P1.04-14 Early Changes in Plasma CXCL2 and MMP2 Levels Predicts the Response to Anti-PD-1 Therapy in Non-Small-Cell Lung Cancer

Previously we reported that changes in the plasma levels of CXCL2 and MMP2, measured by a bead-based multiplex assay: Bio-Plex 200 system, were significantly associated with the clinical outcomes of anti-PD-1 therapy (Matsuo, et al. IJC. 2018). Here we attempted to validate CXCL2 and MMP2, measured by ELISA, as a marker of the effectiveness of anti-PD-1 therapy in expanded patient cohort. Peripheral blood samples were taken from 97 patients with non-small cell lung cancer before nivolumab or pembrolizumab treatment and after 4-10 weeks from the patients who continued these drugs. The levels of CXCL2 and MMP2 were examined before and after anti-PD-1 therapy. We employed Cox regression analysis for CXCL2 and MMP2 as a single explanatory carriable. In comparing the fitness of CXCL2 and MMP2 Cox models, discrimination was assessed by the Harrell’s C-statistic for survival data. Bootstrap methods with 10000 resamplings were used to assess the stability of the regression analysis predictors. The optimal cutoff point was determined as the point at which the Youden index was maximized by ROC curve. Survival curves were generated using the Kaplan–Meier method and comparisons made using the log-rank test. The changes in the plasma levels of CXCL2 after treatment were significantly correlated with PFS (HR 1.003, 95%CI: 1-1.005, P=0.026) and OS (HR 1.004, 95%CI: 1.001-1.007, P=0.003). The C-statistic of the CXCL2 model for PFS and OS were 0.652 (95% CI: 0.437-0.727) and 0.626 (95% CI: 0.528-0.722), respectively. The decreasing revels of CXCL2 tended to be related to better DCR (P=0.134). The changes in the plasma levels of CXCL2 < 29.1 pg/ml was associated with better PFS (HR 2.872, 95%CI: 1.785-4.618, P<0.001) and OS (HR 2.800, 95%CI: 1.633-4.801, P<0.001). The changes in the plasma levels of MMP2 after treatment were also significantly correlated with PFS (HR 0.998, 95%CI: 0.996-0.999, P=0.003) and OS (HR 0.998, 95%CI: 0.996-0.999, P=0.001). The C-statistic of the MMP2 model for PFS and OS were 0.599 (95% CI: 0.515-0.673) and 0.614 (95% CI: 0.523-0.703). The increasing revels of MMP2 was significantly related to better DCR (P=0.020). The changes in the plasma levels of MMP2 > 0.847 ng/ml was associated with better PFS (HR 0.614, 95%CI: 0.388-0.971, P=0.037) and OS (HR 0.501, 95%CI: 0.295-0.852, P=0.011). The early change of CXCL2 and MMP2 were significantly associated with the clinical outcomes of anti-PD-1 therapy. Since these factors in plasma can be easily measured by minimally invasive method, they could be clinically applicable as biomarkers for predicting the clinical benefit of anti-PD-1 therapy for NSCLC patients.