Controversies in the Management of Intracranial Germinomas

Dr. Paulino and his coauthors present a thoughtful, comprehensive survey of the literature regarding the evolving management of intracranial germinoma. They review promising areas of clinical research and acknowledge an increasing role for chemotherapy in the initial management of this tumor.

They also refute the old wisdom that employed a test-screening dose of radiotherapy to distinguish germinoma from other pineal region tumors in order to spare patients the morbidity of surgery. The firm insistence by these radiation oncologists that every patient with a suprasellar or pineal region tumor is best managed with information derived from a generous biopsy is reassuring. I wish to share some additional reflections on how to advance clinical research in this area.

Factors That Promote Therapeutic Conservatism

1. The previous morbidity and mortality associated with operating on patients with pineal region tumors were excessive and deterred the traditional application of histology-based treatment planning.

These concerns are no longer valid. The surgical routes to the pineal (transtentorial or supracerebellar) or posterior third ventricular regions (transcallosal) are safely executed in most pediatric neurosurgical centers. The use of neuronavigation systems guided by magnetic resonance imaging (MRI) and the surgical microscope permits the preoperative identification and intraoperative avoidance of critical vascular and neural anatomy. Endoscopically derived biopsies may be successfully obtained in patients with tumors that extend into the posterior third ventricle, with the added advantage that the surgeon can perform a third ventriculostomy during the same procedure, thereby avoiding the need for a systemic shunt.

2. The disease is readily curable with radiotherapy.

Because the prognosis is so favorable with high doses and volumes of radiotherapy, clinicians have been reluctant to alter the conventional approach, which includes the ventricular field with or without the craniospinal field for isolated pineal or suprasellar primary tumors and craniospinal irradiation for disseminated disease at diagnosis.

The authors of this review, in fact, concur that germinomas are best treated in this fashion. This experience was based on an older clinical experience in which patients developed recurrences outside an involved field following either an inadequate or technically suboptimal staging procedure that most likely underestimated regional and metastatic involvement and included patients with tumors other than pure germinomas.

3. Intracranial germinomas are uncommon.

Intracranial germ-cell tumors constitute less than 1% to 2% of primary central nervous system (CNS) tumors in children, and < 60% of germ-cell tumors are pure germinomas. Since the incidence of germinomas peaks in the second and third decades of life, adult neuro-oncology teams will manage 40% of patients. Not until recently have multi-institution collaborations emerged, and these have involved primarily pediatric consortia in the United States and Europe and one cooperative group (the Pediatric Oncology Group [POG]). To date, the only feasible protocol design has been a single-arm, response-oriented, phase II study.

Management Issues for Future Study

1. Will MRI technology permit earlier diagnosis of patients with smaller primary tumors of lower stage?

The endocrine manifestations of suprasellar germ-cell tumors may be subtle and not easily recognized. Even when patients present with diabetes insipidus, computed tomographic (CT) scans of the head may be normal for several years before a contrast-enhancing suprasellar mass is evident. Magnetic resonance imaging, which includes thin sections through the suprasellar region, identifies pathologic enhancement better than CT, especially in the region of the infundibulum.

Since most germinomas will eventually disseminate, either within the ventricles or subarachnoid space, the earlier the diagnosis, the lower is the incidence of dissemination. Early diagnosis alone will spare patients significant treatment-related morbidity.

2. What is the highest cerebrospinal (CSF) level of beta-human chorionic gonadotropin (beta-HCG) consistent with a favorable-risk germinoma?

Presently, we triage patients with primary CNS germ-cell tumors into two categories based on histology and serum/CSF tumor markers (alpha-fetoprotein [AFP]/beta-HCG). The high-risk group consists of patients with a nongerminoma tumor component, any significant elevation of AFP, or a “modestly elevated” CSF level of HCG. A safe HCG level in the CSF is probably below 50 mIU/mL, but neither this value nor any higher level has been systematically studied with regard to prognosis in patients with histologically verified germinoma.[1]

3. What are the late effects of high-dose and large-volume brain irradiation in long-term germinoma survivors?

The majority of germinoma patients will survive their brain tumors and remain at continued risk for the progressive cognitive effects of radiotherapy. There are very few long-term follow-up data on which to assess the quality of these patients’ survival. Most of the cognitive predictions are extrapolated from studies of other groups of patients with favorable outcomes, such as those with medulloblastomas. Such patients tend to be younger at diagnosis than germinoma patients and to receive lower doses of cerebral irradiation (36 Gy).

The risks of irradiating an adolescent and young adult population are often minimized, but it is well recognized that older patients can acquire similar cognitive deficits.[2] Since many patients with germinomas, even those with apparent pineal primary tumors, present with endocrine deficiencies, it is difficult to assess the late effects of radiation therapy alone on endocrine function.

4. What assumptions can be made regarding the similarity of seminoma and dysgerminoma to their CNS counterpart, the germinoma?

If one could prove that the molecular biology and treatment sensitivities of the systemic and CNS tumors are identical, the clinical lore accrued in the management of a much larger group of patients with systemic germ-cell tumors could be more readily transferred.[3]

We already know that germinomas resemble of seminomas with respect to their high radiation sensitivity and chemosensitivity, but we still employ much higher doses of radiotherapy for germinomas than we do for seminomas and dysgerminomas, and multimodality therapy for CNS tumors is still in the experimental stages. Can germinoma be cured with radiation doses in the range of 30 Gy? What is the smallest radiation volume that will provide a high cure rate? Can we identify on a molecular basis which germinomas can be cured with chemotherapy alone? If so, what clinical and molecular factors are most reliable in predicting responsivity to treatment?

5. What is the optimal role of chemotherapy—to facilitate the use of a reduced radiotherapy dose or to replace radiotherapy altogether?

Chemotherapy is likely to be highly effective in patients with germinoma for several reasons. Patients with disseminated seminomas can be cured with chemotherapy alone.[4] Germinomas tend to arise in CNS locations devoid of the blood-brain barrier, such as the pineal region and infundibulum, so that the pharmacokinetics of chemotherapy within the CNS should be similar to that in a systemic organ.

Pilot studies in patients with newly diagnosed or recurrent germinoma have confirmed the high response rates to both single chemotherapy agents and combination regimens. Special precautions need to be taken when cisplatin (Platinol) and cyclophosphamide (Cytoxan, Neosar) are administered to patients with diabetes insipidus, however. Although most germinoma patients can be cured with either larger doses of radiotherapy or chemotherapy alone, the morbidity or mortality will be greater. I favor a treatment regimen that combines radiotherapy and chemotherapy, but in doses that minimize acute and late effects and produce cure rates equivalent to or better than historical expectations.

In pilot studies conducted in patients with localized disease, a complete response to neoadjuvant, single-agent chemotherapy, such as cyclophosphamide or carboplatin (Paraplatin), has permitted reductions in the dose and volume of radiation therapy in over 70% of patients without compromising survival.[5,6] The ventricles were not routinely irradiated in these pilot studies, and the assumption is that chemotherapy may have eliminated micrometastases within the ventricles and subarachnoid space. It is also theoretically possible that chemotherapy may reduce systemic micrometastases in patients with ventricular-peritoneal (VP) shunts.

6. What type of consortium can best conduct prospective clinical trials?

Clearly, the rarity of germinomas is the rate-limiting factor in the pace of clinical research. In the past, the cooperative groups have been reluctant to engage in a phase III trial that would require more than 5 years of accrual and another 5 years of surveillance before a survival advantage could be confirmed. Prospective clinical trials using either multimodality or single-modality (radiotherapy or chemotherapy) approaches require either the development of a loosely organized consortium or an institution with an unusually large volume of patients, such as may exist in Japan.

A theoretically compelling phase III study that could be undertaken by a multicooperative group consortium would be a comparison of high-dose/volume radiation therapy alone vs reduced-dose/volume radiation therapy and chemotherapy. The end points in this study would be quality-of-life parameters, since the 5-year survival rate in both arms is likely to exceed 90%. This concept is not likely to be widely accepted because “quality-of-life” measures are not particularly quantitative, and the pediatric cooperative groups are just learning to use them. The most productive clinical trials pursuit is to combine the expertise and accrual potential of the Children’s Cancer Group (CCG) and POG or a similar multi-institution consortium, as has been done successfully in Europe.[7]

At present, phase II pilot studies are still the most feasible. A logical next step could be to further escalate the intensity of chemotherapy and reduce the dose and volume of radiotherapy. Outcome measures should include quality of life and cognitive function, as well as response and survival.

References:

1. Sawamura Y, Ikeda J, Shirato H, et al: Germ cell tumors of the CNS: Treatment considerations based on 111 cases and their long-term clinical outcomes. Eur J Cancer 34:104-110, 1998.

2. Armstrong C, Mollman J, Corn B, et al: Effects of radiation therapy on adult brain behavior. Neurology 43:1961-1965, 1993.

3. Bosl G, Motzer R: Testicular germ-cell cancer. N Engl J Med 337:242-253, 1997.

4. Bajorin D, Sarosdy D, Mazumdar M, et al: Randomized trial of etoposide and cisplatin vs etoposide and carboplatin in patients with good-risk germ cell tumors: A multi-institutional study. J Clin Oncol 11:598-606, 1993.

5. Allen J, Kim J, Packer R: Neoadjuvant chemotherapy for newly diagnosed germ-cell tumors of the central nervous system. J Neurosurg 67:65-70, 1987.

6. Allen J, DaRosso R, Donahue B: A phase II trial of preradiation carboplatin in newly diagnosed germinoma of the central nervous system. Cancer 74:940-944, 1994.

7. Calaminus G, Bamberg M, Baranzelli M: Intracranial germ cell tumors: A comprehensive update of the European data. Neuropediatrics 25:26-32, 1994.