Clinical Outcomes after Treatment of Germ Cell Tumors

Intracranial GCTs are a heterogeneous group of neoplasms most commonly diagnosed in the pediatric population. Germinomas are exquisitely radiosensitive with long-term survival rates in excess of 90% with radiotherapy alone. NGGCTs are associated with a poorer prognosis and are typically treated with a combination of radiation and chemotherapy. Given the young age of these patients, achieving optimal outcomes will ultimately require a careful balance of maximizing disease control while minimizing adverse treatment effects. Here we review the management of intracranial GCTs and discuss the clinical outcomes of patients who undergo treatment for these rare and fascinating tumors.

Germ cell tumors (GCTs) are neoplasms of gonadal origin and are broadly classified as germinomas, nongerminomatous germ cell tumors (NGGCTs), or mixed tumors. Germinomas arise from primordial germ cells, whereas NGGCTs (teratoma, embryonal carcinoma, endodermal sinus tumor, and choriocarcinoma) arise from cells of the differentiated embryo, undifferentiated embryo, yolk sac endoderm, and trophoblast, respectively. With the exception of mature teratomas, GCTs are uniformly malignant. Among pure tumors, germinomas are the least aggressive, followed by immature teratomas, embryonal carcinomas, and choriocarcinomas. For mixed tumors, aggressiveness is determined by the proportion of the tumor mass contributed by each cell type. GCTs can be further classified as secreting or nonsecreting. Secreting tumors are characterized by elevation of alpha-fetoprotein (AFP) and/or beta-human chorionic gonadotropin (beta-hCG) in the cerebrospinal fluid (CSF) and/or serum. Although there is considerable variability in expression of AFP and beta-hCG among GCTs, secreting tumors are generally considered more aggressive than their nonsecreting counterparts. One notable exception is a subset of germinomas that contain syncytiotrophoblasts, which secrete low levels of beta-hCG, but clinically behave like pure germinomas.

GCTs typically arise in the gonads; however, 2% to 3% of primary GCTs are extragonadal. Because of the pattern of migration of primitive germ cells during embryonic development, extragonadal GCTs characteristically form in midline structures within the retroperitoneum, anterior mediastinum, and central nervous system (CNS). In the CNS, GCTs exhibit the same midline pattern of distribution observed in other body sites, with most tumors developing in the pineal region (56%) or suprasellar compartment (28%). Less frequent sites of primary intracranial GCTs include the basal ganglia, thalamus, cerebellum, lateral ventricle, cerebellopontine angle, corpus callosum, and spinal cord. GCTs may also be bifocal in 6% to 41% of cases and are most frequently observed simultaneously in the pineal and suprasellar regions. Approximately 65% of intracranial GCTs are classified as germinomas, 18% are teratomas, 5% are embryonal carcinomas, 7% are endodermal sinus tumors, and 5% are choriocarcinomas. In practice, only teratoma and germinoma are frequently encountered as pure neoplasms and approximately 25% of all GCTs are mixed tumors.

GCTs account for 0.3% to 3.4% of all intracranial tumors in Western countries; however, they are considerably more common in Asian countries and represent 4.8% to 15.0% of pediatric brain tumors in Japan. GCTs are typically diagnosed in pediatric or adolescent patients with the peak incidence of intracranial GCTs early in the second decade of life. Males are at an increased risk of developing intracranial GCTs with a male-to-female ratio reported between 1.5:1.0 and 3.0:1.0. An increased incidence of both intracranial and extracranial GCTs has also been reported in patients with certain chromosomal disorders, including Klinefelter syndrome and Down syndrome. Although the incidence of intracranial GCTs remains relatively low, the proportion of intracranial tumors represented by GCTs has been gradually, but steadily increasing worldwide. This observation, taken with the well-documented geographic distribution of these tumors and the observed association with chromosomal abnormalities, lends support to the hypothesis that environmental and genetic factors likely play a role in the pathogenesis of this diverse group of neoplasms.

The presenting signs and symptoms of intracranial GCTs vary with tumor location. Pineal tumors frequently occlude the cerebral aqueduct and patients present with signs and symptoms of elevated intracranial pressure secondary to hydrocephalus. When the tumor compresses the midbrain, patients frequently present with oculomotor abnormalities, including pupils that react better to accommodation than light, lid retraction, convergence abnormalities, and failure of upward gaze (Parinaud syndrome). Suprasellar tumors commonly present with hypothalamic/pituitary axis dysfunction, including diabetes insipidis, delayed or precocious puberty, adrenal insufficiency, or symptoms of growth hormone deficiency.

Although intracranial GCTs can be readily detected with computed tomography (CT), magnetic resonance imaging (MRI) remains the imaging modality of choice for detecting, characterizing, and staging these tumors ( Fig. 1 ). GCTs are generally isointense or low signal intensity on T1-weighted images and isointense or high signal intensity on T2-weighted images. It has been suggested that certain GCT subtypes may have characteristic MRI signals ; however, MRI alone has not been shown to be adequately sensitive or specific to obviate the need for histologic diagnosis. If a GCT is suspected, MRI of the spine is a critical step in staging, as spinal cord involvement is present in up to 10% of patients at the time of diagnosis. CSF cytology should also be obtained for staging, as patients with positive CSF cytology are considered to have metastatic disease even in the absence of MRI findings. Hematogenous spread to extraneural sites at the time of diagnosis has also been described, but this is an exceedingly rare occurrence.

Measurement of CSF and serum beta-hCG and AFP may also aid in diagnosis and some investigators have suggested that significant elevation of these markers in the setting of imaging findings suggestive of aggressive disease may justify forgoing biopsy in favor of prompt treatment with radiotherapy and chemotherapy. Similarly, imaging findings suggestive of pure germinoma in the absence of positive tumor markers may justify forgoing histologic diagnosis to begin radiotherapy immediately. This approach, however, remains controversial. Unless tumor markers and radiology unequivocally indicate a specific diagnosis, the consensus is that tissue should be obtained for histologic evaluation in all patients who are able to tolerate the procedure. The preferred method for obtaining tissue is also controversial. Some investigators have suggested that open biopsy may facilitate more accurate diagnosis. Others maintain that stereotactic biopsy interpreted in the context of serum and CSF tumor markers is preferred to avoid the inherent risks of surgery in the pineal and suprasellar regions. With either approach, discordance between histology and tumor markers likely represents a mixed tumor and should be treated as the more aggressive subtype.

Treatment approaches and clinical outcomes

Identifying the optimal treatment approach for primary intracranial GCTs remains a topic of active research in the field of neuro-oncology. Because of the diversity of this rare group of neoplasms, there is a paucity of well-controlled prospective trials. Available data from retrospective studies is often difficult to interpret, as these studies tend to be relatively small and heterogeneous. An algorithm for the diagnosis and treatment of GCTs is depicted in Fig. 2 . Here we review the management of intracranial GCTs and discuss the clinical outcomes of patients who undergo treatment for these rare and fascinating tumors.

Germinomas

Germinomas are extraordinarily radiosensitive, and 5-year to 10-year survival rates in excess of 90% have been consistently reported with radiotherapy alone. Given the high success rate of established treatment regimens at controlling local and distant disease progression, the focus has more recently shifted to minimizing late radiation effects in this generally young patient population. Historically, the standard of care has been 45 to 55 Gy administered to the primary tumor site with 30 to 36 Gy to the entire craniospinal axis to reduce the risk of leptomeningeal spread. However, with 20-year survival rates in excess of 80% reported in some series, even relatively low doses of radiation raise concern for long-term sequelae. Neurocognitive dysfunction has been well-documented in patients receiving as little as 20 to 25 Gy for treatment of brain tumors. Adverse physical effects of radiotherapy have also been reported, including hypothalamic-pituitary dysfunction, radiation-induced occlusive vasculopathy of large intracranial arteries, an increased estimated stroke rate (11.7%, 16 years after treatment), development of arteriovenous malformations, and a high estimated rate of secondary neoplasms (16.8%, 19 years after treatment). Broad quality-of-life measures, such as marriage, graduation from high school, and employment, are also negatively impacted by exposure to radiotherapy at a young age. In addition, a study examining the self-reported physical and psychological well-being of patients treated for intracranial germinomas reported a direct correlation between poor functional status and early age of diagnosis. These findings suggest that young patients may be especially vulnerable to the late effects of radiotherapy.

A variety of protocols have been designed to decrease radiation exposure in patients being treated for intracranial germinomas. Common strategies include chemotherapy, stereotactic radiosurgery, dose reduction, field reduction, volume-based dose selection, or a combination of these approaches. Although some investigators have reported improved control of intracranial and spinal disease with radiation doses in excess of 40 Gy and upfront craniospinal radiation, this effect has not been consistently demonstrated. Therefore, the current standard of care at most institutions is 21 to 24 Gy to the whole ventricle and an additional boost of 40 to 45 Gy to the primary tumor with craniospinal irradiation reserved for patients with evidence of disseminated disease based on imaging or CSF histology. Chemotherapy may be added to this regimen for patients with extensive or recurrent disease. Regardless of the initial treatment regimen, patients who achieve complete tumor remission rarely experience recurrence. When tumors do recur, they may be highly aggressive ; however, salvage radiotherapy with or without chemotherapy induces lasting remission in the vast majority of cases.

A limited number of trials have evaluated the effectiveness of stereotactic radiosurgery in the treatment of intracranial germinomas. A case report by Regine and colleagues, which describes a patient who underwent radiosurgery and refused further treatment, reported no disease recurrence at 6 months. This case highlights the potential of radiosurgery for controlling the primary lesion; however, a combined approach with 10 to 12 Gy delivered to the primary lesion with stereotactic radiosurgery plus whole-ventricular irradiation (24 Gy) is preferred to reduce the risk of leptomeningeal spread as well as interventricular recurrence. This protocol has been reported to provide control rates comparable to traditional radiotherapy regimens. It has also been suggested that stereotactic radiosurgery may reduce length of hospital stay and overall treatment costs. Additional studies are warranted to evaluate the effectiveness and economic impact of stereotactic radiosurgery in these patients.

Trials of chemotherapy alone have consistently reported lower rates of remission as compared with radiotherapy or a combination of radiotherapy and chemotherapy. Chemotherapy-only regimens have been demonstrated to induce remission at rates in excess of 80% ; however, these effects are transient and recurrence rates in the range of 50% have been consistently reported. A recent international study reported that 7 of 11 patients treated with 4 to 6 cycles of carboplatin/etoposide alternating with cyclophosphamide/etoposide relapsed between 13 months and 5 years after diagnosis. This study represented the third of 3 consecutive international cohort studies evaluating a chemotherapy-only approach for intracranial GCTs. Based on the cumulative results of these trials, the investigators concluded that chemotherapy alone offered unacceptably low rates of durable tumor control as compared with either radiotherapy alone or in combination with chemotherapy.

Although chemotherapy alone offers inferior tumor control rates compared with radiotherapy, several investigators have reported that neoadjuvant chemotherapy may reduce the dose of radiation required to induce lasting remission. Available data indicate that the addition of chemotherapy allows for a slight dose reduction in whole-ventricle irradiation (21.6–25.5 Gy) and a larger reduction in primary site boost (30.0–30.6 Gy) without compromising tumor control rates. Severe chemotherapy-related toxicities reported in these studies have primarily been related to myelosuppression. One study of 17 patients treated with etoposide and cisplatin reported neutropenia requiring administration of granulocyte-colony stimulating factor in 5 patients with leukocyte nadirs below 1000/μL in 3 patients. Thrombocytopenia requiring platelet transfusions was reported in 3 patients and 1 patient dropped below 20,000/μL. Nonhematologic toxicities included mild ototoxicity, vomiting, renal insufficiency, and neuropathy. Other studies reported similar rates of adverse events with reported toxicities specific to the chosen chemotherapeutic regimen. Although available data indicate that the radiation doses used in these trials may reduce the rate of early neurocognitive sequelae, to date there is inadequate follow-up to draw meaningful conclusions regarding the late neurocognitive effects of these regimens. Given the established efficacy of whole-ventricle radiotherapy and local radiation boost with the addition of craniospinal radiation for extensive disease, combination approaches of chemotherapy and reduced radiation doses should likely remain restricted to experimental protocols at this time. In the setting of recurrent or extensive disease, however, the addition of chemotherapy to a standard radiotherapy regimen has been shown to improve remission rates. Chemotherapy may also be indicated in patients with a ventriculoperitoneal shunt to decrease the risk of extraneural metastases.

Given the radiosensitivity of pure germinomas and the risk of surgery in the pineal and suprasellar regions, the role of surgical resection in the treatment of this disease has historically been limited. Advances in microsurgery and endoscopic neurosurgery, however, have led to a reexamination of the role of surgery in the management of these tumors. Although stereotactic biopsy can be prone to diagnostic inaccuracy, endoscopic biopsy has been reported to offer diagnostic yields as high as 90% to 98%. Endoscopic surgery may also afford more accurate assessment of ventricular tumor extension than MRI alone. Given the documented cases of extraneural spread from VP shunts, there has been some concern regarding iatrogenic seeding of endoscopic tracts. Although additional studies are needed to evaluate this potential risk more thoroughly, early reports indicate that rates of tumor dissemination are not increased with endoscopic surgery.

Although extensive resection of germinomas is not currently standard of care, some investigators suggest that the increased diagnostic accuracy afforded by generous tumor sampling should be reevaluated in light of the reduced operative risks associated with modern endoscopic and microsurgical techniques. Furthermore, studies of chemotherapy-only therapy for intracranial germinomas have demonstrated a reduction in recurrence for smaller tumors as compared with larger tumors. Although the rate of recurrence is still higher than that observed in radiation protocols, these findings suggest that there may be a future role for cytoreductive surgery with or without chemotherapy, as the field continues to move toward minimizing radiation exposure in these patients.

Nongerminomatous Germ Cell Tumors

NGGCTs are a heterogeneous group of neoplasms that are less common than germinomas and, as a result, available data are sparse and somewhat difficult to interpret. Unlike germinomas, NGGCTs are relatively insensitive to radiotherapy and carry a graver prognosis. Patients treated with a regimen of whole-ventricle irradiation and boost to the lesion have 5-year survival rates in the range of 35% to 60%. Nevertheless, radiotherapy remains a critical component of treatment for NGGCTs, as chemotherapy-only regimens have been reported to offer only slightly improved 2-year survival over radiotherapy alone and comparable 5-year survival rates. Combination regimens of chemotherapy plus radiotherapy offer improved overall survival with 5-year survival rates consistently reported in the range of 50% to 70% and higher rates reported in some smaller series. Importantly, in a series of 153 histologically confirmed intracranial germ cell tumors (90 NGGCTs), Matsutani and colleagues reported that survival varied significantly based on the histologic subtype of NGGCT. Patients with pure endodermal sinus tumors, choriocarcinomas, or embryonal carcinomas had a 3-year overall survival rate of 27.3%, whereas patients with teratomas or mixed germinomas had a 10-year survival rate of 70.7%. One-year survival rates also varied widely with 0% survival for choriocarcinomas, 33% for endodermal sinus tumors, and 80% for embryonal carcinomas. This finding further highlights the heterogeneity of these tumors and suggests that additional studies evaluating treatment of specific tumor subtypes may improve results.

Chemotherapeutic regimens in NGGCT treatment protocols vary, but frequently include cisplatin or carboplatin and etoposide. Radiation doses are typically in the range of 30 to 36 Gy delivered to the entire craniospinal axis with a tumor boost of 20 to 54 Gy. The rate of metastasis to the spinal cord or distant sites has been reported to be as high as 45% and, unfortunately, the long-term effects of radiotherapy in these patients are of less concern given their poor overall prognosis. As a result, craniospinal irradiation is currently standard-of-care for all patients with intracranial NGGCTs. However, some investigators have suggested that spinal irradiation may not improve outcomes in patients with certain nonmetastatic NGGCTs. More research is needed to determine the role of craniospinal irradiation in the treatment of localized NGGCTs.

Although there are no definitive data to suggest that upfront gross total resection of NGGCTs improves outcomes, the frequency of mixed tumors in this group can lead to an especially high rate of diagnostic error with stereotactic biopsy. This risk, coupled with the importance of accurate subtype identification for treatment and prognosis has led some investigators to suggest that the diagnostic accuracy afforded by maximal resection at the time of diagnosis is worth the inherent risks of surgery. Once a diagnosis has been established, resection has been shown to play an important role in the treatment of some NGGCT subtypes. Surgical resection is the treatment of choice for mature teratomas, and also plays an important role in the treatment of mixed tumors when a component of mature teratoma remains after chemotherapy and radiotherapy have been completed.

In addition, surgery is the only potentially curative treatment for a distinct phenomenon first described by Logothetis and colleagues in 1982 known as growing teratoma syndrome. This syndrome is characterized by rapid, paradoxic growth of a GCT—often in the setting of normalizing tumor markers—in spite of treatment with chemotherapy and radiotherapy. Growing teratoma syndrome is more common in tumors outside the CNS, but has also been reported in intracranial lesions. A recent study estimated that this phenomenon was observed in up to 21% of intracranial NGGCTs. A growing teratoma is potentially curable with surgical resection and second-look surgery should be considered for any NGGCT that continues to grow in spite of adequate treatment with chemotherapy and radiotherapy.

Surgery plays an important palliative role in the care of patients with intracranial NGGCTs who develop hydrocephalus as a result of tumor expansion. Third ventriculostomy, which involves endoscopically perforating the floor of the third ventricle to allow CSF to flow directly into the basal cisterns, is the treatment of choice for hydrocephalus in these patients unless tumor occupies the floor of the third ventricle. Ventriculoperitoneal shunting may also be used, but has been associated with a higher complication rate and an increased risk of extraneural metastasis. Third ventriculostomy can easily be accomplished at the time of endoscopic tumor sampling for patients who present with obstructive hydrocephalus.

The prognosis for patients with recurrent NGGCTs is extremely poor. Only 1 in 4 patients achieves full remission when relapsing after a complete response to initial therapy and very few patients who have a partial response to initial treatment achieve remission from recurrent disease. Myeloablative chemotherapy with autologous hematopoietic stem cell rescue has been proposed as a potential salvage therapy for these patients. Early reports indicate that 5-year overall survival in the range of 50% may be possible using this approach. The optimal chemotherapeutic regimen as well as the role of radiotherapy and surgery for recurrent NGGCTs, however, remains unclear. Future studies will also need to assess the effectiveness of myeloablative chemotherapy for specific subtypes of recurrent NGGCTs.

Only gold members can continue reading. Log In or Register to continue