Predictors of vaginal relapse in stage I endometrial cancer

www.elsevier.com/locate/ygyno

Gynecologic Oncology

Predictors of vaginal relapse in stage I endometrial canceri

Andrea Mariania, Sean C. Dowdya, Gary L. Keeneyb, Michael G. Haddockc,

Timothy G. Lesnickd, Karl C. Podratza,*

aSection of Gynecologic Surgery, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USAbDivision of Anatomic Pathology, Mayo Clinic, Rochester, MN 55905, USAcDivision of Radiation Oncology, Mayo Clinic, Rochester, MN 55905, USA

dDivision of Biostatistics, Mayo Clinic, Rochester, MN 55905, USA

Received 5 November 2004

Available online 13 May 2005

Abstract

Objective. To identify factors predictive of vaginal relapse in stage I endometrial cancer, thereby potentially facilitating the selection of

patients who may benefit from vaginal brachytherapy.

Methods. The study population included 632 patients with stage I endometrial cancer managed with hysterectomy at our institution

between 1984 and 1996. Median follow-up was 73 months; 122 patients (19%) received adjuvant radiotherapy.

Results. Overall, 2.9% of the stage I cohort developed vaginal relapse at 5 years. Vaginal relapse was observed in 1.7% of patients

who received radiotherapy and in 3.0% of those whose treatment did not include radiotherapy (P = 0.36). Cox regression analysis

(including radiotherapy) identified only grade 3 differentiation (hazard ratio = 3.83, P = 0.007) as an independent predictor of vaginal

relapse. Patients with a low-grade tumor had a 5-year vaginal relapse rate of 2%, compared with 7% for those with a grade 3 tumor. When

only patients who did not receive adjuvant radiotherapy were considered, both grade 3 tumor and lymphovascular invasion were

significant predictors of vaginal relapse (P < 0.05). When neither variable was present, 2% of patients experienced vaginal relapse at 5

years, compared with 11% when either 1 was present (P < 0.001). Depth of myometrial invasion was not a significant predictor of vaginal

recurrence.

Conclusion. Histologic grade 3 tumor and lymphovascular invasion were the cogent predictors of vaginal relapse in our population. The

cost and morbidity of vaginal brachytherapy should be balanced against the potential risk of vaginal relapse in this group of patients.

D 2005 Elsevier Inc. All rights reserved.

Keywords: Endometrial cancer; Vaginal relapse; Lymphovascular invasion

Introduction

Endometrial cancer is the most common malignancy of

the female reproductive tract in the United States and is

exceeded annually in overall frequency only by cancers of

the breast, colon, and lung. It has been estimated that

0090-8258/$ - see front matter D 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.ygyno.2005.03.008

Abbreviations: FIGO, Federation of International Gynecology and

Obstetrics; LVI, lymphovascular invasion.i Presented at the 34th Annual Meeting of the Society of Gynecologic

Oncologists (SGO),NewOrleans, Louisiana, January 31 to February 4, 2003.

* Corresponding author.

E-mail address: [email protected] (K.C. Podratz).

during calendar year 2005, 40,880 new cases of endome-

trial cancer will be diagnosed and 7310 deaths will occur

[1].

Vaginal relapse is a relatively common pattern of

treatment failure for patients with endometrial cancer

[2] and can be effectively prevented by the adminis-

tration of vaginal brachytherapy [3]. Although more than

60% of isolated vaginal relapses occurring in patients

with stage I disease are treated successfully at the time

of recurrence [4,5], treatment-associated morbidity is

significant.

Traditional predictors of vaginal relapse have included

the presence of deep myometrial invasion and poor tumor

97 (2005) 820 – 827

A. Mariani et al. / Gynecologic Oncology 97 (2005) 820–827 821

differentiation [3]. Recently, lymphovascular invasion (LVI)

has been identified as a predictor of vaginal relapse in

patients with highly differentiated stage I endometrial cancer

with minimal myometrial invasion [6]. The identification of

patients at risk for vaginal relapse is especially important in

stage I endometrial cancer, for which locoregional therapy

(i.e., surgery and vaginal brachytherapy) may possibly be

the only treatment needed to achieve complete cure of the

disease [7]. Assessing the natural history of occult local

extension may suggest ways to successfully decrease

recurrence rates with locoregional therapy.

The objective of the present analysis was to identify

factors predictive of vaginal relapse in stage I endometrial

cancer, thereby potentially facilitating the selection of

patients who may benefit from vaginal brachytherapy.

Patients and methods

From 1984 to 1996, 1109 patients with endometrial

cancer were managed surgically at Mayo Clinic (Rochester,

Minnesota). Based on data from the medical records, 632

patients with epithelial endometrial cancer met the follow-

ing inclusion criteria: (1) treatment included hysterectomy

and removal of existing adnexal structures, (2) the disease

was stage I, and (3) no other malignancy was diagnosed

within 5 years before or after the diagnosis of endometrial

cancer (except for carcinoma in situ or skin cancer other

than melanoma).

Staging was defined according to the International

Federation of Gynecology and Obstetrics (FIGO) surgical

staging system [8]. For patients who received treatment

before 1988, stage was determined retrospectively on the

basis of the surgical and pathologic assessments. Histologic

classification was according to the World Health Organ-

ization classification [9]. Architectural grading was based on

the degree of glandular differentiation in accordance with

the FIGO guidelines [8]. LVI was considered present when

tumor cells were within or attached to the wall of a

capillary-like space. Tumor diameter was assessed as

previously reported [10].

All hematoxylin–eosin-stained slides of the tumor were

reviewed retrospectively by one of us (G.L.K.) to confirm

the original diagnosis of adenocarcinoma and to determine

FIGO grade, histologic subtype, and LVI.

All surgical procedures were the responsibility of a

gynecologic oncologist. Hysterectomy was generally per-

formed via the abdominal route. During the 13-year period,

several of the surgeons routinely used a modified radical

(type II) hysterectomy for the treatment of endometrial

cancer. The decision to perform a simple hysterectomy or a

wide extrafascial (with unroofing of the ureters) hysterec-

tomy usually depended on the personal preference of the

gynecologic surgeon rather than on histologic or clinical

variables. Several patients underwent vaginal hysterectomy,

followed by laparotomy for surgical staging (when indi-

cated), after the diagnosis of endometrial cancer was made

by means of frozen-section analysis. Selected patients with a

preoperative diagnosis of suspected cervical invasion under-

went radical (type III) hysterectomies. Lymphadenectomy

usually was performed only in patients considered by the

surgeon to be at risk for lymph node metastasis, according

to the histologic grade of the tumor and depth of myometrial

invasion as determined by an intraoperative analysis of

frozen tissue sections.

Postoperative adjuvant radiotherapy consisted of external

beam pelvic radiation or vaginal brachytherapy or both. The

decision to deliver adjuvant radiotherapy depended pre-

dominantly on the assessment by gynecologic oncologists

and radiation oncologists of the risks of local or regional (or

both) recurrence after pathologic evaluation of the surgical

specimen. This decision usually was dictated by the

presence of grade 3 differentiation, nonendometrioid histo-

logic subtype, or deep myometrial invasion, or a combina-

tion of these pathologic features.

If sufficient follow-up information about survival and

recurrence was not available in the clinical records, death

certificates were obtained and letters were sent or telephone

calls were made to patients and family physicians to obtain

the information.

‘‘Vaginal relapse’’ was defined as disease relapse on or

within the vaginal walls or cuff, generally detected by

physical examination or vaginal cytology and confirmed

histologically. Usually, at the time of recurrence, imaging

studies of the chest and/or abdomen and/or pelvis were

obtained to identify possible concomitant pelvic or distant

sites of recurrence. For the purpose of the present analysis,

we considered only the first site (or concomitant sites) of

detected primary recurrence.

For statistical purposes, endometrioid tumors, endome-

trioid tumors with squamous differentiation, and adenosqu-

amous tumors were considered together. Grades 1 and 2

tumors were combined and compared with grade 3 tumors.

The cutoff for tumor diameter was chosen as 2 cm on the

basis of our previous experience [10].

Statistical analysis was performed with the Fisher exact

test and v2 analysis to test for relations between pairs of

categorical variables. The Kaplan–Meier method and log-

rank test were used for time-to-event analyses, with

recurrence as the end point. Differences were considered

statistically significant at P < 0.05. SAS System 8.2

statistical software and JMP statistical program (version

4.0.4, SAS Institute, Inc., Cary, North Carolina) were used

for the analysis.

Results

Descriptive statistics

For the population of 632 patients with stage I

endometrial cancer, the mean age T SD was 64.5 T 11.0

A. Mariani et al. / Gynecologic Oncology 97 (2005) 820–827822

years (range, 22–95 years) and the mean T SD body mass

index was 30.7 T 0.3 kg/m2 (range, 16.6–64.7 kg/m2). The

clinical and pathologic characteristics of these patients are

summarized in Table 1.

Lymph node dissection was performed in 315 patients

(50%). More precisely, 310 (49%) patients had pelvic

lymphadenectomy and 57 (9%) had para-aortic lymphade-

nectomy (3 had para-aortic lymphadenectomy only and 54

had concomitant pelvic and para-aortic lymphadenectomy);

Table 1

Characteristics of the 632 patients with stage I endometrial cancer

Characteristic Patients

No. %a

Age, years

�60 224 35

>60 408 65

BMI, kg/m2

<30 357 57

�30 271 43

NA 4

Stage

IA 154 24

IB 369 59

IC 106 17

NA 3

Myometrial invasion, %

�50 544 86

>50 85 14

NA 3

Invasion of LUS

No 596 94

Yes 36 6

Lymphadenectomy

No 317 50

Yes 315 50

Histologic subtype

Endometrioid 599 95

Nonendometrioid 33 5

LVI

Yes 25 4

No 561 96

NA 46

Associated hyperplasia

No 181 51

Yes 177 49

NA 274

Primary tumor diameter, cm

�2 241 43

>2 325 57

NA 66

Adjuvant radiotherapy

No 508 81

Yes 122 19

NA 2

Histologic grade

1 347 55

2 201 32

3 83 13

NA 1

BMI, body mass index; LUS, lower uterine segment; LVI, lymphovascular

invasion; NA, not available.a Percentages are calculated excluding missing patients.

the site from which the nodes were harvested was not

identified in 2 patients. Adjuvant radiotherapy was admin-

istered to 122 patients (19%), and 7 (1%) received adjuvant

hormonal therapy. Information about administration of

adjuvant radiotherapy was not available for 2 patients. Of

the 122 patients who received adjuvant radiotherapy, 69

(57%) had external beam pelvic radiotherapy only, 16 (13%)

had vaginal brachytherapy only, and 37 (30%) had

combined vaginal brachytherapy and external beam radio-

therapy. The median dose of external beam radiotherapy

administered to the pelvis was 5040 cGy. Brachytherapy

was delivered with low-dose-rate afterloading applicators

with cesium 137. The median dose at the 5-mm depth was

1500 cGy when brachytherapy was performed in addition to

external radiotherapy and 6000 cGy to the vaginal surface

when brachytherapy alone was used.

Overall, of 543 patients with myometrial invasion of

50% or less, 57 (10.5%) had grade 3 tumors, compared with

26 of 85 patients (31%) with myometrial invasion greater

than 50% (P < 0.001). Furthermore, of 511 patients with

myometrial invasion of 50% or less, 13 (3%) had LVI,

compared with 12 of 72 patients (17%) with myometrial

invasion greater than 50% (P < 0.001).

Predictors of vaginal relapse in the whole cohort of 632

patients

Median follow-up was 73 months. Overall, 20 patients in

the stage I cohort developed vaginal relapse (2.9% at 5

years): isolated primary vaginal relapse in 19 and concom-

itant vaginal and hematogenous (lung and bone) relapses in

1. Vaginal relapse at 5 years was observed in 1.7% of

patients who had adjuvant radiotherapy and in 3.0% of those

who did not (P = 0.36). Importantly, no vaginal relapse was

observed at 5 years in the 53 patients who had vaginal

brachytherapy (alone or in association with external pelvic

radiotherapy), compared with 2.9% at 5 years for patients

who had external pelvic radiotherapy only.

On the basis of the univariate analysis of the entire

cohort, grade 3 tumor and nonendometrioid histologic

subtype were identified as significant predictors of vaginal

relapse (P < 0.05), whereas invasion of the lower uterine

segment approached significance (P = 0.056). However, the

following were not significant predictors of vaginal relapse

(P > 0.05): age older than 60 years, body mass index greater

than 30 kg/m2, myometrial invasion greater than 50%,

tumor diameter larger than 2 cm, LVI, hyperplasia asso-

ciated with the tumor, or performance of lymph node

dissection. Patients without invasion of the lower uterine

segment had a vaginal relapse rate of 2.5% at 5 years,

compared with 10% for patients with invasion of the lower

uterine segment (P = 0.056). Moreover, 7% of patients with

a grade 3 tumor had vaginal relapse at 5 years, compared

with 2% of those with a grade 1 or grade 2 tumor (P =

0.007). Furthermore, 2.5% of patients with the endometrioid

histologic subtype had vaginal relapse at 5 years, compared

Fig. 1. Adjuvant radiotherapy for stage I endometrioid endometrial cancer.

The histogram shows the percentage of patients who had adjuvant

radiotherapy according to substage (stage IA–IB vs. IC), presence of

lymphovascular invasion (LVI), and histologic grade (G1–2 vs. G3). The

administration of adjuvant radiotherapy to patients with stage IA or IB

disease was directed by the presence of a grade 3 (G3) tumor. Note that only

patients with endometrioid tumor were considered. For 3 patients, no

information was available on the depth of myometrial invasion or histologic

grade. For 2 patients, no information was available about the administration


with 9% of patients with nonendometrioid histologic

subtype (P = 0.03). However, Cox regression analysis

(including adjuvant radiotherapy) identified only grade 3

differentiation (P = 0.007; hazard ratio [HR] = 3.83; 95%

confidence interval [CI], 1.435–10.22) as an independent

predictor of vaginal relapse. When performance of lymph

node dissection was added to the Cox regression analysis,

grade 3 differentiation was still the only independent

predictor of vaginal recurrence.

As expected, patients with high-risk features, such as

deep myometrial invasion, nonendometrioid histologic

subtype, LVI, no hyperplasia associated with the tumor,

primary tumor diameter larger than 2 cm, or grade 3

differentiation, were significantly more likely to receive

adjuvant radiotherapy (Table 2, Fig. 1).

Predictors of vaginal relapse in patients who did not receive

adjuvant radiotherapy

If only the 508 patients who did not receive adjuvant

radiotherapy are considered, both grade 3 differentiation and

Table 2

Percentage of patients receiving radiotherapy according to clinicopatho-

logic risk factor

Characteristic Radiotherapy,a % of patients P

Age, years 0.25

�60 17

>60 21

BMI, kg/m2 0.48

<30 20

�30 18

Stage <0.001

IA 2

IB 14

IC 66

Myometrial invasion, % <0.001

�50 12

>50 69

Invasion of LUS 0.1

No 19

Yes 31

Histologic subtype 0.02

Endometrioid 18

Nonendometrioid 36

LVI 0.01

Yes 40

No 17

Associated hyperplasia <0.001

No 27

Yes 10

Primary tumor diameter, cm <0.001

�2 5

>2 29

Histologic grade <0.001

1 7

2 31

3 44


invasion.a Information about adjuvant radiotherapy was not available for 2 patients.

of adjuvant radiotherapy.

LVI were significant predictors of vaginal relapse (P <

0.05), and nonendometrioid histologic subtype approached

significance (P = 0.08). However, the following were not

significant predictors of vaginal relapse (P > 0.05): age

older than 60 years, body mass index greater than 30 kg/m2,

invasion of the lower uterine segment, myometrial invasion

greater than 50%, tumor diameter larger than 2 cm,

hyperplasia associated with the tumor, and performance of

lymph node dissection. Results are summarized in Table 3.

Specifically, 2% of patients with a grade 1 or grade 2 tumor

had vaginal relapse at 5 years, compared with 11% of those

with a grade 3 tumor (P < 0.001). Of note, there was no

significant difference in the 5-year rate of vaginal relapse

between patients with a grade 1 tumor (2.2% rate) and those

with a grade 2 tumor (1.5% rate). Moreover, 3% of patients

with the endometrioid histologic subtype had vaginal

relapse at 5 years, compared with 10% of those with a

nonendometrioid histologic subtype (P = 0.08). Further-

more, 7% of patients with LVI had vaginal relapse at 5

years, compared with 3% of those without LVI (P = 0.02)

(Table 3). Cox regression analysis identified only grade 3

differentiation (P = 0.004; HR = 5.45; 95% CI, 1.70–17.48)

and LVI (P = 0.01; HR = 6.94; 95% CI, 1.55–31.03) as

independent predictors of vaginal relapse. When neither of

these 2 variables was present, 2% of patients experienced

vaginal relapse at 5 years, compared with 11% when either

variable was present (P < 0.001). Of the 2 patients who did

not have adjuvant radiotherapy but in whom both variables

were present, 1 had vaginal relapse 63 months after

diagnosis and the other 1 was alive with no evidence of

disease after 91 months of follow-up (Table 4). Again, when

performance of lymph node dissection was added to the Cox

regression model, grade 3 differentiation and LVI retained

Table 3

Percentage of vaginal recurrences at 5 years according to clinicopathologic

risk factor in the 508 patients who did not receive adjuvant radiotherapy

Characteristic Vaginal recurrence at 5 years, % P

Age, years 0.90

�60 3

>60 3

BMI, kg/m2 0.39

<30 4

�30 2

Myometrial invasion, % 0.39

�50 3

>50 0

Invasion of LUS 0.19

No 3

Yes 10

Histologic subtype 0.08

Endometrioid 3

Nonendometrioid 10

LVI 0.02

Yes 7

No 3

Associated hyperplasia 0.96

No 2

Yes 3

Primary tumor diameter, cm 0.63

�2 4

>2 3

Histologic grade <0.001

1–2 2

3 11

Lymph node dissection 0.66

Yes 3

No 3


invasion.

Table 4

Vaginal recurrence rates according to risk factorsa in 508 patients with stage

I endometrial cancer who did not receive adjuvant radiotherapy

Risk factora No. of patientsb 5-Year failure rate, %

None 448 2

Either one 57 11

Both 2 -

a Histologic grade 3 or lymphovascular invasion or both.b Neither histologic grade nor information about lymphovascular invasion

was available for 1 patient.- Of the 2 patients, 1 had vaginal relapse 63 months after primary surgery.


their status as the only independent predictors of vaginal

relapse.

Of note, 85% of patients who had myometrial invasion

>50% had lymph node dissection, compared with 41% of

those with myometrial invasion �50% (P < 0.001). Lymph

node dissection did not significantly influence the percent-

age of vaginal recurrences at 5 years in patients deemed at

high risk for vaginal relapse. In fact, if only patients with

grade 3 tumor are considered, 16% of those who did not

have lymph node dissection had vaginal recurrence at 5

years, compared with 6% in appropriately staged patients

(P = 0.24). Furthermore, if only patients with LVI are

considered, 0% of those who did not have lymph node

dissection had vaginal recurrence at 5 years, compared with

8% in appropriately staged patients (P = 0.53).

Vaginal relapses in patients with nonendometrioid tumors

Among the 33 patients with a nonendometrioid tumor, 12

(36%) received adjuvant radiotherapy (8 had external pelvic

radiotherapy only, 3 had external pelvic radiotherapy in

association with vaginal brachytherapy, and 1 had vaginal

brachytherapy only). The overall rate of vaginal relapse at 5

years was 9%. No discernible difference in the frequency of

vaginal relapse was observed between patients who received

adjuvant radiotherapy (8%) and those who did not (10%)

(P = 0.85). Vaginal relapse was not observed in the 4

patients who had vaginal brachytherapy.

Predictors of vaginal relapse in patients with endometrioid

tumor: stratified analysis according to stage, LVI, histologic

grade, and adjuvant radiotherapy

The 599 patients with an endometrioid tumor were

stratified on the basis of substage and the presence of LVI

and grade 3 differentiation. We observed that adjuvant

radiotherapy was generally administered to patients with

stage IC disease independently of LVI or grade. By

contrast, only a small number of patients with stage IA

or IB disease received radiotherapy. In patients with limited

myometrial invasion, the decision to administer adjuvant

radiotherapy was usually dictated by the presence of grade

3 differentiation (Fig. 1). Among patients with an

endometrioid tumor who did not receive adjuvant radiation,

most vaginal relapses were confined to the subgroup with

stage IA or IB disease and either LVI or grade 3

differentiation. For this group of patients, the rate of

vaginal relapse at 5 years was 12%, compared with 2.1%

for patients with stage IA or IB disease and neither LVI nor

grade 3 differentiation. Of interest, no vaginal relapse was

observed in the 34 patients with stage IC disease who did

not receive any adjuvant radiation (8 patients had grade 3

differentiation or LVI) (Fig. 2).

Discussion

In the present study, we analyzed factors predictive of

vaginal recurrence in endometrial cancer. The rate of vaginal

relapse in our series of patients with stage I disease who did

not have adjuvant radiotherapy was 3% at 5 years. This rate

is similar to that reported in the literature [11,12]. However,

because of the retrospective design of our study, the true

overall risk of vaginal recurrence in patients with stage I

disease may have been underestimated [13]. In fact, patients

with risk factors predictive of recurrence were more likely to

receive adjuvant radiotherapy, which would be expected to

lower the estimated risk of vaginal relapse (Table 2, Fig. 1).

Fig. 2. Vaginal relapse in stage I endometrioid endometrial cancer by risk

group. Only patients who did not receive adjuvant radiotherapy were

considered. The histogram shows the percentage of patients who had

vaginal relapse according to substage (IA– IB vs. IC), presence of

lymphovascular invasion (LVI), and histologic grade (G1–2 vs. G3). Note

that only patients with endometrioid tumor were considered. For 3 patients,

no information was available about the depth of myometrial invasion or

histologic grade; also, no information was available about the admin-

istration of adjuvant radiotherapy to 2 patients.


Our study included patients who had complete surgical

staging and those who did not. We reasoned that the

inclusion of only stage I patients who had lymph node

dissection would potentially select the tumors at higher risk

for lymph node involvement (and for recurrence). More-

over, in our series, the performance of lymphadenectomy

did not significantly influence the rate of vaginal recur-

rences at multivariate analysis and stratifying for risk factors

(i.e., grade 3, LVI). The disadvantage of including patients

who had incomplete surgical staging may have been

inclusion of occult stage III tumors in our analysis.

In accord with other reports [14–19], we observed that

vaginal brachytherapy was adequate for the prevention of

vaginal relapse. In fact, no vaginal relapse has been

observed in the 53 patients who had vaginal brachytherapy.

In contrast, vaginal relapse still occurred in 2.9% of patients

who had external pelvic radiotherapy alone.

In our series, histologic grade 3 subtype and LVI were

the strongest predictors of vaginal relapse (Tables 3 and 4).

In fact, patients with endometrioid stage IA or IB disease

who had LVI or a grade 3 tumor had a 5-year vaginal

relapse rate of 12% (Fig. 2). A comparison with the data

reported for patients with stage IA or IB disease and a

grade 3 tumor confirms our finding [11,12,20]. In

particular, poorly differentiated tumors have previously

been reported to portend a high risk of vaginal relapse in

patients not treated with adjuvant radiotherapy [11].

However, with few exceptions [6], the presence of LVI

has not been generally analyzed as a possible risk factor for

vaginal relapse [2,11,17,21,22]. The present association

between LVI and vaginal relapse could indicate that the

lymphatic or vascular route is involved in the mechanism

of vaginal recurrence.

Myometrial invasion generally has been reported as an

important predictor of vaginal relapse [3,23], indicating the

need for vaginal brachytherapy. The reported rate of vaginal

failure in patients with stage IC disease who did not have

adjuvant radiotherapy varies from 5% in appropriately

staged patients [12] to 10% in unstaged or incompletely

staged patients [2,24]. In our analysis, 85% of patients with

deep myometrial infiltration were in fact appropriately

staged, and myometrial invasion was not a significant

predictor of vaginal relapse, with 0% vaginal recurrence rate

at 5 years (Table 3, Fig. 2). Moreover, depth of myometrial

invasion did not predict the rate of vaginal recurrences even

after stratifying for lymph node dissection. In our previous

analysis, deep myometrial invasion was an important

predictor of distant hematogenous relapse [25]. Therefore,

the presence of deep myometrial invasion may be more

representative of a tumor that is prone to hematogenous

distant recurrences than to local relapse.

It should be noted that, in our series, most patients with

deep myometrial invasion had received adjuvant radio-

therapy (Table 2, Fig. 1), and many of them had been treated

with modified radical hysterectomy; hence, the above

findings may simply be a consequence of the retrospective

design of the study. In fact, it has been previously described

that patients who are at high risk for local relapse are the

ones who also have high risk of distant disease [26].

Therefore, it is theoretically possible that the reason only

distant relapse is evident in patients who have deep

myometrial invasion [25] is because local therapy has

achieved local control, allowing the distant disease to

manifest itself first. Nevertheless, in the present series,

patients who had deep myometrial invasion were more

likely to present with grade 3 tumor or LVI than patients

who had only superficial myometrial invasion. Thus, we

suggest that myometrial invasion may simply be a surrogate

marker for grade 3 differentiation or LVI, noting also that

LVI has not been generally analyzed as a possible risk factor

for vaginal relapse in the literature [2,11,17,21,22].

Owing to the relatively limited number of patients with

stage IC disease who did not receive any adjuvant treatment,

both in the literature [2,12,24] and in our series (Fig. 2), our

observation about the limited role of vaginal brachytherapy

in patients with deep myometrial invasion as the only risk

factor must be verified prospectively.

In the present study, a nonendometrioid histologic

subtype was a strong predictor of vaginal relapse (10% at

5 years if radiotherapy was not administered) (Table 3). The

lack of difference in the rate of vaginal relapse between

patients with nonendometrioid tumor who received adjuvant

radiotherapy and those who did not may be artifactual

because of the small number of patients involved. However,

it may indicate that external beam radiotherapy is less

effective than brachytherapy in preventing vaginal relapse in

this group. Consistent with this supposition was the absence

of vaginal relapse in the 4 patients who received vaginal

brachytherapy.


We previously reported predictors of hematogenous [25],

lymphatic [27], and peritoneal [28] relapse in endometrial

cancer. The present study, with the identification of

predictors of vaginal relapse, completes the analysis on

the different routes of metastasis in endometrial cancer. All

our findings may help in the decision to implement a target-

based plan of adjuvant therapy, based on the predicted

routes of tumor spread. In the presence of predictors of

hematogenous or peritoneal relapse, systemic therapy may

be efficacious [29]. Predictors of lymphatic relapse may

dictate the need for external adjuvant radiotherapy [30], and

predictors of vaginal relapse would suggest adjuvant vaginal

brachytherapy [14–19]. Hence, site-specific therapy would

be based on predictors of the four routes of relapse rather

than only on the stage of disease.

In conclusion, grade 3 differentiation and LVI were the

strongest predictors of vaginal relapse in our population.

The cost and morbidity of vaginal brachytherapy should be

balanced against the potential risk of vaginal relapse when

counseling patients who have stage I disease: a 2% risk of

vaginal relapse with low-grade lesions without LVI and 11%

with either grade 3 differentiation or LVI (or both).

Although the role of deep myometrial invasion in predicting

vaginal relapse appears to be a surrogate for grade 3 tumor

differentiation or LVI or both, verification awaits prospec-

tive analysis. Therefore, we currently recommend post-

operative vaginal brachytherapy only to patients with stage I

disease who have either one (or both) of the following 2

characteristics: grade 3 tumor differentiation or LVI.

Acknowledgments

Supported by the Mayo Cancer Center (P30CA15083)

and the Rochester Research Committee, Mayo Foundation.

References

[1] Jemal A, Murray T, Ward E, Samuels A, Tiwari RC, Ghafoor A, et al.

Cancer statistics, 2005. CA Cancer J Clin 2005;55:10–30.

[2] Creutzberg CL, et al. for the PORTEC Study Group. Surgery and

postoperative radiotherapy versus surgery alone for patients with

stage-1 endometrial carcinoma: multicentre randomised trial. Lancet

2000;355:1404–11.

[3] Elliott P, Green D, Coates A, Kreiger M, Russell P, Coppleson M,

et al. The efficacy of postoperative vaginal irradiation in preventing

vaginal recurrence in endometrial cancer. Int J Gynecol Cancer

1994;4:84–93.

[4] Tewari K, Cappuccini F, Brewster WR, DiSaia PJ, Berman ML,

Manetta A, et al. Interstitial brachytherapy for vaginal recurrences of

endometrial carcinoma. Gynecol Oncol 1999;74:416–22.

[5] Creutzberg CL, et al. for the PORTEC Study Group. Survival after

relapse in patients with endometrial cancer: results from a randomized

trial. Gynecol Oncol 2003;89:201–9.

[6] Alektiar KM, McKee A, Venkatraman E, McKee B, Zelefsky MJ,

Mychalczak BR, et al. Intravaginal high-dose-rate brachytherapy for

Stage IB (FIGO Grade 1, 2) endometrial cancer. Int J Radiat Oncol

Biol Phys 2002;53:707–13.

[7] Podratz KC, Mariani A, Webb MJ. Staging and therapeutic value of

lymphadenectomy in endometrial cancer. Gynecol Oncol 1998;70:

163–4.

[8] Creasman WT, Odicino F, Maisonneuve P, Beller U, Benedet JL,

Heintz AP, et al. Carcinoma of the corpus uteri. J Epidemiol Biostat

2001;6:47–86.

[9] Scully RE, Bonfiglio TA, Kurman RJ, Silverberg SG, Wilkenson EJ,

editors. Histological Typing of Female Genital Tract Tumours. 2nd ed.

Berlin’ Springer-Verlag; 1994. p. 13–8.

[10] Mariani A, Webb MJ, Keeney GL, Haddock MG, Calori G, Podratz

KC. Low-risk corpus cancer: is lymphadenectomy or radiotherapy

necessary? Am J Obstet Gynecol 2000;182:1506–19.

[11] Straughn Jr JM, Huh WK, Kelly FJ, Leath III CA, Kleinberg

MJ, Hyde Jr J, et al. Conservative management of stage I endo-

metrial carcinoma after surgical staging. Gynecol Oncol 2002;84:

194–200.

[12] Straughn JM, Huh WK, Orr Jr JW , Kelly FJ, Roland PY, Gold MA,

et al. Stage IC adenocarcinoma of the endometrium: survival com-

parisons of surgically staged patients with and without adjuvant

radiation therapy. Gynecol Oncol 2003;89:295–300.

[13] Keys HM, et al. Gynecologic Oncology Group. A phase III trial of

surgery with or without adjunctive external pelvic radiation therapy in

intermediate risk endometrial adenocarcinoma: a Gynecologic Onco-

logy Group study. Gynecol Oncol 2004;92:744–51 [Erratum in:

Gynecol Oncol. 94 (2004) 241–2].

[14] MacLeod C, Fowler A, Duval P, D_Costa I, Dalrymple C, Elliott P,

et al. Adjuvant high-dose rate brachytherapy with or without external

beam radiotherapy post-hysterectomy for endometrial cancer. Int J

Gynecol Cancer 1999;9:247–55.

[15] Mohan DS, Samuels MA, Selim MA, Shalodi AD, Ellis RJ, Samuels

JR, et al. Long-term outcomes of therapeutic pelvic lymphadenectomy

for stage I endometrial adenocarcinoma. Gynecol Oncol

1998;70:165–71.

[16] Seago DP, Raman A, Lele S. Potential benefit of lymphadenectomy

for the treatment of node-negative locally advanced uterine cancers.

Gynecol Oncol 2001;83:282–5.

[17] Fanning J. Long-term survival of intermediate risk endometrial cancer

(stage IG3, IC, II) treated with full lymphadenectomy and brachyther-

apy without teletherapy. Gynecol Oncol 2001;82:371–4.

[18] Chadha M. Gynecologic brachytherapy—II: intravaginal brachyther-

apy for carcinoma of the endometrium. Semin Radiat Oncol

2002;12:53–61.

[19] Bliss P, Cowie VJ. Endometrial carcinoma: does the addition of

intracavitary vault caesium to external beam therapy postoperatively

result in improved control or increased morbidity? Clin Oncol (R Coll

Radiol) 1992;4:373–6.

[20] Orr Jr JW, Holimon JL, Orr PF. Stage I corpus cancer: is teletherapy

necessary? Am J Obstet Gynecol 1997;176:777–88.

[21] Larson DM, Broste SK, Krawisz BR. Surgery without radiotherapy for

primary treatment of endometrial cancer. Obstet Gynecol

1998;91:355–9.

[22] Ng TY, Perrin LC, Nicklin JL, Cheuk R, Crandon AJ. Local

recurrence in high-risk node-negative stage I endometrial carcinoma

treated with postoperative vaginal vault brachytherapy. Gynecol Oncol

2000;79:490–4.

[23] Morrow CP, Bundy BN, Kurman RJ, Creasman WT, Heller P,

Homesley HD, et al. Relationship between surgical-pathological risk

factors and outcome in clinical stage I and II carcinoma of the

endometrium: a Gynecologic Oncology Group study. Gynecol Oncol

1991;40:55–65.

[24] Hogberg T, et al. Southern Swedish Gynecologic Oncology Group. A

prospective population-based management program including primary

surgery and postoperative risk assessment by means of DNA ploidy

and histopathology: adjuvant radiotherapy is not necessary for the

majority of patients with FIGO stage I– II endometrial cancer. Int J

Gynecol Cancer 2004;14:437–50.

[25] Mariani A, Webb MJ, Keeney GL, Calori G, Podratz KC. Hema-


togenous dissemination in corpus cancer. Gynecol Oncol 2001;80:

233–8.

[26] Corn BW, Lanciano RM, D_agostino R, Kiggundu E, Dunton CJ,

Purser P, et al. The relationship of local and distant failure from endo-

metrial cancer: defining a clinical paradigm. Gynecol Oncol 1997;

66:411–6.

[27] Mariani A, Webb MJ, Keeney GL, Aletti G, Podratz KC. Predictors of

lymphatic relapse in endometrial cancer. Gynecol Oncol 2002;84:

437–42.

[28] Mariani A, Webb MJ, Keeney GL, Aletti G, Podratz KC.

Endometrial cancer: predictors of peritoneal relapse. Gynecol Oncol

2003;89:236–42.

[29] Mariani A, Dowdy SC, Keeney GL, Long HJ, Lesnick TG, Podratz

KC. High-risk endometrial cancer subgroups: candidates for target-

based adjuvant therapy. Gynecol Oncol 2004;95:120–6.

[30] Mariani A, Webb MJ, Keeney GL, Podratz KC. Efficacy of

adjuvant radiotherapy (RT) in preventing pelvic and paraaortic

failure in patients with endometrial cancer (ENDOCA) and

positive lymph nodes. Int J Gynecol Cancer 2003;13(Suppl 1):75

[abstract].

Documents

Predictors of vaginal relapse in stage I endometrial cancer