Adult type testicular granulosa cell tumors (GCT) are rare sex cord–stromal tumors. However, subsequent to a thorough study of the literature we were able to find 17 cases above and beyond those referred to in one of the recent literature reviews (28 cases)
. Fifteen out of the 45 (33%) reported cases were published after the year 2000, which does not imply that incidence actually increased, but rather, that the tumor is more frequently being recognized.
Morphological diagnosis is based primarily on the typical morphology of the granulosa cells with their coffee-bean like, angulated and grooved nuclei. Macrofollicles ought to be present, the presence of the Call Exner bodies makes correct diagnosis easier; however, they are not always found and thus, are not indispensable to diagnosis
. Young and Scully
 require that “all or almost all” of the tumor should be composed of granulosa cells. In fact, undifferentiated Sertoli cell tumors can have a very similar cytomorphological features but form predominantly ill-defined tubular structures resembling primitive testicular tubules and do not have follicular structures. If the morphological features are ambiguous, the term “incompletely differentiated” (WHO) is preferable. The very rare fibrothecoma of the testis can be also mistaken for undifferentiated sex cord tumor since it manifests the same immunoreactivity as all other stromal tumors. This tumor, however, is composed of highly monotone spindle cells which are embedded in an acellular fibrous stroma
To the best of our knowledge, this is the first case of GCT with heterologous sarcomatous changes observed in the testis. The sarcomatous cells did not show any differentiation, they were negative for desmin and SMA. As in many soft tissue sarcomas, the cells were also positive for pancytokeratin antibodies (AE1:AE3) in our case. With the exception of one case which showed “some evidence of rhabdomyosarcomatous differentiation
”, the sarcoma cells in all other cases described in the ovaries did not show any special differentiation
One could argue that the observed sarcomatous component originates from a concomitant germ cell tumor growing simultaneously with the granulosa tumor. Quite apart from the fact that the tumor has been entirely processed and no germ cell tumor remnant could be found, a smooth transition from the differentiated to the sarcomatous component could be observed. Moreover, the simultaneous occurrence of germ cell and stromal tumors in the unilateral testis is extremely rare without any common genetic or oncogenetic background
. In contrast to the germ cells, the cells of the gonadal stroma tumors are differentiated and not pluripotent; can therefore not give rise to somatic type malignancies (carcinomas, differentiated sarcomas) as the cells of teratomas do. The somatic type malignancies arising in teratomas are well differentiated, completely identical to the tumors originally developing in other organs e.g. soft and nervous tissue, intestine etc.
, whereas the sarcomas described to date in stromal tumors
[32–35] were all composed of undifferentiated spindle cells. It can be assumed that these tumors develop from undifferentiated gonadal stroma.
A molecular characterization of sarcomatous changes in ovarian GCT showed a down-regulation of the GCT-specific genes such as inhibin, estrogen receptor and FSH receptor
. A significant up-regulation of genes with an inflammatory response was consistent with the presence “of a marked inflammatory infiltrate”, which was a striking morphological feature in our case as well. However, contrary to these findings of McNeilage et al.
, the PR receptor was not down-regulated in our case, but immunohistochemically strongly expressed.
Some authors claim that testicular GCT can be mistaken for lymphoma, metastatic carcinoma or even for melanoma
[12, 35, 36], but lymphoma and metastatic carcinoma cells tend to spread in the interstitium between testicular tubules and usually do not form solid tumor masses
Immunohistochemistry is obviously helpful for achieving a correct diagnosis. One must keep in mind, however, that with few exceptions all types of sex cord/gonadal stroma tumor cells show reactivity for the same antibodies. By and large, but not in all cases, the tumor cells of GCT are positive for inhibin, calretinin, vimentin and CD 99
[20, 31]. According to Renshaw et al.
 SMA and S-100 are constantly expressed in all testicular sex cord/gonadal stroma tumors, but other authors cannot confirm these findings
[20, 30]. CD 99 is consistently expressed, whereas EMA has never been detected in the different types of sex cord/gonadal stroma tumors of the testis; these antibodies are thus useful for the differential diagnosis stromal tumor vs. carcinoma metastasis
[20, 31, 40]. Pancytokeratin and desmin are often but not constantly detected
[12, 20, 31]. Also very useful to the diagnosis of these tumors is the detection of estrogen (ER) and progesterone receptors (PR), and in this connection PR are detected more often than ER (75% vs. 50%)
. PlAP immunohistochemistry is obviously the best method to distinguish the stromal from the germ cell tumors of the testes. The PR positivity in the sarcomatous part of the tumor presented here is the only evidence that these cells derive from the granulosa cells and not from the supporting stroma.
A more difficult task is the correct classification of poorly differentiated sex cord/stroma tumors, since the granulosa and Sertoli cells can have a similar appearance and can also express all of the immunohistochemical markers discussed above. Almost all Leydig and granulosa-cell tumors, but only about half of Sertoli-cell tumors, react with inhibin and calretinin antibodies
. A valuable marker could be the melan-A/MART-1 (A 103) melanoma antibody, which is positive in all Leydig and in many Seroli-cell tumors, but negative in GCTs
. CD 56 is constantly expressed in GCT, but has never been tested in the diagnostics of the testicular GCTs. CD56 is a regulator of growth and differentiation in ovarian folliculogenesis
One of the major problems of histological diagnosis of testicular sex cord/stroma tumors is the prediction of malignancy. Atypical mitotic figures occur more frequently and the mitotic index is significantly higher in malignant tumors; vascular invasion, infiltrative margins, and necrosis are also associated with malignancy, but are not very reliable
 when used as a single criterion. The only reliable predictor of malignancy in GCT as in all other stromal tumors is a tumor diameter > 5 cm
[1, 17, 31].
Due to the few cases reported, our knowledge of the biological behavior and the clinical features of this tumor is still fragmentary. The average age of the diseased men is 45 years (Range 12 – 83 years) with an accumulation (45%) of cases in the 5th and 6th decade (Table
1). In 65.5% of cases (21/32) the tumors were smaller than 2 cm (1.7 cm in median, range: 0.7 – 15 cm). Endocrine symptoms (gynecomastia, loss of libido and potency) were observed only in six cases, 26 cases did not have such symptoms and in the other cases this clinical information is missing (for references, see Table
1). In the known cases the duration of clinical signs ranged from a few months to 10 years! Nineteen case reports included a follow up: 2 patients died of disease, one of them 11 years after diagnosis
[7, 17], 3 patients died of unrelated diseases
[12, 20], 3 patients were alive with disease, one of them 6 years after surgery
[16, 17, 23]. The longest follow-up of one of the 13 patients who were alive without evidence of disease was 14 years
Even this small series of cases shows that testicular adult type GCTs behave in an aggressive manner in one quarter of cases. The clinical problem is that there is no specific therapy for these tumors. Various chemotherapeutic agents and regimens have been used for patients with metastatic disease with limited or no success. The “prophylactic” retroperitoneal lymphadenectomy is an option which is very successful in stage I diseases, but not in the higher stages when the lymph nodes are already affected
In summary, we present a unique case of a testicular granulosa cell tumor of adult type with a sarcomatous component whose impact on the clinical course is unknown.