The authors have no conflict of interest
Serum TRACP 5b Is a Useful Marker for Monitoring Alendronate Treatment: Comparison With Other Markers of Bone Turnover†
Article first published online: 4 APR 2005
Copyright © 2005 ASBMR
Journal of Bone and Mineral Research
Volume 20, Issue 10, pages 1804–1812, October 2005
How to Cite
Nenonen, A., Cheng, S., Ivaska, K. K., Alatalo, S. L., Lehtimäki, T., Schmidt-Gayk, H., Uusi-Rasi, K., Heinonen, A., Kannus, P., Sievänen, H., Vuori, I., Väänänen, H. K. and Halleen, J. M. (2005), Serum TRACP 5b Is a Useful Marker for Monitoring Alendronate Treatment: Comparison With Other Markers of Bone Turnover. J Bone Miner Res, 20: 1804–1812. doi: 10.1359/JBMR.050403
- Issue published online: 4 DEC 2009
- Article first published online: 4 APR 2005
- Manuscript Accepted: 4 APR 2005
- Manuscript Revised: 23 MAR 2005
- Manuscript Received: 18 JUN 2004
- TRACP 5b;
- bone markers;
- treatment monitoring;
- randomized study
We studied clinical performance of serum TRACP 5b and other bone turnover markers, including S-CTX, U-DPD, S-PINP, S-BALP, and S-OC, for monitoring alendronate treatment. TRACP 5b had higher clinical sensitivity, area under the ROC curve, and signal-to-noise ratio than the other markers.
Introduction: The purpose of this study was to compare the clinical performance of serum TRACP 5b (S-TRACP5b) with that of other markers of bone turnover in the monitoring of alendronate treatment.
Materials and Methods: This double-blinded study included 148 healthy postmenopausal women that were randomly assigned into two groups: one receiving 5 mg alendronate daily (n = 75) and the other receiving placebo (n = 73) for 12 months. All individuals in both groups received calcium and vitamin D daily. The bone resorption markers S-TRACP5b, serum C-terminal cross-linked telopeptides of type I collagen (S-CTX), and total urinary deoxypyridinoline (U-DPD), and the serum markers of bone formation procollagen I N-terminal propeptide (S-PINP), bone-specific alkaline phosphatase (S-BALP), and total osteocalcin (S-OC) were assessed at baseline and at 3, 6, and 12 months after initiation of treatment. Lumbar spine BMD (LBMD) was measured at baseline and 12 months.
Results: Compared with the placebo group, LBMD increased, and all bone markers decreased significantly more in the alendronate group (p < 0.001 for each parameter). The decrease of S-TRACP5b after first 3 months of alendronate treatment correlated significantly with the changes of all other markers except S-OC, the best correlation being with S-CTX (r = 0.60, p < 0.0001). The changes of LBMD at 12 months only correlated significantly with the changes of S-TRACP5b (r = −0.32, p = 0.005) and S-CTX (r = −0.24, p = 0.037) at 3 months. Based on clinical sensitivity, receiver operating characteristic (ROC) curves, and signal-to-noise ratio, S-TRACP5b, S-CTX, and S-PINP were the best markers for monitoring alendronate treatment. Clinical sensitivity, area under the ROC curve, and signal-to-noise ratio were higher for S-TRACP5b than for the other markers.
Conclusion: These results show that S-TRACP5b, S-CTX, and S-PINP are useful markers for monitoring alendronate treatment.
IN NORMAL HUMAN tissues, high amounts of type-5 acid phosphatase (ACP5, EC 126.96.36.199), also known as TRACP, are found in bone-resorbing osteoclasts, activated macrophages, and dendritic cells. (1–3) TRACP is a glycoprotein with a molecular weight of 32 kDa that is active as a phosphatase and as a generator of reactive oxygen species. The polypeptide chain of TRACP can be cleaved with proteases such as trypsin and cathepsins, which generates two disulfide-bridged subunits and activates phosphatase activity. (4–6)
Two differentially glycosylated isoforms of TRACP, a sialic acid-containing isoform TRACP 5a and a nonsialylated isoform TRACP 5b, circulate in human blood. (7, 8) TRACP 5a is the noncleaved form with a low specific activity, and it is derived from activated macrophages. (9, 10) TRACP 5b is the cleaved two-subunit form with a high specific activity, and it is derived from osteoclasts. (9–12) Three different methods have been published for the measurement of serum TRACP 5b activity (S-TRACP5b) as a marker of bone resorption: one kinetic method and two immunoassays. (13–15) The kinetic method uses a pattern of inhibitors, including tartrate, fluoride, and heparin, to obtain the specificity for TRACP 5b. (13) The immunoassays use antibodies specific for type-5 TRACP. Although these antibodies are specific for TRACP, they can not distinguish between TRACP 5a and TRACP 5b. One of the immunoassays uses naphthol-ASBI-phosphate as a selective substrate for TRACP 5b, (14) and the other uses a selective reaction pH where TRACP 5a is considerably less active than TRACP 5b. (15)
The pH-selective immunoassay is available commercially (SBA-Sciences, Oulu, Finland), and it has been used widely to study S-TRACP5b as a marker of bone resorption. With this assay, it has been shown that S-TRACP5b is elevated in patients with bone diseases, (16–20) decreased after antiresorptive treatment, (15, 21–25) correlates with other markers of bone turnover and BMD, (26) and predicts future fractures. (27) Moreover, S-TRACP5b is a reliable bone marker because it is derived specifically from osteoclasts, its serum levels are not affected by feeding, it has low biological and analytical variability, and it is not accumulated into the circulation in renal or hepatic failure. (12, 16, 25) The only concern when measuring S-TRACP5b is that the enzyme is relatively labile when serum samples are stored at −20°C for >1 month, and for longer periods, serum samples should be stored at −70°C where the enzyme is stable. (15)
One important application for the use of bone turnover markers is their use for monitoring antiresorptive treatment. Here we have studied the use of S-TRACP5b as a marker of bone resorption for monitoring alendronate treatment and compared it with other commonly used and well-known markers of bone turnover.
MATERIALS AND METHODS
Study population and design
A questionnaire was sent to a random population sample of 3000 women from a cohort born from 1942 to 1947 living in the city of Tampere, Finland, asking their interest in participating in the study. (28) Two hundred sixty-three women were invited to a screening examination, and 164 of them were selected to the study. Inclusion criteria were no previous bone fractures, 1–5 years after menopause, no current or previous use of estrogen, corticosteroids, bisphosphonates, or other drugs affecting bone metabolism, no current or previous illnesses affecting bone metabolism, no contraindication to alendronate, femoral neck BMD >0.650 g/cm2 (which is 2.5 SD below the young normal women's reference value as determined by DXA at the UKK Institute), and follicle-stimulating hormone (FSH) level >30 IU/liter. All subjects provided a written informed consent before the study, and the study protocol was approved by a local research committee and independent medical ethics committee at the Tampere University Hospital. After randomization, 5 subjects were no longer willing to participate, and 159 subjects started the study. A total of 11 subjects failed to give all necessary blood samples to allow measuring the bone markers at all time-points. Thus, the remaining 148 subjects (75 in the alendronate group and 73 in the placebo group) were included in the study. The study was a 1-year double-blinded randomized placebo-controlled intervention trial with two experimental groups: one receiving daily 5 mg of alendronate (Fosamax; Merck & Co., Rahway, NJ, USA) and the other receiving placebo (placebo pills, identical with the effective ones donated by Merck & Co). All subjects in both groups received a daily supplement of calcium carbonate (630 mg) and vitamin D (200 IU = 5 μg; Citracal + D; Mission Pharmacal, San Antonio, TX, USA). Originally, the study contained the following four experimental groups as described previously(28): (1) alendronate + exercise; (2) alendronate; (3) placebo + exercise; (4) placebo. Because exercise had no effect on any of the bone markers, (28) groups 1 and 2 were pooled to one group receiving alendronate, and groups 3 and 4 were pooled to one group receiving placebo in this study.
Biochemical markers and BMD
Venous blood samples and 24-h urine samples were obtained after 12-h fasting at baseline and at 3, 6, and 12 months. Serum was separated by centrifugation, aliquoted, and stored at −70°C until the analyses. Serum C-terminal cross-linked telopeptides of type I collagen (S-CTX; Roche Diagnostics, Mannheim, Germany), serum bone-specific alkaline phosphatase (S-BALP; Metra Biosystems, Mountain View, CA, USA), and serum procollagen I N-terminal propeptide (S-PINP; Orion Diagnostica, Espoo, Finland) were determined using commercial immunoassays. S-CTX was measured with an Elecsys 1010 Immunoanalyser (Roche). S-TRACP5b and serum total osteocalcin (S-OC) were measured using in-house immunoassays as described. (15, 29) The commercially available TRACP 5b immunoassay (SBA-Sciences, Turku, Finland) is based on the in-house assay used in this study. Urinary total deoxypyridinoline (U-DPD) was determined by high-performance liquid chromatography as described. (30) All bone marker measurements were performed blinded and in duplicates, with the exception of S-CTX, which was performed in single measurements. Lumbar spine BMD (LBMD) was measured with DXA (XR-26; Norland, Fort Atkinson, WI, USA) at baseline and at the end of the study according to our standard procedures. (31)
Statistical analysis was carried out on SPSS 11.5 for Windows software (SPSS, Chicago, IL, USA) in a microcomputer. Normal distribution and homogeneity of variance were checked before further analyses. Overall differences in the responses of bone markers and LBMD over time between the treatment groups were compared with ANOVA for repeated measurements. Within the study groups, dependent t-test was used to test the significance of differences over time (at baseline versus 3 months for bone markers and 12 months for LBMD). Continuous and normally distributed parameters (i.e., age, BMI) were compared between study groups by t-test for independent samples. In addition, Pearson correlation coefficients were calculated. Analytical variability (CVa) of each bone marker was determined as the mean CV of all duplicated measurements performed in this study. Because S-CTX was analyzed with single measurements, the CVa value of S-CTX was determined separately. Biological variability (CVi) was determined as the mean CV of the changes observed at 3 months compared with baseline in the placebo group. For LBMD, CVa was 0.7%, (31) and CVi was determined as the mean CV of the changes observed at 12 months compared with baseline in the placebo group. One-sided least significant change (LSC) at p < 0.05 was determined for each marker and LBMD based on their analytical and biological variability using the equation: LSC = 2.33 x √(CVa2 + CVi2). (25) For each marker, those subjects that showed a decrease of more than LSC at 3 months (increase at 12 months for LBMD) were considered responders to the treatment. Clinical specificity was determined as the percentage of nonresponders in the placebo group, and clinical sensitivity as the percentage of responders in the alendronate group. Area under curve (AUC) values were determined from receiver operating characteristic (ROC) curves. Signal-to-noise ratio was determined by dividing the change of each marker in the alendronate group at 3 months (signal) with the marker's analytical and biological variability (noise) as calculated from the equation √(CVa2 + CVi2). Data are presented as mean ± SD unless otherwise stated. A p value of <0.05 was considered statistically significant.
Baseline characteristics of the placebo and alendronate groups are shown in Table 1. No statistically significant differences were observed between the groups in any of the parameters shown, including all bone marker and LBMD values. All bone markers studied decreased significantly in both study groups. LBMD increased significantly from baseline only in the alendronate group. The profiles of the changes in bone resorption markers are shown in Fig. 1 and the profiles of the changes in bone formation markers in Fig. 2. All bone markers decreased significantly more and LBMD increased significantly more in the alendronate group than in the placebo group (p < 0.001 for all parameters). S-TRACP5b decreased 40.2% in the alendronate group and 11.3% in the placebo group. S-CTX showed the highest decrease in both groups, being 63.7% in the alendronate group and 21.6% in the placebo group.
Table 2 summarizes correlation analyses of S-TRACP5b with the other bone markers studied. Baseline S-TRACP5b values in all subjects from both study groups correlated significantly with the baseline values of all other markers, the best correlation being with S-CTX (r = 0.66, p < 0.0001). The changes in S-TRACP5b values during the first 3 months of alendronate treatment correlated significantly with the changes in all other markers except S-OC, the best correlation being again with S-CTX (r = 0.60, p < 0.0001). Table 3 shows correlations of the bone markers with LBMD. Baseline LBMD values correlated significantly with baseline S-TRACP5b (r = −0.24, p = 0.003), S-CTX (r = −0.19, p = 0.021), and S-OC (r = −0.18, p = 0.032). Changes of LBMD at 12 months after start of alendronate treatment correlated significantly with changes of S-TRACP5b (r = −0.32, p = 0.005) and S-CTX (r = −0.24, p = 0.037) at 3 months.
The bone markers studied were compared for their clinical performance in monitoring alendronate treatment. Figures 3 (resorption markers) and 4 (formation markers) show changes (%) in the bone marker values after 3 months of treatment in all individual subjects in the alendronate and placebo groups. ROC curves of the bone markers are shown in Fig. 5. CVa, CVi, LSC, clinical specificity, clinical sensitivity, AUC values from ROC curves, and signal-to-noise ratio of each marker and LBMD are summarized in Table 4. All markers had high clinical specificity, the highest value being for S-CTX (98.6%). The highest clinical sensitivities were observed for S-TRACP5b (82.7%), S-CTX (78.7%), and S-PINP (73.3%), the highest AUC values for S-TRACP5b (0.879), S-CTX (0.868), and S-PINP (0.862), and the highest signal-to-noise ratios for S-TRACP5b (3.17), S-PINP (2.95), and S-CTX (2.84). The clinical sensitivity, AUC, and signal-to-noise ratio of LBMD were all substantially lower than those of all bone markers except S-OC.
In the past few years, a lot of new information has been obtained about the use of serum TRACP as a marker of bone resorption. The most important finding was that the nonsialylated 5b isoform of TRACP is released from osteoclasts, whereas another isoform, sialylated TRACP 5a, is derived from activated macrophages and dendritic cells. (7–12, 32) Importantly, it has been shown that, whereas ∼50% of total TRACP activity in normal human serum is of form 5a, ∼90% of total TRACP protein is of form 5a and nonosteoclastic origin. (12) Before this was realized, several assays for TRACP were published that measured both isoforms, 5a and 5b. These assays showed that TRACP is clearly associated with bone turnover, but its performance as a resorption marker was not very convincing compared with several other known markers. (33)
Proteases such as trypsin and cathepsins cleave the polypeptide chain of recombinant TRACP to two subunits that are linked together by a disulfide bridge. (4, 5) The noncleaved form has a low specific activity and a pH-optimum of 5.2 (similar to serum TRACP 5a), and the cleaved form has a substantially higher specific activity and a pH-optimum of 5.8 (similar to serum TRACP 5b). This suggests that the macrophage-derived TRACP 5a circulates as the noncleaved low-activity form with a pH-optimum of 5.2, and the osteoclast-derived TRACP 5b circulates as the cleaved high-activity form with a pH-optimum of 5.8. (32) The reason for this difference is not known, but it may be that TRACP is cleaved by proteases in osteoclasts, but not in macrophages before it is released into the circulation. The sialic acid present in TRACP 5a may also protect the enzyme against proteolytic cleavage.
It has been suggested that TRACP would be released from osteoclasts during the resorption process as an index of osteoclast activity. (1) However, recent studies have shown that S-TRACP5b is a marker of the number of osteoclasts rather than their activity. (17, 34–37) This was supported already in 1983 by the results of Stepan et al., (38) who showed that plasma TRACP activity decreased to normal limits after removal of parathyroid adenoma from primary hyperparathyroidism patients. Later in 2002, Stepan and Burckhardt(35) came to a similar conclusion in their study of osteoporotic men with Klinefelter's syndrome treated with ibandronate. This conclusion was also supported by our previous findings with in vitro cultures of nonresorbing osteoclasts cultured on plastic surface, where TRACP 5b activity secreted into the culture medium correlated significantly with the number of multinucleated osteoclasts formed from bone marrow-derived osteoclast precursor cells during the culture period. (34) Also, Chu et al. (17) have shown that serum TRACP 5b activity correlates with osteoclast number determined histomorphometrically from bone biopsies of patients with renal bone disease. The most convincing evidence to support this hypothesis was obtained from studies with various osteopetrotic rat strains and human patients of type-2 autosomal dominant osteopetrosis (ADO2). (36, 37) In both studies, S-TRACP5b was significantly elevated in subjects with osteopetrosis who had highly elevated numbers of osteoclasts that were functionally inactive and not able to resorb bone.
Recently, Hannon et al. (25) published a study where they analyzed clinical performance of S-TRACP5b. Similarly to this study, their study included a comparison of S-TRACP5b and S-CTX for monitoring alendronate treatment, but with a smaller number of subjects. Their results were in parallel to what we show here (i.e., greater decrease for S-CTX, but a higher signal-to-noise ratio for S-TRACP5b). They concluded that the reason for this result was the observed lower biological variability of S-TRACP5b. In this study, both the signal-to-noise ratio and the clinical sensitivity of S-CTX were lower than those of S-TRACP5b. Our results together with the results of Hannon et al. show that the markers that decrease most are not necessarily the best markers for monitoring the response to antiresorptive treatment, because high analytical and biological variability of a marker may disturb its clinical performance.
The study described here was a placebo-controlled study where all subjects in both the alendronate and placebo groups received daily supplementation of vitamin D and calcium. Thus, it was expected that a reduction in bone turnover would occur also in the placebo group, although to a lesser extent than in the alendronate group. When we calculated the CVi values, we used the change of marker values at 3 months compared with baseline from the individuals in the placebo group. This way we were able to eliminate the decrease of bone turnover caused by vitamin D and calcium supplementation from the total change of bone turnover observed in the alendronate group, allowing us to calculate the effects caused by alendronate treatment alone. Because of this elimination, the signal-to-noise ratios that we observed for S-TRACP5b and S-CTX were lower than those observed by Hannon et al., (25) because they used CVi values from healthy untreated individuals in their calculations. In fact, the “noise” we used is a combination of the analytical and biological variability of the markers and their change caused by vitamin D and calcium, whereas the noise used by Hannon et al. is only the analytical and biological variability of the markers.
In this study, four of the markers measured, S-TRACP5b, S-PINP, S-OC and S-BALP, were measured with manual immunoassays. U-DPD was determined with the “gold standard” high-performance liquid chromatography method, and S-CTX with an automated analyzer. All six markers had an excellent clinical specificity. Although the specificity of S-TRACP5b (89.0%) was the lowest of all markers measured, it was still very good and considerably higher than the best clinical sensitivity observed. Clinical sensitivity, AUC, and signal-to-noise ratio were higher for S-TRACP5b than for any of the other markers tested.
We also compared the use of LBMD with bone markers for monitoring alendronate treatment, and noticed that the clinical sensitivity, AUC, and signal-to-noise ratio of LBMD were all substantially lower than those of the best bone markers, even though the LBMD measurements were performed at 12 months after the initiation of treatment compared with the bone marker measurements performed at 3 months. This clearly shows that bone markers are better tools than BMD for monitoring short-term effects of antiresorptive treatment on skeletal homeostasis, although it has to be kept in mind that long-term changes in bone structure are always needed to explain the changes in risk of fracture.
Based on the clinical performance of the markers, S-TRACP5b, S-CTX, and S-PINP were the best markers for monitoring alendronate treatment in this study. Using S-TRACP5b, S-CTX, and S-PINP together for monitoring the efficacy of antiresorptive treatment could be recommended, because they all show an excellent clinical performance, and they all measure different aspects of bone turnover, S-TRACP5b being a marker of osteoclast number, S-CTX a marker of bone collagen degradation, and S-PINP a marker of bone formation. Measuring bone formation markers together with bone resorption markers would detect disturbances in coupling of bone formation and resorption, and by dividing S-CTX values by S-TRACP5b values, it might be possible to obtain an index reflecting the mean activity of an individual osteoclast.
These results show that S-TRACP5b is a useful marker for monitoring alendronate treatment. However, more studies are needed to elucidate the clinical value of S-TRACP5b as a marker of bone resorption.
The authors thank Merck & Co for grant support and for providing the alendronate for the study and the Medical Research Fund of Tampere University Hospital and the Academy of Finland for grant support.
- 11982 Bone acid phosphatase: Tartrate-resistant acid phosphatase as a marker of osteoclast function. Calcif Tissue Int 34:285–290.
- 21995 Immunohistochemical detection of tartrate-resistant acid phosphatase in non-hematopoietic human tissues. Am J Clin Pathol 104:397–402., , , ,
- 32001 Tartrate-resistant acid phosphatase (Acp 5): Identification in diverse human tissues and dendritic cells. J Histochem Cytochem 49:675–684., , ,
- 41993 Purification and properties of the native form of the purple acid phosphatase from bovine spleen. Biochemistry 32:8120–8129., , ,
- 51999 Tartrate-resistant purple acid phosphatase is synthesized as a latent proenzyme and activated by cysteine proteinases. Biochem J 343:63–69., ,
- 62001 The highly exposed loop region in mammalian purple acid phosphatase controls the catalytic activity. ChemBioChem 2:355–363., , , ,
- 71978 Biochemical properties of tartrate-resistant acid phosphatase in serum of adults and children. Clin Chem 24:1105–1108., , ,
- 81981 Comparison of the tartrate-resistant acid phosphatase in Gaucher's disease and leukemic reticuloendotheliosis. Clin Biochem 14:177–181., , ,
- 92002 Serum tartrate-resistant acid phosphatase isoforms in rheumatoid arthritis. Clin Chim Acta 320:49–58., , , , ,
- 102002 Stable expression of human tartrate-resistant acid phosphatase isoforms by CHO cells. Clin Chim Acta 326:113–122., , , ,
- 111996 Human bone tartrate-resistant acid phosphatase: Purification and development of an immunoassay. J Bone Miner Res 11:1444–1452., , ,
- 122001 Tartrate-resistant acid phosphatase isoform 5b as serum marker for osteoclastic activity. Clin Chem 47:74–80., , ,
- 132000 Development of a kinetic assay for band 5b tartrate-resistant acid phosphatase activity in serum. Clin Chem 46:469–473., , , , ,
- 142001 Naphtol-ASBI phosphate as a preferred substrate for tartrate-resistant acid phosphatase isoform 5b. J Bone Miner Res 16:788–793., , ,
- 152000 Tartrate-resistant acid phosphatase 5b: A novel serum marker of bone resorption. J Bone Miner Res 15:1337–1345., , , , ,
- 162001 Serum Tartrate-resistant acid phosphatase is a specific and sensitive marker of bone resorption. Clin Chem 47:597–600., , , , ,
- 172003 Correlation between histomorphometric parameters of bone resorption and serum type 5b tartrate-resistant acid phosphatase in uremic patients on maintenance hemodialysis. Am J Kidney Dis 41:1052–1059., , , ,
- 182003 Comparison of serum bone resorption markers in the diagnosis of skeletal metastasis. Anticancer Res 23:4095–4099., ,
- 192003 Tartrate-resistant acid phosphatase as serum marker of bone metabolism in cancer patients. Anticancer Res 23:2783–2788., , , , , ,
- 202003 Tartrate-resistant acid phosphatase isoform 5b; a novel serum marker for monitoring bone disease in multiple myeloma. Int J Cancer 106:455–457., , , , , , , ,
- 212002 Serum levels of tartrate-resistant acid phosphatase-5b in breast cancer patients treated with pamidronate. Int J Biol Markers 17:253–258., , , , , , , , ,
- 222003 Bone metabolic markers in bisphosphonate therapy for skeletal metastases in patients with breast cancer. Breast Cancer 10:21–27., ,
- 232003 Pamidronate is superior to ibandronate in decreasing bone resorption, interleukin-6 and beta 2-microglobulin in multiple myeloma. Eur J Haematol 70:34–42., , , , , , , , , ,
- 242003 Pamidronate is an effective treatment for osteoporosis in patients with beta-thalassaemia. Br J Haematol 1234:730–737., , , , , ,
- 252004 Clinical performance of immunoreactive tartrate resistant acid phosphatase isoform 5b as a marker of bone resorption. Bone 34:187–194., , , , ,
- 262002 Serum tartrate-resistant acid phosphatase 5b, but not 5a, correlates with other markers of bone turnover and bone mineral density. Calcif Tissue Int 71:20–25., , , , , , ,
- 272004 Biochemical markers of bone metabolism and prediction of fracture in elderly women. J Bone Miner Res 19:386–393., , , , , , , , ,
- 282003 Effect of alendronate and exercise on bone and physical performance of postmenopausal women: A randomized controlled trial. Bone 33:132–143., , , , , , , , , ,
- 292000 Development and evaluation of three immunofluorometric assays that measure different forms of osteocalcin in serum. Clin Chem 46:332–337., , , , , , ,
- 301988 Quantitative analysis of the pyridinium crosslinks of collagen in urine using ion-paired reversed-phase high-performance liquid chromatography. Anal Biochem 169:197–203., ,
- 311996 Estimation of various mechanical characteristics of human bone using dual energy x-ray absorptiometry. Bone 18:17–28., , , , ,
- 322003 Disease-specific expression of tartrate-resistant acid phosphatase isoforms. J Bone Miner Res 18:1916–1919., , ,
- 331998 Response of biochemical markers of bone turnover to hormone replacement therapy: Impact of biological variability. J Bone Miner Res 13:1124–1133., , ,
- 342000 Rapid screening method for osteoclast differentiation in vitro that measures tartrate-resistant acid phosphatase 5b activity secreted into the culture medium. Clin Chem 46:1751–1754., , , ,
- 352002 Serum activity of type 5b ACP and biochemical markers of type I collagen degradation in osteoporotic men with Klinefelter's syndrome treated with an intravenous ibandronate. Calcif Tissue Int 70:279.,
- 362003 Serum tartrate-resistant acid phosphatase 5b and osteocalcin in naturally occurring osteopetrotic rats. J Bone Miner Res 18S1:SA095., , , , ,
- 372003 Osteoclast-derived serum tartrate-resistant acid phosphatase 5b in Albers-Schönbergs disease (type II autosomal dominant osteopetrosis). Clin Chem 50:883–890., , , , ,
- 381983 Relationship of plasma tartrate resistant acid phosphatase to the bone isoenzyme of serum alkaline phosphatase in hyperparathyroidism. Clin Chim Acta 133:189–200., , , , , , , ,