The Glasgow Prognostic Score Predicts Poor Survival in Cisplatin-Based Treated Patients with Metastatic Nasopharyngeal Carcinoma

Abstract

Background

Several inflammation-based prognostic scoring systems, including Glasgow Prognostic Score (GPS), neutrophil to lymphocyte ratio (NLR) and platelet to lymphocyte ratio (PLR) have been reported to predict survival in many malignancies, whereas their role in metastatic nasopharyngeal carcinoma (NPC) remains unclear. The aim of this study is to evaluate the clinical value of these prognostic scoring systems in a cohort of cisplatin-based treated patients with metastatic NPC.

Methods

Two hundred and eleven patients with histologically proven metastatic NPC treated with first-line cisplatin-based chemotherapy were retrospectively evaluated. Demographics, disease-related characteristics and relevant laboratory data before treatment were recorded. GPS, NLR and PLR were calculated as described previously. Response to first-line therapy and survival data were also collected. Survival was analyzed in Cox regressions and stability of the models was examined by bootstrap resampling. The area under the receiver operating characteristics curve (AUC) was calculated to compare the discriminatory ability of each scoring system.

Results

Among the above three inflammation-based prognostic scoring systems, GPS ( P<0.001) and NLR ( P = 0.019) were independently associated with overall survival, which showed to be stable in a bootstrap resampling study. The GPS consistently showed a higher AUC value at 6-month (0.805), 12-month (0.705), and 24-month (0.705) in comparison with NLR and PLR. Further analysis of the association of GPS with progression-free survival showed GPS was also associated independently with progression-free survival ( P<0.001).

Conclusions

Our study demonstrated that the GPS may be of prognostic value in metastatic NPC patients treated with cisplatin-based palliative chemotherapy and facilitate individualized treatment. However a prospective study to validate this prognostic model is still needed.