Download Clinicopathological Factors Predicting Retroperitoneal Lymph Node

Jpn J
cu« OncoI1998;28(11)673-678
Clinicopathological Factors Predicting Retroperitoneal Lymph
Node Metastasis and Survival in Endometrial Cancer
Xiaohai Tang 1,*, Kenjiro Tanemura', Weimin Ye2, Kazuo Ohml', Ryuichiro Tsunematsu', Takuro Yamada 1,
Noriyuki Katsumatal and Takahiko Sonoda 1
of Gynecology, 2Cancer Information and Epidemiology Division and 3Division of Medical Oncology,
National Cancer Center, Tokyo, Japan
1Division
Background: By clarifying the significance of clinicopathological factors for retroperitoneal lymph
node metastasis and survival of patients with endometrial cancer, we suggest ideas for optimal
treatment of this disease.
Methods: A retrospective study was conducted in 310 women with endometrial cancer who
underwent surgery with retroperitoneal lymphadenectomy. To evaluate retroperitoneal lymph
node metastasis, age-adjusted and multivariable analyses were carried out for six clinicopathological factors including pathological grade, myometrial invasion, cervical invasion, peritoneal
cytology, lymphatic permeation and vascular invasion. To evaluate survival, besides the above
factors, a positive rate of metastasis of dissected retroperitoneal lymph nodes was included.
Results: In 40 patients (13%) with nodal metastasis, the average positive rate of metastasis of
dissected retroperitoneal lymph nodes was 22%. For retroperitoneal lymph node metastasis, the
odds ratio of deep myometrial invasion, cervical invasion and severe lymphatic permeation were
5.97, 2.72 and 12.01, respectively. For survival, the hazard ratios of the positive rates of
metastasis of dissected retroperitoneal lymph nodes (both 25% and <25%), positive peritoneal
cytology and poor pathological grade were 7.10, 3.24, 3.82 and 3.27, respectively, and 5-year
survival rates for them were 0, 50, 72 and 77%, respectively.
Conclusions: For retroperitoneal lymph node metastasis, lymphatic permeation, deep myometrial invasion and cervical invasion were the independent prognostic factors. For survival,
retroperitoneal lymph metastasis, poorpathological grades and positive peritoneal cytology were
the independent prognostic factors. The positive rate of metastasis of dissected retroperitoneal
lymph metastasis playsan important rolein predicting survival of endometrial cancer. Lymph node
biopsy is insufficient in treatment of this disease.
Key words: clinicopathological factors - lymph node metastasis - survival - endometrial cancer
INTRODUCTION
In 1988, the International Federation of Obstetrics and Gynecology (FIGO) introduced a new system for the clinicopathological
staging of endometrial cancer. The staging system includes the
assessment of pathological grade, myometrial invasion, cervical
Received June 10, 1998; accepted August 14, 1998
*A Sasakawa Medical Fellow from the Department of Gynecological
Oncology, Sichuan Cancer Hospital, No. 55, Sect. 4, Ren Min Nan Road,
Chengdu, Sichuan 610041, China
For reprints and all correspondence: Kenjiro Tanemura, Division of
Gynecology, National Cancer Center Hospital, 1-1, Tsukiji 5-chome,
Chuo-Ku, Tokyo 104, Japan
Abbreviations: FIGO, International Federation of Obstetrics and
Gynecology; RLN, retroperitoneal lymph node; P, P value; PAN,
paraaortic lymph node; PLN, pelvic lymph node
invasion, peritoneal cytology and retroperitoneal lymph node
(RLN) metastasis. Since then, numerous reports have elucidated
the prognostic significance of these factors (1-13). In addition,
lymphatic permeation and vascular invasion were discussed for
their prognostic significance in this disease (14,15). However, the
discriminatory value of these is the subject of further investigation.
Patients with cervical invasion have a higher incidence of RLN
metastasis. In general, cervical invasion is frequently associated
with deep myometrial invasion that is in tum strongly related to
RLN metastasis. However, why patients with cervical invasion
have a higher incidence of nodal metastasis needs further
investigation (1-3,12).
The number of metastatic nodes among the total dissected
RLNs has always been individualized in each patient. To
understand its significance for survival, we proposed the
evaluation of positive rate of metastasis of the dissected RLNs.
674
Prognostic factors of endometrial cancer
Table 1. Patients' characteristics
Age
yr
Mean
55
Range
25-82
Stage
IA-IC
No.
195
IIA-IIB
48
IlIA-IlIC
67
Pathology
Adenocarcinoma
Adenosquamous
H ysterectom y
No.
294
16
No.
Simple
227
Modified radical
10
Radical
73
Dissected RLN in each patient
No.
Mean
24
Range
10-62
RLN metastasis
No.
40
Positive rate of metastasis of dissected RLNs
from the study, as their biological properties were different. The
number of dissected RLNs was counted. There was not a
significant difference in the number of dissected RLNs between
the patients with and those without RLN metastasis. The
percentage of metastatic nodes in the total dissected lymph nodes
was defined as a positive rate of metastasis of dissected RLNs.
The living status of the patients was registered up to December
1996 for survival analysis. Distribution of age, FIGO stage,
pathology, type of hysterectomy, number of dissected lymph
nodes in each patient, number of patients with RLN metastasis
and positive rate of metastasis of dissected RLNs are given in
Table 1.
%
Mean
22
Median
10
RLN, retroperitoneallymph node.
As RLN metastasis is associated with a poor prognosis
(1,2,8-17), routine RLN dissection has been suggested for
patients with endometrial cancer. However, the extent of
lymphadenectomy remains controversial because the significance of lymphatic metastasis is yet to be defined (l,8). In
addition, the significance of retroperitoneal lymph node biopsy
also needs further investigation.
The aims of this retrospective study were to analyze the
relationship between RLN metastasis and the clinicopathological
factors, to select independent clinicopathological prognostic
factors of RLN metastasis and to clarify the significance of these
factors for survival.
PATIENTS AND METHODS
Between January 1979 and December 1996, 310 women with
primary endometrial cancer were treated with hysterectomy and
lymphadenectomy of RLN at the Division of Gynecology of the
National Cancer Center Hospital, Japan. All patients were
diagnosed to have FIGO stage IA to mc primary endometrioid
adenocarcinoma and adenosquamous cancer during the mentioned period. Patients with high-risk endometrial cancers, such
as papillary serous cancer and clear cell cancer, were excluded
SURGICAL PROCEDURE
During a period of nearly 20 years, five of the authors steadily
performed all operations on the patients according to routine
methods set up by them at the Division of Gynecology. Patients
without cervical invasion underwent simple abdominal hysterectomy and bilateral oophorectomy. Patients who were at a
relatively advanced stage of the disease and had no cervical
invasion underwent modified radical hysterectomy. Patients who
were identified as having cervical invasion by preoperative
endocervical curettage underwent radical hysterectomy. Peritoneal cytology was obtained immediately on opening the peritoneal cavity. For the lymphadenectomy, the retroperitoneal space
in the pelvis was entered after dividing the round ligaments. The
vessels were outlined and the lymphatic tissue along the external
and internal iliac veins and arteries was removed from inguinal
ligament to the common iliac vessels. In the obturator fossae, the
iliac vessels and the obturator nerves were cleaned of the
lymphatic tissue down to the levator muscle. Lymphatic tissue
overlying the lower segment of the aorta (from aortic bifurcation
up to the level of inferior mesentery) was removed after pelvic
lymphadenectomy. The visual lymph nodes were picked from the
dissected lymphatic tissue and divided into anatomical groups
(aortic nodes, common iliac nodes, external iliac nodes, internal
iliac nodes, obturator nodes, medial inguinal nodes, lateral
inguinal nodes and cardinal ligament nodes) before pathological
examination.
STATISTICAL ANALYSIS
Age-adjusted logistic regression was used to analyze the significance of each of the clinicopathological factors on RLN
metastasis and then multivariate logistic regression was carried
out to determine which of these were the independent prognostic
factors. The Cox regression model was employed in analyzing the
significance of each of the clinicopathological factors on survival
and in determining which of these were the independent
prognostic factors. Spearman analysis was used in correlation
studies of the clinicopathological factors. Survival rates were
calculated by the Kaplan-Meier method and differences in
survival rates were examined using the log-rank test.
Jpn J Clin OncoI1998;28(ll)
675
Table 2. Relation between RLN metastasis and clinicopathological factors
Patients
(11 = 319)
OR*
RLN metastasis
Cases
95%CI
ORt
95%CI
%
Pathological grade
Gl
143
8
6
G2+G3
167
32
19
191
4
2
>1/2-2/3
31
4
>2/3
88
Negative
Positive
3.88
1.79-9.36
1.42
0.50-4.29
13
6.88
1.55-30.68
4.27
0.87-20.70
32
36
26.34
9.90-91.46
5.97
1.84 -23.66
240
17
7
70
23
33
6.43
3.20-13.18
2.72
1.06-7.05
Negative
251
25
10
Positive
59
15
25
3.10
1.49-6.32
2.43
0.88-6.73
Myometrial invasion
0-1/2
Cervical invasion
Peritoneal cytology
Lymphatic permeation
181
4
2
A few + several
98
17
17
9.33
3.32-33.28
3.04
0.92-11.93
Many
31
19
61
70.70
22.42-278.87
12.01
2.72-59.90
Negative
Vascular invasion
Negative
248
17
7
53
26
49
5.68
2.59-12.49
1.88
0.67-5.24
9
7
78
47.47
10.54-335.96
2.43
0.33-23.72
A few + several
Many
OR, odds ratio; CI, confidence interval. *Age-adjusted odds ratios for factors other than age, estimated by logistic regression. t Age variable and other factors.
RESULTS
RELATIONSHIP BETWEEN
RLN
METASTASIS AND
CLINICOPATHOLOGICAL FACTORS
Table 2 shows the results of logistic analysis in the 310 patients.
Patients who had any of the clinicopathological factors in its
positive status or severer grade had a higher frequency of RLN
metastasis compared with patients who had any of these factors
in its negative status or milder grade. As age was related to RLN
metastasis, it was adjusted in the following analysis. In age-adjusted analysis, any of the clinicopathological factors in its
positive status or severer grade had a higher odds ratio compared
with any of these factors in its negative status or milder grade. In
multivariate analysis, severe lymphatic permeation (many lymphatic capillaries filled with cancer cells on microscopy), deep
myometrial invasion and cervical invasion were found to be
independent clinicopathological prognostic factors for RLN
metastasis. Vascular invasion was related to a high frequency of
RLN metastasis. However, it was not an independent clinicopathological prognostic factor for RLN metastasis. To investigate
the reason, Spearman analyses were performed for correlation. A
strong correlation was found between the lymphatic permeation
and vascular invasion in which the correlation coefficient was
0.53 (P = 0.0001). The rest of the factors, including pathological
grade and peritoneal cytology, were not significantly related to
the RLN metastasis.
RELATIONSHIP BETWEEN SURVIVAL AND CLINICOPATHOLOGICAL
FACTORS
The results of Cox regression analysis are given in Table 3. In the
age-adjusted analysis, all the clinicopathological factors except
cervical invasion were found to be related to survival. To
determine the independent prognostic factors for survival,
multivariate analysis was performed. RLN metastasis was
identified as being the most significant prognostic factor for
survival. In addition, pathological grade and peritoneal cytology
were also identified as being independent prognostic factors for
survival. However, myometrial invasion, cervical invasion,
lymphatic permeation and vessel invasion were not significantly
related to survival.
676
Prognostic factors of endometrial cancer
Table 3. Relation between survival and clinicopathological factors
Patients
= 310)
HR*
95%CI
HRt
95% CI
1.87-7.62
3.27
1.49-7.22
(11
Pathological grade
G1
143
1
G2+G3
167
3.78
191
1
Myometrial invasion
0-1/2
>1/2-2/3
31
1.80
2.26--7.73
1.27
0.43-3.71
>2/3
88
4.18
2.26--7.73
1.56
0.67-3.73
1.75
0.96--3.19
0.62
0.29-1.34
1.68-5.44
3.82
1.92-7.62
Cervical invasion
Negative
240
Positive
70
Peritoneal cytology
Negative
251
Positive
59
1
3.02
Lymphatic permeation
Negative
181
1
A few + several
98
2.22
1.11-4.43
1.34
0.60-3.00
Many
31
8.51
4.11-17.62
1.75
0.55-5.05
Vessel invasion
Negative
A few + several
248
1
53
2.78
1.45-5.33
0.91
0.40-2.08
9
17.21
6.91-42.84
2.72
0.72-10.24
Many
Lymph node metastasis
Negative
270
1
<25%
33
6.71
3.58-12.55
3.24
1.41-7.41
~25%
7
17.75
5.68-55.44
7.10
2.29-27.32
HR, hazard ratio; CI, confidence interval. *Age-adjusted hazard ratios. t Age and other factors in the table were adjusted.
To study the influence of the positive rate of metastasis of
dissected lymph nodes on survival, 40 patients (13%) with RLN
metastasis were divided into two sub-groups. The dividing point
was selected at a positive rate of 25%, according to the average
positive rate of 22% of metastasis of dissected lymph nodes. The
first sub-group had 33 patients with <25% of the positive rate of
metastasis of dissected RLNs (the mean number of dissected
nodes was 23). The second sub-group had seven patients with
~25% of the positive rate (the mean number of dissected nodes
was 26). When compared with the sub-group of <25% of positive
rate of nodal metastasis, the sub-group of ~25% of the positive
rate had a significant high hazard ratio to survival (Table 3).
In Kaplan-Meier analysis, all the independent prognostic
factors for survival influenced significantly the survival probability of patients with endometrial cancer (Fig. I). Marked
differences in survival probability were found during the early
years of follow-up compared with the later years of follow-up. At
the fifth year after surgery, all patients who had a ~25% positive
rate of metastasis of dissected RLNs had died, compared with
55% of patients who had a <25% positive rate and 92% of patients
who had no RLN metastasis who were alive. The five-year
survival rates of the patients with negative RLN metastasis and
positive RLN metastasis, negative peritoneal cytology and
positive peritoneal cytology and well differentiated pathology
and moderate to poor differentiated pathologies were 92 and 50%
(P =0.001),90 and 72% (P =0.001) and 96 and 77% (P =0.001),
respectively.
The Kaplan-Meier plot did not show significant influences of
the distribution and number of metastatic sites of the dissected
RLNs on survival.
DISCUSSION
The generally accepted prognostic factors for RLN metastasis
and survival of endometrial cancer were evaluated in this study
(1- i 5). Further, the new proposed positive rate of metastasis of
dissected RLNs was evaluated for discriminating the prognosis in
the patients with RLN metastasis, so that the optimal treatment
protocol could be suggested. As would be expected, each of the
clinicopathological factors was related to RLN metastasis and
survival of this disease. Nevertheless, only a few of them had
independent prognostic significance. Additionally, the independence of these factors was variable, depending on the prognostic
significance judged by the RLN metastasis or the survival.
Jpn J Clin OncoI1998;28(11)
1.00
N
Rl.N metastasis:
N: Negative
(0=270)
P: Positive (0=40)
..
0.60
.~
;:
P=O.OOOI
~
Positive rate of
metastasis of
dissected RLNs:
Rl: <25%
R2: >25%
'0
'"
.",
~
~
8 0.40
c,
P=O.OOOI
R
R2
P
0.20
0.00
~...L.w....c-,-,-,-,-.......w.'-U.J..~...L.w....c-"-'-U--'--'-'-'--,-,-,-"-'-U.J..~..LW-LJ..Lu..Ll..LL.L"-'--U...u..w...........-,-,-,-"-w....L~...w
o
1
2
3
4
5
6
7
9
8
10
11
13
12
14
15
16
17
Years after Surgery
b
1,00
0.80
~
.~
0.60
"l
"0
'"
~8
0,40
""'
N: Gl (0=143)
P: G2 and G3 (0=167)
P=O.OOI
0,20
0.00
~~~~~~~~~~~~~~~~~~~~~
o
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Years after Surgery
1.00
N
0.80
0.60
18
18
677
In this study, deep myometrial invasion (>2/3), cervical
invasion and severe lymphatic permeation (many) were found to
be independent prognostic factors for RLN metastasis and in tum
the RLN metastasis was an independent prognostic factor for
survival. Hence their influence on survival was rooted in the RLN
metastasis. They were harmful to survival, consistent with
previously reported findings (1-3,5,12-15).
As gross myometrial invasion could be assessed preoperatively
(12,18), it provided a marker for preoperative prediction ofRLN
metastasis. In the patients with cervical invasion, the cancerous
lesion in the cervix was easy to metastasize to the paracervical
area. Therefore, radical hysterectomy should be performed in
these patients so that the possible paracervical involvement could
be cleaned. Radiation therapy was another choice in treatment of
cervical invasion (4,12,17). Once lymphatic permeation has
occurred, cancer cells have the potential to metastasize to the
lymph node. Vascular invasion was strongly correlated with
lymphatic permeation. In Inoue et al.'s study, lymphatic permeation and vascular invasion were combined as vessel invasion
and were found to be a reliable prognostic factor for survival (14).
On the other hand, poor pathological grade, positive peritoneal
cytology and RLN metastasis were found to be independent
prognostic factors for survival. Patients with endometrial cancer
of poor pathological grade advanced quickly because such
cancerous lesion tended to be associated with deep myometrial
invasion, lymphatic permeation and cervical invasion from which
RLN metastasis occurred (12,13,19). Hence, once endometrial
cancer of poor pathological grade has been diagnosed, treatment
should be initiated as soon as possible. Positive peritoneal
cytology significantly influences survival of this disease, consistent with results in previous reports (4,5). Other reports revealed
inconsistent results that peritoneal cytology did not play an
important role in the prognosis of patients who had well
differentiated endometrial cancer (6,7). We noted that their
patients had no RLN metastasis.
The extent ofRLN assessment in the new FIGO staging system
for endometrial cancer can vary from simple transperitoneal
palpation with excisional biopsy of enlarged lymph nodes, via
directed biopsies of one or two lymph nodes from each lymph
node area, to formal pelvic and paraaortic lymphadenectomy (1).
However, the average positive rate of metastasis of dissected
RLNs was found to be only 22% in this study and Creasman et
al. (16) reported that even in stage I disease there are still 7-13 %
patients with positive pelvic lymph nodes. Additionally, at
present there is no reliable method for accurate preoperative or
intraoperative assessment of RLN metastasis. Hence lymph node
biopsy alone is insufficient in the treatment of this disease. To
0.40
Peritoneal cytology
N:Negative(n=251)
P:Positive(0=59)
0.20
P=O.OOI
0.00 '-'-'-'-'.........................u..w-"-'-'-'-'-'-'-'-' ...........U-W...L~...u..w.~~..........w......w-W-W..'-'-'-U.~'-'-'-'-"~~
o
1
2
3
4
5
6
7
8
9
10
It
YearsafterSurgery
12
13
14
15
16
17
18
Figure 1. Kaplan-Meier estimation of survival probability in 310 patients with
endometrial cancer, according to (a) RLN metastasis, (b) pathological grade and
(c) peritoneal cytology. Panel (a) shows that RLN metastases, especially with
a status of 225% positive rate of metastasis of dissected RLNs, significantly
influenced the survival probability in patients with this disease compared with
those who had no RLN metastasis. Panels (b) and (c) shows that poorer
differentiated pathologies and positive peritoneal cytology significantly
influenced the survival probability in these patients compared with those who
had well differentiated pathology and negative peritoneal cytology, respectively.
678
Prognostic factors of endometrial cancer
obtain accurate staging, pelvic and paraaortic lymphadenectomy
is necessary in the treatment of this disease.
Instead of the number of metastatic nodes in total dissected
RLNs, the positive rate of metastasis of dissected RLNs was used
to evaluate the survival rate of patients with this disease. This is
because there was a variation in the number of dissected RLNs in
each patient. There are at least three causes of this variation. First,
there may be variations in the actual number of RLNs in each
patient. Second, sometimes there is difficulty in performing
lymphadenectomy in obese patients or in patients with other
severe medical problems (1). Third, there are variations in the
technique to pick up the lymph nodes from the dissected
lipo-Iymphatic tissue. The positive rate of metastasis in dissected
RLNs was considered to be a more scientific evaluation for
prediction of survival than the number of metastatic nodes in
dissected RLNs, e.g. five of 15 or five of 30 are completely
different.
Patients with RLN metastasis were classified into FIGO stage
lIlC and had a poorer survival rate than those without it. However,
even in these patients, survival was distinguishable by the positive
rate of metastasis of dissected RLNs. The survival curve of
patients with a 2::25% positive rate of metastasis of dissected
RLNs fell steeply soon after surgery. It is presumed that the high
positive rate of nodal metastasis represents high malignant
biological properties of the cancer and predisposes the patients to
local and distant relapse. On the other hand, patients with a low
positive rate of nodal metastasis had much better 5-year survival
rates than patients with a high positive rate. The explanation for
this phenomenon is as follows. In the patients with a low positive
rate of nodal metastasis, as the number of metastatic nodes was
small, it was fairly easy to dissect them thoroughly. As a result,
local recurrence was relatively low in this subgroup of patients
with RLN metastasis. There is evidence that distant recurrence
was low when local recurrence was low (17). Second, the low
positive rate of nodal metastasis may reflect a relatively early
period in the natural history of the disease or a relatively low
malignant biology of the cancer.
The results in this study did not show an influence on survival
of site and distribution ofRLN metastasis. This was in accordance
with Kadar et al. 's report that the difference in survival between
women with paraaortic lymph node (PAN) and pelvic lymph node
(PLN) metastasis was not significant (4). However, it was
contrary to Yokoyama et al. 's result that the survival of patients
with PAN metastasis was poorer than that of patients with only
PLN metastasis (8). Our view is that, in endometrial cancer, low
segmental PAN and PLN metastasis are nodal metastasis of the
same location so that they have a similar influence on survival. In
any case, PAN metastasis suggests that the disease has progressed
to a severer extent.
The 5-year survival rate was not satisfactory with surgery alone
in the treatment of patients with endometrial cancer and RLN
metastasis, poorly differentiated lesion and positive peritoneal
cytology. Hence postoperative radiation therapy and/or chemotherapy should be employed to improve the survival rate of these
patients (9). Additionally, for patients with deep myometrial
invasion, cervical invasion, lymphatic permeation and vessel
invasion, postoperative treatment should be initiated if RLN
metastasis has been identified.
Acknowledgment
This study was supported in part by the Japan-China Medical
Association Sasakawa Medical Scholarship Fund.
References
1. Larson DM, Connor GP, Broste SK, Krawisz BR, Johnson KK. Prognostic
significance of gross myometrial invasion with endometrial cancer. Obstet
Gynecol 1996;88:394-8.
2. Kaku T, Tsuruchi N, Tsukamoto N, Hirakawa T, Kamura T, Nakano H.
Reassessment of myometrial invasion in endometrial cancer. Obstet Gynecol
1994;84:979-82.
3. Cohen CJ. Advanced (FIGO stages III and IV) and recurrent cancer of the
endometrium. In: Coppleson M, Monaghan JM, Morrow CP, Tattersall MH,
editors. Gynecologic Oncology, 2nd ed. Edinburgh: Churchill Livingstone,
1992;778-98.
4. Kadar N, Homesley HD, Malfetano JH. Prognostic factors in surgical stage
III and IV cancer of the endometrium. Obstet Gynecol 1994;84:983-6.
5. Creasman WT, Rutledge FN. The prognostic value of peritoneal cytology in
gynecologic malignant disease. Am J Obstet Gynecol 1976; 110:773-81.
6. Vecek N, Marinovic T, Ivic J, Jukic S, Nola M, Dzanic-Cemalovic N, et al.
Prognostic impact of peritoneal cytology in patients with endometrial cancer.
Eur J Gynaecol OncoI1993;14:380-5.
7. Kato T, Yamada T, Tanemura K, Tsunematsu R, Ohmi K, Sonoda T. Clinical
significance of peritoneal cytology in patients with endometrial adenocarcinoma. SankaTo Fujinka 1989;56:1759-63 (in Japanese).
8. Yokoyama Y, Maruyama H, Sato S, Saito Y Indispensability of pelvic and
paraaortic lymphadenectomy in endometrial cancers. Gynecol Oncol
1997;64:411-7.
9. Onda T, Yoshikawa H, Mizutani K, Mishima M, Yokota H, Nagano H, et ai.
Treatment of node-positive endometrial cancer with complete node dissection, chemotherapy and radiation therapy. Br J Cancer 1997;75:1836-41.
10. Kodama S, Kase H, Tanaka K, Matsui K. Multivariate analysis of prognostic
factors in patients with endometrial cancer. lnt J Gynaecol Obstet
1996;53:23-30.
11. Behbakht K, Yordan EL, Casey C, DeGeest K, Massad LS, Kirschner CV, et
ai. Prognostic indicators of survival in advanced endometrial cancer. Gynecol
OncoI1994;55:363-7.
12. Lampe B, Kurzl R, Hantschmann P. Prognostic factors that predict pelvic
lymph node metastasis from endometrial cancer. Cancer 1994;74:2502-8.
13. Kosary CL. FlGO stage, histology, histologic grade, age and race as
prognostic factors on determining survival for cancers of the female
gynecological system: an analysis of 1973-87 SEER cases of cancers of the
endometrium, cervix, ovary. vulva and vagina. Semin Sure Oncol 1994
10:31-46.
14. Inoue Y, Obata L, Abe K, Ohmura G, Doh K, Yoshioka T, et ai. The
prognostic significance of vascular invasion by endometrial cancer. Cancer
1996;78:1447-51.
15. Tsuruchi N, Kaku T, Kamura T, Tsukamoto N, Tsuneyoshi M, Akazawa K et
ai. The prognostic significance of Iymphovascular space invasion in
endometrial cancer when conventional hemotoxylin and eosin staining is
compared to immunohistochemical staining. GynecoIOnco/1995;57:307-12.
16. Creasman WT, Morrow CP, Bundy L. The surgical pathological spread
pattern of endometrial cancer. A Gynecologic Oncology Group Study.
Cancer 1987;60:2035-41.
17. Com BW, Lanciano RM, Agostino RD, Kiggundu E, Dunton CJ, Purser P,et
al. The relationship of local and distant failure from endometrial cancer:
defining a clinical paradigm. Gynecol Oneal 1997 66:411-6.
18. Tang X, Muramatsu Y, Yajima M, Sonoda T. Endometrium-myometrium
ratio: a newly proposed diagnostic parameter on magnetic resonance imaging
assessment of myometrial invasion by endometrial cancer. Jpn J Clin Oncol
1993;23:278-83.
19. Sonoda T. Clinical diagnosis of endometrial cancer: especially for discussion
about primary lesion. Sanka To Fujinka 1987;56:2117-25 (in Japanese).