Vitamin D and metabolites measurement by tandem ... - Springer Link

8 downloads 0 Views 485KB Size Report
Mar 29, 2013 - 2001;15(2):133–. 40. doi:10.1002/bmc.43. 89. Bouillon R, Okamura WH, Norman AW. .... Sharpless KE, et al. Development and certification of a ...
Rev Endocr Metab Disord (2013) 14:159–184 DOI 10.1007/s11154-013-9241-0

Vitamin D and metabolites measurement by tandem mass spectrometry Johannes M. W. van den Ouweland & Michael Vogeser & Silvia Bächer

Published online: 29 March 2013 # Springer Science+Business Media New York 2013

Abstract The prevalence of vitamin D deficiency in the general population has become a major public health problem. Vitamin D deficiency might have significant consequences not only to bone health but possibly to autoimmune-, infectious and cardiovascular disease. This has resulted in increased clinical testing for 25-hydroxyvitamin D (25(OH)D) in serum, as circulating 25(OH)D is regarded as the best indicator of adequate exposure to sunlight and dietary intake of vitamin D. There are reportedly over 50 vitamin D metabolites of which 25(OH)D and 1,25(OH)2D are well known to provide clinical information. More recently, there is increasing interest in measuring the C3-epimer of 25(OH)D, which has shown to contribute significantly to the 25(OH)D concentration, particularly in infant populations, and in 24,25(OH)2D, a major catabolite of 25(OH)D metabolism. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is an analytical tool that allows the specific determination of all relevant vitamin D metabolites, with the potential of performing multiple analyte analysis in a single experimental setting, creating a vitamin D profile. This article reviews recent advances in the quantification of vitamin D metabolites using LC-MS/MS. Keywords Vitamin D and metabolites . Liquid chromatography-tandem mass spectrometry (LC-MS/MS) Abbreviations 1,25(OH)2D 24R,25(OH)2D3 25(OH)D

1,25-dihydroxyvitamin D 24R,25-dihydroxyvitamin D3 25-hydroxyvitamin D

J. M. W. van den Ouweland Department of Clinical Chemistry, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands M. Vogeser (*) : S. Bächer Institute of Laboratory Medicine, Hospital of the University of Munich, Marchioninistr. 15, 81377, Munich, Germany e-mail: [email protected]

3-epi-25(OH)D APCI APPI CV DBS DEQAS ESI GC-MS ID LC-MS/MS LLE LLOQ MRM NIST PFP PTH PTAD RIA RRA SPE SRM VDBP

3-epi-25-hydroxyvitamin D Atmospheric pressure chemical ionisation Atmospheric pressure photo ionisation Coefficient of variation Dried blood spots Vitamin D external quality assessment scheme Electrospray ionisation Gas chromatography mass spectrometry Isotope dilution Liquid chromatography tandem mass spectrometry Liquid liquid extraction Lower limit of quantification Multi reaction monitoring National Institute of Standards and Technology Pentafluorophenyl Parathyroid hormone 4-phenly-1,2,4-triazoline-3,5-dione Radio immunoassay Radio receptor assay Solid phase extraction Standard reference material Vitamin D binding protein

1 General introduction Worldwide there is a continuously growing interest in the characterisation of the vitamin D status by laboratory tests, both on an individual and epidemiological level. This is in particular based on accumulating evidence that vitamin D deficiency—often defined as a decrease of relatively high plasma parathyroid hormone (PTH) levels in response to vitamin D administration—is highly prevalent in particular

160

in winter months[1]. This is not at all restricted to northern regions [2]. Furthermore, a plethora of data has become available suggesting non-calcemic, extra-skeletal or even pleiotropic effects of vitamin D deficiency. However, the extremely large body of epidemiological data and results of studies on vitamin D supplementation is very difficult to interpret and conclusions about causality extrapolated from observational data seem premature [3]. Nevertheless, there are recommendations communicated by important scientific bodies, like the Institute of Medicine (IOM) and Endocrine Society, which include—besides recommended daily intakes—also target concentrations of 25-hydroxyvitamin D (25(OH)D). This intermediate metabolite of vitamin D [4] is the most widely assessed marker of the vitamin D status, however, it must be recognized as one of several potential surrogate markers of the vitamin D status. Vitamin D, the parent compound, has two forms, vitamin D3 (cholecalciferol) and vitamin D2 (ergocalciferol). Vitamin D3 can be synthesized from 7-dehydrocholesterol when skin is exposed to UV irradiation from sunlight (with thermal isomerization), while vitamin D2 can be produced by plants and fungi by the

Fig. 1 Overview of the vitamin D metabolism

Rev Endocr Metab Disord (2013) 14:159–184

solar irradiation of ergosterol. Both forms of vitamin D are transported bound to vitamin D binding protein (VDBP), are hydroxylated twice, with the initial step of 25-hydroxylation in the liver to produce 25(OH)D, while the second step occurs both in the kidney and extra-renal sites (e.g. skin, prostate, intestine, and breast tissue) to form biologically active 1,25dihydroxyvitamin D (1,25(OH)2D)) (Fig. 1). 1,25(OH)2D becomes further oxidized into 1,25(OH)2D3-26,23-lactone and calcitroic acid [5]. There are reportedly over 50 vitamin D metabolites [6] of which only five (vitamin D, 25(OH)D, 1,25(OH)2D, 24R,25-dihydroxyvitamin D3 (24R,25(OH)2D) and 3-epi-25-hydroxyvitamin D (3-epi-25(OH)D)) have been more or less extensively quantitated (Table 1). To date, only two of these, namely, 25(OH)D and 1,25(OH)2D, provide clinically relevant information. However, the quantitation of 24R,25(OH)2D, vitamin D and 3-epi-25(OH)D can provide important information in a research environment. Potential limitations in the diagnostic reliability of 25(OH)D include inter-individual and genetically determined variation in the protein binding and thus the bioavailability of the compound

Rev Endocr Metab Disord (2013) 14:159–184

[7]; potential genetic polymorphism in the affinity of vitamin D receptors; the probable inability of single 25-hydroxyvitamin D determination to characterize the long-term vitamin D status throughout the year with substantial variation of UV intensity of sun light. As a further and substantial potential confounder of individual and epidemiological data, limited accuracy and reproducibility of routine 25(OH)D methods have been recognized [8]. These tests were initially based on competitive ligand binding methods involving both, antibodies or vitamin D-binding protein with radioactive labelled 25(OH)D as the tracer and manual handling. Consequent to the increasing number of test ordering automated ligand binding tests were introduced from 2000 on. While in manual test protocols release of the target analyte from its highly avid bond to vitamin D-binding protein can be achieved reliably by e.g. acetonitrile, this release is difficult to achieve when using closed automated ligand binding tests which do not allow an evaporation step. In regions with prevalent use of ergocalciferol (vitamin D2) inconsistent cross-reactivity with 25-hydroxyvitamin D2 further contributes to between-method bias in individual samples. Furthermore, the susceptibility of competitive ligand binding methods to poorly definable matrix effects with a high degree of between-individual variation is well recognized. Beyond these analytical issues inherent to ligand binding test the reliability of 25(OH)D measurement was critically limited by the fact that neither a reference method nor—as a consequence—reference material preparations were available until recently. This was in contrast to most of the steroid hormones for which gas chromatography mass spectrometry (GC-MS)based reference methods were already available for many years. In contrast the development of robust and convenient GC-MS methods for 25(OH)D methods proved extremely difficult. Only the technology liquid chromatography tandem mass spectrometry (LC-MS/MS) enabled the development of mass

161

spectrometric methods for this analyte, enabling to involve the principle of stable istotope dilution internal standardisation which can help to realize highest analytical reliability. A first LC-MS/MS method for serum 25(OH)D measurement involving derivatisation was described in 2001 [9], while a first derivatisation-free method was reported in 2004 [10]. Meanwhile two LC-MS/MS procedures have obtained the status of reference methods [11, 12]. It is recognized as stateof-the-art that routine ligand binding tests for 25(OH)D are standardized to and validated according to LC-MS/MS methods establishing traceability of measurements with routine laboratory methods to the SI-unit [13]. Thanks to the rather good practicability of LC-MS/MS this technology is also increasingly used for routine diagnostic purposes. Thus LC-MS/MS has gained a crucial role in the characterisation of the vitamin D status during recent years. Several recent reviews have addressed a wide range of issues related to the quantification of serum or plasma 25(OH)D. However, meanwhile several novel analytical issues have been recognized in the context of this analytical field with increased interest in measuring 24,25(OH)2D and the C3-epimer of 25(OH)D. Furthermore, LC-MS/MS has the potential of performing multiple analyte analysis in a single experimental setting, creating a vitamin D profile (e.g. measuring CYP24A1 activity by simultaneous quantitation of 25(OH)D, 1,25(OH)2D and 24,25(OH)2D), an achievement not feasible by ligand binding assays (Fig. 2). This present review is focussed on these recent topics in the measurement of vitamin D metabolites.

2 Vitamin D Vitamin D, the parent compound, is a poor indicator of nutritional status because of its short circulating half-life. The

Table 1 Concentrations of vitamin D metabolites in human serum and their percentage immunoassay cross-reactivity Analytea

Vitamin D3 Vitamin D2 25(OH)D3 25(OH)D2 1,25(OH)2D3 1,25(OH)2D2 24R,25(OH)2D3 3-epi-25(OH)D3

Circulation levels in human serum

18–29 nmol/L [115] 8–165 nmol/L [75] < 7 nmol/L [75] 48–168 pmol/L (0.1 % of 25(OH)D) [55] 355, 3) 417 > 363, IS 407 > 263

a) Waters Quattro LC b) ESI+ c) 1) 401 > 159, IS 405 > 159

a) ThermoQuest LCQ b) APCIc) 620

a) Finnigan MAT LCQ b) APCI+ c) 739, IS: 745

a) ThermoQuest LCQ b) APCI+ c) 1) 746 > 746, 2) 758 > 758, IS: 760 > 760

MS: a) Instrument b) Ionization c) Mass transition

Table 2 Overview of LC-MS/MS methods for quantification of different vitamin D metabolites in human biological fluids (up to April 2012) Chromatographic run time

9 min

Linear range: 8 min 1) 4–256 nmol/L 2) 5–316 nmol/L LLOQ: 1) 4 nmol/L 2) 5 nmol/L

1) 62 nmol/L 2) 61 nmol/L 3) 60 nmol/L

Linear range: 11 min 1) 62–250 nmol/L 2) 61–242 nmol/L 3) 60–240 nmol/L LLOQ:

Linear range: 25–363 nmol/L LLOQ: not disclosed

Linear range: 7 min 1 + 2) 0.05–1 ng/tube LLOQ: 1) 7.5 nmol/L 2) 7.3 nmol/L

LLOQ: 1) 0.55 nmol/L

Linear range: 10 min 1) 0.05 ng–1.2 ng/tube

LLOQ: 1) 7.5 nmol/L 2) 7.3 nmol/L

Linear range: 7 min 1 + 2) 0.05–1 ng/tube

Linear range, LLOQ

164 Rev Endocr Metab Disord (2013) 14:159–184

Analyte, Internal standard (IS)

1) 25OHD3 2) 25OHD2 IS: 2H3-Δ9-THC

1) 25OHD3 2) 25OHD2 3) 3-Epi-25OHD3 4) 3-Epi-25OHD2 IS: 2H6-25OHD3

1) 25OHD3 2) 25OHD2 IS: 1-alpha-OHD3

1) D3 2) D2 3) 25OHD3 4) 25OHD2 1,25(OH)2D3

1) 25OHD3 2) 25OHD2 IS: 2H6-25OHD3

Reference

Saenger et al., 2006 [43]

Singh et al., 2006 [67]

Gören et al., 2007 [117]

Priego Capote et al., 2007 [118]

Chen et al., 2008 [48]

Table 2 (continued) LC: a) Column b) Mobile phase c) Flow rate a) XTerra (3.5 μm, 50 × 2.1 mm, Waters) b) 2 mM ammonium acetate in MeOH + 0.1 % formic acid c) 0.1 mL/min

a) Chirex-PGLY and DNB (250 × 4.6 mm, Phenomenex) b) A)MeOH, B) 0.005% formic acid (gradient elution) c) 0.9 mL/min a) ZivakDs (3 μm, 100 × 4.6 mm, Zivak) b) MeOH:H2O + 5 % formic acid (95:5 v/v) c) 0.5 mL/min

a) Zorbax Eclipse XDB-C18 (5 μm, 150 × 4.6 mm, Agilent) b) A) 5 mM ammonium formate in MeCN:H2O (90:10 v/v), B) 5 mM ammonim formate in MeOH (gradient elution) c) 1 mL/min

a) Supelcosil LC-18 (3 μm, 30 × 3 mm, Supelco) b) EtOH:H2O (83:17 v/v) c) 0.5 mL/min

Sample type (sample volume), Sample preparation

Serum (200 μL), LLE (heptane)

Serum (200 μL), PP (MeCN) → on-line turboflow extraction (Cyclone-P 50 × 1 mm, Cohesive Technologies)

Serum/Plasma (500 μL), PP (MeOH:MeCN) → LLE (heptane)

Serum (1,000 μL), PP (MeOH:) → LLE (hexane)

Serum (200 μL), PP (MeCN) → on-line SPE (OASIS HLB, 25 μm, 20 × 2.1 mm, Waters)

a) Waters Micromass Quattro micro b) APCI+ c) 1) 401 > 383, 2) 413 > 395, IS: 407 > 389

a) Agilent 6410 b) ESI+ c) 1) 385 > 159, 2) 397 > 159, 3) 383 > 159, 4) 413 > 395, 5) 399 > 226

a) Thermo Finnigan LCQ Deca b) APCI+ c) 1) 383 > 365, 2) 395 > 377, IS 401

a) API 4000, Applied Biosystems b) APCI+ c) 1 + 3) 401 > 383, 2 + 4) 413 > 395, IS: 407 > 389

a) Waters Micromass Quattro b) ESI+ c) 1) 401 > 365, 2) 413 > 355, IS 318 > 196

MS: a) Instrument b) Ionization c) Mass transition

Chromatographic run time

7 min

Linear range: 7 min 1) 12.5–250 nmol/L 2) 12.1–243 nmol/L LLOQ: not disclosed

Linear range: 27 min 1) 2.6–1,302 nmol/L 2) 2.5–1,260 nmol/L 3) 2.5–1,250 nmol/L 4) 2.4–1,213 nmol/L 5) 1.2–60.1 nmol/L LLOQ: 1) 0.8 nmol/L 2) 0.5 nmol/L 3) 1.5 nmol/L 4) 1.0 nmol/L 5) 1.2 nmol/L

LLOQ: 1) not dislosed 2) 24 nmol/L

2) 24–1,211 nmol/L

Linear range: 6 min 1) 25–1,250 nmol/L

Linear range: 1–4) not disclosed LLOQ: 1–4) not disclosed

Linear range: 6 min 1) 2.5–250 nmol/L 2) 2.4–242 nmol/L

Linear range, LLOQ

Rev Endocr Metab Disord (2013) 14:159–184 165

1) 25OHD3 2) 25OHD2 IS: 2H6-25OHD3 3) 1,25(OH)2D3 4) 1.25(OH)2D2 5) 24,25(OH)2D3 IS 2H61.25(OH)2D3 25OHD3 IS: 25OHD4

Aronov et al., 2008 [60]

1) 25OHD3 2) 25OHD2 IS: 2H6-25OHD3

1) 25OHD3 2) 25OHD2 IS: 2H3-25OHD3

1) 25OHD3 2) 25OHD2 IS: 2H6-25OHD3

Knox et al., 2009 [51]

Eyles et al., 2009 [96]

Newman et al., 2009 [97]

Higashi et al., 2008 [40]

Analyte, Internal standard (IS)

Reference

Table 2 (continued)

a) ACQUITY TQD, Waters b) ESI+ c) 1) 401 > 159, 2) 413 > 83, IS 407 > 159

a) API 4000 QTRAP, Applied Biosystems b) ESI+ c) 1) 558 > 298, 2) 570 > 298, IS 561 > 301

a) ACQUITY UPLC BEH C8 (1.7 μm, 50 × 2.1 mm, Waters) b) A) 2 mM ammonium acetate + 0.1 % formic acid, B) MeOH (gradient elution) c) 0.35 mL/min a) Zorbax SB-C18 (5 μm, 50 × 2.1 mm, Agilent) b) MeCN:0.1 % formic acid (63:37 v/v) c) 0.35 mL/min

Plasma/Serum (200 μL), PP (MeOH) → automated SPE (10 mg Orochem C8)

Dried blood spot (6 mm disk), a) Varian Pursuit 3u PFP Sonication → PP (MeOH) → LLE (hexane) (3 μm, 50 × 2.0 mm, Varian) Serum (10 μL), PP (MeOH) → LLE (hexane) b) A) 0.1 % formic acid + 2 mM ammonium acetate, B) MeOH (gradient elution) c) 0.2 mL/min

Dried blood spot (3.2 mm disk), PP (MeCN) → Derivatization (PTAD)

a) API 2000 b) ESI+

a) YMC-Pack Pro C18 RS (5 μm, 150 × 2.0 mm, YMC) b) MeOH: 10 mM ammonium formate (5:1 v/v) + 5 mM methylamine c) 0.2 mL/min

Saliva (1 mL), PP (MeCN) → off-line SPE → Derivatization (PTAD)

a) Varian 320 MS TQ b) ESI+ c) 1) 401 > (365 + 383), 2) 413 > (377 + 395), 3) 407 > (371 + 389)

c) 607 > 298, IS 621 > 298

a) Quattro Premier, Waters b) ESI+ c) 1) 558 > 298, 2) 570 > 298, IS: 564 > 298, 3) 574 > 314, 4) 586 > 314, 5) 574 > 298, IS 580 > 314

a) UPLC BEH C18 (1.7 μm, 100 × 2.1 mm, Waters) b) A) MeCN:H2O (10:90 v/v) + 0.1 % formic acid, B) MeOH (gradient elution) c) 0.4 mL/min

Serum (500 μL), PP (MeCN) → LLE (MTBE) or off-line SPE (OASIS HLB, Waters) → Derivatization (PTAD)

MS: a) Instrument b) Ionization c) Mass transition

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

Chromatographic run time

~8 min

1 + 2) not disclosed

LLOQ:

Linear range: 10 min 1) 8.7–187 nmol/L 2) n.d.

LLOQ: 1) 7.7 nmol/L 2) 10.7 nmol/L

Linear range: ~3 min 1) 7.7–125 nmol/L 2) 10.7–121.3 nmol/L

Linear range: 5 min 1) 4–2,500 nmol/L 2) 7.5–2,500 nmol/L LLOQ: 1) 4 nmol/L 2) 7.5 nmol/L

Linear range: 5–125 pmol LLLOQ: 5 pmol/L

Linear range: 12 min not disclosed LLOQ: 1) 62 pmol/L 2) 61 pmol/L 3 + 5) 60 pmol/L 4) 58 pmol/L

Linear range, LLOQ

166 Rev Endocr Metab Disord (2013) 14:159–184

Analyte, Internal standard (IS)

1) 25OHD3 IS: 2H6-25OHD3 2) 25OHD2

1) 25OHD3 2) 25OHD2 IS: 2H6-25OHD3

1) 25OHD3 IS: 2H6-25OHD3 2) 25OHD2 IS: 2H6-25OHD2

1) 25OHD3 IS: 2H625OHD3 IS: 2H625OHD2 3) 24,25(OH)2D3 4) 1,25(OH)2D3 IS: 2H61.25(OH)2D3 5) 1.25(OH)2D2 IS 2H61.25(OH)2D2

Reference

Bunch et al., 2009 [44]

Hojskov et al., 2010 [53]

Herrmann et al., 2010 [36]

Ding et al., 2010 [62]

Table 2 (continued)

a) Synergi MAX-RP (4 μm, 50 × 2 mm, Phenomenex) b) MeOH:2 mM ammonium acetate (85:15 v/v) c) 0.35 mL/min

a) Supelcosil LC-8 (3 μm, 33 × 3.0 mm, Supelco) b) A) H2O, B) MeOH, C) 2 % (v/v) MeOH in H2O, D) toluene (gradient elution) c) 0.6 mL/min

a) ACQUITY BEH C18 (1.7 μm, a) QTRAP 4000, Applied 50 × 2.1 mm, Waters) Biosystems b) ESI+ b) A) 0.1 % formic acid in H2O + 5 mM methylamine, B) 0.1 % formic c) 1) 607 > 298, IS: 613 > 298, 2) 619 acid in MeOH (gradient elution) c) 0.3 mL/min > 298, IS 625 > 298, 3) 623 > 298, 4) 623 > 314, IS: 629 > 314. 5) 635 > 314, IS: 641 > 314

Serum (100 μL), PP (MeCN) → automated LLE (heptane)

Serum (100 μL), PP (MeCN)

Plasma (200 μL, 50 μL), PP (MeCN) → SPE (OASIS HLB, Waters) or LLE (ethyl acetate) or μelution SPE (OASIS HLB μelution plates, Waters) → Derivatization (PTAD)

a) API 5000 b) APPI+ c) 1) 383 > 365, IS: 389 > 371, 2) 395 > 377, IS 401 > 383

a) API 3000, Applied Biosystems b) not disclosed c) 1) 401 > 365, 2) 413 > 395, IS 407 > 371

a) Thermo TSQ Quantum Access b) APCI+ c) 1) 401 > 383, 2) 413 > 395, IS 407 > 389

a) Hypersil GOLD aQ (5 μm, 50 × 2.1 mm, Thermo Fisher) b) H2O:MeOH (5:95 v/v) c) 0.85 mL/min

Serum (100 μL), PP (MeCN) → online turboflow extraction (Cyclone-P 50 × 1 mm, Thermo Fisher)

MS: a) Instrument b) Ionization c) Mass transition

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

4 min

3 min (using two LC channels) 5.5 min (using one LC channel)

Chromatographic run time

5) 23 pmol/L

3) 0.2–720 pmol/L 4) 0.5–120 pmol/L 5) 0.2–117 pmol/L LLOQ: 1) 25 pmol/L 2) 24 pmol/L 3) not disclosed 4) 48 pmol/L

Linear range: 365, IS: 407 > 371, 2) 413 > 337, IS: 419 > 337

c) 1) 401 > 159, 2) 413 > 83, IS: 407 > 159

1) 3.5 nmol/L, 2) 2.0 nmol/L

LLOQ:

2) 2.0–550 nmol/L

a) ACQUITY TQ, Waters Linear range: 5 min b) ESI+ 1) 3.5–550 nmol/L

Linear range: not disclosed 1) 2.5–2,496 nmol/L

Linear range: >15 min 29–4,800 pmol/L LLOQ: 36 pmol/L

1) 2) 3) 4)

Linear range: 27 min 1) 0.25–125 nmol/L 2) 0.24–121 nmol/L 3) 0.12–60 nmol/L 4) 0.012–6 nmol/L LLOQ:

Linear range, LLOQ

a) API 4000 b) APCI+

a) API 5500 Qtrap, Applied Biosystems b) ESI+ c) 423 >369, IS: 429 > 393

a) Quantum Ultra EMR, Thermo Scientific b) ESI+ c) 1) 558 > 298, 2) 570 > 298, 3) 574 > 298, 4) 574 > 314

MS: a) Instrument b) Ionization c) Mass transition

168 Rev Endocr Metab Disord (2013) 14:159–184

1) 25OHD3 IS: 2H6-25OHD3 2) 25OHD2

Stepman et al., 2011 [11]

IS: stanozol-D3

4) 1-alpha-OHD3 5) 7-alphahydroxy-4cholesten-3-one

3) 25OHD2

1) 25OHD3 2) 3-Epi-25OHD3

v.d. Ouweland et 1) 25OHD3 al., 2011 2) 3-Epi-25OHD3 [69] 3) 25OHD2 IS: 2H6-25OHD3

Shah et al., 2011 [42]

1) 25OHD3 2) 3-Epi-25OHD3 IS: 2H3-25OHD3 3) 25OHD2 4) 3-Epi-25OHD2 IS: 2H3-25OHD2

Tai et al., 2010 [12]

IS: 2H6-25OHD2

Analyte, Internal standard (IS)

Reference

Table 2 (continued)

Serum (250 μL), Additon of NaOH → PP (MeCN:MeOH 9:1 v/v) → off-line SPE

a) ACQUTY CSH fluorophenyl (1.7 μm, 100 × 2.1 mm, Waters) b) A) 2 mM ammonium acetate + 0.1 % formic acid in H2O, B) 0.3 % formic acid in MeOH (gradient elution) c) 0.35 mL/min

LLOQ: 1–5) not disclosed

2 + 4 + 5) not disclosed

Linear range: 1) 1.3–210.6 nmol/L 3) 1.2–204.5 nmol/L

Linear range: 1 + 2) not disclosed LLOQ: 1) 1.2 nmol/L 2) 1.5 nmol/L

Linear range: 1 + 2) not disclosed LLOQ: 1 + 2) not disclosed

Linear range, LLOQ

17 min

1) 22 min 2) 15 min

> 40 min

Chromatographic run time

c) 1 + 2) 401 > 159, 2) 413 > 159, IS 407 > 159

LLOQ: 1–3) not disclosed

a) ACQUITY TQ, Waters Linear range: 6.5 min b) ESI+ 1–3) not disclosed

a) API 3000, Applied Biosystems b) ESI+ c) 1 + 2 + 4 + 5) 401 > (383 + 365 + 159), 3) 413 > (395 + 377 + 355), IS: 322 > 81

a) 1) Zorbax SB-CN (5 μm, 2.1 × a) ACQUITY TQ, Waters 250 mm, Agilent), 2) ACQUITY BEH b) ESI+ C18 (1.7 μm, 50 × 2.1 mm, Waters) c) 1) 401 > 159, IS 407 b) A) H2O : MeOH : formic acid > 159, 2) 413 > 159, (50:50:0.025 v/v/v), B) H2O : MeOH : IS 419 > 159 formic acid (5:95:0.025 v/v/v) (gradient elution) c) 1) 0.4 mL/min, 2) 0.2 mL/min

Serum (250 μL), Alkalinization (NaOH) → LLE (n-hexane) → fractionation (Sephadex LC-20 chromatography) → on-line SPE (ACQUITY BEH 300 C4, 1.7 μm, 50 × 2.1 mm)

a) 1) Zorbax SB C18 RRHD (1.8 μm, 100 × 2.1 mm, Agilent), 2) ULTRON ES-OVM Chiral column (5 μm, 150 × 2 mm) b) A) 0.1 % formic acid in MeCN, B) 0.1 % formic acid in H2O (gradient elution) c) 0.2 mL/min

a) API 4000, Applied Biosystems b) APCI+ c) 1 + 2) 401 > 383, IS: 404 > 386, 3 + 4) 413 > 395, IS: 416 > 398

a) Zorbax SB CN (5 μm, 250 × 4.6 mm, Agilent) b) H2O:MeOH (34:66 v/v) c) 1.0 mL/min

Serum (2 g), LLE (hexane:ethyl acetate 50:50 v/v)

Serum (not disclosed), PP (MeOH:isopropanol 1:1 v/v) → LLE (hexane:dichloromethane)

MS: a) Instrument b) Ionization c) Mass transition

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

Rev Endocr Metab Disord (2013) 14:159–184 169

a) Onyx Monolithic C18 (100 × 3.0 mm, Phenomenex) b) A) MeOH:H2O:50 mM lithium acetate (70:29:1 v/v/v), B) MeOH:50 mM lithium acetate (99:1 v/v) (gradient elution) c) 0.5 mL/min

a) ACE3C8 (3 μm, 75 × 4.2 mm, VWR) b) A) H2O, B) 1 % toluene in MeOH (gradient elution) c) 0.3 mL/min

a) Hypersil Gold (1.9 μm, 100 × 2.1 mm, Thermo Scientific) b) A) 0.1 % formic acid in H2O, B) MeCN (gradient elution) c) 0.2 mL/min

Serum (500 μL), Method 1: Immunoaffinity extraction with ImmunoTubes Method 2: PP (MeCN) → off-line SPE (OASIS HLB, Waters) → immunoaffinity extraction with IDS antibody

Serum (100 μL), LLE (acetone)

Plasma (1,000 μL), PP (MeCN) → LLE (ethyl acetate) → Derivatization (PTAD)

1) 1,25(OH)2D3 IS: 2H61.25(OH)2D3

Yuan et al., 2011 [47]

1) D3 IS: 2H6-D3 2) D2 IS: 2H6-D2 3) 25OHD3 IS: 2H6-25OHD3 4) 25OHD2 IS: 2H6-25OHD2

1) 25OHD3 IS: 2H6-25OHD3 2) 24,25(OH)2D3 3) 1,25(OH)2D3 4) 4β,25(OH)2D3 IS: 2H61.25(OH)2D3

Adamec et al., 2011 [14]

Wang et al., 2011 [61]

IS: 2H61.25(OH)2D2 3) 24R,25(OH)2D3 4) 24S,25(OH)2D3 5) 23R,25(OH)2D3

2) 1.25(OH)2D2

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

Analyte, Internal standard (IS)

Reference

Table 2 (continued) Chromatographic run time

Linear range: 25 min 1) 6.2–200 nmol/L

2) 10.1–756 nmol/L 3) 2.5–749 nmol/L 4) 2.4–727 nmol/L LLOQ: 1) 5.2 nmol/L 2) 5.0 nmol/L 3) 2.5 nmol/L 4) 2.4 nmol/L

Linear range: 25 min 1) 5.2–780 nmol/L

LLOQ: 1) 8 pmol/L, 2) 9 pmol/L

3–5) not disclosed

2) 9–496 pmol/L

Linear range: 10 min 1) 8–495 pmol/L

Linear range, LLOQ

c) 1) 558 > 298, IS: 564 2) 1.2–38 nmol/L > 298, 2) 574 > 298, 3) 0.06–1.9 nmol/L 3 + 4) 574 > 314, IS: 4) not disclosed 580 > 314 LLOQ: 1) 0.13 nmol/L 2) 0.12 nmol/L 3) 0.06 nmol/L 4) not disclosed

a) Agilent 6410 b) ESI+

a) Agilent 6460 b) APPI+ c) 1) 385 > (367 + 259 + 247), IS: 391 > (373 + 265 + 253), 2) 397 > (379 + 271 + 255), IS: 403 > (385 + 277 + 255), 3) 401 > (383 + 365 + 257), IS: 407 > (389 + 371 + 263), 4) 413 > (395 + 355 + 337), IS: 419 > (401 + 355 + 337)

a) TSQ Quantum Vantage, ThermoFisher b) ESI+ c) 1) 423 > 369, IS: 429 > 374, 2) 435 > 381, IS: 441 > 386, 3 + 4) 423 > 367, 5) 423 > (369 + 349)

MS: a) Instrument b) Ionization c) Mass transition

170 Rev Endocr Metab Disord (2013) 14:159–184

a) API 3000, Applied Biosystems b) heated nebulizer+ c) 560 > 298, IS: 566 > 298

a) Synergi Polar-RP (4 μm, 50 × 2.0 mm, Phenomenex) b) MeOH : 10 mM ammonium acetate (88:12 v/v) c) 0.3 mL/min

Method 1: Serum (1,000 μL), In-tube LLE (MTBE) → Derivatization (PTAD) Method 2: Serum (400 μL), 96-well plate LLE (MTBE) → Derivatization (PTAD) Method 3: Serum (100 μL), In-tip SPME → Derivatization (PTAD)

D3IS: 2H6-D3 IS: 2H6-D3

Xie et al., 2011 [119]

c) 0.5 mL/min

System 1: a) Quattro Premier XE, Waters b) ESI+ c) 1) 401 > (383 + 365), IS: 407 > (389 + 105), 2) 413 > (395 + 107), IS: 419 > (401 + 337) System 2: a) Quantum TSQ Ultra AM, Thermo Scientific b) APCI+ c) 1) 383 > (365 + 211), IS: 389 > (371 + 211), 2) 395 > (209 + 269), IS: 401 > (383 + 209)

System 1: a) Hypersil Gold C18 (1.9 μm, 50 × 2.1 mm, Thermo Scientific) b) A) ammonium formate pH 3, B) MeOH (gradient elution) c) 0.5 mL/min System 2: a) Kinetex C18 (2.6 μm, 50 × 2.1 mm, Phenomenex) b) A) ammonium formate pH 3. B) MeOH (gradient elution)

Plasma (100 μL), PP (MeOH:MeCN: MeOH:0.05 M ZnSO4 1.5:1:5:2 v/v/v/v)

1) 25OHD3 IS: 2H6-25OHD3 2) 25OHD2 IS: 2H6-25OHD2

Bogusz et al., 2011 [35]

IS 2H61.25(OH)2D2

2) 1.25(OH)2D2

a) Xevo, Waters b) ESI+ c) 1) (574 + 592) > 314, IS: 580 > 314, 2) 586 > 314, IS: 592 > 314

a) ACQUITY BEH C18 (1.7 μm, 50 × 2.1 mm, Waters) b) A) 0.1 % formic acid in H2O, B) 0.1 % formic acid in MeCN (gradient elution) c) flow gradient

Serum/Plasma (400 μL), PP (MeOH:MeCN 80:20 v/v) → immunoextraction with IDS antibody → Derivatization (PTAD)

1) 1,25(OH)2D3 IS: 2H61.25(OH)2D3

Strathmann et al., 2011 [74]

MS: a) Instrument b) Ionization c) Mass transition

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

Analyte, Internal standard (IS)

Reference

Table 2 (continued)

4.6 min

Chromatographic run time

Linear range: Method 1 + 2: 1.3–65 nmol/L Method 3: 13– 650 nmol/L LLOQ: Methods 1 + 2: 1.3 nmol/L Method 3: 13 nmol/L

2) 1.5 nmol/L

5 min

Linear range: 6 min System 1 + 2: 1) 3–157 nmol/L, 2) 1–122 nmol/L LLOQ: System 1: 1) 3.6 nmol/L 2) 1.5 nmol/L System 2: 1) 3.0 nmol/L

1) 3.0 pmol/L 2) 1.5 pmol/L

LLOQ:

Linear range: not disclosed

Linear range, LLOQ

Rev Endocr Metab Disord (2013) 14:159–184 171

1) 25OHD3 2) 3-Epi-25OHD3 IS: 25OHD4

1) 25OHD3 2) 3-Epi-25OHD3 IS: 2H6-25OHD3 3) 25OHD2 IS: 2H3-25OHD2

1) 25OHD3 2) 25OHD2 3) 24,25(OH)2D3 IS: 2H6-25OHD3

1) 25OHD3 IS: 2H6-25OHD3

Higashi et al., 2011 [41]

Schleicher et al., 2011 [71]

Wagner et al., 2011 [83]

Thibeault et al., 2012 [45]

2) 25OHD2 IS: 2H3-25OHD2

Analyte, Internal standard (IS)

Reference

Table 2 (continued)

a) Kinetex PFP (1.7 μm, 100 × 2.1 mm, a) TSQ Vantage, Thermo Phenomenex) Scientific b) MeOH : H2O (72:28 v/v) b) APCI+ c) 0.4 mL/min c) 1 + 2) 383 > 365, IS: 389 > 371, 3) 395 > 377, IS: 398 > 380

a) Eclipse C8 (1.8 μm, 50 × 3.0 mm) b) A) MeOH, B) H2O (gradient elution) c) 0.8 mL/min

a) Sunfire C-18 (3.5 μm, 50 × 2.1 mm, Waters) b) A) MeOH:H2O (98:2 v/v) + 0.1 % formic acid +2 mM ammoniumacetate, B) H2O + 0.1 % formic acid + 2 mM ammoniumacetat (85:15 v/v) c) 0.8 mL/min

Serum (100 μL), LLE (hexane)

Serum (200 μL), LLE (MTBE) → LLE (heptane)

Serum (200 μL), PP (MeCN) with filtration → on-line SPE (X-Terra C18 (5 µm, 20 x 2.1 mm, Waters)

Chromatographic run time

Linear range: 5 min 1) 4–160 nmol/L 2) 3–75 nmol/L LLOQ: 1) 4 nmol/L 2) 3 nmol/L

c) 1) 401 > 365, 2) 413 > 355, IS 407 > 371, 419 > 355

Linear range: 6.5 min 1) 0–437 nmol/L 2) 0–403 nmol/L 3) 0–60 nmol/L LLOQ: 1) 0.83 nmol/L 2) 0.57 nmol/L 3) 0.8 nmol/L

Linear range: 14 min 1) 12–150 nmol/L 2) 4–50 nmol/L 3) 8–100 nmol/L LLOQ: 1–3) not disclosed

Linear range: 12 min 1) 7.5–150 nmol/L 2) not disclosed LLOQ: 1) 7.5 nmol/L 2) not disclosed

Linear range, LLOQ

a) API 4000 Qtrap b) ESI+

a) API 5000, Applied Biosystems b) APCI+ c) 1) 401 > 383, 2) 413 > 395, 39 417 > 399, IS: 407 > 389

a) YMC-Pack Pro C18 RS (5 μm, a) API 2000, Applied 150 × 2.0 mm, YMC) Biosystems b) 10 mM ammonium formate : MeOH b) ESI+ c) 1 + 2) 649 > 340, IS: (5:1 v/v) + 5 mM methylamine c) 0.2 mL/min 663 > 340

DBS (3 mm diameter), Extraction (MeOH) → offline SPE → Derivatization (PTAD)

MS: a) Instrument b) Ionization c) Mass transition

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

172 Rev Endocr Metab Disord (2013) 14:159–184

1) 25OHD3 2) 25OHD2 3) 24,25OH2D3 4) 3-epi-25OHD3 IS: 2H6-25OHD3

1) 25OHD3 IS: 2H6-25OHD3 2) 25OHD2 IS: 2H3-25OHD2

1) 25OHD3 2) 25OHD2 3) 3-epi-25OHD3 IS: not disclosed

Baecher et al., 2012 [75]

Farrell et al., 2012 [27]

Lensmeyer et al.,2012 [73]

a) Method A) Acquity BEH C8 (1.7 µm, a) Method A) XE 50 x 2.1 mm, Waters), Method B) Premier (Waters), Acquity BEH phenyl (1.7 µm, Method B) TQD 50 x 2.1 mm, Waters) (Waters) b) A) H2O + 0.1 % formic acid + b) A + B) ESI+ 2 mM ammoniumacetat, B) c) 1) 401 > 159, 2) 413 MeOH + 0.1 % formic acid + > 395 (355 (B)), IS 407 2 mM ammoniumacetate > 159, 416 > 398 (358 (B)) c) Method A) 0.4 mL/min, Method B) 0.45 mL/min a) Zorbax SB cyanopropyl (3.5 μm, 100 × 4.6 mm, Agilent) b) MeOH:H2O:formic acid (670:330:2 v/v/v) c) 1.2 mL/min

Serum (150 μL), Method A: PP (Zn sulfate:MeOH) → off-line SPE (Oasis μelution HLB plate, Waters), Method B: PP (Zn sulfate:MeOH) → LLE (hexane)

Serum (300 μL), PP (MeCN/Zn sulfate/H2O) → off-line SPE (Strata C18 96-well plate, Phenomenex)

Table format adapted from [31]

24R,25(OH)2D3 = 2.4

1,25(OH)2D2 = 2.33

1,25(OH)2D3 = 2.40

25OHD2 = 2.42

25OHD3 = 2.5

D2 = 2.52

D3 = 2.60

Conversion factors:

a) API 4000 (MDS Sciex) b) APCI c) 1 and 3) 383 > 211, 2) 395 > 209, IS not disclosed

a) TSQ Quantum Ultra II, ThermoFisher b) APCI+ c) 1) 401 > 159, 2) 413 > 159, 3) 417 > 381, 4) 401 > 159, IS 407 > 159

a) Kinetex PFP (2.6 μm, 150 × 4.6 mm, Phenomenex) b) MeOH:0.5 mM ammoniumacetate (gradient elution) c) 0.8 mL/min

Serum (200 μL), PP (MeCN) → on-line SPE (LiChrospher RP-4 ADS 25 μm, Merck)

MS: a) Instrument b) Ionization c) Mass transition

LC: a) Column b) Mobile phase c) Flow rate

Sample type (sample volume), Sample preparation

Conversion from conventional units to SI units: Conversion factors x conventional units = SI units

Analyte, Internal standard (IS)

Reference

Table 2 (continued) Chromatographic run time

Linear range: not disclosed 1–3) not disclosed LLOQ: 1–3) not disclosed

1 + 2) not disclosed

LLOQ:

Linear range: not disclosed 1 + 2) not disclosed

Linear range: 21 min 1) 4.0–265.3 nmol/L 2) 3.9–183.6 nmol/L 3) 2.8–129.9 nmol/L 4) 2.0–133.8 nmol/L LLOQ: 1) 4.0 nmol/L 2) 3.9 nmol/L 3) 2.8 nmol/L 4) 2.0 nmol/L

Linear range, LLOQ

Rev Endocr Metab Disord (2013) 14:159–184 173

174

calibration curve is made in analyte-free serum, however, this is hard to obtain. Some use charcoal treatment of human serum [47], or an artificial matrix like phosphate-buffered saline with bovine serum albumin [25, 48, 49], use ethanolic calibrators [43, 44, 50] or commercially available human serum-based calibrator material [51]. Recently, ethanolbased calibrators have become available (National Institute of Standards and Technology (NIST)). Driven by financial restraints and/or high test volumes laboratories seek for ways to increase the throughput on LCMS/MS systems. Increased productivity can be achieved by innovations in sample preparation, chromatography as well as in mass detection, but require stringent conditions on the robustness and reliability of such approaches [52, 53]. Some LC-MS/MS methods for 25(OH)D measurement use derivatization procedure. Derivatization enhances the poor ionization efficiencies of steroids like vitamin D metabolites, leading to higher sensitivity and more specific detection. Disadvantages often include more laborious sample preparation requirements [32].

4 1α,25-dihydroxyvitamin D2/3 1,25(OH)2D3 (calcitriol) is the physiologically active form of vitamin D3. It is a product of the enzymatic hydroxylation of 25(OH)D3 (calcidiol) by CYP27B1 (1α-hydroxylase) that takes place primarily in the kidneys, but also to a lesser extent in extra-renal tissues such as placenta, bone and macrophages [54]. The plasma half life of 1,25(OH)2D is 4 to 8 h, as compared to approximately 2 to 3 weeks for 25(OH)D. The physiological regulation of 1,25(OH)2D is tightly regulated so that as vitamin D stores decline, 1,25(OH)2D is maintained due to increasing PTH action on 1-alpha-hydroxylation activity. As a consequence, circulating serum levels of 1,25(OH)2D are maintained despite declining substrate levels making measurements of 1,25(OH)2D levels not always clinically useful. Measuring serum levels of 1,25(OH)2D should be considered upon suspicion of deficiency or excess of 1,25(OH)2D leading to hypo- or hypercalcemia, or in monitoring of therapy with active vitamin D metabolites. Decreased concentrations are found in patients with renal insufficiency, vitamin Ddependent rickets type I (1-alphahydroxylase deficiency), X-linked hypophosphatemia, hypoparathyroidism and hypomagnesemia. Increased 1,25(OH)2D concentrations are found in Vitamin D-dependent rickets type 2 (end-organ resistance), hereditary hypophosphataemic rickets with hypercalcuria, primary hyperparathyroidism, granulomatous disorders and lymphoma (extra-renal production sites of 1,25(OH)2D). It is also important to remember that circulating 1,25(OH)2D provides essentially no information with respect to the patient’s nutritional vitamin D status

Rev Endocr Metab Disord (2013) 14:159–184

Off all the steroid hormones, 1,25(OH)2D represented the most difficult challenge to the analytical biochemist with respect to its quantitation. 1,25(OH)2D circulates at concentrations too low for direct UV quantitation (being in the low picomolar range (48–168 pmol/L; [55]), is highly lipophilic, and is relatively unstable. The first radioreceptor assay (RRA) for 1,25(OH) 2 D was introduced in 1974 [56], followed by RIA (DiaSorin and IDS). These all require extensive sample purification to minimize contribution of other vitamin D metabolites, as the 1,25(OH)2D antibody lacks specificity. The IDS Gamma-B RIA operates under the assumption that 1,25(OH)2D is the only quantitatively significant circulating 1α-hydroxylated vitamin D metabolite, an assumption that is clearly in error [57]. Given the low circulating levels of 1,25(OH)2D, antibody specificity is of main importance. In vitamin D intoxicated cases, using certain radioimmunoassays, falsely elevated 1,25(OH)2D levels may be found from co-measurement of various other circulating vitamin D metabolites. Also when using physical detection methods, chromatographic separation of other dihydroxylated metabolites is mandatory to avoid falsely increased concentrations of 1,25(OH)2D. The quantitation of this analyte by LC-MS/MS is challenging due to its low circulating serum concentration, presence of interfering substances in serum, and its poor ionization efficiency in ESI and APCI due to the lack of ionizable polar groups. To increase the ionization efficiency some researchers have used ammonium or lithium adduct formation [47, 58, 59]. The first report of a LC-MS/MS method for measurement of 1,25(OH) 2 D was from Kissmeyer et al. in rat and pig serum. They used ammonium adduct formation to achieve a lower limit of quantification (LLOQ) of 48 pmol/L by using 1 mL sample. However, no human serum data were reported. More recently, Casetta et al. [59] developed a method for quantitation of 1,25(OH)2D in human serum with the generation of stable Li-adducts. They used a rather complicated LC plumbing configuration with a perfusion column for online sample extraction, and two monolithic columns to separate 1,25(OH)2D3 from isobaric interferences. They achieved a LLOQ of 36 pmol/L with acceptable coefficients of variation (CV’s) of 5–15 % at physiological concentration levels. Another tool to enhance the ionization efficiency is the Diels-Alder derivatization. By the application of Cookson-type reagents, which rapidly and quantitatively react with the s-cis-diene structure of D compounds to give the Diels-Alder adduct, LOQs of 3– 60 pmol/L were obtained in different methods [60 pmol/L [60], 12 pmol/L [49], 60 pmol/L [61], 3 pmol/L in combination with immunoaffinity purification [55], 23 (D2) and 48 (D3) pmol/L [62]]. Recently, a variant Cookson type reagent with a quaternary amine as ionization enhancing group (QAO-Cookson) has been used for the quantification of 1,25(OH) 2 D achieving an 165 fold sensitivity

Rev Endocr Metab Disord (2013) 14:159–184

enhancement [LLOQ 10 pmol/L) [S. Dey, B. Williamson, S. Pillai, S. Purkayastha, MSACL Meeting 2011, Abstract, 2011, p. 73, Available from: https://www.msacl.org/msacl conference2011.php.]. Instead of increasing effort on optimizing the MSconditions to further improve assay sensitivity, in the last year, focus has shifted towards improvements in sample preparation. An elegant way for analyte enrichment is the use of immunoaffinity purification [63]. The immunoaffinity extraction removes isobaric interferences and matrix effects present in patient serum with significant improvement in MS detection. Yuan et al. [47] used the 1,25(OH) 2 D 2 / 3 ImmunoTube LC-MS/MS kit from Immundiagnostik for immunoaffinity purification in combination with lithium adduct formation to improve ionization efficiency. The determined LLOQs were 8.2 pmol/L for 1,25(OH) 2 D 3 and 9.1 pm ol/L for 1,25(OH) 2 D 2 . Strathmann et al. [55] performed protein precipitation, followed by immunoextraction with solid-phase anti1,25(OH)2D antibody from IDS and finally performed a derivatization with 4-phenyl-1,2,4-triazoline-3,5-dione (PTAD). LLOQ was 3 pmol/L. Both the Strathmann and Yuan methods showed excellent agreement to a reference LC-MS/MS method for 1,25(OH)2D (Mayo Clinics, Rochester), whereas in comparison with the DiaSorin RIA the LC-MS/MS gave systematically lower results (bias −27.1 %) indicating cross-reacting substances that have not been removed by the RIA sample preparation [47]. So, by combining immunoaffinity purification with either lithium-adduct formation or Diels-Alder derivatization, sensitive and specific methods for quantitation of 1,25(OH)2D have been developed and these are particularly usefull for accurate measurement of lowered 1,25(OH)2D concentrations in various disease states, e.g. in Multiple Sclerosis.

5 3-Epi-25-hydroxyvitamin D3 It was recently shown that 25(OH)D3, like 1,25(OH)2D3 and 2 4, 25 ( O H ) 2 D 3 , is m e tab oli ze d th ro ug h the C3 epimerization pathway [64]. The 3-epi-1,25(OH)2D3 is nearly as potent as 1,25(OH)2D3 in suppressing PTH secretion [65], but has only reduced calcemic properties [66]. Still, the biological relevance of 3-epi-25(OH)D remains to be elucidated. The C3-epimer has received attention from its detection in 23 % of infant sera less than 1 year old with 3epi-25(OH)D contributing 9–61 % of the total 25(OH)D using a LC-MS/MS method which partially separated 25(OH)D from its C3-epimer [67]. There was a modest inverse relationship between patient age in days and the percentage of 3-epi-25(OH)D. The 3-epi-25(OH)D metabolite was not detected in children from 1 to 18 years of age or

175

in adults. This led the authors to postulate that C3epimerisation originates from immature vitamin D metabolism. A limitation of their study was that 3-epi-25(OH)D was only partially separated from the major 25(OH)D peak which did not allow them to quantify low (20). The concentration of 3-epi-(OH)D3 ranged from 0.25 to 59.3 nmol/L (mean 3.75 nmol/L). Plotting the concentration of 3epi-(OH)D3 to the concentration of 25(OH)D3 it could be observed that the epimer amount increased as 25(OH)D3 increased in a nonlinear mode [75]. Even if the physiological role of the C3-epimerisation pathway has not yet been elucidated, it may become important in the future resulting in routine measurement of 3epi-(OH)D as well [68]. At present, most focus is on eliminating potential contribution of the C3-epimer in assays for 25(OH)D measurement, at least in infants (100 nmol/l) [75, 83]. When measuring 24R,25(OH)2D it is important to realize that it is the R-form, and not the unnatural epimer 24S,25(OH)2D, that is metabolically active [86]. The quantification of this analyte is challenging because of its low serum level and low ionization efficiency in ESI and APCI [60]. In some LC-MS/MS methods for 25(OH)D not using derivatization, 24R,25(OH)2D is also being measured, albeit at suboptimal sensitivity [83, 87]. Higashi et al. [88] has described a dedicated LC-MS method for determination of 24R,25(OH)2D3 in human plasma using derivatization with a Cookson-type reagent (4-[4-(6-methoxy-2-benzoxazolyl)

When measuring vitamin D metabolites by LC-MS/MS one should be aware of interferences from isomeric and isobaric compounds. Isomeric compounds have identical molecular formula and nominal mass, but different structure (e.g. 25(OH)D3 versus 1α(OH)D3). Epimers are diastereomers that differ in configuration of only one stereogenic center. Isobaric compounds are compounds with the same nominal mass but with a different molecular formula (e.g. 25(OH)D3 versus 7α-hydroxy-4cholesten-3-one). Discrimination of isomers and isobars might be achieved from differences in chromatographic behavior and fragmentation in MS. As many as 40–50 metabolites of vitamin D have been identified [89], apart from numerous chemically synthesized analogs [90]. Main isobaric interferences in LC-MS/MS analysis of 25(OH)D3 are by naturally existing 3-epi25(OH)D3 [67] and 7α-hydroxy-4cholesten-3-one [42, 50], which is an intermediate in the synthesis of bile acids from cholesterol [91], and analog 1α(OH)D3 (alfacalcidol) [42, 50], which is sometimes used for the treatment of osteoporosis, rickets and hypocalcemia [92]. In the case of LCMS/MS analysis of 1,25(OH)2D3, care should be taken to chromatographic separation of several naturally circulating di-hydroxylated vitamin D compounds (e.g. 23,25(OH)2D, 25,26(OH)2D, 24,25(OH)2D, 4β,25(OH)2D), as they share the same molecular weight and may form the same mass to charge parent and product ion pairs upon ionization. Wang et al. [61] identified 4ß,25-(OH)2D3, an isomer to 1α,25dihydroxyvitamin D3 which was found at concentrations comparable to 1α,25-dihydroxyvitamin D3. The authors postulate that in some published methods for 1,25(OH)2D3 the 1,25(OH)2D3concentration might be overestimated from coelution of 4β,25(OH)2D3 and 1,25(OH)2D3. Yuan et al. described 23,25(OH)2D3 as a potential interfering compound, although no signification concentrations were found in patient samples [47]. Upon vitamin D intoxication or high dose vitamin D supplementation, other pathways of 25(OH)D3 metabolism emerge as alternatives to 1- and 24-hydroxylation, with

Rev Endocr Metab Disord (2013) 14:159–184

continuous rise of 25(OH)D3-26,23-lactone [93, 94]. It is well known that some immunoassays for 1,25(OH) 2D overestimate true 1,25(OH)2D concentrations from crossreactivity to 25(OH)D 3 -26,23-lactone. As 25(OH)D 3 26,23-lactone differs from 1,25(OH) 2 D 3 in molecular weight (no isobaric compound), it will however not interfere in the determination of 1,25(OH)D3 by LC-MS/MS.

8 Determination of vitamin D metabolites in alternative biological samples

177

neonatal DBS without interference from 3-epi-25(OH)D3 [41]. The method employed two steps of derivatization, a Diels-Alder reaction with PTAD followed by acetylation, to enhance the detectability of 25(OH)D3 in ESI-MS/MS (LLOQ 7.5 nmol/L) and to separate 25(OH)D3 from its C3-epimer. The growing awareness that developmental vitamin D deficiency can lead to adverse health outcomes in the offspring has led to an increased focus on the neonatal assessment of vitamin D status [98], and may result in growing interest of DBS for 25(OH)D measurement. 8.2 Saliva

Although 25(OH)D is usually measured in serum from venous blood, a number of other sample types have been exploited, including dried blood spots, saliva and cerebrospinal fluid. 8.1 Dried blood spot (DBS) The DBS technique has been in use for a number of years, particularly for the diagnosis of inborn errors of metabolism [72, 95]. DBS sampling is usually from heel- or fingerprick reflecting capillary blood. DBS sampling offers advantages for the collection, storage and shipping of blood samples, but potential problems with sample extraction do exist [95]. A single 3.2 mm DBS punch is equivalent to only 3.28 μL whole blood requiring sensitive assay technology. Eyles et al. [96] have described an LC-MS/MS assay for measurement 25(OH)D in DBS. 25(OH)D was extracted from 3.2 mm DBS punches with addition of deuterated 25(OH)D3 as internal standard, derivatized with PTAD prior to analysis by LC-MS/MS. The assay had good accuracy and precision, and was highly sensitive, being capable of detecting