Ovarian YST is encountered most frequently in the second and third decades of life, followed by the first and fourth decades, and is extremely rare in postmenopausal women. Only 58 cases of ovarian YST among postmenopausal females have been recorded in the English literature during the past 45 years based on a MEDLINE database search [3–32]. We retrospectively reviewed the clinicopathological characteristics of all the published literatures of postmenopausal YSTs in Table 1. Many studies have postulated that postmenopausal YSTs have distinct molecular pathways that are different from those occurred in younger patients, leading to different clinicopathologic features and prognosis.
Table 1
The clinicopathologic characteristics of 58 cases of ovarian yolk sac tumors in postmenopausal women.
Characteristic
|
|
Number of patients
|
%
|
Age,years
|
|
|
|
Median=60
|
|
|
|
Range=50-86
|
|
|
|
Stage
|
|
|
|
Ⅰ
|
|
23
|
40.3
|
Ⅱ
|
|
6
|
10.5
|
Ⅲ
|
|
18
|
31.5
|
Ⅳ
|
|
2
|
3.5
|
N/A
|
|
8
|
14
|
Histological subtype
|
|
|
|
Pure YST
|
|
21
|
36.8
|
Mixed:YST+epithelial ovarian cancer
|
29
|
50.9
|
YST+MMMT
|
|
2
|
3.5
|
YST+EC/IT
|
|
3
|
5.7
|
YST+NET
|
|
2
|
3.5
|
Preoperatively Serum AFP
|
|
|
|
elevated
|
|
28
|
49.1
|
normal
|
|
5
|
8.8
|
N/A
|
|
24
|
41.4
|
Preoperatively Serum CA125
|
|
|
|
elevated
|
|
17
|
29.8
|
normal
|
|
6
|
10.5
|
N/A
|
|
34
|
59.6
|
Site of tumor
|
|
|
|
left ovary
|
|
18
|
31.6
|
right ovary
|
|
29
|
50.9
|
Bilateral ovary
|
|
6
|
10.5
|
Uterus
|
|
1
|
1.8
|
N/A
|
|
3
|
5.3
|
Tumor size (median=20cm)
|
|
|
|
<20cm
|
|
21
|
36.2
|
≧20cm
|
|
7
|
12.1
|
N/A
|
|
30
|
51.7
|
Adjuvant Chemotheraphy
|
|
|
|
Yes
|
|
33
|
56.9
|
BEP
|
|
10
|
17.2
|
TC
|
|
4
|
6.9
|
BEP and TC
|
|
4
|
6.9
|
Others
|
|
15
|
25.9
|
No
|
|
6
|
10.3
|
N/A
|
|
19
|
32.8
|
Outcome(DFS median=24 months; OS median=10 months)
|
|
DFS<24 months
|
|
10
|
17.2
|
DFS≧24 months
|
|
6
|
10.3
|
DOD<10 months
|
|
12
|
20.7
|
DOD≧10 months
|
|
15
|
25.7
|
N/A
|
|
15
|
25.9
|
YST: Yolk sac tumor; MMMT: malignant Müllerian mixed tumor; EC: embryonal carcinoma; IT: immature teratoma; NET: neuroendocrine tumor; DFS: disease-free survival; OS: overall survival; DOD: died of disease; BEP: bleomycin, etoposide, cisplatin; TC: paclitaxel, carboplatin; N/A: not avaiable. |
Postmenopausal YSTs occur most often in association with a somatic-type malignancy. The most commonly associated somatic neoplasm is endometrioid carcinoma. secondly, other associated tumors were high-grade serous carcinoma, clear cell carcinoma, mucinous tumors, and MMMT. Up to now, the pathogenesis of postmenopausal YSTs has not been well understood. There are four theories to recommend: the teratoma theory, retrodifferentiation, the collision theory, and the neometaplasia theory [6, 16, 18, 30]. The theory that YSTs originate from germ cells is highly unlikely at an older age, as germ cells are not identifiable histologically in the ovaries of the postmenopausal female. Nogales [11]postulated that both components of postmenopausal YSTs arise from a common, non-germ-cell precursor. Neometaplasia, known as aberrant differentiation, refers to the germ cell component that arises through a process of transformation of the epithelial precursor neoplasm [6, 30]. The most frequently reported epithelial precursor neoplasm is endometrioid carcinoma, which sometimes develops from endometriosis or endometriotic cysts [33]. This is in line with the hypothesis that endometriosis or endometriotic cysts may be a risk factor for YST in postmenopausal women [24]. However, the molecular events involved in this transformation have not yet been elucidated. Many reports consider that the germ cells are developed from teratoma to be unlikely, as most of the tumors lack the histologic diversity and presence of immature embryonic tissue seen in immature teratoma or the more mature fetal or adult tissue seen in mature cystic teratoma associated with carcinoma. Herein, we report a new case of ovarian YST in a 64-year-old woman. We identified diverse differentiation of the tumor based on morphologic features and the immunohistochemical profile, including YST, Müllerian epithelial components, heterologous elements including chondrosarcoma and rhabdomyosarcoma, and the intestinal variant. A variety of tumor components are also considered to be derived from the somatic component by a process of “neometaplasia” or “retrodifferentiation”.
In fetal development, AFP is produced by the yolk sac, liver, and upper gastrointestinal tract. It has been demonstrated that germ cell tumors in patients with elevated serum AFP are either composed entirely of or contain yolk sac tumor elements [34]. Some studies have shown that serum AFP always elevated pre-operatively in YSTs. In our study, the serum AFP was measured pre-operatively only in 33 cases, in which elevated AFP was in 28 cases, and normal in 5 cases. Boussios suggested that in older women with an ovarian mass and an elevated serum AFP level, a diagnosis of ovarian YST should be suspected [3]. Meanwhile, we found that the serum AFP level was not different between pure YSTs and mixed YSTs in postmenopausal women (P>0.05). There were 2 cases of ovarian YST mixed with a MMMT reported in the previous English literature. The serum AFP level was elevated slightly to 23µg/L (normal 0-20µg/L) in one case, and corresponding data in another case were not available. It is interesting that the serum AFP of our case reported here did not elevate pre-operatively or post-operatively. It should be noted that the serum AFP determination is useful diagnostically and in monitoring therapy, but a normal result may not always indicate the absence of active disease, especially in postmenopausal females. CA-125 has been found to be a valuable tumor marker in adenocarcinoma patients. The reported cases of ovarian YSTs had elevated serum CA-125 in 17 patients, and most of them were mixed YSTs which were associated with epithelial cancer. The serum CA-125 elevated to 65.10 U/ml (normal, 0-35 U/ml) in our case. It should be borne in mind that the diagnosis of ovarian YST even if the serum CA-125 elevated but AFP was negative result in older women.
The morphological and immunophenotypical features of such tumors markedly overlapped with epithelial neoplasms such as clear cell carcinoma, endometrioid carcinoma, serous carcinoma and malignant Müllerian mixed tumor which can make diagnosis difficult. An immunohistochemical panel, including CK7, EMA, PAX-8, WT-1, estrogen receptor (ER), progesterone receptor (PR), HNF-1, and Napsin A, should be applied for identifying the diverse tumor components. It is generally accepted that the positive immunostains for AFP, SALL4, and glypican-3 are valuable markers in identifying YSTs components histologically [35]. SALL4, a zinc finger transcription factor, is a specific marker for germ cell tumors, but in addition to YSTs, it is positive in dysgerminoma and embryonal carcinoma, among other tumors. Although, SALL4 appears to be the more sensitive marker for distinguishing YST from other epithelial tumors [21], the aberrant staining patterns were always likely to occur and the positivity of SALL4 was found in more than 20% of ovarian serous carcinomas [36]. AFP is more specific for YSTs, and can be identified not only in the tumor cells but also in the eosinophilic globules present both inside and outside the tumor cells, but the immunostains of AFP always was weak or negative. Glypican-3, an oncofetal protein expressed in the fetal liver and malignant tumors of hepatocytic lineage, is more sensitive than AFP but not as specific [37]. Positive glypican-3 appears variable in clear cell carcinoma, and when positive, it does not reliably distinguish it from YSTs [38]. There are significant overlap in the immunophenotype between postmenopausal YSTs and epithelial neoplasms. CK7 and EMA were often positive in glandular YSTs, sometimes diffusely, which may reflect the probable derivation of the YST components from the somatic epithelial neoplasms [30]. Other germ cell tumors, such as dysgerminoma, express OCT3/4, CD117 and D2-40, in addition to SALL4, which can be differentiated from YSTs. Embryonal carcinoma is usually positive for OCT-3/4 and CD30 and negative for SALL4, AFP and glypican-3, while YST typically displays the opposite pattern. Positiveа-inhibin and calretinin can be used to distinguish adult granulosa cell tumors from YSTs. The morphological and immunohistochemical features of our case were favour of a diagnosis of ovarian YST associated with MMMT. For the ovarian tumors occurred in older women with complex morphologically components, a set of immunophenotypes should be used for diagnosis and differential diagnosis.
Due to the malignancy of postmenopausal YSTs, the prognosis is poor, and the three-year survival rate is 13% [28]. Some studies have demonstrated that the tumor stage [39], serum AFP decline rate[40], residual tumor [41], volume of ascites [40] were prognosis-related factors in ovarian YSTs. Since the introduction of several combination chemotherapy regimens, including CBC (cisplatin-based chemotherapy), VBP (vinblastine, bleomycin, cisplatin)., PVP (cisplatin, vinblastine, peplomycin), TC (paclitaxel, carboplatin) and BEP, et al, the prognosis of ovarian YSTs has dramatically improved. For women with an ovarian YST treated with the BEP regimen, the 5-year survival rate approaches 94%, and even in patients with advanced-stage disease [39], consequently, the BEP regimen has been taken for the standard of care for ovarian YSTs. Platinum-based adjuvant chemotherapy aiming to treat both epithelial ovarian neoplasms and germ cell tumors, which are components that are identifiable or postulated to be a part of the tumor at one time, has also been used. It was reported that postmenopausal YSTs may be less sensitive to chemotherapy than those that arise de novo [18], as the tumor component arises by transformation of an epithelial ovarian neoplasm. To date, there wasn’t a report on the analysis of prognosis-related factors for postmenopausal YSTs, moreover, there was no effective chemotherapy regimens recommended due to the rarity. Herein, we showed that advanced-stage disease (HR=5.367, P=0.0205) and different adjuvant chemotherapy regimen (HR=3.992, P=0.0004) were associated with poor overall survival by using univariate and multivariate analysis, and the selection of adjuvant chemotherapy regimen (HR=1.091, P=0.003) was a favorable independent prognostic factor for postmenopausal YSTs, whether the somatic-type malignancies were mixed with or not. Age, serum AFP, serum CA125, tumor size and tumor location were not associated with overall survival (P>0.05). What calls for special attention is that the first-line adjuvant chemotheraphy regimen with BEP or TC or the combination of BEP and TC may be quite effective against postmenopausal YSTs, but more clinical data will need to be accumulated in the future.