The table-5 contains data from the largest studies of patients with similar diseases who were followed at different institutions, and treated with different modalities. The median survival of 6.4 months reported in our study is consistent with other reports describing the natural history and treatment with WBRT alone of breast carcinoma metastatic to the brain [19,20]. Researchers reporting on different treatment modalities in similar groups of patients have noted median survival times in the range of 4 – 16 months after surgery with or without WBRT, SRS or WBRT alone [19-25], as showed in table-5.

The RTOG has evaluated a number of different radiation fractionation schemes, but median survival seems independent of the dose and schedule [26-29]. In our study, total dose of WBRT was not a statistically significant predictor of overall survival. Surgery is an important modality for patients with a single brain metastasis, particularly when favorable prognostic factors and systemic disease control are present [30,31]. Our data showed that patients undergoing resection of brain lesion followed by WBRT was associated with significantly better overall survival (p < 0.0001) than patients submitted to biopsy or WBRT alone. Patchell et al [15], randomly assigned 48 patients with single brain metastases (10% with breast primaries) to surgery followed by WBRT versus WBRT alone. Patients in the combined arm experienced a longer duration of functional independence (38 v 8 weeks), and improved survival (40 v 15 weeks; P < .01). Noordijk et al [30] conducted a randomized trial of 63 patients (19% with breast primaries) that confirmed and extended these findings. Importantly, in this study, the benefit of combinedmodality therapy was seen only in patients with stable or absent extracranial disease. Patients with progressive extracranial disease at study entry achieved a median survival of only 5 months, irrespective of the allocated treatment. One additional trial failed to demonstrate a survival or quality-of-life benefit [31]. Nearly half of the patients in this trial had extracranial disease, and 10 of 43 patients randomly assigned to radiotherapy underwent surgical resection.

The end point of this cohort was to evaluate the different prognostic factors related with overall survival and to analyze the importance of recursive partitioning analysis (RPA) class (RTOG) in patients with brain metastasis. In our data, the prognostic factors in the univariate analysis associated with better survival were: Higher KPS, solitarymetastasis, surgical resection, RPA class I, BS-BM -3, control primary tumor, and absence of extracranial metastases; in the multivariate analysis RPA class I and surgical resection maintained associated with better survival, being all these prognostic factors were showed for others authors in previous studies [9,14,16].

In recent publication, the Radiation Enhancing Allosteric Compound for Hypoxic Brain Metastases (REACH) study [32] tested the hypothesis that adding efaproxiral to WBRT plus supplemental oxygen would improve survival better than WBRT with supplemental oxygen alone. The results of this study suggest that efaproxiral, may improve response rates to WBRT and survival in patients with brain metastases, mainly metastases from breast cancer. Moreover, in this phase III study; KPS, number of extracranial metastatic sites, and sex had the highest statistical significance in multivariate analysis. In our study, the others factors (age, chemotherapy, dose and fractionation schedule) analyzed were not associated with any effect in survival. RPA class in this study showed similar results to RTOG protocols to identify patients with different results [9], with the median survival time for class I (11.5 months), II (6.2 months) and III (3.0 months) (p = 0.0001), respectively. In this series, the BS-BM was effective in identifying patients with different outcomes in a simple and easy manner. A BS-BM of 0 had greater specificity but lower sensitivity BS-BM. However, in our study BS-BM when compared to RPA class in multivariate analyses did not achieved significant statistical in Cox regression backward method, this data shows that RPA class is more powerful and precise than BS-BM in to predict survival for patients with brain metastases from breast cancer. Thus, theses results do not invalidated its use as a system for predict survival, only it confirms that the RPA is a more efficient system for this. But, which was the reason for this to occur? Probably this fact occurred because the BS-BM takes into account only three variables (i.e., KPS, primary tumor control, and the presence of extracranial metastases), which have been found in most studies, as well as in our own evaluation, to be the most important prognostic factors for survival. Thus it seems that less important factors had been affected indirectly by the other main factor as extracranial metastases or surgical resection of lesions. Patients with three or more BMs had a greater proportion of extracranial metastases and smaller than surgical resection of lesions than those with one or two BMs (48%vs 22 % and 26% vs 58 %, respectively). In this way, our data showed that BS-BM system may be used effectively in patients with brain metastases treated by WBRT alone or combined with surgery.

Conclusion

In conclusion, both the stratification systems examined were able to identify quite well those patients who might or might not benefit from WBRT. RPA class was shown to be the most reliable indicators of survival. BS-BM has the advantage of focusing on only three major factors for survival. Our data suggest that patients with brain metastases from breast cancer classified as RPA class I may be effectively treated with local resection followed by WBRT, mainly in those patients with single metastases, higher KPS and cranial extra disease controlled. We believe that patients presenting with a RPA Class III or BS-BM of 0 are clearly not good candidates for surgical resection followed by WBRT. Patients with RPA Class II or BS-BM of 1 in general have a poor outcome, and, in these patients, the decision concerning treatment remains difficult. Despite the generally ominous prognosis, some patients still benefit from surgical ressection. Brain metastases from breast cancer pose numerous challenges. Future areas of research may include characterization of risk factors and in this way to evaluated new approaches for the treatment of brain metastases.