Endometrial cancer is the most common invasive neoplasm of the female genital tract and the fourth most common cancer in women in the United States 1. Over the last 30 yr, endometrial tumors have broadly been divided into 2 main categories, type I and type II 1–7. Historically, endometrial tumors have primarily been classified on the basis of their histology alone. However, as our appreciation for the varied morphology and prognoses of these neoplasms deepens, ancillary studies are increasingly used for routine diagnoses, particularly in cases where there is morphologic ambiguity or tumor heterogeneity.
Mixed endometrial carcinomas represent a particularly interesting group of endometrial tumors. Defined as a combination of 2 or more distinct histologic subtypes, one component of which must be a type II tumor comprising at least 5% of the tumor volume, they have been a source of both great interest and diagnostic difficulty 8. Recent estimates show that mixed endometrial carcinomas account for anywhere between 3% and 10% of all endometrial carcinomas 8, although this is likely variable between institutions given their morphologic complexity. In the literature, the most commonly observed histologic combination is serous carcinoma (SC) and endometrioid carcinoma (EC) followed by mixed EC and clear cell carcinoma (CC) 9,10.
Given the increased recognition of morphologic overlap between different morphologic types of endometrial cancer, routine use of immunohistochemistry has been used in many cases where there is a diagnostic question based on morphology alone. Through exhaustive efforts over the years, the immunohistochemical prototypes for SC, EC, and CC have become fairly well-established (Table 1) 1,8,11. Stains including p53, p16, Ki67, estrogen receptor (ER), and progesterone receptor (PR) have routinely been used in recent years to assist in the diagnosis of these tumors. More recently, Napsin A and HNF1B have been established as reliable markers of CC and their utility continues to be investigated 11–13.
Our goal was to evaluate the staining pattern of a panel of commonly used immunohistochemical markers in mixed endometrial carcinomas of classic morphology. Our aims were 2-fold: (1) to assess how reliable prototypical staining patterns are in mixed tumors, and (2) to assess the staining patterns within morphologically distinct regions of the same tumor. We sought to define the importance of these findings in the setting of routine clinical diagnosis and highlight issues the pathologist should be aware of when interpreting both biopsy and resection specimens.
MATERIALS AND METHODS
We identified 18 cases of mixed endometrial carcinomas from the surgical pathology archives at New York Presbyterian Hospital—Weill Cornell Medicine. Cases included 14 hysterectomy specimens, 13 of which also had a corresponding endometrial biopsy or curettage, and 4 endometrial curettage specimens available as formalin-fixed paraffin-embedded tissue. All slides were reviewed by 3 pathologists with expertise in Gynecologic Pathology (L.H.E., E.C.P., C.E.M.) to confirm the diagnoses. All cases included at least 2 histologic subtypes (EC, SC, and/or CC) each comprising at least 5% of the tumor burden. Cases were assessed for classic morphology of each component. Cases with ambiguous, undifferentiated, and malignant stromal morphology were excluded from the study. Mucinous tumors were also excluded.
Immunohistochemical staining for p53, p16, Ki67, ER, PR, and Napsin A was performed on selected blocks from each case using commercially available antibodies as detailed in Table 1. An average of 2 formalin-fixed paraffin-embedded tissue blocks were selected from each case and sections were cut at 5 µm. Staining was carried out on whole tissue sections using the Bond III Autostainer (Leica Microsystems, Illinois) (Table 2).
For p53, ER, PR, and Ki67, only nuclear positivity was considered contributory. Both cytoplasmic and nuclear reactivity were assessed with respect to p16 and cytoplasmic reactivity was assessed with respect to Napsin A. Intensity and extent of staining were evaluated on a 4-tiered (0, 1+, 2+, 3+) and a 5-tiered scale (0=null, 1=<1%–24%, 2=25%–49%, 3=50%–74%, 4=75%–100%) scale, respectively. Strong (3+) staining was considered to be the same intensity as strong staining seen in positive controls. Moderate (2+) staining was slightly weaker but still appreciated on low-power, weak (1+) was faint and required confirmation on higher power, and 0 was considered to be the absence of any staining. Extent of staining was based on the percentage of tumor cells staining. p53 staining was considered aberrant when >75% or 0% of tumor nuclei were positive, percentages highly likely to correlate with TP53 gene mutation 14,15.
The study was performed under an institutional review board—approved protocol.
Eighteen cases of mixed endometrial carcinomas were evaluated using the immunohistochemical panel as detailed above. The study included 9 cases of mixed EC/SC (2 mixed grade 1 EC/SC, 3 mixed grade 2 EC/SC, and 4 mixed grade 3 EC/SC), 4 mixed SC/CC, 2 mixed EC/CC (grade 3 and grade 2 EC), and 3 mixed SC/EC/CC (2 grade 1 and 1 grade 2 EC) as detailed in Table 3.
Patient age ranged from 52 to 95 yr (mean 69.6 yr). All patients were postmenopausal. The average age of the EC/SC group was 60.7 yr, SC/CC was 82.5 yr, EC/CC was 68.5 yr, and EC/SC/CC was 80.0 yr. Clinical stage ranged from stage 1A to 4B. Of the 13 cases with hysterectomy and corresponding endometrial sampling, 7 were diagnosed as mixed carcinomas on biopsy and 1 case was qualified as “suggestive of a mixed tumor.” Of the 5 cases in which a second component was not detected on biopsy, 2 showed the SC component of a mixed SC/EC tumor, 1 SC component of a mixed SC/CC tumor, 1 CC component of a mixed EC/CC carcinoma, and 1 case showed detection of only the EC component of an EC/SC tumor.
For cases comprised of SC and EC, the percentage of components ranged from 10% to 90% for each. In mixed SC/CC cases, the proportion of SC ranged from 40% to 80%. In mixed EC/CC cases, the EC proportion ranged from 60% to 70%. Triply mixed cases showed a SC component ranging from 25% to 50%, EC from 10% to 25%, and CC from 30% to 50%.
The typical immunostaining pattern expected for the individual tumor components was seen in only 3 cases, while in 15 cases an unexpected pattern was observed in at least one immunomarker (summarized in Table 3). Overall, by tumor type, the most common unexpected finding in EC/SC cases was diffuse positivity for p16 and/or ER/PR in both components; in SC/CC, diffuse positivity for p53 in both components was most frequently seen; and in SC/CC/EC, Napsin A negativity was most commonly observed. Table 4 outlines the immunohistochemical scoring for the individual tumor components in all cases and Table 5 highlights the most common unexpected results for each marker.
Results of p53 Immunostaining
There were 7 cases that demonstrated unexpected p53 staining patterns. Two of these were mixed EC/SC cases; 1 showed a null mutation pattern in both components and 1 showed wild-type pattern in both components (case #8, SC/EC, Fig. 1). Four cases were mixed SC/CC cases and 1 case was a SC/ CC/EC case, all of which showed p53 mutation pattern in the CC component. Interestingly, all clear cell components of mixed SC/CC cases showed diffuse p53 positivity while no mixed EC/CC cases showed diffuse p53 positivity. For example, case #11 (SC/CC, Fig. 2) showed diffuse p53 positivity in both components while the clear cell component was entirely negative for p16. In the EC/SC case with a null staining pattern, the EC component was FIGO grade 3.
Results of p16 Immunostaining
Nine cases showed unexpected p16 immunohistochemistry. Five of these cases included mixed EC/SC cases, with diffusely positive p16 staining in both components despite wild-type p53 staining in the EC component, and 2 cases were mixed SC/CC cases which showed p16 negativity in the CC. Two mixed SC/CC/EC showed patchy p16 positivity in all components, including the SC component. All unexpected diffuse p16 positivity in EC cases was only seen in cases that also had an SC component. For example, in case #1, both the SC and EC components showed diffuse p16 positivity. However, additional immunomarkers showed divergent staining patterns (Fig. 3).
Results of ER/PR Immunostaining
Unexpected hormone receptor staining included 5/16 cases (31%) with positive ER and/or PR staining in the SC component, 4/9 (44%) cases with positivity in the CC component and 2/14 cases (14%) with negative staining in the EC component. Most unexpected ER/PR positivity in SC was seen in association with an EC component which was also positive. EC cases with ER and/or PR negativity were seen in association with an SC component. Three of the 4 positive CC cases were associated with an EC component which was also positive. Figure 4 illustrates these findings in case #15.
Results of Napsin A Immunostaining
Eight cases showed unexpected Napsin A staining patterns, including 5 cases where the CC component was negative (3 SC/CC/EC cases, 1 SC/CC case, and 1 EC/CC case) and 3 cases in which the SC component was positive (2 EC/SC and 1 SC/CC). Overall, Napsin A showed unexpected staining patterns in both SC and CC, the most common aberration of which was CC negativity (5/9 cases; 56%).
Results of Ki67 Immunostaining
Ki67 positivity pattern did not significantly diverge from expected staining, with an average score (intensity×proportion) of 7.6 in SC components, 4.3 in EC components and 3.5 in CC components, respectively, although CC components had proliferative indices lower than expected.
To summarize the findings based on individual morphologic components, we have seen EC components with mutational p53 pattern in 1/14 (7%) cases, diffuse p16 in 5/14 (36%) cases, negative ER/PR in 2/14 (14%) cases, and positive Napsin A staining in 2/14 (14%) cases. SC components showed wild-type p53 pattern in 1/16 (6%) case, focal p16 in 2/16 (13%) cases, positive ER/PR in 4/16 (25%) cases and positive Napsin A in 2/16 (13%) cases. CC component showed mutational p53 pattern in 5/9 (56%) cases, negative p16 in 2/9 (22%) cases and positive ER/PR in 4/9 (44%) cases, and Napsin A negative in 5/9 (56%) cases.
Mixed endometrial carcinomas have long posed a diagnostic dilemma to pathologists. Not only is there morphologic heterogeneity within the individual components, the histologic overlap between subtypes and their frequently intermingled nature make them a unique and challenging entity. In recent years, promising data has come to light regarding the phenotypic and molecular landscape of these tumors 9,16,17. Our current project sought to further characterize the immunohistochemical staining patterns of histologically distinct mixed endometrial carcinomas and their potential clinical and pathologic implications. We found that despite displaying diagnostic morphology, many cases of mixed endometrial carcinomas demonstrate overlapping immunophenotypes. This makes this diagnosis particularly challenging, especially when only small portions of the tumor are represented on sampling.
Within our cohort, all markers showed unexpected staining patterns. Unexpected p16 or ER/PR staining was the most common abnormality, with 9 cases showing abnormal staining patterns in each group. Diffuse p16 positivity has been observed in a preponderance of both SC and ovarian high-grade serous carcinoma. The molecular mechanisms that cause diffuse expression are unclear, though association with Rb1 loss and inactivation has been suggested in a portion of these tumors 18,19. In this study, 36% of EC cases showed strong, diffuse positivity, a number notably higher than the 3% to 10% previously observed in the literature for EC alone 19–21. Interestingly, all unexpected diffuse p16 positivity in EC cases was seen only in cases that also harbored an SC component.
Previous studies have demonstrated the utility of ER/PR in distinguishing pure EC from pure SC, specifically regarding low-grade tumors (G1/G2 EC). It was shown that these stains become less reliable when dealing with high-grade (G3) endometrioid tumors 6,20,22. We also found overlapping ER/PR staining patterns in mixed cases, though our study did not show a distinction related to the grade of the EC tumor, unlike that seen in pure cases. In our mixed cases, both low-grade and high-grade endometrioid tumors showed unexpected staining patterns.
Uterine CCs are less common than their serous and endometrioid counterparts, and less is known about their histogenesis and immunophenotype, in part due to grouping in early endometrial studies 23,24. In addition, endometrioid and less commonly serous tumors can display areas of clear cell change that may be confused with CC, particularly to the untrained observer 23,25. In an attempt to avoid the challenging issue of clear cell change, we sought to limit the study to those cases that would be diagnosed as CC if seen in isolation. Of the 5 cases containing both EC and CC components, 4 showed foci of clearing in the EC component but no definitive secretory change, although this is admittedly a challenging distinction. We observed that our CC cases showed higher rates of mutant p53 positivity (56%) and ER/PR positivity (44%) than that previously observed in pure CC (37.5% and 0%–28%, respectively) 13,26–29. Mutation-associated p53 expression in CC was always seen in association with a serous component in this study.
We also observed a significantly lower percentage (44%) of CC components positive for Napsin A than that previously described in cases of pure clear cell endometrial carcinoma which has been cited in the literature ranging from 67% to 93% 13. Studies have shown Napsin A positivity in 6% to 10% of endometrioid carcinomas of the endometrium. Positivity is exceedingly rare in SCs 13, although we observed positivity in three cases (19%). Interestingly, case #8 also showed wild-type p53 staining pattern in both components, raising the possibility that perhaps the serous component was actually an eosinophilic variant of CC and not truly a SC, despite displaying classic morphologic features.
The vast majority of cases with aberrant immunophenotypes showed overlapping staining patterns with the other component(s) of the same tumor. As we learn more about the biology and behavior of endometrial carcinoma, misclassification becomes of even greater importance. Endometrioid and serous histologies are traditionally regarded as the hallmarks of the dualistic approach to endometrial cancer. Uterine CCs are most commonly grouped with type 2 tumors, as these patients present with more frequent extrauterine spread and decreased survival in comparison to their endometrioid counterparts, although this remains somewhat controversial 29–31. Mixed carcinomas with variant immunohistochemical staining patterns pose a higher risk of biopsy-resection specimen discordance. Given that alternate treatment strategies are used according to morphology, correct diagnosis is imperative, particularly in cases with an endometrioid versus serous/clear cell histology. Within our cohort of tumors, which were notably restricted to those with classical appearance to the best of our ability, 5 cases were not detected on biopsy, although this may be due to sampling, as not all biopsies were available for review. However, it is also prudent to point out that the majority of cases had a limited panel or no ancillary immunohistochemical stains performed at diagnosis. Our findings highlight the importance of remaining circumspect when reviewing small specimens of mixed histology, as they may not demonstrate staining patterns of their prototypes.
Discussion of these cases begs the question of where these tumors will fall within the changing landscape of endometrial tumors, particularly with respect to the newly proposed TCGA classification describing 4 molecular categories of endometrial carcinomas: (1) ultramutated [polymerase epsilon (POLE) mutated], (2) hypermutated (microsatellite instability), (3) copy number alterations—low, and (4) copy number alterations—high 32. A handful of studies have evaluated the reproducibility of these categories in clinical practice 33–35. Talhouk et al. 33 proposed a classification scheme utilizing immunohistochemistry for mismatch repair proteins and p53, as well as sequencing of the exonuclease domain of POLE (PRoMisE) to designate tumors according to the 4 molecular categories, a method that was later used by Hoang et al. 34, who found increased reproducibility using this molecular scheme. Although these papers illustrate the marked progress made in addressing issues of diagnostic reproducibility in endometrial cancer, mixed endometrial carcinomas may continue to present a unique challenge to this classification.
Overall, our data provides immunohistochemical support for the complex evolution of mixed carcinomas. To date, the literature has shown that mixed carcinomas include both true collision tumors, albeit rarely, as well as more commonly, tumors with divergent differentiation from a common precursor. Our findings show that it is not uncommon for immunohistochemical staining patterns to deviate from established prototypes in these tumors. These findings become particularly important in the setting of biopsy specimens, which may only contain small fragments of morphologically heterogeneous tumors, as misdiagnosis can have significant repercussions for the patient. Further molecular work would help discern the pathways by which these tumors develop and into what category they would be best placed in the continually developing landscape of endometrial cancer.
1. Kurman RJ, Ellenson L, Ronnett BM. Blaustein's Pathology of the Female Genital Tract Vol 7
. New York: Springer; 2011.
2. Lax SF, Kurman RJ. A dualistic model for endometrial carcinogenesis based on immunohistochemical and molecular genetic analyses. Verh Dtsch Ges Pathol 1997;81:228–32.
3. Lax SF, Kendall B, Tashiro H, et al. The frequency of p53, K-ras mutations, and microsatellite instability differs in uterine endometrioid and serous carcinoma: evidence of distinct molecular genetic pathways. Cancer 2000;88:814–24.
4. Sherman ME. Theories of endometrial carcinogenesis: a multidisciplinary approach. Mod Pathol 2000;13:295–308.
5. Alkushi A, Abdul-Rahman ZH, Lim P, et al. Description of a novel system for grading of endometrial carcinoma and comparison with existing grading systems. Am J Surg Pathol 2005;29:295–304.
6. Alkushi A, Clarke BA, Akbari M, et al. Identification of prognostically relevant and reproducible subsets of endometrial adenocarcinoma based on clustering analysis of immunostaining data. Mod Pathol 2007;20:1156–65.
7. Bokhman JV. Two pathogenetic types of endometrial carcinoma. Gynecol Oncol 1983;15:10–7.
8. Kurman RJ, Carcangiu ML, Herrington CS, et al. WHO Classification of Tumors of Female Reproductive Organs. Lyon: International Agency for Research on Cancer; 2014.
9. Köbel M, Meng B, Hoang LN, et al. Molecular analysis of mixed endometrial carcinomas shows clonality in most cases. Am J Surg Pathol 2015;40:166–80.
10. Quddus MR, Sung CJ, Zhang C, et al. Minor serous and clear cell components adversely affect prognosis in “mixed-type’’ endometrial carcinomas: a clinicopathologic study of 36 stage-I cases. Reprod Sci 2010;17:673–8.
11. Bartosch C, Manuel Lopes J, Oliva E. Endometrial carcinomas. Adv Anat Pathol 2011;18:415–37.
12. Lim D, Ip PPC, Cheung ANY, et al. Immunohistochemical comparison of ovarian and uterine endometrioid carcinoma, endometrioid carcinoma with clear cell change, and clear cell carcinoma. Am J Surg Pathol 2015;39:1061–9.
13. Iwamoto M, Nakatani Y, Fugo K, et al. Napsin A is frequently expressed in clear cell carcinoma of the ovary and endometrium. Hum Pathol 2015;46:957–62.
14. Yemelyanova A, Vang R, Kshirsagar M, et al. Immunohistochemical staining patterns of p53 can serve as a surrogate marker for TP53 mutations in ovarian carcinoma: an immunohistochemical and nucleotide sequencing analysis. Mod Pathol 2011;24:1248–53.
15. Cole AJ, Dwight T, Gill AJ, et al. Assessing mutant p53 in primary high-grade serous ovarian cancer using immunohistochemistry
and massively parallel sequencing. Sci Rep 2016;6:26191. Doi:10.1038/srep26191.
16. Espinosa I, D’Angelo E, Palacios J, et al. Mixed and ambiguous endometrial carcinomas: a heterogenous group of tumors with different clinicopathologic and molecular genetic features. Am J Surg Pathol 2016;40:972–81.
17. Lawrenson K, Pakzamir E, Liu B, et al. Molecular Analysis of Mixed Endometrioid and Serous Adenocarcinoma of the Endometrium. PLoS One 2015;10:e0130909.
18. Milea A, George SHL, Matevski D, et al. Retinoblastoma pathway deregulatory mechanisms determine clinical outcome in high-grade serous ovarian carcinoma. Mod Pathol 2014;27:991–1001.
19. Yemelyanova A, Ji H, Shih I-M, et al. Utility of p16 expression for distinction of uterine serous carcinomas from endometrial endometrioid and endocervical adenocarcinomas: immunohistochemical analysis of 201 cases. Am J Surg Pathol 2009;33:1504–14.
20. Reid-Nicholson M, Iyengar P, Hummer AJ, et al. Immunophenotypic diversity of endometrial adenocarcinomas: implications for differential diagnosis. Mod Pathol 2006;19:1091–100.
21. Chiesa-Vottero AG, Malpica A, Deavers MT, et al. Immunohistochemical overexpression of p16 and p53 in uterine serous carcinoma and ovarian high-grade serous carcinoma. Int J Gynecol Pathol 2007;26:328–33.
22. Alkushi A, Köbel M, Kalloger SE, et al. High-grade endometrial carcinoma: serous and grade 3 endometrioid carcinomas have different immunophenotypes and outcomes. Int J Gynecol Pathol 2010;29:343–50.
23. Offman SL, Longacre TA. Clear cell carcinoma of the female genital tract (not everything is as clear as it seems). Adv Anat Pathol 2012;19:296–312.
24. Fadare O, Zheng W, Crispens MA, et al. Morphologic and other clinicopathologic features of endometrial clear cell carcinoma: a comprehensive analysis of 50 rigorously classified cases. Am J Cancer Res 2013;3:70–95.
25. Gatius S, Matias-Guiu X. Practical issues in the diagnosis of serous carcinoma of the endometrium. Mod Pathol 2016;29 (suppl 1):S45–58.
26. Hoang LN, Han G, McConechy M, et al. Immunohistochemical characterization of prototypical endometrial clear cell carcinoma—diagnostic utility of HNF-1β and oestrogen receptor. Histopathology 2014;64:585–96.
27. Mhawech-Fauceglia P, Yan L, Liu S, et al. ER+ /PR+ /TFF3+ /IMP3− immunoprofile distinguishes endometrioid from serous and clear cell carcinomas of the endometrium: a study of 401 cases. Histopathology 2013;62:976–85.
28. Lax SF, Pizer ES, Ronnett BM, et al. Clear cell carcinoma of the endometrium is characterized by a distinctive profile of p53, Ki-67, estrogen, and progesterone receptor expression. Hum Pathol 1998;29:551–8.
29. Bae HS, Kim H, Young Kwon S, et al. Should endometrial clear cell carcinoma be classified as type II endometrial carcinoma? Int J Gynecol Pathol 2015;34:74–84.
30. Olawaiye AB, Boruta DM. Management of women with clear cell endometrial cancer: a Society of Gynecologic Oncology (SGO) review. Gynecol Oncol 2009;113:277–83.
31. McGunigal M, Liu J, Kalir T, et al. Survival differences among uterine papillary serous, clear cell and grade 3 endometrioid adenocarcinoma endometrial cancers: a National Cancer Database Analysis. Int J Gynecol Cancer 2017;27:85–92.
32. Getz G, Gabriel SB, Cibulskis K, et al. Integrated genomic characterization of endometrial carcinoma. Nature 2013;497:67–73.
33. Talhouk A, McConechy MK, Leung S, et al. A clinically applicable molecular-based classification for endometrial cancers. Br J Cancer 2015;113:299–310.
34. Hoang LN, Kinloch MA, Leo JM, et al. Interobserver agreement in endometrial carcinoma histotype diagnosis varies depending on The Cancer Genome Atlas (TCGA)-based molecular subgroup. Am J Surg Pathol 2017;41:245–52.
35. Stelloo E, Nout RA, Osse EM, et al. Improved risk assessment by integrating molecular and clinicopathological factors in early-stage endometrial cancer-combined analysis of the PORTEC Cohorts. Clin Cancer Res 2016;22:4215–24.
Keywords:©2018International Society of Gynecological Pathologists
Mixed endometrial carcinoma; Immunohistochemistry; Histology