- Email: [email protected]
Mucinous histology is a risk factor for nodal metastases in endometrial cancer
Gynecologic Oncology 125 (2012) 541–545
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Mucinous histology is a risk factor for nodal metastases in endometrial cancer Fernanda Musa a,⁎, Marilyn Huang b, Brandi Adams a, Edyta Pirog c, Kevin Holcomb d a
Department of Obstetrics & Gynecology, Weill Cornell Medical College, New York, NY, USA Division of Gynecologic Oncology, MD Anderson Cancer Center, Houston, TX, USA c Department of Pathology, Weill Cornell Medical College, New York, NY, USA d Division of Gynecologic Oncology, Weill Cornell Medical College, New York, NY, USA b
a r t i c l e
i n f o
Article history: Received 13 October 2011 Available online 9 March 2012 Keywords: Endometrial cancer Mucinous differentiation Mucinous adenocarcinoma of the endometrium lymph node metastases
a b s t r a c t Objectives. Mucinous adenocarcinoma of the endometrium (MUC) is a rare histological variant of endometrial carcinoma accounting for 1–9% of endometrioid tumors. Few studies have characterized its clinical behavior. This is a case–control study at a single institution comparing the risk factors and clinical course of MUC relative to endometrioid adenocarcinoma. Methods. A case–control study was performed including patients treated for endometrial cancer between 1996 and 2006. 41 cases of mucinous adenocarcinoma were identified. Each case was matched with two controls of endometrioid histology by age and histological grade. Cases and controls were compared with regard to known risk factors for endometrial cancer and the extent of disease at diagnosis. Chi-square tests were used to compare proportions and Student's t-tests for the comparison of means. Multivariate regression was used to identify the independent predictors of lymph node metastases. Overall survival was calculated using the Kaplan–Meier method and compared with the Log-rank test. p b .05 was considered significant for all tests. Results. Cases and controls were matched by age and FIGO grade and were found to be similar in regard to ethnicity, body mass index and medical history. No significant difference in myometrial invasion (MI) > 50% or the presence of lymph-vascular space invasion was found between cases and controls, however, 17% of patients with MUC had lymph node metastases compared to 3% of controls (p = .01). Multivariate analysis controlling for both tumor grade and depth of MI identified mucinous histology as an independent predictor of lymph node metastasis (p = .02). There was no difference in adjuvant treatment, recurrence rate or survival between the two groups. Conclusion. Mucinous differentiation was found to be an independent predictor of lymph node metastasis in the study population. Comprehensive surgical staging including retroperitoneal node dissection should be strongly considered in all endometrial cancer patients with predominantly mucinous histology. © 2012 Elsevier Inc. All rights reserved.
Introduction Endometrial cancer is the most common gynecologic malignancy in the United States with approximately 43,470 new cases diagnosed yearly and approximately 8000 annual deaths attributed to the disease [1]. Although the majority of cases are identified early and are associated with a favorable prognosis, there are several well established poor prognostic factors for endometrial cancer which reliably predict a patient's risk of recurrence and death [2,3]. Most of these factors, such as myometrial invasion, tumor grade, presence of lympho-vascular space invasion and regional node metastases are determined only after the results of comprehensive surgical staging are available. Furthermore, obesity, age and medical co-morbidities may limit the feasibility of such staging
⁎ Corresponding author. E-mail address: [email protected] (F. Musa). 0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2012.03.004
in some patients with endometrial cancer, making preoperative prognostic variables even more valuable. Endometrial cancer is a heterogeneous neoplasm comprised of two major histopathologic subtypes: endometrioid and non endometrioid. Often identified on a preoperative biopsy specimen, histology may be an independent predictor of survival. For instance, uterine papillary serous carcinomas are more aggressive tumors, associated with a 50% recurrence rate and decreased overall survival [4]. Mucinous adenocarcinoma of the endometrium [MUC] is a rare variant of endometrial carcinoma, accounting for less than 10% of endometrial tumors [4–7]. In the largest series published to date, MUC was only found in 21/256 cases (9%) [5]. Much of the literature on MUC was obtained prior to the practice of routine surgical staging, and is limited by the use of pre-operative pelvic irradiation. There remains limited information regarding mucinous differentiation as a prognostic factor in endometrial cancer. The aim of this study was to determine the incidence of MUC in a single institution, to evaluate potential risk factors for its
542
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545
development, and to examine the association between MUC and known prognostic factors in endometrial cancer as well as survival. Materials and methods Case selection and data retrieval Following approval by our institutional review board (IRB), the computerized files of the Pathology Department at the New York Presbyterian Hospital–Weill Cornell Medical Center were searched for the diagnosis of “endometrial adenocarcinoma” from 1996 to 2006. Women who did not undergo hysterectomy at the New York Presbyterian Hospital were excluded as were women with tumors with serous or clear cell differentiation, uterine sarcomas, or synchronous ovarian tumors. Three hundred and forty-four patients were identified, of which 41 had tumors with predominant mucinous differentiation (MUC). MUC was defined by the presence of intracytoplasmic mucin in >50% of the total tumor cell population (Fig. 1A) [8]. Lesions with less than 50% of mucinous differentiation were labeled as having “focal mucin production” or “focal mucinous features” and were not included in this study. Tumors that were predominantly mucinous were differentiated from mucinous metaplasia of the endometrium by the presence of architectural features of carcinoma such as
villoglandular or cribriform architecture (Fig. 1B). All cases of MUC were reviewed by a senior gynecologic pathologist (ECP) to confirm the diagnosis. Each case of mucinous endometrial adenocarcinoma was paired with two controls of usual endometrioid histology (without focal mucinous, clear cell or serous differentiation) matched by age and grade. The medical records for all cases in the study group and matched controls were retrieved and analyzed. Information extracted from the medical records included date and type of surgery performed, age at diagnosis, race, BMI, parity, history of hormone use, history of tobacco use and medical co-morbidities such as diabetes mellitus or hypertension. History of hormonal use was further subdivided into history of oral contraceptive use, use of hormonal replacement therapy, selective estrogen receptor modulators (tamoxifen or raloxifene) or progestin therapy. Tumor characteristics extracted from the records included histology, grade, depth of myometrial invasion, pelvic and para-aortic lymph node status, presence of lymph-vascular space invasion and positive washings. Follow up information included adjuvant treatment, date of last follow up or death, presence of recurrence and site of recurrent disease. Cause of death was determined from autopsy reports, hospital records or death certificates. Immunohistochemistry Immunohistochemical stains for estrogen and progesterone receptor statuses were performed on 4 mm formalin-fixed, paraffinembedded tissue sections, which were subjected to heat-induced antigen retrieval and incubated in an automated stainer with the ER (Novocastra, dilution 1:50) and PR (Novocastra, dilution 1:500) antibodies. Slides were then stained with diaminobenzidine chromogen and counterstained with hematoxylin. The nuclear staining for the receptors was scored as the following: 0 = no staining; 1 = less than 25% cell staining; 2 = more than 25% but less than 75% cell staining; and 3 = more than 75% cells staining. Appropriate negative and positive controls were used. Statistical analysis Statistical analyses were performed with SPSS (Statistical Package for the Social Sciences, version 16). Cases and controls were compared with regard to known risk factors for endometrial cancer and the extent of disease at diagnosis. Chi-square and Fisher's exact tests were used to compare proportions and Student's t-tests were used for the comparison of means. Multivariate analysis was used to identify the independent predictors of lymph node metastases. Recurrence free survival and overall survival were calculated using the Kaplan–Meier method and compared with the Log-rank test. p b 0.05 was considered significant for all tests. Results
Fig. 1. A. Mucinous adenocarcinoma of the endometrium with tumor cells showing abundant intracytoplasmic mucin, bland nuclear features and villoglandular architecture. B. Mucinous adenocarcinoma of the endometrium with tumor cells showing moderate intracytoplasmic mucin, bland nuclear features and cribriform architecture.
A review of the files of the Pathology Department at the New York Presbyterian Hospital showed that the ratio of MUC (tumors with predominant mucinous differentiation) to endometrioid adenocarcinomas without or with focal mucinous differentiation was 41 to 303 cases (14:100). The mean age for MUC cases was 65.3 (range = 43–80 yrs) and 64.5 (range = 43–82 yrs) for controls. The mean BMI was 28 for both cases and controls. Cases and controls were similar with regard to ethnicity and reflected the predominantly Caucasian patient population at our institution (Table 1). Race/ethnicity information was unknown for 10 controls (12%) and 1 case (2%). The patients in the control group were 70% Caucasian, 7% African American, 2% Hispanic and 9% Asian. In the cases group, patients were 83% Caucasian, 5% African American, 7% Hispanic and 2% Asian (p = NS).
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545 Table 1 Clinical characteristics of patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls). Demographic characteristics
Controls
t-Test Age at diagnosis (yrs) Body Mass Index (BMI)
Mean 64.46 28.08
SEM 1.147 0.739
Mean 65.32 28.10
Cases SEM 1.660 1.149
p 0.992 0.673
p
Chi-square Ethnicity Race — unknown White Black Hispanic Asian
n
%
n
%
p
10 57 6 2 7
12.2 69.5 7.3 2.4 8.5
1 34 2 3 1
2.4 82.9 4.9 7.3 2.4
0.137
SEM — standard error mean.
Regarding patient characteristics and medical histories, there was no difference in the proportion of smokers in either group (Table 2). Seventy-five percent of patients with mucinous histology had never smoked cigarettes compared to 66% of patients in the control group. Similarly, no difference was found between the proportion of medical co-morbidities such as diabetes mellitus and hypertension between cases and controls. Approximately 12% of cases and controls had a history of diabetes and 51% of cases and 54% of controls had a history of hypertension. Use of hormone replacement therapy and selective estrogen modulators was similar between the two groups (Table 2). However, use of the oral contraceptive pill was more frequent in endometrioid controls versus mucinous cases (32 versus 7%, p = 0.02). Both groups underwent similar surgical procedures. Approximately 70% of patients with mucinous carcinomas were staged
Table 2 Etiologic and pathologic characteristics of patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls). Patient characteristics
Controls
Cases
Chi-square tests
n
%
n
%
70 10 21 10 39 13 6 0 54 28 72 10 38 44
87.5 12.5 67.7 32.3 67.2 22.4 10.3 0 65.9 34.1 87.8 12.2 46.3 53.7
33 3 29 2 25 8 7 1 30 10 36 5 20 21
91.7 8.3 93.5 6.5 61 19.5 17.1 2.4 75 25 87.8 12.2 48.8 51.2
56 22 4 68 14 69 13 17 65 63 2 75 7 75 7
68.3 26.8 4.9 82.9 17.1 84.1 15.9 20.7 79.3 96.9 3.1 91.5 8.5 91.5 8.5
28 11 2 32 9 34 7 12 29 24 5 34 6 36 3
68.3 26.8 4.9 78 22 82.9 17.1 29.3 70.7 82.8 17.2 85 15 92.3 7.7
Etiologic risk factors Tamoxifen use — never Tamoxifen use — ever OCP use — never OCP use — ever HRT use — never Estrogen only Estrogen and progesterone Progesterone only Cigarette smoking — never Cigarette smoking — ever Diabetes — no Diabetes — yes Hypertension — no Hypertension — yes Prognostic factors Tumor grade I Tumor grade II Tumor grade III Myometrial invasion — inner half Myometrial invasion — outer half Lymphatic invasion — no Lymphatic invasion — yes Lymph node sampling — no Lymph node sampling — yes Lymph node status — no metastases Lymph node status — positive for metastases FIGO stages I–II FIGO stages III–IV Recurrences — no Recurrences — yes
Bold emphasis signifies p values b 0.05. FIGO — FIGO, International Federation of Gynecology and Obstetrics
p
0.379 0.010 0.478
0.208 1.000 0.474
1.000
0.336
543
with total hysterectomy, bilateral salpingo-oophorectomy, pelvic and para-aortic lymph node sampling and omentectomy (when appropriate) versus 79% of endometrioid cases (p = NS). In patients who were staged, the mean number of pelvic and para-aortic lymph nodes retrieved was similar between cases and controls. The average number of nodes sampled was 15.6 for cases, and 19.4 for controls (p = 0.09). All patients found to have advanced disease (stages III–IV) in the control group had their nodes sampled. In the mucinous group, one patient with probable stage IIIa disease (focal adnexal involvement) did not undergo nodal sampling due to medical co-morbidities. She was treated with chemotherapy and radiation and was free of recurrence at 43 months of follow up. The six remaining patients with advanced disease (stages III–IV) had lymph node sampling. Tumor grade was matched between cases and controls: 68% (28 cases and 56 controls) were Grade I, 27% (11 cases and 22 controls) were Grade II and 5% (2 cases and 4 controls) were Grade III. Approximately 80% of all patients had b50% myometrial invasion at diagnosis in both groups (83% of controls and 78% of cases, p = NS). There was no difference in the status of pelvic washings or presence of lymph-vascular space invasion between cases and controls (p = NS). Despite the similar tumor grades and depths of myometrial invasion seen in cases and controls, mucinous carcinomas were significantly more likely to present with regional nodal metastases compared to endometrioid tumors(17% versus 3%, p = 0.027). The association remained statistically significant on univariate logistic regression (p = 0.002). On multivariate analysis controlling for grade and depth of myometrial invasion, mucinous histology remained an independent predictor of nodal metastases (p = 0.022) (Table S2). Patients with mucinous histology were 12 times more likely to present with lymph node metastases compared to those with endometrioid tumors. There was no statistically significant difference between cases and controls with regard to adjuvant therapy. Approximately 36.4% of controls underwent adjuvant radiation versus 50% of cases of mucinous differentiation (p = NS). Of all patients who received radiation therapy, there was no significant difference between the proportion that underwent whole pelvic radiation versus brachytherapy versus a combination of both modalities (Table 3). Regarding systemic treatment, only 5.2% of patients in the control group received adjuvant chemotherapy versus 10% in the study group (p = NS) (Table 2). Median follow up was 43 months for cases (0–142 months) and 42 months for controls (range 0–133 months). The mean overall survival in months for cases and controls was similar, 116 versus 115 months (Fig. 2). A total of ten patients recurred during follow up and the median time to recurrence was 23 months (23 months for control group, 14 months for MUC). Three patients with MUC had recurrent disease (8%) versus 7 patients in the endometrioid control group (9%), p = NS. There were 6 pelvic recurrences (including lymph nodes and vaginal masses), 3 pulmonary and 1 hepatic recurrence. All three recurrences in the MUC group were pelvic (Table S1). Table 3 Adjuvant radiation versus chemotherapy for patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls).
0.526 0.203 0.027 0.217 0.591
Patient characteristics
Controls
Chi-square tests
n
%
n
%
18 2 11 0
58 6 35 0
11 2 7 2
50 9 32 9
73 4
95 5
36 4
90 10
Radiation therapy Brachytherapy Whole pelvic XRT Pelvic XRT and brachytherapy Unknown Chemotherapy No Yes
Cases
p
0.370
0.271
544
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545
mucinous. The immunostaining score was determined separately for the areas of mucinous differentiation and separately for the endometrioid areas. The results are shown in Table S3 and Fig. 3. Overall, all cases showed at least focal staining for ER and PR. However, the distribution of the scores indicated lower level of expressions of ER and PR in the mucinous areas as compared to the endometrioid areas. Discussion
Fig. 2. Kaplan–Meier survival curves for patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls).
ER and PR immunostainings were performed in 30 cases of mucinous adenocarcinoma. In 22 of the stained cases the tumor had mucinous (predominant) differentiation in addition to focal areas of endometrioid differentiation; in 8 cases the tumors were purely
100.0 90.0 80.0 70.0 % of cases 60.0 50.0 40.0 30.0 20.0 10.0 0.0 ER MUC ER EM
3 2 1 staining score 0 PR MUC PR EM
Fig. 3. Results of estrogen receptor (ER) and progesterone receptor (PR) immunostaining. Staining score: 0 = no staining; 1 = less than 25% cell staining; 2 = more than 25% but less than 75% cell staining; 3 = more than 75% cell staining.
Mucinous adenocarcinoma of the endometrium [MUC] is a relatively understudied and rare histological variant of endometrial carcinoma, accounting for less than 10% of endometrial tumors [4–7]. During the studied time period (1996–2006) mucinous adenocarcinoma comprised approximately 10% of all endometrial cancers treated at our institution. Controlling for grade of tumor and depth of myometrial invasion, mucinous histology was found to be an independent predictor of lymph node metastases. Despite the known association between nodal metastases and poor prognosis in endometrial cancer [3], no difference between recurrence free or overall survival was observed between endometrioid and mucinous adenocarcinoma in our study. The largest series on MUC was a retrospective review published in 1983 by Ross et al. at Stanford. Of 256 patients with carcinoma confined to the uterine corpus identified in the study period from 1959 to 1975, 98 (21%) were found to have focal intracytoplasmic mucin production, and only 21 (9%) were predominantly mucinous tumors (>50% mucinous features) [5]. Ross et al. found that frequency of myometrial invasion and recurrence rates were not different between the predominantly mucinous tumors and endometrioid cancers and concluded that mucin production did not have an independent impact on prognosis. However, this study was limited by the use of pre-operative radiation in 25% of patients. Additionally, the study utilized clinical staging and the rates of nodal metastases were not compared between mucinous and endometrioid carcinomas. Melhem and Tobon published a series of 18 cases of mucinous adenocarcinoma that occurred between 1969 and 1985 at a single institution [9]. They also concluded that MUC tends to be well differentiated and associated with a good prognosis. However, their patients were also clinically staged without an assessment of nodal status. Of the 18 patients, 3 died of “recurrent” disease: the first at 7 months after surgery, with lung metastases; the second at 11 months after surgery with positive para-aortic lymph nodes; and the last died of disease 54 months post-op with regional metastases to the bladder. Of the three patients with MUC and >50% myometrial invasion in their series, two died of the disease. Large clinical studies on MUC after 1987 are lacking. Thus, epidemiologic and prognostic risk factors have not been established. One study hypothesized tamoxifen and other synthetic gestagens to be risk factors in the development of MUC [10]. Our study did not confirm this association nor did it identify clear risk factors for MUC in our study population. Although most of our patients had never used oral contraceptives, those who did were more likely to develop tumors without mucinous features. None of the patients who used prior oral contraceptives had lymph node involvement or recurrences. The retrospective nature of our study and the relatively small numbers prevent any definitive statements on this finding. Perhaps oral contraceptive use offers greater protection against mucinous endometrial cancer compared to endometrioid tumors. It has been proposed that exogenous estrogen use such as the use of tamoxifen increases progesterone receptor (PR) concentrations in the endometrium, augmenting its own gestagen-like effectiveness in promoting mucinous metaplasia and subsequently carcinoma of the endometrium, but this theory has not yet been proven [10]. Our findings suggest at least focal ER and PR expressions on all mucinous cases; however, in cases where mucinous differentiation is coexistent with endometrioid areas, the expressions of ER and PR are higher in
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545
the endometrioid areas. PR expression by immunohistochemistry has been independently associated with disease free status in endometrioid cancers. Fukuda et al. identified negative PR status as an independent predictor of lymph node metastasis in endometrial cancer [11]. Creasman has suggested that the status of PR appeared to have a greater significance on survival than other well recognized prognostic factors in endometrial carcinoma [12]. It is plausible that in the rare cases where endometrial cancer develops despite prior oral contraceptive use, the tumors are indolent and lack mucinous features. Contrary to the favorable prognosis associated with PR expression and mucinous histology, the presence of lymph node metastasis has been associated with poor prognosis in multiple endometrial cancer studies [2,13–15]. Although most series of MUC are retrospective and too small to assess prognosis, we are not the first to suggest that these tumors seem more likely to metastasize. In a small series of 16 patients, Fujiwara and Longacre described a cytologically bland and architecturally well differentiated mucinous tumor of the endometrium with voluminous extracellular mucin [16]. In their series, 44% of patients had a documented history of hormonal therapy. Despite the bland histologic features, extensive LVSI was seen in 1/ 16 cases (6%), lymph node metastasis in 2/16 cases (13%), omental metastasis in 1/16 cases (6%) and cervical stromal involvement in 3/16 cases (19%). Their results are supportive of our findings of an increased rate of lymph node involvement in mucinous tumors (17 versus 3%). Recent evidence in endometrial cancer suggests that uterine risk factors are important determinants of prognosis. Kwon et al. retrospectively reviewed 316 Canadian women diagnosed with endometrial cancer between 1996 and 2000 with particular attention to 5 year survival [17]. 12% of the cohorts studied had positive lymph nodes. The group demonstrated that age > 60, grade III histology, deep myometrial invasion and cervical stromal invasion were all associated with increased mortality. Five year OS for node negative patients with more than two uterine risk factors was 53% compared to 75% of patients who were node positive without uterine risk factors. They concluded that pelvic node status was not an independent determinant of survival for their study population and that uterine risk factors should be considered regardless of nodal status when offering systemic therapy to maximize survival. This finding is consistent with the results of the current study where survival between cases and controls is similar despite the significantly increased rate of nodal metastases in women with mucinous endometrial cancers. A retrospective review of 27,000 women from the SEER database also addressed this issue [18]. In this study, patients were stratified according to uterine risk factors such as myometrial invasion, cervical stromal invasion, grade III histology and nodal status. The group observed that although uterine risk factors influenced survival, nodal disease was still a more important negative prognostic factor. The 5 year OS for node positive patients without uterine risk factors in their study was 68%, similar to node negative patients with more than 2 uterine risk factors. Node positive patients with concurrent uterine risk factors had a 58% 5 year survival. Unfortunately, it is not possible to determine the proportion of MUC on the SEER database and to calculate their survival, but it is possible that for this subset
545
of patients, uterine risk factors may be more important than nodal status. Endometrial cancer is a heterogeneous disease with several identified intrauterine and extrauterine prognostic factors. The current study adds to the limited body of literature examining the significance of mucinous histology in endometrial carcinoma. This is the largest series published to date on MUC and the only study designed to assess its impact on lymph node status. Our data suggest that mucinous histology is an independent risk factor for regional node metastases. In light of this finding, we recommend comprehensive surgical staging including retroperitoneal node sampling in patients with endometrial cancer of predominantly mucinous histology. Supplementary materials related to this article can be found online at doi:10.1016/j.ygyno.2012.03.004.
Conflict of interest statement The authors have no conflicts of interest to report.
References [1] Cancer Facts & Figures. American Cancer Society; 2010. [2] Creasman WT, et al. Surgical pathologic spread patterns of endometrial cancer. A Gynecologic Oncology Group study. Cancer 1987;60(8 Suppl):2035–41. [3] Keys HM, et al. A phase III trial of surgery with or without adjunctive external pelvic radiation therapy in intermediate risk endometrial adenocarcinoma: a Gynecologic Oncology Group study. Gynecol Oncol 2004;92(3):744–51. [4] Hendrickson M, et al. Adenocarcinoma of the endometrium: analysis of 256 cases with carcinoma limited to the uterine corpus. Pathology review and analysis of prognostic variables. Gynecol Oncol 1982;13(3):373–92. [5] Ross JC, et al. Primary mucinous adenocarcinoma of the endometrium. A clinicopathologic and histochemical study. Am J Surg Pathol 1983;7(8):715–29. [6] Sorvari TE. A histochemical study of epithelial mucosubstances in endometrial and cervical adenocarcinomas. With reference to normal endometrium and cervical mucosa. Acta Pathol Microbiol Scand Suppl 1969:207 p. Suppl 207:1+. [7] Salm R. Mucin production of normal and abnormal endometrium. Arch Pathol 1962;73:30–9. [8] Blaustein A, Kurman RJ. Blaustein's pathology of the female genital tract. 5th ed. New York: Springer-Verlag; 2002. xv, 1391 p. [9] Melhem MF, Tobon H. Mucinous adenocarcinoma of the endometrium: a clinicopathological review of 18 cases. Int J Gynecol Pathol 1987;6(4):347–55. [10] Dallenbach-Hellweg G, Hahn U. Mucinous and clear cell adenocarcinomas of the endometrium in patients receiving antiestrogens (tamoxifen) and gestagens. Int J Gynecol Pathol 1995;14(1):7–15. [11] Fukuda K, et al. Prognostic significance of progesterone receptor immunohistochemistry in endometrial carcinoma. Gynecol Oncol 1998;69(3):220–5. [12] Creasman WT, et al. Influence of cytoplasmic steroid receptor content on prognosis of early stage endometrial carcinoma. Am J Obstet Gynecol 1985;151(7): 922–32. [13] Morrow CP, et al. Relationship between surgical–pathological risk factors and outcome in clinical stage I and II carcinoma of the endometrium: a Gynecologic Oncology Group study. Gynecol Oncol 1991;40(1):55–65. [14] Boronow RC, et al. Surgical staging in endometrial cancer: clinical–pathologic findings of a prospective study. Obstet Gynecol 1984;63(6):825–32. [15] DiSaia PJ, et al. Risk factors and recurrent patterns in stage I endometrial cancer. Am J Obstet Gynecol 1985;151(8):1009–15. [16] Fujiwara M, Longacre TA. Low-grade mucinous adenocarcinoma of the uterine corpus: a rare and deceptively bland form of endometrial carcinoma. Am J Surg Pathol 2011;35(4):537–44. [17] Kwon JS, et al. Are uterine risk factors more important than nodal status in predicting survival in endometrial cancer? Obstet Gynecol 2009;114(4):736–43. [18] Barrena Medel NI. Comparison of the prognostic significance of uterine factors and nodal status for endometrial cancer. Am J Obstet Gynecol 2011;204(3):248 e1-7.
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Mucinous histology is a risk factor for nodal metastases in endometrial cancer Fernanda Musa a,⁎, Marilyn Huang b, Brandi Adams a, Edyta Pirog c, Kevin Holcomb d a
Department of Obstetrics & Gynecology, Weill Cornell Medical College, New York, NY, USA Division of Gynecologic Oncology, MD Anderson Cancer Center, Houston, TX, USA c Department of Pathology, Weill Cornell Medical College, New York, NY, USA d Division of Gynecologic Oncology, Weill Cornell Medical College, New York, NY, USA b
a r t i c l e
i n f o
Article history: Received 13 October 2011 Available online 9 March 2012 Keywords: Endometrial cancer Mucinous differentiation Mucinous adenocarcinoma of the endometrium lymph node metastases
a b s t r a c t Objectives. Mucinous adenocarcinoma of the endometrium (MUC) is a rare histological variant of endometrial carcinoma accounting for 1–9% of endometrioid tumors. Few studies have characterized its clinical behavior. This is a case–control study at a single institution comparing the risk factors and clinical course of MUC relative to endometrioid adenocarcinoma. Methods. A case–control study was performed including patients treated for endometrial cancer between 1996 and 2006. 41 cases of mucinous adenocarcinoma were identified. Each case was matched with two controls of endometrioid histology by age and histological grade. Cases and controls were compared with regard to known risk factors for endometrial cancer and the extent of disease at diagnosis. Chi-square tests were used to compare proportions and Student's t-tests for the comparison of means. Multivariate regression was used to identify the independent predictors of lymph node metastases. Overall survival was calculated using the Kaplan–Meier method and compared with the Log-rank test. p b .05 was considered significant for all tests. Results. Cases and controls were matched by age and FIGO grade and were found to be similar in regard to ethnicity, body mass index and medical history. No significant difference in myometrial invasion (MI) > 50% or the presence of lymph-vascular space invasion was found between cases and controls, however, 17% of patients with MUC had lymph node metastases compared to 3% of controls (p = .01). Multivariate analysis controlling for both tumor grade and depth of MI identified mucinous histology as an independent predictor of lymph node metastasis (p = .02). There was no difference in adjuvant treatment, recurrence rate or survival between the two groups. Conclusion. Mucinous differentiation was found to be an independent predictor of lymph node metastasis in the study population. Comprehensive surgical staging including retroperitoneal node dissection should be strongly considered in all endometrial cancer patients with predominantly mucinous histology. © 2012 Elsevier Inc. All rights reserved.
Introduction Endometrial cancer is the most common gynecologic malignancy in the United States with approximately 43,470 new cases diagnosed yearly and approximately 8000 annual deaths attributed to the disease [1]. Although the majority of cases are identified early and are associated with a favorable prognosis, there are several well established poor prognostic factors for endometrial cancer which reliably predict a patient's risk of recurrence and death [2,3]. Most of these factors, such as myometrial invasion, tumor grade, presence of lympho-vascular space invasion and regional node metastases are determined only after the results of comprehensive surgical staging are available. Furthermore, obesity, age and medical co-morbidities may limit the feasibility of such staging
⁎ Corresponding author. E-mail address: [email protected] (F. Musa). 0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2012.03.004
in some patients with endometrial cancer, making preoperative prognostic variables even more valuable. Endometrial cancer is a heterogeneous neoplasm comprised of two major histopathologic subtypes: endometrioid and non endometrioid. Often identified on a preoperative biopsy specimen, histology may be an independent predictor of survival. For instance, uterine papillary serous carcinomas are more aggressive tumors, associated with a 50% recurrence rate and decreased overall survival [4]. Mucinous adenocarcinoma of the endometrium [MUC] is a rare variant of endometrial carcinoma, accounting for less than 10% of endometrial tumors [4–7]. In the largest series published to date, MUC was only found in 21/256 cases (9%) [5]. Much of the literature on MUC was obtained prior to the practice of routine surgical staging, and is limited by the use of pre-operative pelvic irradiation. There remains limited information regarding mucinous differentiation as a prognostic factor in endometrial cancer. The aim of this study was to determine the incidence of MUC in a single institution, to evaluate potential risk factors for its
542
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545
development, and to examine the association between MUC and known prognostic factors in endometrial cancer as well as survival. Materials and methods Case selection and data retrieval Following approval by our institutional review board (IRB), the computerized files of the Pathology Department at the New York Presbyterian Hospital–Weill Cornell Medical Center were searched for the diagnosis of “endometrial adenocarcinoma” from 1996 to 2006. Women who did not undergo hysterectomy at the New York Presbyterian Hospital were excluded as were women with tumors with serous or clear cell differentiation, uterine sarcomas, or synchronous ovarian tumors. Three hundred and forty-four patients were identified, of which 41 had tumors with predominant mucinous differentiation (MUC). MUC was defined by the presence of intracytoplasmic mucin in >50% of the total tumor cell population (Fig. 1A) [8]. Lesions with less than 50% of mucinous differentiation were labeled as having “focal mucin production” or “focal mucinous features” and were not included in this study. Tumors that were predominantly mucinous were differentiated from mucinous metaplasia of the endometrium by the presence of architectural features of carcinoma such as
villoglandular or cribriform architecture (Fig. 1B). All cases of MUC were reviewed by a senior gynecologic pathologist (ECP) to confirm the diagnosis. Each case of mucinous endometrial adenocarcinoma was paired with two controls of usual endometrioid histology (without focal mucinous, clear cell or serous differentiation) matched by age and grade. The medical records for all cases in the study group and matched controls were retrieved and analyzed. Information extracted from the medical records included date and type of surgery performed, age at diagnosis, race, BMI, parity, history of hormone use, history of tobacco use and medical co-morbidities such as diabetes mellitus or hypertension. History of hormonal use was further subdivided into history of oral contraceptive use, use of hormonal replacement therapy, selective estrogen receptor modulators (tamoxifen or raloxifene) or progestin therapy. Tumor characteristics extracted from the records included histology, grade, depth of myometrial invasion, pelvic and para-aortic lymph node status, presence of lymph-vascular space invasion and positive washings. Follow up information included adjuvant treatment, date of last follow up or death, presence of recurrence and site of recurrent disease. Cause of death was determined from autopsy reports, hospital records or death certificates. Immunohistochemistry Immunohistochemical stains for estrogen and progesterone receptor statuses were performed on 4 mm formalin-fixed, paraffinembedded tissue sections, which were subjected to heat-induced antigen retrieval and incubated in an automated stainer with the ER (Novocastra, dilution 1:50) and PR (Novocastra, dilution 1:500) antibodies. Slides were then stained with diaminobenzidine chromogen and counterstained with hematoxylin. The nuclear staining for the receptors was scored as the following: 0 = no staining; 1 = less than 25% cell staining; 2 = more than 25% but less than 75% cell staining; and 3 = more than 75% cells staining. Appropriate negative and positive controls were used. Statistical analysis Statistical analyses were performed with SPSS (Statistical Package for the Social Sciences, version 16). Cases and controls were compared with regard to known risk factors for endometrial cancer and the extent of disease at diagnosis. Chi-square and Fisher's exact tests were used to compare proportions and Student's t-tests were used for the comparison of means. Multivariate analysis was used to identify the independent predictors of lymph node metastases. Recurrence free survival and overall survival were calculated using the Kaplan–Meier method and compared with the Log-rank test. p b 0.05 was considered significant for all tests. Results
Fig. 1. A. Mucinous adenocarcinoma of the endometrium with tumor cells showing abundant intracytoplasmic mucin, bland nuclear features and villoglandular architecture. B. Mucinous adenocarcinoma of the endometrium with tumor cells showing moderate intracytoplasmic mucin, bland nuclear features and cribriform architecture.
A review of the files of the Pathology Department at the New York Presbyterian Hospital showed that the ratio of MUC (tumors with predominant mucinous differentiation) to endometrioid adenocarcinomas without or with focal mucinous differentiation was 41 to 303 cases (14:100). The mean age for MUC cases was 65.3 (range = 43–80 yrs) and 64.5 (range = 43–82 yrs) for controls. The mean BMI was 28 for both cases and controls. Cases and controls were similar with regard to ethnicity and reflected the predominantly Caucasian patient population at our institution (Table 1). Race/ethnicity information was unknown for 10 controls (12%) and 1 case (2%). The patients in the control group were 70% Caucasian, 7% African American, 2% Hispanic and 9% Asian. In the cases group, patients were 83% Caucasian, 5% African American, 7% Hispanic and 2% Asian (p = NS).
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545 Table 1 Clinical characteristics of patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls). Demographic characteristics
Controls
t-Test Age at diagnosis (yrs) Body Mass Index (BMI)
Mean 64.46 28.08
SEM 1.147 0.739
Mean 65.32 28.10
Cases SEM 1.660 1.149
p 0.992 0.673
p
Chi-square Ethnicity Race — unknown White Black Hispanic Asian
n
%
n
%
p
10 57 6 2 7
12.2 69.5 7.3 2.4 8.5
1 34 2 3 1
2.4 82.9 4.9 7.3 2.4
0.137
SEM — standard error mean.
Regarding patient characteristics and medical histories, there was no difference in the proportion of smokers in either group (Table 2). Seventy-five percent of patients with mucinous histology had never smoked cigarettes compared to 66% of patients in the control group. Similarly, no difference was found between the proportion of medical co-morbidities such as diabetes mellitus and hypertension between cases and controls. Approximately 12% of cases and controls had a history of diabetes and 51% of cases and 54% of controls had a history of hypertension. Use of hormone replacement therapy and selective estrogen modulators was similar between the two groups (Table 2). However, use of the oral contraceptive pill was more frequent in endometrioid controls versus mucinous cases (32 versus 7%, p = 0.02). Both groups underwent similar surgical procedures. Approximately 70% of patients with mucinous carcinomas were staged
Table 2 Etiologic and pathologic characteristics of patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls). Patient characteristics
Controls
Cases
Chi-square tests
n
%
n
%
70 10 21 10 39 13 6 0 54 28 72 10 38 44
87.5 12.5 67.7 32.3 67.2 22.4 10.3 0 65.9 34.1 87.8 12.2 46.3 53.7
33 3 29 2 25 8 7 1 30 10 36 5 20 21
91.7 8.3 93.5 6.5 61 19.5 17.1 2.4 75 25 87.8 12.2 48.8 51.2
56 22 4 68 14 69 13 17 65 63 2 75 7 75 7
68.3 26.8 4.9 82.9 17.1 84.1 15.9 20.7 79.3 96.9 3.1 91.5 8.5 91.5 8.5
28 11 2 32 9 34 7 12 29 24 5 34 6 36 3
68.3 26.8 4.9 78 22 82.9 17.1 29.3 70.7 82.8 17.2 85 15 92.3 7.7
Etiologic risk factors Tamoxifen use — never Tamoxifen use — ever OCP use — never OCP use — ever HRT use — never Estrogen only Estrogen and progesterone Progesterone only Cigarette smoking — never Cigarette smoking — ever Diabetes — no Diabetes — yes Hypertension — no Hypertension — yes Prognostic factors Tumor grade I Tumor grade II Tumor grade III Myometrial invasion — inner half Myometrial invasion — outer half Lymphatic invasion — no Lymphatic invasion — yes Lymph node sampling — no Lymph node sampling — yes Lymph node status — no metastases Lymph node status — positive for metastases FIGO stages I–II FIGO stages III–IV Recurrences — no Recurrences — yes
Bold emphasis signifies p values b 0.05. FIGO — FIGO, International Federation of Gynecology and Obstetrics
p
0.379 0.010 0.478
0.208 1.000 0.474
1.000
0.336
543
with total hysterectomy, bilateral salpingo-oophorectomy, pelvic and para-aortic lymph node sampling and omentectomy (when appropriate) versus 79% of endometrioid cases (p = NS). In patients who were staged, the mean number of pelvic and para-aortic lymph nodes retrieved was similar between cases and controls. The average number of nodes sampled was 15.6 for cases, and 19.4 for controls (p = 0.09). All patients found to have advanced disease (stages III–IV) in the control group had their nodes sampled. In the mucinous group, one patient with probable stage IIIa disease (focal adnexal involvement) did not undergo nodal sampling due to medical co-morbidities. She was treated with chemotherapy and radiation and was free of recurrence at 43 months of follow up. The six remaining patients with advanced disease (stages III–IV) had lymph node sampling. Tumor grade was matched between cases and controls: 68% (28 cases and 56 controls) were Grade I, 27% (11 cases and 22 controls) were Grade II and 5% (2 cases and 4 controls) were Grade III. Approximately 80% of all patients had b50% myometrial invasion at diagnosis in both groups (83% of controls and 78% of cases, p = NS). There was no difference in the status of pelvic washings or presence of lymph-vascular space invasion between cases and controls (p = NS). Despite the similar tumor grades and depths of myometrial invasion seen in cases and controls, mucinous carcinomas were significantly more likely to present with regional nodal metastases compared to endometrioid tumors(17% versus 3%, p = 0.027). The association remained statistically significant on univariate logistic regression (p = 0.002). On multivariate analysis controlling for grade and depth of myometrial invasion, mucinous histology remained an independent predictor of nodal metastases (p = 0.022) (Table S2). Patients with mucinous histology were 12 times more likely to present with lymph node metastases compared to those with endometrioid tumors. There was no statistically significant difference between cases and controls with regard to adjuvant therapy. Approximately 36.4% of controls underwent adjuvant radiation versus 50% of cases of mucinous differentiation (p = NS). Of all patients who received radiation therapy, there was no significant difference between the proportion that underwent whole pelvic radiation versus brachytherapy versus a combination of both modalities (Table 3). Regarding systemic treatment, only 5.2% of patients in the control group received adjuvant chemotherapy versus 10% in the study group (p = NS) (Table 2). Median follow up was 43 months for cases (0–142 months) and 42 months for controls (range 0–133 months). The mean overall survival in months for cases and controls was similar, 116 versus 115 months (Fig. 2). A total of ten patients recurred during follow up and the median time to recurrence was 23 months (23 months for control group, 14 months for MUC). Three patients with MUC had recurrent disease (8%) versus 7 patients in the endometrioid control group (9%), p = NS. There were 6 pelvic recurrences (including lymph nodes and vaginal masses), 3 pulmonary and 1 hepatic recurrence. All three recurrences in the MUC group were pelvic (Table S1). Table 3 Adjuvant radiation versus chemotherapy for patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls).
0.526 0.203 0.027 0.217 0.591
Patient characteristics
Controls
Chi-square tests
n
%
n
%
18 2 11 0
58 6 35 0
11 2 7 2
50 9 32 9
73 4
95 5
36 4
90 10
Radiation therapy Brachytherapy Whole pelvic XRT Pelvic XRT and brachytherapy Unknown Chemotherapy No Yes
Cases
p
0.370
0.271
544
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545
mucinous. The immunostaining score was determined separately for the areas of mucinous differentiation and separately for the endometrioid areas. The results are shown in Table S3 and Fig. 3. Overall, all cases showed at least focal staining for ER and PR. However, the distribution of the scores indicated lower level of expressions of ER and PR in the mucinous areas as compared to the endometrioid areas. Discussion
Fig. 2. Kaplan–Meier survival curves for patients with mucinous adenocarcinoma (cases) versus patients with endometrioid adenocarcinoma (controls).
ER and PR immunostainings were performed in 30 cases of mucinous adenocarcinoma. In 22 of the stained cases the tumor had mucinous (predominant) differentiation in addition to focal areas of endometrioid differentiation; in 8 cases the tumors were purely
100.0 90.0 80.0 70.0 % of cases 60.0 50.0 40.0 30.0 20.0 10.0 0.0 ER MUC ER EM
3 2 1 staining score 0 PR MUC PR EM
Fig. 3. Results of estrogen receptor (ER) and progesterone receptor (PR) immunostaining. Staining score: 0 = no staining; 1 = less than 25% cell staining; 2 = more than 25% but less than 75% cell staining; 3 = more than 75% cell staining.
Mucinous adenocarcinoma of the endometrium [MUC] is a relatively understudied and rare histological variant of endometrial carcinoma, accounting for less than 10% of endometrial tumors [4–7]. During the studied time period (1996–2006) mucinous adenocarcinoma comprised approximately 10% of all endometrial cancers treated at our institution. Controlling for grade of tumor and depth of myometrial invasion, mucinous histology was found to be an independent predictor of lymph node metastases. Despite the known association between nodal metastases and poor prognosis in endometrial cancer [3], no difference between recurrence free or overall survival was observed between endometrioid and mucinous adenocarcinoma in our study. The largest series on MUC was a retrospective review published in 1983 by Ross et al. at Stanford. Of 256 patients with carcinoma confined to the uterine corpus identified in the study period from 1959 to 1975, 98 (21%) were found to have focal intracytoplasmic mucin production, and only 21 (9%) were predominantly mucinous tumors (>50% mucinous features) [5]. Ross et al. found that frequency of myometrial invasion and recurrence rates were not different between the predominantly mucinous tumors and endometrioid cancers and concluded that mucin production did not have an independent impact on prognosis. However, this study was limited by the use of pre-operative radiation in 25% of patients. Additionally, the study utilized clinical staging and the rates of nodal metastases were not compared between mucinous and endometrioid carcinomas. Melhem and Tobon published a series of 18 cases of mucinous adenocarcinoma that occurred between 1969 and 1985 at a single institution [9]. They also concluded that MUC tends to be well differentiated and associated with a good prognosis. However, their patients were also clinically staged without an assessment of nodal status. Of the 18 patients, 3 died of “recurrent” disease: the first at 7 months after surgery, with lung metastases; the second at 11 months after surgery with positive para-aortic lymph nodes; and the last died of disease 54 months post-op with regional metastases to the bladder. Of the three patients with MUC and >50% myometrial invasion in their series, two died of the disease. Large clinical studies on MUC after 1987 are lacking. Thus, epidemiologic and prognostic risk factors have not been established. One study hypothesized tamoxifen and other synthetic gestagens to be risk factors in the development of MUC [10]. Our study did not confirm this association nor did it identify clear risk factors for MUC in our study population. Although most of our patients had never used oral contraceptives, those who did were more likely to develop tumors without mucinous features. None of the patients who used prior oral contraceptives had lymph node involvement or recurrences. The retrospective nature of our study and the relatively small numbers prevent any definitive statements on this finding. Perhaps oral contraceptive use offers greater protection against mucinous endometrial cancer compared to endometrioid tumors. It has been proposed that exogenous estrogen use such as the use of tamoxifen increases progesterone receptor (PR) concentrations in the endometrium, augmenting its own gestagen-like effectiveness in promoting mucinous metaplasia and subsequently carcinoma of the endometrium, but this theory has not yet been proven [10]. Our findings suggest at least focal ER and PR expressions on all mucinous cases; however, in cases where mucinous differentiation is coexistent with endometrioid areas, the expressions of ER and PR are higher in
F. Musa et al. / Gynecologic Oncology 125 (2012) 541–545
the endometrioid areas. PR expression by immunohistochemistry has been independently associated with disease free status in endometrioid cancers. Fukuda et al. identified negative PR status as an independent predictor of lymph node metastasis in endometrial cancer [11]. Creasman has suggested that the status of PR appeared to have a greater significance on survival than other well recognized prognostic factors in endometrial carcinoma [12]. It is plausible that in the rare cases where endometrial cancer develops despite prior oral contraceptive use, the tumors are indolent and lack mucinous features. Contrary to the favorable prognosis associated with PR expression and mucinous histology, the presence of lymph node metastasis has been associated with poor prognosis in multiple endometrial cancer studies [2,13–15]. Although most series of MUC are retrospective and too small to assess prognosis, we are not the first to suggest that these tumors seem more likely to metastasize. In a small series of 16 patients, Fujiwara and Longacre described a cytologically bland and architecturally well differentiated mucinous tumor of the endometrium with voluminous extracellular mucin [16]. In their series, 44% of patients had a documented history of hormonal therapy. Despite the bland histologic features, extensive LVSI was seen in 1/ 16 cases (6%), lymph node metastasis in 2/16 cases (13%), omental metastasis in 1/16 cases (6%) and cervical stromal involvement in 3/16 cases (19%). Their results are supportive of our findings of an increased rate of lymph node involvement in mucinous tumors (17 versus 3%). Recent evidence in endometrial cancer suggests that uterine risk factors are important determinants of prognosis. Kwon et al. retrospectively reviewed 316 Canadian women diagnosed with endometrial cancer between 1996 and 2000 with particular attention to 5 year survival [17]. 12% of the cohorts studied had positive lymph nodes. The group demonstrated that age > 60, grade III histology, deep myometrial invasion and cervical stromal invasion were all associated with increased mortality. Five year OS for node negative patients with more than two uterine risk factors was 53% compared to 75% of patients who were node positive without uterine risk factors. They concluded that pelvic node status was not an independent determinant of survival for their study population and that uterine risk factors should be considered regardless of nodal status when offering systemic therapy to maximize survival. This finding is consistent with the results of the current study where survival between cases and controls is similar despite the significantly increased rate of nodal metastases in women with mucinous endometrial cancers. A retrospective review of 27,000 women from the SEER database also addressed this issue [18]. In this study, patients were stratified according to uterine risk factors such as myometrial invasion, cervical stromal invasion, grade III histology and nodal status. The group observed that although uterine risk factors influenced survival, nodal disease was still a more important negative prognostic factor. The 5 year OS for node positive patients without uterine risk factors in their study was 68%, similar to node negative patients with more than 2 uterine risk factors. Node positive patients with concurrent uterine risk factors had a 58% 5 year survival. Unfortunately, it is not possible to determine the proportion of MUC on the SEER database and to calculate their survival, but it is possible that for this subset
545
of patients, uterine risk factors may be more important than nodal status. Endometrial cancer is a heterogeneous disease with several identified intrauterine and extrauterine prognostic factors. The current study adds to the limited body of literature examining the significance of mucinous histology in endometrial carcinoma. This is the largest series published to date on MUC and the only study designed to assess its impact on lymph node status. Our data suggest that mucinous histology is an independent risk factor for regional node metastases. In light of this finding, we recommend comprehensive surgical staging including retroperitoneal node sampling in patients with endometrial cancer of predominantly mucinous histology. Supplementary materials related to this article can be found online at doi:10.1016/j.ygyno.2012.03.004.
Conflict of interest statement The authors have no conflicts of interest to report.
References [1] Cancer Facts & Figures. American Cancer Society; 2010. [2] Creasman WT, et al. Surgical pathologic spread patterns of endometrial cancer. A Gynecologic Oncology Group study. Cancer 1987;60(8 Suppl):2035–41. [3] Keys HM, et al. A phase III trial of surgery with or without adjunctive external pelvic radiation therapy in intermediate risk endometrial adenocarcinoma: a Gynecologic Oncology Group study. Gynecol Oncol 2004;92(3):744–51. [4] Hendrickson M, et al. Adenocarcinoma of the endometrium: analysis of 256 cases with carcinoma limited to the uterine corpus. Pathology review and analysis of prognostic variables. Gynecol Oncol 1982;13(3):373–92. [5] Ross JC, et al. Primary mucinous adenocarcinoma of the endometrium. A clinicopathologic and histochemical study. Am J Surg Pathol 1983;7(8):715–29. [6] Sorvari TE. A histochemical study of epithelial mucosubstances in endometrial and cervical adenocarcinomas. With reference to normal endometrium and cervical mucosa. Acta Pathol Microbiol Scand Suppl 1969:207 p. Suppl 207:1+. [7] Salm R. Mucin production of normal and abnormal endometrium. Arch Pathol 1962;73:30–9. [8] Blaustein A, Kurman RJ. Blaustein's pathology of the female genital tract. 5th ed. New York: Springer-Verlag; 2002. xv, 1391 p. [9] Melhem MF, Tobon H. Mucinous adenocarcinoma of the endometrium: a clinicopathological review of 18 cases. Int J Gynecol Pathol 1987;6(4):347–55. [10] Dallenbach-Hellweg G, Hahn U. Mucinous and clear cell adenocarcinomas of the endometrium in patients receiving antiestrogens (tamoxifen) and gestagens. Int J Gynecol Pathol 1995;14(1):7–15. [11] Fukuda K, et al. Prognostic significance of progesterone receptor immunohistochemistry in endometrial carcinoma. Gynecol Oncol 1998;69(3):220–5. [12] Creasman WT, et al. Influence of cytoplasmic steroid receptor content on prognosis of early stage endometrial carcinoma. Am J Obstet Gynecol 1985;151(7): 922–32. [13] Morrow CP, et al. Relationship between surgical–pathological risk factors and outcome in clinical stage I and II carcinoma of the endometrium: a Gynecologic Oncology Group study. Gynecol Oncol 1991;40(1):55–65. [14] Boronow RC, et al. Surgical staging in endometrial cancer: clinical–pathologic findings of a prospective study. Obstet Gynecol 1984;63(6):825–32. [15] DiSaia PJ, et al. Risk factors and recurrent patterns in stage I endometrial cancer. Am J Obstet Gynecol 1985;151(8):1009–15. [16] Fujiwara M, Longacre TA. Low-grade mucinous adenocarcinoma of the uterine corpus: a rare and deceptively bland form of endometrial carcinoma. Am J Surg Pathol 2011;35(4):537–44. [17] Kwon JS, et al. Are uterine risk factors more important than nodal status in predicting survival in endometrial cancer? Obstet Gynecol 2009;114(4):736–43. [18] Barrena Medel NI. Comparison of the prognostic significance of uterine factors and nodal status for endometrial cancer. Am J Obstet Gynecol 2011;204(3):248 e1-7.