Preoperative CA 125 Tumour Marker in Endometrial Cancer by yaofenjin



Preoperative CA 125 Tumour Marker
in Endometrial Cancer: Correlation
With Advanced Stage Disease
Alexandra Sebastianelli, MD, FRCSC, Marie-Claude Renaud, MD, FRCSC, Jean Grégoire, MD, FRCSC,
Michel Roy, MD, FRCSC, Marie Plante MD, FRCSC
Department of Gynecologic Oncology, Centre Hospitalier Universitaire de Québec, pavillon Hôtel-Dieu de Québec, Quebec City QC

Abstract                                                                    préopératoire de CA 125 et de données cliniques et histopathologiques
                                                                            exhaustives. Le test du chi carré, le test exact de Fisher, une
Objective: To evaluate if a preoperative serum CA 125 level ³ 35 kU/L       régression logistique multivariée et la courbe d’efficacité du
  in patients with endometrial cancer correlates with a surgical stage      récepteur ont été utilisés aux fins de l’analyse statistique.
  III or IV and poor histopathological prognostic factors.
                                                                         Résultats : Un total de 186 (73 %) patientes présentaient une
Methods: We conducted a retrospective cohort study of 254 patients         pathologie de stade 1 ou 2; 68 patientes (27 %) présentaient une
  who underwent hysterectomy and full staging for endometrial cancer.      pathologie de stade 3 ou 4. Un nombre statistiquement plus élevé
  Preoperative serum CA 125 was available for each patient as well         de patientes du groupe « stade 3 ou 4 » présentaient un taux
  as complete clinical and histopathological data. Chi-square, Fisher      sérique de CA 125 ³ 35 kU/l (58 %), par comparaison avec les
  exact test, multivariate logistic regression, and receiver operating     patientes du groupe « stade 1 ou 2 » (16 %) (RC, 7,44; P < 0,001).
  characteristic curve curves were used for statistical analysis.          Aucune corrélation n’a été constatée entre le taux sérique de CA
Results: A total of 186 (73%) patients had stage I or II disease and       125 et le sous-type histologique. Les patientes qui présentaient
  68 (27%) had stage III or IV disease. A statistically higher number      une pathologie de stade 1 ou 2 et un taux sérique de CA 125 ³
  of patients from the stage III or IV group had a serum CA 125            35 kU/l (46 %) connaissaient une invasion myométriale profonde
  level ³ 35 kU/L (58%) compared with the stage I or II group (16%)        considérablement plus souvent (> 50%) que les patientes qui
  (OR 7.44; P < 0.001). There was no correlation between serum             présentaient un taux sérique de CA 125 < 35 kU/l (18 %) (RC,
  CA 125 level and histological subtype. Patients with stage I or II       3,68; P = 0,006).
  disease and serum CA 125 ³ 35 kU/L (46%) had significantly             Conclusion : La détermination du taux sérique préopératoire de CA
  more frequent deep myometrial invasion (> 50%) than did those            125 est un test très simple qui permet de détecter les patientes
  with serum CA 125 < 35 kU/L (18%) (OR 3.68; P = 0.006).                  présentant un adénocarcinome endométrial de stade plus avancé.
Conclusion: Assay of the preoperative serum CA 125 level is a very         Son utilisation systématique pourrait contribuer au triage
  simple test to detect patients with more advanced stage endometrial      préopératoire des patientes exposées à des risques élevés.
  adenocarcinoma. Its routine use could help triage high risk patients    J Obstet Gynaecol Can 2010;32(9):856–860

                                                                              ndometrial cancer is the most common gynaecologic
Objectif : Évaluer si la présence d’un taux sérique préopératoire
  de CA 125 ³ 35 kU/l chez les patientes qui présentent un cancer
  de l’endomètre est en corrélation avec la présence d’un stade
                                                                         E    cancer.1 Its reported incidence is 14.5/100 000 women,
  chirurgical 3 ou 4 et de piètres facteurs pronostiques                 and it consequently affects 3900 Canadian women each
  histopathologiques.                                                    year. Endometrial cancer is also known to be the fourth
Méthodes : Nous avons mené une étude de cohorte rétrospective            commonest cancer overall and the seventh leading cause of
  auprès de 254 patientes ayant subi une hystérectomie et une
  stadification complète en raison d’un cancer de l’endomètre.
                                                                         death by cancer in North America.2 It is well recognized
  Pour chacune des patientes, nous disposions du taux sérique            that the prognosis for this disease is intimately related to the
                                                                         surgical stage. Histological subtype, grade, and involvement
                                                                         of the lymph vascular space are also important features to
 Key Words: CA 125, endometrial cancer, prognostic factors,
                                                                         consider with regard to the need for adjuvant treatment.
 tumour markers                                                          CA 125 is a cell surface antigen expressed by different types
 Competing Interests: None declared.                                     of tumour cells. Its serum concentration is generally
 Received on March 28, 2010
                                                                         proportional to the size of the tumour mass. Measurement
                                                                         of serum CA 125 is commonly used in evaluation and
 Accepted on June 15, 2010
                                                                         follow-up of patients with ovarian cancer.3 However there

                                Preoperative CA 125 Tumour Marker in Endometrial Cancer: Correlation With Advanced Stage Disease

is little published information regarding preoperative             clinicopathological factors on serum CA 125 levels. P <
CA 125 levels in patients with endometrial cancer and the          0.05 was considered statistically significant. A receiver oper-
relationship of these levels to stage of disease.4                 ating characteristic curve was used to determine the optimal
Approximately 75% of patients with endometrial cancer are          CA 125 level that would best predict metastatic disease. For
diagnosed at an early stage, when the prognosis is usually         the purpose of this analysis, the early stage disease group
excellent. However, 25% of patients are diagnosed at more          included patients with stage I and II carcinoma, and the
advanced stages.5 In these cases, thorough surgical staging        advanced stage disease group included patients with stage
becomes very important to guide the need for adjuvant              III and IV carcinoma.
treatment. The serum CA 125 level is thus of growing interest in
the preoperative investigation of women with endometrial
cancer because of its potential to predict the presence of         A total of 351 patients were initially entered in the study.
more advanced stage disease. Patients suspected of having          Ninety-two patients were excluded because there was no
more advanced stage disease could then potentially be              tumour marker recorded in their medical chart. Five more
triaged and referred preoperatively to gynaecologic                patients were also excluded because of potentially con-
oncologists for complete surgical staging.                         founding medical conditions, leaving 254 patients for final
The primary objective of this study was to determine               analysis. The mean age of the group was 63.3 ± 11.3 years.
                                                                   One hundred eighty-five patients (73%) had stage I or II
whether a preoperative CA 125 level ³ 35 kU/L correlates
                                                                   disease and 68 (27%) had stage III or IV disease. Among
with an endometrial cancer of surgical stage III or IV. The
                                                                   the patients with early stage disease, 53 were stage IA, 92
secondary objective was to determine whether a preoperative
                                                                   were stage IB, and 22 were stage IC. Stage II patients were
CA 125 level ³ 35 kU/L was associated with a more aggressive       distributed as follows: eight were stage IIA and 10
histological subtype (clear cell or papillary serous) or with      were stage IIB. A total of 49 patients were stage III and
other poor histopathological prognostic factors.                   were distributed as follows: 27 were stage IIIA, none were
                                                                   stage IIIB, and 22 were stage IIIC. A total of 14 patients had
                                                                   positive pelvic washings, but only six patients were classi-
We conducted a retrospective cohort study of all patients          fied as stage IIIA solely on the basis of cytology. Of the 27
with a diagnosis of primary endometrial cancer at                  stage IIIA patients, six were classified as such because of the
L’Hôtel-Dieu de Québec between October 1993 and                    positive washings and the remainder had either serosal or
October 2003. The diagnosis of endometrial carcinoma was           adnexal involvement. Stage IIIC patients had positive lymph
established preoperatively by endometrial biopsy or                nodes. Five of these patients had positive para-aortic lymph
fractional curettage. Preoperative investigation, including a      nodes, and the others had positive pelvic nodes. However,
complete history and physical examination, a chest X-ray,          the para-aortic lymph node status was known for only 84
and abdominal and pelvic ultrasound, was carried out for           patients, because the others did not undergo para-aortic
each patient, and this work-up included a serum CA 125             lymph node dissection. For purposes of analysis, all of these
measurement.                                                       patients were grouped as node positive patients. A total of
Patients were staged according to the FIGO surgical staging        19 patients were classified as stage IV, and all of these were
system adopted in 1998.6 All had peritoneal washings for           stage IVB. Two were categorized as stage 4B because of
cytology, a total abdominal or laparoscopically assisted           bone metastasis; the remainder were categorized on the
vaginal hysterectomy, bilateral salpingo-oophorectomy,             basis of abdominal tissue involvement (Table 1).
and pelvic node sampling. Omentectomy and para-aortic              With regard to serum CA 125 levels, 186 patients (73%) had
node sampling were performed for clear cell and papillary          values < 35kU/L, and 68 (27%) had levels ³ 35 kU/L.
serous subtypes and for grade 3 tumours. Detailed                  There were no differences between patients with low and
histopathological data were available for each patient.            high CA 125 levels in preoperative characteristics, including
Patients who did not have a CA 125 level recorded preoper-         menopausal status, cigarette smoking, and parity (Table 2).
atively or who underwent surgery in another hospital were          The median follow-up time was 45 months, and four patients
excluded from the study. Women with medical problems               were lost to follow-up. Twenty-five patients (9.8%) experienced
which could potentially elevate CA 125 levels                      a recurrence, and a total of 19 (7.4%) died of the disease.
(endometriosis, other past or actual neoplasia) were also          Of the 68 patients with advanced stage disease, 39 (58%)
excluded.                                                          had a CA 125 level ³ 35 kU/L, compared with 30 (16%) of the
Fisher exact test, chi-square testing, and multivariate logistic   186 patients with early stage carcinoma (OR 7.44; P < 0.001).
regression were performed to examine the effects of                The association between poor histopathological prognostic

                                                                                       SEPTEMBER JOGC SEPTEMBRE 2010 l        857

 Table 1. Stage III and IV patients                                 Table 2. Preoperative patient characteristics

 Stage                                              n                                     CA 125            CA 125
                                                                                         < 35 kU/ L        ³ 35 kU/L
 Stage III: Total 49                                                Characteristics        n (%)             n (%)          P
      Stage IIIA                                  27/49
         Cytology                                  6/49
                                                                        No                33/184             12/68         NS
         Serosal or adnexal disease               21/49                                    (17.9)            (17.6)
      Stage IIIB                                   0/49                 Yes               151/184            56/68
      Stage IIIC (pelvic or para-aortic nodes)    22/49                                    (82.1)            (82.4)

 Stage IV: Total 19                                                 Smoker

      Stage IVA                                     0                   No                159/185            60/68         NS
                                                                                           (85.9)            (88.3)
      Stage IVB                                   19/19
                                                                        Yes               26/185              8/68
                                                                                           (14.1)            (11.8)

factors and CA 125 level is shown in Table 2. A significant             0                 46/183             27/66         NS
                                                                                           (25.1)            (40.9)
increase in serum CA 125 level was noted in patients
                                                                        ³1                137/183            39/66
with grade 3 tumours, deep myometrial invasion,                                            (74.9)            (59.1)
lymphovascular space involvement, cervical invasion,
positive cytology, adnexal involvement, and distant metastasis
(Table 3).
Among the group of patients with early stage disease, a             Table 3. Relation between serum CA 125 level and
                                                                    histopathological prognostic factors
significant increase in serum CA 125 was noted in patients
with deep myometrial invasion. However, there was no                                      CA 125           CA 125
correlation with lymphovascular space invasion or cervical                               < 35 kU/L        ³ 35 kU/L
                                                                    Factor                 n (%)            n (%)          P
involvement (Table 4).
                                                                    Grade 3                33/167           23/61        < 0.001
With regard to the histological subtype, among the 186 patients                            (19.8)           (37.7)
with a CA 125 level ³ 35 kU/L, 28% had a non-                       Myometrial             28/137           24/61        0.005
endometrioid histological tumour sub-type, compared with            invasion > 50%         (20.4)           (39.3)
20% of those with a CA 125 level < 35 kU/L (P = 0.2).               Lymphovascular         36/186           31/68        < 0.001
                                                                    space involvement      (19.4)           (45.6)
The sensitivity of the CA 125 level in predicting advanced          Cervical               14/186           20/68        < 0.001
stage carcinoma using a cut-off value of 35 kU/L was                involvement            (12.9)           (29.4)
58.2% and the associated specificity was 84%. The positive          Positive cytology      14/186           16/67        < 0.001
predictive value of the CA 125 level using 35 kU/L as a                                     (7.5)           (23.9)
cut-off was 57% while the negative predictive value was             Adnexal                6/186            18/68        < 0.001
                                                                    involvement            (3.2)            (26.5)
83%. The receiver operating characteristic curve demon-
                                                                    Positive nodes         14/186           26/68        < 0.001
strates that the sensitivity could be improved using a cut-off                              (7.5)           (38.2)
value of 25 kU/L. Using this cut-off, the calculated sensitivity    Distant metastasis     8/185            18/68        < 0.001
increased to 67.2%, while the specificity decreased to 77.2%                               (4.3)            (26.5)
(Figure). The positive predictive value with the modified
cut-off value was 49%, and the negative predictive value
was 86%.
                                                                   cytology will probably not benefit from a lymph node dis-
DISCUSSION                                                         section and staging procedure, in contrast to patients cate-
                                                                   gorized as having stage IIIA disease on the basis of serosal
We conducted a retrospective review of 254 patients with a         or adnexal involvement. There is little chance that this
preoperative diagnosis of endometrial cancer. Our data             cohort of patients influenced our data, since they represented
clearly show a significant increase in preoperative serum CA       only six out of the 68 patients with advanced stage disease.
125 levels for patients with a stage III or IV endometrial
carcinoma compared with patients who had stage I and II            A significant rise in serum CA 125 levels was also noted in
disease (P < 0.001). We know that patients categorized as          the subgroup of patients with early stage disease in the pres-
having stage IIIA disease solely on the basis of positive          ence of deep myometrial invasion (P < 0.006). This finding

                                Preoperative CA 125 Tumour Marker in Endometrial Cancer: Correlation With Advanced Stage Disease

                                                                  Receiver operative curve: optimal CA 125 level
 Table 4. Relation between serum CA 125 level and
 histopathological prognostic factors in early stage
                                 CA 125 ³ 35kU/L
 Factor                               n (%)               P
 Lymphovascular space
     Yes                             22/152              NS
    No                                 7/32
 Myometrial invasion
    < 50%                            13/100             0.006
    > 50%                             11/31
 Cervical involvement
    Yes                              25/166              NS
    No                                 4/18

has not previously been reported, and could be of substantial
clinical interest because patients with deep myometrial
invasion benefit from complete surgical staging and require
                                                                  markers had a recurrence of disease, and CA 125 was
adjuvant treatment and rigorous follow-up.
                                                                  elevated in six of these seven patients.16 According to
Our data compare favourably with previously published             Denschlag et al.,18 an elevated CA 125 level in stage III
data. In 2002, Hsieh et al.7 associated elevated preoperative     endometrial cancer patients is a bad prognostic sign.
CA 125 levels with an increased risk of metastatic disease at
the time of surgery and concluded that elevated preoperative
CA 125 levels could be a good indication for performing           The optimal CA 125 level associated with an advanced stage
full pelvic lymphadenectomy.7 Following a review of               endometrial carcinoma is not established. Sensitivity of the
92 patients, Chung et al.8 reported in 2006 that an elevated      commonly used cut-off value of 35 kU/L was not found to
CA 125 level was a good predictor of lymph node metastasis        be very high (57%). Our study suggests that lowering the
in women with endometrial cancer. Our study also agrees           cut-off level to 25 kU/L could increase its diagnostic sensi-
with the findings of Patsner et al.,9 as well as others.10–14     tivity but at the expense of a lower specificity. This finding
Among their 89 patients, Patsner et al.9 found 20 of 23           is different from that reported by Hsieh et al.7 In their
patients with extrauterine disease were found to have an          cohort of 124 patients, a value of 40 kU/L was determined
elevated serum CA 125. However, in 1998, Koper et al.15           to be optimal using the receiver operating characteristic
stated that this single measurement was not sensitive             curve. We acknowledge that the sensitivity and positive
enough to warrant performing a lymph node dissection.             pretictive value of CA 125 measurements are low. How-
Dotters4 also reported that CA 125 lacked sensitivity in          ever, measurement of serum CA 125 is inexpensive and
identifying patients requiring staging.                           easy. Our study showed that this simple blood test identified
The main strength of this study is the large number of            77% of patients who required a staging procedure, with an
patients included. To our knowledge, it is the largest series     87% negative predictive value, using 25 kU/L as reference
on the subject. Its retrospective design reduces its strength,    value. In comparison, patients with stage I epithelial ovarian
although only four patients were lost to follow-up. The role      carcinoma exhibit an elevated serum CA 125 in only 50% of
of measuring CA 125 in the follow-up of ovarian cancer            cases.19 Despite this low sensitivity, finding an elevated CA
patients is well known. The accuracy of CA 125 levels in          125 level is considered to be an indication for referral in
predicting recurrence of endometrial cancer in patients with      postmenopausal woman. We believe that an elevated CA 125
preoperative elevated tumour markers has been shown in            level in a patient with endometrial cancer should be a factor
the study of Lo et al.16 and Rose et al.17 Lo et al. found that   in deciding to refer a patient for lymph node dissection and
seven of 12 patients with preoperative elevated tumour            staging purposes.

                                                                                     SEPTEMBER JOGC SEPTEMBRE 2010 l        859

CONCLUSION                                                                            and prognosis in endometrial cancer. Acta Obstet Gynecol Scand
Measurement of serum CA 125 is a very simple test to perform                       9. Patsner B, Mann WJ, Cohen H, Loesch M. Predictive value of preoperative
in the preoperative evaluation of women with endometrial                              serum CA 125 levels in clinically localized and advanced endometrial
                                                                                      carcinoma. Am J Obstet Gynecol 1988;158:399–402.
cancer. Its routine use could help triage high-risk patients
                                                                                  10. Ebina Y, Sakuragi N, Hareyama H, Todo Y, Nomura E, Takeda M, et al.
for referral to a gynaecologic oncologist, because an ele-                            Para-aortic lymph node metastasis in relation to serum CA 125 levels and
vated level is associated with an increased risk of advanced                          nuclear grade in endometrial carcinoma. Acta Obstet Gynecol Scand
stage disease.                                                                        2002;81:458–65.
                                                                                  11. Schneider J, Centeno M, Saez F, Genolla J, Ruibal A. Preoperative CA-125,
AKNOWLEDGEMENTS                                                                       CA 19–9 and nuclear magnetic resonance in endometrial carcinoma:
                                                                                      correlation with surgical stage. Tumour Biol 1999;20:25–9.
The authors wish to thank François Harel, MSc, for statistical                    12. Santala M, Talvensaari-Mattila A, Kauppila A. Peritoneal cytology and
support.                                                                              preoperative serum CA 125 level are important prognostic indicator
                                                                                      of overall survival in advanced endometrial cancer. Anticancer Res
REFERENCES                                                                            2003;23(3C):3097–104.

 1. Ben-Shachar I, Pavelka J, Cohn DE, Copeland LJ, Ramirez N, Manolitsas         13. Lo SS, Cheng DK, Ng TY, Wong LC, Ngan HY. Prognostic significance
    T, et al. Surgical staging for patients presenting with grade 1 endometrial       of tumour markers in endometrial cancer. Tumor Biol 1997;18:241–9.
    carcinoma. Obstet Gynecol 2005;105:487–93.                                    14. Takac I, Gisek B. Serum CA 125 levels and lymph node metastasis in patient
 2. Chand JK, WU H, Heung MK, Shin JY, Osann K, Kapp DS. The                          with endometrial cancer. Wien Klin Wochenschr 2006;118(Suppl 2):62–5.
    outcomes of 27,063 women with unstaged endometrioid uterine cancer.
                                                                                  15. Koper NP, Massuger LF, Thomas CM Kiemeney LA, Verbeek AL.
    Gynecol Oncol 2007;106: 282–8.
                                                                                      Serum CA 125 measurement to identify patients with endometrial cancer
 3. Cannistra SA. Cancer of the ovary. N Engl J Med 2004;351:2519–29.                 who require lymphadenectomy. Anticancer Res 1998;18(3B):1897–902.
 4. Dotters DJ. Preoperative CA 125 in endometrial cancer: is it useful?          16. Lo SS, Khoo US, Cheng DK, Ng TY, Wong LC, Ngan HY. Role of serial
    Am J Obstet Gynecol 2000;182:1328–34.
                                                                                      tumor markers in the surveillance for recurrence in endometrial cancer.
 5. Sorosky JI. Endometrial cancer. Obstet Gynecol 2008;111:436–47.                   Cancer Detect Prev 1999;23:397–400.
 6. International Federation of Gynecology and Obstetrics (FIGO). Annual          17. Rose PG, Sommers RM, Reale FR, Hunter RE, Fournier L, Nelson BE.
    report on the results of treatment in gynaecologic cancer. Int J Gynecol          Serial serum CA 125 measurement for evaluation of recurrence in patients
    Obstet 1989;28:189–93.
                                                                                      with endometrial carcinoma. Obstet Gynecol 1994;84:2–16.
 7. Hsieh CH, Changchien CC, Lin H, Huang EY, Huang CC, Lan KC, et al.
    Can a preoperative CA 125 level be a criterion for full pelvic                18. Denschlag D, Tan l, Patel S, Kerim-Dikeni A, Souhami L, Gilbert L.
    lymphadenectomy in surgical staging of endometrial cancer? Gynecol                Stage III endometrial cancer: preoperative predictability, prognostic factors,
    Oncol 2002;86:28–33.                                                              and treatment outcome. Am J Obstet Gynecol 2007;196:546.e1–7.
 8. Chung HH, Kim JW, Park NH, Song YS, Kang SB, Lee HP. Use of                   19. Carlson KJ, Skates SJ, Singer DE. Screening for ovarian cancer. Ann Intern
    preoperative serum CA-125 levels for prediction of lymph node metastasis          Med 1994;121:124–32.


To top