|
|
||||||||
Radiation Oncology |
1 From the Departments of Oncology, Institute for Frontier Medical Sciences (Y.S.), and Therapeutic Radiology and Oncology (K.S., N.O., M.H.), Kyoto University, Sakyo-ku, Kyoto 606-8507, Japan. From the 1999 RSNA scientific assembly. Received November 30, 1999; revision requested December 30; final revision received April 3, 2000; accepted May 1. Supported in part by the Grant-in-Aid for Scientific Research (B) from the Japanese Ministry of Education, Science and Culture (10557087, 07455339, 11877152). Address correspondence to Y.S. (e-mail: yuta@frontier.kyoto-u.ac.jp).
| ABSTRACT |
|---|
|
|
|---|
MATERIALS AND METHODS: Between 1985 and 1995, 38 patients with intracranial germinoma diagnosed histologically or with established criteria were enrolled. Total radiation doses to the primary tumor site were 36 Gy after total removal, 40 Gy for tumors less than 2.5 cm in diameter, 45 Gy for those 2.54.0 cm, and 50 Gy for those greater than 4.0 cm, with 1.61.8-Gy daily fractions. Patients underwent irradiation of the primary tumor site or cerebrospinal axis (2024 Gy), depending on findings at diagnosis. No chemotherapy was allowed.
RESULTS: All patients completed radiation therapy. Thirty-five patients were treated according to protocol, and three with relatively slow tumor regression or presence of a cyst received additional radiation (57 Gy, 5052 Gy total). Ten-year overall and relapse-free survival rates were 91% and 95%, respectively. Two patients developed meningeal dissemination, but none had local failure. Treatment complications included chordoma in one patient and internal carotid artery occlusion in another. No treatment-related decline of performance status was observed in the other patients.
CONCLUSION: All tumor volumebased radiation doses were effective, without risk of local failure. Intracranial germinoma 4 cm or less in diameter can be cured with doses of 4045 Gy. Investigation of further dose reduction seems worthwhile. Radiation therapy alone with these doses should be compared with ongoing chemotherapeutic protocols plus low-dose (2430-Gy) irradiation in future studies.
Index terms: Brain neoplasms, 10.361 Brain neoplasms, therapeutic radiology, 10.1269 Dosimetry Germ cell neoplasm, 10.361 Spinal cord, therapeutic radiology, 30.1269
| INTRODUCTION |
|---|
|
|
|---|
In view of these observations, we conducted a prospective study to investigate whether the radiation dose could be reduced for small or middle-sized germinomas without jeopardizing the high cure rate. Preliminary results of this study have been reported previously, together with our retrospective series (10). The purpose of this study was to describe mature results of this study after a median follow-up of 10 years.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Pretreatment contrast materialenhanced MR imaging and/or CT brain examinations both before and after surgery (if any) were mandatory. MR imaging of the whole spine was not mandatory and was performed in only four patients because of limited machine time. Cytologic examination of the cerebrospinal fluid (CSF) was performed whenever possible, except in patients suspected of having elevated intracranial pressure. Tumor markers such as human chronic gonadotropin, carcinoembryonic antigen, and
-fetoprotein, were routinely checked after 1987. Patients with elevated carcinoembryonic antigen and/or
-fetoprotein levels were excluded, but those with a human chorionic gonadotropin level up to 1,000 IU/mL were eligible.
Accrual of 3540 patients was planned; with this patient number, a decrease in the survival and/or relapse-free rates exceeding 20% from the standard 5-year rates of 95% would be detectable, with a one-sided significance level of 0.05 and a power of 0.8 (12). Between April 1985 and July 1995, 38 patients were enrolled in the study. The patient and tumor characteristics are shown in Table 1. Pretreatment performance status of the patients was evaluated before radiation therapy after surgery (if any). Pituitary function was evaluated in patients with a suprasellar lesion.
|
Our methods for focal and craniospinal irradiation were described previously (10,11). For focal irradiation, the anteroposterior and the four lateral opposing fields were usually used unless other techniques were considered better. The treatment volume included the tumor volume determined at MR imaging and/or CT before surgery (if any) plus at least 2-cm margins and the lateral and third ventricles to 3 cm from the tumor. Most of the rest of the ventricles were usually covered with anteroposterior beams. In patients receiving craniospinal irradiation, focal boost either followed or preceded craniospinal irradiation. The whole craniospinal fields were treated on the same day. Focal irradiation after craniospinal irradiation was directed at the volume as it was presented at diagnosis.
The daily dose used was generally 1.8 Gy for focal irradiation and 1.6 Gy for craniospinal irradiation, but depending on the patients condition, modification of the daily dose between 1.5 and 1.8 Gy was allowed. The total radiation dose was determined on the basis of tumor size before radiation therapy (after surgery, if any). The total dose to the primary tumor site was 40 Gy for tumors less than 2.5 cm in maximum diameter, 45 Gy for tumors 2.54.0 cm, and 50 Gy for tumors greater than 4.0 cm. Depending on the actual response of the tumors to radiation therapy, addition or subtraction of one fraction was allowed. In patients undergoing gross total tumor removal, the total dose was 36 Gy. Total dose to the cerebrospinal axis was 20 Gy (standard dose was 20.8 Gy in 13 fractions) in patients with negative (five patients) or equivocal (five patients) CSF cytologic results or 24 Gy when CSF cytologic results were positive (11 patients) or not obtained (four patients).
Details of actual treatment are shown in Table 2. All patients completed radiation therapy. Thirteen patients underwent focal irradiation, while 25 patients underwent cerebrospinal axis irradiation. The total radiation doses were delivered as initially planned in 35 patients. In the remaining three patients, who had been scheduled to receive 45 Gy, an additional 57 Gy was administered because of the relatively slow disappearance of the tumor mass in two patients and the persistence of a cyst associated with the tumor in one patient. The total radiation dose was 36 Gy in two patients, 40 Gy (range, 3842 Gy) in 13 patients, 45 Gy (range, 4347 Gy) in 13 patients, and 50 Gy (range, 4852 Gy) in 10 patients. Although chemotherapy was not allowed, one patient received two courses of cisplatin and etoposide after 40 Gy of radiation administered by a neurosurgeon at a referring hospital (13).
|
| RESULTS |
|---|
|
|
|---|
The Figure shows overall survival and relapse-free survival curves for the 38 patients. The 5- and 10-year survival rates were 97% (standard error, 3%) and 91% (standard error, 5%), respectively. The corresponding relapse-free rates were both 95% (standard error, 4%). No patient developed local recurrence, but two developed CSF dissemination.
|
One patient developed chordoma within the irradiated volume in the frontal lobe 66 months after radiation therapy (45 Gy). The tumor was resected, and the patient survived without recurrence at 133 months. Another patient who had received 40 Gy developed occlusion of the internal carotid artery 74 months after radiation therapy and died of multiple organ failure thereafter. This patient developed encephalitis 3 years after radiation therapy and since then had a poor status. Therefore, it is unclear whether irradiation was the sole cause of the vascular event. In the other patients, treatment-related decline of performance status or severe neurocognitive deficits were not observed. No patient who had normal anterior pituitary function before radiation therapy developed clinical symptoms of pituitary dysfunction. Two of four patients who had amenorrhea regained a spontaneous menstrual cycle after radiation therapy.
| DISCUSSION |
|---|
|
|
|---|
Previous findings suggest that radiation doses lower than 50 Gy are associated with an increased local recurrence rate. Sung et al (15) found two local recurrences in six patients receiving less than 50 Gy, while there was no local recurrence in eight patients receiving 50 Gy or more. However, in their study, the incidence of CSF dissemination was high in both dose groups, and appropriate assessment of local control might have been precluded. Amendola et al (16) reported four local recurrences among nine patients receiving less than 50 Gy, but most of these patients had been treated before 1965, and the quality of radiation therapy might not have been as high as the current standard.
Dattoli and Newall (5) reported local and spinal recurrences in one patient receiving 39.6 Gy, while no recurrence was noted in 11 patients receiving 49 Gy or more. However, the reason why only this patient received the low dose is not described and remains unclear. In a retrospective analysis, Haddock et al (17) reported that patients receiving 40 Gy or less had worse outcomes, but marginal recurrences and failures with histiologic findings not consistent with germinoma were included in this group. Therefore, from these study findings, it may not be concluded that local recurrence would substantially increase with radiation doses lower than 50 Gy.
On the other hand, recent findings suggest that intracranial germinoma can generally be cured with doses below 50 Gy. Hardenbergh et al (2) found no recurrence in 14 patients receiving less than 50 Gy. Shirato et al (18) also found no local recurrence in 29 patients receiving 4048 Gy with 2.0-, 2.5-, or 3.0-Gy daily fractions, although two patients developed spinal metastases. In the study of Ono et al (19), there was no recurrence in 13 patients treated with doses less than 45 Gy (mean, 39 Gy). In more recent German studies (20), no local recurrence was detected with a dose of 45 Gy. A few other investigators (9,21,22) have reported similar findings. These results of past studies and those of the current study suggest that a dose of 50 Gy is generally unnecessary for intracranial germinomas, except large ones.
The dose for CSF prophylaxis is a separate issue. Since the radiation dose for the primary tumor seems to be reducible, the dose for the cerebrospinal axis may also be reduced. We used 20 Gy for patients with negative or equivocal CSF cytologic results and have not found CSF dissemination. However, we found CSF dissemination in a patient who definitely had positive CSF cytologic results and who was treated with 23.7 Gy in the CSF space. Because of the limited number of patients, it is unclear whether this recurrence was due to underdosage or other unknown causes (a small proportion of germinomas do recur even after higher-dose radiation therapy), and further experience is required to determine the optimal dose for CSF prophylaxis.
Since all radiation doses based on tumor volume used in this study were effective, without any risk of local failure, the possibility of further dose reduction may be a topic of further investigations. However, protocols including those for systemic chemotherapy have been established recently in many institutions. Chemotherapy alone seems to be associated with an unacceptably high recurrence rate (13,23,24), and this approach is not justified. Instead, the combination of chemotherapy and low-dose irradiation is now increasingly being investigated (2528). Findings of small studies (25,27) suggest generally favorable outcomes, but the follow-up periods are still short.
Recently, Matsutani (29) reported an interim result of a Japanese multicenter trial in which three courses of carboplatin and etoposide were administered before focal irradiation with 24 Gy. Of 71 patients with germinoma treated with this protocol, five patients developed recurrence within relatively short follow-up periods (median, 25 months). Four of the five recurrences occurred outside the radiation field; this fact would suggest that these anticancer agents are relatively ineffective for CSF prophylaxis due to their inability to penetrate the blood-brain barrier. When the relatively short follow-up periods are considered, the recurrence rate may still increase, and this treatment may prove to be inferior to standard radiation therapy with craniospinal fields in high-risk patients in the future. Even if the combined treatment proves to produce cure rates equal to those obtained with higher doses of radiation therapy, toxic effects and quality of life should be compared in future studies.
In addition, the use of radiosurgery is also being investigated to minimize unnecessary irradiation of normal brain tissue (30). Thus, the optimal treatment for intracranial germinoma in terms of the best control rate with the fewest toxic effects remains to be determined. When the relatively low incidence of this tumor is considered, it is expected that ongoing and future multicenter cooperative studies will provide us with information about the optimal treatment for this highly curable disease.
In conclusion, the radiation doses used in this study, 36 Gy after total removal, 40 Gy for tumors less than 2.5 cm in diameter, 45 Gy for 2.54.0-cm tumors, and 50 Gy for tumors greater than 4.0 cm, were effective for the local control of intracranial germinomas. These results form a basis from which to launch a randomized phase 3 trial to further investigate the best treatment for this tumor.
| FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, Y.S.; study concepts and design, Y.S.; definition of intellectual content, Y.S.; literature research, Y.S.; clinical studies, all authors; data acquisition, Y.S., K.S., N.O.; data analysis, Y.S.; statistical analysis, Y.S.; manuscript preparation and editing, Y.S.; manuscript review and final version approval, all authors.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. T. Reddy, J. C. Wellons III, J. C. Allen, J. B. Fiveash, H. Abdullatif, K. W. Braune, and P. A. Grabb Refining the staging evaluation of pineal region germinoma using neuroendoscopy and the presence of preoperative diabetes insipidus Neuro-oncol, April 1, 2004; 6(2): 127 - 133. [Abstract] [PDF] |
||||
Read all eLetters
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |