|Home | About | Journals | Submit | Contact Us | Français|
25-hydroxyvitamin D (25OHD) determination is of diagnostic importance for the investigation of vitamin D deficiency and much more rarely, intoxication. Despite the name, vitamin D is a pre-hormone, being endogenously synthesised provided there is adequate sunlight. Its biological function, exerted through the active form 1,25 dihydroxyvitamin D3 (1,25(OH)2D) is to maintain calcium and phosphate levels in the blood.1 In addition, it has important roles in immune regulation.2
Demand for 25OHD and 1,25(OH)2D assays has increased substantially worldwide since the introduction of commercial kit assays, but performance is relatively poor with only just over half of laboratories achieving acceptable performance.3 A recently emerging problem is that some immunoassays underestimate 25-hydroxyvitamin D2 metabolites due to differences in affinity between the antibodies or D-binding proteins employed.4
It is likely that achievement of reliable, clinically appropriate vitamin D results will require a combination of careful use of commercial systems by skilled operators, validated reference ranges, good quality control schemes and the availability of reference methods of analysis.
Vitamin D3 (cholecalciferol) is formed in the skin by photolysis of 7-dehydrocholesterol by ultraviolet radiation from sunlight. Cholecalciferol is transported in the circulation bound to vitamin D binding protein and is hydroxylated in the liver to 25OHD. A further hydroxylation reaction occurs in the kidney to form the active hormone 1,25(OH)2D. This reaction is tightly regulated by induction of the CYP27B1 enzyme which is stimulated by PTH and inhibited by hyperphosphataemia and 1,25(OH)2D.5
Ergocalciferol (vitamin D2) is produced commercially by ultraviolet irradiation of a provitamin D sterol (ergosterol) that occurs in plants.6 It differs from vitamin D3 in having an additional methyl group at C24 and a double bond at C22-23. It undergoes the same hydroxylation reactions as cholecalciferol to form 25OHD2 and 1,25(OH)2D2. Ergocalciferol is the only prescription pharmaceutical available in Australia7 and the only high dose preparation available in the USA8 so D2 metabolites can represent a significant fraction of the circulating hormone in patients treated with these preparations (see later).
Notwithstanding the International Union of Pure and Applied Chemistry (IUPAC) recommendation9 for the use of the trivial names calcidiol (25-hydroxyvitamin D, 25OHD) and calcitriol (1,25-dihydroxyvitamin D, 1,25(OH)2D), these two terms have not come into widespread use. As a consequence the abbreviations 25OHD and 1,25(OH)2D are used in this review.
Although cholecalciferol and 1,25(OH)2D can be measured in the circulation, the best estimates of vitamin D status are provided by measurement of 25OHD 10,11 This is due to its long serum half-life (approximately 3 weeks) and because the 25-hydroxylation step is unregulated, thus reflecting substrate availability. A number of commercial kit assays are available for the clinical laboratory and will be discussed further.
In contrast, cholecalciferol has a short half-life (approximately 24 h) so that serum levels depend on recent sunlight exposure and vitamin D ingestion. The assay is difficult due to the lipophilic nature of the molecule and no commercial versions are available.12
Since production of 1,25(OH)2D is tightly regulated and serum half life is 4–6 h, circulating levels provide limited information about nutritional vitamin D status. Although commercial radioimmunoassays and ELISAs are now available, measurement of 1,25(OH)2D is principally of interest only in renal disease and remains the preserve of specialist laboratories.
Approximately 85% of 25OHD is bound to D binding protein (DBP), 15% to albumin, and 0.03% free.13 Thus chromatographic separation requires an extraction step to release 25OHD from its binding proteins and this can be subject to variable co-precipitation.14 On the other hand, non-extracted immunoassay methods may be susceptible to matrix effects, particularly due to the lipophilic nature of 25OHD.15 Taken together with the low (nanomolar) levels of vitamin D metabolites in serum, these factors have made the routine measurement of 25OHD an analytical challenge. Clinical assessment of vitamin D status in patients receiving ergocalciferol requires measurement that is equally reactive to both the 25OHD3 and the 25OHD2 metabolites. It has recently become clear that not all of the currently available commercial immunoassays will detect 25OHD2 with adequate sensitivity.15 Although D2 is used in the USA and Australia, D3 is the form given in Europe and so these problems may be confined to Australia and the USA. However, these analytical issues may be exacerbated by the less efficient conversion of D2 to 25OHD2 compared to the D3 forms.16 Furthermore, the half life of 25OHD2 is shorter and clinical potency is less than one third.8 Together, this means that the clinical picture in an ergocalciferol-treated patient may be highly uncertain.
Definitive methods employ GC coupled with mass spectrometry detection.17,18 Recently a candidate reference method using LC-tandem mass spectrometry was published.14 Whilst these reference methods are suitable for validating the recovery and accuracy of routine methods, their complexity and derivatisation requirements mitigate against regular use.
A large number of HPLC methods for vitamin D determination have been published. Chromatographic separation is suitable as a reference method, since it is capable of resolving D2 and D3 forms as well as the 25OH, 1,25(OH)2 and 24,25(OH)2 metabolites. Early assays utilised normal phase separation19,20 and these have been followed by reverse phase separation.21 Many adaptations and variations have been published (reviewed15,22). The most widely adopted methods use liquid-liquid or liquid-solid pre-sample cleanup with UV detection after column separation. A C18 reverse phase column with isocratic or gradient elution using acetonitrile/water is now the standard procedure and diode array detection following hexane extraction gives results of sufficient clinical sensitivity.23 Thus determination of 25OHD by HPLC with UV detection can be considered the gold standard method.15 However, it is accepted that these methods are unsuitable for routine, clinical laboratory use.
The earliest immunoassay method used DBP and 3H-25OHD tracer.24 An RIA using antibody that is cospecific to both 25OHD2 and 25OHD3 was developed by Hollis25 and was improved with the use of 125I-labelled tracer.26 More recently, chemiluminescent assays have utilised both DBP27 and antibody-based binding.28 A range of immunoassays is now commercially available (in Australia, the DiaSorin, IDS, and Nichols Institute assays comprise 90% of the market29).
The Hollis assay25,26 was released commercially with FDA approval and has become widely used. Acetonitrile extraction is followed by competitive radioimmunoassay using 125I-labelled 25OHD and antibody to 25OHD. A second antibody is used as precipitating agent. Although Hollis found that the primary antibody recognises 25OHD2 and 25OHD3 equally,4 other authors have published data indicating that the assay under-recognises 25OHD2.30
The sample obtained after acetonitrile extraction is incubated with antibody to 25OHD in competition with 125I-labelled 25OHD. A second antibody coupled to cellulose is used for separation of bound radioactivity. The manufacturers state that there is 75% cross reactivity of the antibody with 25OHD2 compared to 100% for the 25OHD3 form.31 Interestingly, data from the DEQAS indicate that the EIA version of this IDS assay performed better than the RIA, despite using the same antibody.32
This assay first separates 25OHD from DBP using a denaturing agent. Competition is then established in the same sample well between the sample 25OHD and 25OHD on magnetic particles for human DBP. Separation utilises the magnetic particles and detection is by chemiluminescence using acridinium-ester.27
Despite the Nichols Advantage manufacturers earlier claims for 100% reactivity with 25OHD2, it has become apparent that the assay is unable to measure samples containing substantial amounts of 25OHD2 reliably.15 Recently, data from the UK quality assurance program has shown that the assay displays a positive bias (~31%) with 25OHD3 and a negative one with 25OHD2.32 This under-recovery of 25OHD2 in some patient samples is now acknowledged in the information supplied with the kit.
This chemiluminescent assay has recently become available. Serum is incubated with antivitamin-D coated microparticles and isoluminol derivative-conjugated 25OHD before measurement of the chemiluminescent signal.28 The antibody is said to be same as that used in the DiaSorin RIA33 and the results correlate well with the radioimmunoassay.34 However, this assay was found to recognise 25OHD2 more than 25OHD3,35 an anomaly that remains to be explained.
Standardization of results between methods and laboratories remains a significant problem.30,36,37 This may be due to variability with temperature of antigen-antibody or protein-binding protein interactions as well as to differences in recognition of the 25OHD2 and 25OHD3 forms. Furthermore, the Diasorin RIA showed good correlation with HPLC in experienced hands but compared poorly in a laboratory that had not validated the method,37 emphasising the need for good quality assurance. The Nichols Advantage showed poor comparison with HPLC and overestimated basal 25OHD levels whilst underestimating exogenous 25OHD3. Recently this kit underwent a re-standardisation exercise due to the results having drifted downwards in the Australian QAP. This resulted in values increasing by 17% in an unselected patient population (Martin P, 2004 personal communication).
An international quality assurance scheme (DEQAS, UK) has been in operation since 1989.3 At present there are 16 Australian laboratories participating (DEQAS, personal communication). Additionally in Australia, the RCPA-AACB QAP has 36 laboratories participating in the Vitamin D Endocrine program.29 Results from the DEQAS scheme have highlighted the 25OHD2 detection problem32 as well as confirming that only half the participants are achieving acceptable performance (80% of results within 30% of All Laboratories Trimmed Mean).
Serum is the preferred specimen although plasma (EDTA and Li-heparin) samples are satisfactory.34 Vitamin D analytes have been shown to be stable for up to 2 weeks at 30°C3 and to be unaffected by up to 4 freeze-thaw cycles.38 Storage of frozen serum samples at −20°C for up to one year has also been reported to cause no loss in vitamin D metabolites.12
Measurement of 25OHD by immunoassay will remain the method of choice for reasons of convenience, speed, turnaround and cost. Although the current generation of commercial assays have shortcomings in defining the vitamin D status of patients being treated with ergocalciferol, now that this problem has been recognised, performance will undoubtedly improve with the development of better assays. Given the problems caused by ergocalciferol outlined here, it may also be reasonable to recommend a change to using D3 as the pharmaceutical of choice.
Continuing efforts to improve laboratory performance and vigilance with quality assurance programs are required. Laboratories offering vitamin D assays are urged to participate in the local and international programs.
The establishment and maintenance of reference assays for 25OHD are required and have been supported by the AACB. An HPLC assay is being established at RMIT to provide estimation of both 25OHD2 and 25OHD3 metabolites in plasma samples so that the performance of routine immunoassays on quality assurance samples containing mixtures of these metabolites can be monitored.
The contents of articles or advertisements in The Clinical Biochemist – Reviews are not to be construed as official statements, evaluations or endorsements by the AACB, its official bodies or its agents. Statements of opinion in AACB publications are those of the contributors. Print Post Approved - PP255003/01665.
No literary matter in The Clinical Biochemist – Reviews is to be reproduced, stored in a retrieval system or transmitted in any form by electronic or mechanical means, photocopying or recording, without permission. Requests to do so should be addressed to the Editor. ISSN 0159 – 8090.