Patients with metastatic brain disease have a poor prognosis and curative treatment is not achievable in most clinical situations, with 50% of patients dying from their neurological cancer rather than their extracranial cancer . Due to this unfortunate outlook, maximizing patient's period of survival and comfort level is of great importance. Although several Phase III studies have been published assessing the efficacy of different treatment modalities, many questions still remain unanswered and further randomized evidence is needed not only to prove superior treatments in comparison studies, but to identify optimal courses of treatment in unique patient subsets [6–9, 13–17]. Our comprehensive analysis evaluates the clinical effects treatment regimen, age, performance status, primary tumor histology, number of brain metastases, and total volume of brain metastases have on patient survival.
Perhaps the most questionable matter in the management of patients with brain metastases is whether the addition of WBRT to SRS will provide patients with a superior prognosis when compared to patients treated with SRS alone . Our study did not find statistically significant survival differences between the SRS alone treatment group and the SRS with WBRT treatment group in both univariate and multivariate analysis. In the randomized controlled trial published by Aoyama et al. , the authors evaluated the clinical outcomes of patients treated with SRS with or without WBRT and also witnessed no significant (p = 0.4) differences in survival between the two treatment arms. However, the patients treated with WBRT with SRS had a substantially better 12-month brain tumor recurrence rate (p < 0.001) and underwent salvage therapy (p < 0.001) less often than the patients treated with SRS alone, but these increases in tumor control did not affect patient survival. Several retrospective cohort studies published in the last ten years have also reported that the addition of WBRT to SRS does not result in superior levels of patient survival [11, 18–21].
On multivariate analysis, we found that the survival of the SRS alone treatment arm did not statistically differ when compared to the survival of the resection with WBRT treatment arm. These data correlate with the Phase III randomized trial conducted by Muacevic et al. . A total of 64 patients with a single, surgically accessible brain metastasis ≤30 mm in diameter, a Karnofsky Performance Score (KPS) ≥70, and a controlled primary cancer were randomized into a GK radiosurgery alone group (31 patients) and a surgery with WBRT group (33 patients). The authors reported non-significant differences in survival between the two treatment groups. Rades et al.  retrospectively compared SRS alone and surgery with WBRT in 260 patients classified in RPA class 1 or 2  that were diagnosed with 1 to 2 brain metastases and also reported that the two groups did not differ in survival. Our multivariate analysis also found superior levels of survival in patients treated with resection with SRS when compared to patients treated with SRS alone. The body of world literature lacks sufficient studies comparing patients treated with SRS alone against patients treated with resection with SRS. However, survival differences between patients treated with SRS alone and patients treated with resection with SRS was recently reported in another study by Rades et al. . The authors analyzed the clinical outcomes of 164 patients of advanced age (≥65 years). Specifically, 34 patients were treated with WBRT alone, 43 patients were treated with SRS alone, 41 patients were treated with resection + SRS, and 46 patients were treated with resection + WBRT+ SRS boost. In contrast to our results, which favored the resection with SRS treatment group, the authors reported that treatment regimen influenced survival, with the SRS alone treatment group surviving a greater time than the resection + SRS treatment group. The results reported by Rades et al.  can be explained when considering the risks of surgery in elderly patients. This data permits the treatment of select patients who are <65 years of age and are functionally independent with resection in combination with SRS.
In subset analysis, patients treated with WBRT alone at our institution exhibited the shortest period of survival, with each of the other five treatment arms surviving a substantially greater time than the WBRT alone treatment arm. Although it is likely that the treatment arms consisted of very different patient subsets with respect to ECOG-PS class, tumor number, tumor volume, and extent of systemic disease, both univariate and multivariate analysis found statistically significant differences between the hazard ratio of patients treated with WBRT and the hazard ratio of patients treated with SRS alone. No randomized controlled trials have been conducted assessing patients treated with SRS alone compared with patients treated with WBRT alone. However, in a recent literature review, Linskey et al.  found level 3 evidence indicating that patients with 1 to 3 brain metastases that are treated with SRS alone have superior levels of survival when compared to patients treated with WBRT alone.
As expected, we found that age and performance status are both significant predictors in determining patient prognosis, as survival statistically favored patients <65 years old in univariate analysis and patients in a lower ECOG-PS class in both univariate and multivariate analysis. Several comparison studies have reported a survival dependency on patient age and performance status. Sanghavi et al.  retrospectively analyzed the outcomes and potential prognostic factors of a total of 502 patients treated with SRS with WBRT and 1200 patients treated with WBRT alone and found that survival was more pronounced in patients with a higher KPS (p = 0.0001), a controlled primary cancer (p = 0.0023), the absence of extracranial cancer (p = 0.0001), and a lower RPA class (p = 0.000007). Kocher et al.  compared the efficacy of SRS alone against WBRT alone in 255 patients with 1 to 3 brain metastases and reported statistically significant increases in median survival in patients categorized in RPA class 1 (p < 0.0001) and RPA class 2 (p < 0.04). Frazier et al.  retrospectively analyzed 237 patients treated with SRS ± WBRT and also found that survival statistically favored patients that were <65 years of age (p = 0.008) with KPS values >70 (p = 0.034).
The number and volume of brain metastases patients possess at the time of diagnosis are crucial factors in prescribing the most advantageous course of treatment in select patient groups. When evaluating our six treatment arms in univariate analysis; however, the number and size of brain metastases did not influence patient survival. Tumor resection in combination with WBRT and/or SRS in treating patients with a single brain metastasis is recommended for those who present with severe neurologic deficits, a ventricular obstruction, or a tumor of a large intracranial volume (which often produces mass effect) . When the patient has controlled neurological symptoms, a tumor/s of a small intracranial volume, a single brain metastasis, a surgically inoperable brain metastasis, or multiple brain metastases, SRS alone or in combination with WBRT is often the recommended course of treatment . Questions remain regarding the survival dependency on the number and size of brain metastases patient groups possess. Studies have shown increased survival levels in patients with a single brain metastasis that were treated with radiosurgery [6, 26]. However, other publications have reported that tumor volume has a greater impact on patient survival than number of brain metastases and primary tumor histology, with patients possessing small tumor volumes surviving a greater period of time [27–30]. Further study and research is needed on how the number and total volume of brain metastases affect patient survival.
The histologic subtype of the primary tumor may be an essential predictor in assessing the survival advantage of specific patient subsets. NSCLC is known to produce the greatest amount of metastatic brain lesions [31, 32]. In univariate analysis, survival statistically favored patients with NSCLC when compared to patients with SCLC and patients classified in the other primary histology group. In multivariate analysis; however, survival statistically favored patients in the breast cancer group when compared to patients in the NSCLC group. Increases in the survival of breast cancer patients when compared to NSCLC patients was also recently reported in the survival analysis of 237 patients treated with radiosurgery by Frazier et al. . These results are likely due to advances in the surgical and chemotherapeutic care of breast cancer patients . It was also observed in multivariate analysis that survival statistically favored patients with NSCLC when compared to the combined melanoma and renal-cell carcinoma group. Traditionally, melanoma and renal-cell carcinoma have been classified as "radioresistant" tumor histologies because of their negative response to standard radiation treatment. However, several studies have reported positive outcomes when treating patients with melanoma and renal-cell carcinoma primaries with radiosurgery [34–40]. In a phase II trial conducted by Manon et al. , 31 patients diagnosed with melanoma, renal-cell carcinoma, and sarcoma primary cancers with 1 to 3 brain metastases were treated with SRS alone. The 3 and 6 month intracranial failure rate for the evaluated patients was found to be 25.8 and 48.3%, respectively. The authors concluded that delaying WBRT for patients with melanoma, renal-cell carcinoma, and sarcoma primary cancers may be appropriate for specific subgroups of patients, but must be approached with caution.