Japanese Journal of Clinical Oncology

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 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 turn 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.

Table 1. Patients' characteristics

Age	yr
Mean	55
Range	25-82
Stage	No.
IA-IC	195
IIA-IIB	48
IIIA-IIIC	67
Pathology	No.
Adenocarcinoma	294
Adenosquamous	16
Hysterectomy	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	%
Mean	22
Median	10

RLN, retroperitoneal lymph 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 (1,8). 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 IIIC 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 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.

RESULTS

DISCUSSION

The generally accepted prognostic factors for RLN metastasis and survival of endometrial cancer were evaluated in this study (1-15). 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.

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 turn 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 of RLN 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 of RLN 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 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-lymphatic 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 IIIC 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 [ge]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 of RLN 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.