Comparison of high performance liquid chromatography, radio immunoassay and electrochemiluminescence immunoassay for quantification of serum 25 hydroxy vitamin D

https://doi.org/10.1016/j.clinbiochem.2011.04.020Get rights and content

Abstract

Objective

To compare the performance of radioimmunoassay (RIA) with high-performance liquid chromatography (HPLC) and electrochemiluminescence (ECLIA) for the quantification of vitamin D (25OHD).

Methods

HPLC method for the determination of 25OHD in human biological samples was developed and compared in terms of accuracy and precision with a commercially available RIA assay. Performance of RIA assay with ECLIA technology for 25OHD analysis was further compared.

Results

Median 25OHD levels with HPLC vs. RIA were 50.1 nmol/L (IQ = 17.7–199.4 nmol/L) and 51.1 nmol/L (IQ = 12.5–187.2 nmol/L) respectively, whereas median 25OHD concentration with RIA vs. ECLIA was 32.4 nmol/L (9.98–199.7 nmol/L) and 29.9 nmol/L (4.9–214.6 nmol/L), respectively. Comparison data for HPLC vs. RIA showed RIA =  1.13 + 1.01 (HPLC) (RMSE = 11.2 nmol/L) and for RIA vs. ECLIA revealed, ECLIA = 3.21 + 0.9 (RIA) (RMSE9.6 nmol/L).

Conclusion

Acceptable correlation was observed among HPLC and RIA and also with RIA and ECLIA in quantification of 25OHD.

Introduction

Recent research has identified vital role of vitamin D for muscle functioning, prevention of type 1 diabetes mellitus, hypertension, and many common cancers [1]. Overall vitamin D status is best assessed by measuring circulating 25-hydroxy vitamin D (25OHD) which is a complex task [2], [3]. 25OHD measurement has also become essential in the management of various metabolic bone diseases, neonatal hypocalcaemia, pregnancy, hypoparathyroidism and post menopausal osteoporosis [4], [5].

The two major forms of vitamin D are cholecalciferol (vitamin D3) from animal sources and ergocalciferol (vitamin D2) from plant sources. Cholecalciferol is derived mainly from actions of ultraviolet light on the skin (at wavelengths of 295–315 nm) [6]. In the body, it binds to vitamin D binding protein in plasma and is transported to the liver where, both are hydroxylated forming 25OHD. Unlike parent vitamin D, 25OHD is the key storage form and is considered to be the marker of overall nutritional status of vitamin D [7].

Quantification of 25OHD is not a trouble-free mission [8]. Around 85% of 25OHD is bound to D binding protein, 15% to albumin, and 0.03% free. 25OHD strongly binds with D binding protein and this link must be broken for its extraction in high performance liquid chromatography (HPLC) methods [9], [10], [11]. Non extracted immunoassay methods are prone to matrix effects due to the lipophillic nature of 25OHD metabolite. This prevents binding of 25OHD with the antibody in the sample or standard and affects the validity of the assay.

Historically, the development of assays for the measurement of vitamin D metabolites in body fluid has progressed very slowly. Initial assays in the 1970s for 25OHD were based on competitive binding principle [12]. The major shortcoming of these assays was long incubation period and cumbersome extraction procedures. Later in 1978, HPLC assay was introduced [13]. However, this method has the benefit of separately quantifying vitamin D2 and vitamin D3 and is long been recognized as the gold standard. The disadvantages of HPLC included tedious technique, high cost, large sample volume and proficient expertise to perform this type of analysis.

In 1985 the first commercially available RIA for measurement of 25OHD was approved by the Food and Drug Administration for clinical use [14]. Both RIA and HPLC give precise results but required tedious preanalytical preparations and restrict sample throughput in large clinical laboratories. Recently many new RIA and enzyme immunoassays have been developed, with ease of automation.

Lack of standardization of 25OHD assays leads to errors in assessment of vitamin D status. In this paper we have compared three different assays for the quantification of 25OHD so that we can have a comparative validation in terms of accuracy and precision used in our laboratory and identify the best procedure suitable for a commercial laboratory.

Section snippets

Materials and methods

The analytical performance of 25OHD methods performed by a manual radioimmunoassay (RIA) (kit from DiaSorin,) was compared with a high performance liquid chromatographic (HPLC) method on Perkin Elmer series 200 with ultraviolet detection (n = 40) and RIA (kit from DiaSorin) with electro chemiluminescence immunoassay (ECLIA) on a Roche Modular E-170 (n = 20).

Samples

Out of all human blood samples collected for 25OHD analysis in the clinical laboratory two subsets were randomly picked using Statistical Package for the Social Sciences (SPSS). The analysis of 25OHD by HPLC was conducted at Research Lab at Aga Khan University, while samples were analyzed with RIA and ECLIA at Section of Chemical Pathology, Aga Khan University Karachi.

  • I.

    Methodology for RIA:

  • i.

    Equipment: Genesys multi well gamma scintillation counter was used to record radioactivity of the molecules.

  • ii.

Statistical analysis

SPSS version 16 was used for statistical analysis of the results. Regression analysis was performed as part of the method to evaluate the performance of the methods and to measure to what extent there is a linear relationship between two variables. Linear regression was done using Passing Bablok. The data was plotted and carefully examined for non linearity. Agreement between the two methodologies was assessed using graphical plot as described by Bland and Altman.

Results for HPLC and RIA

Forty samples for 25OHD were analyzed on HPLC and RIA simultaneously. Median serum 25OHD using RIA and HPLC were 51.1 nmol/L (IQ = 12.5–187.2 nmol/L) and 50.1 nmol/L (IQ = 17.7–199.4 nmol/L) respectively. Paired t test revealed that there was no statistically significant difference between the two methods (p = 0.8). Passing Bablok regression analysis between the two methods yielded RIA =  1.13 + 1.01 (HPLC) nmol/L. Confidence interval for intercept and slope was − 9.4–5.1 nmol/L and 0.9–1.1 nmol/L respectively.

Results for RIA and ECLIA

Twenty blood samples for 25OHD were analyzed using RIA and ECLIA methodologies. Median 25OHD concentration by RIA and ECLIA were 32.4 nmol/L (IQ = 9.98–199.7 nmol/L), and 29.9 nmol/L (4.9–214.6 nmol/L), respectively.

Paired t test was applied and no statistically significant difference between the two methods was found (p = 0.37).

Passing Bablok regression analysis generated ECLIA = 3.21 + 0.9 (RIA) nmol/L. Confidence interval for intercept was 0–6.24 nmol/L and for slope was calculated to be 0.8–1.0 nmol/L. (

Discussion

The Passing-Bablok analysis of agreement between HPLC and RIA and also for RIA and ECLIA showed that the observations were close to the “perfect fit” slope of 1. The Bland–Altman analysis of bias revealed that there was no significant bias between HPLC and RIA 25OHD values (observed bias, 0.42; 95% confidence intervals, − 21.5 to 22.4 nmol/L). Similarly Bland–Altman analysis between RIA and ECLIA revealed that there was no significant bias between the two methodologies for 25OHD quantification

References (22)

  • M.F. Holick

    High prevalence of vitamin D inadequacy and implications for health

    Mayo Clin Proc.

    (2006 Mar)
  • J.C. Souberbielle et al.

    Update on vitamin D and evaluation of vitamin D status

    Ann Endocrinol (Paris).

    (2008 Dec)
  • B.W. Hollis

    Measuring 25-hydroxyvitamin D in a clinical environment: challenges and needs

    Am J Clin Nutr.

    (2008 Aug)
  • S. Mansoor et al.

    Jabbar

    Prevalence and significance of Vitamin D deficiency and insufficiency among apparently healthy adults. Clinical Biochemistry.

    (2010)
  • M.F. Holick

    Vitamin D: a D-Lightful health perspective

    Nutr Rev

    (2008 Oct)
  • M.F. Holick et al.

    Deficiency

    N Engl J Med

    (2007)
  • Y. Inoue et al.

    Vitamin D status affects osteopenia in postmenopausal patients with primary hyperparathyroidism

    Endocr J.

    (2008 Mar)
  • S. Masuda

    Vitamin D: its chemistry, metabolism and biological functions

    Nippon Rinsho.

    (1999 Oct)
  • B.W. Hollis

    Editorial: the determination of circulating 25-hydroxyvitamin D: no easy task

    J Clin Endocrinol Metab.

    (2004 Jul)
  • R. Vieth

    Problems with direct 25-hydroxyvitamin D assays, and the target amount of vitamin D nutrition desirable for patients with osteoporosis

    Osteoporos Int.

    (2000)
  • R. Bouillon et al.

    Two direct (nonchromatographic) assays for 25-hydroxyvitamin D

    Clin Chem.

    (1984 Nov)
  • Cited by (43)

    • An ultra-sensitive and highly selective impedimetric sensor for vitamin D measurement based on a novel imprinted polymer synthesized utilizing template-derived functional monomer

      2022, Analytica Chimica Acta
      Citation Excerpt :

      Hence, vitamin D is a key molecule for the prevention of various chronic illnesses, leading to an increased requirement for regular tracing of vitamin D levels in patients. The existing techniques available for the determination of VD concentration in biological samples are high-performance liquid chromatography-mass spectrometry (HPLC-MS) [4] radioimmunoassay (RIA) [5] and chemiluminescence [6]. Although the methods cited above are highly precise and sensitive, these techniques are expensive, time-consuming, and need a highly skilled person.

    • Gold‑platinum bimetallic nanoparticles coated 3-(aminopropyl)triethoxysilane (APTES) based electrochemical immunosensor for vitamin D estimation

      2020, Journal of Electroanalytical Chemistry
      Citation Excerpt :

      The detection of 25(OH)D3, the major circulating form of vitamin D, is carried out to evaluate the concentration of vitamin D in human serum [16]. Various conventional methods such as radioimmunoassay [17], HPLC-MS [18], chemiluminescence [19] methods have been employed for the detection of vitamin D. Though, these methods are accurate, reliable and have the required sensitivity but their expensiveness and longer analysis time make these techniques unsuitable for routine use [20]. On the other hand, electrochemical immunosensing technique can be used owing to its cost effectiveness, short analysis time, sensitivity and point of care diagnosis.

    • An efficient electrochemical biosensor for Vitamin-D<inf>3</inf> detection based on aspartic acid functionalized gadolinium oxide nanorods

      2019, Journal of Materials Research and Technology
      Citation Excerpt :

      The various reports on Vit-D suggest the required level of Vit-D is about 30 ng mL-1 and the value lower than 30 ng mL−1 is associated with Vit-D deficiency [11,12]. The conventional techniques used for detecting Vit-D concentration are enzyme-linked immunosorbent assay, mass spectrometry, chromatography, and radioimmunoassay etc. [13–15]. These techniques consume more time, require sophisticated instrumentation and are expensive so there is an increased demand for analytical techniques for Vit-D determination which should be reliable, selective, easy to operate, rapid and economical.

    View all citing articles on Scopus
    View full text