Prognostic value of the modified systemic inflammation score in non-small-cell lung cancer with brain metastasis

Background Brain metastases (BM) from non-small-cell lung cancer (NSCLC) is the most common brain malignancy. Systemic inflammation biomarkers have recently been evaluated as prognosis indicators in several tumors. The combination of these markers has not been evaluated in NSCLC with BM yet. Here, we explored the predictive value of pretreatment inflammatory biomarkers and established a novel, clinically applicable prognostic index for NSCLC patients with BM. Methods A retrospective investigation of 951 NSCLC patients newly diagnosed with BM at Sun Yat-sen University Cancer Center was conducted. We randomly divided patients into a training cohort (n = 674) or validation cohort (n = 277). Receiver operating characteristic (ROC) curve analysis was carried out to obtain the optimal cut-off values of pretreatment systemic inflammatory indexes. The associations between serum biomarkers and overall survival (OS) were analyzed by Kaplan-Meier curves and Cox proportional models. The resulting prediction model has been externally verified through the validation cohort. Results The optimal cut-off value of the neutrophil-lymphocyte ratio (NLR) in predicting OS was 4.71, while the clinical standard of 40 mg/L was chosen as the optimal cut-off value of albumin. Univariate and multivariate analyses revealed that patients receiving local treatment, chemotherapy, a NLR < 4.71 and albumin [greater than or equal to] 40 mg/l independently predicted improved survival. We combined the two inflammatory indexes (NLR and albumin level) to establish the modified systemic inflammation score (mSIS) which divides patients into low risk, medium risk or high-risk groups. The 1-year OS rates of three groups were 59.7%, 40.5% and 29.4%, respectively in the training cohort. The same result was verified in the validation cohort with the 1-year OS rates 69.7%, 47.0% and 7.7%, respectively. The mSIS exhibited better discrimination power than the American Joint Committee on Cancer's (AJCC) 7th T + N staging system in the training cohort (Harrell's concordance index (C-index): 0.744 vs 0.502, P < 0.05), and the discrimination was also superior to that of AJCC's 7th T + N staging system in the validation cohort (C-index: 0.724 vs 0.527, P < 0.05). The 1-year and 2-year OS rates of the AUC also exhibited superior survival predictive ability to that of the AJCC's 7th T + N staging system in NSCLC patients with BM. Conclusion The pretreatment mSIS may be an independent prognostic factor for OS in