Serum carcinoembryonic antigen as a tumour marker in patients with endometrial cancer.

BACKGROUND
No potential tumour markers have been validated for prognosis in endometrial cancer. However, carcinoembryonic antigen (cea) is one of the most widely used tumour markers in various types of cancer. Although cea expression in endometrial cancer has been investigated, its prognostic value remains controversial, and no studies have investigated serum cea levels in large case series. In the present study, we investigated diagnostic and prognostic applications of serum cea for endometrial cancer.


METHODS
This prospective study was approved by our Institutional Review Board. Between January 2006 and December 2012, serum cea was measured prospectively in 215 patients with endometrial cancer and was subsequently measured during treatment and at scheduled follow-up examinations in patients with elevated baseline serum cea.


RESULTS
During the study period, 215 patients (142 stage i, 19 stage ii, 32 stage iii, 22 stage iv) were treated for endometrial cancer. By the time of last follow-up, 52 had relapsed (24.2%), and the median follow-up duration was 45 months (range: 1-95 months). Elevated serum cea was identified in 25 patients (11.6%) and was associated with histologic type (p = 0.04), histologic grade (p = 0.03), and myometrial invasion depth (p = 0.01). Elevated serum cea was not related to clinical stage, lymph node metastasis, distant metastasis, age, menopausal status, or body mass index. Relapse of disease was related to elevated serum cea (p = 0.006).


CONCLUSIONS
Serum cea is a potential prognostic indicator for endometrial cancer.


INTRODUCTION
Endometrial cancer is the most common cancer of the female genital tract, and its incidence is increasing globally. Although several biomarkers have been associated with clinical characteristics and prognosis in endometrial cancer [1][2][3][4][5][6] , none have been implemented in clinical practice. Novel biomarkers for endometrial cancer are therefore required.
Carcinoembryonic antigen (cea) is a 200-kDa glycoprotein, described first in 1965 by Gold and Freedman 7 , when they identified the antigen that was present in both fatal colon and colon adenocarcinoma, but that appeared to be absent from healthy adult colon. Serum cea is one of the most widely used tumour markers, having been reported to be elevated in many cancers, such as those of colon and rectum, breast, lung, pancreas, liver, stomach, thyroid, prostate, bladder, and ovary [8][9][10] .
Although cea expression has been investigated in endometrial cancer, controversy continues to revolve around its associations with that malignancy. Neunteufel et al. 11 and Duk et al. 12 reported that cea was expressed in 58% and 62.5% of endometrial carcinoma patients respectively. A few reports that investigated serum cea levels revealed that serum cea was elevated in endometrial cancer patients in 22% of cases 13,14 . Only those few small studies have investigated serum cea in patients with endometrial cancer; no large studies of endometrial cancer have reported cea levels.
In the present work, we investigated the diagnostic and prognostic value of serum cea in patients with endometrial cancer.   15 . Histologic diagnoses were confirmed by microscopy examination of hematoxylin-and eosin-stained sections according to World Health Organization criteria. In operable cases, patients received total abdominal or radical hysterectomy plus bilateral salpingooophorectomy. Peritoneal fluid samples were obtained for cytology testing, and systemic pelvic lymphadenectomy was performed in most cases. Para-aortic lymph node sampling was performed in patients with intermediate-or high-risk disease. In patients without lymph node adenectomy, lymph nodes larger than 1 cm were detected by computed tomography or magnetic resonance imaging (or both) and were considered positive. After surgery, adjuvant chemotherapy was provided for patients with intermediate-or high-risk disease. No hormonal therapies or targeted treatments such as monoclonal antibodies or tyrosine kinase inhibitors were administered.
Treatment was followed by gynecologic examinations, including transvaginal or abdominal ultrasonography (or both) and cytology testing of vaginal cut edges, and by laboratory examinations, including assessments of tumour marker expression. Further computed tomography or magnetic resonance imaging exams were performed for patients with clinically suspicious symptoms or elevated tumour marker levels. Recurrent disease was diagnosed by biopsy or imaging methods.
Serum cea was measured before therapy, during the pre-treatment examinations. A lower level of 5.0 ng/mL was used as the cut-off for normal serum cea. In patients with elevated baseline serum cea, cea was measured again during treatment and at scheduled follow-up examinations. The intervals between follow-up appointments were every 2 months during years 1-3 after treatment, every 6 months during years 4-5 after treatment, and annually from year 6 after treatment.

Statistical Analyses
Relationships between clinical groups were analyzed using the Fisher exact probability test. In the analyses of disease relapse, survival distributions were calculated using the Kaplan-Meier method, and relapse-free patients included those with no evidence of disease recurrence and those who died from unrelated causes. Univariate and multivariate Cox regression analyses were used to identify variables associated with relapse-free survival (histologic type, histologic grade, clinical stage, myometrial invasion depth, lymph node metastasis, distant metastasis, age, menopausal status, and body mass index), and associations were considered significant at p < 0.05.

Serum CEA in Patients with Endometrial Cancer
Table ii shows serum cea measurements and clinical characteristics for the patients. Elevated serum cea was detected in 25 patients (11.6%). Compared with patients having endometrioid adenocarcinoma, those with other histologic disease types had significantly more elevated serum cea measurements (25.0% vs. 9.9%, p = 0.04); similarly, serum cea was significantly elevated in patients with histologic grade 3 disease than in those with grade 1 or 2 disease (19.7% vs. 8.4%, p = 0.03). Serum cea was also significantly elevated in patients with a myometrial invasion depth exceeding 1/2 (19.5%) than in those with a myometrial invasion depth less than 1/2 (7.2%, p = 0.01). Elevated serum cea was not associated with clinical stage, lymph node metastasis, distant metastasis, age, menopausal status, or body mass index.
Of the 25 patients with elevated serum cea, 17 (68.0%) achieved remission. Although serum cea fell to within the defined normal range in 11 of those patients, it did not fall in 6 patients. However, 12 of the 25 patients (48.0%) relapsed, with a concomitant increase in serum cea in every case. Of the 6 patients whose serum cea did not fall into the defined normal range, none experienced disease recurrence. In the patients with elevated serum cea, relapse of disease was significantly more frequent than it was in the patients with normal serum cea (23.1% vs. 8.0%, p = 0.006).

DISCUSSION
Numerous studies have investigated biomarkers for endometrial cancer 1-6 . In particular, measurement of cancer antigen 125 (CA125) in serum has been investigated as a tumour marker in patients with endometrial cancer. Duk et al. 12 reported that CA125 is a useful marker in endometrial carcinoma, and Patsner et al. 15 demonstrated its utility as a marker for monitoring treatment effects in patients with advanced endometrial cancer. However, the precise role of CA125 in preoperative evaluations of endometrial cancer patients remains uncharacterized and controversial 16 . Nonetheless, no other potential tumour markers have been validated for predicting prognosis and classifying endometrial cancers, warranting identification of tumour markers that can inform the management of patients with endometrial cancer.
On the other hand, serum cea, one of the most widely used tumour markers, is reported to be elevated in many cancers, such as those of colon and rectum, breast, lung, pancreas, liver, stomach, thyroid, prostate, bladder, and ovary [8][9][10] . The identification of abnormal pre-and postoperative serum cea could be a useful auxiliary for prognosis or postoperative surveillance in colorectal cancer patients 10 . Although some publications have investigated cea expression or serum cea level, they included only small numbers of patients with endometrial cancer [11][12][13][14] . One report investigated serum cea for detection of recurrent disease in endometrial carcinoma patients, and although that report concluded that serum cea was useful for the detection of recurrent disease, only stage i patients were included, and the postoperative therapy was radiotherapy 17 . Taken together, those reports indicate the requirement for a comprehensive investigation, in a large series of endometrial cancer patients, of serum cea as a tumour marker for preoperative diagnosis and subsequent management.
To the best of our knowledge, the present study is the first to address the foregoing issues and to clarify the diagnostic and prognostic value of serum cea as a tumour marker for endometrial cancer. Elevated serum cea was identified in 25 of 215 patients with endometrial cancer (11.6%). In general, 15%-25% of patients with disease that is clinically confined to the uterus have elevated serum CA125; approximately 75% of patients with metastatic disease have elevated serum CA125 15 . Thus, elevated serum cea in 10.6% of our patients with stages i and ii disease indicates diagnostic value comparable to that for serum CA125 in patients with early-stage disease. The present study also shows that significantly elevated serum cea is more common among patients with histologic grade 3 disease (19.7%) than in those with grades 1 and 2 disease combined (8.4%, p = 0.03). Moreover, elevated serum cea was significantly more prevalent in patients with a histologic disease type other than endometrioid adenocarcinoma (25.0% vs. 9.9% in those with endometrioid adenocarcinoma, p = 0.04). The latter finding reveals a significant correlation of serum cea with loss of tumour differentiation. Immunohistochemical studies by Neunteufel et al. 11 showed that cea expression is associated with the histologic grade of endometrial cancers. Our data accord with the data in their report. Furthermore, serum cea was also significantly elevated in patients with a myometrial invasion depth greater than 1/2 (19.5%) than in those with myometrial invasion depth less than 1/2 (7.2%, p = 0.01). No report has investigated the relationship between serum level or expression of cea and depth of myometrial invasion. Elevated serum cea was not associated with clinical stage, lymph node metastasis, distant metastasis, age, menopausal status, or body mass index. Although Cherchi et al. showed that there was no relationship between serum cea and surgical pathologic stage 13 , Heyer et al. 18 showed that serum cea measurements differed significantly between patients with and without hepatic and pulmonary metastases at the time of initial diagnosis.
The relationship between serum cea and disease status in patients with endometrial cancer remains controversial. On the other hand, in our study, patients with elevated serum cea experienced relapse of disease significantly more frequently than did the patients with normal serum cea (23.1% vs. 8.0%, p = 0.006). Therefore, measurements of serum cea might provide further prognostic information after surgical risk stratification.
In summar y, serum cea was determined before therapy during the pre-treatment examinations, and then during treatment and at follow-up for patients with elevated serum cea. Subsequent analyses indicated that serum cea, as a marker of endometrial cancer risk, might facilitate prognosis prediction and could inform clinical management. Of 25 patients with elevated serum cea, 17 (68.0%) achieved remission. Although serum cea fell to within the normal range in 11 of those patients, it did not fall in 6 patients. Nevertheless, 12 of the 25 patients (48.0%) relapsed, all with concomitant increases in serum cea, indicating that serum cea might reflect disease status. Serum cea might therefore facilitate the management of endometrial cancer patients with elevated serum cea.

CONCLUSIONS
The present study is the first to investigate serum cea at follow-up and strongly indicates the potential of cea to be a clinically useful prognostic factor in endometrial cancer.