Archives of Gynecology and Obstetrics

https://doi.org/10.1007/s00404-018-4789-2

GYNECOLOGIC ONCOLOGY

Does tumor grade influence the rate of lymph node metastasis

in apparent early stage ovarian cancer?
Dimitrios Nasioudis1   · Spyridon A. Mastroyannis1 · Emily M. Ko1 · Nawar A. Latif1

Received: 25 December 2017 / Accepted: 16 May 2018

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

Abstract
Purpose  To evaluate the prevalence of regional lymph node (LN) metastasis in patients with non-clear cell epithelial ovarian
cancer apparently confined to the ovary, stratified by tumor grade.
Methods  The National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) database was accessed
(1988–2014). We identified patients with epithelial ovarian carcinoma of serous, endometrioid and mucinous histology
apparently confined to the ovary who underwent extensive lymphadenectomy (defined as at least 20 lymph nodes removed).
Demographics, tumor histology, grade and lymph node status were collected. Comparisons were made with Chi square and
Mann–Whitney U tests.
Results  A total of 1242 women met the inclusion criteria. Endometrioid adenocarcinoma was the most common histology
(564 patients (45.4%)) while 443 (35.7%) and 235 (18.9%) patients had serous, and mucinous adenocarcinoma, respectively.
The rate of LN metastasis in low-grade serous was 9.0% (6/67) vs. 14.4% (54/376) in high-grade serous histology (OR, 1.71,
95% CI 0.70, 4.14, p = 0.24). In patients with low-grade endometrioid tumors, the rate of LN metastasis was 1.7% (7/407)
vs. 5.1% (8/157) observed in those with high-grade tumors (OR: 3.07, 95% CI 1.09, 8.61, p = 0.033). Lastly, the rate of LN
metastasis in mucinous histology was 1.7% (3/177) in low-grade vs. 8.6% (5/58) in high-grade tumors (OR: 5.47, 95% CI
1.27, 23.66, p = 0.024).
Conclusions  Regional LN metastasis in apparent stage I low-grade mucinous and endometrioid ovarian tumors is infrequent.

Keywords  Ovarian cancer · Grade · Lymph node · Lymphadenectomy

Introduction recommendations. Patients with favorable characteristics

(stage IA, low-grade, non-clear cell histology) can be safely
Ovarian cancer is currently the fifth leading cause of can- spared of adjuvant chemotherapy [3]. However, systematic
cer-related death among women in the United States [1]. LN dissection of pelvic and para-aortic LNs can be associ-
Approximately 30% of the patients with (EOC) present ated with a prolonged operative time, increased blood loss
with disease confined to the ovary or pelvis [2]. According and significant morbidity [4]. The aim of the present report
to FIGO guidelines every patient diagnosed with ovarian was to investigate the prevalence of regional LN metas-
cancer should undergo a complete surgical staging proce- tasis in apparent early-stage non-clear cell EOC stratified
dure that includes omentectomy, biopsy of all suspicious by tumor histology and grade, using a multi-institutional
lesions, peritoneal washings and lymph node sampling/dis- population-based database.
section (LND). Surgical staging provides valuable informa-
tion on tumor dissemination thus guiding adjuvant treatment
Materials and methods
* Dimitrios Nasioudis The National Cancer Institute’s SEER database was accessed
dimitrios.nasioudis@uphs.upenn.edu
and using the ICD-O-3/WHO 2008 site code “C.569/ovary”
1
Department of Obstetrics and Gynecology, Hospital a cohort of patients with ovarian cancer diagnosis between
of the University of Pennsylvania, Helen O’Dickens 1988 and 2014 was abstracted. Data deriving from 18 can-
Women’s Health Center, 3400 Spruce Street, Philadelphia, cer registries as released on April 2017 were used [5]. All
PA, USA

13
Vol.:(0123456789)
 Archives of Gynecology and Obstetrics

patient data are de-identified and available to the public for (49.9%) compared to those with endometrioid (34.8%) and
research purposes. The ICD-O-3 morphology codes “8441, mucinous (29.4%) tumors, p < 0.001.
8460–8463, 9014”, “8470–8490, 9015” and “8380–8381, The median number of LN removed and examined was 27
8560, 8570” as grouped by the International Agency for (IQR:10). No statistically significant difference in the extent
Research on Cancer (IARC) were used to identify women of LND was found based on tumor grade following stratifi-
with serous, mucinous and endometrioid ovarian tumors, cation by histology. Rate of LN metastasis was 6.7% while
respectively [6]. Using the Collaborative staging (CS) and median number of positive LNs removed was 2 (IQR:4).
the Extent of Disease (EOD)-10 fields, those with disease Median lymph node ratio (LNR, ratio of the number of
apparently confined to the ovary and documented absence of positive lymph nodes to the total number of lymph nodes
distant metastasis were selected for further analysis. Patients removed) was 0.048 (IQR: 0.13). Figure 1 depicts the dis-
who did not undergo cancer-directed surgery, those who tribution of LNR in our cohort. The rate of LN metastasis in
were upstaged due to microscopic peritoneal implants, cases low-grade serous tumors was 9.0% (6/67) vs. 14.4% (54/376)
without histologic confirmation or without tumor grade in high-grade serous histology (OR, 171, 95% CI 0.70, 4.14,
information were excluded. In addition, from the present p = 0.24). For patients with low-grade endometrioid tumors,
study we opted to exclude clear cell tumors given that they the rate of LN metastasis was 1.7% (7/407) vs. 5.1% (8/157)
are currently not deemed histologically gradable. Tumor observed in those with high-grade tumors (OR: 3.07, 95%
grade was recoded from a 4-tier into a 3-tier system that CI 1.09, 8.61, p = 0.033). Lastly, the rate of LN metastasis in
included grade 1 (well differentiated), grade 2 (moderately mucinous histology was 1.7% (3/177) in low-grade vs. 8.6%
differentiated) and grade 3 (poorly differentiated and undif- (5/58) in high-grade tumors (OR: 5.47, 95% CI 1.27, 23.66,
ferentiated, combined). We opted to exclude from our analy- p = 0.024). Table 2 summarizes the rate of LN metastasis
sis cases with grade 2 tumors given significant variations in stratified by tumor histology and grade.
their classification into low and high grade [7]. Information
derived from the histopathology report was employed to
identify women who underwent lymphadenectomy (LND). Discussion
In the present report, we opted to include only patients who
had extensive LND defined as at least 20 LNs removed and The results of the present report further support the grow-
examined. Comparison of categorical and continuous vari- ing body of evidence that the prevalence of occult regional
ables was made with the Fisher’s exact or Chi square tests LN metastases in low grade, endometrioid and mucinous
and Mann–Whitney U test, respectively while odds ratio and EOC apparently confined to the ovary is very low. A recent
95% confidence intervals were calculated. Statistical analysis meta-analysis of retrospective studies that included data
was performed with the SPSS v.24 statistical package. The from 273 patients with apparent stage I–II, low-grade
alpha level of statistical significance was set at 0.05. ovarian tumors (of any histology) who underwent sys-
tematic LND revealed that rate of occult LN metastases
was 2.9% [8]. However, in that study, limited information
Results on tumor histology was available. In a multi-center study
that examined 107 patients with apparent stage I disease
A total of 1242 women met the inclusion criteria. The and low-grade mucinous and endometriod histology who
median age at diagnosis was 53 years (range 18–88), and the underwent systematic LND, LN metastasis was extremely
majority of patients were White (84.9%). Endometrioid ade- rare occurring in only one patient [9]. In the same study,
nocarcinoma was the most common histology [564 patients contrary to our results, the prevalence of LN metasta-
(45.4%)] while 443 (35.7%) and 235 (18.9%) patients had sis among patients with low-grade serous and apparent
serous, and mucinous adenocarcinoma, respectively. In our stage I disease was 2.4%, which was significantly lower
cohort, 84.9% of serous tumors were high grade compared than the rate observed in our cohort (9.0%). It should be
to 27.8% of endometrioid and 24.7% of mucinous tumors underlined that low and high grade serous tumors are
(p < 0.001). A total of 633 (51%) patients had apparent stage characterized by unique molecular profiles and clinico-
IA disease while 486 (39.1%), 105 (8.5%) and 18 (1.4%) had pathological characteristics indicating that these tumors
apparent IC, IB and INOS (not otherwise specified) disease, are two distinct entities and not the ends of the same dis-
respectively. Table 1 summarizes the clinico-pathological ease [10]. Interestingly, a previous analysis of the SEER
characteristics of the study population stratified by tumor database failed to demonstrate a survival benefit of LND
grade. Patients with serous histology were older (median for patients with apparent stage I mucinous tumors [11].
age 57 years) than those with endometrioid and mucinous As such, omitting LND (in the absence of grossly abnor-
tumors (51 and 50 years, respectively p < 0.001) and pre- mal lymph nodes) when staging low grade mucinous and
sented more commonly with apparent stage IC disease endometrioid EOCs may potentially shorten operative time

13
Archives of Gynecology and Obstetrics

Table 1  Clinical characteristics Low-grade High-grade p value* Overall

of study cohort stratified by
tumor grade Age (median, IQR, years) 51 (16) 56 (15) < 0.001 53 (16)
Age (years) < 0.001
 <50 290 (44.5%) 155 (26.2%) 445 (35.8%)
 ≥ 50 361 (55.1%) 436 (73.8%) 797 (64.2%)
Race 0.88
 White 551 (84.6%) 504 (85.3%) 1055 (84.9%)
 Black 24 (3.7%) 23 (3.9%) 47 (3.8%)
 Other/unknown 76 (11.7%) 64 (10.8%) 140 (11.3%)
Histology < 0.001
 Serous 67 (10.3%) 376 (63.6%) 443 (35.7%)
 Endometrioid 407 (62.5%) 157 (26.6%) 564 (45.4%)
 Mucinous 177 (27.2%) 58 (9.8%) 235 (18.9%)
Laterality < 0.001
 Unilateral 589 (90.5%) 442 (74.8%) 1031 (83%)
 Bilateral 62 (9.5%) 149 (25.2%) 211 (17%)
Substage
 IA 411 (63.1%) 222 (37.6%) < 0.001 633 (51%)
 IB 33 (5.1%) 72 (12.2%) 105 (8.5%)
 IC 197 (30.3%) 289 (48.9%) 486 (39.1%)
 INOS 10 (1.5%) 8 (1.4%) 18 (1.4%)
LN removed (median, IQR) 27 (11) 27 (10) 0.28 27 (10)
LN status
 Negative 635 (97.5%) 524 (88.7%) < 0.001 1159 (93.3%)
 Positive 16 (2.5%) 67 (11.3%) 83 (6.7%)
LN positive (median, IQR) 1 (1) 2 (4) 0.64 2 (4)
LN ratio (median, IQR) 0.045 (0.05) 0.071 (0.16) 0.14 0.048 (0.13)

Bold values indicate p < 0.05

*p value from Chi square or Mann–Whitney U test

and minimize post-operative complications without jeop-

ardizing the oncological outcomes of these patients. The
above data support omitting a second surgery to perform
LND after a first procedure in which low-grade endometri-
oid or muinous ovarian cancer is discovered in an adnexal
mass. The challenge, however, may be reliably diagnosing
low-grade histology based on frozen section analysis at
the time of surgery. A complete staging procedure should
be performed when a malignancy is diagnosed, but there
is uncertainty regarding tumor histology or grade on the
frozen section. The limitations of the current study is the
retrospective nature and the lack of specific information on
the surgical staging technique employed at each reporting
site as well as the exact anatomical location (pelvic and/
or para-aortic) of the harvested LNs as well as their gross
appearance. Lastly, variations in specimen handling by
different pathology laboratories may have influenced the
final reported LN count. While given the lack of central
pathology report possible tumor misclassification cannot
be excluded. Regional lymph node involvement in appar-
Fig. 1  Lymph node ratio distribution in study population ent stage I low-grade mucinous and endometrioid ovarian

13
 Archives of Gynecology and Obstetrics

Table 2  Rate of regional lymph Serous Endometrioid Mucinous

node (LN) metastasis stratified
by tumor histology and grade Low grade High grade Low grade High grade Low grade High grade

LN status
 Positive 6 (9%) 54 (14.4%) 7 (1.7%) 8 (5.1%) 3 (1.7%) 5 (8.6%)
 Negative 61 (91%) 322 (85.6%) 400 (98.3%) 149 (94.9%) 174 (98.3%) 53 (91.4%)

cancer is infrequent. International collaborations may aid 3. Winter-Roach BA, Kitchener HC, Dickinson HO (2009) Adjuvant
in identifying patients with apparent early stage EOC who (post-surgery) chemotherapy for early stage epithelial ovarian can-
cer. Cochrane Database Syst Rev. https:​ //doi.org/10.1002/14651​
may be safely spared of systematic LND. 858.CD004​706.pub3
4. Kleppe M, Wang T, Van Gorp T et al (2011) Lymph node metas-
Author contributions  DN: Conception, Protocol/project development, tasis in stages I and II ovarian cancer: a review. Gynecol Oncol
Data collection and management, Data analysis, Manuscript writing/ 123(3):610–614
editing, SAM: Data analysis, Manuscript writing/editing, EMK: Data 5. Surveillance, Epidemiology, and End Results (SEER) Program
analysis, Manuscript writing/editing, NAL: Protocol/project develop- Research Data (1973–2015), National Cancer Institute, DCCPS,
ment, Data analysis, Manuscript writing/editing. Surveillance Research Program. https​://www.seer.cance​r.gov
6. Cancer Incidence in Five Continents Vol.X IARC Scientific Pub-
Funding  The authors declare no source of funding for the research lication No 164, 2014 Chapter 4, p84
conducted. 7. Gockley A, Melamed A, Bregar AJ, Clemmer JT, Birrer M,
Schorge JO, Del Carmen MG, Rauh-Hain JA (2017) Outcomes
of women with high-grade and low-grade advanced-stage serous
Compliance with ethical standards  epithelial ovarian cancer. Obstet Gynecol 129(3):439–447
8. Lago V, Minig L, Fotopoulou C (2016) Incidence of lymph
Conflict of interest  All authors declare no potential conflict of interest, node metastases in apparent early-stage low-grade epithelial
including and financial or personal relationship which could viewed ovarian cancer: a comprehensive review. Int J Gynecol Cancer.
as one. 26(8):1407–1414
9. Minig L, Heitz F, Cibula D et al (2017) Patterns of lymph node
Ethical standards  The SEER is a public dataset available to the public metastases in apparent stage i low-grade epithelial ovarian cancer:
for research purpose. All patient data are de-identified. The present a multicenter study. Ann Surg Oncol 24(9):2720–2726
study was deemed exempt by the institutional review board. 10. Kaldawy A, Segev Y, Lavie O, Auslender R, Sopik V, Narod SA
(2016) Low-grade serous ovarian cancer: a review. Gynecol Oncol
143(2):433–438
11. Nasioudis D, Chapman-Davis E, Witkin SS et al (2017) Prognostic
References significance of lymphadenectomy and prevalence of lymph node
metastasis in clinically-apparent stage I endometrioid and muci-
1. Siegel RL, Miller KD, Jemal A (2015) Cancer statistics, 2015. CA nous ovarian carcinoma. Gynecol Oncol 144(2):414–419
Cancer J Clin 65:5–29. https​://doi.org/10.3322/caac.21314​
2. Young RC (2003) Early-stage ovarian cancer: to treat or not to
treat. J Natl Cancer Inst 95(2):94–95

13