Toward manipulating sex hormones to modulate the adaptive antitumor immune response in liver cancer.

e16246 Background: The incidence and mortality of hepatocellular carcinoma (HCC) is three to seven times greater in men than women after accounting for known environmental risk factors. This longstanding unexplained observation hints at sex-based biological differences, which are thought to be due to male hormone signaling mediated by the transcription factor androgen receptor (AR). AR is detected in 70% of HCC and increased AR activity corresponds with worse HCC outcomes. While AR inhibition is highly effective for the treatment of prostate cancer, clinical trials to inhibit AR signaling in advanced HCC have been unsuccessful to date. Understanding the role of AR in HCC is expected to inform new therapeutic opportunities for this rapidly expanding patient population for which treatment options are currently limited. Methods: Transgenic animal development, molecular biology, RNA-sequencing, in-vitro and ex-vivo assessments. Results: A lack of human-relevant genetically engineered mouse models (GEMMs) of HCC has hindered progress in this area for decades. Therefore, guided by HCC sequencing databases, we developed a new genetically engineered mouse model of HCC using cre-LoxP technology that recapitulates the sex differences observed in human HCC. We demonstrate that AR inhibition via surgical castration suppresses liver tumors while treatment with dihydrotestosterone (DHT) results in large liver tumors in both male and female mice, making the model an ideal platform to functionally assess the role of AR and male sex hormones in a genetically defined organismal model system of HCC. We generated cell lines from these liver tumors and unexpectedly found that while they express AR, they are not overtly responsive to AR signaling manipulations ex-vivo, suggesting the outcomes associated with active AR in HCC may not be due to AR signaling in the tumor cells themselves. In support of this idea, we found that these tumor cell lines engraft and grow at the same rates in gondadectomized immunocompromised nod-scid gamma (NSG) mice with or without DHT supplementation, suggesting a non-cell-autonomous mechanism may account for the greater tumor burden in male mice. Collectively, these observations guided us in developing our central hypothesis that AR signaling suppresses the adaptive antitumor immune response to accelerate HCC progression via immune evasion. Our ongoing efforts to characterize the impact of hormonal manipulations in HCC are expected to lead to new treatments to complement and extend current standard-of-care immunotherapies for HCC. Conclusions: We developed and phenotypically characterized a new genetically engineered mouse model of HCC that recapitulates the sex differences observed in human HCC. This model will serve as a useful platform to study the role of sex hormones in adaptive antitumor immunity in HCC.