PS02.207: NEUTROPHIL-LYMPHOCYTE RATIO HOLDS NO PROGNOSTIC VALUE FOR OESOPHAGEAL AND JUNCTIONAL ADENOCARCINOMA IN PATIENTS TREATED WITH NEOADJUVANT CHEMOTHERAPY.

Background Several studies have been suggesting that neutrophil-lymphocyte ratio (NLR), as it reflects systemic inflammation, could help predict survival in oesophageal and junctional carcinomas. Therefore, we aimed to determine whether baseline NLR holds prognostic and predictive value in oesophageal and junctional adenocarcinomas (OAC) for patients treated with neoadjuvant chemotherapy (nCT) followed by surgery. Methods We studied the data of 144 included patients that received nCT, all identified from a prospectively maintained database. Pre-treatment haematology reports were used to calculate the baseline NLR, dividing absolute neutrophil count by absolute lymphocyte count. Multiple ways of grouping patients based on NLR were tried, including determining the optimal cut-off value based off a ROC-curve and the standard threshold for elevated NLR (> 5). NLR quartiles were used to display possible differences between groups in relation to overall survival (OS), disease-free survival (DFS) and pathological response according to Mandard score. Cox regression analysis was performed to determine independent prognostic factors for OS. Results The ROC-curve showed that NLR has no discriminating power for survival status (area under the curve = 0.460) and therefore no optimal cut-off value could be determined. Also, using the most frequently used threshold for elevated NLR (≥ 5) to group patients did not lead to a difference in OS (P = 0.112). Median OS times for NLR quartiles were 65 (Q1), 32 (Q2), 45 (Q3) and 46 months (Q4), with no significant difference (P = 0.926). DFS showed no difference between groups either, with median DFS times of 30 (Q1), 22 (Q2), 38 (Q3) and 23 months (Q4, P = 0.973). Pathological response according to Mandard score did not vary between NLR quartiles (P = 0.925). In addition, NLR was not associated with OS in univariate analysis (P = 0.518). Multivariate analysis showed that both pathological N- and M-stage, and number of involved nodes were independent prognostic indicators for OS. Conclusion The present study shows that, in contrast to other recently published papers, baseline NLR holds no prognostic or predictive value for OAC patients treated with nCT. This result strongly questions the validity of NLR as a prognostic indicator and its clinical usefulness. Disclosure All authors have declared no conflicts of interest. [ABSTRACT FROM AUTHOR]