BACKGROUND. Multiinstitutional experience with the management of cerebral m
etastases from malignant germ cell tumors (MGCT) is presented.
METHODS, Clinical data regarding brain metastases from MGCT at diagnosis (G
roup 1 [56 patients]) or after cisplatin-based chemotherapy (Group 2 [83 pa
tients]) were collected retrospectively All patients in Group I received "c
onventional" cisplatin-based chemotherapy supplemented by cerebral radiothe
rapy (36 patients) and/or neurosurgery (10 patients). In the patients in Gr
oup 2 cerebral metastases were detected a median of 9 months after the init
iation of chemotherapy. Thirty-five patients received chemotherapy, 59 pati
ents received radiotherapy, and 25 patients underwent neurosurgery.
RESULTS. The 5-year cause specific survival rate in Group 1 was 45% (95% co
nfidence interval [CI], 31-59%). Neurosurgery and the absence of extracereb
ral, nonpulmonary visceral disease, but not cerebral radiotherapy, were ind
ependent predictors of good prognosis. The 5-year cause specific survival r
ate in Group 2 tvas 12% (95% CI, 4-20%), but was 39% among patients with an
isolated brain recurrence (24 patients). Radiotherapy, but not chemotherap
y, represented an independent predictor of good prognosis together with bra
in metastases at first recurrence and the absence of extracerebral recurren
ce.
CONCLUSIONS. Among patients with brain metastases at the time of diagnosis
of an MGCT, cisplatin-based chemotherapy resulted in a 5-year cause specifi
c survival rate of 45%, with cerebral radiotherapy having limited impact. T
he 5-year cause specific survival rate for all patients with brain metastas
es after cisplatin-based chemotherapy was 12%, but increased to 39% in pati
ents with an isolated brain recurrence. Cerebral radiotherapy (and neurosur
gery) represent essential treatment modalities for patients in whom brain m
etastases are diagnosed after induction chemotherapy. (C) 1999 American Can
cer Society.