ORIGINAL ARTICLE

Morphological and Immunohistochemical Reevaluation of

Tumors Initially Diagnosed as Ovarian Endometrioid
Carcinoma With Emphasis on High-grade Tumors
Diana Lim, MD,* Rajmohan Murali, MBBS, MD, FRCPA,wz Melissa P. Murray, MD,w
Emanuela Veras, MD,w Kay J. Park, MD,w and Robert A. Soslow, MDw

transitional cell–like architecture) and were immunophenotypi-

Abstract: Ovarian endometrioid carcinomas (OEC) of low grade cally indistinguishable from OSCs with papillary architecture.
have characteristic morphologic features, but high-grade tumors Sixty percent of OSC overexpressed p16, 50% overexpressed
can mimic high-grade serous and undifferentiated carcinomas. p53, and 82% expressed WT-1. At last follow-up, 52% had died
We reviewed tumors initially diagnosed as OEC to determine of disease. Compared with OSC, OEC patients more frequently
whether a combination of pathologic and immunohistochemical presented below 60 years of age (P = 0.046), had low-stage tu-
features can improve histologic subclassification. Tumors ini- mors (P < 0.001), were more frequently unilateral (P < 0.001),
tially diagnosed as OEC were reviewed using World Health more frequently had synchronous endometrial endometrioid
Organization criteria. We also noted the presence of associated carcinomas (P < 0.001); and had no evidence of disease at last
confirmatory endometrioid features (CEFs): (i) squamous met- follow-up (P < 0.001). Their tumors were of lower grade
aplasia; (ii) endometriosis; (iii) adenofibromatous background; (P < 0.001), had more CEFs (P < 0.001), and less frequently
and (iv) borderline endometrioid or mixed Mullerian compo- overexpressed p16 and p53 (P = 0.003 and P < 0.001, re-
nent. A tissue microarray was constructed from 27 representa- spectively) and less frequently expressed WT-1 (P < 0.001). This
tive tumors with CEF and 14 without CEF, and sections were analysis emphasizes the diagnostic value of CEFs, the presence
stained for WT-1, p16, and p53. Of 109 tumors initially diag- of a low-grade gland-forming endometrioid component, and
nosed as OEC, 76 (70%) tumors were classified as OEC. The WT-1 negativity, as valid, clinically relevant criteria for a di-
median patient age was 55 years, and 75% of patients were agnosis of OEC. Glandular and/or cribriform architecture alone
younger than 60 years. Ninety-two percent presented with dis- may be seen in both OECs and OSCs and are therefore not
ease confined to the pelvis, and 87% of tumors were unilateral. informative of diagnosis. Further study is needed to elaborate
The median tumor size was 11.8 cm. Only 3% of tumors were the characteristics of the exceedingly rare high-grade OEC.
high grade (grade 3of 3). Eighty percent of cases had at least 1
CEF, and 59% had at least 2 CEFs. Eleven percent overex- Key Words: diagnosis, endometrioid carcinoma, immunohisto-
pressed p16, 0% overexpressed p53, and 3% expressed WT-1. chemistry, ovary, pathology, serous carcinoma
Only 10% of patients died of disease at last follow-up. Thirty- (Am J Surg Pathol 2016;40:302–312)
three (33) tumors, or 30% of tumors originally classified as
endometrioid, were reclassified as serous carcinoma (OSC). The
median patient age was 54.5 years, and 59% of patients were
younger than 60 years of age. Only 27% had disease confined to
the pelvis at presentation, 52% of tumors were unilateral, and
P rimary ovarian endometrioid carcinomas (OEC) rep-
resent approximately 10% of all ovarian carcinomas
and are the second or third most commonly encountered
the median tumor size was 8 cm. Associated squamous differ- primary ovarian malignancy.1–5 They are generally asso-
entiation, endometrioid adenofibroma, and endometrioid or ciated with a good prognosis, as the majority are low
mixed Mullerian borderline tumor (CEFs) were not present in grade, and patients often present with stage I or II dis-
any case, but 6% of patients had endometriosis. Approximately ease.6 In most instances, low-grade OEC can readily be
one half of the reclassified OSC demonstrated SET-pattern distinguished from other subtypes of ovarian carcinoma
morphology (combinations of glandular, cribriform, solid, and based solely on histopathologic features. Morphologic
features commonly associated with OEC (confirmatory
From the *Department of Pathology, National University Health System, endometrioid features, CEFs) include: squamous meta-
Singapore, Singapore; wDepartment of Pathology; and zMarie-Josée plasia, endometriosis, adenofibromatous background,
and Henry R. Kravis Center for Molecular Oncology, Memorial Sloan and associated borderline endometrioid component.1
Kettering Cancer Center, New York, NY.
D.L., R.M., and M.P.M. contributed equally. High-grade OECs are infrequently encountered, but they
Conflicts of Interest and Source of Funding: The authors have disclosed are a cause of significant interobserver disagreement, as
that they have no significant relationships with, or financial interest they may be confused with high-grade ovarian serous
in, any commercial companies pertaining to this article. carcinoma (OSC) or mixed epithelial carcinoma.6 Recent
Correspondence: Rajmohan Murali, MBBS, MD, FRCPA, Department
of Pathology, Memorial Sloan Kettering Cancer Center, 1275 York
molecular and protein studies have shown that a pro-
Avenue, New York, NY 10065 (e-mail: MuraliR@mskcc.org). portion of tumors previously classified as “high-grade
Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. endometrioid carcinomas” display similar genetic and

302 | www.ajsp.com Am J Surg Pathol  Volume 40, Number 3, March 2016

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Am J Surg Pathol  Volume 40, Number 3, March 2016 Ovarian Endometrioid Carcinoma

FIGURE 1. Endometrioid carcinoma showing squamous FIGURE 2. High-grade serous carcinomas with SET features
morule and low-grade endometrioid glandular differentiation. (originally diagnosed as FIGO grade 3 endometrioid carcino-
ma) showing solid architecture.

immunophenotypic profiles to high-grade OSC,6 and a

subset of these tumors demonstrate solid, pseudoendo- of OEC, treated between 1998 and 2010 with available
metrioid, and/or transitional cell–like growth patterns slides in the department archives. The cases included both
(SET pattern)7,8 that are increasingly thought to be part in-house surgical resections for which both glass slides and
of the spectrum of OSC, particularly in the presence of a blocks were available and consultation cases that included
BRCA1 mutation.8 This highlights the diagnostic chal- only glass slides.
lenges faced by many pathologists in trying to subtype All available hematoxylin and eosin–stained slides
these tumors using morphologic criteria. (median 10 slides per case, range 1 to 36) were reviewed
Using modern diagnostic criteria we reevaluated a independently by up to 4 gynecologic pathologists who
large series of unselected ovarian carcinomas classified as were blinded to the initial diagnoses and histologic reports,
endometrioid to determine the prevalence of high-grade imunohistochemical findings, and clinical outcomes. Sec-
endometrioid carcinoma and misclassification as high- tions of fallopian tube were examined if present, but none
grade endometrioid carcinoma. To do so, we synthesized of the cases were submitted for “sectioning and extensively
a variety of morphologic, immunohistochemical, muta- examining the fimbriated end” (SEE-FIM).
tional, and clinical data. On the basis of this, we make The 2014 World Health Organization histologic
recommendations about the application of morphologic criteria, briefly outlined here, were largely used in tumor
and immunohistochemical features that should allow re- classification.6 We also paid particular attention to the
producible and clinically meaningful diagnosis and clas- presence of morphologic features commonly associated
sification of ovarian carcinomas. with OEC (Fig. 1). These CEFs were: (i) squamous
metaplasia—presence of keratinization, intercellular
MATERIALS AND METHODS bridges, or solid growth of cells with abundant dense eo-
sinophilic cytoplasm and distinct cellular borders, present
Case Selection and Review either at the stromal interface or as morules separating
The study was approved by the institutional review adjacent glands (Fig. 1); (ii) endometriosis—presence of
board of Memorial Sloan Kettering Cancer Center. In- endometrial-type glands and/or stroma, often associated
stitutional databases were queried for cases with a diagnosis with evidence of hemorrhage; (iii) adenofibromatous

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.ajsp.com | 303

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Lim et al Am J Surg Pathol  Volume 40, Number 3, March 2016

FIGURE 3. High-grade serous carcinomas with SET features FIGURE 4. High-grade serous carcinomas with SET features
(originally diagnosed as FIGO grade 3 endometrioid carcino- (originally diagnosed as FIGO grade 3 endometrioid carcino-
ma) showing a pseudoendometrioid growth pattern. ma) showing transitional cell–like architecture.

background—presence of tubular and cystic endometrial- SETs are cytologically high-grade tumors that form
type glands surrounded by prominent fibrous stroma; and glands and cribriform structures, with or without solid
(iv) borderline endometrioid or mixed Mullerian architecture, with or without blunt papillae reminiscent of
component—this may exhibit 2 growth patterns, either transitional cell carcinoma. Many tumors have high
adenofibomatous or intracystic, and is composed of numbers of tumor-infiltrating lymphocytes, but these were
glandular proliferation similar to that observed in atypical not required for diagnosis as OSC-SET in this study.
hyperplasia of the endometrium.6 Tumors lacking CEFs Importantly, the presence of any CEF other than endo-
were still classified as endometrioid if the tumor contained metriosis was considered exclusionary, as was the presence
components that were histologically identical to the tu- of any component that resembled the tubules of low-grade
bules of low-grade endometrioid carcinoma of endome- endometrioid adenocarcinoma.
trium (Fig. 1) (ie, the tubules are lined by stratified tall Tumors were graded using the International Feder-
columnar cells with eosinophilic cytoplasm).9 ation of Gynecology and Obstetrics (FIGO) grading sys-
Tumors that were classified as “nonendometrioid” tem,6 identical to the grading system used for endometrioid
were further studied to specify tumor type and determine carcinomas of the uterine corpus. Tumors were also as-
the prevalence of SET-pattern high-grade serous carci- signed a Shimizu-Silverberg grade, which makes use of
noma (OSC-SET),8 a known mimic of high-grade endo- tumor architecture, nuclear grade, and mitotic index.10
metrioid and transitional cell carcinomas. OSC-SETs are Clinical and follow-up information was obtained
characterized by 1 or more of the following architectural from the patients’ electronic medical records. Available
patterns: solid, cribriform (pseudoendometrioid), or paraffin-embedded blocks from tumors resected at Me-
transitional cell carcinoma-like.8One pathologist (R.A.S.) morial Sloan Kettering Cancer Center were used for the
examined all available slides from the nonendometrioid construction of a tissue microarray (TMA). Representa-
carcinomas and subclassified them further as follows: tive tumor areas were annotated by a gynecologic patho-
OSC-SET (Figs. 2–5); high-grade papillary (conventional) logist, and three 0.6 mm cores from different areas of each
serous carcinoma (OSC-P); and other/unclassified. OSC- tumor were sampled for the TMA (Manual Tissue

304 | www.ajsp.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Am J Surg Pathol  Volume 40, Number 3, March 2016 Ovarian Endometrioid Carcinoma

FIGURE 5. High-grade serous carcinomas with SET features FIGURE 6. Low-grade endometrioid carcinoma exhibiting
(originally diagnosed as FIGO grade 3 endometrioid carcino- papillary architecture.
ma) showing numerous tumor-infiltrating lymphocytes.

Arrayer MTA-1; Beecher Instruments, Sun Prairie, WI). RESULTS

Immunohistochemical studies were preformed on sections Hematoxylin and eosin slides were available for 118
from the TMA using commercially available antibodies patients, whose tumors were diagnosed over a 12-year
for WT-1 (clone C-19; Santa Cruz Biotechnology Inc., period spanning 1998 to 2010. Nine cases were excluded
Dallas, TX), p16 (INK4A; Ventana Medical Systems Inc., from analysis; of these, 6 were associated with a con-
Tucson, AZ), and p53 (clone DO-7, Ventana Medical current uterine endometrioid tumor, and it was unclear
Systems Inc.). WT-1 was considered positive if there was whether the ovarian tumors represented metastatic or
diffuse reactivity throughout the tumor. p16 was consid- synchronous disease; 1 was reclassified as an endome-
ered positive if >50% of the tumor was immunoreactive trioid borderline tumor; and 2 were poorly differentiated
without intervening single or aggregated cells lacking tumors that were not classifiable.
staining. Aberrant p53 expression was defined as nuclear
positivity in >50% of tumor cells or complete absence of
staining in the presence of a positive internal control. Endometrioid Carcinomas
Seventy-six (76) tumors were classified as OEC
(Figs. 1, 6, 7), representing 70% of tumors originally di-
agnosed as endometrioid. The median patient age was
Statistical Analysis 55.0 years, with a range of 25 to 81 years; 75% of patients
Statistical analysis was performed using PASW SPSS were younger than 60 years of age. Sixty-seven of 73
18.0 software (IBM Corporation, Armonk, NY). Fisher staged patients (92%) presented with disease confined to
exact or w2 tests were used to test associations of histologic the pelvis (53 stage I and 14 stage II), whereas only 6
type with nominal and ordinal covariates, and non- patients had extrapelvic disease at presentation (5 stage
parametric tests were used for comparison of medians for III and 1 stage IV). Three patients were unstaged. Eighty-
continuous covariates. All statistical tests were 2-tailed, and seven percent of tumors were unilateral. The median
a P-value of <0.05 was considered statistically significant. tumor size was 11.8 cm, with a range of 0.4 to 30 cm.

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.ajsp.com | 305

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Lim et al Am J Surg Pathol  Volume 40, Number 3, March 2016

TABLE 1. Associations of Tumor Type With Clinicopathologic

Features
Endometrioid Nonendometrioid P
N
109 76 33
Age
< 60 y 60 20 0.046
Z60 y 16 13
Age (median, range) 55 (25-81) 54.5 (37-80) 0.22
Tumor size (median, range) 11.8 (0.4-30) 8 (3.5-18.5) 0.02
pT stage
I 53 6 < 0.001
IA 24 5
IB 2 0
IC 27 1
II 14 3
IIA 4 0
IIB 3 0
IIC 7 3
III 5 20
IIIA 1 3
IIIB 0 0
IIIC 4 17
IV 1 4
Laterality
Unilateral 66 17 < 0.001
Bilateral 10 16
Status at last follow-up
NED 59 12 < 0.001
AWD 3 3
Dead 7 16
Architectural pattern
Glandular 56 5 < 0.001
Papillary 12 6
FIGURE 7. Low-grade endometrioid carcinoma exhibiting a Solid 7 17
sex cord-like pattern. Mitotic rate (per 10 hpf) 1 (1-24) 7.5 (1-96) < 0.001
(median, range)
FIGO grade
1 54 0 < 0.001
2 18 8
Microscopically, 54 (71%) were FIGO grade 1, 18 3 4 25
(24%) were grade 2, and 4 (5%) were grade 3. Using the Shimizu-Silverberg grade
Shimizu-Silverberg system, 64 (84%) were grade 1, 9 1 64 3 < 0.001
2 9 6
(12%) were grade 2, and 2 (3%) were grade 3. Therefore, 3 2 20
between 84% and 97% of OECs are “low-grade,” with Squamous differentiation
the higher value grouping together grade 1 and grade 2 Absent 26 33 < 0.001
tumors. The median mitotic index was 1 per 10 high- Present 50 0
Endometriosis
power fields, with a range of 1 to 24. Squamous differ- Absent 39 31 < 0.001
entiation was present in 50 cases (76%), endometriosis in Present 37 2
37 cases (49%), endometrioid or mixed Mullerian bor- Adenofibroma
derline tumor in 32 cases (42%), endometrioid adenofi- Absent 60 33 0.004
broma in 16 cases (21%), and a synchronous endometrial Present 16 0
Borderline tumor
endometrioid carcinoma in 28 cases (37%) (Table 1). Absent 44 33 < 0.001
Eighty percent of cases had at least 1 CEF, and 59% had Present 32 0
at least 2 CEFs (Table 2). All cases contained histologi- No. CEFs
cally low-grade gland-forming endometrioid components. 0 15 31 < 0.001
1 16 2
For a diagnosis of OEC, the presence of any CEF yielded 2 24 0
sensitivity, specificity, positive predictive values, and neg- 3 13 0
ative predictive values of 80%, 94%, 97%, and 67%, re- 4 8 0
spectively, whereas the presence of >1 CEF yielded values Concurrent endometrial tumor
of 59%, 100%, 100%, and 59%, respectively (Table 3). Absent 46 29 < 0.001
Present 28 1
Immunohistochemistry was attempted in 31 tumors P16
with available material. Three of 31 cases (11%) overex- Negative 28 4 0.003
pressed p16, 0 of 30 cases (0%) overexpressed p53, and 1 of Positive 3 6

306 | www.ajsp.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Am J Surg Pathol  Volume 40, Number 3, March 2016 Ovarian Endometrioid Carcinoma

TABLE 1. (continued) TABLE 2. Associations of Clinicopathologic Features With

Endometrioid Nonendometrioid P Presence or Absence of CEFs
Absent Present P
P53
Negative 30 5 < 0.001 N
Positive 0 5 109 46 63
WT-1 Tumor classification
Endometrioid 15 61 < 0.001
Negative 29 1 < 0.001
Nonendometrioid 31 2
Positive 1 9 Age
< 60 y 26 54 0.001
AWD indicates alive with disease; hpf, high-power field; NED, no evidence of
Z60 y 20 9
disease.
Age (median, range) 62.5 (37- 51 (25-78) 0.003
81)
Tumor size, cm (median, range) 9 (3.5-18.5) 11.2 (0.4- 0.23
30)
pT stage
I 14 45 < 0.001
30 cases (3%) expressed WT-1 (Table 1). After a median II 7 10
follow-up period of 65 months (range, 6 to 180 mo), 59 pa- III 21 4
tients (86%) had no evidence of disease, 3 (4%) were alive IV 3 2
Laterality
with disease, and 7 (10%) had died of disease (Table 1). Unilateral 30 53 0.04
Bilateral 16 10
Status at last follow-up
High-grade Endometrioid Carcinomas NED 23 48 0.002
Only 2 of 76 endometrioid carcinomas (Figs. 8, 9), AWD 3 3
details of which are presented in Table 4, were classified Dead 17 6
Architectural pattern
as Shimizu-Silverberg grade 3 or high-grade. Glandular 15 46 < 0.001
Papillary 9 9
Nonendometrioid Carcinomas Solid 17 7
Nuclear grade
Thirty-three (33) tumors were classified as non- 1 1 14 < 0.001
endometrioid carcinoma, representing 30% of tumors 2 15 44
3 25 4
originally diagnosed as endometrioid. The median patient Mitotic rate (per 10 hpf) (median, 3 (1-96) 1 (1-17) < 0.001
age was 54.5 years, with a range of 37 to 80 years; 59% of range)
patients were younger than 60 years of age. Just 9 staged FIGO grade
1 8 46 < 0.001
patients (27%) presented with disease confined to the 2 13 13
pelvis (6 stage I and 3 stage II), whereas 24 patients (73%) 3 25 4
Shimizu-Silverberg grade
had extrapelvic disease at presentation (20 stage III and 4 1 13 54 < 0.001
stage IV). Fifty-two percent of tumors were unilateral, 2 9 6
and 48% were bilateral. The median tumor size was 8 cm, 3 20 2
Squamous differentiation
with a range of 3.5 to 18.5 cm (Table 1). Absent 46 13 < 0.001
Microscopically, none (0%) was FIGO grade 1, 8 Present 0 50
(24%) were grade 2, and 25 (76%) were grade 3. Using Endometriosis
Absent 46 24 < 0.001
the Shimizu-Silverberg system, 3 (10%) were grade 1, 6 Present 0 39
(21%) were grade 2, and 20 (69%) were grade 3. The Adenofibroma
Absent 46 47 < 0.001
median mitotic index was 7.5 per 10 high-power fields, Present 0 16
with a range of 1 to 96. Squamous differentiation, endo- Borderline tumor
metrioid adenofibroma, and endometrioid or mixed Absent 46 31 < 0.001
Present 0 32
Mullerian borderline tumor were not present in any case; Concurrent endometrial tumor
2 patients (6%) had endometriosis. Six of 10 cases (60%) Absent 34 41 0.10
overexpressed p16, 5 of 10 cases (50%) overexpressed Present 8 21
P16
p53, and 9 of 11 cases (82%) expressed WT-1 (Table 1). Negative 12 20 0.45
After a median follow-up duration of 80 months Positive 5 4
P53
(range, 18 to 158 mo), 12 patients (38%) had no evidence Negative 13 22 0.14
of disease, 3 (10%) were alive with disease, and 16 (52%) Positive 4 1
had died of disease. WT-1
Negative 8 22 0.001
Positive 9 1
Nonendometrioid Subsets AWD indicates alive with disease; hpf, high-power field; NED, no evidence of
One pathologist (R.A.S.) examined all available slides disease.
from 21 nonendometrioid carcinomas, which were sub-
classified further as follows: OSC-P (n = 8); OSC-SET
(n = 11) (Figs. 2–5) and other/unclassified (n = 2) (Table 5). SET, although there were twice as many patients under 60
There were no statistically significant clinical or im- years of age in the OSC-SET group and more than twice as
munohistochemical differences between OSC-P and OSC- many OSC-SET patients with bilateral tumors than patients

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.ajsp.com | 307

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Lim et al Am J Surg Pathol  Volume 40, Number 3, March 2016

with OSC-P. However, the numbers in these subgroups were

small (Table 5).

Comparison of Endometrioid and Serous

Carcinomas Originally Diagnosed as
Endometrioid Carcinomas
Compared with OSC patients, OEC patients more
frequently presented under age 60 (P = 0.046), had low-
stage tumors (P < 0.001), were more frequently unilateral
(P < 0.001), more frequently had synchronous endometrial
endometrioid carcinomas (P < 0.001), and had no evidence
of disease at last follow-up (P < 0.001). Their tumors were
of lower grade (P < 0.001), had more CEFs (P < 0.001),
and less frequently overexpressed p16 and p53 (P = 0.003
and P < 0.001, respectively) (Figs. 10–12) and less
frequently expressed WT-1 (P < 0.001) (Figs. 13–15)
(Table 1). Because of the small numbers of high-grade
OECs in the cohort (Table 5), it was not meaningful to
statistically compare them with OSCs.

DISCUSSION
Our criteria identified 2 distinct groups of patients,
30% of whose tumors were reclassified as OSC despite
originally being diagnosed as OEC; immunophenotyping
validated this morphology-based classification. Approx-
imately one half of OSCs exhibited a SET pattern,8 a
newly described pattern that has historically been con-
sidered to represent high-grade endometrioid or transi- FIGURE 8. High-grade (FIGO grade 3) endometrioid carci-
tional cell carcinoma. Clinicopathologic and immuno- noma with a predominantly solid growth pattern.
histochemical comparison with papillary serous carcinomas
confirms that these are serous carcinoma variants, not high-
grade endometrioid carcinomas.8 Patients with OSC-SET and positive predictive values but only modest sensitivity
showed a trend toward younger ages and more frequent and negative predictive values.
bilaterality of their tumors than those with OSC-P.7 Immunohistochemical markers may serve as useful
In this analysis, we emphasize the diagnostic value adjuncts to the classification of ovarian carcinomas.11,12
of CEFs and the presence of a low-grade gland-forming The immunoprofile of OEC is broadly similar to its ute-
endometrioid component as central to a diagnosis of rine counterparts and is characterized by frequent im-
OEC. Glandular and/or cribriform architecture alone is munoreactivity for estrogen receptor and progesterone
not a valid criterion, as these architectural features were receptor and absent or focal positivity for WT-1 and p16
found in nearly every case examined. However, 80% of as well as “wild-type” p53 expression.6 Low-grade tu-
OECs have at least 1 CEF, compared with 23% of OSCs mors, particularly those with squamous morules can also
(which represented only 2 patients, both of whom had exhibit nuclear b-catenin expression.13–17 In contrast, al-
endometriosis). For a diagnosis of OEC, the presence of though estrogen and progesterone receptors are also
any CEF yields high sensitivity, specificity, and positive commonly expressed in OSC, the majority demonstrates
predictive values and modest negative predictive values, diffuse WT-1 and p16 positivity as well as aberrant p53
whereas the presence of >1 CEF yields perfect specificity expression.18–21 In a study of ovarian carcinomas by

TABLE 3. Number of CEFs and Tumor Classification

CEFs Endometrioid Nonendometrioid P Sn (%) Sp (%) PPV (%) NPV (%)
0-1 CEFs 31 45 < 0.001 59 100 100 59
2-4 CEFs 45 0
Absent 15 31 < 0.001 80 94 97 67
Present 61 2
NPV indicates negative predictive value (uncertain cases excluded from these calculations); PPV, positive predictive value; Sn, sensitivity; Sp, specificity.

308 | www.ajsp.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Am J Surg Pathol  Volume 40, Number 3, March 2016 Ovarian Endometrioid Carcinoma

TABLE 5. Serous Carcinomas—Comparison of Conventional

and SET-associated Tumors
OSC-P (n [%]) OSC-SET (n [%]) P
N 8 (42) 11 (58)
Age
< 60 y 3 (33) 6 (67) 0.65
> 60 y 5 (50) 5 (50)
Stage
I 2 (67) 1 (33) 0.46
II 0 (0) 1 (100)
III 4 (33) 8 (67)
IV 2 (67) 1 (33)
Laterality
Unilateral 5 (62) 3 (38) 0.18
Bilateral 3 (27) 8 (73)
P16 IHC
Negative 2 (50) 2 (50) 1.00
Positive 3 (50) 3 (50)
P53 IHC
Negative 3 (60) 2 (40) 1.00
Positive 2 (40) 3 (60)
WT-1 IHC
Negative 1 (100) 0 (0) 1.00
Positive 4 (44) 5 (56)
IHC indicates immunohistochemistry.

Madore et al23 identified WT-1 positivity in 31% (10/32)

of grade 1 and grade 2 tumors. In our cohort, only rare
OECs overexpressed p16, 1 was WT-1 positive and none

FIGURE 9. High-grade (FIGO grade 3) endometrioid carci-

noma with focal low-grade endometrioid glandular differ-
entiation.

Gilks et al,22 50 cases originally diagnosed as OEC were

reclassified as OSC using morphologic criteria. These
tumors were predominantly high-grade (16 grade 2 tu-
mors and 33 grade 3 tumors), and 68% of these tumors
were WT-1 positive, a result comparable to that observed
in their cohort of OSC (78%) but significantly different
from OEC (4%). Although studies have shown that WT-
1-positive OECs tend to be high grade, immunoreactivity
in low-grade tumors has been reported in the literature.

TABLE 4. Characteristics of 2 High-grade Endometrioid

Carcinomas
Case 1 Case 2
Age 59 67
Stage IIC IIIC
Laterality Unilateral Unilateral
CEFs — —
P16 IHC — NA
P53 IHC — NA
WT-1 IHC — NA
Follow-up status No evidence of disease No evidence of disease
(duration) (173 mo) (135 mo)
IHC indicates immunohistochemistry. FIGURE 10. Immunohistochemistry—p53 expression in en-
dometrioid carcinoma.

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.ajsp.com | 309

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Lim et al Am J Surg Pathol  Volume 40, Number 3, March 2016

FIGURE 11. Immunohistochemistry—p53 expression in OSC-SET. FIGURE 12. Null pattern of p53 expression in an OSC-SET.
Note the rare non-neoplastic cells that display weak p53 ex-
pression; this serves as a positive internal control.

overexpressed p53, confirming our morphologic classi- tigators to hypothesize that OEC may arise through a dual
fication scheme with immunopositivity rates that are in pathogenic pathway or that some “high-grade OEC” may
line with most previously published values. in fact represent misclassified high-grade OSC,14,23 a
Given these data, one can make the following gen- conclusion that our data support. It has been shown that
eralizations about OEC: they typically present as large, the distinction of high-grade OECs from OSC is often
unilateral, low-grade ovarian tumors, confined to the associated with significant interobserver disagreement.25,26
pelvis, with excellent clinical outcomes. This information, However, there is immunohistochemical and molecular
particularly the size and laterality information, can be evidence to support the existence of a true high-grade
used as adjuncts for differential diagnosis, not only to OEC.14,23 Madore et al23 identified 2 high-grade, high-
separate OECs and OSCs, but also to distinguish primary stage, WT-1-negative OECs with concurrent b-catenin and
OECs from metastatic endometrioid carcinomas. Only TP53 aberrations. One of the tumors was spontaneously
2% of tumors studied were high-grade OEC. Therefore a immortalized in cell culture and gave rise to an aggressive
diagnosis of high-grade OEC should be approached with endometrioid carcinoma cell line. Nevertheless, the true in-
caution, particularly when SET patterns are present and cidence of high-grade OEC is likely to be substantially lower
CEFs are lacking. Gene expression studies have revealed than that reported in earlier literature.4,27 Unfortunately, we
that some high-grade OECs demonstrate expression pro- did not have genotype data on any of the cases in our co-
files similar to that of high-grade OSC.14,24 In a study of hort, but it is clear from our findings and the literature that
29 OECs (10 grade 1, 11 grade 2, and 8 grade 3), Geyer further research on high-grade OECs is required.
et al14 observed p53 overexpression in 6 grade 3 tumors The distinction between high-grade OEC and OSC
categorized as OECs with the majority lacking molecular can have important consequences. For example, in-
alterations characteristic of lower-grade tumors, such as b- correctly classifying a high-grade OSC as OEC might
catenin and KRAS mutations. This difference in biology deprive the patient of the opportunity to be tested for
between low-grade and high-grade OECs is also reflected germline BRCA1/BRCA2 mutations, which may have
in their clinical behavior; high-grade tumors tend to be several downstream repercussions, including (i) affecting
associated with significantly poorer clinical outcomes than the choice of adjuvant chemotherapeutic agents such as
low-grade tumors.14 These findings have led some inves- PARP inhibitors28,29; (ii) prognostic assessment; and (iii)

310 | www.ajsp.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Am J Surg Pathol  Volume 40, Number 3, March 2016 Ovarian Endometrioid Carcinoma

FIGURE 13. Immunohistochemistry—WT-1 expression in en- FIGURE 14. Immunohistochemistry—WT-1 expression in OSC-
dometrioid carcinoma. SET.

appropriate screening and surveillance of other family literature on the topic reports that OSC-SETs constitute a
members. Accurate diagnosis is also important for variant of OSC with association with BRCA1 mutation or
the correct assignment of patients to appropriate clinical promoter methylation,7,8 presentation at younger age,
trials. better prognosis, and infrequent association with serous
Of the 30% of tumors in this cohort that were re- tubal intraepithelial carcinoma.7 Although these tumors
classified as OSCs, approximately 50% were OSC-SETs. represented approximately 50% of OSCs in this cohort
It is understandable that these tumors, specifically, were that was enriched for tumors with an endometrioid ap-
originally classified as high-grade OECs because of the pearance, this almost definitely overestimates the preva-
frequent combinations of glandular, cribriform, and solid lence of OSC-SET among unselected OSCs.
growth patterns. The OSC-Ps had combinations of pap- This study’s strengths include a large number of
illary and glandular architecture with high nuclear grade, relatively rare tumors with good clinical annotation and
making it difficult to understand why these tumors had follow-up, as well as sufficient numbers of reclassified
been classified as endometrioid; perhaps the long time- OSCs that revealed heterogeneity within this category.
frame over which the cases were originally diagnosed However, there were insufficient numbers of bona fide
(1998 to 2010) and subsequent refinements in diagnostic high-grade OECs for statistical comparison with tumors
criteria are factors that contributed to their initial classi- reclassified as OSC. Many cases were consultation cases
fication as OEC. Clinical and immunohistochemical lacking tissue blocks, which limited the number of cases
data reported herein support that OSC-SETs are un- studied with immunohistochemistry.
equivocally OSCs, not OECs. Immunohistochemistry can
be used as a diagnostic adjunct when a diagnosis of high-
grade OEC is under consideration, with WT-1 being the CONCLUSIONS
most robust discriminator; all tested OSC-SETs were This analysis emphasizes the diagnostic value of
WT-1 positive, as compared with only 3% of OECs. CEFs, the presence of a low-grade gland-forming endo-
Several recent papers have reported that SET-pat- metrioid component, and WT-1 negativity, as valid, re-
tern tumors belong to the family of OSC. The most recent producible, and clinically relevant criteria for a diagnosis

Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved. www.ajsp.com | 311

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.

Lim et al Am J Surg Pathol  Volume 40, Number 3, March 2016

10. Shimizu Y, Kamoi S, Amada S, et al. Toward the development of a

universal grading system for ovarian epithelial carcinoma: testing of
a proposed system in a series of 461 patients with uniform treatment
and follow-up. Cancer. 1998;82:893–901.
11. Darvishian F, Hummer AJ, Thaler HT, et al. Serous endometrial
cancers that mimic endometrioid adenocarcinomas: a clinicopatho-
logic and immunohistochemical study of a group of problematic
cases. Am J Surg Pathol. 2004;28:1568–1578.
12. Yemelyanova A, Ji H, Shih IeM, et al. Utility of p16 expression for
distinction of uterine serous carcinomas from endometrial endome-
trioid and endocervical adenocarcinomas: immunohistochemical
analysis of 201 cases. Am J Surg Pathol. 2009;33:1504–1514.
13. Catasus L, Bussaglia E, Rodrguez I, et al. Molecular genetic
alterations in endometrioid carcinomas of the ovary: similar
frequency of beta-catenin abnormalities but lower rate of micro-
satellite instability and PTEN alterations than in uterine endome-
trioid carcinomas. Hum Pathol. 2004;35:1360–1368.
14. Geyer JT, Lopez-Garcia MA, Sanchez-Estevez C, et al. Pathoge-
netic pathways in ovarian endometrioid adenocarcinoma: a molec-
ular study of 29 cases. Am J Surg Pathol. 2009;33:1157–1163.
15. Irving JA, Catasus L, Gallardo A, et al. Synchronous endometrioid
carcinomas of the uterine corpus and ovary: alterations in the beta-
catenin (CTNNB1) pathway are associated with independent
primary tumors and favorable prognosis. Hum Pathol. 2005;36:
605–619.
16. Moreno-Bueno G, Gamallo C, Perez-Gallego L, et al. beta-Catenin
expression pattern, beta-catenin gene mutations, and microsatellite
instability in endometrioid ovarian carcinomas and synchronous
endometrial carcinomas. Diagn Mol Pathol. 2001;10:116–122.
17. Oliva E, Sarrio D, Brachtel EF, et al. High frequency of beta-catenin
mutations in borderline endometrioid tumours of the ovary.
J Pathol. 2006;208:708–713.
18. Acs G, Pasha T, Zhang PJ. WT1 is differentially expressed in serous,
endometrioid, clear cell, and mucinous carcinomas of the perito-
neum, fallopian tube, ovary, and endometrium. Int J Gynecol
FIGURE 15. Rare WT-1-positive cells in a low-grade endo- Pathol. 2004;23:110–118.
metrioid carcinoma. 19. Hirschowitz L, Ganesan R, McCluggage WG. WT1, p53 and
hormone receptor expression in uterine serous carcinoma. Histo-
pathology. 2009;55:478–482.
of OEC. Glandular and/or cribriform architecture alone 20. Nofech-Mozes S, Khalifa MA, Ismiil N, et al. Immunophenotyping
may be seen in both OECs and OSCs and are therefore of serous carcinoma of the female genital tract. Mod Pathol.
not informative of diagnosis. Further study is needed to 2008;21:1147–1155.
21. Waldstrom M, Grove A. Immunohistochemical expression of wilms
elaborate the characteristics of the exceedingly rare high- tumor gene protein in different histologic subtypes of ovarian
grade endometrioid carcinomas of the ovary. carcinomas. Arch Pathol Lab Med. 2005;129:85–88.
22. Gilks CB, Ionescu DN, Kalloger SE, et al. Tumor cell type can be
REFERENCES reproducibly diagnosed and is of independent prognostic signifi-
1. Cho KR. Ovarian cancer update: lessons from morphology, cance in patients with maximally debulked ovarian carcinoma. Hum
molecules, and mice. Arch Pathol Lab Med. 2009;133:1775–1781. Pathol. 2008;39:1239–1251.
2. Long ME, Taylor HC Jr. Endometrioid carcinoma of the ovary. 23. Madore J, Ren F, Filali-Mouhim A, et al. Characterization of the
Am J Obstet Gynecol. 1964;90:936–950. molecular differences between ovarian endometrioid carcinoma and
3. Gray LA, Barnes ML. Endometrioid carcinoma of the ovary. Obstet ovarian serous carcinoma. J Pathol. 2010;220:392–400.
Gynecol. 1967;29:694–701. 24. Wu R, Hendrix-Lucas N, Kuick R, et al. Mouse model of human
4. Kline RC, Wharton JT, Atkinson EN, et al. Endometrioid ovarian endometrioid adenocarcinoma based on somatic defects in
carcinoma of the ovary: retrospective review of 145 cases. Gynecol the Wnt/beta-catenin and PI3K/Pten signaling pathways. Cancer
Oncol. 1990;39:337–346. Cell. 2007;11:321–333.
5. Kurman RJ, Craig JM. Endometrioid and clear cell carcinoma of 25. Kobel M, Kalloger SE, Baker PM, et al. Diagnosis of ovarian
the ovary. Cancer. 1972;29:1653–1664. carcinoma cell type is highly reproducible: a transcanadian study.
6. Kurman RJ, Carcangiu ML, Herrington CS, et al. WHO Am J Surg Pathol. 2010;34:984–993.
Classification of Tumours of Female Reproductive Organs. Lyon: 26. Stalsberg H, Abeler V, Blom GP, et al. Observer variation in
IARC Press; 2014. histologic classification of malignant and borderline ovarian tumors.
7. Howitt BE, Hanamornroongruang S, Lin DI, et al. Evidence for a Hum Pathol. 1988;19:1030–1035.
dualistic model of high-grade serous carcinoma: BRCA mutation 27. DePriest PD, Banks ER, Powell DE, et al. Endometrioid carcinoma
status, histology, and tubal intraepithelial carcinoma. Am J Surg of the ovary and endometriosis: the association in postmenopausal
Pathol. 2015;39:287–293. women. Gynecol Oncol. 1992;47:71–75.
8. Soslow RA, Han G, Park KJ, et al. Morphologic patterns associated 28. Bryant HE, Schultz N, Thomas HD, et al. Specific killing of
with BRCA1 and BRCA2 genotype in ovarian carcinoma. Mod BRCA2-deficient tumours with inhibitors of poly(ADP-ribose)
Pathol. 2012;25:625–636. polymerase. Nature. 2005;434:913–917.
9. Hoang LN, McConechy MK, Kobel M, et al. Histotype-genotype 29. Schreiber V, Dantzer F, Ame JC, et al. Poly(ADP-ribose): novel
correlation in 36 high-grade endometrial carcinomas. Am J Surg functions for an old molecule. Nat Rev Mol Cell Biol. 2006;7:
Pathol. 2013;37:1421–1432. 517–528.

312 | www.ajsp.com Copyright r 2015 Wolters Kluwer Health, Inc. All rights reserved.

Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.