Evaluation of Serum Bone Turnover Markers in Renal Dialysis Patients

E.T. Leary¹, M.K. McLaughlin¹, D. Swezey¹, T. Aggoune¹, T.H. Carlson¹, A.P. Foster²

¹Pacific Biometrics, Inc., Seattle, WA ²Roche Diagnostics, Corporation, Indianapolis, IN

Association of Bone and Mineral Research
ASBMR 22nd Annual Meeting, Toronto, Canada
September 2000
Poster Presentation
Abstract: JBMR V15 S1, Sept 2000;S525


Chronic renal dialysis patients have compromised bone status and increased bone turnover. This patient group presents special challenges in bone metabolism assessment. More recently, bone turnover markers (BTM) have demonstrated clinical utility in renal patients in spite of the retention and accumulation of some markers. This study compares recently available serum BTMs in 50 renal dialysis patients (21 males and 29 females) age 20 to 85 yrs with dialysis duration of 1 mo to 19 yrs (most under 3 yrs). Fasting serum and plasma samples were collected in the morning. The samples were processed and frozen within 3 hrs and stored at -70°C until analysis for PTH and the following BTMs: serum N-telopeptides by ELISA (sNTx; Ostex International); serum C-telopeptides by Elecsys [sCTx(r); Roche Diagnostics] and by ELISA [sCTx(o); Osteometer BioTech A/S]; bone specific alkaline phosphatase by ELISA (BSAP; Metra Biosystems); osteocalcin by Elecsys [OC(r); Roche Diagnostics] and by IRMA [OC(c); ELSA-OSTEO; CIS bio International]. PTH was by Elecsys [PTH(r); Roche Diagnostics] and by chemiluminescence [PTH(n); Nichols Diagnostics]. In addition, albumin, calcium, creatinine, phosphorous and alkaline phosphatase were measured.

PTH and BTM values in this population were greatly elevated. Overall, the frequency of marker results exceeding the healthy normal ranges were >90% in sNTx, sCTx and OC, >70% in PTH and >30% in BSAP. Comparisons among the BTMs and PTH were listed below as correlation coefficients (r).

[table “table1” not found /]

The correlation between the resorption markers (sNTX and sCTx) was strong. The correlation between the formation markers, BSAP and OC, and between the formation and resorption markers is poorer, partly due to different renal retention. Assays for the same analyte agreed well, however, the greatly elevated marker values present special demands on the assay design, such as analyte dynamic range, dilution linearity, assay precision and specificity and user friendliness. For example, further dilution was required in 80% of the samples in the sNTx assay, 26% in sCTx (o) and only 2% (one sample) in sCTx(r).

In summary, recently available serum BTMs may be useful in monitoring bone turnover status in the renal dialysis patients. Well-defined sample collection protocol and properly selected assays are important. Stratifying the patients using PTH concentration and normalizing marker values with creatinine may enhance the discriminatory power of the BTM.


Chronic renal failure patients on hemodialysis have compromised bone status. The standard tools for monitoring bone metabolism of these patients are invasive and expensive. This study investigates the potential of emerging serum markers of bone turnover as monitoring tools using a cohort of 50 renal dialysis patients.



  • 50 chronic hemodialysis patients with renal failure (21 males and 29 females) age 20-85 yr participated in the study.
  • Dialysis treatment periods spanned 1 mo to 19 yrs with most subjects undergoing treatment for less than 3 yr.
  • Fasting blood samples were collected in the morning prior to one of three times weekly dialysis sessions.
  • Serum, sodium heparin plasma and K3EDTA plasma were obtained from each subject, processed within 3 hr of collection and stored at -70ºC until analysis.
  • Urine specimens were not available.

Analytical Methods 
Biomarker Methods Compared

[table “table2” not found /]

* For investigational use only.

  • In addition to the analytes listed above, albumin, calcium, creatinine and phosphorous were analyzed


Method Comparison

  • In spite of different test formats, the correlation between marker methods that measure the same epitopes was excellent.
  • Linear regression analysis gave a correlation coefficients (r) of 0.954 for sCTx (r) and sCTx (o), 0.966 for OC (r) and OC (c), and 0.992 for PTH (r) and PTH (n).
  • The assays, however, differed in test design and performance specifics (e.g. sample stability, assay dynamic range, precision and specificity, calibration frequency, turnaround time), and these have marked practical significance.
  • Because of possible renal retention of the analytes of interest and because of potential interferents (e.g., osteocalcin or PTH fragments), renal samples present special analytical challenges, for example, assay specificity and the need for wide a dynamic range.
  • OC (r) and OC (c) both demonstrated good dilution linearity in this cohort and sample stability (24 hr at 4°C).

Marker Concentrations in Renal Dialysis Patients 

  • Circulating bone turnover markers that are excreted through the kidneys were greatly elevated in the hemodialysis patients.
  • Predictably, PTH was greatly elevated.
  • Four patients with elevated TAP values (in red) also exhibited greatly elevated GGT, but not necessarily increased BSAP.
  • The percent of samples that required dilution and reanalysis in this study was significant (listed below). These sample treatments have considerable practical implications.

sNTx-80%; sCTx (o) – 26%; OC (r) – 20%; OC (c) – 12%; sCTx (r), PTH (r) and PTH (c) – 2%; BSAP and TAP – 0%.

Marker Comparison [shown as correlation coefficients (r) on 50 subjects]

[table “table3” not found /]

*n=49. One subject with PTH level>20,000 pg/mL by both PTH assays was excluded from the statistical analysis (these data were included in the table in the abstract). This subject had normal BAP (27.6 U/L), TAP (70 IU/L), but elevated CTx, NTx and OC. **n=46. Four subjects with elevated TAP that did not fit the among-test correlation pattern were excluded from the estatistical analysis.

Correlation Among Markers with and without Renal Retention

  • sNTx ans sCTx correlated well (r = 0.852).
  • The correlation between serum telopeptides and BSAP was similar to that between OC and BSAP (r ~ 0.60). Renally cleared markers (telopeptide and OC) gave better correlations (r=0.79 – 0.90).
  • TAP correlation with BSAP was similar to that of other markers with BSAP (r ~0.60).
  • When the ‘outlier’ TAP specimens were excluded, correlation between TAP and all other markers significantly improved. The excluded patients had greatly elevated GGT, indicating hepatic involvement.
  • Adjustment of OC results by normalization with serum creatinine did not improve the correlation with BSAP or other markers.
  • In contrast to earlier reports, little correlation between PTH and BSAP or TAP was observed.

Marker Comparison – Renal vs Non-renally Impaired

  • The relationship among markers
  • was simliar for the current dialysis population and the non-renally impaired population of comparable marker concentrations (n=185). Correlation (r) between OC and BSAP from healthy males, and premenopausal and postmenopausal women, and Paget’s patients generally ranged below 0.4.


  • The resorption markers, serum CTx and NTx correlated well and are greatly elevated in the dialysis cohort.
  • The formation markers, BSAP and TAP, which are not retained by the kidney and OC, which is retained by the kidney are moderately correlated with the resorption markers
  • TAP may not be as useful as BSAP as a marker.
  • The extent of marker elevation that is secondary to renal retention is unknown. This complicates interpretation of the data.
  • Similar marker patterns between sCTx, sNTx, OC and BSAP are observed in this dialysis cohort and patients wihout renal impairment. Bone markers are potentially useful in assessing bone turnover status in renally impaired subjects.
  • The bone marker assay design is especially important for use in this cohort. The Elecsys assays are precise, robust and easy to use.