Abstract 372P
Background
Brain metastases (BM) from lung cancer is a common clinical scenario in which, prognostic assessment is vital for clinical decision-making. The updated diagnostic-specific graded prognostic assessment (Lung-molGPA) is one of the most useful indices for prognosis in this scenario. Even though there is growing evidence that liver metastases (LM) confer worst survival, this is not considered in any prognostic score for BM.
Methods
From a single-center retrospective database of 782 patients treated for NSCLC between 2011-2016, patients with newly diagnosed BM were identified. We examined the effect of LM in the survival of our cohort of patients with BM and according to stratification using Lung-molGPA.
Results
A total of 134 patients presented BM during the period specified. The median age at BM diagnosis was 62.2 years. According to histologic classification, 68.7% were classified as adenocarcinoma, and 26.9% as non-adenocarcinoma. BM were present at diagnosis in 42.5%. LM were present in 20.9% at the moment of which BM were diagnosed. Gene alterations were detected in 18.6%. The overall survival (OS) from the time of the initial diagnosis of the disease was 19.7 months (95% confidence interval (CI): 15.5-24). We found that the free interval of BM (≥ 6 months) had prognostic significance (HR 0.46; P<0.00001). From the time of diagnosis of BM, the OS was 11.8 months (95% CI: 7.1-16.4). A better KPS (>70), younger age (<60), and number of BM (1) were significantly related to decreased risk of death. We also found that the patients with LM had a median OS of 4.1 vs 14.3 months in the group without LM (HR 1.92; P<0.001). Finally, we validate the Lung-molGPA in our cohort of patients showing a significant separation between groups (P<0.00001). We found that for patients without LM the Lung-molGPA scores kept the good stratification power (P<0.00001), but no differences were observed in the group of patients with LM (P=0.532).
Conclusions
When using BM prognosis indices in NSCLC patients, the individual presence of LM or the free interval of BM should be considered for an accurate stratification, to individualize treatment options and better select patients for clinical trials.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.