The sensitivity and specificity of a new formula to distinguish endometrioid type endometrial carcinoma from ovarian endometrial carcinoma

https://doi.org/10.1016/j.ejogrb.2009.09.016Get rights and content

Abstract

Objectives

Endometrioid type adenocarcinoma sometimes occupies both endometrium and ovary and in some cases the origin cannot be determined.

Study design

In this study, we established a formula to distinguish ovarian endometrioid cancer (EOC) from endometrioid type endometrial cancer (EEC), based on our previous report of cyclin and KI67 expression pattern by immunohistochemistry of 36 EECc and 37 OECc by the logistic regression. We calculated the diagnostic accuracy using 92 test samples retrospectively and finally could diagnose the origin of 16 cases in whom endometrioid type adenocarcinoma arose in both ovary and endometrium and could be determined by Scully's criteria, and 15 cases in whom endometrioid type adenocarcinoma arose in both ovary and endometrium and Scully's criteria were not usuful retrospectively.

Results

The estimated formula is as follows: Logit(Prob(EOC)) = −1.1437  0.0853 CNA + 0.0423 CNB + 0.173 CND1 + 0.0129 CNE + 0.0224 CNF + 0.0508 KI67, where Prob(EOC) is the probability that a clinical sample is EOC. If Prob(EOC) is larger than 0.5, the diagnosis is ovarian cancer; if less than 0.5 it is endometrial cancer. Finally, using the formula, 37 of 48 EECs (77.1%) and 33 of 44 EOCs (75.0%) were correctly classified, with an accuracy of 76.1% (p < 0.0001), retrospectively. In 12 of the 16 cases (75%) who could be determined by Scully's criteria, the origin determined by Scully's criteria was concordant with the origin determined by the formula retrospectively. In the other 15 cases, 12 cases were judged as ovary/ovary, 2 cases were judged as uterus/uterus and 1 case was judged as uterus/ovary.

Conclusion

The formula we established was thought to be useful to distinguish the origin of the cases in whom endometrioid type adenocarcinoma arises in both ovary and endometrium.

Introduction

Endometrial and ovarian cancers are the most common cancers of the female genital tract. Both can be subdivided into different histologic types, which include endometrioid type adenocarcinomas. Endometrioid type endometrial cancer (EEC) is the most common form of cancer of the endometrium, comprising 75–80% of cases (1). In 75% of all cases, the tumor is confined to the uterine corpus at the time of diagnosis, and uncorrected survival rates of 75% or more are expected [1]. Ovarian endometrioid cancer (EOC) accounts for a relatively small proportion (8%) of epithelial ovarian cancers [1]. Approximately 75–85% of patients are diagnosed at the time when their disease has spread throughout the peritoneal cavity [1]. When controlled for stage, the survival rate for endometrioid tumors is poor and similar to that for serous tumors [2]. EOCs morphologically resemble EECs, though their biological behaviors are very different [1].

Endometrioid type adenocarcinoma sometimes occupies both endometrium and ovary. Patients with disease of both the endometrium and the ovary can be classified into three groups: (A) endometrial cancer with metastasis to the adnexa, (B) ovarian cancer with metastasis to the endometrium, or (C) synchronous primary cancers of the endometrium and ovary. It is important to distinguish these three statuses to determine the therapeutic schedule. Scully et al. described a similar but more extensive list of clinicopathologic features used to differentiate endometrial disease with metastasis to the ovary, ovarian disease with metastasis to the endometrium, and independent primary cancers [3]. However, in some cases the origin could not be determined, because this criterion is not enough for clinical practice. Shedden et al. performed microarray analysis for 33 EOCs and 13 EECs and found specific genes for EOC or EEC [4]. The microarray is thought to be useful to distinguish EEC and EOC, but this method cannot easily be performed in clinical practice, because it has several problems as follows: (A) expensive, (B) difficult to treat samples, (C) reproducibility, and (D) no standard of microarray tip. We previously compared cyclin A, B1, D1, E, F and KI67 expression in 37 EOCs and 38 EECs and found that the expression pattern differed between EOC and EEC [5]. In this study, we established a formula to distinguish EOC from EEC, based on the difference of cyclin and KI67 expression, and calculated the diagnostic accuracy using test samples retrospectively. We finally attempted to diagnose the origin of 15 cases in whom endometrioid type adenocarcinoma arose in both ovary and endometrium and Scully's criteria were not useful retrospectively.

Section snippets

Materials and methods

We performed this study based on the STARD methods [6].

Test samples

To calculate the diagnostic accuracy, we applied this formula for test samples retrospectively. All patient-derived specimens were collected and archived under protocols approved by the institutional review boards (IRBs) of the Osaka City General Hospital and Kagoshima City Hospital. All tissues were obtained as archival specimens that had been formalin-fixed and processed for routine pathological analysis. A total of 48 EECs and 44 EOCs were randomly recruited in this study. Fifty-seven cases

Accuracy of formula for test samples

This study was performed between July 2005 and June 2006. There were no significant differences in age and stage between EEC and EOC, however, grade 3 tumor was found more often in EOC than EEC (20.5% vs. 2.1%, p = 0.006). Estrogen receptors were present in 17 of 44 EOCs and 39 of 48 EECs (38.6% vs. 81.3%, p = 0.006). The comparison of each cyclin and KI67 protein expression between EEC and EOC is shown in Table 3. Cyclin F and KI67 expression were higher in EOC than EEC, however, cyclin A

Discussion

Endometrioid type adenocarcinoma sometimes occupies both endometrium and ovary. The clinical staging system differs between EEC and EOC. The standard operation method also differs between EOC and EEC. In EOC, postoperative adjuvant therapy is thought to be combination chemotherapy consisting of carboplatin plus paclitaxel, however, in EEC, postoperative adjuvant therapy is thought to be radiation therapy or combination chemotherapy consisting of cisplatin and doxorubicin. It is important to

References (12)

  • Y. Hashiguchi et al.

    Combined analysis of p53 and RB pathways in epithelial ovarian cancer

    Human Pathol

    (2001)
  • J.H. William et al.

    Principles and practice of gynecologic oncology

    (2000)
  • J. Zwart et al.

    Five-year survival in patients with endometrioid carcinoma of the ovary versus those with serous carcinoma

    Eur J Gynaecol Oncol

    (1998)
  • R.E. Scully et al.

    Tumors of the ovary, maldeveloped gonads, fallopian tube, and broad ligament. Atlas of tumor pathology

    (1998)
  • K.A. Shedden et al.

    Histologic type, organ of origin, and Wnt pathway status: effect of gene expression in ovarian and uterine carcinoma

    Clin Cancer Res

    (2005)
  • S. Nishimura et al.

    Comparison of cyclin expression between endometrioid-type endometrial cancer and endometrioid ovarian cancer

    J Clin Oncol

    (2004)
There are more references available in the full text version of this article.

Cited by (1)

  • Expression of Estrogen Receptors (α, β), Cyclooxygenase-2 and Aromatase in normal endometrium and endometrioid cancer of uterus

    2013, Advances in Medical Sciences
    Citation Excerpt :

    It is well recognized that estrogens (Estrone - E1, Estradiol – E2) play an important role in the development and progression of EC [2]. Results from many clinical and epidemiological studies demonstrated that excessive and/or prolonged exposure to estrogen (not opposed by progesterone) is an important risk factor for the development of EC, especially that of the endometrioid type (type 1 by Bokhman's classification) [3–5]. However, many other risk factors for EC development have been recently implied including: high postmenopausal body mass (obesity), hypertension, hormone replacement therapy, anovulatory states, early menarche, late menopause and null parity [6–9].

View full text