Background: The metastasis of prostate cancer to bone is associated with a substantial increase in bone matrix turnover. Matrix metalloproteinases (MMPs) play roles in both normal bone remodeling and invasion and metastasis of prostate cancer. This study was designed to determine the role of MMP activity in prostate cancer that has metastasized to bone. Methods: Single human fetal bone fragments were implanted subcutaneously in immunodeficient mice. Four weeks later, PC3 human prostate cancer cells were injected directly into some of the implants, and daily treatment was begun with batimastat (a broad-spectrum MMP inhibitor). There were six mice (i.e., six implants) in each of four experimental arms: bone alone with and without batimastat and bone injected with PC3 cells with and without batimastat. Bone implants were harvested after 14 days of treatment and analyzed for MMP expression, bone histomorphometry, osteoclast counts, blood vessel density, and tumor cell proliferation and apoptosis. Complementary data were obtained from bone biopsy samples from patients and a bone organ coculture system. All statistical tests were two-sided. Results: MMPs were detected in tumor and stromal cells of clinical specimens and experimental bone implants. In vivo, MMP inhibition reduced the number of osteoclasts per millimeter in PC3-injected implants—from 8.2 (95% confidence interval [CI] = 7.9 to 8.5) to 3.0 (95% CI = 2.3 to 3.7) (P = .006). In addition, it prevented degradation of marrow trabeculae within the bone implants (cross-sectional area of implant occupied by mineralized trabeculae: untreated implant = 29.1% [95% CI = 27.1% to 31.1%], PC3-injected implant = 14.0% [95% CI = 10.9% to 17.1%] [P = .005 versus untreated], and batimastat-treated PC3-injected implant = 27.2% [95% CI = 22.4% to 32.0%] [P = .03 versus PC3 injected alone]). MMP inhibition reduced proliferating tumor cells from 20.8% (95% CI = 19.9% to 21.7%) to 7.4% (95% CI = 5.2% to 9.6%) (P = .006), without affecting angiogenesis or apoptosis. In vitro, MMP inhibition had no toxic effect on PC3 cells but prevented calcium release from bone fragments cocultured with PC3 cells. Conclusions: MMP activity appears to play an important role in bone matrix turnover when prostate cancer cells are present in bone. Bone matrix turnover and metastatic tumor growth appear to be involved in a mutually supportive cycle that is disrupted by MMP inhibition.

Introduction: The chemokine CXCL12, also known as SDF-1, and its receptor, CXCR4, are overexpressed in prostate cancers and in animal models of prostate-specific PTEN deletion, but their regulation is poorly understood. Loss of the tumor suppressor PTEN (phosphatase and tensin homolog) is frequently observed in cancer, resulting in the deregulation of cell survival, growth, and proliferation. We hypothesize that loss of PTEN and subsequent activation of Akt, frequent occurrences in prostate cancer, regulate the CXCL12/CXCR4 signaling axis in tumor growth and bone metastasis. Methods: Murine prostate epithelial cells from PTEN^+/+, PTEN ^+/− , and PTEN^−/− (prostate specific knockdown) mice as well as human prostate cancer cell lines C4-2B, PC3, and DU145 were used in gene expression and invasion studies with Akt inhibition. Additionally, HA-tagged Akt1 was overexpressed in DU145, and tumor growth in subcutaneous and intra-tibia bone metastasis models were analyzed. Results: Loss of PTEN resulted in increased expression of CXCR4 and CXCL12 and Akt inhibition reversed expression and cellular invasion. These results suggest that loss of PTEN may play a key role in the regulation of this chemokine activity in prostate cancer. Overexpression of Akt1 in DU145 resulted in increased CXCR4 expression, as well as increased proliferation and cell cycle progression. Subcutaneous injection of these cells also resulted in increased tumor growth as compared to neo controls. Akt1 overexpression reversed the osteosclerotic phenotype associated with DU145 cells to an osteolytic phenotype and enhanced intra-osseous tumor growth. Conclusions: These results suggest the basis for activation of CXCL12 signaling through CXCR4 in prostate cancer driven by the loss of PTEN and subsequent activation of Akt. Akt1-associated CXCL12/CXCR4 signaling promotes tumor growth, suggesting that Akt inhibitors may potentially be employed as anticancer agents to target expansion of PC bone metastases.

The CXCL12/CXCR4 axis transactivates HER2 and promotes intraosseous tumor growth. To further explore the transactivation of HER2 by CXCL12, we investigated the role of small GTP protein Gαi2 in Src and HER2 phosphorylation in lipid raft membrane microdomains and the significance of CXCR4 in prostate cancer bone tumor growth. We used a variety of methods such as lipid raft isolation, invasion assays, an in vivo model of intratibial tumor growth, bone histomorphometry, and immunohistochemistry to determine the role of CXCR4 signaling in lipid raft membrane microdomains and effects of targeting of CXCR4 for bone tumor growth. We determined that (a) CXCL12/CXCR4 transactivation of EGFR and HER2 is confined to lipid raft membrane microdomains, (b) CXCL12 activation of HER2 and Src is mediated by small GTP proteins in lipid rafts, (c) inhibition of the CXCL12/CXCR4 axis through plerixafor abrogates the initial establishment of tumor growth without affecting the growth of established bone tumors, and (d) inhibition of EGFR signaling through gefitinib leads to inhibition of established bone tumor growth. These data suggest that lipid raft membrane microdomains are key sites for CXCL12/CXCR4 transactivation of HER2 via small GTP binding protein Gαi2 and Src kinase. The initial establishment of prostate cancer is supported by the endosteal niche, and blocking the CXCL12/CXCR4 axis of this niche along with its downstream signaling severely compromises initial establishment of tumors in the bone microenvironment, whereas expanding bone tumors are sensitive only to the members of growth factor receptor inhibition.

The adoption of active surveillance varies widely across urological communities, which suggests a need for more consistency in the counseling of patients. To address this need we used the RAND/UCLA Appropriateness Method to develop appropriateness criteria and counseling statements for active surveillance. Panelists were recruited from MUSIC urology practices. Combinations of parameters thought to influence decision making were used to create and score 160 theoretical clinical scenarios for appropriateness of active surveillance. Recent rates of active surveillance among real patients across the state were assessed using the MUSIC registry. Low volume Gleason 6 was deemed highly appropriate for active surveillance whereas high volume Gleason 6 and low volume Gleason 3+4 were deemed appropriate to uncertain. No scenario was deemed inappropriate or highly inappropriate. Prostate specific antigen density, race and life expectancy impacted scores for intermediate and high volume Gleason 6 and low volume Gleason 3+4. The greatest degree of score dispersion (disagreement) occurred in scenarios with long life expectancy, high volume Gleason 6 and low volume Gleason 3+4. Recent rates of active surveillance use among real patients ranged from 0% to 100% at the provider level for low or intermediate biopsy volume Gleason 6, demonstrating a clear opportunity for quality improvement. By virtue of this work urologists have the opportunity to present specific recommendations from the panel to their individual patients. Community-wide efforts aimed at increasing rates of active surveillance and reducing practice and physician level variation in the choice of active surveillance vs treatment are warranted.

Why does prostate cancer metastasize to bone? Why is there increased turnover of the bone matrix in the presence of prostate cancer? A recent autopsy study supports a traditional hypothesis that gross, anatomic patterns of blood flow influence the distribution of metastatic deposits. On the other hand, recent developments in animal models of prostate cancer bone metastasis have renewed interest in the traditional ‘seed and soil’ hypothesis: several studies point to specific biological interactions between prostate cancer cells and the bone microenvironment that can explain the tendency of prostate cancer cells to colonize bone and grow into clinically relevant metastatic deposits. Some studies implicate mechanisms including chemoattraction and enhanced adherence to bone endothelium. Additional data suggest that prostate cancer cells are ‘osteomimetic’, that is, they take on the properties and behaviors of osteoblasts or osteoclasts upon arrival in bone. These activities lead to enhanced turnover of the bone matrix and may explain the propensity of prostate cancer to grow in bone. Finally, a series of studies have implicated other molecular markers as distinguishing characteristics of bone-metastatic prostate cancer tissue.