Purpose: To evaluate in terms of probabilities of local-regional control an
d survival, as well as of treatment-related toxicity, results of radiosurge
ry for brain metastasis arising in previously irradiated territory.
Patients and methods: Between January 1994 and March 2000, 54 consecutive p
atients presenting with 97 metastases relapsing after whole brain radiother
apy (WBRT) were treated with stereotactic radiotherapy. Median interval bet
ween the end of WBRT and radiosurgery was 9 months (range 2-70). Median age
was 53 years (24-80), and median Karnofski performance status (KPS) 70 (60
-100). Forty-seven patients had one radiosurgery, five had two and two had
three. Median metastasis diameter and volume were 21 mm (6-59) and 1.2 cc (
0.1-95.2), respectively. A Leksell stereotactic head frame (Leksell Model G
, Elektra, Instrument, Tucker, GA) was applied under local anesthesia. Irra
diation was delivered by a gantry mounted linear accelerator (linacs) (Satu
rne, General Electric). Median minimal dose delivered to the gross disease
was 16.2 Gy (11.8-23), and median maximal dose 21.2 Gy (14-42).
Results: Median follow-up was 9 months (1-57). Five metastases recurred. On
e- and 2-year metastasis local control rates were 91.3 and 84% and 1- and 2
-year brain control rates were 65 and 57%, respectively. Six patients died
of brain metastasis evolution, and three of leptomeningeal carcinomatosis.
One- and 2-year overall survival rates were 31 and 28%, respectively. Accor
ding to univariate analysis, KPS, RPA class, SIR score and interval between
WBRT and radiosurgery were prognostic factors of overall survival and brai
n free-disease survival. According to multivariate analysis, RPA was an ind
ependent factor of overall survival and brain free-disease survival, and th
e interval between WBRT and radiosurgery longer than 14 months was associat
ed with longer brain free-disease survival. Side effects were minimal, with
only two cases of headaches and two of grade 2 alopecia.
Conclusion: Salvage radiosurgery of metastasis recurring after whole brain
irradiation is an effective and accurate treatment which could be proposed
to patients with a KPS > 70 and a primary tumour controlled or indolent. We
recommend that a dose not exceeding 14 Gy should be delivered to an isodos
e representing 70% of the maximal dose since local control observed rate wa
s similar to that previously published in literature with upper dose and si
de effects were minimal. (C) 2001 Elsevier Science Ireland Ltd. All rights
reserved.