Survival following whole brain radiation treatment for cerebral metastases: an audit of 474 patients

https://doi.org/10.1016/j.radonc.2004.02.019Get rights and content

Abstract

Background and purpose

To report the outcome of patients with brain metastases from solid tumors treated with whole brain radiotherapy (WBRT) in a single institution. Given the high proportion of melanoma patients, a secondary aim was to compare our outcomes for patients with melanoma to those with other cancers.

Patients and methods

A retrospective audit identified 474 patients treated between January 1983 and December 1999. Survival was calculated using the Kaplan–Meier method. Cox regression modeling was used for multivariate analysis.

Results

Four hundred and fifty nine patients have died from their disease. The median survival was 4.1 months for the whole group and 3.6 months for the 42% of patients with melanoma. The 1 and 2 year survival was 15 and 5%. Six patients lived beyond 5 years. 105 of 186 patients with a single brain metastasis underwent surgery plus WBRT, and 81 received WBRT alone. Median survival was 8 and 4 months, respectively, (P<0.0001). 30 Gy in 10 fractions was used more commonly in the early part of the study compared to 20 Gy in 5 fractions more recently. There was no difference in survival by time period.

Conclusions

The survival in this series was comparable to other studies. Performance status, resection, dose, and the presence of extracranial disease appeared to be significant prognostic factors. The survival for the large number of patients with melanoma did not differ from the rest of the cohort.

Introduction

The treatment of brain metastases from solid tumors with whole brain radiotherapy (WBRT) and corticosteroids is a recognized standard treatment. Despite reported response rates of 50–75%, survival is poor, with a median survival of 3–6 months after radiotherapy for those with multiple metastases ([14], [17], [18], [20]). Response to radiotherapy may take up to one month and during this time up to 20–30% of patients will die from their systemic disease ([8]). Better outcomes are achieved in those with a single metastasis that can be surgically removed ([7], [20], [22], [25]). Studies also suggest improved outcomes with the use of stereotactic radiosurgery ([16], [17]).

Prognostic factors such as performance status, systemic tumor activity, age, site of primary and number of brain metastases have been identified ([2], [7], [10], [19], [21], [22], [30]). These prognostic factors may allow better selection of individual treatments.

Gaspar ([10]) retrospectively examined a multicenter database comprising 1200 patients pooled from three consecutive randomized RTOG trials performed between 1979 and 1993. After analysis by recursive partition analysis (RPA), three classes of patients were derived based on the four following prognostic factors: Karnofsky performance status (KPS), primary tumor status, presence of extracranial metastases and patient age. These classes were subsequently validated ([11]), and corroborated by Neider ([21]).

These classes were defined as follows: class I is age<65, KPS≥70, controlled primary and no extracranial metastases, class III is KPS<70 and class II is defined as those patients who were neither I nor III.

This paper reports a series of 474 patients with brain metastases who were treated with WBRT at the Department of Radiation Oncology, Royal Prince Alfred Hospital (RPAH), between January 1983 and December 1999.

The aim of this study was to document the overall survival of these patients. Secondary aims included:

  • (1)

    to determine whether our results were comparable to reported series in the literature

  • (2)

    to identify prognostic factors affecting outcomes for our patients

  • (3)

    to compare outcomes for those patients with melanoma versus other cancers.

Section snippets

Methods and materials

An audit of the database of the Department of Radiation Oncology identified 700 patients who were referred for WBRT from 1983 to 1999 inclusive.

A total of 171 patients with diagnoses such as acute lymphoblastic leukemia, cerebral primaries, leptomeningeal disease and skull base tumors were excluded. Twenty three overseas patients were omitted from the study due to inadequate follow up. Patients under the age of 16 were not included, neither were those patients who were seen but not treated nor

Statistics

Survival analyses were made using the Kaplan–Meier method. SPSS version 9.0 and SPIDA version 6 packages were used for statistical analysis. Cox regression modeling was used for multivariate analysis.

Dose fractionation schedules were categorized as >30, 30 Gy in 10 fractions, 20 Gy in 5 fractions or other, in order to allow comparison in the multivariate analysis. For the multivariate analysis, cases with missing data were omitted. As statistical tests revealed that the proportional hazards

Results

Four hundred and seventy four patients were included in this study, 313 (66%) men and 161 (34%) women. The mean age was 58 years (median 59 years, range 16–89). Performance status was recorded as follows: 251 (53%) were ECOG 0 or 1, 84 (18%) ECOG 2, 57 (12%) ECOG 3 or 4, and unknown in 82 (17%).

The primary site of malignancy is listed in Table 1. Notably, 42% had known primary cutaneous melanoma and 30% had a lung primary. Histopathology was available for 452 patients, of whom 201 (42%) had

Discussion

This audit presents the outcomes for 474 patients with brain metastases from solid tumors irradiated at a single institution. This study is unusual due to the very large proportion of melanoma patients (42%). In other series, lung and breast metastases account for the majority of patients treated. In an analysis of 1200 patients from the RTOG database, 17% had melanoma primaries, whereas 61% had lung, and 12% had breast cancer ([10]). In 1999, Lagerwaard ([19]) published data on 1292 patients.

Conclusion

This study of 474 patients treated at a single institution has survival comparable to other studies. The survival for the large number of patients with melanoma did not differ from the rest of the cohort.

References (31)

Cited by (0)

View full text