Journal of Cancer Research and Clinical Oncology

https://doi.org/10.1007/s00432-018-2628-2

ORIGINAL ARTICLE – CLINICAL ONCOLOGY

Ovarian metastasis in patients with endometrial cancer: risk factors

and impact on survival
Tanja Ignatov1 · Holm Eggemann1 · Elke Burger2 · Olaf Ortmann3 · Serban Dan Costa1 · Atanas Ignatov1,3 

Received: 17 February 2018 / Accepted: 12 March 2018

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Abstract
Background  Oophorectomy is generally performed in patients with endometrial cancer despite the rate of ovarian metastasis
being relatively low.
Patients and methods  A multicenter retrospective registry-based study was performed in 2329 patients with endometrial
cancer. The outcome measures were the incidence of ovarian metastasis and the impact on overall survival.
Results  Median follow-up was performed at 84 months. A total of 2158 women were eligible for analysis, of which 131
(6.1%) had ovarian metastasis. Women with ovarian metastasis were more likely to have > 50% myometrial invasion, undif-
ferentiated nonendometrioid tumors, and lymph and vascular space invasion. The presence of < 50% myometrial invasion,
endometrioid histology, well-differentiated cancer, and negative lymph and vascular space invasion were associated with a
very low rate (0.5%) of ovarian metastasis. Notably, after matching for tumor histology and grade, myometrial invasion, and
lymph and vascular space invasion, ovarian metastasis was not associated with a reduced median overall survival.
Conclusions  Ovarian preservation should be offered to premenopausal women with endometrial cancer in whom myome-
trial invasion is less than 50%, the histological type is endometrioid and well-differentiated, and lymph and vascular space
invasion is not involved.

Keywords  Endometrial cancer · Ovarian · Metastasis

Introduction reports have demonstrated that preservation of the ovaries is

safe and not associated with increased mortality in women
Endometrial cancer (EC) is the most common cancer in with early stage EC (Lau et al. 2014; Lee et al. 2007; Wright
women in developed countries (Dumas et al. 2016), and et al. 2009); however, the decision to preserve the ovaries has
surgery with hysterectomy and oophorectomy is the stand- to be taken carefully, as ovarian metastasis has been reported
ard treatment. The latter is performed to stage the EC and in up to 10% of cases (Gemer et al. 2004; Gilani; Modaress
to reduce estrogen production, as EC is estrogen-dependent. et al. 2011; Hetu et al. 2009; Takeshima et al. 1998). Efforts
Up to 10% of patients with EC are premenopausal (Lee et al. to identify the risk factors associated with ovarian metastasis
2007; Pellerin and Finan 2005; Wright et al. 2009). Some in patients with EC have been undertaken in retrospective
series with limited numbers of patients (Gemer et al. 2004;
Gilani; Modaress et al. 2011; Takeshima et al. 1998).
Tanja Ignatov and Holm Eggemann have equally contributed. Therefore the aims of this study were to analyze the risk
factors associated with ovarian metastasis in patients with
* Atanas Ignatov
atanas.ignatov@gmail.com EC in a large retrospective cohort register study and to inves-
tigate the impact on overall survival.
1
Department of Gynecology and Obstetrics, Otto-
von-Guericke University, G.‑Hauptmann Str. 35,
39108 Magdeburg, Germany
2
Institute of Biometry and Medical Informatics,
Otto-von-Guericke University, Magdeburg, Germany
3
Department of Gynecology and Obstetrics, University
Medical Center Regensburg, Regensburg, Germany

13
Vol.:(0123456789)

Patients and methods
The cancer registry of Saxony-Anhalt, a federal state in Ger-
many, was reviewed to determine all patients with EC treated
between 2000 and 2014. This tumor registry holds informa-
tion on the diagnosis, age at diagnosis, date of diagnosis,
tumor stage, tumor grade, lymph node status, date of disease
recurrence, date of death, suggested treatment regimens, and
treatment regimens used (Eggemann et al. 2017). Informa-
tion about the date and cause of death is routinely entered
into the system shortly after death. This cohort study aimed
to analyze women with EC diagnosed between 2000 and
2014 in ten hospitals in Saxony-Anhalt: University Hospital
Magdeburg, Harzklinikum Dorothea Christiane Erxleben,
Johanniter Clinic Genthin-Stendal, HELIOS Clinic Burg,
Klinikum Olvenstedt Magdeburg, Klinik Marienstift Magde-
burg, and the AMEOS Clinics in Aschersleben, Halberstadt,
Haldensleben, and Schönebeck. Patients were excluded if
no information was available regarding ovarian metastasis.
Written informed consent was obtained from all patients
before treatment. Patient data were recorded whilst blind
to the patient name and date of birth. In accordance with a
statement from the Research and Ethical Committee, Otto-
von-Guericke University, Magdeburg, Germany, additional
individual consent was not required for this analysis. The
manuscript was prepared in accordance with the STROBE
statement criteria (von Elm et al. 2007).
To avoid selection bias, patients who underwent surgery
prior to 2009 were restaged according to the 2009 FIGO
criteria (Eggemann et al. 2016). To avoid further selection
bias regarding treatment in the survival analysis, matching
analysis was performed for patients with different demo-
graphic and clinical characteristics. The matching process
was based on five prognostic criteria: depth of myometrial
invasion, histological type, histological grade, lymph space
involvement, and vascular space involvement. The match-
ing procedure was conducted at random and without any
information on patient outcome.
The primary outcome measures were the risk factors asso-
ciated with ovarian metastasis and overall survival (OS). OS
was used as a primary outcome, because information about
a patient’s death and its cause was automatically recorded in
the cancer registry via the civil registry office, thus leading
to minimal loss of follow-up and keeping transfer bias to a
minimum (Eggemann et al. 2016). OS was defined as the
time from the date of diagnosis to the date of death (from
any cause). The follow-up ended with the patient’s death, the
last available information in the tumor registry, or the last
follow-up on July 19, 2017.
13
Journal of Cancer Research and Clinical Oncology
Statistical analysis
The study was designed as a retrospective cohort study
with high external validity. The statistical calculations
were performed using SPSS version 22.0 (SPSS, Chicago,
IL, USA). The correlation of variables and the distribu-
tion of clinical, pathological, and treatment characteris-
tics were assessed using Chi-squared tests. Survival prob-
ability was studied using the Kaplan–Meier method. The
equality of survival curves was tested using the log-rank
test. Cox proportional hazards models were used to assess
the influence of different clinical and pathological factors
on ovarian metastasis. Statistical analyses were two-sided,
and P < 0.05 was considered statistically significant.
Results
A total of 2392 women with EC were treated in the afore-
mentioned hospitals between January 2000, and Decem-
ber 2014, and were, therefore, eligible for analysis. The
flow diagram in Fig. 1 shows the study design and the
number of patients at each stage of the study. Of the total
number of patients, 234 were excluded from the analysis
because of missing information regarding ovarian involve-
ment. Therefore, 2158 women were eligible for analysis,
of which 131 (6.1%) had ovarian metastasis. The matching
analysis was performed in 142 patients (Fig. 1).
The median age of the study population was 69 years
(range 29–96 years), and the median follow-up time was
84 months (range 0–213 months). The clinical and patho-
logical characteristics are presented in Table 1. Women
with ovarian metastasis were more likely to have > 50%
myometrial invasion (92.9%), undifferentiated grade 3
tumors (55.9%), nonendometrioid histology (16%), and
lymph (78.7%) and vascular (37.3%) space invasion when
compared to subjects without ovarian metastasis. No sig-
nificant difference was found regarding the age of diagno-
sis (P = 0.394).
The associations between the risk of ovarian metastasis
and the depth of myometrial invasion, tumor histology,
tumor grade, lymph space invasion, and vascular space
invasion were investigated using multivariate analysis.
Myometrial invasion, histological type, and lymph space
invasion were associated with an increased risk of ovar-
ian metastasis (Table 2). In the case of deep myometrial
invasion (> 50%), the risk of ovarian metastasis was
increased 66-fold [odds ratio (OR) 66.46; 95% confidence
interval (CI) 15.87–278.33]. The risk of ovarian metasta-
sis was increased 3.7-fold [OR 3.74; 95% CI 1.49–9.39]
and 2.6-fold [OR 2.61; 95% CI 1.37–4.99] in women with
Journal of Cancer Research and Clinical Oncology

Fig. 1  Study design
2392 patients assessed for inclusion

234 had no information about

ovarian metastases

2158 eligible for analysis

2027 without ovarian metastasis 131 with ovarian metastasis

matching for myometrial invasion, histological type and grade, lymph and
vascular space involvement

142 eligible for survival analysis

71 without ovarian metastais 71 with ovarian metastasis

Table 1  Clinical and pathological characteristics Table 2  Multivariate analysis of risk factors for ovarian metastasis
Varaible Ovarian metastasis P value Variable OR (CI 95%) P value
Negative Positive Myometrial invasion
Total 2027 (84.7%) 131 (6.1%)  < 50% Reference
Age at diagnosis (years) 67.8 (31–96) 68.6 (29–95) 0.394  > 50% 66.46 (15.87–278.33) 0.0001
Myometrial ivasion Histological grade
 < 50% 1528 (75.5%) 9 (7.1%) 0.0001  1 Reference
 > 50% 499 (24.6%) 118 (92.9%)  2, 3 1.13 (0.74–1.72) 0.580
Histological grade Histological type
 1 717 (35.6%) 16 (12.6%) 0.0001  Endometrioid Reference
 2 908 (45.1%) 40 (31.5%)  Nonendometrioid 3.74 (1.49–9.39) 0.005
 3 389 (19.3%) 71 (55.9%) Lymph space invasion
Histological type  Negative Reference
 Endometrioid 1938 (95.8%) 110 (84.0%) 0.0001  Positive 2.61 (1.37–4.99) 0.004
 Nonendometrioid 85 (4.2%) 21 (16.0%) Vascular space invasion
Lymph space invasion  Negative Reference
 Negative 890 (73.9%) 19 (21.3%) 0.0001  Positive 1.79 (0.94–3.42) 0.077
 Positive 314 (26.1%) 70 (78.7%)
Vascular space invasion
 Negative 1083 (93.4%) 47 (62.7%) performed. The predictive value of different risk factors
 Positive 76 (6.6%) 28 (37.3%) 0.0001 was also analyzed. The presence of myometrial invasion
(< 50%), endometrioid histological type, well-differenti-
ated cancer, and negative lymph and vascular space inva-
nonendometrioid histology and lymph space invasion, sion was associated with a very low rate (0.5%) of ovar-
respectively (Table 2); however, histological grade and ian metastasis (Table 3). Conversely, > 50% myometrial
vascular space invasion were not independent prognostic invasion, nonendometrioid histological type, tumor grade
factors after adjustment for all prognostic variables was 2 or 3, and positive lymph and vascular space invasion

13
 Journal of Cancer Research and Clinical Oncology

Table 3  Predictive value of different risk factors for ovarian metas- and 42.3%, respectively. This difference was not statistically
tasis significant.
Risk factors Ovarian metastasis P value
Negative Positive
Discussion
Risk group 371 (99.5%) 2 (0.5%) 0.0001
 Low risk This large register study showed that 6.1% of patients with
  Myometrial invasion < 50%
  Endometrioid histology EC had ovarian metastasis. It also strongly demonstrated
  Grade 1 that < 50% myometrial invasion, endometrioid histology,
  Lymph space invasion nega- well-differentiated cancer, and negative lymph and vascular
tive space invasion were associated with a very low rate of ovar-
  Vascular space invasion nega-
tive ian metastasis.
 High risk 1427 (91.7%) 129 (8.3%) Deep myometrial invasion was the strongest predictive
  Myometrial invasion > 50% factor for ovarian metastasis, with > 50% invasion being
  Nonendometrioid histology associated with a 66-fold increased risk of ovarian metas-
  Grade 2, 3 tasis. Similar results were also observed in a retrospective
  Lymph space invasion nega-
tive study of 210 patients with EC (Gilani Modaress et al. 2011).
  Vascular space invasion Gemer et al. also found that the risk of ovarian metastasis
negative significantly increased in women with > 50% myometrial
Low risk vs high risk 16.77 (95%CI 4.13–68.10) 0.0001 invasion (Gemer et al. 2004). Interestingly, these groups
also showed that tumor grade was an additional prognostic
factor for ovarian metastasis. In the present cohort, tumor
100
grade was predictive in a univariate analysis but not in a
p=0.079 multivariate analysis. This discrepancy could be due to all
investigated tumors in the aforementioned studies being
80 endometrioid and mostly well differentiated (Gemer et al.
Overall survival (%)

2004). This is important as nonendometrioid pathology is

60
40
20
Without ovarian metastasis
With ovarian metastasis
0
0 2 4 6 8 10
Time (years)
Fig. 2  Overall survival for women with endometrial cancer depend-
ing on ovarian metastasis
increased the risk of ovarian metastasis 16.8-fold, and
ovarian metastasis was observed in 129 (8.3%) of 1556
patients (Table 3).
The effect of ovarian involvement on OS was investigated
using Kaplan–Meier analysis (Fig. 2). To avoid confounder
and selection bias, matching analysis was performed based
on the aforementioned prognostic criteria. A total of 142
patients were identified and included in a further survival
analysis. Notably, the presence of ovarian metastasis was
not associated with reduced OS in this population. Patients
with and without ovarian metastasis had an OS of 28.2%
an independent predictive factor and a well-known adverse
prognostic factor (Eggemann et al. 2017). The third inde-
pendent risk factor in the multivariate analysis was lymph
space invasion. It was associated with a 2.6-fold increased
risk of ovarian metastasis and was in accordance with other
studies (Gemer et al. 2004; Gilani; Modaress et al. 2011).
Thus myometrial invasion, histological type, and lymph
space invasion are strong predictors for ovarian metastasis
12 14 16 18 20 in patients with EC. The absence of these factors reduced
the risk of ovarian metastasis to below 0.5%. This should be
taken in consideration if ovarian preservation is requested.
Recently, Wright et al. demonstrated that ovarian preserva-
tion in premenopausal patients with EC is safe and does not
increase the cancer-specific mortality (Wright et al. 2009).
Additional two studies also suggested that ovarian preserva-
tion may be considered in premenopausal women with the
early stage EC as it was not associated with an increased risk
of mortality (Lau et al. 2014; Lee et al. 2007). This finding,
along with the observation that the early oophorectomy is
associated with increased all-cause mortality (Larson 2011),
should be considered in the case of operative treatment of
premenopausal women with EC. Moreover, ovarian metas-
tasis was found to not be an independent prognostic factor
for OS. In the matching analysis, the presence of ovarian
metastasis did not significantly influence the OS of patients
with EC. Therefore, it appears that ovarian metastasis is only

13
Journal of Cancer Research and Clinical Oncology
a confounder with regards to OS and it does not appear to
be associated with adverse survival effects (Lau et al. 2014;
Lee et al. 2007; Wright et al. 2009). In our study, the sur-
vival of patients with and without ovarian metastasis was
28.2% and 42.3%, respectively. Interestingly, this difference
was not statistically significant and could be due to the low
number of patient investigated. Nevertheless, the preserva-
tion of ovaries is relevant for premenopausal women. In the
case of postmenopausal women, the ovariectomy should be
considered as standard.
The limitations of this study include: (1) its retrospective
nature; (2) its lack of central reference pathology; and (3)
that the disease-free survival and the rate of recurrence were
not investigated. Conversely, the strengths of the study were:
(1) the large sample size and long follow-up; (2) the high
level of external validity, as treatment decisions were investi-
gated under real clinical conditions (e.g., it was multicentric,
the study population was similar to the general population,
and the exclusion criteria were kept to a minimum); (3) the
minimal loss to follow-up regarding OS; and (4) the docu-
mentation of the clinical and pathological characteristics of
the patients.
In conclusion, the present study suggests that the ovaries
might be preserved in women with the early stage endome-
trioid EC if the disease is well differentiated, without lymph
and vascular space invasion, and involves < 50% myometrial
involvement. The treatment decision should balance the risk
factors and should be discussed with the patient.
Funding  This study was not funded.
Compliance with ethical standards  of age or younger: a clinicopathological analysis. Am J Obstet
Conflict of interest  All authors declare that they have no conflict of
interest.
Ethical approval  This article does not contain any studies with animals
performed by any of the authors. In accordance with the statement of
the Research and Ethical Committee of the Otto-von-Guericke Uni-
versity, Magdeburg, Germany, additional individual consent for this
analysis was not needed. All procedures performed in studies involving
human participants were in accordance with the ethical standards of
the institutional and/or national research committee and with the 1964
Helsinki declaration and its later amendments or comparable ethical
standards.
Informed consent  Written informed consent was obtained from all
patients before treatment. An additional individual consent for this
analysis was not needed.
References
Dumas L, Ring A, Butler J, Kalsi T, Harari D, Banerjee S (2016)
Improving outcomes for older women with gynaecological malig-
nancies. Cancer Treat Rev 50:99–108
Eggemann H, Ignatov T, Kaiser K, Burger E, Costa SD, Ignatov A
(2016) Survival advantage of lymphadenectomy in endometrial
cancer. J Cancer Res Clin Oncol 142(5):1051–1060
Eggemann H, Ignatov T, Burger E, Costa SD, Ignatov A (2017) Man-
agement of elderly women with endometrial cancer. Gynecol
Oncol 146(3):519–524
Gemer O, Bergman M, Segal S (2004) Ovarian metastasis in women
with clinical stage I endometrial carcinoma. Acta Obstet Gynecol
Scand 83(2):208–210
Hetu V, Petignat P, Wu Y, Drouin P, Sauthier P, Provencher D, Gauth-
ier P (2009) Positive adnexal or uterine serosal involvement in
stage IIIC endometrial cancer is an adverse factor for recurrence.
Gynecol Obstet Invest 67(3):173–177
Larson CA (2011) Evidence-based medicine: an analysis of prophy-
lactic bilateral oophorectomy at time of hysterectomy for benign
conditions. Curr Oncol 18(1):13–15
Lau HY, Twu NF, Yen MS, Tsai HW, Wang PH, Chuang CM, Wu
HH, Chao KC, Chen YJ (2014) Impact of ovarian preserva-
tion in women with endometrial cancer. J Chin Med Assoc
77(7):379–384
Lee NK, Cheung MK, Shin JY, Husain A, Teng NN, Berek JS, Kapp
DS, Osann K, Chan JK (2007) Prognostic factors for uterine can-
cer in reproductive-aged women. Obstet Gynecol 109(3):655–662
Modaress MG, Cheraghi F, Zamani N (2011) Ovarian metasta-
sis in endometriod type endometrial cancer. Int J Fertil Steril
5(3):148–151
Pellerin GP, Finan MA (2005) Endometrial cancer in women 45 years
Gynecol 193(5):1640–1644
Takeshima N, Hirai Y, Yano K, Tanaka N, Yamauchi K, Hasumi K
(1998) Ovarian metastasis in endometrial carcinoma. Gynecol
Oncol 70(2):183–187
von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Van-
denbroucke JP, Initiative S (2007) The strengthening the report-
ing of observational studies in epidemiology (STROBE) state-
ment: guidelines for reporting observational studies. Lancet
370(9596):1453–1457
Wright JD, Buck AM, Shah M, Burke WM, Schiff PB, Herzog TJ
(2009) Safety of ovarian preservation in premenopausal women
with endometrial cancer. J Clin Oncol 27(8):1214–1219
13