 2-tailed Fisher exact test.
similar frequency of somatic mutations in CTNNB1, FBXW7, PIK3CA, PPP2R1A, and TP53 as compared
with EMC. In contrast, it appeared that there was a higher frequency of KRAS mutation in EMC (26%) than in UC, in which it was not found in any case examined in this study (P < 0.001, 2-tailed Fisher exact test). In this report, UCs with CTNNB1 mutations did not have TP53 mutations in the analyzed samples, and thus there was a negative correlation between CTNNB1 and TP53 mutations (P < 0.05, 2-tailed Fisher exact test, Table 4). Similarly, by analyzing the TCGA data, we found that EMC (excluding those with hypermutation) with CTNNB1 mutations tended not to harbor TP53 mutations (P < 0.001, 2-tailed Fisher exact test). No
FIGURE 1. Representative photomicrographs from case 4. A, Hematoxylin and eosin stain shows the morphologic features characteristic of low-grade EMC and UC component (top panels). Immunostaining of b-catenin reveals diffuse nuclear accumulation of b-catenin in the UC component, whereas the staining is faint in the EMC component (bottom panels). B, Sequence chromatograms of CTNNB1 demonstrate a wild-type sequence in the EMC component and a somatic mutation in UC (G to A point mutation). r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
663
Kuhn et al
Am J Surg Pathol
TABLE 3. Correlation Between Mutation Frequency in UC Compared With EMC and Uterine Serous Carcinoma (USC) From the TCGA Data Set Mutation Rate
Gene
EMC w/o UC EMC (TCGA, Hypermutation (n = 20) n = 200) (TCGA, (%) (%) n = 121) (%)
PIK3CA CTNNB1 TP53 FBXW7 PPP2R1A KRAS
50 30 30 20 20 0
55 37 14 12 7 26
USC (TCGA, n = 53) (%)
55 46 18 14 6 20
36 0 74 26 23 2
evidence of statistically significant correlation was observed among other mutated genes.
DISCUSSION One of the most intriguing observations in studying the pathogenesis of uterine and ovarian EMCs is its occasional association with UC, which exhibits distinct morphologic features from its adjacent EMC. Those EMCs with the UC component usually behave in a more aggressive manner than those without. Thus, delineation of the molecular changes in UC may have significant biological and clinical implications. In this study, we provide new molecular evidence that UC arises from preexistent EMC because all the molecular changes detected in EMC can be found in the associated UC from the same patients, and, more importantly, additional molecular changes are recorded in UC as compared with the concurrent EMC. These findings accord to the paradigm of the tumor progression model describing that tumor progression is characterized by accumulated molecular genetic alterations.20 The finding that somatic mutations including CTNNB1, FBXW7, PIK3CA, PPP2R1A, and TP53 detected in EMC were also identified in concurrent UC is unlikely a random event and, in fact, strongly supports the clonal relationship between EMC and associated UC. This view is further supported by the immunostaining finding that PTEN loss, presumably due to an inactivating mutation/deletion, is observed in both UC and EMC from the same patients. If UC represents a tumor progression from EMC, one would expect to detect additional mutations in UC but not in EMC. By carefully analyzing the mutational profile between these 2 components from the same tumor, we demonstrated that this is TABLE 4. Correlation Between CTNNB1 and TP53 Somatic Mutations CTNNB1 Mutation
CTNNB1 Wild Type
Total
0 8 8
9 15 24
9 23 32
TP53 mutation TP53 WT Total
P < 0.05, 2-tailed Fisher exact test.
664 | www.ajsp.com
Volume 38, Number 5, May 2014
the case as additional somatic mutations were found in the UC component that could not be identified in the concurrent EMC in 42% matched EMC and UC pairs. This observation not only suggests that EMC is the precursor lesion of UC but also provides new insight into the molecular pathogenesis of how EMC progresses to UC. In particular, additional somatic mutations in CTNNB1, PPP2R1A, and TP53 were identified in UCs. It is plausible that those mutations may play a role in driving the transition from EMC to UC in some of the EMC cases, but it should not be construed that EMCs with those mutations are destined to become UC. This is because mutations in CTNNB1, PPP2R1A, and TP53 can also be detected in many pure (conventional) EMC cases without the UC component. A likely explanation is that tumor progression from EMC to UC depends on specific molecular genetic context in the EMC, so that additional mutations of those genes, and thus the dysregulation of the pathways they control, may increase the “oncogenic dosage” to EMCs and drive the progression from EMC to UC. The concept of “oncogenic dosage” describes the overall accumulated tumor-promoting effects contributed by individual oncogenic mutation and/or tumor suppressor–inactivating mutation. It is the “dosage” rather than the type of constituent molecular genetic alterations that matters. It has been shown in the engineered animal models that PTEN deletion alone is, in general, insufficient to drive ovarian tumor formation, and it requires one of the additional molecular alterations—Apc deletion, Pik3ca mutation, Kras mutation, or Arid1a—to have the sufficient “dosage” for tumor initiation.21–23 Another interesting finding is that we did not detect KRAS mutations in any of the UCs or associated EMCs, as opposed to a relatively high frequency of KRAS mutations in EMC reported by TCGA (26%). This finding raises the possibility that those EMCs that progress to UC may be molecularly different from those that are not associated with UC (eg, pure EMC). Therefore, only a minority of EMCs is prone to progress to UC, and later acquisition of mutations, like CTNNB1 and TP53 mutations, in EMC subclones may clonally expand and develop UC. On the basis of this observation, the role of CTNNB1 mutation is context dependent. We propose that additional CTNNB1 mutations may be associated with tumor progression of UC. Therefore, unlike conventional EMC, CTNNB1 mutations seem to be acquired late in the pathogenesis of EMC that progresses into UC. Future genome-wide studies are required to comprehensively catalog molecular genetic alterations in UC and associated EMC to identify the potential genes and the pathways that drive progression from EMC to UC. The highly malignant nature of UC raises a question whether UC is molecularly close to uterine serous carcinoma, the prototype of type II uterine carcinoma, and should be classified as such.24,25 By comparing the mutation profiles of CTNNB1, FBXW7, PIK3CA, PPP2R1A, and TP53 and those from the TCGA data, we show that UC has a very similar mutation profile to uterine EMC but not to uterine serous carcinoma r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 5, May 2014
(Table 3).11,14 Likewise, loss of PTEN expression, presumably because of inactivating mutations/deletions, is detected in 40% of UC, but PTEN mutations rarely occur in uterine serous carcinoma. This is expected because UC may likely represent a subsequent step of tumor progression from EMC and therefore has a mutation profile characteristic of EMC. What is surprising is that the mutation frequency of TP53 in UC is strikingly low as compared with other highly aggressive gynecologic carcinomas including uterine serous carcinoma and high-grade ovarian serous carcinoma. Finally, the fact that TP53 mutations do not co-occur with CTNNB1 mutations in UC, as well as in the TCGA endometrial cancer data, supports that mutations in both genes may be functionally redundant or may confer negative cell selection. Thus, EMC-UC tumor progression may represent another pathway that results in highly malignant uterine carcinoma besides the pathways leading to uterine serous carcinoma and carcinosarcoma. In summary, UC is a unique neoplastic disease that is frequently associated with uterine or ovarian well-differentiated EMC. Our molecular genetic analyses demonstrate that UC is clonally related to EMC, which likely represents the precursor of UC. The tumor, like EMC, is characterized by the molecular alterations in the pathways involving b-catenin and PTEN, of which molecular changes are uncommon in the type II gynecologic tumors including uterine serous carcinoma and ovarian highgrade serous carcinoma. The results from this study provide new molecular insight into this underrecognized and highly malignant gynecologic cancer. REFERENCES 1. Altrabulsi B, Malpica A, Deavers MT, et al. Undifferentiated carcinoma of the endometrium. Am J Surg Pathol. 2005;29:1316–1321. 2. Silva EG, Deavers MT, Malpica A. Undifferentiated carcinoma of the endometrium: a review. Pathology. 2007;39:134–138. 3. Tafe LJ, Garg K, Chew I, et al. Endometrial and ovarian carcinomas with undifferentiated components: clinically aggressive and frequently underrecognized neoplasms. Mod Pathol. 2010;23:781–789. 4. Abeler VM, Kjorstad KE, Nesland JM. Undifferentiated carcinoma of the endometrium. A histopathologic and clinical study of 31 cases. Cancer. 1991;68:98–105. 5. Wilson TO, Podratz KC, Gaffey TA, et al. Evaluation of unfavorable histologic subtypes in endometrial adenocarcinoma. Am J Obstet Gynecol. 1990;162:418–423. 6. Robboy SJ, Bradley R. Changing trends and prognostic features in endometrial cancer associated with exogenous estrogen therapy. Obstet Gynecol. 1979;54:269–277. 7. Benediktsdottir KR, Jonasson JG, Hallgrimsson J. Tumours in Iceland: 12. Malignant tumours of the corpus of the uterus. APMIS. 1989;97:781–786.
r
2014 Lippincott Williams & Wilkins
Molecular Characterization of UC Associated With EMC
8. Wu E, Shih IM, Diaz-Montes TP. Dedifferentiated endometrioid adenocarcinoma: an under-recognized but aggressive tumor? Gynecol Oncol Rep. 2013;5:25–27. 9. Giordano G, D’Adda T, Bottarelli L, et al. Two cases of low-grade endometrioid carcinoma associated with undifferentiated carcinoma of the uterus (dedifferentiated carcinoma): a molecular study. Pathol Oncol Res. 2012;18:523–528. 10. Romero-Pe´rez L, Lo´pez-Garcı´ a MA, Dı´ az-Martı´ n J, et al. ZEB1 overexpression associated with E-cadherin and microRNA-200 downregulation is characteristic of undifferentiated endometrial carcinoma. Mod Pathol. 2013. [Epub ahead of print]. 11. Cancer Genome Atlas Research Network. Kandoth C, Schultz N, et al. Integrated genomic characterization of endometrial carcinoma. Nature. 2013;497:67–73. 12. Kobel M, Bell D, Carcangiu ML, et al. Undifferentiated carcinoma. In: Kurman RJ, Carcangiu ML, Herrington CS, et al, eds. World Health Organization Classification of Tumors- Gynecological Malignancy. 4th ed. Lyon, France: IARC Press; 2014. (In press). 13. Kuhn E, Teller L, Piana S, et al. Different clonal origin of bilateral papillary thyroid carcinoma, with a review of the literature. Endocr Pathol. 2012;23:101–107. 14. Kuhn E, Wu RC, Guan B, et al. Identification of molecular pathway aberrations in uterine serous carcinoma by genome-wide analyses. J Natl Cancer Inst. 2012;104:1503–1513. 15. Kuhn E, Ayhan A, IeM Shih, et al. The pathogenesis of atypical proliferative Brenner tumor: an immunohistochemical and molecular genetic analysis. Mod Pathol. 2013. [Epub ahead of print]. 16. Kuhn E, Kurman RJ, Vang R, et al. TP53 mutations in serous tubal intraepithelial carcinoma and concurrent pelvic high-grade serous carcinoma-evidence supporting the clonal relationship of the two lesions. J Pathol. 2012;226:421–426. 17. Kuhn E, Ayhan A, Shih IM, et al. Ovarian Brenner tumour: a morphologic and immunohistochemical analysis suggesting an origin from fallopian tube epithelium. Eur J Cancer. 2013;49: 3839–3849. 18. Djordjevic B, Hennessy BT, Li J, et al. Clinical assessment of PTEN loss in endometrial carcinoma: immunohistochemistry outperforms gene sequencing. Mod Pathol. 2012;25:699–708. 19. Moreno-Bueno G, Hardisson D, Sa´nchez C, et al. Abnormalities of the APC/beta-catenin pathway in endometrial cancer. Oncogene. 2002;21:7981–7990. 20. Merlo LM, Pepper JW, Reid BJ, et al. Cancer as an evolutionary and ecological process. Nat Rev Cancer. 2006;6:924–935. 21. Kinross KM, Montgomery KG, Kleinschmidt M, et al. An activating Pik3ca mutation coupled with Pten loss is sufficient to initiate ovarian tumorigenesis in mice. J Clin Invest. 2012;122:553–557. 22. Dinulescu DM, Ince TA, Quade BJ, et al. Role of K-ras and Pten in the development of mouse models of endometriosis and endometrioid ovarian cancer. Nat Med. 2005;11:63–70. 23. Wu R, Hendrix-Lucas N, Kuick R, et al. Mouse model of human ovarian endometrioid adenocarcinoma based on somatic defects in the Wnt/beta-catenin and PI3K/Pten signaling pathways. Cancer Cell. 2007;11:321–333. 24. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol Oncol. 1983;15:10–17. 25. Kuhn E, Bahadirli-Talbott A, Shih IeM. Frequent CCNE1 amplification in endometrial intraepithelial carcinoma and uterine serous carcinoma. Mod Pathol. 2013. [Epub ahead of print].
www.ajsp.com |
665
