Lack of benefit of spinal irradiation in the primary treatment of intracranial germinoma

A multiinstitutional, retrospective review of 180 patients

Abstract

BACKGROUND

The current study assessed the contribution of spinal irradiation to the treatment outcome of patients with intracranial germinoma.

METHODS

Clinical data from 180 patients with intracranial germinoma, who were treated with radiotherapy and/or chemotherapy from 1980 to 2001, were collected from 6 institutions. The patients' median age was 16 years (range, 1–47 yrs), and the male-to-female ratio was 133:47. Pathologic verification was obtained in 88 patients. A solitary tumor was seen in 129 patients, and multifocal or disseminated tumors were detected in 51 patients. The median tumor size was 2.5 cm (range, 0.6–7.0 cm). Local field and/or whole brain irradiation was performed in 114 patients, and craniospinal irradiation was performed in 66 patients. Fifty-five patients were treated with chemotherapy. The median follow-up time was 89 months (range, 3–297 mos).

RESULTS

Eight-year overall and event-free survival rates were 91% and 89%, respectively. The 8-year recurrence rates at the primary site, intracranial space, and the spinal space were 1%, 6%, and 6%, respectively. Cox regression analysis showed that spinal irradiation (hazard ratio, 1.050; 95% confidence interval [CI], 0.355–3.170) did not contribute to a favorable event-free survival.

CONCLUSIONS

Spinal irradiation did not contribute to favorable event-free survival in patients with intracranial germinoma. Cancer 2005. © 2005 American Cancer Society.

Primary intracranial germinoma is relatively rare, composing 2.1% of all primary intracranial neoplasms and 9.8% of brain tumors in children in Japan.1 It usually arises in the midline of the brain, such as the suprasellar and pineal regions, and the peak incidence is in the early pubertal years.2–8 On computed tomography (CT) and magnetic resonance imaging (MRI), primary intracranial germinoma is detected as a small solitary tumor arising in the midline of the brain. However, some patients show disseminated disease at their initial evaluation, and some develop disseminated disease along the craniospinal space after treatment.5–9 Primary intracranial germinoma tends to disseminate into the craniospinal space despite the tumor size and location, and thus prevention of dissemination is key to successful management.3, 10–12 These tumors are very sensitive to radiation and chemotherapy, and both radiotherapy and/or chemotherapy have been used as the initial treatment.13–17 Craniospinal irradiation has been used in clinical practice to avoid dissemination into the craniospinal space. However, there is still controversy regarding the appropriate irradiation volume.5, 7, 8, 13, 18, 19 A large treatment volume with high-dose radiation leads to late adverse effects, such as growth deficits, mental retardation, hormonal abnormalities, and secondary malignancies.5, 7, 8, 13, 14, 16, 17 Thus, a less toxic treatment regimen should be established for children and adolescents, especially young children in the early pubertal years and younger.20, 21

We reviewed the clinical records of patients with intracranial germinomas who were treated with radiotherapy and/or chemotherapy at six Japanese institutions to assess the contribution of spinal irradiation to treatment outcome.

MATERIALS AND METHODS

Patient Characteristics

We collected the clinical records of 180 patients with intracranial germinomas who were treated with radiotherapy and/or chemotherapy from 1980 to 2001 at 6 Japanese institutions. All patients were evaluated by cranial CT at initial treatment, and MRI was used from the late 1980s. The patients' median age at the time of diagnosis was 16 years (range, 1–47 yrs), and the male-to-female ratio was 133:47. The performance status (PS) score according to the Eastern Cooperative Oncology Group (ECOG) was 0 for 106 patients, 1 for 30 patients, 2 for 12 patients, 3 for 7 patients, and 4 for 9 patients. PS data were not obtained in 16 patients. Pathologic verification was obtained in 88 patients, and diagnosis was based on clinical findings in 92 patients. Clinical diagnosis was made using the following parameters: young age, tumor location, and rapid response to radiotherapy showing > 80% reduction in average diameter at a dose of approximately 20 Gy.3

The primary site was the pineal region in 60 patients, suprasellar region in 53 patients, and basal ganglia in 16 patients, whereas 51 patients had multifocal or disseminated tumors. We grouped multiple tumors, multiple midline tumors, and disseminated tumors together as multifocal or disseminated tumors because it was not possible to define these tumors clearly. Among the 51 patients with multifocal or disseminated tumors, 25 had 2 tumors in the suprasellar region and pineal region, 14 had 2 tumors in the pineal region and other sites, 7 had > 2 tumors in the suprasellar and pineal regions and other sites, 4 had 2 tumors in the suprasellar region and other sites, and 1 patient had 2 tumors in the basal ganglia and the lateral ventricle. The median of the maximum tumor diameter was 2.5 cm (range, 0.6–7.0 cm). All patients were evaluated for spinal space dissemination by cytologic examination of cerebrospinal fluid (CSF), and spinal MRI was performed in 125 patients. CSF cytology was examined once for each patient. Cytologic examination of CSF demonstrated the incidence of tumors in two patients, and spinal MRI showed spinal dissemination in three patients. Serum tumor markers, such as human chorionic gonadotropin β-subunit (β-HCG), was evaluated before treatment in 73 patients, and slight elevation of β-HCG (1.1–120 mIU/mL) was found in 15 patients (20%). None of the 33 patients who were evaluated at the initial treatment showed elevated α-fetoprotein (AFP) or carcinoembryonic antigen (CEA) titer.

Table 1 shows clinical characteristics of the patients treated with and without spinal irradiation. A significant difference was observed only with regard to the number of patients who received chemotherapy (P = 0.002), and there were no differences between the two groups with regard to other clinical parameters, such as age, gender, performance status, hormonal abnormality, location and number of tumors, and tumor size.

Table 1. Clinical Characteristics of Patients with Intracranial Germinoma Treated with Spinal-Axis Irradiation and Those without Spinal-Axis Irradiation

Characteristics	Spinal irradiation
	Without (n = 114)	With (n = 66)
	No. of patients (%)	No. of patients (%)
a ECOG: Eastern Cooperative Oncology Group.
Age, yrs
Median	16	16
Range	1–47	6–36
≥ 16 yrs	61 (54)	38 (58)
< 16 yrs	53 (46)	28 (42)
Gender
Male	81 (71)	52 (79)
Female	33 (29)	14 (21)
Performance status (ECOGa)
0–1	84 (74)	52 (79)
2–4	19 (16)	9 (14)
Unknown	11 (10)	5 (7)
Hormonal abnormality
Normal	54 (47)	37 (56)
Deficit	59 (52)	29 (44)
Unknown	1 (1)	0 (0)
Location and no. of tumors
Solitary tumor
Pineal lesion	38 (33)	22 (33)
Suprasellar	36 (32)	17 (26)
Basal ganglia	11 (10)	5 (8)
Multiple or disseminated tumors	29 (25)	22 (33)
Diagnosis
Pathologic	59 (52)	29 (44)
Clinical	55 (48)	37 (56)
Tumor bulk
> 2 cm	23 (20)	15 (22)
≥ 2–> 4 cm	60 (53)	36 (55)
≥ 4 cm–> 6 cm	20 (17)	12 (18)
≥ 6 cm	1 (1)	0 (0)
Unknown	10 (9)	3 (5)
Chemotherapy
None	70 (61)	55 (83)
Administered	44 (39)	11 (17)

Radiotherapy

Super-voltage X-rays (4–10 MV) or gamma-ray from a ⁶⁰Co unit were used. The fraction size for the primary site was 1.8–2.0 Gy, and that for the whole craniospinal region was 1.6–2.0 Gy. Patients received 5 fractions per week. The treatment volume of radiotherapy varied according to the policy of each radiation oncologist. The local irradiation field included the primary tumor with a margin of 1–2 cm. Local irradiation of the primary site was performed using the two-opposed lateral fields, four-field technique (anteroposterior–posteroanterior plus opposed lateral fields), or arc or rotational techniques. Whole-brain irradiation was performed using two-opposed lateral fields, and spinal irradiation using a posterior–anterior field. The reference point of the prescribed radiation dose was the midplane of the opposed fields and the intersection of the field center (isocenter) for multiple fields. The craniospinal field included the entire cranial contents and entire spinal space. In the craniospinal field, the cranial contents were treated with the opposed lateral fields with a margin of about 1 cm between the skull base and the shaped inferior margin of the field. The spinal space field was treated with one or more posterior fields. The inferior border of the spinal space field was at the inferior border of the S3 vertebra, and the lateral borders were located 1–1.5 cm lateral to the edge of the vertebral bodies.

Fourteen patients were treated with local irradiation field only, 100 patients with local field plus whole-brain or whole-ventricle irradiation, and 66 patients received craniospinal irradiation. The median and mean radiation doses for the primary site were 50 Gy and 46 Gy (range, 24–64 Gy), and those for the whole brain or whole ventricles were 30 Gy and 30 Gy (range, 12–54 Gy), respectively. The median and mean radiation doses for the whole craniospinal region were 30 Gy and 27 Gy (range, 7.2–37.5 Gy), respectively. Among 14 patients treated with local irradiation field alone, 9 received chemotherapy, and 5 did not. Among the 100 patients treated with local field plus whole-brain or whole-ventricle irradiation, 35 received chemotherapy, and 65 did not. Among the 66 patients treated with craniospinal irradiation, 11 received chemotherapy, and 55 did not.

Chemotherapy

Chemotherapy was given according to the policy of each physician. Fifty-five patients were treated with neoadjuvant, concurrent, or adjuvant chemotherapy. Three patients were treated concurrently with chemotherapy and radiotherapy, and the other 52 patients were administered these therapies in sequential order. Among the 51 patients with multiple or disseminated tumors, 23 received chemotherapy. Thirty-two of the 129 patients with solitary tumors received chemotherapy. Multiagent chemotherapy included mainly platinum-based chemotherapy.22 Forty patients were treated with 1 to 5 cycles (mainly, 2 or 3 cycles) of etoposide plus cisplatin or carboplatin. The EP regimen included both etoposide (100 mg/m²) and cisplatin (20 mg/m²) for 5 consecutive days every 4 weeks. Another 15 patients were treated with cisplatin, bleomycin, nitrosourea, and other agents. Three patients received intrathecal methotrexate.

Statistical Analysis

The overall survival was defined as death due to any cause, and the event-free survival was defined as the recurrence of diseases or death due to any cause. Local recurrence was defined as the initial or secondary recurrence at the primary site or surrounding area (1 cm margin); intracranial recurrence was defined as the local recurrence and the initial or secondary recurrence in the intracranial space above the level of the second cervical vertebra (C2); and spinal recurrence was defined as the initial or secondary recurrence in the spinal canal below C2. The survival rate was calculated using the Kaplan–Meier Method.23 Cox regression analysis was used to adjust the observed treatment effect for the influence of various prognostic factors, including gender, tumor size, number of tumors, spinal irradiation, and chemotherapy, on event-free survival. The median follow-up time of the 180 patients was 89 months (range, 3–297 mos). Statistical analysis was performed using SAS version 8.2 (SAS Institute, Cary, NC).

RESULTS

Survival Rate

The 8-year overall and event-free survival rates of the 180 patients were 91% and 89%, respectively. The 8-year recurrence rates at the primary site and intracranial space were 1% and 6%, respectively. The 8-year recurrence rate in the spinal space was only 6%. Figure 1 shows event-free survival curves of the 180 patients, and there were no differences between the two curves. Figure 2 shows event-free survival curves of the patients with solitary tumors and of those with multifocal or disseminated tumors. Of the 17 patients with recurrent disease, 10 (58%) and 14 (82%) developed recurrent disease within the first 2 years or within the first 5 years after the initial treatment, respectively. Only 3 patients developed recurrent disease more than 5 years after the initial treatment.

Figure 1. Event-free survival curves of the 180 patients. SXRT = 0, patients who did not receive spinal irradiation; SXRT = 1, patients who did receive spinal irradiation.

Figure 2. Event-free survival curves of patients with solitary tumors and of patients with multifocal or disseminated tumors.

For all 180 patients, Cox regression analysis of clinical parameters showed that spinal irradiation did not contribute to a favorable event-free survival (Table 2). The hazard ratio of spinal irradiation was 1.050 (95% confidence interval [CI]: 0.355–3.170). Female gender, tumor size of ≥ 4 cm, and multifocal or disseminated tumors were not associated with an unfavorable event-free survival.

Table 2. Estimated Hazard Ratios from Cox Regression Analysis of Disease-Free Survival

Characteristic	No. of patients	Multivariate analyses			P
		HRa	95% CIb	χ^2
χ^2: chi-square values. a HR: hazard ratio. b 95% CI: 95% confidence interval.
Gender
Male	133	1.000
Female	47	2.859	0.982–8.321	3.7150	0.0539
Tumor bulk
> 4cm	134	1.000
≤ 4 cm	33	2.585	0.841–7.941	2.7495	0.0973
No. of tumors
Solitary	129	1.000
Multifocal or disseminated tumors	51	2.351	0.742–7.450	2.1084	0.1465
Spinal irradiation
None	114	1.000
Administered	66	1.134	0.365–3.525	0.0476	0.8274
Chemotherapy
None	125	1.000
Administered	55	1.518	0.487–4.736	0.5173	0.4720

Recurrence and Salvage

Among 17 patients who developed recurrence, pathologic verification was obtained in 3 patients; all 3 patients had recurrent germinoma. Recurrence at the primary site occurred in 1 patient and, in the intracranial space, in 10 patients, excluding 1 patient with local recurrence. In the spinal space, recurrence occurred in 9 patients as the initial or secondary recurrence (Table 3). Two patients developed recurrence beyond the central nervous system (peritoneum and bone). Of the 10 patients who developed recurrent disease within the first 2 years after the initial treatment, only 3 patients (Cases 1, 12, 13) survived after recurrence (4 yrs 8 mos–17 yrs 10 mos) (Table 3). Of the 7 patients who had recurrent disease beyond 2 years, 6 of them (Cases 2, 3, 6, 10, 15, 16) survived with or without disease after recurrence (2 yrs 6 mos–12 yrs).

Table 3. Clinical Data on Seventeen Patients with Recurrent Disease after Treatment

Case #	Age (yrs)	Gender	Initial location	Pathologic confirmation	Maximum diameter of tumor (cm)	Serum tumor marker		CSF cytology	RT dose (Gy)			CTX	Site of first failure (mos)	Salvage therapy	Outcome
						bHCG	AFP		Primary site	Whole brain	Whole spine
bHCG: human chorionic gonadotropin beta-subunit; AFP: alpha-fetoprotein; CSF: cerebrospinal fluid; RT: radiotherapy; CTX: chemotherapy. a She received whole ventricle irradiation, and marginal recurrence around the radiation field developed.
1	10	Female	Pineal, suprasellar	No	4	79 mIU/mL	> 10 ng/mL	Negative	40	-	-	Yes	Spine (5)	RT + CTX	Alive at 112 mos
2	15	Male	Pineal, ventricle	Yes	3	> 1.0	> 10	Negative	40	-	-	Yes	Ventricle (29)	RT + CTX	Alive at 59 mos
3	11	Female	Suprasellar	Yes	2.9	> 1.0	?	Negative	52	-	-	No	Cerebrum (49)	RT + CTX	Alive at 206 mos
4	3	Female	Thalamus	No	3	> 1.0	> 10	Negative	54	-	-	No	Suprasella (9)	None	Died at 14 mos
5	14	Male	Suprasellar, pons	No	?	?	?	Negative	48.5	12	-	Yes	Spine (17)	RT + CTX	Died at 68 mos
6	12	Female	Suprasellar	Yes	?	24	?	Negative	40	20	-	Yes	Cerebrum, spine (73)	RT + CTX	Alive at 154 mos
7	18	Female	Pineal	No	1.5	> 1.0	> 10	Negative	50	30	-	No	Abdomen (4), cerebrum, spine (71)	RT + CTX	Died at 121 mos
8	18	Male	Hypothalamus	Yes	5.8	> 1.0	> 10	Negative	50	30	-	No	Ventricle (44)	CTX	Died at 53 mos
9	16	Male	Pineal	Yes	5	> 1.0	> 10	Negative	50	30	-	No	Spine (10)	Unknown	Died at 52 mos
10	20	Male	Pineal, suprasellar, ventricle	Yes	3	> 1.0	> 10	Negative	50	30	-	Yes	Spine (62)	RT + CTX	Alive at 110 mos
11	24	Female	Suprasellar	Yes	4.2	?	?	Negative	40	40a	-	No	Cerebrum (3)	RT + CTX	Died at 22 mos
12	13	Male	Hypothalamus	No	4	?	?	Negative	59.5	39.5	13.5	Yes	Spine (5)	RT + CTX	Alive at 62 mos
13	14	Female	Suprasellar	Yes	3.4	?	> 10	Negative	44	30	22	Yes	Cerebrum (6)	RT	Alive at 220 mos
14	9	Male	Pineal, ventricle, chiasma	Yes	3	> 1.0	> 10	Negative	30.4	30.4	22.8	Yes	Primary site (7)	None	Died at 8 mos
15	20	Male	Pineal, suprasellar	No	2.5	> 1.0	> 10	Negative	48.7	28.7	28.7	No	Bone (38)	CTX	Alive at 123 mos
16	17	Female	Suprasellar, ventricle	Yes	3	> 1.0	> 10	Negative	40	20	30	No	Ant. Horn, spine (74)	RT + CTX	Alive at 125 mos
17	16	Male	Suprasellar	Yes	1	> 1.0	> 10	Negative	55.8	30.6	30.4	No	Cerebrum, spine (18)	RT + CTX	Died at 69 mos

Recurrent disease in the intracranial space was the initial recurrence in 8 of the 11 patients who developed recurrence in this area. Of the eight patients with initial intracranial recurrence, four received irradiation and chemotherapy after recurrence, one received irradiation alone, and three received no salvage therapy. Of the eight patients, three survived (2 yrs 6 mos–17 yrs 10 mos). Nine patients developed recurrence in the spinal space, and in all of these patients, this was the initial recurrence. Of these nine patients, six received irradiation and chemotherapy as salvage therapy; details regarding salvage therapy in the other three patients were not known. Of the nine patients with spinal recurrence, six survived (4 yrs–8 yrs 10 mos).

Adverse Effects

Lethal late adverse effects, convulsions, hormonal abnormalities, hemorrhage, and secondary neoplasm were found in four patients. One patient, a 13-year-old boy, suffered from uncontrolled convulsions after treatment. He received craniospinal irradiation (primary site, 54.4 Gy; whole brain, 30.4 Gy; whole spine, 30.4 Gy) without chemotherapy. He died because of repeated uncontrolled convulsions 4 years 3 months after initial treatment. One patient, an 18-year-old man, developed lethal hormonal deficit. He had a tumor in the suprasellar region and received craniospinal irradiation (primary site, 49.4 Gy; whole brain, 23.4 Gy; whole spine, 30.6 Gy) without chemotherapy. Management of the hormonal deficit was poor, and he died of hypernatremia without evidence of disease 5 years 5 months after initial treatment. One patient, a 17-year-old boy, developed a cerebral hemorrhage after treatment. He received craniospinal irradiation (primary site, 55.2 Gy; whole brain, 35.2 Gy; whole spine, 35.2 Gy) without chemotherapy. He developed cerebral hemorrhage 9 years after initial treatment. One patient, an 18-year-old man, developed a second malignancy in the intracranial space. He received cranial irradiation (primary site, 50 Gy; whole brain, 30 Gy) without chemotherapy. Astrocytoma (Grade 2) occurred in the cerebrum 5 years after initial treatment.

DISCUSSION

The treatment volume of radiotherapy for intracranial germinoma is still controversial. In many institutions craniospinal irradiation has been used for clinical treatment of this disease.3, 7, 8, 18 The outcome after craniospinal irradiation is excellent, with 10-year overall survival rates ranging from 87% to 91%.3, 7, 8, 18 The German Cooperative Prospective Trials MAKEI 83/86/89 were conducted to assess the outcome of patients with intracranial germinoma after craniospinal irradiation alone.10 The 5-year relapse-free survival rate of all 60 patients was 91%, and only 1 patient developed spinal recurrence. Haas-Kogan evaluated 35 patients with localized germinoma who did not receive spinal irradiation and reported that no isolated spinal recurrence occurred.15 Wolden evaluated 48 patients with primary intracranial germ cell tumors including 24 patients with germinomas.13 Only four of these patients received spinal irradiation. The rate of solitary spinal recurrence was only 2%, and spinal irradiation was not necessary after complete diagnostic craniospinal evaluation. Linstadt reported that craniospinal irradiation was indicated in a patient with high-risk factors, including gross tumor spill causing contamination of CSF, malignant CSF cytology, and subependymal or subarachnoid dissemination.14 Hardenbergh reported that craniospinal irradiation might be indicated for patients with multifocal midline germinoma or patients with evidence of spinal space seeding.7 Our study was limited by a lack of pathologic confirmation, no evaluation of serum or CSF tumor markers, and no evaluation of spinal MRI. However, the results of the current study indicated that spinal irradiation was not associated with a favorable treatment outcome. The current study was a nonrandomized retrospective study, and we could not exclude the possibilities of bias and limitations to the results. However, spinal irradiation may be not necessary for patients with localized germinoma after complete diagnostic craniospinal evaluation.

Whole-brain or whole-ventricle irradiation may eradicate the subclinical disease surrounding the primary tumor and prevent marginal recurrence as well as recurrence in the ventricles.18, 24 Haddock reported that the incidence of intracranial recurrence in patients treated with local irradiation was 36%, and incidence of intracranial recurrence in patients treated with whole-brain irradiation or craniospinal irradiation were less than 10%.8 Aoyama evaluated patients with pathologically confirmed germinoma and reported a 10-year event-free survival rate of only 22% in patients treated with limited local irradiation.18 The event-free survival rates after whole-brain irradiation or craniospinal irradiation ranged from 91% to 100%.3, 5, 7, 8, 13, 14 In cases to be treated with definitive radiotherapy alone, whole-brain or whole-ventricle irradiation should be used for patients with solitary localized tumors.

Local recurrence after definitive radiotherapy was rare. Previous reports showed that radiation doses < 50 Gy were associated with high local recurrence.25, 26 However, these cases of recurrence were treated about 30 years ago, when the quality of radiotherapy was not as good as the current standard. Dattoli and other investigators reported that some patients who received low-dose radiotherapy (< 40 Gy) developed local recurrence.8, 15, 19 Many recent studies have shown that patients who received the appropriate radiation dose (45 Gy or 50 Gy) for the primary lesion did not develop local recurrence.3, 5–8

Cranial irradiation leads to late adverse effects, including intellectual deficits, hormonal abnormalities, arterial obstruction, and secondary malignancies.5, 11, 19, 27–30 Diabetes insipidus, panhypopituitarism, and delayed gonadal function are observed frequently in patients with intracranial germinoma at initial evaluation.4, 7, 8, 25, 28 Few authors have evaluated the frequency of radiation-induced adverse effects correctly by comparing intellectual status and hormonal abnormalities before and after treatment.10, 16, 19, 30–32 Merchant evaluated pediatric patients with primary brain tumors for evidence of endocrinopathy before radiotherapy.32 He reported that 66% of these patients had evidence of endocrinopathy before radiotherapy, and these observations suggested overestimation of the incidence of radiation-induced endocrinopathy. The incidence of pretreatment hormonal abnormalities in patients with germinomas ranged from 25% to 70%, and the incidence of patients with newly developed hormonal abnormalities after treatment ranged from 0% to 12%.10, 13, 16, 19, 24 Patients should be examined for the presence of hormonal abnormalities before and after treatment to correctly evaluate the influence of treatment on the occurrence of adverse effects.

In only a few studies, intellectual state was evaluated before and after treatment. Radiation-induced intellectual deficits are associated with patient age.31, 33–36 Clinical experience involving patients suffering from leukemia who received prophylactic cranial irradiation showed that the incidence of intellectual deficits in patients younger than 6 or 8 years of age was higher than that in older subjects.35, 36 Fortunately, intracranial germinomas usually arise in patients older than 8 years of age. In the current study, only 6 (4%) patients were younger than 8 years of age. In this study, we could neither evaluate the intellectual sequelae in patients after treatment, nor could we evaluate treatment-induced intellectual deficits. Shirato reported that 5 of 51 patients who received radiotherapy presented with intellectual deficits.5 Of these five patients, four received whole-ventricle irradiation at a dose of > 39 Gy. There is no agreement on the optimal radiation dose for the whole brain, but most investigators have applied doses of 25–30 Gy to eradicate microscopic disease.8, 37–39 Shibamoto recommended a low-dose craniospinal irradiation (20–24 Gy) for patients with positive or negative CSF cytology.9 There have been no recent reports showing that treatment outcome after high-dose radiation at doses of > 30 Gy for the whole brain is superior to that of low-dose radiation at < 30 Gy, and thus it has been suggested that low-dose radiation at < 30 Gy should be used for the whole brain.8, 37–39

New treatment strategies, including chemotherapy alone without radiotherapy or combined local irradiation with systemic chemotherapy, have been performed to avoid radiation-induced brain damage.16, 17, 22, 40, 41 Allen conducted a Phase II study that was designed to selectively reduce the radiation dose in patients showing a complete response to neoadjuvant carboplatin.17 Seventy percent of these patients showed a complete response to carboplatin, and the good responders received radiotherapy at a dose of 30 Gy in the involved field. The treatment outcome was excellent, but this study was performed in only 11 patients. Buckner conducted a Phase II study that evaluated the response rate and survival effects of treatment using etoposide plus cisplatin followed by radiotherapy.16 Of these 14 patients, 11 achieved a complete response after neoadjuvant chemotherapy. The good responders received limited local irradiation at a dose of 30 Gy. Only one patient developed recurrence in the spinal leptomeninges, and he was free from disease > 5 years after spinal irradiation. Aoyama reported the outcome of a prospective study that assessed the efficacy of induction chemotherapy followed by small-field radiotherapy.22 Twenty-seven patients with germinoma received etoposide plus cisplatin. The 5-year event-free survival rate of the patients with pure germinoma was 86%, but that of patients with syncytiotrophoblastic giant cells (STGC) was 44%. Five of 11 patients with STGC developed local or marginal recurrence, and marginal recurrence occurred in patients who received radiotherapy over an inadequate, small field. Even if radiotherapy combined with chemotherapy is applied, use of an adequate radiotherapy technique may be very important. After the establishment of adequate radiotherapy for local irradiation and optimized systemic chemotherapy, combination therapy, including induction chemotherapy followed by small-field radiotherapy, will become a standard strategy.

Some prognostic factors for intracranial germinoma after radiotherapy and/or chemotherapy have been reported.18, 24, 42–44 Our study could not show that a tumor size of 4 cm or more, female gender, or multifocal or disseminated tumors were associated with unfavorable event-free survival. However, there have been few studies showing the clinical value of tumor size for the prognosis of patients with intracranial germinoma.3, 5, 9 Shibamoto performed a single-institution prospective study in which the radiation dose was escalated according to tumor size.3 However, no other studies have evaluated the clinical significance of tumor size in patients with intracranial germinoma. Uematsu and other investigators reported that intracranial germinoma with STGC showed a greater tendency to recur than pure germinoma.42, 45 However, Ogawa and other investigators reported that there was no significant difference between the survival of patients with and without STGC.24, 43 It will be important in future studies to evaluate whether these clinical parameters are truly associated with the treatment outcome.