Stereotactic Radiosurgery with Whole Brain Radiation Therapy Improves Survival in Patients with Brain Metastases: Report of Radiation Therapy Oncology Group Phase III Study 95-08

Reviewer: Ryan Smith, MD
Last Modified: October 7, 2002

Presenter: Paul Sperduto
Presenter's Affiliation: RTOG
Type of Session: Plenary


    Radiation therapy to the whole brain has been the standard treatment for brain metastases. Recently, in good performance status patients, there has been some evidence for better outcomes with the additional use of stereotactic radiosurgery (SRS). This study investigates the the addition of SRS onto whole brain radiation therapy for the control of brain metastases.

Materials and Methods

  • This is a prospective randomized study of 334 patients with 1-3 brain metastases, randomized to whole brain radiation therapy (WBRT) alone vs. WBRT + SRS.
  • Patients were stratified according to 1 brain metastasis vs. 2-3 metastases and whether they had systemic disease or not
  • Patients were required to have an MRI showing 1-3 metastases, size < 4.1 cm, KPS 70 or greater.
  • Treatment characteristics were well-balanced between groups. 64% of the patients had lung cancer, with 50% with adenocarcinoma
  • 32% had no evidence of extracranial metastases
  • 56% had a solitary metastasis, 24% had 2 metastases, 20% had 3 metastases
  • WBRT was given in 15 fractions of 2.5 Gy for a total dose of 37.5 Gy
  • Radiosurgery was given as per a size criteria: <2.1 cm-24 Gy; 2.1-3 cm-18 Gy; 3.1-4 cm-15 Gy
  • 19% did not recieve SRS due mainly to either disease progression or patient refusal


  • Median survival was 6.5 mo (SRS) vs. 5.7 mo (WBRT alone) (p=.13)
  • In patients with solitary brain metastases, SRS patients did better, with median survival of 6.5 mo vs. 4.9 mo
  • In patients with 1-3 metastases and age <50, median survival was 9.9 months (SRS) vs. 8.3 months (WBRT alone), p=.04
  • In patients with 1-3 metastases with NSCLC, survival was 5.9 months (SRS) vs. 3.9 months (WBRT alone), p=.04
  • In those patients with KPS 90-100, there was an increased median survival in patients treated with SRS (10.2 months vs. 7.4 months, p=.07)
  • In patients that are RPA class I (reflecting control of systemic disease), median survival was 11.6 months (SRS) vs. 9.6 months (WBRT), p=.04
  • On multivariate analysis, RPA class was most predictive, with KPS, age <65, and the presence of extracranial metastases also predictive.
  • Other endpoints included: improved KPS and less dependence on steroids in the SRS group
  • Local control was 82% in the SRS group vs. 71% in the WBRT alone group
  • Acute/late toxicity was <3% in each group
  • 17% received salvage SRS in WBRT alone group, with additional 4-5% surgical salvage

Author's Conclusions

  • WBRT + SRS improved survival in patients with: A solitary metastasis, 1-3 brain metastases and age <50, 1-3 metastases and NSCLC, and RPA class I patients
  • OS was not improved in the group as a whole because of 19% of the patients randomized to receive SRS not receiving it and the high salvage rate with SRS in the WBRT alone group
  • The addition of SRS improves local control, achieves improved or stable KPS, and decreases steroid dependence
  • This is the new standard of care in patients with brain metastases fitting the eligibility characteristics of RTOG 95-08

Clinical/Scientific Implications

    It is difficult to accept WBRT + SRS as the universal standard of care for patients with 1-3 brain metastases. As per the original study design, stratification was only for those patients with 1 brain metastasis vs those patients with 2-3 metastases and for those patients with controlled systemic disease or not. Therefore, any other subset analysis reported on is statistically flawed. Therefore, the results of this study must be viewed with some caution. A previous report of this study last year showed no benefit to the addition of SRS for patients with 2-3 brain metastases. Again, as per the stratification, the group with a solitary metastasis did benefit from SRS and therefore, patients fitting this criteria should be strongly considered for SRS boost following WBRT. However, the other groups that are claimed to have a benefit should not be universally given a SRS boost based on the other subset analyses. Additionally, the improved survival is on the order of a month, or at the most two months, which may be statistically significant, but which must be improved upon to be clinically useful. The 1 year local control numbers should also be taken with a note of caution, given that the median survival for the group is approximately 6 months. Granted, the palliation of symptoms, including increasing KPS and decreasing steroid dependence is important, though whether that small benefit in these areas alone warrants the use of SRS in these patients should be decided clinically-not as a broad recommendation. SRS is definitely an interesting tool in treating patients with brain metastases, but should not be used in all patients based on this study alone.

Oncolink's ASTRO Coverage made possible by an unrestricted Educational Grant from Ortho Biotech.