Objective. The aim of this study was to analyze FIGO Stage IIIc endometrial
cancer (EC) patients to better define clinicopathologic associations, patt
erns of failure, and survival.
Methods. Charts were abstracted from EC patients with lymph node metastasis
from 1989 to 1998. Data on clinicopathologic variables, adjuvant treatment
, site of first recurrence, and survival were collected. Associations betwe
en variables were tested by chi (2) and Wilcoxon rank sums. Survival analys
es were performed by the Kaplan-Meier method, and multiple regression analy
sis was done by the Cox proportional hazards model.
Results. From 607 EC patients evaluated, 47 (8%) were identified with FIGO
Stage IIIc disease. All 47 underwent hysterectomy and pelvic lymph node (PL
N) sampling, and 42/47 had para-aortic lymph node (PALN) sampling. Stage II
Ic disease was defined by positive PLN alone in 38%, positive PLN and PALN
in 41%, and positive PALN alone in 17%. Twelve of 47 also had positive peri
toneal cytology and/or adnexal metastases. Grade III tumors were present in
56% and > 50% myometrial invasion in 61%. No association between depth of
invasion (DOI) and grade was seen, however. Nearly 1/3 of cases had papilla
ry serous or clear cell histology. Postoperative adjuvant treatment include
d whole abdominal radiation (36%), pelvic radiation with (19%) and without
(17%) extended field, chemotherapy (17%), and oral progestins (11%). The 3-
year and 5-year survival estimates for all patients were 77 and 65%, respec
tively. At a median follow-up of 37 months, 5 patients are alive with disea
se, and 10 are dead of disease. A distant site of first recurrence was most
common (21%), followed by pelvic failure (9%). Only 1 patient has had an a
bdominal recurrence. Univariate predictors of survival included age, DOI, a
nd extranodal disease, but not grade, histology, or PALN involvement. For t
he 12 patients with nodal disease and positive cytology and/or adnexa, 3-ye
ar survival was 39% versus 93% for those patients without evidence of extra
nodal disease. In a multivariate analysis only DOI was an independent predi
ctor of survival (P = 0.03).
Conclusions. Once lymph node involvement occurs, the importance of addition
al extranodal disease increases. Consideration of substaging Stage IIIc pat
ients based on positive adnexa or cytology is supported by the data. The ex
tent which adjuvant treatments contributed to the 77% 3-year survival remai
ns to be defined. The patterns of failure suggest a possible role for combi
ned modalities in future treatments. (C) 2001 Academic Press.