Pierre Lukas

Evaluation of automated immunoassays for 25(OH)-vitamin D determination in different critical populations before and after standardization of the assays

INTRODUCTION Standardization of immunoassays for 25(OH)-vitamin D determination is a major problem in clinical practice. A worldwide standardization program has started to address this and will reduce the bias observed between immunoassays. We aimed to calibrate 5 immunoassays on a LC-MS/MS traceable to the SRM 2972 and the ID-LC-MS/MS 25(OH)D Reference Method Procedure to see if the re-standardization would be efficient in a population of 3rd trimester pregnant women (PW), hemodialysis (HD) and osteoporosis (OP) patient. MATERIAL AND METHODS 184 serum samples (25(OH)D: 8.4-87 ng/ml) were selected to calibrate the immunoassays (Abbott-Architect, Roche-Elecsys, DiaSorin-Liaison, Siemens-Centaur and IDS-iSYS). Chromsystems MassChrom method was used as the referenced. Serum obtained in 34 PW, 25 HD and 34 OP patients were used as comparatives. RESULTS After adjusting to LC-MS/MS, immunoassays had regression slopes nearly identical to 1.0 with intercepts <0.5 ng/ml. However, in special populations, a systematic bias was still observed, except for iSYS. CONCLUSIONS Re-standardization of 25(OH)D immunoassay will globally improve the differences. However, patients with a different serum matrix will still present significantly different results when they will be run with different methods. For those patients, the LC-MS/MS method seems to be the method of choice, even if some immunoassays are less influenced than others.

Evaluation of the cross-reactivity of 25-hydroxyvitamin D2 on seven commercial immunoassays on native samples

Introduction: In serum, 25-hydroxy-vitamin D (25(OH)D) can be found in two forms, namely 25(OH)D2 and 25(OH)D3. We recently published a mathematical method to estimate the 25(OH)D2 recovery without spiking the samples. Since then, new “total” vitamin D immunoassays have appeared on the market (Roche “Total” vitamin D, Siemens Centaur vitamin D “total”, “total” vitamin D on DiaSorin Liaison XL, Abbott Architect Vitamin D). We aimed to study the 25(OH)D2 recovery of these new immunoassays and re-evaluate the cross-reactivity of previously studied assays (IDS iSYS Vitamin D and DiaSorin RIA).

Analytical and clinical evaluation of the new Fujirebio Lumipulse G non-competitive assay for 25(OH)-vitamin D and three immunoassays for 25(OH)D in healthy subjects, osteoporotic patients, third trimester pregnant women, healthy African subjects, hemodialyzed and intensive care patients

Abstract Background: In this study, we provide a short analytical evaluation of the new Fujirebio Lumipulse®G non-competitive immunoassay for 25(OH)D. Clinical performance was compared with three commercial competitive automated immunoassays against a Vitamin D Standardization Program (VDSP)-traceable liquid chromatography-tandem mass spectrometry (LC-MS/MS) in six different clinically relevant populations. Methods: Lumipulse®G 25(OH)D precision, measurement uncertainty, recovery, limit of quantification were assessed, as well as 25(OH)D2 and C3-epimer recovery. For method comparison, 250 serum samples obtained in healthy Caucasians and Africans, osteoporotic, hemodialyzed and intensive care patients and 3rd trimester pregnant women were analyzed by all methods. Correlation was studied using Passing-Bablok and Bland-Altman analysis. Concordance correlation coefficient (CCC) was calculated to evaluate agreement between immunoassays and the LC-MS/MS. Results: The Lumipulse®G 25(OH)D assay presented interesting analytical features and showed excellent correlation to the LC-MS/MS results (y=1.00×–1.35 ng/mL), as obtained in healthy Caucasian individuals. In the other special populations, Lumipulse®G presented a concordance with LC-MS/MS which was generally higher than competitors, even if all methods significantly under-recovered 25(OH)D in hemodialyzed patients. Intra-assay CV ranged from 12.1% at 9.6 ng/mL to 2.1% at 103.7 ng/mL and inter-assay CV ranged from 16.2 to 3.7% at the same concentrations, respectively. Measurement uncertainty, with a probability of 95%, were respectively 33.1 and 7.6% at these concentrations. LOQ was found to be at 4.6 ng/mL. Mean (95% CI) 25(OH)D2 revovery was 77% (74–81) and no cross-reactivity was observed with C3-epimer. Conclusions: Fujirebio Lumipulse®G 25-OH Vitamin D Total assay is therefore considered suitable for assessment of vitamin D status in clinical routine.

Standardization of DiaSorin and Roche automated third generation PTH assays with an international standard: impact on clinical populations

Abstract Background: Standardization of parathyroid hormone (PTH) assays is a major issue, especially in hemodialyzed (HD) patients. Two automated third generation PTH assays (Roche Elecsys and DiaSorin Liaison) are now available. These assays are specific for the (1-84) PTH and do not cross-react with the (7-84) fragment, contrary to second generation (intact) assays. We aimed to calibrate the two methods against the WHO International PTH Standard (IS) 95/646 to see if the two assays could provide comparable results in a population of healthy subjects, HD patients and patients suffering from primary hyperparathyroidism (PHP). Methods: We selected 79 healthy subjects and two populations of patients presenting PTH disorders: 56 HD and 27 PHP patients. We reconstituted the IS in a pool of human serum containing undetectable levels of 1-84 PTH and prepared 13 serum standards ranging from 0 to 2000 pg/mL. The standards were run on the two instruments to calibrate the assays on the IS. The different populations were run before and after restandardization. Results: As these kits were differently calibrated, the results obtained after restandarization were significantly different. Restandardization process improved concordance between assays and, taking the analytical variability of the two kits into account, the results could be considered to be similar. Conclusions: Restandardization of automated third generation PTH assays with the WHO 1-84 PTH Standard significantly reduces inter-method variability. Reference ranges and raw values are totally transposable from one method to the other in healthy subjects, but also in diseased patients, e.g., with HD or those suffering from PHP.

Analytical and clinical validation of the new Abbot Architect 25(OH)D assay: fit for purpose?

Abstract Background: We provide a clinical and analytical evaluation of the reformulated version of the Abbott Architect 25-hydroxyvitamin D assay. We compared this assay with three commercial automated immunoassays and against a VDSP-traceable liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) in six different populations. We also supplemented 40 healthy volunteers with either 600,000 IU of vitamin D2 or 100,000 of vitamin D3 to evaluate the performance of the immunoassays vs. the LC-MS/MS. Methods: Precision and limit of quantification were assessed, 25(OH)D2 and C3-epimer recovery were calculated. Two hundred and forty samples obtained in healthy Caucasians and Africans, osteoporotic, hemodialyzed and intensive care patients and 3rd trimester pregnant women were analyzed by all methods. Correlation was studied using Passing-Bablok and Bland-Altman analysis. Concordance correlation coefficient (CCC) was calculated to evaluate agreement between immunoassays and LC-MS/MS. We verified if patients were homogeneously classified with the immunoassays when they took vitamin D2 or vitamin D3 after 1, 7 and 28 days. Results: We observed excellent analytical features and showed a very good correlation to the LC-MS/MS results in the overall population. Compared to the other immunoassays, concordance of the new Abbott assay with the LC-MS/MS was at least similar, and often better in diseased populations. Althought the cross-reactivity with 25(OH)D2 was not of 100%, there was no significant difference in the classifications of the patients, either supplemented with D2 or D3 or after 7 or 28 days. Conclusions: This modified version of the Abbott Architect assay is clearly improved compared to the previous one and presents a better agreement with the LC-MS/MS.

Vitamin D status after a high dose of cholecalciferol in healthy and burn subjects

BACKGROUND Burn patients are at risk of vitamin D (VD) deficiency and may benefit from its pleiotropic effects as soon as acute phase. Aim of this observational study was to assess effects of a cholecalciferol (VD3) bolus on VD status in adult burn patients (Group B, GB) after admission, compared to healthy subjects (Group H, GH). METHODS Both groups received an oral dose of 100,000 IU VD3. Blood samples were collected before (D0) and 7 days (D7) after bolus to measure 250H-D, 1,25(OH)2-D, parathyroid hormone (PTH) and fibroblast growth factor 23 (FGF23). Albumin (ALB) and VD binding protein (DBP) were measured and used to calculate free 25OH-D level. Data were expressed as median (min-max) or proportions. RESULTS A total of 49 subjects were included: 29 in GH and 20 in GB. At D0, prevalence of VD deficiency was higher in GB: 25OH-D was 21.5 (10.1-46.3) ng/ml in GH vs 11 (1.8-31.4) ng/ml in GB. DBP and ALB were lower in GB. At D7, DBP was stable in both groups while ALB decreased in GB. 25OH-D increased by 66.6 (13.5-260.3)% in GH. In GB, changes in 25OH-D extended from -36.7% to 333.3% with a median increase of 33.1%. Similar changes were observed in each group for free 25OH-D. High FGF23 levels were observed in GB. CONCLUSIONS This study highlighted the differences in VD status and in response to a high dose VD3 in burn patients when compared to healthy patients. Pitfalls in VD status assessment are numerous during acute burn care: 25OH-D measurement needs cautious interpretation and interest of free 25OH-D is still questionable. They should not prevent burn patients to receive VD supplements during acute care. Higher doses than general recommendations should probably be considered.

Effect of Body Site and Surface on Vitamin D and 25-Hydroxyvitamin D Production after a Single Narrowband UVB Exposure

TO THE EDITOR Skin color, body surface area (BSA) exposed to UVB, ability to tan, latitude, altitude, season, solar zenith angle, clothing, age, moment of the day, and use of sunscreen influence the cutaneous synthesis of vitamin D (Chen et al., 2007; Holick, 2003; Libon et al., 2013; Rizzoli et al., 2013; Touvier et al., 2015). There are still many controversies concerning the impact of the UV-exposed anatomical region and BSA on cutaneous vitamin D production. This study was performed in accordance with the Declaration of Helsinki (World Medical Association, 2013) and approved by the university hospital ethics committee. All study procedures were explained to the volunteers. All participants signed an informed consent form. To determine whether various anatomical regions exhibit different vitamin D production capacities, assessed as production per percentage of BSA, this study initially measured cholecalciferol (vitamin D3, representing cutaneous synthesis) and 25(OH)D3 (the circulating form of vitamin D) levels before and on days 1, 2, and 5 in 72 young healthy volunteers, recruited from among medical students (male 1⁄4 25, female 1⁄4 47, mean age 1⁄4 23.0 2.3 years, age range 1⁄4 19e29 years, body mass index 1⁄4 21.6 1.9 kg/m, body mass index range1⁄4 18e25 kg/m, Fitzpatrick’s phototype III) after a single narrowband (310e315nm) UVB exposure of 0.8 minimal erythematous dose (mean 1⁄4 0.22 0.08 J/cm) to different BSAs. BSAs were determined according to real-life clothing habits: group I (n 1⁄4 15): head, neck, and hands (BSA 1⁄4 9%); group II (n 1⁄4 15): head, neck, arms, and hands (BSA 1⁄4 23%); group III (n 1⁄4 15): head, neck, arms, hands, legs, and feet (BSA 1⁄4 50%); group IV (n 1⁄4 15): the total body (BSA 1⁄4 96%), as well as a noneUVBexposed control group (group 0, n 1⁄4 12). The BSA percentages were calculated according to Wallace’s rule (Hettiaratchy and Papini, 2014). 25(OH)D3 analysis was performed using liquid chromatography/tandem mass spectrometry kits for 25(OH) D3 measurement (MassChrom 25OH-Vitamin D3/D2 [LC-MS/MS], Chromsystems, Gräfelfing, Germany) (Cavalier et al., 2014). Cholecalciferol was determined with an in-house developed liquid chromatography/tandem mass spectrometry method. Statistics were expressed as mean and standard deviation for continuous variables and as frequency tables for categorical variables. On graphs, mean values were plotted with their standard error. The general linear mixed model was used to analyze the evolutions of vitamin D and 25(OH)D3 over time and test for group differences. A linear regression was used to compare the area under the curve of vitamin D and 25(OH)D3 with respect to BSA. Results were considered significant at the 5% critical level (P < 0.05). Calculations were done with SAS, version 9.4 (SAS Institute, Cary, NC). Vitamin D level increased with peak levels on day 1 for groups I and II and on day 2 for groups III and IV, and it decreased subsequently in all groups except in the control group (Figure 1). In contrast, 25(OH)D3 increased steadily at all time points in all groups but not in the control group (Figure 1). The larger the exposed area was (group IV > III > II > I), the higher the increase (Figure 1). Expressed as area under the curve, no difference in 25(OH)D3 level was observed between the groups (P 1⁄4 0.29) at day 0. After UVB irradiation, a steady increase in vitamin D and 25(OH)D3 levels was observed in groups I through IV, with a constant increase according to the body surface exposed (Table 1). This study gave evidence that the vitamin D production capacities of various skin regions are not similar at all. In fact, the relative mean vitamin D production expressed as area under the curve for the entire body (group IV) was 314 nmol/L, corresponding to 96% of BSA. Hence, the production per percentage of BSA was 3.3 nmol/L. Consequently, the production in group I was 25.4 nmol/L; in group II, 10.3 nmol/L; and in group III, 5.36 nmol/L.

Sunscreens block cutaneous vitamin D production with only a minimal effect on circulating 25-hydroxyvitamin D

SummaryA 50+ SPF sunscreen decreased significantly cutaneous vitamin D production following a single narrow-band (nb)UVB exposure, independently from the body surface area exposed. In contrast, the circulating 25(OH)D3 levels were only minimally affected. It is probable that another endogenous source of precursors is selected when skin-originated precursors are lacking.PurposeSunscreen use, highly advocated for preventing cutaneous carcinogenesis, is potentially leading to an aggravation of vitamin D deficiency with its consequences on bone health. The effect of sunscreens on circulating vitamin D levels remains debated. This study investigated the effect of sunscreen on cutaneous vitamin D production and circulating 25(OH)D3 levels, according to different body surface areas (BSA).MethodsVitamin D and 25(OH)D3 levels were measured in four groups exposed to a single nbUVB exposure on 9% (group I: head and hands), 23% (group II: head, hands and arms), 50% (group III: head, hands, arms and legs) and 96% (group IV: total body) of the body surface without and with a 50+ sun protection factor sunscreen.ResultsSunscreen use decreased by 83, 88.3, 75.7 and 92.5% the cutaneous vitamin D production in groups I to IV, respectively, but only by 13.2, 10.5, 7.7 and 10.4% the values of circulating 25(OH)D3, correspondingly.ConclusionsAlthough a 50+ sunscreen decreases significantly cutaneous vitamin D production following a single nbUVB exposure, and independently from the BSA, the circulating 25(OH)D3 levels were only minimally affected. This could be explained by a switch to another endogenous source of precursors. Short-term sunscreen use probably does not affect circulating vitamin D levels and hence does not increase the risk for osteoporosis. The effect of long-term sunscreen use remains however to be determined.

impact of stopping vitamin K antagonist therapy on concentrations of dephospho-uncarboxylated Matrix Gla protein

Several experimental and clinical studies suggest that vitamin K antagonist (VKA) therapy is a risk factor for the development of vascular calcifications and calciphylaxis in hemodialysis (HD) patients [1] . One major pathophysiological mechanism that could explain these observations involves both vitamin K and matrix Gla protein (MGP). Briefly, MGP is an 11 kDa protein secreted by vascular smooth muscle cells, acting as a potent local inhibitor of vascular calcification. In order to be fully active, MGP must be phosphorylated and carboxylated [2] . This carboxylation is highly dependent on availability of vitamin K. HD patients are per se prone to vitamin K deficiency which can be potentiated by VKA therapy. In this situation of vitamin K depletion, the MGP calcification inhibitor activity is decreased leading to vascular calcification [3] . The inactive form of MGP, namely the dephospho-uncarboxylated MGP (dp-ucMGP), is presented as a good biomarker of vitamin K status and vascular calcification. Indeed, higher dp-ucMGP levels have been associated with higher level of vascular calcifications in CKD and HD patients [4] . To sustain the hypothesis of dp-ucMGP as a marker of vitamin K status, several authors have shown in the general population, in CKD and in HD patients that VKA therapy was associated with higher dp-ucMGP levels [4 – 6] . Conversely, recent data in HD patients have shown that dp-ucMGP concentrations were decreasing after vitamin K supplementation [7 – 9] . Whether vitamin K therapy could be of interest in the prevention of vascular calcification in HD patients is currently under investigation [10] . Until recently, few alternative strategies were available to VKA in HD with atrial fibrillation or valve replacement. Recent data in HD suggested that fondaparinux, an indirect factor Xa inhibitor, could be safely used in these patients [11] . In the present study, we have measured dp-ucMGP in patients directly after switching from VKA to fondaparinux. Our goal was to confirm the influence of VKA on dp-ucMGP levels and study the kinetic of these potential changes in MGP concentrations. We studied HD patients treated by VKA in our university center. These patients were all treated by acenocoumarol. Switching from VKA to fondaparinux was considered only in patients anticoagulated for atrial fibrillation. Seven patients, dialyzed three times a week were considered. Two measurements (T1 and T2) were obtained at the beginning of the two dialysis sessions before stopping VKA. The patients stopped VKA therapy the day before the first dialysis session of the next week (on Sunday or Monday). Five measurements were then obtained at the beginning of each of the next five dialysis sessions (T3 – T7). dp-ucMGP was quantified with an automated assay (Ina K tif MGP iSYS kit, IDS, Boldon, UK). dp-ucMGP concentrations were compared using one-way repeated measures analysis of variance by ranks (Friedman test), followed by Dunn ’ s test correcting the α -level for pairwise comparison between time-points. MannWhitney test was used for comparison between cohorts. Main clinical and biological characteristics of the seven patients are described in Table 1 . Before switching from VKA to fondaparinux, median concentrations of dp-ucMGP *Corresponding author: Pierre Delanaye, Service de Dialyse, CHU Sart Tilman, 4000 Li è ge, Belgium, Phone: + 32 43667111, Fax: + 32 43667205, E-mail: pierre_delanaye@yahoo.fr ; and Nephrology-Dialysis-Transplantation, University of Li è ge, CHU Sart Tilman, Li è ge, Belgium Bernard E. Dubois and Jean-Marie Krzesinski: Nephrology-DialysisTransplantation, University of Li è ge, CHU Sart Tilman, Li è ge, Belgium Pierre Lukas, Pierre Peters and Etienne Cavalier: Clinical Chemistry, University of Li è ge, CHU Sart Tilman, Li è ge, Belgium Hans Pottel: Department of Primary Care and Public Health @ Kulak, KU Leuven Kulak, Kortrijk, Belgium