RESUMO
OBJECTIVE: Inflammation plays a crucial role in the initiation and progression of many cancers. This study aimed to investigate the utility of pretreatmentneutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), derived neutrophil-lymphocyte ratio (dNLR), and a combination of PLR and NLR in predicting the risk of death according to clinical staging in cervical cancer (CC) patients. METHODS: A cohort study of women with CC, diagnosed and treated at a single cancer referral center in Brazil, from 2006 to 2009. A multivariate Cox regression analysis and ROC curve analysis accessed the predictive value of inflammatory response biomarkers in overall survival (OS). The median values of the biomarkers were used as cut-off points. RESULTS: A total of 1,266 patients were included in the study, 76.0% with locally advanced disease. After adjusting for clinical variables, NLR > 2.57, PLR ≥ 146.70, dNLR ≥ 1.778 and PLR + NLR in combination had equivalent performance in predicting worse OS, but only among patients with locally advanced disease (adjusted Hazard Ratio [aHR] = 1.453, 95% Confidence Interval [CI] = 1.227-1.722; p < 0.001; aHR = 1.429; 95% CI = 1.209-1.688; p < 0.001; aHR = 1.486, 95% CI = 1.257-1.756, p < 0.001, aHR = 1.731; 95% CI = 1.411-2.123; p < 0.001, respectively). CONCLUSION: In conclusion, PLR, NLR, dNLR and PLR + NLR in combination presented equivalent performance in predicting OS in locally advanced CC patients. They are simple and readily available from routine blood tests, not entailing additional costs. PLR, NLR, dNLR and PLR + NLR in combination are strong prognostic biomarkers candidates in locally advanced CC and should be further explored in prospective trials.
