Physician self-reported treatment of brain metastases according to patients’ clinical and demographic factors and physician practice setting
1 Department of Radiation Oncology, Georgetown University Medical Center, 3800 Reservoir Road, Washington, D.C, USA
2 Department of Radiation Oncology, MD Anderson Cancer Center Orlando, Orlando, Florida, USA
3 Cancer Control Program, Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, Washington D.C, USA
Radiation Oncology 2012, 7:188 doi:10.1186/1748-717X-7-188Published: 8 November 2012
Limited data guide radiotherapy choices for patients with brain metastases. This survey aimed to identify patient, physician, and practice setting variables associated with reported preferences for different treatment techniques.
277 members of the American Society for Radiation Oncology (6% of surveyed physicians) completed a survey regarding treatment preferences for 21 hypothetical patients with brain metastases. Treatment choices included combinations of whole brain radiation therapy (WBRT), stereotactic radiosurgery (SRS), and surgery. Vignettes varied histology, extracranial disease status, Karnofsky Performance Status (KPS), presence of neurologic deficits, lesion size and number. Multivariate generalized estimating equation regression models were used to estimate odds ratios.
For a hypothetical patient with 3 lesions or 8 lesions, 21% and 91% of physicians, respectively, chose WBRT alone, compared with 1% selecting WBRT alone for a patient with 1 lesion. 51% chose WBRT alone for a patient with active extracranial disease or KPS=50%. 40% chose SRS alone for an 80 year-old patient with 1 lesion, compared to 29% for a 55 year-old patient. Multivariate modeling detailed factors associated with SRS use, including availability of SRS within one’s practice (OR 2.22, 95% CI 1.46-3.37).
Poor prognostic factors, such as advanced age, poor performance status, or active extracranial disease, correspond with an increase in physicians’ reported preference for using WBRT. When controlling for clinical factors, equipment access was independently associated with choice of SRS. The large variability in preferences suggests that more information about the relative harms and benefits of these options is needed to guide decision-making.