Introduction. Nephrol Dial Transplant (2012) 27: doi: /ndt/gfr535 Advance Access publication 22 September 2011

Transcription

1 Nephrol Dial Transplant (2012) 27: doi: /ndt/gfr535 Advance Access publication 22 September 2011 Interpretation of serum PTH concentrations with different kits in dialysis according to the KDIGO guidelines: importance of the reference () values Etienne Cavalier 1, Pierre Delanaye 2, Laura Vranken 1, Anne-Catherine Bekaert 1, Agnès Carlisi 1, Jean-Paul Chapelle 1 and Jean-Claude Souberbielle 3 1 Department of Clinical Chemistry, University of Liège, CHU Sart-Tilman, Liège, Belgium, 2 Department of Nephrology, Dialysis and Hypertension, University of Liège, CHU Sart-Tilman, Liège, Belgium and 3 Laboratoire d explorations fonctionnelles, Hôpital Necker- Enfants malades, Assistance Publique des Hôpitaux de Paris (APHP), and INSERM U845, Paris, France Correspondence and offprint requests to: Etienne Cavalier; Etienne.cavalier@chu.ulg.ac.be Abstract Background. The recommended target range for serum parathyroid hormone (PTH) in dialysis has changed from 150 to 300 pg/ml in the KDOQI guidelines to two to nine times the upper limit in the KDIGO ones. Although inclusion/exclusion criteria for the reference population are highly important, they are usually not mentioned in the commercial kits. In this study, we used the same reference population of vitamin D-replete subjects to establish reference values for 10 commercial PTH kits. We evaluated whether this may improve the classification of dialysis according to the KDIGO compared to the use of reference values proposed by the manufacturers. Methods. We measured serum PTH with 10 different kits in 149 haemodialysis, and OH-vitamin D-replete (>75 nmol/l) individuals with an estimated glomerular filtration rate >60 ml/min/1.73 m 2. Results. For the 10 kits, our upper limit was lower than those of the manufacturers. The difference was, however, variable from one kit to another. The two kits that yielded the lowest and the highest absolute concentrations classified differently 84/149 (56.4%) according to the KDOQI and 53/149 (36.2%) according to the KDIGO using the manufacturers values. Using our values significantly decreased the discrepancies with 24/149 (16.1%) being still classified differently. Taking the measurement uncertainty into consideration, 8% of the only remained differently classified by these two kits. Conclusions. Using the same vitamin-d-replete population to establish the reference range for 10 commercial PTH kits significantly improved the classification of haemodialysis according to the KDIGO target range. Keywords: haemodialysis; KDIGO; parathyroid hormone; reference values; vitamin D Introduction Parathyroid hormone (PTH) determination is routinely performed in suffering from chronic kidney diseases (CKDs) for the diagnosis and management of mineral bone diseases. Indeed, both too high and too low PTH concentrations are better avoided in with CKD, leading the experts to propose an optimal range for PTH serum concentrations. The recommended target range for serum PTH in dialysis has changed from 150 to 300 pg/ml in the KDOQI guidelines [1] to two to nine times the upper limit of in the KDIGO ones [2]. One should remember beforehand that the KDOQI target range for serum PTH concentrations had been derived from studies that compared, during the late 80s early 90s, bone biopsy data from dialysis to serum PTH concentrations measured with the (no longer available) Allegro intact PTH assay [3]. Since then, it has been reported that any kind of bone turnover could be found for PTH concentrations between (grossly) 100 and 500 pg/ml [4]. Other studies reported that, in dialysis, PTH concentrations were associated with mortality only for the highest concentrations (> pg/ml) [5, 6]. For these reasons, the KDOQI PTH range ( pg/ ml) has been expanded in the KDIGO (corresponding for example to pg/ml with the Allegro assay when the manufacturer s upper limit of 65 pg/ml is considered). Another point to discuss is why the KDIGO proposes a target PTH range based on multiples of the upper limit rather than absolute concentrations such as in the KDOQI. To understand that, the inter-method variability in PTH measurement [7 9] should be considered. Indeed, while the KDOQI proposed the same PTH target range whatever the assay used, the PTH concentration of a given serum may vary by a factor >2 depending on the assay used, potentially inducing opposite diagnostic/therapeutic attitudes [8]. As the standardization of PTH assays is highly Ó The Author Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oup.com

2 PTH reference range and KDIGO guidelines 1951 difficult to achieve [10], the proposition of the KDIGO work group to use a target range for serum PTH based on the upper limit of the values is a pragmatic and elegant way to overcome this problem of inter-method variability in PTH assays. We think, however, that discussions about PTH reference values are needed. Indeed, reference values for serum PTH concentrations are generally obtained by measuring PTH in a population of apparently healthy subjects. Exclusion criteria for this population are highly important and should correspond to any potential cause of altered PTH secretion, including vitamin D insufficiency. This last point is important because, according to most reports [11], vitamin D insufficiency is very frequent in the general population and thus should be prevalent in an otherwise apparently healthy group recruited to establish reference values for PTH. However, excluding vitamin D-insufficient subjects from the reference group requires measuring the 25-hydroxy vitamin D (25OHD) concentration beforehand in all subjects, a practice which complicates the establishment of PTH reference values and which was not considered in most studies which provided serum PTH reference values for different immunoassays [12 14]. Some years ago, however, we demonstrated that excluding subjects with low serum 25OHD concentrations from a reference population decreased the upper limit for serum PTH by 25 35% depending on the assay considered [15]. Thus, for a given PTH assay, the values (and consequently the KDIGO target range) may significantly vary, depending on the reference population that has been recruited, and especially whether the vitamin D status has been taken into account. In the present study, we used the same reference population of vitamin D-replete subjects to establish reference values for 10 commercial PTH kits. Our aim was to evaluate whether this may improve the classification of dialysis according to the KDIGO target range compared to the use of reference values proposed by the manufacturers. Material and methods Dialysis We studied 149 (63 women, 86 men) aged years (extreme: years), undergoing dialysis three times a week. Blood was obtained just before a dialysis session and was centrifuged within 30 min of blood sampling. Serum was aliquoted and stored at 80 C until assayed. Reference population One hundred and twenty women aged years (min max: years), and 120 men aged years (min max: years) served as a reference population. All were Caucasians, apparently healthy, and were supplemented with vitamin D3 at various doses. Inclusion criteria were a 25OHD concentration (DiaSorin Liaison) of 75 nmol/l, serum calcium and phosphate concentrations comprised between 2.15 and 2.60 and 0.74 and 1.51 mmol/l, respectively, and an estimated glomerular filtration rate (GFR) (MDRD formula) of >60 ml/min/1.73 m 2. The use of drugs known to influence bone and calcium/phosphorus metabolism (bisphosphonate, active vitamin D compounds, anticonvulsants, thiazidic and loop diuretics) was an exclusion criteria. PTH assays Assay characteristics as provided by the manufacturers are indicated in Table 1. With each of these assays, PTH was measured according to the recommendations of the respective manufacturers. All measurements were done at the Clinical Chemistry laboratory of the Liège University hospital. This Laboratory is accredited by ISO Methodology Firstly, we classified the dialysis according to the KDOQI guidelines. Secondly, we show the classification of the with the new KDIGO guidelines, by multiplying by a factor 2 and 9 the upper reference range of each PTH assay as provided by the manufacturers. Thirdly, we will use the upper value of the reference range that we have established in our reference population for each PTH assay to classify the dialysis according to the KDIGO guidelines. Finally, we will take the measurement uncertainty into consideration to classify the with our established reference range. Indeed, we are dealing here with different cut-offs obtained by an analytical process. One should know that the real true value of any analytical process is never known and that the results observed are only a rough estimation of this true value. Instead of printing this rough estimation on their protocols, laboratories could provide the frame where the real true value could be found, with a reasonable probability. This frame is called the measurement uncertainty. Determination of this measurement uncertainty is quite complicated. A rule of thumb consists of using the inter-assay coefficient of variation multiplied by a factor of 2 to have a rough estimation of the measurement uncertainty. In other words, if an analytical process has a coefficient of variation (CV) of 10% and if the result observed with this process is 100, the real true value could roughly be comprised between 80 and 120. Using this approach, we can estimate the grey zone that surrounds the multiples of the upper limits of the values found with the different kits. Statistics In our reference population, the ity of the distribution of the PTH concentrations was assessed by the Kolmogorov Smirnov test. With a given PTH kit, means of unpaired data ( subjects versus dialysis ) were compared with the Mann Whitney test. Percentages were compared with the chi-squared test. Agreement between kits was assessed with the kappa test. A P value <0.05 was considered as significant. Results As shown in Table 2, absolute PTH concentrations in our 149 dialysis varied greatly from one assay to another and were higher (P < ) than in our reference population. The kit that produced the highest values (Abbott Architect) and those that produced the lowest values (DiaSorin N-tact IRMA and DiaSorin Liaison 1-84) differed by a factor >2. According to the KDOQI target values, 84 of 149 (56.4%) were classified differently with the Abbott Architect and with the DiaSorin N-tact IRMA kits: 48 with a PTH <150 pg/ml with the DiaSorin N-tact IRMA had a concentration >150 pg/ml with the Abbott Architect and 14 of them were even >300 pg/ml. Two only were classified similarly by the two methods as being between 150 and 300 pg/ml, whereas 36 were comprised in the pg/ml by the Diasorin N-tact IRMA kit when they were >300 pg/ml with the Abbott Architect (Figure 1a). Using the upper limit provided by the manufacturers to determine the KDIGO range greatly improved the discrepancies in classifying the as shown in Table 2. The Abbott Architect and the DiaSorin N-tact IRMA were still the kits that yielded the highest discrepancy (53, i.e. 36.2%) in the classification of the dialysis according to the KDIGO range: 25 with a PTH below twice the upper limit with the DiaSorin N-tact IRMA were in fact between two and nine times the upper limit with the Architect, while 29 with a PTH concentration within two to nine times the upper limit with the DiaSorin N-tact IRMA had a value above nine times the upper with the Abbott Architect (Figure 1b).

3 Table 1. Characteristics of the PTH assays tested in the study, as provided by the manufacturers Name of the assay Manufacturer Intraassay CV (%) Interassay CV (%) Reference range (manufacturer) Reference population Automated Intact PTH Architect Abbott (Abbott Park, IL) <6.1 < plasma samples from apparently healthy adults Access PTH intact Beckman-Coulter (Brea, CA) <2.6 < paired samples (serum and plasma EDTA) from apparently healthy men and women aged years old. Exclusion of individuals with ab calcium, creatinine and 25-OH vitamin D levels. 2nd or 3rd generation Tracer Epitope, type and origin of coated antibody Epitope, type and orign of labelled antibody Detection limit Highest measurable value Yes 2nd Acridinium ester goat goat Yes 2nd Alkaline goat Monoclonal phosphatase N-tact PTH SP IRMA DiaSorin (Stillwater, MN) <3.6 < serum samples from apparently No 2nd 125 I healthy fasting young adults goat goat 1 34 Liaison N-tact DiaSorin (Stillwater, MN) <5.0 < healthy adults. Yes 2nd Isoluminol Intact PTH Vitros 5600 Ortho Clinical Diagnostics (Rochester, NY) <2.0 < EDTA, heparin plasma or serum from 240 presenting calcium, Yes 2nd Horseradish peroxidase goat 1 34 goat TSH, creatinine and vitamin D levels. <2.7 < Not specified Yes 2nd Ruthenium Monoclonal No 2nd 125 I Elecsys 2010 Roche (Mannheim, Germany) Total intact PTH IRMA Scantibodies (Shantee, CA) <4.8 < EDTA plasma samples from apparently healthy blood donors according to the National Clinical Chemistry Laboratory Standards (NCCLS) recommendations Immulite 2000Xpi Intact PTH Siemens Healthcare Diagnostics (Deerfield, IL). <5.7 < Serum from 255 apparently healthy Liaison 1-84 DiaSorin (Stillwater, MN) <5.9 < individuals with 25-OH vitamin D levels >75 nmol/l and serum calcium levels comprised between and mmol/l Ca-PTH IRMA Scantibodies (Shantee, CA) <4.9 < EDTA plasma samples from apparently healthy blood donors according to the National Clinical Chemistry Laboratory Standards (NCCLS) recommendations Yes 2nd Alkaline phosphatase Monoclonal murine Yes 3rd Isoluminol C-terminal No 3rd 125 I Monoclonal goat goat 1 34 N-terminal N-terminal goat E. Cavalier et al.

4 Table 2. Classification of 149 haemodialysed with 10 different PTH assays according to the KDOQI, the KDIGO (with the manufacturers published reference range) and the KDIGO with the reference values established in the laboratory on the same reference population Methods Abbott Architect Beckman Access DiaSorin N-tact IRMA DiaSorin Liaison N-tact Ortho Vitros Roche Elecsys Scantibodies Total intact PTH Siemens Immulite DiaSorin Liaison 1-84 Scantibodies Ca-PTH IRMA Mean 6 SD observed in the hemodialyzed Lowest highest observed value <150 pg/ml, pg/ml, >300 pg/ml, Reference range (manufacturer) KDIGO range according to the reference values provided by the manufacturer <23 upper 2 93 upper of >93 upper of Lower and upper reference limits (95% confidence interval) obtained in our reference population KDIGO range according to the upper value of the reference range established in our reference population <23 upper of 2 93 upper of >93 upper 2nd generation assays (18.1) 36 (24.2) 86 (57.7) (16.8) 84 (56.4) 40 (26.8) 16.3 ( ) to (16.1) 82 (55.0) 43 (28.9) 64.7 ( ) (26.8) 46 (30.9) 63 (42.3) (34.2) 86 (57.7) 12 (8.1) 10.1 ( ) to (17.5) 82 (55.0) 41 (27.5) 47.4 ( ) (50.3) 38 (25.5) 36 (24.2) (33.6) 88 (59.0) 11 (7.4) 7.2 ( ) to (24.2) 82 (55.0) 31 (20.8) 35.7 ( ) (23.5) 31 (20.8) 83 (55.7) (23.5) 94 (63.1) 20 (13.4) 21.3 ( ) to 68.2 ( ) (22.1) 91 (61.1) 25 (16.8) (28.9) 47 (31.5) 59 (39.6) (23.5) 80 (53.7) 34 (22.8) 10.8 ( ) to 47.5 ( ) (18.1) 84 (56.4) 34 (25.5) (26.9) 48 (32.2) 61 (40.9) (26.2) 85 (57.0) 25 (16.8) 13.7 ( ) to (18.8) 82 (55.0) 39 (26.2) 50.2 ( ) (28.2) 28 (18.8) 70 (53.0) (24.2) 77 (51.7) 36 (24.1) 7.8 ( ) to (18.8) 77 (51.7) 50 (29.5) 49.7 ( ) (24.8) 29 (19.5) 83 (55.7) (20.8) 79 (53.0) 39 (26.2) 5.4 ( ) to 57.1 ( ) 3rd generation assays (48.3) 46 (30.9) 31 (20.8) (26.2) 85 (57.0) 25 (16.8) 4.6 ( ) to 25.8 ( ) (43.6) 39 (26.2) 45 (30.2) (22.8) 77 (51.7) 38 (25.5) 6.8 ( ) to 30.8 ( ) (13.4) 78 (52.4) 51 (34.2) (18.1) 78 (52.4) 44 (29.5) (20.1) 70 (47.0) 49 (32.9) PTH reference range and KDIGO guidelines 1953

5 1954 E. Cavalier et al. There was no significant difference in PTH concentrations nor in age between men and women in our reference population. With all the tested PTH assays, the distribution of the concentrations was Gaussian and we thus used a parametric (mean 2SD) approach to calculate the reference values. These reference values are shown in Table 2 for each PTH kit. They were lower than the reference values provided by the manufacturers with an upper limit only slightly different for three kits (the Abbott Architect, DiaSorin Liaison N-tact and Ortho Vitros kits) but frankly lower ( % lower) for the seven other kits. The difference was most important for the Beckman Access kit ( 46.1%), the DiaSorin N-tact IRMA ( 33.9%) and the DiaSorin Liaison 1-84 kit ( 32.8%). We calculated the KDIGO target range with the upper limit of our reference values to classify the. The kappa test was used to evaluate the inter-rater agreement between each technique, with our upper limit or the manufacturer s published one (Table 3). The mean kappa was significantly higher with our reference range (kappa ¼ versus 0.749, P < 0.05), showing a moderate but significant improvement in the classification of the. The Abbott Architect and the DiaSorin N-tact IRMA kits were still those which showed the most discrepant classification with 24 (16.1%) classified differently: 12 with a PTH value below twice the upper limit with the DiaSorin N-tact IRMA had a value between two and nine times the upper limit with the Abbott Architect kit, while 12 with a PTH within two to nine times the upper limit with the DiaSorin N-tact IRMA kit had a value more than nine times the upper limit with the Abbott Architect kit (Figure 1c). With these two kits, the improvement in the classification of the (24 classified differently with our reference values versus 53 with the manufacturer s reference values) was significant (chi-squared ¼ 13.7; P ¼ ). Using analytical CVs of 5% as indicated by the respective manufacturers for PTH values >50 pg/ml, we calculated the measurement uncertainty to determine the grey zone around the cut-offs (two to nine times upper ) with the DiaSorin IRMA and Abbott Architect (Figure 1d). By doing this, 12 (8%) only remained differently classified by the two kits. Discussion In this study, we confirmed that, depending on the PTH assay kit, a dialysis patient could be classified as having a PTH concentration either below or above the target range of the former KDOQI guidelines ( pg/ml). On the contrary, our data show a better inter-method agreement between Fig. 1. Results obtained in 149 dialysis with the two kits that yield the most discrepant results, the DiaSorin IRMA (x axis) and Abbott Architect (y axis). (a) Shows the classification according to the KDOQI, (b) The classification with the KDIGO with the upper limit provided by the manufacturers and (c)the classification with the KDIGO with the upper limit determined in our reference population. In the uncoloured zones (white zones), the are similarly classified by the two kits. The zones in light grey correspond to the zones where discordances between the classifications are observed. Part (d) corresponds to part (c), but the measurement uncertainty has been taken into consideration. The zones in dark grey correspond to the measurement uncertainty at the two levels (upper reference range multiplied by 2 and 9). In this figure, eight only remain in the light grey zones of discordance.

6 PTH reference range and KDIGO guidelines 1955 Table 3. Kappa tests for agreement between the different methods when the laboratory (bold) or the manufacturer s (italic) reference ranges are used a a The overall agreement is significantly higher among the different methods when the laboratory s reference ranges are used (kappa ¼ versus 0.749, P < 0.05). all the tested PTH assays with regard to the KDIGO guidelines (maintain PTH concentration within two to nine times the upper limit of ). This was even significantly improved when the same vitamin D-repleted population was used to establish reference values for PTH with all the tested kits. Nevertheless, with some of these kits (like the Beckman Access, the DiaSorin N-tact IRMA, Roche Elecsys or Dia- Sorin Liaison 1-84) using our upper limit or the manufacturer s published one had an important impact on the classification of the. Indeed, the classified as low or high (less than two times or more than nine times) greatly varied according to the upper limit of the reference range that was used. In our opinion, these large differences in the obtained upper reference range (and in the classification) are due to the lack of robustness of most of the manufacturers reference ranges. Indeed, as shown in Table 1, the recruited is sometimes insufficient; there is, for some of the assays, no indication on the vitamin D status, on the creatinine concentrations or on the ethnic origin of the. One manufacturer does not even specify how the reference ranges that he proposes have been obtained. In this study, we used the same large and very well-defined reference population to calculate our reference limits. This allowed us to present a better harmonization in the classification of the haemodialysed when we used these robust limits. This assertion is also true whether these PTH kits are automated chemiluminescent or IRMA methods, from the second or the third generation. This point is particularly important from a clinical point of view. Indeed, it means that whatever the PTH method used by the laboratory, and if the are followed-up with the same PTH method, therapeutic decisions can virtually be the same in all the dialysis centres that would follow the KDIGO guidelines. This was not true in the past. At this point of the discussion, several important remarks have to be formulated. Firstly, it was not the aim of this study to evaluate the performance of different guidelines to correctly classify the according to their bone disease status. For that purpose, a bone biopsy is needed. Secondly, one might have the feeling that, as the KDIGO guidelines are much broader than the KDOQI ones, it is obvious that the classification of the is naturally better according to the KDIGO. The thing that has to be pointed out in this study is not that much the importance of the value of the multiple of the upper limit that is used (i.e. two, four or nine times), but rather the fact that, by doing so, the classification of the is greatly harmonized. Indeed, with the KDOQI, 43.6% of the will receive the same treatment if the laboratory used the Abbott Architect or the DiaSorin IRMA, whereas this percentage will increase to 76.5% or even 83.9% with different cut-offs using the upper limits of the kits (two to four times or two to nine times, respectively). Analytical validation is, however, not the only point that should be taken into consideration for the interpretation of a laboratory result. Indeed, the intra-individual variation is a parameter of importance and the intra-individual variability for PTH in haemodialysed has recently been estimated at 25.6% [16]. Taking this biological variation into account is highly important when two consecutive PTH concentrations are compared in a given patient. Establishment of reference values is not an easy task [17]. Indeed, many points must be taken into consideration.

7 1956 E. Cavalier et al. Firstly, it has been shown that using serum or EDTAplasma for PTH determination could lead to some discrepancies [7, 18, 19]. This is particularly true with the Siemens Immulite method, and the manufacturer even proposes different reference ranges for serum or EDTA-plasma samples. All our tests have been performed on serum. Laboratories should thus use one type of samples according to their best routine practice [18] and avoid the use of both types of samples. Secondly, until now, there is no universally recognized International Standard for the calibration of the PTH assay kits. The 95/646 International Standard is currently being evaluated, and studies on its commutability (i.e. its ability to be used with different kits) are ongoing. In the future, if this standard is universally accepted, all the methods could be calibrated against this unique material and the reference ranges should thus be re-evaluated for all the methods. Finally, different authors have also used vitamin D-replete populations to publish PTH reference values for different assay kits [20, 21]. Their results were sometimes different than those presented in this study. Indeed, several other determinants of PTH concentrations should also be taken into consideration, such as age, GFR, calcium intake, ethnic origin and body mass index, as indicated by the expert panel who published the last recommendations for the diagnosis of asymptomatic primary hyperparathyroidism [22]. So, in our opinion, an important multicentre work should be performed to recruit a very extensive reference population of apparently healthy vitamin D-replete subjects in order to establish the PTH reference range for all the available PTH kits, stratifying the data according to the above-mentioned determinants of PTH concentration. Special attention should be paid in recruiting a sufficient subjects from different ethnic groups as, for example, African American and Caucasians are known to differ in term of PTH secretion [23]. Similarly, calcium intake should be recorded as well as any drug or multi-vitamin supplement potentially interfering with PTH concentration. Exclusion criteria should include more products than in the present study, such as nicotinic acid, which has been shown to lower serum phosphate [24] or biotin, which may considerably influence PTH measurement with some kits [25]. In conclusion, we have shown that with 10 different methods for PTH determination, a global overall agreement in the classification of Stage 5 CKD could be achieved with the KDIGO guidelines when the same vitamin D- replete population was used to establish the reference range with these 10 tested methods. Conflict of interest statement. None declared. References 1. KDOQI clinical practice guidelines for bone metabolism and disease in chronic kidney disease. Am J Kidney Dis 2003; 42: S1 S Kidney Disease: Improving Global Outcomes (KDIGO) CKD-MBD work group KDIGO clinical practice guideline for the diagnosis, evaluation, prevention and treatment of Chronic Kidney Disease-Mineral and Bone Disorder (CKD-MBD). Kidney Int 2009; 6: S1 S Sherrard DJ, Hercz G, Pei Y et al. The spectrum of bone disease in endstage renal failure an evolving disorder. Kidney Int 1993; 43: Barreto FC, Barreto DV, Moyses RM et al. KDOQI-recommended intact PTH levels do not prevent low-turnover bone disease in hemodialysis. Kidney Int 2008; 73: Block GA, Klassen PS, Lazarus JM et al. Mineral metabolism, mortality, and morbidity in maintenance hemodialysis. J Am Soc Nephrol 2004; 15: Kalantar-Zadeh K, Kuwae N, Regidor DL et al. Survival predictability of time-varying indicators of bone disease in maintenance hemodialysis. Kidney Int 2006; 70: Joly D, Drueke TB, Alberti C et al. Variation in serum and plasma PTH levels in second-generation assays in hemodialysis : a cross-sectional study. Am J Kidney Dis 2008; 51: Souberbielle JC, Boutten A, Carlier MC et al. Inter-method variability in PTH measurement: implication for the care of CKD. Kidney Int 2006; 70: Cantor T, Yang Z, Caraiani N, Ilamathi E. Lack of comparability of intact parathyroid hormone measurements among commercial assays for end-stage renal disease : implication for treatment decisions. Clin Chem 2006; 52: Souberbielle JC, Roth H, Fouque DP. Parathyroid hormone measurement in CKD. Kidney Int 2010; 77: Mithal A, Wahl DA, Bonjour JP et al. Global vitamin D status and determinants of hypovitaminosis D. Osteoporos Int 2009; 20: Blind E, Schmidt-Gayk H, Scharla S et al. Two-site assay of intact parathyroid hormone in the investigation of primary hyperparathyroidism and other disorders of calcium metabolism compared with a midregion assay. J Clin Endocrinol Metab 1988; 67: Endres DB, Villanueva R, Sharp CF Jr, Singer FR. Immunochemiluminometric and immunoradiometric determinations of intact and immunoreactive parathyrin: performance in the differential diagnosis of hypercalcemia and hypoparathyroidism. Clin Chem 1991; 37: Nussbaum SR, Zahradnik RJ, Lavigne JR et al. Highly sensitive two-site immunoradiometric assay of parathyrin, and its clinical utility in evaluating with hypercalcemia. Clin Chem 1987; 33: Souberbielle JC, Cormier C, Kindermans C et al. Vitamin D status and redefining serum parathyroid hormone reference range in the elderly. J Clin Endocrinol Metab 2001; 86: Gardham C, Stevens PE, Delaney MP et al. Variability of parathyroid hormone and other markers of bone mineral metabolism in receiving hemodialysis. Clin J Am Soc Nephrol 2010; 5: Cavalier E, Delanaye P. Defining a reference population : no easy task. J Bone Miner Res 2009; 24: Cavalier E, Delanaye P, Carlisi A et al. Stability of intact parathyroid hormone in samples from hemodialysis. Kidney Int 2007; 72: Cavalier E, Delanaye P, Hubert P et al. Estimation of the stability of parathyroid hormone when stored at -80 degrees C for a long period. Clin J Am Soc Nephrol 2009; 4: La ulu SL, Roberts WL. Performance characteristics of six intact parathyroid hormone assays. Am J Clin Pathol 2010; 134: Rejnmark L, Vestergaard P, Heickendorff L, Mosekilde L. Determinants of plasma PTH and their implication for defining a reference interval. Clin Endocrinol (Oxf) 2011; 74: Eastell R, Arnold A, Brandi ML et al. Diagnosis of asymptomatic primary hyperparathyroidism: proceedings of the third international workshop. J Clin Endocrinol Metab 2009; 94: Sawaya BP, Butros R, Naqvi S et al. Differences in bone turnover and intact PTH levels between African American and Caucasian with end-stage renal disease. Kidney Int 2003; 64: Müller D, Mehling H, Otto B et al. Niacin lowers serum phosphate and increases HDL cholesterol in dialysis. Clin J Am Soc Nephrol 2007; 2: Meany D, Jan de Beur S, Bill MJ et al. A case of renal osteodystrophy with unexpected serum intact parathyroid hormone concentration. Clin Chem 2009; 55: Received for publication: ; Accepted in revised form: