Abstract
Aim: To assess whether zoledronic acid (ZOL) adds to the effect of combined androgen blockade (CAB) in patients with hormone-naive bone metastatic prostate cancer. Patients and Methods: Patients were treated with either a combination of CAB (luteinizing hormone-releasing hormone agonist and bicalutamide) and ZOL (CAB-Z group) or CAB-alone (historical control patients, CAB-C group). ZOL was injected intravenously at 4 mg every 4 weeks. One hundred and five and 100 patients among 205 enrolled patients were assigned to the CAB-Z group and CAB-C group, respectively. The time to prostate-specific antigen (PSA) failure in patients in the CAB-Z group was compared to that in the CAB-C group. The primary end-point of the study was the time-to-PSA failure. Results: PSA and serum N-telopeptide of type I collagen (NTx) levels were examined before treatment and every 3 months after treatment. PSA failure occurred in 42 (40.0%) patients in the CAB-Z group and 48 (48.0%) patients in the CAB-C group. The biochemical recurrence-free rate was significantly lower in patients in the CAB-C group (p=0.004, by log-rank test). The categorical biopsy Gleason score pre-treatment serum NTx and treatment with ZOL were shown to be independent predictors of PSA failure-free survival time (p=0.040, p=0.005 and p=0.026, respectively). Conclusion: ZOL given with CAB as initial treatment delays the time-to-PSA failure in patients with hormone-naive bone metastatic prostate cancer.
Zoledronic acid (ZOL) received US Food and Drug Administration approval for the treatment of patients with progressive bone metastasis from prostate cancer in 2003 after a phase III placebo-controlled, double-blind trial in 643 patients (1). The clinical end-point was a composite of four skeletal-related events (SREs). ZOL administration resulted in a statistically significant decrease in the proportion of patients with at least 1 SRE (33% vs. 44% for the ZOL and placebo arms, respectively; p=0.021). A statistically significant increase in the median time to first SRE was also observed, favoring ZOL (hazard ratio=0.67; p=0.011) but with no survival benefit.
The overall survival (OS) benefit of ZOL in patients with bone metastasis has been unclear. Recent clinical trials in breast cancer patients showed ZOL to have an antitumor effect (2, 3). Several studies have confirmed the anti-tumor activity of ZOL and combined androgen blockade (CAB) initiated in men with bone-metastatic prostate cancer (4-6).
Malignant bone disease is associated with increased levels of bone resorption marker N-telopeptide of type I collagen (NTx), which correlate with both the presence and extent of bone metastasis. NTx elevation was found to predict a poor outcome in a variety of patients with bone metastases (7-9). ZOL inhibits osteoclast-mediated osteolysis and improves survival in patients with bone metastases from various cancers (10, 11). Decreases in NTx with ZOL administration have been shown to delay the time-to-bone lesion progression. Therefore, NTx may have utility for monitoring therapeutic responses in the bone.
The present study prospectively examined the clinical outcome of ZOL added to CAB in patients with hormone-naive bone metastatic prostate cancer. We compared the results in such patients with the results of CAB-alone in control patients from one Institution. We also assessed the association between NTx levels and clinical outcomes in patients receiving ZOL for bone metastases from hormone-naive bone metastatic prostate cancer.
Materials and Methods
Patients. One hundred and five prostate cancer patients with bone metastasis treated at hospitals in the Tama Bone Metastasis Prostate Cancer group, Japan, between November 2007 and March 2012 were enrolled in the study. None of the patients had undergone radiotherapy or prostatectomy. All cancers were diagnosed histologically by prostatic needle biopsy. All patients received CAB therapy, which is a combination of luteinizing hormone-releasing hormone (LHRH) agonist and bicalutamide. All of the patients received 4 mg ZOL (by intravenous injection) every 4 weeks in addition to CAB therapy (CAB-C group). An additional 100 patients who had received only CAB therapy and were treated at hospitals in the Tama Bone Metastasis Prostate Cancer group between April 1999 and April 2006 (CAB-C group) served as historical controls. Cancer stage was assigned according to the tumor-node-metastasis classification and was based on digital rectal examination, transrectal ultrasonography, computed tomography, magnetic resonance imaging, and bone scintigraphy. The extent of bone metastasis was classified by the extent of disease (EOD) score upon bone scintigraphy according to the method of Soloway et al. (12). An SRE was defined according to Saad et al. as a pathological bone fracture (vertebral or non-vertebral), spinal cord compression, bone surgery, or radiation therapy (including the use of radioisotopes), hypercalcemia of malignancy, or a change in antineoplastic therapy to treat bone pain (1).
The time to prostate-specific antigen (PSA) failure in patients in the CAB-Z group was compared with that in the historical control group (CAB-C group). The primary endpoint of the study was the time to PSA failure. PSA-only failure was defined as three consecutive rises of PSA to >2 ng/ml for 4 weeks after anti-androgen withdrawal despite secondary hormonal manipulation. CAB-C group data were analyzed retrospectively for PSA response and compared with data from the enrolled patients. The secondary endpoint of the study was the causal relationship between the CAB-Z group and the change of bone turnover markers. PSA and serum NTx levels were examined before treatment and every 3 months thereafter. NTx levels were characterized as normal (<17.7 nmol bone collagen equivalents per liter (BCE/l)) or elevated (>17.7 nmol BCE/l).
Patients'clinical characteristics per study group.
This clinical trial was approved by the ethics committees of each hospital in the Tama Bone Metastasis Prostate Cancer group. Statistical analysis. Between-group differences in variables were analyzed by the Mann-Whitney U-test. The chi-square test or Fisher's exact test was used as a test of independence. The biochemical recurrence-free survival rates were calculated by the Kaplan-Meier method and analyzed by the log-rank test. Multivariate analysis was performed using a Cox proportional hazards regression model. A p-value of <0.05 was considered significant. All statistical analyses were performed with SPSS software (version 18.0, link or supplier).
Results
Patients' background. Patients' clinical characteristics are shown per study group (CAB-Z group and CAB-C group) in Table I. Before treatment, there was no significant between-group difference in age, serum PSA level, biopsy Gleason score, EOD score or pre-treatment serum NTx.
PSA failure-free survival in the two groups. Follow-up time was 50.0 months in the CAB-Z group and 97.5 months in the CAB-C group. Kaplan-Meier curves of PSA failure-free survival in the two groups are shown in Figure 1. PSA failure occurred in 42 (40.0%) patients in the CAB-Z group and 48 (48.0%) patients in the CAB-C group. The PSA failure-free survival was significantly shorter in patients in the CAB-C group (p=0.004, by log-rank test). Results of multivariate Cox logistic regression analysis are given in Table II. The categorical biopsy Gleason score pretreatment scrum NTx and treatment with ZOL were shown to be independent predictors of PSA failure-free survival time (p=0.040, p=0.005 and p=0.026, respectively). However, PSA, serum albumin, and alkaline phosphatase (ALP) levels and EOD were not independent predictors (p>0.05).
Kaplan-Meier analysis of PSA failure-free rate in CAB-Z group and CAB-C group. Differences in Kaplan-Meier analysis were tested by the log-rank test. Median PSA failure-free rate in CAB-Z group and CAB-C group was 37.0 and 20.5 months, respectively (p=0.004).
Kaplan-Meier analysis of SREs-free rate in CAB-Z group and CAB-C group. Differences in Kaplan-Meier analysis were tested by the log-rank test. Median skeletal-related events (SREs)-free rate was not reached for patients in CAB-Z group and median SREs-free rate in CAB-Z group and CAB-C group was 60.0 months (p=0.044).
Univariate and multivariate analyses for prediction of PSA failure-free survival.
SRE in the two groups. The SREs were also evaluated. SREs occurred in 18 (18.0%) patients in the CAB-C group and 8 (7.6%) patients in the CAB-Z group. Details of SRE in the CAB-C group were bone radiation therapy (7 patients) and anti-neoplastic therapy to treat bone pain (11 patients). Patients in the CAB-Z group had bone radiation therapy (7 patients) and anti-neoplastic therapy to treat bone pain (1 patient). Statistically significant differences were recognized between the two groups, and it was shown that the occurrence of SREs was delayed by treatment with ZOL (p=0.044) (Figure 2).
Causal relationship between the CAB-Z group and the change of bone turnover markers. Kaplan-Meier estimates of PSA failure-free stratified according to baseline serum NTx levels and NTx levels obtained 3 months after the initiation of therapy are shown in Figure 3. Freedom from PSA failure was compared among four groups of patients: those with <17.7 mmolBCE/l NTx at baseline measurement and 3 months after the start of therapy (group 1, n=21); those with >17.7 mmolBCE/l NTx at baseline measurement but <17.7 mmolBCE/L NTx 3 months after the start of therapy (group 2, n=51); those with <17.7 mmolBCE/l NTx at baseline measurement but >17.7 mmolBCE/l NTx 3 months after the start of therapy (group 3, n=0); and those with >17.7 mmolBCE/l NTx at baseline measurement and 3 months after the start of therapy (group 4, n=33). The survival rates were calculated from the time of the baseline blood draw. The shortest median PSA failure-free period (17.0 months) was seen in group 4, and this failure-free period differed significantly from the median PSA failure-free period of group 1 (28.0 months) (p<0.001), and group 2 (33.5 months) (p<0.001). The difference in failure-free survival time between group 1 and group 2 (p>0.05) was not significant. After 3 months of ZOL treatment, normalized serum NTx correlated with improved PSA failure-free time and persistently elevated serum NTx did not. Moreover, percentage reductions from baseline serum NTx levels correlated with the outcomes regardless of whether patients transitioned from elevated to normal levels.
Kaplan-Meier estimates of PSA failure-free time stratified according to serum NTx levels before (baseline) and 3-months after initiation of therapy. Group 1 (n=21), patients with <17.7 mmolBCE/l NTx before and 3 months after initiation of therapy; group 2 (n=51), patients with >17.7 mmolBCE/l NTx before therapy but <17.7 mmolBCE/l NTX 3 months after initiation of therapy; group 4 (n=33), patients with >17.7 mmolBCE/l NTx before and 3 months after initiation of therapy.
Discussion
ZOL inhibits tumor-associated angiogenesis. ZOL has been shown in vitro to inhibit cancer cell proliferation, adhesion, and invasion, as well as angiogenesis. ZOL has also been shown to indirectly reduce tumor burden by activating γδ T-cells, a T-cell subset that recognizes phosphorylated antigens and possesses anticancer activity (14). In addition, ZOL in combination with docetaxel has shown synergistic cytotoxic or cytostatic effects (14). For example, in patients with metastatic castration-resistant prostate cancer (CRPC) who received androgen blockade, combining ZOL with dexamethasone and octreotide significantly improved median progression-free survival and OS over survival obtained with ZOL alone. In, 45 intermediate-risk patients (23 who received ZOL) with bone metastases from renal cell carcinoma (15), multivariate analysis showed relatively low serum calcium (p=0.0083) and ZOL treatment (p=0.0013) to be independent predictors of prolonged survival. The SRE rate was lower in the ZOL-treated group than in the non-ZOL-treated group (p=0.0453). In addition, OS was significantly longer in this group (p=0.0034). A pilot study involving 40 patients with bone metastasis from bladder cancer showed that ZOL (versus placebo) significantly reduced the risk of SRE by 58% (p=0.0089), reduced the pain score by 1.42 units (p=0.015), and improved the 1-year OS rate 6-fold (p=0.02) (16). Taken together, the results of these studies suggest an antitumor effect of ZOL.
In a large multi-center phase III trial, ZOL (versus placebo) was shown to decrease the risk of SRE at 15 months and to increase the median to 24 months in patients with CRPC (17). Thus, ZOL is recommended for the prevention of skeletal morbidity in patients with CRPC. Most clinical studies investigating ZOL in prostate cancer patients have been performed in patients with CRPC. Few studies of initial ZOL treatment have been conducted in patients with hormone-naive bone metastatic prostate cancer (HSPC). Kamiya et al. examined the effects of ZOL treatment through changes in serum PSA levels in patients treated with ZOL and CAB (4). In patients with prostate cancer and bone metastasis, combination therapy decreased serum PSA and ALP levels more effectively than CAB alone. Nozawa et al. provided essential data on clinical outcomes including SRE-free survival, time to disease progression (TTP), performance status, and safety in patients treated with ZOL and CAB (5). Uemura et al. indicated that the time to nadir PSA in patients given CAB plus ZOL was significantly shorter than that in patients given only CAB (6). Furthermore, significant differences in TTP and OS were recognized between EOD 3+ patients given CAB plus ZOL and those given only CAB. In our study, there was a significant difference in the time-to-PSA failure between the CAB-Z and CAB-C groups. Treatment with ZOL was independent predictor of PSA failure-free survival time. These results suggest the potential anti-tumor activity of ZOL administered as initial treatment in prostate cancer patients with bone metastasis. However, in the high EOD subgroups, there was no significant difference in time-to-PSA failure between the CAB-Z and CAB-C groups (data not shown). Treatment with ZOL for HSPC delayed the occurrence of SREs. OS did not differ significantly between patients treated with CAB-Z and those treated with CAB-C (data not shown), but it is possible that control patients received various therapies after CRPC was diagnosed. In addition, the median follow-up periods of the two groups were not identical. Therefore, further data and careful observations are needed in the future.
Clinical assay of bone tumor biomarkers has become common for patients with bone metastasis. NTx is known as a biomarker of bone tumors and can be assayed in both serum and urine. NTx is a specific marker of bone collagen breakdown (18). In a large study involving patients with prostate cancer and bone metastasis treated with ZOL versus placebo, baseline urine NTx levels were predictive of both SRE and survival (19). Elevated urine NTx is an independent prognostic factor for poor survival in patients with bone metastasis from CRPC. Elevated serum NTx in patients with bone metastases from breast cancer indicates relatively short TTP and OS (20). In our study, after 3 months of ZOL therapy, normalized serum NTx, but not persistently elevated serum NTx, correlated with improved PSA failure-free time. In patients with bone metastasis from HSPC, elevated serum NTx is an independent predictor of PSA failure.
In conclusion, the results of our study indicate the potential anti-tumor activity of ZOL and CAB when used as initial treatment for HSPC; that is, ZOL with CAB delays disease progression. The present study also revealed the importance of bone resorption marker serum NTx as an independent predictor of PSA failure-free time in patients with HSPC.
Footnotes
-
Conflicts of Interest
No Author has any financial disclosures or conflicts of interest to declare.
- Received April 21, 2014.
- Revision received June 16, 2014.
- Accepted June 17, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
