Presentation Authors: Tetsutaro Hayashi, Hiroshima, Japan, Kenichiro Ikeda*, Vancouver, Canada, Roland Seiler, Bern, Switzerland, Robert H Bell, Susan Ettinger, Kendric Wang, Htoo Zarni Oo, Hamidreza Abdi, Wolfgang Jaeger, Tilman Todenhoefer, Colin Collins, Vancouver, Canada, Akio Matsubara, Hiroshima, Japan, Peter C Black, Vancouver, Canada
Introduction: Gemcitabine (GEM) and cisplatin (CDDP) combination chemotherapy (GC) is the standard treatment for patients with advanced bladder cancer (BC), but responses have been reported in only 60% of patients, and these are rarely durable. We aimed to use a genomic analysis to determine mechanisms of resistance to GC.
Methods: Three chemo-sensitive BC cell lines were treated serially with increasing concentrations of CDDP or GEM in order to establish acquired resistance. Gene expression of the resistant cells was compared to the sensitive parental cells. Results were validated in The Cancer Genome Atlas (n=405) and in a patient cohort treated with neoadjuvant GC (n=223). Immunohistochemistry (IHC) was performed in 14 patient tumors before and after neoadjuvant GC and in 37 patients with metastatic BC treated with GC. Correlative in vitro experiments were conducted to explore the mechanism of acquired chemo-resistance.
Results: Gene expression analysis revealed that STAT1 and six interferon-regulated genes were among the most highly up-regulated genes in resistant cells. In the TCGA dataset, STAT1 expression correlated with the expression of the other 6 genes (P < 0.001). Highest STAT1 expression was observed in basal/squamous and luminal infiltrated subtypes. Five-year survival in these patients treated without neoadjuvant GC was 49.7% and 47.7% in tumors with high and low STAT1 expression (compared the median), respectively. In a cohort of patients treated with neoadjuvant GC, the corresponding survival was 62.7% and 78.9%. Nuclear STAT1 expression by IHC was absent in tumors prior to GC but detected in 29% after GC, suggesting that GC activates STAT1 in a subset of patients. In patients with metastatic BC, STAT1 expression was higher in patients with progressive disease (P=0.078) and high STAT1 expression correlated with worse prognosis (P=0.012). Knockdown of STAT1 in resistant cells without CDDP/GEM treatment increased cell growth by cell cycle progression, which was accompanied by increased SKP2 and decreased p27. However, STAT1 knockdown with CDDP/GEM treatment decreased cell growth and increased apoptosis, suggesting that STAT1 silencing restored sensitivity to GC. Conclusions: STAT1 signaling is activated in a subset of BC patients and is associated with acquired chemotherapy resistance. Pending further validation, STAT1 may be considered as potential target in combination with GC, as well as a predictive marker of response to GC.
Conclusions: STAT1 signaling is activated in a subset of BC patients and is associated with acquired chemotherapy resistance. Pending further validation, STAT1 may be considered as potential target in combination with GC, as well as a predictive marker of response to GC.