LAUSR

decreased in 27 cases compared to baseline. SLD decreased in 9 cases (17.0%) in high pre-AMC group, which was significantly lower than that in low pre-AMC group (18 cases, 48.6%, p[0.001). The disease control rate defined by RECIST ver. 1.1 at best response in high pre-AMC group was 30.6%, which was significantly lower than that in their counterpart (61.1%, p[0.003). No association was observed between pre-AMC levels and the occurrence of adverse events (p[0.184). The 12-month progression-free survival (PFS) rate for high pre-AMC group was 11.8%, which was significantly lower than that for their counterpart (41.4%, p<0.001). Multivariate Cox regression analysis revealed that pre-AMC level of >342 (p[0.007), and liver metastases (p[0.028) were the independent indicators for disease progression. Furthermore, the 12-month cancer-specific survival (CSS) rates for high pre-AMC group was 45.9%, which was significantly lower than that for their counterpart (89.7%, p<0.001). Multivariate analysis revealed that the pre-AMC level of >342 was the only independent indicator for cancer-specific death (p<0.001). The change of AMC level before and after pembrolizumab was elevated in 41 cases (45.6%). However, there was no association between clinical outcome of PFS and CSS and the change of AMC level. CONCLUSIONS: Elevated pre-AMC could identify a population with a poor response to pembrolizumab treatment among chemoresistant UC patients.