|
|
||||||||
Articles |
a Address correspondence to this author at: Clinica Medica, Oncologia Medica, Azienda Ospedaliera San Luigi di Orbassano, Regione Gonzole 10, 10043 Orbassano, Italy. Fax 39 11 9026676; e-mail dogliott{at}pasteur.sluigi.unito.it
| Abstract |
|---|
|
|
|---|
Methods: We studied 323 patients with bone metastases from various primary malignancies. We sequentially measured the serum concentrations of bone alkaline phosphatase [by an electrophoretic technique (BALP)], carboxy-terminal telopeptide of type I collagen (ICTP), calcium (CaS), intact parathyroid hormone (PTH), and the fasting urinary excretion of calcium (Ca:Cr). Immunoradiometric serum bone alkaline phosphatase (I-BALP) and urinary excretion of deoxypyridinoline (DPYD) were also assessed in the 175 cases. Data were analyzed as a function of bone pain (assessed by a validated pain questionnaire), the number of radiographically confirmed sites of bone involvement, and the most frequent primary tumor types: breast cancer (BC; 124 patients), prostate cancer (PC; 90 patients), and non-small cell lung cancer (LC; 49 patients).
Results: Serum BALP and I-BALP correlated with the number of radiologically identified blastic bone lesions. BALP and I-BALP were more frequently increased in PC (72% for both measurements) than in BC (50% and 60%, respectively) or LC (3% and 5%, respectively; P <0.001 for BALP and P = 0.001 for I-BALP). ICTP and DPYD values did not differ among PC, BC, and LC, but they did show a direct relationship with the disease extent in bone (P <0.001). CaS and Ca:Cr did not vary significantly according to the bone tumor burden. Bone pain directly correlated with ICTP (P <0.001), DPYD (P = 0.002), CaS (P <0.002), and Ca:Cr (P = 0.001), whereas the relationship was inverse for serum PTH (P = 0.002). When patients were stratified according to the primary tumor, ICTP correlated with the bone pain in all subsets (P <0.005, <0.005, and <0.001 for BC, PC, and LC, respectively), as did CaS and Ca:Cr in LC patients (P = 0.01 and 0.02, respectively) but not in PC and BC patients.
Conclusions: The patterns of bone turnover markers differ among the primary tumor types. Both resorption and formation markers reflect the number of radiographically identified sites of bone metastases, whereas resorption markers and serum calcium but not formation markers correlate with bone pain.© 1999 American Association for Clinical Chemistry
| Introduction |
|---|
|
|
|---|
Severe pain occurs in at least two-thirds of patients and is the symptom of cancer that patients fear most apart from death itself (8). Pain results from stimulation of pain receptors in the periosteum and endosteum, although the underlying mechanisms are poorly understood. Pressure effects of the expanding tumor mass, cytokine and prostaglandin release, or the occurrence of microfractures may all be contributory. Bone pain may be also related to the specific segments affected as well as the involvement of non-bone anatomical structures.
Several studies have reported symptomatic benefits in bone metastatic patients given potent osteoclast inhibitors such as bisphosphonates (9)(10)(11). Any relationship between bone resorption and bone pain, however, still awaits substantial support from studies on large series of patients.
The biochemical markers of bone turnover (12)(13) are admittedly of value in the evaluation of patients with metabolic bone disease, e.g., Paget disease and osteoporosis (14)(15), but their clinical utility in cancer patients whose disease has metastasized to bone is still debated (11)(16)(17)(18)(19)(20)(21)(22).
In the present study involving patients with bone metastases arising from various malignancies, we measured serum bone-specific alkaline phosphatase [with either electrophoretic (BALP) or immunoradiometric (I-BALP) assays (23)(24)], serum carboxy-terminal telopeptide of type I collagen (ICTP) (25), and urinary deoxypyridinoline (DPYD) (26) as markers of osteoblastic and collagen breakdown activities, respectively. The urinary molar ratio of calcium to creatinine (Ca:Cr) in fasting subjects (27), serum calcium (CaS), and serum parathyroid hormone (PTH) were concomitantly measured. The aim of the study was to search for different biochemical patterns in relation not only to primary tumor type but also to painful symptomatology and disease extent.
| Materials and Methods |
|---|
|
|
|---|
|
Diagnosis of bone involvement was performed with bone scans followed by radiological confirmation (x-ray) of hot spots. Computed tomography was performed to discriminate lesions that appeared positive at scintigraphy and negative at x-ray. The whole skeleton was arbitrarily divided into the following areas: skull, cervical, dorsal, lumbar spine, sacrum, right femur, left femur, right humerus, left humerus, right ribs, left ribs, sternum, right pelvis, and left pelvis. Bone pain was evaluated by means of a validated questionnaire according to Coleman (28). The items included the assessment of the performance status, analgesic consumption, and mobility, which produced a pain score of 0 to 16.
According to the disease extent at x-ray evaluation, the patients were
divided as having only one bone site, two or three sites, between four
and six sites, and more than six sites. A further stratification was
performed given a pain score
2, between 3 and 6, between 7 and 10,
and >10. The patient distribution according to the stratifications
performed is outlined in Table 1
.
marker assays
All samples were drawn or collected in the early morning, after an
overnight fast, and included spot urine specimens for measurements of
calcium and creatinine and blood specimens for measurements of BALP,
I-BALP, ICTP, calcium, and PTH. Samples were stored at -70 °C until
assayed. Serum ICTP was measured using a commercially available RIA kit
(FARMOS Diagnostica). The intra- and interassay CVs were ~4% and
5%, respectively. Serum total alkaline phosphatase activity was
measured with a standardized kinetic color test (Merck Diagnostica)
using p-nitrophenyl phosphate as a substrate; the CVs were
always <5% in a whole range of values. BALP was estimated as the
percentage of total alkaline phosphatase after electrophoretic
separation on agarose gel (REP; Helena). The serum I-BALP concentration
was measured with a two-site IRMA (Tandem-OSTASE; Hybritech
Europe); the intra- and interassay CVs were ~4% and 6%,
respectively. Serum PTH (intact molecule) was measured with an IRMA
(Nichols); the intra- and interassay CVs were ~4% and 6%,
respectively. Calcium and creatinine were assessed in serum and urine
with standard automated analyzer techniques.
The CaS concentration was corrected for serum albumin by subtracting or adding a calcium value of 0.02 mmol/L per 1.0 g of albumin above or below 40 g/L albumin. The urinary excretion of calcium was expressed as Ca:Cr (mol/mol). Urinary DPYD was measured in hydrolyzed urine samples with HPLC, using commercial kits (Metra Biosystem) and expressed as the molar ratio to creatinine. The inter- and intraassay CVs were <11% and 12%, respectively.
The reference intervals for the measured analytes were obtained in a group of healthy subjects (115 females and 128 males). All of these subjects were selected as having normal bone mineral density on the lumbar column, evaluated with a dual energy x-ray absorptiometry (Hologic QDR 4500). The upper limits of normal were defined as the means plus 2 SD.
statistical analysis
Differences between groups were tested using the nonparametric
MannWhitney U-test or the KruskalWallis analysis of
variance. Statistical significance was calculated assuming unequal
variance with adjustment for unequal group sizes; P <0.05
was regarded as significant. Differences between proportions were
assessed by the
2 test. All P
values are two-sided. The SPSS package in the MS-DOS version was used
for statistical computation (29).
| Results |
|---|
|
|
|---|
Among the 38 patients who had increased PTH, 4 (3 BC and 1 PC) were hypercalcemic (CaS >2.65 mmol/L) and 24 (16 PC, 7 BC, and 1 urothelial) had increased BALP, normal or decreased CaS, normal or decreased Ca:Cr, and mixed or blastic but not lytic lesions at x-ray. The remaining 10 patients did not show any distinctive feature with respect to patients with PTH within the reference interval.
The concentrations of bone markers in relation to the extent of disease
in bone and bone pain are shown in Table 2
.
|
BALP, I-BALP, ICTP, and DPYD showed a progressive increase with the increase of the appearances in bone, whereas CaS, Ca:Cr, and PTH did not. When patients were stratified according to bone pain, ICTP and DPYD showed a progressive increase with the increase of the bone pain, whereas BALP and I-BALP did not parallel the ICTP pattern. The concentrations of BALP and I-BALP were, in fact, comparable in patients with different pain scores. CaS and Ca:Cr progressively increased with the increase in bone pain, whereas PTH progressively decreased.
patients stratified according to primary malignancy
In the BC subset, lytic appearances were observed in 86 patients
(69%), whereas 37 (30%) had mixed and 1 (1%) had blastic lesions. In
the PC subset, blastic appearances, observed in 53 patients (59%),
prevailed over mixed and lytic lesions, which were recorded in 29
(32%) and 8 cases (9%), respectively. In the LC subset, 47 patients
(96%) had lytic lesions, and 2 (4%) had mixed metastases.
The concentration ranges and medians of bone turnover markers are
outlined in Fig. 1
. BALP and
I-BALP progressively increased in patients with LC, BC, and PC. The
median PTH concentration was significantly lower in LC patients with
respect to BC and PC patients. Serum ICTP, urinary DPYD, urinary Ca:Cr,
and CaS did not change according to different primary tumors. LC
patients suffered from greater pain in the face of the lower extent of
disease in the bone.
|
In the PC subset, 72% of patients had BALP or I-BALP concentrations above the reference intervals, 66% and 63% had ICTP and DPYD concentrations above the reference intervals, 11% and 14% had CaS and Ca:Cr values above the reference intervals, and 21% had PTH concentrations above the reference interval. In the BC subset, increased BALP and I-BALP were observed in 43% and 60% of the patients, whereas the 69% of patients had increased ICTP, 54% had increased DPYD, 12% had increased CaS, 24% had increased Ca:Cr, and 10% had increased PTH. In LC patients, the percentages of patients with values above the reference intervals were 3% for BALP, 5% for I-BALP, 63% for ICTP, 47% for DPYD, 21% for CaS, and 20% for Ca:Cr. In this subset, no case had PTH values above the reference interval.
About one-half of the hypercalcemic BC and LC patients and only one PC patient were symptomatic. Hypocalcemia was observed in 11% of PC cases, 6% of BC cases, and 4% LC cases. Serum PTH <20 µg/L was recorded in 14% of patients with PC, 18% with BC and 38% with LC.
The relationship between bone turnover markers and either the extent of disease in bone or bone pain was also separately evaluated in the patient subsets. The low number of patients with DPYD and I-BALP measurements precluded comparisons for these analytes. ICTP was higher in patients with a greater tumor burden in PC (P <0.001) and BC (P <0.001) subsets. It also paralleled the extent of disease in LC patients without attaining statistical significance. BALP increased significantly with the increase of bone involvement in PC (P <0.001) and BC (P <0.001) patients but not in LC ones. Ca:Cr showed a tendency to increase with the increase of tumor load in LC and BC cases and a tendency to decrease in PC cases. In all subsets, PTH did not show appreciable changes as a function of the extent of disease.
BALP values paralleled bone pain in PC patients (P =
0.05) but not in BC patients, whereas ICTP serum concentrations showed
a significant increase with an increasing pain index in PC and BC
subsets (P <0.005). Ca:Cr (but not CaS and PTH) increased
with the increase of bone pain in BC cases (P = 0.02).
No relationship was observed between bone pain and CaS, Ca:Cr, and PTH
in PC cases. In LC patients, as depicted in Table 3
, ICTP, Ca:Cr, and CaS correlated with the pain score, whereas
BALP did not. In addition, a trend toward a reduction of PTH values
with pain increase was apparent.
|
| Discussion |
|---|
|
|
|---|
In the present study, bone collagen breakdown was evaluated by serum ICTP in all patients. Several reports have suggested ICTP as a good indicator of the progression of bone metastases and of the disease response to systemic treatment (17)(20)(31)(32)(33). Because urinary DPYD was more specific than serum ICTP (34), this analyte was additionally measured in ~50% of cases. ICTP and DPYD concentrations were comparable in patients with lung, breast, and prostate primary malignancies. This observation is consistent with the occurrence of remarkable osteoclast activation irrespective of the primary source of bone metastases.
The positivity of the bone scan in all cases indicated that osteoblast activity was enhanced at the sites of metastases. However, biochemical evidence was attained in a subset of patients unevenly distributed according to the primary malignancy. As reported previously (35), BALP and I-BALP paralleled the number of osteoblast deposits seen at radiology. These markers frequently exceeded the reference intervals in PC and BC patients, who had x-ray appearances often associated with a marked osteoblast reaction, but were increased only slightly in LC cases, who very frequently had "pure" lytic lesions at x-ray examination.
Notwithstanding the general correlation between bone turnover markers and the tumor load in bone, BALP, I-BALP, ICTP, and DPYD varied within a wide range in individual patients with similar disease extent. It is conceivable that tumor-produced paracrine factors, which account for enhanced bone resorption and formation (3)(6), vary to a great extent not only among patients with different cancers but also among patients bearing the same primary malignancy.
The prevalence of hypercalcemia in the present series was higher than that observed previously (36). The greater frequency of hypercalcemia in LC patients with respect to BC and PC patients confirms the results of other studies (36)(37). The most plausible explanation for hypercalcemia in these patients is the production of PTHrP (38). Unfortunately, we were not able to obtain reliable data on serum PTHrP measured retrospectively on frozen specimens. Interestingly enough, hypercalcemia in PC cases was as frequent as that of BC cases, at variance with previous findings (36)(37). It should be noted that ~90% of PC patients had progressive disease to androgen suppression; to our knowledge no data have been published on the occurrence of hypercalcemia in this particular population.
In the present series, a nonnegligible percentage of patients had increased PTH. Some of these patients had concomitantly increased CaS and Ca:Cr; hence, they could be diagnosed as having primary hyperparathyroidism (39). The majority of cases, however, had low values of both CaS and Ca:Cr; all had mixed or blastic bone appearances. These features appear compatible with the so-called "bone hunger syndrome" that has been repeatedly described in PC and is characterized by secondary hyperparathyroidism attributable to calcium entrapment in bone as a consequence of an exceedingly high osteoblast activity (35)(40)(41). Our findings suggest that this metabolic derangement also occurs in some BC patients. In the present series, serum and urinary collagen breakdown products were proportional to the extension of disease in bone, as we suggested previously (35).
When overall patients were stratified according to bone pain, not only ICTP and DPYD, but also CaS and Ca:Cr showed a progressive increase with the increase in pain, whereas serum PTH showed an inverse correlation. When patients were further stratified according to the primary malignancy, serum ICTP correlated with bone pain in all subsets, but the relationship between CaS and bone pain was apparent only in LC patients and not in BC and PC ones.
Induction of hypercalcemia is related to osteoclast activation, but its
maintenance frequently involves increased tubular reabsorption
(42). The relationship between bone pain and CaS suggests
that mediators of hypercalcemia (such as PTHrP and tumor growth
factor-
), which influence renal calcium absorption
(5)(43), also play a role in causing pain,
possibly by stimulating prostaglandin production in bone
(43). As an additional consideration, our data could, on the
one hand, lend support to the use of bisphosphonates for controlling
bone pain
(9)(10)(44)(45); on the
other hand, it may suggest a mechanism for individual refractoriness to
these drugs, which notoriously do not affect the release of soluble
factors such as PTHrP (46)(47)(48).
In conclusion, this study shows differential patterns of bone turnover markers in cancer patients with bone metastases as a function not only of the primary malignancy but also of the tumor load in the skeleton and the pain. Both resorption and formation markers are of value if one considers the number of bone metastases. Resorption markers, but not formation markers, correlate with bone pain. The relationship between bone pain and CaS is a novel finding that suggests a contribution of factors responsible for tubular calcium reabsorption in the pathogenesis of painful bone events.
| Acknowledgments |
|---|
| Footnotes |
|---|
1 Nonstandard abbreviations: PTHrP, parathyroid hormone-related protein; BC, breast cancer; LC, lung cancer; PC, prostate cancer; BALP, bone-specific alkaline phosphatase measured with electrophoretic assays; I-BALP, bone-specific alkaline phosphatase measured with immunoradiometric assays; ICTP, carboxy-terminal telopeptide of type I collagen; DPYD, deoxypyridinoline; Ca:Cr, urinary ratio calcium to creatinine; CaS, serum calcium; and PTH, parathyroid hormone. ![]()
| References |
|---|
|
|
|---|
The following articles in journals at HighWire Press have cited this article:
![]() |
G. G. Schwartz Prostate Cancer, Serum Parathyroid Hormone, and the Progression of Skeletal Metastases Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 478 - 483. [Abstract] [Full Text] [PDF] |
||||
![]() |
J-J Body, M Lichinitser, S Tjulandin, P Garnero, and B Bergstrom Oral ibandronate is as active as intravenous zoledronic acid for reducing bone turnover markers in women with breast cancer and bone metastases Ann. Onc., July 1, 2007; 18(7): 1165 - 1171. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. J. Leeming, M. Koizumi, I. Byrjalsen, B. Li, P. Qvist, and L. B. Tanko The Relative Use of Eight Collagenous and Noncollagenous Markers for Diagnosis of Skeletal Metastases in Breast, Prostate, or Lung Cancer Patients Cancer Epidemiol. Biomarkers Prev., January 1, 2006; 15(1): 32 - 38. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. E. Coleman, P. Major, A. Lipton, J. E. Brown, K.-A. Lee, M. Smith, F. Saad, M. Zheng, Y. J. Hei, J. Seaman, et al. Predictive Value of Bone Resorption and Formation Markers in Cancer Patients With Bone Metastases Receiving the Bisphosphonate Zoledronic Acid J. Clin. Oncol., August 1, 2005; 23(22): 4925 - 4935. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. E. Brown, R. J. Cook, P. Major, A. Lipton, F. Saad, M. Smith, K.-A. Lee, M. Zheng, Y.-J. Hei, and R. E. Coleman Bone Turnover Markers as Predictors of Skeletal Complications in Prostate Cancer, Lung Cancer, and Other Solid Tumors J Natl Cancer Inst, January 5, 2005; 97(1): 59 - 69. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. Fohr, C. R. Dunstan, and M. J. Seibel Markers of Bone Remodeling in Metastatic Bone Disease J. Clin. Endocrinol. Metab., November 1, 2003; 88(11): 5059 - 5075. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. M. Demers Biochemical Markers in the Management of Patients with Metastatic Bone Disease Clin. Chem., August 1, 1999; 45(8): 1131 - 1132. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |