Clinical usefulness of bone markers in prostate cancer with bone metastasis

Authors


Naoto Kamiya M.D., Department of Urology, Toho University Sakura Medical Center, 564-1 Shimoshizu, Sakura, Chiba 285-8741, Japan. Email: naoto.kamiya@med.toho-u.ac.jp

Abstract

Bone metastases occur in approximately 70% of patients with advanced prostate cancer. Skeletal-related events have been correlated with reduced survival and quality of life of patients with prostate cancer. Biochemical markers of bone metabolism (e.g. bone formation, bone resorption, osteoclastogenesis) might meet an unmet need for useful, non-invasive and sensitive surrogate information for following patients' skeletal health. Recently, zoledronic acid and denosumab have been proven to have the potential for preventing skeletal-related events among prostate cancer patients with bone metastasis. An improved understanding of the mechanisms underlying bone metastasis has also led to the recognition of multiple molecular targets and advances in therapy. However, estimating the efficacy of these agents is difficult. A clinical trial for castration-resistant prostate cancer is currently underway based on the definition of The Prostate Cancer Clinical Trials Working Group, and bone turnover markers are being used as conventional end-points for the clinical trial. Bone turnover markers are useful surrogate markers reflecting the effect of new therapeutic drugs and prognosis, as well as assessment of bone metastases. In particular, N-terminal cross-linked telopeptide of type 1 collagen and bone-specific alkaline phosphatase are widely used bone metabolism markers, and offer reliable surrogate markers to detect bone metastatic spread and to predict prognosis for prostate cancer patients with bone metastases.

Abbreviations & Acronyms
ICTP =

carboxyterminal telopeptide of type I collagen

ADT =

androgen-deprivation therapy

ALP =

alkaline phosphatase

BAP =

bone-specific alkaline phosphatase

BMD =

bone-mineral density

BSP =

bone sialoprotein

CI =

confidence interval

CRPC =

castration-resistant prostate cancer

CT =

computed tomography

CTX =

C-terminal cross-linking telopeptide of type I collagen

DPD =

deoxypyridinoline

EAU =

European Association of Urology

EGF =

epidermal growth factor

EOD =

extent of disease

FGF =

fibroblast growth factor

GR =

grade of recommendation

IGF =

insulin-like growth factor

IL-6 =

interleukin-6

MMP =

matrix metalloproteinase

NCCN =

National Comprehensive Cancer Network

NPV =

negative predictive value

NTX =

N-terminal cross-linked telopeptide of type I collagen

OC =

osteocalcin

OPG =

osteoprotegerin

OPN =

osteopontin

P1CP =

C-terminal propeptide of procollagen type 1

P1NP =

N-terminal propeptide of procollagen type 1

PCa =

prostate cancer

PCWG =

Prostate Cancer Clinical Trial Working Group

PPV =

positive predictive value

PSA =

prostate-specific antigen

PTHrP =

parathyroid hormone-related protein

PYD =

pyridinoline

ROC =

receiver operating characteristic

RANK =

receptor activator of nuclear factor κB

RANKL =

receptor activator of nuclear factor-κB ligand

Src =

sarcoma

SRE =

skeletal-related events

TGF =

transforming growth factor

TRAP =

tartrate-resistant acid phosphatase

TRAP-5b =

tartrate-resistant acid phosphatase isoenzyme 5b

u-NTX =

urinary N-terminal cross-linked telopeptide of type I collagen

VEGF =

vascular endothelial growth factor

ZOL =

zoledronic acid

Introduction

Bone is the most frequent site for metastases in patients with advanced solid tumors, such as prostate, breast, lung, thyroid and renal cancers.1 Approximately 70% of patients with advanced PCa will develop bone metastases.2 ADT remains the first-choice therapy for metastatic PCa and is effective. However, a rapid loss of BMD occurs within the first 6–12 months of ADT.3,4 As a result, ADT causes osteoporosis, and is associated with an increased risk of fracture in patients with PCa.5,6 Metastatic bone disease disrupts the normal homeostasis of bone, which is a dynamic process that involves the coupled balancing of osteoclast-mediated osteolysis and osteoblast-mediated osteogenesis.7 The resulting increased and unbalanced bone metabolism leads to a loss of bone integrity, which can result in skeletal complications. These bone metastases are associated with many complications and much morbidity, including severe bone pain, prolonged hospital stay, reduced mobility, hypercalcemia and pathological fractures.8 Furthermore, SRE have been correlated with reduced overall and median survival and quality of life among patients with PCa.8,9

The assessment of bone metastases relies primarily on imaging techniques, and 99mTc-based bone scintigraphy is routinely used for the detection of bone metastases.10–12 Bone scintigraphy offers high sensitivity, but lacks specificity in the detection of skeletal metastases, and the value of bone scanning for detecting disease progression has been questioned on the basis of cost-effectiveness.13 When serum PSA levels are <20 ng/mL, the likelihood of a positive bone scan in asymptomatic patients is estimated to be just 0.8%.14 In addition, precise diagnosis of bone metastases and judgment of response to treatment after damage has occurred in bone is difficult.

The ideal biomarker of bone metabolism would provide both sensitivity to identify patients with bone metastases or patients at high risk of negative clinical outcomes from bone metastases and specificity in monitoring skeletal health.7 Bone metastases cause osteoclastogenesis and bone resorption, disrupting the balance between osteoblast and osteoclast activity. Bone formation markers are direct or indirect products of osteoblast activity, whereas bone resorption markers are derived from the degradation of skeletal collagen.

The uncertain science and clinical utility of biochemical markers of bone metabolism mean that such approaches have not yet been established as surrogate measurements for clinical efficacy. However, serum biochemical markers can be determined frequently and easily, with negligible disturbance to the patient. Several studies have assessed the diagnostic efficacy of both bone formation and resorption markers for detecting bone metastases in PCa.15–21

ZOL, a nitrogen bisphosphonate, is effective against a variety of cancers with both osteolytic and osteoblastic lesions.22–24 ZOL has a broad clinical indication to prevent SRE among all cancer patients with bone lesions. Furthermore, the clinical efficacy of denosumab, a RANKL antibody, against cancers with bone metastases has also been recently reported.25,26 However, estimating the efficacy of these medicines is difficult. Furthermore, problems exist from the viewpoint of radiation exposure and cost to carry out imaging studies such as bone scintigraphy or CT. Surrogate assessments that are simple, and can rapidly and sensitively detect changes in bone health are thus required, so that serial measurements can be easily taken to follow patient progress. Changes in biochemical markers of bone metabolism might provide valuable information for predicting and monitoring response to therapy.

New guidelines for CRPC have recently been published by the second PCWG.27 The aim of the PCWG2 is to update eligibility and outcome measures in clinical trials that evaluate systemic therapy for patients with progressive PCa and post-castration levels of testosterone. Based on PCWG2 guidelines, Armstrong et al. described bone turnover markers as useful surrogate markers for clinical trial end-points.28 Furthermore, the usefulness of pretreatment serum level of bone turnover markers as a prognostic factor in docetaxel therapy for CRPC has been reported.29,30

This review presents the clinically relevant biomarkers of bone metabolism and the available evidence for their use in PCa with bone metastasis.

Mechanisms of bone remodeling and the bone microenvironment

PCa is characterized by osteoplastic changes, but osteolytic changes are also important. This is supported by biochemical analysis. In other words, cases of PCa with bone metastasis show elevated levels of both serum bone formation and bone resorption markers.

Figure 1 shows the bone microenvironment in bone metastases. Osteoclasts cause bone destruction in osteolytic lesions. These cells acidify the bony environment by discharging H+ and decalcifying hydroxyapatite. Signal transduction of RANK-RANKL is necessary for the differentiation of osteoclasts. RANKL binds to RANK on the cell surface of osteoclasts, and bone resorption is accelerated. OPG is produced by osteoblasts or interstitial cells, and competes with RANK as a decoy receptor for RANKL. OPG binds to RANKL with higher affinity than RANK, inhibiting RANKL activity.31 Bone formation ensues in several steps, which include coupling osteoblasts to the site of recent osteolysis, secretion of matrix proteins and mineralization of the osteoid matrix. Accelerator factors of bone formation, such as TGF-β/FGF and IGF-1 and −2 are produced by PCa cells.32 Osteoblasts are activated by these cytokines. OPN and OC are thought to be produced by activated osteoblasts, and bone formation is promoted. According to a recent study, PSA promotes the growth of osteoblasts and decreases osteoclasts through the apoptotic induction of osteoclast progenitor cells.33

Figure 1.

Bone microenvironment in the presence of bone metastases.

Maintenance and repair of normal bone results in the release of enzymes, peptides and mineral components that have been characterized as serum and urinary biomarkers of bone remodeling.2 Many of these markers are also produced by other tissues, and levels of bone turnover markers could also change in response to fluctuations in other metabolic processes.34 However, in patients with bone metastases, acute changes in these bone turnover marker levels typically indicate alterations in skeletal homeostasis.34 Furthermore, most of the clinically relevant markers currently being investigated are more bone-specific factors.2 Biochemical markers of bone remodeling might thus be an ideal tool to evaluate changes in bone turnover, such as those associated with malignant bone lesions and patient response to treatment.

Role of bone formation and resorption markers

Imaging examinations (e.g. bone scintigraphy, plain X-ray, CT and magnetic resonance imaging) are non-invasive procedures that can be carried out on many sites of bone with a relatively short wait for results. However, these examinations do not show acute changes to bone homeostasis, such as those associated with malignant bone lesions.7 Several biochemical markers of bone formation and resorption with varying specificity and sensitivity are used to predict bone turnover, diagnose metabolic bone disease, and monitor anti-resorptive treatment and effectiveness. We have previously assessed the diagnostic accuracy of serum bone turnover markers for detecting bone metastasis in patients with PCa and to assess the usefulness of these markers as predictors of mortality from PCa (Fig. 2).15 These biochemical markers have a potential use in determining appropriate therapy for patients with bone loss and in monitoring response to therapy. Levels of biochemical markers of bone metabolism can be assessed inexpensively from blood or urine samples, allowing relatively easy monitoring of skeletal health, particularly in the case of acute changes.35 Actually, because bone resorption typically precedes bone formation, bone resorption markers might allow earlier detection of changes in bone metabolism than bone formation markers.

Figure 2.

ROC curves for bone turnover markers to detect bone metastasis in PCa patients. inline image, 1CTP; inline image, ALP; inline image, TRAP-5b; inline image, BAP.

Bone formation markers

Bone formation markers include an enzyme (ALP) and by-products of osteogenesis or osteoblast-secreted factors (OC and amino- and carboxy-terminal procollagen 1 extension peptides).13–20Table 1 shows typical bone formation markers.

Table 1. Markers of bone turnover and osteoclastogenesis
MarkerSpecimenClinical outcome
Bone formation marker
 ALPSerumPresence of bone metastases
Predict prognosis
 BAPSerumPresence of bone metastases
Predict prognosis, SRE
 OCSerumPresence of bone metastases
 P1CPSerumPresence of bone metastases
 P1NPSerumPresence of bone metastases
Predict prognosis
Bone resorption marker
 PYDUrinePresence of bone metastases
 DPDUrinePresence of bone metastases
 CTXSerum or urinePresence of bone metastases
Predict prognosis
 NTXSerum or urinePresence of bone metastases
Predict prognosis
 1CTPSerumPresence of bone metastases
Predict prognosis
 TRAP-5bSerumPresence of bone metastases
Predict prognosis
Osteoclastogenesis marker
 OPGSerumPresence of bone metastases
Predict prognosis
 BSPSerumPresence of bone metastases
Predict prognosis

ALP

Osteoblasts are rich in ALP. However, ALP, as an enzyme associated with the plasma membrane of cells, is also found in the liver, intestine and placenta, all of which might contribute to the total amount of ALP found in blood.36 Although the bone and liver isoenzymes contribute approximately equally to the total, the intestinal fraction accounts for less than 10% in adults.35 ALP is still the parameter most widely used for assessing bone metastases, and is as sensitive as BAP, suggesting that this non-specific bone turnover marker remains a valuable index of bone formation in this condition.37 ALP has been used as a non-specific marker of bone metastasis from PCa since 1936, when Gutman et al. showed that its serum level increased with osteoblastosis.38 It has thus stood the test of time and remains a reliable indicator of osteoblastic activity, as with bone metastases.39 Wymenga et al. commented that patients with newly diagnosed and untreated PCa should undergo bone scintigraphy if there is bone pain or if ALP levels are ≥90 U/L.40 In our previous study, the serum level of ALP was higher in PCa with bone metastases than in patients without bone metastases. Furthermore, a pretreatment group with high levels of ALP showed poorer prognosis than a group with low levels.15 Pretreatment of serum ALP levels also independently predicted overall survival in docetaxel-treated patients and forms a central component of pretreatment nomograms used in CRPC.29,30 However, it must be recognized that ALP also accumulates in the circulation with hepatobiliary disease, blood dyscrasias and heart failure.

BAP

Several isoforms of ALP are secreted by various organs into the blood. The first immunoradiometric assay for BAP incorporating two specific monoclonal antibodies was developed in 1990.41 Under normal conditions, BAP, as an indicator of osteoblast metabolism, represents <40% of serum ALP. BAP is a relatively specific marker for osteogenesis. Serum BAP levels could play a complementary role in diagnosing bone metastases of PCa.42 BAP could provide useful clinical information on the extent of skeletal metastasis and represent an easy way of enhancing the clinical utility of PSA. The addition of BAP to PSA as an initial evaluation could permit a staging bone scan to be avoided for PSA ranges of 10–20 ng/mL, with significant implications for cost-saving. In our study, serum BAP levels were a significant predictor of bone metastasis on univariate analysis. As the EOD on bone scintigraphy score increased, serum levels of BAP increased significantly. Serum BAP levels were also identified as significant independent predictors of cause-specific survival on univariate analysis. In particular, serum BAP levels correlated significantly with incidence of SRE.15 BAP is a reliable and established bone formation marker for PCa with bone metastasis, and the usefulness of BAP has been described in the 2012 EAU guidelines for PCa.43

OC

OC is the major non-collagen protein of bone matrix, and is also known as bone gamma-carboxyglutamate protein and bone gamma-carboxyglutamic acid-containing protein.35 The function of osteocalcin is not clear; it might serve as a site for hydroxyapatite crystals. In the process of matrix synthesis, some OC is released and circulates in blood with a short half-life determined mainly by renal clearance.35 Although no intact OC is released during bone resorption, fragments are released in vitro and also during resorption and formation.44–46 OC can be measured by immunoassay of plasma or serum, but is labile in blood. OC is reduced in lipemic serum because of binding to lipids, and might be degraded in vitro by proteolytic enzymes liberated from erythrocytes.35 Assays for OC are not standardized, and different antibodies clearly recognize different fragments.47–49 Antibodies that recognize both the intact molecule and the large N-terminal midmolecule fragment appear to be the most clinically informative.50

Urinary OC levels might also be assayed, but, because of recovery and degradation, urinary OC levels typically reflect only basal bone turnover rather than acute changes in bone metabolism.51

Procollagen extension peptides

Osteoblasts secrete large procollagen molecules that undergo extracellular cleavage at the amino and carboxy termini.35 Collagen type I comprises approximately 90% of the organic bone matrix.52 Extracellular processing occurs before the final collagen fibril assembly, with the N-terminal (P1NP) and C-terminal (P1CP) regions generated in a 1:1 ratio with collagen and released into the serum.7 Levels of P1NP and P1CP thus might reflect the level of osteogenesis. However, type I collagen is synthesized in some other tissues, which might contribute to serum P1NP and P1CP levels. Serum levels of P1CP have been correlated with bone formation, and decreased levels have been reported after bisphosphonate therapy or hormone replacement therapy. Both P1NP and P1CP are removed by the liver, but P1NP can also be deposited directly into bone and has been found to constitute 5% of the non-collagenous protein in bone. However, recent reports have suggested that P1NP offers greater diagnostic validity than P1CP.53 Positive correlations have been described between elevated serum levels of P1NP, PSA and eventual development of bone metastases in PCa.54 In a separate multivariate Cox analysis, P1NP was identified as an independent predictor of survival in patients with PCa.55

Bone resorption markers

Bone resorption markers include an enzyme, TRAP, and byproducts of osteolysis or osteoclast-secreted factors, which include calcium and bone matrix degradation products, such as hydroxyproline, pyridinium cross-links and telopeptides.15,37,56–61Table 1 shows typical bone resorption markers, and Figure 3 shows the structure of the type 1 collagen.

Figure 3.

Structure of type 1 collagen.

PYD and DPD

PYD and DPD are produced from the post-translational modification of lysine, and hydroxylysine produces the non-reducible pyridinium cross-links. PYD and DPD stabilize mature type 1 collagen in all major connective tissues. During bone resorption, PYD and DPD are released from bone in an approximately 3:1 ratio as free molecules or attached to collagen fragments.7 Urinary excretion is closely related to the rate of bone resorption. Levels of PYD and DPD are not influenced by degradation of newly synthesized collagens or dietary collagen intake, and, although these markers are present in other tissues, bone is the major reservoir and has a higher turnover than most connective tissues.7 Several studies have shown that urinary levels of PYD and DPD from PCa with bone metastases were significantly greater than levels from localized PCa.62–64 However, the contribution from soft tissues might make these markers less accurate than other markers, particularly in the case of PYD.

CTX and NTX

CTX and NTX are the carboxyterminal and aminoterminal peptides, respectively, of mature type I collagen with the cross-links attached, and are released during bone resorption.7 Degradation products of collagen are of various sizes and might undergo additional breakdown in the liver or kidney to constituent amino acids. However, osteoclast-derived fragments differ from those formed in non-skeletal tissues. The cross-linked peptide is primarily attached as an alpha-2 isoform for NTX from bone and as an alpha-1 isoform from other tissues.52 CTX peptide exists in alpha and beta isoforms, with beta isoforms found more often in mature bone.

Assays for NTX utilize an antibody to the alpha-2 chain, which can be conveniently measured in urine or serum.7 However, urinary results must be adjusted for urine dilution, which might add to measurement variability. Urinary CTX measurements have poor precision at concentrations lower than 200 µg/L, so serum or plasma samples are often used. Plasma CTX is more stable, but some anticoagulants interfere with the assay results. Serum CTX measurements utilize an antibody to the beta isoform.7

Serum CTX and NTX levels are significantly increased in PCa patients with bone metastases compared with patients without metastases. Furthermore, patients with levels of CTX and NTX above cut-off levels showed significantly shorter survival than patients with low marker levels.65 Rajpar et al. reported that pretreatment of elevated u-NTX levels offered a strong independent predictor of decreased survival in CRPC patients treated using ZOL.66

Pyridinoline cross-linked 1CTP

Another metabolic product of mature type 1 collagen resorption is 1CTP.7 Immunoassays of serum 1CTP detect the telopeptide portion of the collagen fragment that resides between the two alpha-1 chains.67 Increased levels of serum 1CTP correlate well with bone resorption levels in patients with either high or low bone turnover. Fluctuations in serum 1CTP levels did not correlate with BMD changes in postmenopausal women with osteoporosis who were undergoing bisphosphonate treatment.52 In other words, 1CTP provides a reliable marker for diagnosing bone metastases, because 1CTP is unaffected by osteoporosis.

Several studies have shown that serum 1CTP levels in PCa patients with bone metastasis were significantly higher than those in PCa patients without bone metastasis.68,69 Furthermore, serum 1CTP levels correlated significantly with EOD score, showing a significant downward trend in response to hormonal therapy in PCa patients with bone metastasis. In our study, serum 1CTP level was an independent predictor of bone metastasis according to both uni- and multivariate analyses. Furthermore, as EOD score increased, serum levels of 1CTP increased significantly. Serum 1CTP level was also a significant independent predictor of cause-specific survival according to uni- and multivariate analyses.15 However, clinicians should be aware that 1CTP accumulates in the circulation in patients with renal failure.

TRAP-5b

TRAP-5b is secreted primarily by activated osteoclasts and is one of two isoforms detected in human serum. Activated macrophages secrete the second isoform, TRAP-5a. Osteoclasts secrete the active enzyme after attaching to the bone surface. The enzyme then enters the circulation, where it is inactivated and degraded. Catalytically active enzyme levels in the circulation thus reflect levels of enzyme recently released as a result of bone resorption.7 Halleen et al. were the first to purify TRAP-5b from human osteoclasts, and also detected specific osteoclast-derived TRAP-5b activity from human serum.70,71 Serum TRAP-5b has two important advantages over other bone turnover markers. First, all other known serum bone turnover markers accumulate in the circulation under conditions of renal and hepatic failure, potentially leading to false-positive results, whereas TRAP-5b is unaffected by these problems.72 Second, serum TRAP-5b activity shows less diurnal variability than other serum bone turnover markers, probably because the half-life of other markers is shorter.70 This makes TRAP-5b less sensitive to rapid fluctuations in bone resorption rate. The efficacy of TRAP-5b for cases of PCa with bone metastasis has also been reported, with levels of this marker significantly elevated in PCa patients with bone metastases compared with those without bone metastases.18,20 Furthermore, serum TRAP-5b levels correlated significantly with EOD.18 In our study, univariate analysis showed that serum TRAP-5b levels predicted bone metastasis in PCa. Furthermore, as EOD score increased, serum levels of BAP increased significantly. However, serum TRAP-5b levels were not significant independent predictors of cause-specific survival according to univariate analysis.15

Markers of osteoclastogenesis

The balance between osteoblastic and osteoclastic activity in bone is essentially determined by osteoclastogenesis, which is regulated by three proteins: RANK, RANKL and OPG.

RANKL/OPG

The RANKL/RANK/OPG system represents a key regulatory mechanism in osteoclastogenesis. Surgical biopsy specimens of metastatic PCa have shown significantly higher expression of RANKL and OPG than specimens from non-metastatic on immunohistochemical examination.73,74 Elevated serum levels of either protein alone or increases in the ratio of RANKL to OPG have been investigated as prognostic tools in PCa patients with bone metastases.

Serum OPG levels were found to be increased in patients with bone metastases, compared with localized cancer or lymph node metastases in PCa.75 In contrast, serum RANKL levels did not differ among the control, benign prostatic hyperplasia and PCa subgroups. In our study, serum OPG levels were significantly elevated in patients with bone metastasis and significantly predicted bone metastasis on multivariate stepwise logistic regression analysis. Our ROC analyses also showed serum OPG level as the most reliable predictor of bone metastasis.76 Jung et al. measured levels of 10 serum bone turnover markers in PCa patients with bone metastasis.65 They commented that serum OPG levels showed the best discriminatory power to detect bone metastatic spread and predict survival probability in PCa patients with bone metastasis. Our previous study supports their data, in that serum OPG levels were significantly elevated in patients with bone metastasis, and patients with high serum OPG levels showed significantly poorer outcomes than those with low serum OPG levels according to the Kaplan–Meier method and Cox proportional hazards modeling.76 Mountzios et al. assessed levels of RANKL, OPG, and OPN in patients with breast, lung and prostate cancer with newly diagnosed bone metastases, in parallel with bone resorption and bone formation markers. Patients with breast and lung cancers shared increased levels of RANKL, OPG, and OPN, whereas PCa patients showed elevated levels of OPG and BAP only.77 Based on those findings, PCa patients with bone metastases seem to follow a rather different course of bone turnover compared with patients with other cancers, showing a predominance of bone formation. PCa cells seem to provoke profound elevation of OPG alone. The presence of these crucial bone resorption regulators in PCa bone metastases suggests a mechanism whereby PCa cells might modulate the osteoclastogenesis system, a finding that has profound implications for the establishment and development of PCa bone metastases in advanced disease.

OPG might be useful for predicting bone metastasis and might provide important information for counseling patients regarding their clinical classification and the need for imaging. Furthermore, phase I studies using recombinant OPG in patients with multiple myeloma and patients with breast cancer-related bone metastases have been reported, and future studies will show whether OPG also has therapeutic potential in these areas.78

BSP

BSP is a non-collagenous bone matrix protein secreted by osteoclasts and is part of the small integrin-binding ligand N-linked glycoprotein family. It is secreted by other cells and is present in all mineralized tissues.7 Several studies have assessed the diagnostic efficacy of detecting BSP in bone metastatic disease. However, more recent evidence suggests that in the presence of RANKL, BSP might synergistically induce osteoclastogenesis.79 Furthermore, BSP and RANKL have been shown to exert opposite effects on osteoclast survival and apoptosis. BSP contributed to osteoclast survival and decreased apoptosis, and might also play a role in the nucleation of hydroxyapatite in the bone matrix.80 However, BSP has been found bound to complement factor H in serum, and this complex must be disrupted to accurately measure total BSP levels.81 Serum BSP levels in patients with bone metastases secondary to PCa represent an independent prognostic factor for survival.65

Role of bone turnover marker changes under treatment with ZOL

ZOL has broad clinical indications for the prevention of SRE in all cancer patients with bone lesions. Treatment with ZOL can reduce levels of bone markers, and exploratory analyses have shown strong correlations between reductions in bone marker levels and reduced risks of SRE. ZOL is recommended as a GR-A bone-targeted therapy for CRPC according to the 2012 EAU guidelines for PCa,40 and a category 1 bone-targeted therapy in the 2012 NCCN guidelines for PCa.82 However, clinical efficacy cannot be reliably inferred through assessment of markers alone. Recently, large randomized trials of PCa patients with bone metastases have examined whether ZOL can reduce the risk of SRE.83–85 These trials provide an opportunity for investigating changes in bone turnover markers and clinical outcomes during ZOL therapy.86,87

Coleman et al. showed that patients with high and moderate u-NTX levels showed twofold increases in the risk of skeletal complications and disease progression compared with patients showing low u-NTX levels. High u-NTX levels in each solid tumor category were associated with a four- to sixfold increased risk of death, and moderate u-NTX levels with a two- to fourfold increased risk compared with low u-NTX levels. BAP also showed some correlation with risk of negative clinical outcomes.86 The authors commented that u-NTX provides valuable prognostic information in PCa patients with bone metastases receiving ZOL therapy. Similarly, Lipton et al. showed that normalized NTX within 3 months of ZOL therapy, as opposed to persistently elevated NTX, was associated with reduced risks of skeletal complications and death.88 In contrast, Lein et al. showed the usefulness of serum BAP and 1CTP in PCa patients with bone metastases receiving ZOL therapy. Bone markers, except for 1CTP and ALP, decreased to 20–80% of baseline values at week 12 after drug administration, showing a generally higher decline in the non-SRE group, except for NTX. At all time-points during treatment, higher and increasing concentrations of bone markers were observed in the SRE group compared with the non-SRE group.89 Similarly, Izumi et al. showed the usefulness of serum BAP and 1CTP in PCa patients with bone metastases receiving ZOL therapy. They commented that measurement of serum 1CTP and BAP levels in the early phase after starting ZOL treatment might be useful for physicians to inform patients of their prognosis, and to inform determination of the subsequent treatment plan.90 According to Saad et al., current evidence indicates that NTX and BAP levels might provide a convenient, non-invasive means of assessing response to ZOL therapy at the individual level.87 In brief, assessment of bone turnover markers might complement established diagnostic and monitoring paradigms in PCa with bone metastasis.

Clinical outcomes of bone marker levels in patients receiving new therapeutic drug: anti-RANKL antibody, Src-inhibitor and c-MET/VEGF receptor 2 inhibitor

Denosumab is a human monoclonal antibody that specifically binds to RANKL, a key mediator of osteoclast formation, function and survival. Denosumab is recommended as a GR-A bone-targeted therapy for CRPC in the 2012 EAU guidelines for PCa.43 Denosumab is also recommended as a category 1 bone-targeted therapy in the 2012 NCCN guidelines for PCa.82 Recently, we are able to use denosumab for PCa patients with bone metastases in Japan. Denosumab has recently shown efficacy for the prevention of SRE, and potentially represents a novel treatment option in men with bone metastases from CRPC.91 Fizazi K et al. carried out to randomized double-blind study that compared denosumab with ZOL for prevention of SRE in men with bone metastases from CRPC. Both u-NTX and serum BAP were significantly suppressed in the denosumab arm compared with the ZOL arm.91 A phase III clinical trial comparing denosumab with placebo in PCa patients undergoing ADT provided the first evidence of reduced fracture incidence with an antiresorptive agent in this setting.92 Denosumab also significantly decreased serum levels of bone turnover markers compared with placebo: 6 months after the last dose of study drug, serum levels of CTX, P1NP and TRAP-5b were decreased by medians of 45%, 61% and 33%, respectively, compared with baseline. In contrast, placebo-treated patients recorded increases of 8–18% in bone marker levels compared with baseline.92 Significantly greater decreases in bone turnover markers for denosumab were also seen in subgroup analyses based on age, prior ADT and baseline levels of bone turnover markers. Suppression of bone turnover markers was consistent with the marked increases in BMD reported previously.93

Src also controls normal and abnormal bone physiology, and has been implicated in the development and progression of bone metastases. Dasatinib is a tyrosine kinase inhibitor that potently inhibits Src-family kinases, and shows preclinical activity relevant to PCa. This agent is thus undergoing clinical trials.94,95 Levels of u-NTX were reduced by 40% from baseline in 22 of 43 (51%) CRPC patients receiving dasatinib, and BAP was reduced in 26 of 44 (59%) CRPC patients receiving dasatinib.96 The combination of docetaxel and dasatinib was identified as safe and active, leading to an ongoing phase III trial.97 In this trial, u-NTX was reduced in 33 of 38 (87%) CRPC patients receiving dasatinib, and BAP was reduced in 26 of 33 (76%) CRPC patients receiving dasatinib. Median decreases from baseline (range) in patients with reductions were 36% (6–81%) for u-NTX and 26% (6–85%) for BAP. Among evaluable patients with improved results of bone scans, 12 of 13 (92%) showed a u-NTX decrease, 10 of 12 (83%) had a BAP decrease and 13 of 14 (93%) had a PSA decrease. Among evaluable patients with stable bone scan, u-NTX, BAP or PSA decreases occurred in 15 of 16 (94%), 12 of 13 (92%) and 17 of 19 (90%) patients, respectively.

Cabozantinib is an orally bioavailable novel tyrosine kinase inhibitor of c-MET and VEGF receptor 2, and is currently being tested in a randomized discontinuation study of adult patients with advanced malignant disease.98 A recent abstract described 100 evaluable patients with metastatic CRPC. Effects on osteoclasts and osteoblasts were also observed, with 55% of patients experiencing a ≥50% reduction in u-NTX and 56% of patients experiencing a ≥50% reduction in ALP.99

Conclusions

Although biochemical bone turnover markers provide insights into ongoing rates of bone metabolism, and potentially offer non-invasive techniques for monitoring bone turnover in patients with PCa, these biochemical markers are not sufficiently characterized surrogate measurements to definitively predict clinical outcomes in individual patients. Also, diagnosis of bone metastases from PCa using only biochemical bone turnover markers seems rare at present. However, several biochemical bone marker assessments might provide a useful complement to established techniques for identifying patients at high risk of skeletal involvement and for monitoring progression of bone disease.91 We might avoid the need for frequent CT or bone scintigraphy by using reliable biochemical bone turnover markers. In conclusion, measurement of baseline and serial levels of bone turnover markers provides prognostic information across a wide range of systemic therapies. Ongoing studies are evaluating the utility of changes in bone-labeled level as support for clinical decision-making in the individual patient. These studies seem likely to obtain significant penetration for the use of bone turnover markers to evaluate disease progression and the risk of skeletal morbidity.

Acknowledgment

This work was supported by a Grant-in-Aid from the Ministry of Education, Culture, Sports, Science and Technology (contract grant numbers: 22591761, 23791792).

Conflict of interest

None declared.

Ancillary