- Email: [email protected]
Comparison of the analytical and clinical performance characteristics of an N-MID versus an intact osteocalcin immunoradiometric assay
Clinica Chimica Acta 294 (2000) 67–76 www.elsevier.com / locate / clinchim
Comparison of the analytical and clinical performance characteristics of an N-MID versus an intact osteocalcin immunoradiometric assay a, a a Masaaki Takahashi *, Kazuhiro Kushida , Akira Nagano , b Tetsuo Inoue a
Department of Orthopaedic Surgery, Hamamatsu University School of Medicine, 3600 Handa, Hamamatsu 431 -3192, Japan b Aoyama General Hospital, Hoi-Gun, Kosakai-Cho 441 -0195, Japan
Received 29 July 1999; received in revised form 23 November 1999; accepted 29 November 1999
Abstract Osteocalcin is the most abundant non-collagenous protein in bone, reflecting its formation. It was reported that the instability of intact osteocalcin results from the cleavage of the C-terminal sequence of osteocalcin to produce a large N-terminal osteocalcin fragment. A two-site immunoassay for detecting both the N-terminal osteocalcin fragment and the intact osteocalcin was developed that were both independent of an unstable C-terminal sequence. The aim of this study is to investigate the performance of an N-MID osteocalcin immunoradiometric assay and to compare it with an intact osteocalcin assay. Ten serum samples were repeatedly frozen and thawed up to seven times. The variability of the values of N-MID osteocalcin was less than that of the intact osteocalcin. For stability of osteocalcin in serum after storage, the mean value of N-MID was 94.3% of the initial value after 7 days at 48C, whereas the intact was 73.4%. The reduction of intact values were significantly larger than that of N-MID after 2, 5 and 7 days. At 2 308C, the values of N-MID did not change for up to 10 weeks. The concentrations of osteocalcin measured by an N-MID osteocalcin and an intact osteocalcin were investigated in 27 premenopausal subjects, 27 postmenopausal subjects, and 68 osteoporotic patients (23 with vertebral fractures and 45 with hip fractures). The percent mean increase of osteocalcin in postmenopausal subjects over premenopausal subjects was 98% in N-MID versus 42% in the intact assay. The z-scores of N-MID and intact showed similar results in all groups. N-MID osteocalcin significantly correlated with intact osteocalcin (r 5 0.755), and other biochemical markers for bone formation, such as
*Corresponding author. Tel.: 181-53-435-2299; fax: 181-53-435-2296. E-mail address: [email protected] (M. Takahashi) 0009-8981 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 99 )00251-X
68
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
bone specific alkaline phosphatase (r 5 0.606) and C-terminal propeptide of type I procollagen (PICP) (r 5 0.568). An N-MID IRMA had better stability during storage than intact and had the discriminative ability which is similar to the intact assay in postmenopause and osteoporosis. Therefore, an N-MID osteocalcin IRMA could improve the clinical utility and evaluation of osteocalcin. 2000 Elsevier Science B.V. All rights reserved. Keywords: Osteocalcin; Bone marker; Bone formation; Vertebral fracture; Hip fracture
1. Introduction In the past decade, markers for bone formation have been developed. Osteocalcin, also called bone gla protein (BGP), a 49 amino acid protein, is the most abundant non-collagenous protein in bone [1]. Although the biological function of osteocalcin is still unknown, osteocalcin is synthesized by osteoblast and is only located in the bone. Evidence has shown that it reflects osteoblastic activity and bone formation [2,3]. Although most of the newly synthesized osteocalcin is incorporated into the bone matrix, some fraction is released into blood circulation [4]. Since circulating osteocalcin concentrations were initially measured using conventional radioimmunoassays (RIAs), a variety of immunometric assays, using polyclonal or monoclonal antibodies, have been developed with better analytical sensitivity and specificity than conventional RIAs. However, those immunoassays have given discordant results in diverse clinical settings [5,6]. This discordance is caused by epitopic specificity and differential reactivity with circulating fragments of osteocalcin [7,8]. To overcome this problem, two-site immunoassays for the measurement of the intact molecule of osteocalcin were developed [9]. However, the assays for intact osteocalcin have not been any better than the conventional assays. It was reported that one third of circulating osteocalcin was intact osteocalcin, one third was a large Nterminal osteocalcin fragment and the rest were various smaller fragments [10]. The instability of intact osteocalcin is mainly caused by the cleavage of the C-terminal sequence of osteocalcin, producing a large N-terminal midfragment [10]. Accordingly, intact molecules of osteocalcin in vitro such as, blood, serum and plasma, are thought to be unstable. Therefore, use of these assays requires precise control of temperature, duration of storage, and the avoidance of freezing and thawing. A two-site enzyme-immunoassay for detecting both the large N-terminal osteocalcin and intact osteocalcin was developed to be independent of an unstable C-terminal sequence. The performance of this enzyme-immunoassay have been reported [11]. The aim of this study is to investigate the analytical and clinical performance of an N-MID osteocalcin immunoradiometric assay (IRMA).
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
69
2. Materials and methods
2.1. Subjects All subjects were females. The premenopausal group (PRE) consisted of 27 women aged 32–51 years (mean ages 6S.D.; 42.664.2 years) who had regular menstrual cycles. The postmenopausal group (POST) consisted of 21 postmenopausal females aged 51–90 years (64.069.7). No subjects had any previous history of any disease known to affect bone metabolism (such as hyperthyroidism, renal disease, collagen disease or ovarian tumor), and had not previously used estrogen or any other hormonal medications. All were volunteers and gave their informed consent prior to the study. The vertebral fracture group (VX) consisted of 23 patients aged 51–92 years (mean ages6S.D.; 73.16 8.5 years) diagnosed by X-ray films of the thoracic and lumbar spine, which deformed more than 20% reduction in anterior, middle, and / or posterior height as vertebral fracture [12]. There were no fractures sustained within the prior 6 months. The hip fracture group (HX) consisted of 45 patients aged 62–90 years (mean ages 6S.D.; 77.268.0 years). None of the patients in either group had been receiving any medications that might have affected calcium metabolism during the study. Secondary osteoporosis patients were excluded from the study.
2.2. Sample collection Blood was kept at room temperature, centrifuged within 1 h, and sera was stored at 2 808C until use. Prior to analysis, the frozen samples were thawed and the measurements were initiated within 1 h after thawing. For sera in the hip fracture group, samples were collected within 48 h after fracture [13].
2.3. Measurements of N-MID osteocalcin IRMA N-MID osteocalcin concentrations were determined by an IRMA kit following the manufacturer’s instructions (N-MID osteocalcinE IRMA, Osteometer BioTech A / S, Rodovre, Denmark). Two monoclonal antibodies (MAbs) which recognize the midregion human osteocalcin (aa20–43) and N-terminus (aa1–19) were used in this assay. Duplicate measurements were performed for each assay. The intra- and interassay coefficients of variation (CV) were 4.4% (15.8 ng / ml) and 3.5% (from the manufacturer’s insert), respectively. The detection limit was 0.5 ng / ml.
70
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
2.4. Measurements of the other biochemical markers for bone formation A commercially available intact osteocalcin kit, BGP IRMA: MitsubishiE from Mitsubishi-Kagaku (Tokyo, Japan), were used for comparison studies. The antibodies used in this assay were two MAbs which recognize the midportion (aa12–33) and C-terminus (aa30–49) of osteocalcin [6]. The intra- and interassay CVs were 6.3% (at 6.3 ng / ml) and 4.0% (6.7 ng / ml), respectively. Bone specific alkaline phosphatase activity (Bone ALP) was measured by an EIA kit with ALKPHASE-BE (Metra Biosystems Inc., Mountain View, CA, USA) [14]. The monoclonal antibody against human bone ALP was used in this assay. CV was 4.0–8.3%. C-terminal propeptide of type I procollagen (PICP) was assayed by a RIA kit from Orion Diagnostica (Oulunsalo, Finland) [15]. CV was 2.0–8.0%.
2.5. Statistical analysis Data were analyzed using a StatView II program (Abacus Concepts, Inc., Berkeley, CA) on a Macintosh computer. The statistical significance was determined with nonparametric statistics. Mann–Whitney U-test was used to compare between two groups. Simple regression was performed for univariate correlation and the statistical significance of correlation was determined with Spearman rank correlation test. Values of P , 0.05 were considered to be significant.
3. Results All studies were performed using sera kept at 2 808C. The preparation of sera was described above in sample collection. Ten serum samples collected from ten volunteers were repeatedly frozen and thawed up to seven times and were measured by an N-MID IRMA and an intact IRMA at odd thawings (Fig. 1). The variability of the values of N-MID osteocalcin was less than that of the intact osteocalcin. The stability of osteocalcin in serum during storage at 4 and 2 308C was tested. After thawing of ten serum samples, an aliquots of batches of each serum sample was kept at 48C for 2, 5 and 7 days, or kept at 2 308C for 2, 4, 10 weeks. After those interval, each batch of serum was re-frozen at 2 808C. All frozen samples were thawed and assayed immediately by the N-MID IRMA and the intact IRMA. Serum osteocalcin concentrations were expressed as percentages of the initial values (Fig. 2). When stored at 48C, the mean value of N-MID was 94.3% of the initial value in 7 days, whereas the mean value of the intact assay was 73.4% of the initial value after 7 days. The reduction of intact
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
71
Fig. 1. Variability after seven freeze / thaw cycles (measured at odd thawings). The values (n510) were expressed as percentages of the initial values. Upper panel, N-MID IRMA; lower panel, intact IRMA.
values were significantly larger than that of N-MID values after 2, 5 and 7 days. When stored at 2 308C, the N-MID values did not change for up to 10 weeks. The intact values showed the greater variation from 89.1 to 132.4% of the initial value in 10 weeks. Table 1 shows the mean6S.D. of the concentrations of osteocalcin measured by the N-MID osteocalcin and the intact osteocalcin in PRE, POST, VX and HX.
72
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
Fig. 2. The stability of osteocalcin in serum during storage at 4 and 2308C (n510). Upper panel: N-MID IRMA; lower panel: intact IRMA; left panel: 2308C; right panel: 48C. Serum osteocalcin concentrations were expressed as percentages of the initial values. After storage at 48C, there were significant differences in the reduction percentage between N-MID and intact at 2 days (P5 0.012), 5 days (P50.002), and 7 days (P50.0011) by Mann–Whitney U-test.
The percent mean increase of osteocalcin in POST over PRE was 98% in N-MID and 42% in intact. To evaluate the discrimination power of N-MID in postmenopausal subjects (POST) and osteoporotic patients (VX 1 HX), the z-scores of N-MID and intact, for comparison, were calculated in POST against PRE, and in VX and HX against POST (Fig. 3). The formula for z-scores was z 5 (Xi 2 m) /SD, where Xi is the independent value, and where m and SD are the mean and standard
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
73
Table 1 The concentrations of N-MID and intact osteocalcin in premenopausal (PRE) and postmenopausal subjects (POST) and in osteoporotic patients with vertebral fractures (VX) and hip fracture (HX), and the characteristics of the groups
n Age, range Mean6S.D. N-MID Intact
PRE
POST
VX
HX
27 32–51 42.664.2 8.165.1 3.861.9
21 51–90 64.069.7 16.164.7* 5.463.2
23 51–92 73.168.5 18.4611.6 9.166.5***
45 62–90 77.268.0 11.267.5** 5.063.7
* P,0.001 vs. PRE; ** P,0.005 vs. POST; *** P,0.05 vs. POST.
deviations of the against group. Similar results were obtained from both assays in each group. Table 2 shows the correlation between N-MID osteocalcin and the other
Fig. 3. The z-scores of an N-MID osteocalcin and an intact osteocalcin in the postmenopausal subjects (POST) against the premenopausal subjects, and vertebral fractures (VX) and hip fracture (HX) against POST. (j) N-MID IRMA; (9) intact IRMA.
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
74
Table 2 Correlation between N-MID osteocalcin IRMA and other biochemical markers for bone formation a
Intact osteocalcin Bone ALP PICP a
r
p
0.755 0.606 0.568
0.0001 0.0001 0.0001
p values were calculated by Spearman rank correlation test.
biochemical markers for bone formation. All correlations were significant in the total of the subjects and patients.
4. Discussion The present study demonstrates that the N-MID osteocalcin is more stable in human serum for up to 7 days at 48C and up to 10 weeks at 2308C than the intact osteocalcin. With respect to the long-term handling, samples was stable even after seven freeze–thaw cycles. The stability of osteocalcin during short time storage at 48C measured by an N-MID IRMA was also better than the intact osteocalcin assay. Osteocalcin was stable during long time storage at 2308C when measured by the N-MID IRMA. Intact osteocalcin is reported to be cleaved at the site of C-terminal sequence of osteocalcin to produce a large N-terminal midfragment. Therefore, the concentrations of intact assay greatly were reduced by the storage at 48C. When stored at 2308C, the concentrations of intact osteocalcin did not simply decrease, but showed a great variation over 10 weeks. Since re-frozen and thawing was performed two times during the testing, the freeze and thaw might have more affected the concentrations of intact osteocalcin than the storage did. Under routine clinical laboratory conditions, an N-MID osteocalcin IRMA could have an advantage in storage and clinical utility over other osteocalcin assays. The percent mean increase of N-MID IRMA osteocalcin in postmenopause over premenopause was 98% in the present study. That is greater than previous report (56%) for the N-MID ELISA [11] and a conventional RIA [16]. Comparison of z-scores of N-MID osteocalcin and intact osteocalcin showed the similar trends in postmenopausal subjects against premenopausal subjects and in vertebral fractures and hip fractures against postmenopausal subjects. Z-scores of N-MID and intact were plus values in postmenopausal subjects and vertebral fractures and were minus values in hip fractures. Therefore, the discriminative ability of an N-MID osteocalcin and intact osteocalcin are compatible in both postmenopausal subjects and osteoporotic patients. However, the extent of
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
75
z-scores values were somewhat different between N-MID and intact. The degree of an increase of the measurements in postmenopausal subjects might mainly cause the difference of the extent of z-scores. N-MID osteocalcin significantly correlated the other biochemical markers for bone formation. It moderately correlated to intact osteocalcin. The proportion of intact molecule and an N-MID fragment in blood circulation in subjects and patients may produce this result. However, the stability of the assay might also have affected the correlation. Bone specific ALP is produced by osteoblasts and reflects osteoblastic activity, whereas PICP is a degradation product from type I procollagen and is secreted into blood circulation during the synthesis of bone matrix. Although the biochemical character is different between these two bone formation markers, the correlation of those two assays to N-MID osteocalcin were similar, with a moderate degree of correlation.
Acknowledgements We are grateful to Mika Matsumoto for her technical help.
References [1] Hauschka PV, Lian JB, Cole DEC, Gundberg CM. Osteocalcin and matrix gla protein: Vitamin K-dependent proteins in bone. Physiol Rev 1989;69:990–1047. [2] Epstein S. Serum and urinary markers of bone remodeling: assessment of bone turnover. Endocr Rev 1988;9:437–49. [3] Lian JB, Gundberg CM. Osteocalcin biochemical considerations and clinical applications. Clin Orthop 1988;226:267–91. [4] Price PA, Nishimoto SK. Radioimmunoassay for the vitamin-K dependent protein of bone and its discovery in plasma. Proc Natl Acad Sci USA 1980;77:2234–8. [5] Tracy RP, Andrianorivo A, Riggs BL, Mann KG. Comparison of monoclonal and polyclonal antibody-based immunoassays for osteocalcin: a study of sources of variation in assay results. J Bone Miner Res 1990;5:451–61. [6] Masters PW, Jones RG, Purves DA, Cooper EH, Cooney JM. Commercial assays for serum osteocalcin give clinically discordant results. Clin Chem 1994;40:358–63. [7] Gundberg CM, Wilson MS, Gallop PM, Parfitt AM. Determination of osteocalcin in human serum: results with two kits compared with those by a well-characterized assay. Clin Chem 1985;31:1720–3. [8] Gundberg CM, Weinstein RS. Multiple immunoreactive forms of osteocalcin in uremic serum. J Clin Invest 1986;77:1762–7. [9] Garnero P, Grimaux M, Demiaux B, Preaudat C, Seguin P, Delmas PD. Measurement of serum osteocalcin with a human specific two-site immunoradiometric assay. J Bone Miner Res 1992;7:1389–98. [10] Garnero P, Grimaux M, Seguin P, Delmas PD. Characterization of immunoreactive forms of human osteocalcin generated in vivo and in vitro. J Bone Miner Res 1994;9:255–64.
76
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
[11] Rosenquist C, Qvist P, Bjarnason N, Cristiansen C. Measurement of a more stable region of osteocalcin in serum by ELISA with two monoclonal antibodies. Clin Chem 1995;10:1439– 45. [12] Genant HK, Wu CY, Van Kuijk C, Nevitt MC. Vertebral fracture assessment using a semiquantitative technique. J Bone Miner Res 1993;8:1137–48. [13] Takahashi M, Kushida K, Hoshino H, Aoshima H, Ohishi T, Inoue T. Effect of fracture to biochemical markers and vitamin K in acute phase of patients with hip fracture. Clin Chem 1998;44:1583–4. [14] Gomez BJ, Ardakani S, Ju J et al. Monoclonal antibody assay for measuring bone specific alkaline phosphatase activity in serum. Clin Chem 1995;41:1560–6. [15] Parfitt AM, Simon LS, Villanueva AR, Simon SM. Procollagen type I carboxyterminal extension peptide in serum as a marker of collagen biosynthesis in bone. Correlation with iliac bone formation rates and comparison with total alkaline phosphatase. J Bone Miner Res 1987;2:427–36. [16] Pino JD, Martin-Gomez E, Martin-Rodriguez M et al. Influence of sex, age, and menopause in serum osteocalcin (BGP) levels. Klin Wochenschr 1991;69:1135–8.
Comparison of the analytical and clinical performance characteristics of an N-MID versus an intact osteocalcin immunoradiometric assay a, a a Masaaki Takahashi *, Kazuhiro Kushida , Akira Nagano , b Tetsuo Inoue a
Department of Orthopaedic Surgery, Hamamatsu University School of Medicine, 3600 Handa, Hamamatsu 431 -3192, Japan b Aoyama General Hospital, Hoi-Gun, Kosakai-Cho 441 -0195, Japan
Received 29 July 1999; received in revised form 23 November 1999; accepted 29 November 1999
Abstract Osteocalcin is the most abundant non-collagenous protein in bone, reflecting its formation. It was reported that the instability of intact osteocalcin results from the cleavage of the C-terminal sequence of osteocalcin to produce a large N-terminal osteocalcin fragment. A two-site immunoassay for detecting both the N-terminal osteocalcin fragment and the intact osteocalcin was developed that were both independent of an unstable C-terminal sequence. The aim of this study is to investigate the performance of an N-MID osteocalcin immunoradiometric assay and to compare it with an intact osteocalcin assay. Ten serum samples were repeatedly frozen and thawed up to seven times. The variability of the values of N-MID osteocalcin was less than that of the intact osteocalcin. For stability of osteocalcin in serum after storage, the mean value of N-MID was 94.3% of the initial value after 7 days at 48C, whereas the intact was 73.4%. The reduction of intact values were significantly larger than that of N-MID after 2, 5 and 7 days. At 2 308C, the values of N-MID did not change for up to 10 weeks. The concentrations of osteocalcin measured by an N-MID osteocalcin and an intact osteocalcin were investigated in 27 premenopausal subjects, 27 postmenopausal subjects, and 68 osteoporotic patients (23 with vertebral fractures and 45 with hip fractures). The percent mean increase of osteocalcin in postmenopausal subjects over premenopausal subjects was 98% in N-MID versus 42% in the intact assay. The z-scores of N-MID and intact showed similar results in all groups. N-MID osteocalcin significantly correlated with intact osteocalcin (r 5 0.755), and other biochemical markers for bone formation, such as
*Corresponding author. Tel.: 181-53-435-2299; fax: 181-53-435-2296. E-mail address: [email protected] (M. Takahashi) 0009-8981 / 00 / $ – see front matter 2000 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 99 )00251-X
68
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
bone specific alkaline phosphatase (r 5 0.606) and C-terminal propeptide of type I procollagen (PICP) (r 5 0.568). An N-MID IRMA had better stability during storage than intact and had the discriminative ability which is similar to the intact assay in postmenopause and osteoporosis. Therefore, an N-MID osteocalcin IRMA could improve the clinical utility and evaluation of osteocalcin. 2000 Elsevier Science B.V. All rights reserved. Keywords: Osteocalcin; Bone marker; Bone formation; Vertebral fracture; Hip fracture
1. Introduction In the past decade, markers for bone formation have been developed. Osteocalcin, also called bone gla protein (BGP), a 49 amino acid protein, is the most abundant non-collagenous protein in bone [1]. Although the biological function of osteocalcin is still unknown, osteocalcin is synthesized by osteoblast and is only located in the bone. Evidence has shown that it reflects osteoblastic activity and bone formation [2,3]. Although most of the newly synthesized osteocalcin is incorporated into the bone matrix, some fraction is released into blood circulation [4]. Since circulating osteocalcin concentrations were initially measured using conventional radioimmunoassays (RIAs), a variety of immunometric assays, using polyclonal or monoclonal antibodies, have been developed with better analytical sensitivity and specificity than conventional RIAs. However, those immunoassays have given discordant results in diverse clinical settings [5,6]. This discordance is caused by epitopic specificity and differential reactivity with circulating fragments of osteocalcin [7,8]. To overcome this problem, two-site immunoassays for the measurement of the intact molecule of osteocalcin were developed [9]. However, the assays for intact osteocalcin have not been any better than the conventional assays. It was reported that one third of circulating osteocalcin was intact osteocalcin, one third was a large Nterminal osteocalcin fragment and the rest were various smaller fragments [10]. The instability of intact osteocalcin is mainly caused by the cleavage of the C-terminal sequence of osteocalcin, producing a large N-terminal midfragment [10]. Accordingly, intact molecules of osteocalcin in vitro such as, blood, serum and plasma, are thought to be unstable. Therefore, use of these assays requires precise control of temperature, duration of storage, and the avoidance of freezing and thawing. A two-site enzyme-immunoassay for detecting both the large N-terminal osteocalcin and intact osteocalcin was developed to be independent of an unstable C-terminal sequence. The performance of this enzyme-immunoassay have been reported [11]. The aim of this study is to investigate the analytical and clinical performance of an N-MID osteocalcin immunoradiometric assay (IRMA).
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
69
2. Materials and methods
2.1. Subjects All subjects were females. The premenopausal group (PRE) consisted of 27 women aged 32–51 years (mean ages 6S.D.; 42.664.2 years) who had regular menstrual cycles. The postmenopausal group (POST) consisted of 21 postmenopausal females aged 51–90 years (64.069.7). No subjects had any previous history of any disease known to affect bone metabolism (such as hyperthyroidism, renal disease, collagen disease or ovarian tumor), and had not previously used estrogen or any other hormonal medications. All were volunteers and gave their informed consent prior to the study. The vertebral fracture group (VX) consisted of 23 patients aged 51–92 years (mean ages6S.D.; 73.16 8.5 years) diagnosed by X-ray films of the thoracic and lumbar spine, which deformed more than 20% reduction in anterior, middle, and / or posterior height as vertebral fracture [12]. There were no fractures sustained within the prior 6 months. The hip fracture group (HX) consisted of 45 patients aged 62–90 years (mean ages 6S.D.; 77.268.0 years). None of the patients in either group had been receiving any medications that might have affected calcium metabolism during the study. Secondary osteoporosis patients were excluded from the study.
2.2. Sample collection Blood was kept at room temperature, centrifuged within 1 h, and sera was stored at 2 808C until use. Prior to analysis, the frozen samples were thawed and the measurements were initiated within 1 h after thawing. For sera in the hip fracture group, samples were collected within 48 h after fracture [13].
2.3. Measurements of N-MID osteocalcin IRMA N-MID osteocalcin concentrations were determined by an IRMA kit following the manufacturer’s instructions (N-MID osteocalcinE IRMA, Osteometer BioTech A / S, Rodovre, Denmark). Two monoclonal antibodies (MAbs) which recognize the midregion human osteocalcin (aa20–43) and N-terminus (aa1–19) were used in this assay. Duplicate measurements were performed for each assay. The intra- and interassay coefficients of variation (CV) were 4.4% (15.8 ng / ml) and 3.5% (from the manufacturer’s insert), respectively. The detection limit was 0.5 ng / ml.
70
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
2.4. Measurements of the other biochemical markers for bone formation A commercially available intact osteocalcin kit, BGP IRMA: MitsubishiE from Mitsubishi-Kagaku (Tokyo, Japan), were used for comparison studies. The antibodies used in this assay were two MAbs which recognize the midportion (aa12–33) and C-terminus (aa30–49) of osteocalcin [6]. The intra- and interassay CVs were 6.3% (at 6.3 ng / ml) and 4.0% (6.7 ng / ml), respectively. Bone specific alkaline phosphatase activity (Bone ALP) was measured by an EIA kit with ALKPHASE-BE (Metra Biosystems Inc., Mountain View, CA, USA) [14]. The monoclonal antibody against human bone ALP was used in this assay. CV was 4.0–8.3%. C-terminal propeptide of type I procollagen (PICP) was assayed by a RIA kit from Orion Diagnostica (Oulunsalo, Finland) [15]. CV was 2.0–8.0%.
2.5. Statistical analysis Data were analyzed using a StatView II program (Abacus Concepts, Inc., Berkeley, CA) on a Macintosh computer. The statistical significance was determined with nonparametric statistics. Mann–Whitney U-test was used to compare between two groups. Simple regression was performed for univariate correlation and the statistical significance of correlation was determined with Spearman rank correlation test. Values of P , 0.05 were considered to be significant.
3. Results All studies were performed using sera kept at 2 808C. The preparation of sera was described above in sample collection. Ten serum samples collected from ten volunteers were repeatedly frozen and thawed up to seven times and were measured by an N-MID IRMA and an intact IRMA at odd thawings (Fig. 1). The variability of the values of N-MID osteocalcin was less than that of the intact osteocalcin. The stability of osteocalcin in serum during storage at 4 and 2 308C was tested. After thawing of ten serum samples, an aliquots of batches of each serum sample was kept at 48C for 2, 5 and 7 days, or kept at 2 308C for 2, 4, 10 weeks. After those interval, each batch of serum was re-frozen at 2 808C. All frozen samples were thawed and assayed immediately by the N-MID IRMA and the intact IRMA. Serum osteocalcin concentrations were expressed as percentages of the initial values (Fig. 2). When stored at 48C, the mean value of N-MID was 94.3% of the initial value in 7 days, whereas the mean value of the intact assay was 73.4% of the initial value after 7 days. The reduction of intact
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
71
Fig. 1. Variability after seven freeze / thaw cycles (measured at odd thawings). The values (n510) were expressed as percentages of the initial values. Upper panel, N-MID IRMA; lower panel, intact IRMA.
values were significantly larger than that of N-MID values after 2, 5 and 7 days. When stored at 2 308C, the N-MID values did not change for up to 10 weeks. The intact values showed the greater variation from 89.1 to 132.4% of the initial value in 10 weeks. Table 1 shows the mean6S.D. of the concentrations of osteocalcin measured by the N-MID osteocalcin and the intact osteocalcin in PRE, POST, VX and HX.
72
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
Fig. 2. The stability of osteocalcin in serum during storage at 4 and 2308C (n510). Upper panel: N-MID IRMA; lower panel: intact IRMA; left panel: 2308C; right panel: 48C. Serum osteocalcin concentrations were expressed as percentages of the initial values. After storage at 48C, there were significant differences in the reduction percentage between N-MID and intact at 2 days (P5 0.012), 5 days (P50.002), and 7 days (P50.0011) by Mann–Whitney U-test.
The percent mean increase of osteocalcin in POST over PRE was 98% in N-MID and 42% in intact. To evaluate the discrimination power of N-MID in postmenopausal subjects (POST) and osteoporotic patients (VX 1 HX), the z-scores of N-MID and intact, for comparison, were calculated in POST against PRE, and in VX and HX against POST (Fig. 3). The formula for z-scores was z 5 (Xi 2 m) /SD, where Xi is the independent value, and where m and SD are the mean and standard
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
73
Table 1 The concentrations of N-MID and intact osteocalcin in premenopausal (PRE) and postmenopausal subjects (POST) and in osteoporotic patients with vertebral fractures (VX) and hip fracture (HX), and the characteristics of the groups
n Age, range Mean6S.D. N-MID Intact
PRE
POST
VX
HX
27 32–51 42.664.2 8.165.1 3.861.9
21 51–90 64.069.7 16.164.7* 5.463.2
23 51–92 73.168.5 18.4611.6 9.166.5***
45 62–90 77.268.0 11.267.5** 5.063.7
* P,0.001 vs. PRE; ** P,0.005 vs. POST; *** P,0.05 vs. POST.
deviations of the against group. Similar results were obtained from both assays in each group. Table 2 shows the correlation between N-MID osteocalcin and the other
Fig. 3. The z-scores of an N-MID osteocalcin and an intact osteocalcin in the postmenopausal subjects (POST) against the premenopausal subjects, and vertebral fractures (VX) and hip fracture (HX) against POST. (j) N-MID IRMA; (9) intact IRMA.
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
74
Table 2 Correlation between N-MID osteocalcin IRMA and other biochemical markers for bone formation a
Intact osteocalcin Bone ALP PICP a
r
p
0.755 0.606 0.568
0.0001 0.0001 0.0001
p values were calculated by Spearman rank correlation test.
biochemical markers for bone formation. All correlations were significant in the total of the subjects and patients.
4. Discussion The present study demonstrates that the N-MID osteocalcin is more stable in human serum for up to 7 days at 48C and up to 10 weeks at 2308C than the intact osteocalcin. With respect to the long-term handling, samples was stable even after seven freeze–thaw cycles. The stability of osteocalcin during short time storage at 48C measured by an N-MID IRMA was also better than the intact osteocalcin assay. Osteocalcin was stable during long time storage at 2308C when measured by the N-MID IRMA. Intact osteocalcin is reported to be cleaved at the site of C-terminal sequence of osteocalcin to produce a large N-terminal midfragment. Therefore, the concentrations of intact assay greatly were reduced by the storage at 48C. When stored at 2308C, the concentrations of intact osteocalcin did not simply decrease, but showed a great variation over 10 weeks. Since re-frozen and thawing was performed two times during the testing, the freeze and thaw might have more affected the concentrations of intact osteocalcin than the storage did. Under routine clinical laboratory conditions, an N-MID osteocalcin IRMA could have an advantage in storage and clinical utility over other osteocalcin assays. The percent mean increase of N-MID IRMA osteocalcin in postmenopause over premenopause was 98% in the present study. That is greater than previous report (56%) for the N-MID ELISA [11] and a conventional RIA [16]. Comparison of z-scores of N-MID osteocalcin and intact osteocalcin showed the similar trends in postmenopausal subjects against premenopausal subjects and in vertebral fractures and hip fractures against postmenopausal subjects. Z-scores of N-MID and intact were plus values in postmenopausal subjects and vertebral fractures and were minus values in hip fractures. Therefore, the discriminative ability of an N-MID osteocalcin and intact osteocalcin are compatible in both postmenopausal subjects and osteoporotic patients. However, the extent of
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
75
z-scores values were somewhat different between N-MID and intact. The degree of an increase of the measurements in postmenopausal subjects might mainly cause the difference of the extent of z-scores. N-MID osteocalcin significantly correlated the other biochemical markers for bone formation. It moderately correlated to intact osteocalcin. The proportion of intact molecule and an N-MID fragment in blood circulation in subjects and patients may produce this result. However, the stability of the assay might also have affected the correlation. Bone specific ALP is produced by osteoblasts and reflects osteoblastic activity, whereas PICP is a degradation product from type I procollagen and is secreted into blood circulation during the synthesis of bone matrix. Although the biochemical character is different between these two bone formation markers, the correlation of those two assays to N-MID osteocalcin were similar, with a moderate degree of correlation.
Acknowledgements We are grateful to Mika Matsumoto for her technical help.
References [1] Hauschka PV, Lian JB, Cole DEC, Gundberg CM. Osteocalcin and matrix gla protein: Vitamin K-dependent proteins in bone. Physiol Rev 1989;69:990–1047. [2] Epstein S. Serum and urinary markers of bone remodeling: assessment of bone turnover. Endocr Rev 1988;9:437–49. [3] Lian JB, Gundberg CM. Osteocalcin biochemical considerations and clinical applications. Clin Orthop 1988;226:267–91. [4] Price PA, Nishimoto SK. Radioimmunoassay for the vitamin-K dependent protein of bone and its discovery in plasma. Proc Natl Acad Sci USA 1980;77:2234–8. [5] Tracy RP, Andrianorivo A, Riggs BL, Mann KG. Comparison of monoclonal and polyclonal antibody-based immunoassays for osteocalcin: a study of sources of variation in assay results. J Bone Miner Res 1990;5:451–61. [6] Masters PW, Jones RG, Purves DA, Cooper EH, Cooney JM. Commercial assays for serum osteocalcin give clinically discordant results. Clin Chem 1994;40:358–63. [7] Gundberg CM, Wilson MS, Gallop PM, Parfitt AM. Determination of osteocalcin in human serum: results with two kits compared with those by a well-characterized assay. Clin Chem 1985;31:1720–3. [8] Gundberg CM, Weinstein RS. Multiple immunoreactive forms of osteocalcin in uremic serum. J Clin Invest 1986;77:1762–7. [9] Garnero P, Grimaux M, Demiaux B, Preaudat C, Seguin P, Delmas PD. Measurement of serum osteocalcin with a human specific two-site immunoradiometric assay. J Bone Miner Res 1992;7:1389–98. [10] Garnero P, Grimaux M, Seguin P, Delmas PD. Characterization of immunoreactive forms of human osteocalcin generated in vivo and in vitro. J Bone Miner Res 1994;9:255–64.
76
M. Takahashi et al. / Clinica Chimica Acta 294 (2000) 67 – 76
[11] Rosenquist C, Qvist P, Bjarnason N, Cristiansen C. Measurement of a more stable region of osteocalcin in serum by ELISA with two monoclonal antibodies. Clin Chem 1995;10:1439– 45. [12] Genant HK, Wu CY, Van Kuijk C, Nevitt MC. Vertebral fracture assessment using a semiquantitative technique. J Bone Miner Res 1993;8:1137–48. [13] Takahashi M, Kushida K, Hoshino H, Aoshima H, Ohishi T, Inoue T. Effect of fracture to biochemical markers and vitamin K in acute phase of patients with hip fracture. Clin Chem 1998;44:1583–4. [14] Gomez BJ, Ardakani S, Ju J et al. Monoclonal antibody assay for measuring bone specific alkaline phosphatase activity in serum. Clin Chem 1995;41:1560–6. [15] Parfitt AM, Simon LS, Villanueva AR, Simon SM. Procollagen type I carboxyterminal extension peptide in serum as a marker of collagen biosynthesis in bone. Correlation with iliac bone formation rates and comparison with total alkaline phosphatase. J Bone Miner Res 1987;2:427–36. [16] Pino JD, Martin-Gomez E, Martin-Rodriguez M et al. Influence of sex, age, and menopause in serum osteocalcin (BGP) levels. Klin Wochenschr 1991;69:1135–8.