Germinomas are rare malignant primary brain tumors thought to be arising from primordial germ cells of the yolk sac endoderm. They form part of the spectrum of germ cell tumors (GCTs) and are the commonest type of GCT in the central nervous system (CNS). In the CNS, the histopathological appearances are termed germinoma and account for approximately 60% of all primary intracranial GCTs. Intracranial GCTs are histopathologically identical to their gonadal counterparts with a testicular primary being referred to as a seminoma and an ovarian primary as a dysgerminoma.
Germinomas account for 1–3% of all intracranial neoplasms and commonly arise in midline structures adjacent to the third ventricle. The most common regions involved are the pineal (∼45%) and suprasellar (∼35%) areas. In around 15% of cases, concurrent lesions are detected in both the pineal gland and the suprasellar area or very rarely the basal ganglia or the brainstem. These tumors are termed bifocal or multiple midline GCTs.
Germinomas have a tendency to spread through the subependymal lining and cerebrospinal fluid (CSF). Approximately, 5–10% of patients present with either microscopic or macroscopic metastatic disease in the CSF at the time of diagnosis. Extraneural metastases at diagnosis are extremely uncommon.
There is a marked geographical variation in the incidence of germinomas. In the West, they account for 0.4–3.4% of primary central nervous system tumors while in Japan and the Far East they are five to eight times more frequent.
Germinomas are more commonly found in males with a male to female ratio of approximately 2:1 and predominantly affect patients in their teens with approximately 75% of patients’ diagnosed with a primary CNS GCT being in the age range of 10–20 and are a recognized tumor entity in CNS childhood cancer. The WHO classification divides germ cell tumors into subtypes depending on the final differentiated cell. With regards to their management, they are divided into three major subgroups: pure germinomas, secreting (malignant) non-germinomatous germ cell tumors, and teratomas. Classically, germinomas consist of large uniform cells with clear cytoplasm. Syncytiotrophoblastic cells can be present and may secrete low levels of beta human chorionic gonadotrophin (β-HCG). Despite this, germinomas are often referred to as non- secreting germ cell tumors if β-HCG levels are ≤50 IU/l in either serum or CSF. There is no evidence to suggest that a slight elevation of β-HCG levels is associated with a worse outcome.
Presenting symptoms of a germinoma depend on the anatomical site of the primary tumor and its growth rate. The classical symptom of pineal primaries is Parinaud syndrome (paralysis of upward gaze, headache, and impaired pupillary constriction to light with preservation of accommodation). In addition, tumors in this location frequently compress the Sylvian aqueduct leading to obstructive hydrocephalus, indicated by symptoms of raised intracranial pressure (diurnal headache, vomiting, and lethargy). The commonest presenting symptoms of suprasellar tumors are obstructive hydrocephalus, endocrine and visual field defects, and/or reduced visual acuity. The frequent involvement of the pituitary stalk and the proximity of the tumor to the hypothalamopituitary axis lead to diabetes insipidus (DI). This symptom can precede the radiological and pathological diagnosis of a suprasellar germinoma sometimes by up to a few years. Other symptoms include hypothalamic damage, for example, anorexia or weight gain, somnolence, mood swings, disrupted sleep pattern, electrolyte imbalances, temperature dysregulation, failure to thrive, precocious puberty, secondary amenorrhea, and panhypopituitarism.
When a germinoma is suspected based on the clinical history and physical examination, magnetic resonance imaging (MRI) with intravenous contrast enhancement of the brain and spine is the diagnostic imaging of choice to establish the precise location and extent of the disease. The classic appearance of a germinoma on MRI is an iso- or hyperdense well-defined lesion that uniformly enhances with contrast. If the clinical picture and radiological findings are suspicious of a primary CNS GCT, tumor marker analysis is important to differentiate between secreting GCTs and germinomas as pathologically elevated tumor markers establish a diagnosis without the requirement for tissue diagnosis. When the tumor marker assays are negative, a histopathological verification of primary tumor is mandatory. In addition, an examination of the CSF must be performed in all patients, to confirm or refute microscopic tumor dissemination (M1 disease), and establish the presence or absence of elevated CSF tumor markers (αFP, β-HCG). An isolated elevation of CSF tumor markers in the presence of negative serum markers establishes the diagnosis of a secreting germ cell tumor and subsequently defines the appropriate management strategy. If CSF cannot be obtained intraoperatively a lumbar puncture should only be performed at diagnosis, if the ventricles are not obstructed or an adequate CSF diversion has been performed in patients with obstructive hydrocephalus. Alternatively, a lumbar puncture should be performed within 1–2 days of the surgical intervention to relieve raised intracranial pressure as the half-life of βHCG is only 5 days after complete macroscopic resection. It has to be acknowledged that there is currently no uniformly validated laboratory test for the measurement of CSF tumor markers although this is an internationally agreed mandatory requirement for staging CNS germinomas. To complete staging, a CSF analysis for cytology should be performed 14 days after the surgical intervention to avoid false positive results.
The differential diagnosis of a pineal mass includes pinealoblastoma, pineocytoma, glioma, or benign cysts. At a suprasellar location, the possible differential diagnoses include, predominantly, optic chiasm and hypothalamic gliomas, craniopharyngiomas, and Rathke’s pouch cysts.
In an emergency situation, an endoscopic ventriculostomy or ventriculo-peritoneal shunt with or without a prior external ventricular drainage (EVD) will alleviate raised intracranial pressure, secondary to obstructive hydrocephalus, and/or could be used to obtain diagnostic tissue. Otherwise, the standard diagnostic neurosurgical procedure is an open biopsy of the mass lesion. There is, however, increasing use of less invasive techniques such as stereotactic and endoscopic biopsy. Modern neurosurgical techniques using a stereotactic approach are deemed safe and mortality is in the region of 0.5% in experienced hands with similar diagnostic yield rates.
There is no advantage in attempting a gross total resection in intracranial germinomas as it does not alter management or outcome, but it can be associated with significant surgical-related morbidity. Radical surgery is reserved for the very few patients with a significant residual mass following completion of treatment which usually represents a residual mature teratoma component on histopathological assessment.
Histological subtype is the single most important prognostic factor for outcome as the natural history and treatment of different histopathological subtypes are quite distinct.
Excess of 90% of germinomas are radiocurable when treated in a tertiary specialist neuro-oncology unit with close links to an endocrine service. Historically, the gold standard treatment for germinomas has been conventionally fractionated craniospinal radiotherapy (using external beam ionizing radiation therapy) followed by a boost to the primary site given the extremely high radiation sensitivity of these tumors. The pattern of relapse in localized germinomas is dominated by ventricular recurrences, and it is unusual to develop an isolated spinal cord relapse. A recent literature review confirmed that there is good evidence to suggest that in completely staged patients with localized germinomas irradiation of the ventricles, followed by a boost to the primary tumor area gives equivalent long-term control rates compared to wide-field craniospinal radiotherapy. Originally, the craniospinal axis was treated with a dose of 30–35 Gy followed by a boost of 10–20 Gy in 7–12 fractions to the primary site. Over the last two decades, consecutive studies have demonstrated that a reduction of the dose to the craniospinal axis to 24 Gy is equivalent with respect to long-term disease-free survival. In addition, there was no loss of local control when reducing the primary tumor dose from 50–54 to 40–45 Gy. The use of three-dimensional planning and conformal radiotherapy, in conjunction with reduced volumes, is likely to minimize the amount of normal tissue irradiated to high doses of radiotherapy. This is likely to reduce treatment-related late sequelae, for example, growth impairment, neurocognitive disabilities, cerebrovascualr events, and second malignant neoplasms.
There is some evidence to suggest that the primary tumor dose can be further reduced if a complete response can be demonstrated following multiagent induction chemotherapy in primary localized germinomas. However, there is controversy as to what volume requires irradiation in these patients. In addition, controversy exists how to classify bifocal tumors. In Europe, bifocal tumors are regarded as localized disease whilst in the United States these patients are treated using strategies designed for primary metastatic disease.
There is currently no controversy over the volume that should be irradiated in patients with proven evidence of metastatic germinoma at diagnosis. CSF-positive disease is a risk factor for spinal seeding, but it does not predict for recurrence when treated with craniospinal radiotherapy. Long-term control in excess of 90–95% at 5 years is achieved even in the presence of widespread macroscopic metastatic disease (M 2/3) when craniospinal irradiation to dose levels of 24–30 Gy and boosts to all sites of macroscopic disease up to a dose of 40–45 Gy is used.
Germinomas are inherently chemosensitive. As primary radiotherapeutic strategies in very young children are associated with noticeable late morbidity, focus in the scientific pediatric oncology community for a long time has been to develop chemotherapeutic strategies for germinomas. This is either as a primary treatment strategy or in combination with risk-adapted reduced doses or irradiation. However, despite being chemosensitive, only a limited number of patients with localized germinoma are primarily chemocurable as the delivery of conventional chemotherapy is to a degree limited in these patients by the blood–brain barrier. Most patients require salvage treatment including radiotherapy. In addition, a chemotherapy only approach comes at the expense of significant treatment-related morbidity and mortality, not associated with radiotherapy alone. It might be anticipated that with continuing improvements in research and supportive care, as well as the availability of novel chemotherapeutic agents, some of the current limitation maybe overcome in the future.
At present, chemotherapy has been successfully used in a number of combined modality treatment approaches (chemoradiotherapy) for patients with localized disease. This aims to reduce the volume and/or the dose of radiotherapy with a reduction of late morbidity associated with radiotherapy in very young children. There is no proven benefit for using chemotherapy with respect to event free or overall survival in patients with metastatic disease at diagnosis.
The backbone of most reported multiagent chemotherapy are platinum derivatives (particularly carboplatinum), epipodophyllotoxins (e.g. etoposide), alkylating agents (e.g. cyclofosfamide, ifosfamide), and/or antibiotics (e.g. bleomycin). The commonest chemotherapy-associated side effects are short term but can be life threatening. These include hematological morbidities with the risk of bleeding and infection (particularly neutropenic sepsis), renal and hearing impairment, hemorrhagic cystitis, electrolyte disturbances (particularly in patients with known DI), and infertility to name a few. While recurrences are rare, after combined modality treatment (∼15%) they are potentially salvageable by further chemo and/or radiotherapy.
Little data is available regarding critical genetic mutations in the pathogenesis of intracranial GCTs. Speculation regarding a genetic link has arisen due to case reports of intracranial GCTs occurring in children with Down’s syndrome. Cytogenetic studies have repeatedly demonstrated abnormalities of chromosomes 1 and 12, mirroring the changes seen in extracranial GCTs. Very rarely, patients with intracranial GCTs have been reported to subsequently develop gonadal GCTs, perhaps indicating an unidentified genetic predisposition. CNS germinomas, in particular, often display sex chromosome abnormalities, usually an increased number of copies of chromosome X. Along with the increased incidence in peri-pubertal patients and its association with the site of origin being near the diencephalic nuclei, which regulates gonadotropin activity, has formed the basis for a speculated link to gonadotropins. For example, an excess of intracranial GCTs has been reported in patients with Klinefelter’s syndrome, which is associated with high levels of gonadotropins. However, this association may possibly be more likely linked to the underlying chromosomal anomalies rather than the endocrine stimuli.
Recent studies demonstrated frequent mutations of in germinomas (∼25%). KIT is a tyrosine kinase whose mutations may result in constitutive activation and has been implicated in chronic myelogenous leukemia (CML) and gastrointestinal stromal tumors (GISTs). Such a mutation may offer new therapeutic opportunities in this tumor group in the future.