Germ cell tumors (GCT) of the central nervous system (CNS) are thought to be derived from totipotent primordial germ cells, capable of both embryonic and extraembryonic differentiation. Based on the histologic components and the variable degree of differentiation, CNS GCTs are classified as germinomatous and nongerminomatous germ cell tumors (NGGCT). Germinomas comprise two-thirds of the CNS GCTs, and NGGCTs account for the remaining third.
The NGGCTs may be composed of elements of choriocarcinoma, endodermal sinus (or yolk sac) tumor, embryonal carcinoma or teratoma (mature or immature). Often, the NGGCTs are a mixture of the above elements. This variability and complexity of CNS GCTs leads to controversy in both diagnosis and management.
In addition, the rarity of CNS GCTs, comprising 1% to 2% of all primary CNS neoplasms, adds to the difficulty in determining optimal treatment. Very few prospective studies are available, and retrospective studies are limited based on the low number of patients involved, the variability in tumor size and location, histology, surgical approach, chemotherapy, and/or irradiation. In the past 2 decades, international cooperative trials have been conducted and advances have been made in treatment and prognosis.
CNS GCTs are typically midline tumors, most commonly seen in the pineal and/or suprasellar regions. Peak age at diagnosis is 10 to 12 years; however, CNS GCTs may be seen throughout childhood, adolescence, and young adulthood. The clinical presentation is dependent on the location of the tumor, whether suprasellar, pineal, or both. Common presentations include symptoms from increased intracranial pressure, visual tract involvement, and/or endocrine abnormalities.
If a CNS GCT is suspected, extent of disease evaluation should include: (1) high-resolution magnetic resonance imaging of the head and spine, with and without gadolinium; (2) evaluation of the cerebrospinal fluid (CSF) for cytology by lumbar puncture or sampling of ventricular fluid at time of shunt placement; (3) CSF and serum measurement of alpha-fetoprotein (AFP) and beta-human chorionic gonadotropin (BHCG); (4) baseline endocrine and neuropsychologic evaluations; and (5) a formal visual examination.
Radiologically, CNS GCTs cannot be distinguished from other CNS tumors. In the past, if patients had a pineal and/or suprasellar tumor, and suspected GCT, they were given a diagnostic trial of radiotherapy. If an early complete clinical response was seen, the patient was diagnosed with a germinoma. However, other pineal region tumors, as well as mixed NGGCT, may respond initially in the same manner and require very different treatment in order to prevent relapse. Therefore, this practice is no longer used.
The issue then arises regarding the necessity of a biopsy. A histologic specimen is unnecessary if the patient has a positive AFP or BHCG in the CSF and/or serum. Germinomas are generally negative for tumor markers, although they may secrete low levels of BHCG in the CSF (less than 100 mIU/mL). In NGGCTs, endodermal sinus tumors are associated with increased levels of AFP, while choriocarcinomas are associated with raised levels of BHCG. The secretion of these tumor markers in the CSF is pathognomonic for NGGCT, and no further histologic specimen is indicated. When low levels of BHCG are detected in the CSF, it is likely a germinoma with syncytiotrophoblastic cells, and the need for a histologic specimen is debatable. In the French Society of Pediatric Oncology (SFOP) experience, four out of nine patients with secreting germinomas were treated without a histologic diagnosis, and the outcome was the same as for germinomas with a histologic diagnosis.
All other patients with suspected GCT and negative tumor markers require a histologic specimen for diagnosis and treatment. Germinomas are exquisitely sensitive to radiotherapy with excellent cure rates, whereas NGGCTs have a poorer prognosis and require more intensive chemotherapy and irradiation. Given the differing natural histories and responses to treatment of germinomas and NGGCTs, histopathology to determine an optimal treatment strategy in tumor marker-negative patients is important.
It is possible, however, to make an erroneous diagnosis from a small biopsy specimen due to sampling error of a mixed GCT. Specifically, a diagnosis of a germinoma may be made from a small biopsy of a mixed GCT containing germinomatous elements (usually immature or mature teratoma). A gross total resection may provide greater tissue for histologic diagnosis, but given the location of these tumors and the resultant risk of postsurgical morbidity, a partial or total resection for tissue diagnosis is currently not recommended. Due to this risk of histologic sampling error, if any residual radiographic abnormality is present after two to four cycles of chemotherapy, the patient should undergo a "second-look" surgery.
Role of Surgery
Again, given the location of GCTs and the high postsurgical morbidity, the risks and benefits of surgery must be considered in light of the excellent response to irradiation and chemotherapy in germinoma patients. Based on the retrospective study of Sawamura et al, no further benefit was found in performing a resection of any kind--partial or complete--beyond treatment with irradiation and chemotherapy.
Unlike germinomas, the role of radical resection in NGGCTs is unclear, with no definitive studies having been conducted. It is possible that radical resection may increase survival rates in NGGCT. Current studies have supported the use of delayed resective surgery, or "second-look" surgery, if residual radiographic abnormalities are seen after chemotherapy and tumor markers have normalized. In this case, the residual lesion is likely to be teratoma or necrosis/fibrosis devoid of tumor. If it is a mature teratoma, surgery may be curative. These patients are then spared any further radiation therapy by performing the "secondlook" surgery.
If immature teratoma is present, then local-field irradiation is initiated without further chemotherapy.[ 3,5,6] In patients whose tumor markers have not normalized, the pathology from "second-look" surgery was also often consistent with either fibrosis or teratoma. However, the risk of subsequent recurrence or progression of disease was significant. Therefore, "second-look" surgery was not supported in cases with any elevation of tumor markers, as the surgery did not improve outcome or allow for a change in therapy.