A novel risk stratification model for patients with brain metastasis: Incorporating with physiological signals.

38 Background: Brain metastasis (BM) develops in 10%-40% of patients with malignancies. Survival expectancy is critical for choosing an optimal treatment for BM patients. Previous reports have proposed different risk stratification systems based on clinical and image-related factors. Heart rate variability (HRV), a physiological factor, has also been reported as a novel prognosticator for BM patients. In this study, we proposed a novel risk stratification model for BM patients by incorporating HRV signals.

METHODS: Patients with BM referred to our department for palliative RT were prospective recruited. Those who were not suitable for HRV evaluation were excluded for analysis. A 5-minute ECG was acquired before treatment started for each patient for subsequent HRV signal analysis. All patients underwent WBRT for 30-37.5 Gy in 10-15 daily fractions. Local boost to gross intracranial tumors were performed selectively. The primary end-point for this study was overall survival (OS). Survival analysis was carried out by Kaplan-Meier method and factors were compared by log-rank test and cox-regression. A novel risk stratification model combining HRV index SDNN and KPS was proposed.

RESULTS: Total 62 patients were recruited from Jan 2010. The median follow-up time was 40.7 months for patients who were still alive. Only 1 patient (1.6%) met the criterion of RPA class I. The median OS for all patients was 4.5 months. Univariate and multivariate analysis revealed that SDNN ≥ 10ms and KPS ≥ 70 were independent survival prognosticators. For a selected group of patients (n = 29, Group 1, SDNN ≥ 10ms and KPS ≥ 70) the median OS was 9.9 months, which is significantly higher than the rest 33 patients (Group 2) as of 2.4 months. The 3-month/6-month OS rate in Group 1 and Group 2 were 82.8%/72.4% and 36.4%/24.2% (p < 0.001).

CONCLUSIONS: Using physiological factors for risk stratification is novel. Our results demonstrated that a group of patients with potential longer survival could be selected by using HRV index and KPS. This group of patients should be considered differently in treatment selection. Further studies are warranted to validate our findings and improve the quality of care in patients with BM.