Study of the mechanisms behind the additive effect of neoadjuvant castration on radiotherapy for prostate cancer

Abstract

Castration improves responses to radiotherapy (RT) in prostate cancer with unknown mechanism. An understanding of what happens at the cellular and molecular level in prostate cancer cells, while reducing their access to androgens and then exposing them to ionizing radiation (IR), would give us an opportunity to optimize the treatment and may also inspire novel therapeutic approaches.

Paper I: Growth of solid tumours such as prostate cancer is characterized by neovascularization and increased glycolysis as a result of the hypoxic microenvironment of the tumour. HIF-1α is an important transcription factor that regulates cell adaptation to hypoxia and transcription of genes involved in angiogenesis, cell survival, glucose metabolism, and tumour invasion. To test whether any connection between castration therapy and intra-tumoural hypoxia, measured by HIF-1α expression, prostate biopsy specimens from 14 patients with prostate cancer were investigated. Downregulation of HIF-1α expression after castration was observed in five patients with initial high HIF-1α expression. HIF-1α expression was also reduced in two of three patients with initial low HIF-1α expression. These data suggest that neoadjuvant castration reduces tumour cell hypoxia in prostate cancer, which may contribute to the increased radiosensitivity after castration.

Paper II: We investigated whether castration impairs non-homologous end-joining (NHEJ) repair of DNA double-strand breaks (DSBs) by downregulation of Ku70 expression. The same cohort of patients used in paper I was analysed. After castration, the nuclear Ku70 levels were reduced in 12 patients (levels varied from 43% to 97% after castration, p <0.001). The reduction in Ku70 expression correlated significantly with the decrease of serum PSA level after castration, suggesting that AR activity regulates Ku70 protein levels in prostate cancer tissue. Our conclusion was that castration results in decreased levels of Ku70 protein. Since Ku70 protein is necessary for NHEJ repair of DSBs, a downregulation of DNA repair leads to increased radiosensitivity.

Paper III: Emerging data demonstrate homologous recombination (HR) defects in castration resistant malignant prostate tumours, rendering these sensitive to PARP inhibitor treatments. Here, we demonstrate a direct link between the androgen receptor (AR) being required for maintenance of HR gene expression and activity in prostate cancer cells, as well as in maintenance of DNA damage response signalling. As a consequence, we show PARP-mediated backup repair pathway is upregulated in prostate cancer tissues in patients following androgen-deprivation therapy (ADT). Furthermore, upregulation of PARP activity is essential for prostate cancer survival, and we demonstrate a synthetic lethality between ADT and PARP inhibitors in vivo. These data demonstrate that HR may be functionally impaired earlier in prostate cancer etiology as a consequence of ADT; prior to emerging castration resistance and that this potentially can be exploited therapeutically using PARP inhibitors in combination with an ADT upfront in advanced or high risk prostate cancer.

Paper IV: Since castration improves responses to radiotherapy (RT) in prostate cancer, we hypothesized that this radiosensitization is caused by castration-mediated down-regulation of non homologous end joining (NHEJ) repair of DNA double-strand breaks (DSBs). To test this, forty-eight patients with localized prostate cancer were enrolled in two arms, either treating with RT upfront or after receiving neo-adjuvant castration. We biopsied patients at diagnosis, before and after castration and RT treatments to monitor androgen receptor (AR), NHEJ and DSB repair in verified cancer tissue. We show that patients receiving neoadjuvant castration prior to RT had reduced levels of the NHEJ protein Ku70, impaired RTinduced NHEJ activity and a higher level of unrepaired DSBs, measured by γ-H2AX foci in cancer tissues. This study demonstrates that castration impairs NHEJ activity in prostate cancer tissue, explaining improved RT responses in tumours.

Paper V: Despite the early diagnosis and subsequently effective treatment of intermediate and high-risk prostate cancer, the recurrence rate remains regrettably high. Here, we wanted to investigate how effectively castration suppresses androgen receptor (AR) signalling, thereby affecting DNA damage repair in primary hormone-naïve prostate cancer. From the same cohort of patients as in Paper IV, four patients from arm 1 and one patient from arm 2 were analysed. The levels of AR, Ku70, phosphorylated DNA-PKcs and PAR were measured. A significant correlated reduction in the mean intensity of nuclear AR was observed in four patients whose serum PSA was reduced to the greatest extent, about 90% (ρ=1, p <0.001). Although complete castration was obtained using serum testosterone levels in these patients, the levels of AR, and consequently of Ku70 and phosphorylated DNAPKcs remained high and positively co-varied in clusters of cells throughout prostate tumour areas. Meanwhile, a tendency towards an inverse correlation was observed between AR, Ku70 and phosphorylated DNA-PKcs as compared with PARP-1 activity. In conclusion, we are first to demonstrate the heterogeneous landscape of AR and the consequent divergent although co-varied response of DNA damage repair. To date, it remains unclear whether the emergence of these castration-resistant cells in hormone-naïve prostate cancer, is due to the high levels of intra-tumoural residual androgen following castration and consequently suboptimal androgen suppression of otherwise androgen-dependent cells or caused by quiescent castration-resistant cells that promote progression according to clonal selection pressure. The current finding certainly warrants further investigation in the future.