Weber, Sarah K. MD*; Sauerwald, Axel MD*; Pölcher, Martin MD*; Braun, Michael MD*; Debald, Manuel MD*; Serce, Nuran Bektas MD‡; Kuhn, Walther MD*; Brunagel-Walgenbach, Giesela MD†; Rudlowski, Christian MD*
Endometrial cancer is the most frequent gynecologic malignancy in the Western world. In North America, endometrial cancer is the eighth most common cause of death from cancer in the female population and in Europe, an estimated 9000 women die of endometrial cancer each year.1 The death rate has increased during the recent decades, probably because of an increase in life span and coexisting medical comorbidities.2
Important prognostic factors in endometrial cancer are histological type, deep myometrial invasion, grade, and lymph node status.3 Five-year survival is approximately 90% in disease confined to the uterus, but drops to 60% when lymph nodes are positive.4 Surgical assessment of lymph node status in endometrial cancer, especially in apparent early stages, remains an area of significant debate in the gynecologic oncology community.5 To date, it remains unclear whether systematic lymph node dissection is of any therapeutic value. Also, morbidity and mortality rates of an exact surgical staging in early endometrial cancer need to be justified. Thus, based on relevant prognostic parameters, patients who will benefit from complete systematic lymphadenectomy have to be selected to minimize overtreatment and undertreatment.
Lymph node metastasis is a complex multistep biological process initialized by tumor cells that involve stromal invasion, new lymphatic vessel formation, and spread to the lymph nodes. Reliable markers predictive of lymph node metastasis might improve prognostication and be useful for therapeutic decision making in early cancers.
In several solid tumors, lymph vessel invasion (LVI) has been shown to be predictive of lymph node involvement and a prognostic factor for survival.6–10 Further, LVI has been included as a novel adverse prognostic factor for early breast cancer.11
In endometrial cancer, LVI has previously been investigated and was suggested to be prognostically relevant.3,12–14 The standard method for assessing LVI is light microscopic examination of hematoxylin and eosin (H-E)–stained sections. However, in the absence of specific markers for lymph vessel endothelium, distinguishing lymphatic from blood vessels and retraction artifacts can be challenging.15 A precise staining of lymphatic endothelium would help to overcome those problems in the detection of LVI. During the last decade, several specific markers for lymphatic endothelium have been explored, but most of these parameters lack specificity. The endothelial marker D2-40 (podoplanin), however, is exclusively detectable in lymphatic but not blood endothelial cells.6 Therefore, D2-40, which is recognized by monoclonal D2-40 antibodies, is regarded as highly sensitive and specific for LVI assessment in several tumor.10,16–18 Up to now, only limited data are available for D2-40–based LVI assessment in endometrial cancer.19
In the present study, we thus evaluated LVI by D2-40 immunostaining in primary endometrial cancer and correlated the findings with lymph node metastasis and clinical outcome.
PATIENTS AND METHODS
Patient Selection and Histological Findings
This retrospective analysis was performed to evaluate consecutively records of 182 patients with histologically confirmed endometrioid adenocarcinoma of the uterine corpus. The study was approved by the ethics committee of the University of Bonn, Germany. Patients’ data were collected retrospectively and blinded (name and date of birth). Patients underwent surgery at the Department of Obstetrics and Gynecology, University of Bonn Medical Centre, between January 1995 and December 2008.
Surgical treatment consisted of abdominal hysterectomy and bilateral salpingo-oophorectomy with or without complete systematic lymph node dissection (pelvic and para-aortic lymph nodes). Before 2002, a complete lymphadenectomy was performed in case of clinically and computed tomography–based enlarged lymph nodes suspiciously for metastasis. After 2002, a complete systematic lymphadenectomy was conducted and defined by the removal of at least 10 para-aortic and 15 pelvic lymph nodes confirmed by pathological assessment in accordance to the German guidelines.20 Only patients with histologically confirmed endometrial adenocarcinoma and at least myometrial invasion were included. Tumors worse than FIGO stage IIIC and other histopathology subtypes, for example, papillary serous or clear cell carcinoma, were excluded. Data on tumor characteristics, especially regarding LVI in H-E staining were obtained from the original histopathological reports. In all cases, a histopathological workup of tumor specimens (histological type, tumor size, tumor grading, lymph node involvement, etc) was performed. The clinical parameters were collected from the patients’ charts. For all patients, pathological data and formalin-fixed paraffin-embedded tumor blocks were available.
The original H-E–stained sections of the primary tumors were reviewed, and the associated tumor blocks were obtained from the Department of Pathology at the University of Bonn Medical Centre. From these blocks, 4-mm sections were cut consecutively: 1 section was stained with H-E and 2 sections underwent D2-40 immunohistochemistry as follows. Monoclonal antibodies antihuman D2-40 (1:50; Signet, Dedham, MA) were used for immunohistochemical detection. Sections were dried overnight at room temperature and were deparaffinized and rehydrated by decreasing concentrations of ethanol followed by Tris buffer.
For antigen retrieval, the sections were incubated in a 10-mM citrate buffer at pH 6.0 and heated by microwave radiation (600 W) for 2 × 15 minutes. The following steps were performed semiautomatically using a streptavidin-biotin-peroxidase technique and a DAKO TechMate 500 (DAKO A/S, Denmark) according to the provider’s instructions. Positive and negative control experiments were performed. Each tumor, stained with H-E, and D2-40 was blinded for clinicopathological information and follow-up and then screened for the presence of LVI. Lymph vessel invasion was defined as the presence of tumor cells in blank spaces enclosed from the endothelium. In the case of LVI identification, the presence of lymphovascular invasion was controlled by an additional D2-40–immunostained slide and on H-E–stained sections. In addition to the standard pathological assessment of the H-E slides in the University Pathology Department, all cases of this study were reviewed by a specialist gynecological pathologist (B.-S.N.). Evaluation of immunohistochemistry was performed by B.-S.N. and R.C. independent from each other in a blinded manner.
Statistical analyses were performed using the Statistical Package for Social Sciences, version 14.0 (SPSS, Chicago, IL). Overall survival (OS) estimations were calculated according to Kaplan-Meier and log-rank tests. Overall survival was considered as the period from first diagnosis until death of any disease. Multivariate analysis was done by Cox regression analysis. Comparisons between the clinical parameters and D2-40 LVI were performed using Fisher exact and χ2 tests. P < 0.05 was considered statistically significant.
A total of 182 patients were available for immunohistochemical analysis. D2-40 LVI was confirmed only if tumor clusters were detectable within a D2-40–positive lymph vessel (Fig. 1). Detailed patient and tumor characteristics are listed in Table 1. The median age at diagnosis of endometrial cancer was 63 years (range, 41-93 years). The median follow-up period was 60 months. Most patients (136/182, 74.6%) were diagnosed at stages confined to the myometrium (FIGO I). A total of 18 patients (9.9%) were histologically staged as FIGO II and 7 (3.8%) as FIGO IIIA, B. Complete lymph node dissection was performed in 86 (47.3%) of 182 patients. Moreover, 21 patients (11.5%) had a positive (FIGO IIIC) and 65 patients (35.7%) patients had a negative lymph node status. The median number of removed lymph nodes was 36 (range, 25-90). Most tumors (n = 132, 72.5%) had moderate tumor differentiation (grade 2), whereas 35 patients (19.2%) revealed a poor tumor grading (grade 3).
Lymph vessel invasion was identified by D2-40 (D2-40 LVI) in 53 tumor specimens (29.1%) compared with 34 tumor specimens (18.3%) by routine H-E staining (H-E LVI; P = 0.001). All cases of D2-40 LVI could be reproduced retrospectively on the serially retrieved H-E–slides. In 3 cases (1.6%), LVI assessed by H-E could not be detected retrospectively by D2-40 immunostaining. In all 3 cases, a false H-ELVI positivity was the result of retraction artifacts, which could be easily identified by D2-40 immunostaining.
D2-40 LVI was statistically correlated with advanced tumor stages, lymph node involvement, and poor tumor differentiation (P < 0.001 each; Table 1). Of 95 patients with staged FIGO IA (tumors confined to the mucosa were excluded), 15 (15.8%) showed LVI detected by D2-40 compared with 8 (19.5%) of 41 patients with D2-40–LVI positivity in FIGO tumor stage IB and 7 (49.9%) of 18 with FIGO stage II. In advanced tumor stages, LVI was observed more frequently. Of 21 patients with lymph node metastasis (FIGO IIIC), 17 (81.0%) were positive for D2-40 LVI. In tumors with moderate differentiation (grade 2), LVI could be detected by D2-40 in 26 (19.7%) of 132 patients compared with 22 (62.9%) of 35 patients with tumor differentiation grade 3. Of 15 patients with grade 1 tumors, 2 (13.3%) showed LVI. Of 63 patients with negative nodal status, 14 (22.2%) showed D2-40–based LVI, whereas in the subgroup with unknown lymph node status, 16 (17.6%) of 91 patients showed D2-40 LVI positivity.
Overall survival of the study group was significantly dependent on the established prognostic parameters like tumor stage, nodal involvement, and tumor differentiation (Fig. 2 and Table 2). D2-40 LVI–positive tumors were significantly associated with reduced OS in univariate and multivariate analyses (Table 2 and Fig. 2). In the subgroup of patients with tumors confined to the inner half of the myometrium (FIGO IA) and an absence of D2-40 LVI, we observed an excellent 5-year OS (97.8%) compared to patients with low myometrial invasion and D2-40 LVI positivity (77.8%, P = 0.007; Fig. 3). D2-40 LVI–negative tumors with unknown lymph node status had a similar 5-year OS (96.3%) compared with D2-40 LVI-negative tumors with negative lymph node status (93.6%; Fig. 3). No survival differences were found between D2-40 LVI status and more advanced tumor stages or G3 tumor differentiation.
Women with endometrial cancer frequently experience serious comorbidities, and the surgical efforts to perform a full lymphadenectomy increase the risks for severe short-term and long-term morbidity. The search for new predictors for lymph node metastases in women with endometrial cancer has outstanding clinical importance. Identifying tumors with high risk for lymph node involvement would distinguish patients in whom a systematic lymphadenectomy might improve survival from others in whom a lymphadenectomy would be without clinical benefit and, moreover, implicate the risk for major surgical complications.
To the best of our knowledge, this study is the first to demonstrate the clinical relevance of D2-40–based evaluation of LVI in endometrial cancer. Our study clearly indicates that LVI predicts lymph node metastasis independent from tumor extension and differentiation. Furthermore, LVI assessed by D2-40 staining was of prognostic value and correlated with reduced OS. D2-40–based evaluation of LVI was superior to conventional H-E–stained analysis. It is remarkable that, in our study, tumor-filled lymph vessels marked by D2-40 could be discovered by H-E retrospectively. This reflects the pathologists’ limited ability to detect lymphatic invasion by routine H-E staining. In the situation that tumor cell groups obliterate the vascular lumen, distinguishing vascular from lymphatic vessels may be particularly difficult on H-E–stained sections.21 Our present results may suggest that D2-40 immunostaining allows a more robust differentiation between lymphoid and blood vessels as well as retraction artifacts.
Immunohistochemical staining of several tumor types with D2-40 could help to detect lymph vessels more accurately and more frequently compared with conventional H-E–based LVI assessment.10,16,22 Our study could detect a D2-40 LVI rate of 29%, which differs significantly from the conventional H-E–based LVI assessment (18%). Data from the literature found a broad range of LVI (14-52%) in endometrial cancer. These different results may be affected by the use of different LVI detection systems and the analysis of varying patient groups with heterogeneous tumor characteristics.
Lymphatic tumor spread in endometrial cancer is a result of tumor progression, which is reflected by a deep myometrial invasion and/or poor tumor differentiation. Lymph node metastasis is found in more than 25% in FIGO IB, grade 3 endometrial carcinoma compared with a lack of lymph node involvement in FIGO IA, grade 1 tumors.23 For studies, a detailed clinicopathological evaluation of the uterus and pelvic and para-aortic lymph nodes (LNDs) is an important precondition to analyze the clinical significance of LVI. However, most of the studies reported have a lack of sufficient data regarding LND assessment.15,24
In our study, patients could be divided into 3 groups: nodal-positive patients with complete LND, nodal-negative patients with complete LND, and patients in whom a complete LND was omitted. More than 80% of the lymph node–positive patients showed D2-40 LVI in our study, whereas 21% of the nodal-negative patients had D2-40 LVI involvement. It is of clinical importance that nodal-negative patients with D2-40 LVI had a significantly reduced OS compared to patients without D2-40 LVI.
In less than 50% of the patients, LND was not performed which reflects the heterogeneity of our study group. Because of a different preoperative clinical and histopathological workup, the indication for a complete systematic LND was inconsistent. However, in these patients, a D2-40 LVI rate of 20% demonstrates that this cohort is more comparable to the nodal-negative subgroup.
Our study indicates that LVI is a feature for malignant aggravation of endometrial cancer. There was a low incidence of LVI in early invasive and well-differentiated tumors, whereas deep myometrial invasion and poorly differentiated tumors were statistically and significantly associated with LVI.
There are conflicting data in the literature regarding the role of LVI in endometrial cancer for the prediction of lymph node metastasis.13,19 In histopathological studies for endometrial cancer, an increased detection rate of D2-40 LVI was reported.24 A recent study from Vandenput et al19 reported no correlation between D2-40 LVI and lymph node metastasis in 62 patients with endometrial cancer. They found 52% of the tumors exhibiting D2-40 LVI. Their results are in contrast to the findings of this study and might be the result of a patient group more likely to have advanced tumor stages and a high percentage of lymph node metastasis. However, these contradictory results might also reflect the methodical problems of LVI assessment for the routine clinical use. D2-40 (podoplanin) is the most reliable immunohistochemical marker for lymph vessels, which has a high specificity and sensitivity.6 However, to identify tumor cells within the lymph vessel might be challenging. Therefore, it is of great importance to count only lymph vessels filled with tumor clusters.
To focus more strongly on a routine pathological workup on lymphatic tumor involvement of endometrial cancer, it may in the future be helpful to select patients at risk. Recently, a distinctive pattern of myometrial invasion has been described25 and designated by the acronym MELF (microcystic, elongated, and fragmented), reflecting a feature of more aggressive and lymph-invading endometrial cancer.26 For standardized LVI evaluation, it would be useful to select patients at risk for LVI. If MELF would be routinely be assessed, tumors with the feature of MELF may be candidates to analyze LVI in more detail, for example, by using D2-40 immunohistochemistry.
The results of our study suggest D2-40–based LVI assessment to be an important feature for endometrial cancer. It might contribute to the therapeutic concept to perform systematic lymphadenectomy only secondarily after the pathological workup of the uterus would predict lymph node metastasis. Therefore, upcoming prospective trials have to analyze prospectively several markers like tumor size, extent of myometrial invasion, tumor differentiation, and D2-40 LVI to predict lymph node involvement.
1. Jemal A, Siegel R, Xu J, et al.. Cancer statistics, 2010. CA Cancer J Clin
. 2010; 60: 277–300.
2. Amant F, Moerman P, Neven P, et al.. Endometrial cancer. Lancet
. 2005; 366: 491–505.
3. Mariani A, Webb MJ, Keeney GL, et al.. Predictors of lymphatic failure in endometrial cancer. Gynecol Oncol
. 2002; 84: 437–442.
4. Hoekstra AV, Kim RJ, Small W Jr, et al.. FIGO stage IIIC endometrial carcinoma: prognostic factors and outcomes. Gynecol Oncol
. 2009; 114: 273–278.
5. Leitao MM, Barakat RR. Advances in the management of endometrial carcinoma. Gynecol Oncol
. 2011; 120: 489–492.
6. Van der Auwera I, Cao Y, Tille JC, et al.. First international consensus on the methodology of lymphangiogenesis quantification in solid human tumours. Br J Cancer
. 2006; 95: 1611–1625.
7. Van der Auwera I, Colpaert C, Van Marck E, et al.. Lymphangiogenesis in breast cancer. Am J Surg Pathol
. 2006; 30: 1055–1056; author reply 1056–1057.
8. Yonemura Y, Endou Y, Tabachi K, et al.. Evaluation of lymphatic invasion in primary gastric cancer by a new monoclonal antibody, D2-40. Hum Pathol
. 2006; 37: 1193–1199.
9. Quek ML, Stein JP, Nichols PW, et al.. Prognostic significance of lymphovascular invasion of bladder cancer treated with radical cystectomy. J Urol
. 2005; 174: 103–106.
10. Braun M, Wardelmann E, Debald M, et al.. Detection of lymphovascular invasion in vulvar cancer by D2-40 (podoplanin) as a predictor for inguinal lymph node metastases. Onkologie
. 2009; 32: 732–738.
11. Goldhirsch A, Glick JH, Gelber RD, et al.. Meeting highlights: international expert consensus on the primary therapy of early breast cancer 2005. Ann Oncol
. 2005; 16: 1569–1583.
12. Tsuruchi N, Kaku T, Kamura T, et al.. The prognostic significance of lymphovascular space invasion in endometrial cancer when conventional hemotoxylin and eosin staining is compared to immunohistochemical staining. Gynecol Oncol
. 1995; 57: 307–312.
13. Briet JM, Hollema H, Reesink N, et al.. Lymphvascular space involvement: an independent prognostic factor in endometrial cancer. Gynecol Oncol
. 2005; 96: 799–804.
14. Watari H, Todo Y, Takeda M, et al.. Lymph-vascular space invasion and number of positive para-aortic node groups predict survival in node-positive patients with endometrial cancer. Gynecol Oncol
. 2005; 96: 651–657.
15. Alexander-Sefre F, Singh N, Ayhan A, et al.. Detection of tumour lymphovascular space invasion using dual cytokeratin and CD31 immunohistochemistry. J Clin Pathol
. 2003; 56: 786–788.
16. Kahn HJ, Marks A. A new monoclonal antibody, D2-40, for detection of lymphatic invasion in primary tumors. Lab Invest
. 2002; 82: 1255–1257.
17. Chu AY, Litzky LA, Pasha TL, et al.. Utility of D2-40, a novel mesothelial marker, in the diagnosis of malignant mesothelioma. Mod Pathol
. 2005; 18: 105–110.
18. Braun M, Flucke U, Debald M, et al.. Detection of lymphovascular invasion in early breast cancer by D2-40 (podoplanin): a clinically useful predictor for axillary lymph node metastases. Breast Cancer Res Treat
. 2008; 112: 503–511.
19. Vandenput I, Vanhove T, Calster BV, et al.. The use of lymph vessel markers to predict endometrial cancer outcome. Int J Gynecol Cancer
. 2010; 20: 363–367.
20. Emons G, Kimmig R. Interdisciplinary S2k guidelines on the diagnosis and treatment of endometrial carcinoma. J Cancer Res Clin Oncol
. 2009; 135: 1387–1391.
21. Niakosari F, Kahn HJ, Marks A, et al.. Detection of lymphatic invasion in primary melanoma with monoclonal antibody D2-40: a new selective immunohistochemical marker of lymphatic endothelium. Arch Dermatol
. 2005; 141: 440–444.
22. Arigami T, Natsugoe S, Uenosono Y, et al.. Lymphatic invasion using D2-40 monoclonal antibody and its relationship to lymph node micrometastasis in pN0 gastric cancer. Br J Cancer
. 2005; 93: 688–693.
23. Chi DS, Barakat RR, Palayekar MJ, et al.. The incidence of pelvic lymph node metastasis by FIGO staging for patients with adequately surgically staged endometrial adenocarcinoma of endometrioid histology. Int J Gynecol Cancer
. 2008; 18: 269–273.
24. Mannelqvist M, Stefansson I, Salvesen HB, et al.. Importance of tumour cell invasion in blood and lymphatic vasculature among patients with endometrial carcinoma. Histopathology
. 2009; 54: 174–183.
25. Murray SK, Young RH, Scully RE. Unusual epithelial and stromal changes in myoinvasive endometrioid adenocarcinoma: a study of their frequency, associated diagnostic problems, and prognostic significance. Int J Gynecol Pathol
. 2003; 22: 324–333.
26. Quick CM, May T, Horowitz NS, et al.. Low-grade, low-stage endometrioid endometrial adenocarcinoma: a clinicopathologic analysis of 324 cases focusing on frequency and pattern of myoinvasion. Int J Gynecol Pathol
. 2012; 31: 337–343.