Treatment of Five or More Brain Metastases with Stereotactic Radiosurgery

Abstract

Purpose/Objective(s)To examine the outcomes of patients with five or more brain metastases treated in a single session with stereotactic radiosurgery (SRS).Materials/MethodsSixty-four patients with radiographically diagnosed brain metastases treated with LINAC or Gamma Knife based SRS to 5 or more lesions in a single session were reviewed. Primary disease type, number of lesions treated, SRS records, whole brain radiation (WBRT) records, Karnofsky performance score (KPS) at SRS, and status of primary disease and systemic disease at SRS were included. Patients were treated using standard dosing defined by RTOG 90-05, with adjustments for critical structures and/or brainstem. We defined prior WBRT as WBRT completed >1 month prior to SRS and concurrent WBRT as WBRT completed within 1 month before or after SRS. Kaplan-Meier estimates and Cox proportional hazard regression were used to determine which patient and treatment factors predicted overall survival.ResultsThe median overall survival (OS) after SRS was 7.5 months. The median KPS was 80 (range, 60–100). KPS of >80 significantly influenced OS (median OS = 4.8 months for KPS <70 vs 8.8 months for KPS >80, p = 0.0097). The number of lesions treated did not significantly influence OS (median OS = 6.6 months for <8 vs 9.9 months for >8, p = ns). Primary site histology did not significantly influence median OS and was broken down as the following: breast (n = 13) vs other (11.5 months for breast vs 6.8 months, p = ns), NSCLC (n = 23) vs other (9.6 months for NSCLC vs 6.8 months, p = ns), or radioresistant (RadR) histology (renal cell/melanoma, n = 20) vs. other (4.2 months for RadR vs 9.6 months, p = ns). On multivariate Cox modeling, KPS and prior WBRT significantly predicted for OS. WBRT prior to SRS compared to concurrent WBRT significantly influenced survival, with a risk ratio (RR) of 0.423 (CI = 0.191–0.936, p = 0.0338). No WBRT group compared to concurrent WBRT had a RR of 0.447 (CI = 0.176–1.134, p = ns), and no WBRT group compared to prior WBRT had a RR of 1.051 (CI = 0.534–2.070, p = ns). KPS of <70 predicted for poorer outcomes, with a RR of 2.164 (CI = 1.157–4.049, p = 0.0157).ConclusionsStereotactic radiosurgery to 5 or more lesions is an effective palliative option, especially for patients previously treated with WBRT. A KPS of 80 and above is predictive of better outcomes. Purpose/Objective(s)To examine the outcomes of patients with five or more brain metastases treated in a single session with stereotactic radiosurgery (SRS). To examine the outcomes of patients with five or more brain metastases treated in a single session with stereotactic radiosurgery (SRS). Materials/MethodsSixty-four patients with radiographically diagnosed brain metastases treated with LINAC or Gamma Knife based SRS to 5 or more lesions in a single session were reviewed. Primary disease type, number of lesions treated, SRS records, whole brain radiation (WBRT) records, Karnofsky performance score (KPS) at SRS, and status of primary disease and systemic disease at SRS were included. Patients were treated using standard dosing defined by RTOG 90-05, with adjustments for critical structures and/or brainstem. We defined prior WBRT as WBRT completed >1 month prior to SRS and concurrent WBRT as WBRT completed within 1 month before or after SRS. Kaplan-Meier estimates and Cox proportional hazard regression were used to determine which patient and treatment factors predicted overall survival. Sixty-four patients with radiographically diagnosed brain metastases treated with LINAC or Gamma Knife based SRS to 5 or more lesions in a single session were reviewed. Primary disease type, number of lesions treated, SRS records, whole brain radiation (WBRT) records, Karnofsky performance score (KPS) at SRS, and status of primary disease and systemic disease at SRS were included. Patients were treated using standard dosing defined by RTOG 90-05, with adjustments for critical structures and/or brainstem. We defined prior WBRT as WBRT completed >1 month prior to SRS and concurrent WBRT as WBRT completed within 1 month before or after SRS. Kaplan-Meier estimates and Cox proportional hazard regression were used to determine which patient and treatment factors predicted overall survival. ResultsThe median overall survival (OS) after SRS was 7.5 months. The median KPS was 80 (range, 60–100). KPS of >80 significantly influenced OS (median OS = 4.8 months for KPS <70 vs 8.8 months for KPS >80, p = 0.0097). The number of lesions treated did not significantly influence OS (median OS = 6.6 months for <8 vs 9.9 months for >8, p = ns). Primary site histology did not significantly influence median OS and was broken down as the following: breast (n = 13) vs other (11.5 months for breast vs 6.8 months, p = ns), NSCLC (n = 23) vs other (9.6 months for NSCLC vs 6.8 months, p = ns), or radioresistant (RadR) histology (renal cell/melanoma, n = 20) vs. other (4.2 months for RadR vs 9.6 months, p = ns). On multivariate Cox modeling, KPS and prior WBRT significantly predicted for OS. WBRT prior to SRS compared to concurrent WBRT significantly influenced survival, with a risk ratio (RR) of 0.423 (CI = 0.191–0.936, p = 0.0338). No WBRT group compared to concurrent WBRT had a RR of 0.447 (CI = 0.176–1.134, p = ns), and no WBRT group compared to prior WBRT had a RR of 1.051 (CI = 0.534–2.070, p = ns). KPS of <70 predicted for poorer outcomes, with a RR of 2.164 (CI = 1.157–4.049, p = 0.0157). The median overall survival (OS) after SRS was 7.5 months. The median KPS was 80 (range, 60–100). KPS of >80 significantly influenced OS (median OS = 4.8 months for KPS <70 vs 8.8 months for KPS >80, p = 0.0097). The number of lesions treated did not significantly influence OS (median OS = 6.6 months for <8 vs 9.9 months for >8, p = ns). Primary site histology did not significantly influence median OS and was broken down as the following: breast (n = 13) vs other (11.5 months for breast vs 6.8 months, p = ns), NSCLC (n = 23) vs other (9.6 months for NSCLC vs 6.8 months, p = ns), or radioresistant (RadR) histology (renal cell/melanoma, n = 20) vs. other (4.2 months for RadR vs 9.6 months, p = ns). On multivariate Cox modeling, KPS and prior WBRT significantly predicted for OS. WBRT prior to SRS compared to concurrent WBRT significantly influenced survival, with a risk ratio (RR) of 0.423 (CI = 0.191–0.936, p = 0.0338). No WBRT group compared to concurrent WBRT had a RR of 0.447 (CI = 0.176–1.134, p = ns), and no WBRT group compared to prior WBRT had a RR of 1.051 (CI = 0.534–2.070, p = ns). KPS of <70 predicted for poorer outcomes, with a RR of 2.164 (CI = 1.157–4.049, p = 0.0157). ConclusionsStereotactic radiosurgery to 5 or more lesions is an effective palliative option, especially for patients previously treated with WBRT. A KPS of 80 and above is predictive of better outcomes. Stereotactic radiosurgery to 5 or more lesions is an effective palliative option, especially for patients previously treated with WBRT. A KPS of 80 and above is predictive of better outcomes.