Rapid supercritical fluid chromatography–tandem mass spectrometry (SFC-MS/MS) for the routine quantification of retinol and α-tocopherol in human serum and plasma

As interest in vitamins A and E has increased in recent years, so too has the need for rapid and reliable methods by which these compounds can be assayed. The development of a supercritical fluid chromatography–tandem mass spectrometry (SFC-MS/MS)-based method with an automated extraction procedure provides faster analysis than other chromatographic-based methods and improved specificity.


In the clinical laboratory, retinol and α-tocopherol are most commonly measured in serum (or plasma) to estimate the nutritional status of vitamins A and E, respectively. Chromatographic-based methods, such as high-performance liquid chromatography (HPLC) and liquid chromatography–tandem mass spectrometry (LC-MS/MS), have been used since the 1980s to perform the analysis but involve long and tedious extraction procedures, and the use of large volumes of solvents and serum. Our objective was to develop a selective and sensitive rapid method based on supercritical fluid chromatography–tandem mass spectrometry (SFC-MS/MS) with an automated extraction procedure that required the minimal use of solvent and patient material.

Biological functions of vitamins A and E Vitamin A

Vitamin A occurs in a variety of forms that can be divided into two groups: retinoids and provitamin A carotenoids. Naturally occurring retinoids include retinol (Fig. 1), retinol esters, retinaldehyde and retinoic acid, whereas the major provitamin A carotenoid is β-carotene [1]. Retinol is found in the circulation and transported by retinol binding proteins (RBP) [2]. It is a precursor for two essential molecules: 11-cis-retinol and all-transretinoic acid. The most understood action of vitamin A is its role in vision. In brief, retinol is transported in the blood by RBP to the retina of the eye. Retinol is then esterified to form retinyl esters, which can be stored in retinal pigment epithelial cells. When needed, retinyl esters are hydrolysed and isomerized to 11-cis-retinol and then oxidized to 11-cis-retinal. The 11-cis form of retinal is a component of the visual pigment rhodopsin, which is found in the rods of the retina. On absorption of light, rhodopsin isomerizes to transretinal and subsequently releases opsin that results in a series of conformational and other changes, leading to the generation of an electrical signal to the optic nerve to relay visual imagery to the brain [3]. All-transretinoic acid modulates gene expression by activation of nuclear receptors, of which there are two groups: retinoic acid receptors (RAR) that bind all-trans-retinoic acid (and some other retinoids); and retinoid X receptors (RXR) that bind 9-cis-retinoic acid. Other metabolic roles of vitamin A include maintenance of epithelial barriers, immune competence, reproduction and embryonic growth and development [3].

Vitamin E

Compounds with vitamin E activity contain a chromanol ring attached to a long phytyl side chain (Fig. 2). There are eight naturally occurring forms: α-, β-, γ-, and δ-tocopherols and -tocotrienols [1]. α-Tocopherol is the most abundant form of vitamin E in the circulation, and accounts for about 90% of the vitamin E in human tissue [4]. It is a potent antioxidant and aids the prevention of lipid peroxidation, such as phospholipids within cell membranes. The mechanism by which it exerts its protective effect is through the absorption of peroxyl radicals, which results in the termination of the free radical chain reaction. Vitamin E is transported to and from the liver via various lipoproteins and incorporated into cell membranes where it is anchored by its hydrophobic tail. It is suggested that vitamin E antioxidant properties help to prevent or delay disease and disorders linked to oxygen free radicals. These include cardiovascular disease in certain patient subgroups such as those on hemodialysis or in diabetic patients [1] and patients with cognitive impairment and neurological diseases such as Alzheimer’s in which amyloid-β protein causes cytotoxicity through oxidative stress [4].

Vitamin E also modulates the transcription of certain genes, inhibits platelet aggregation and vascular smooth muscle proliferation and has an effect on cell signalling in the immune system. Of note, α-tocopherol has a significant role in the prevention of miscarriage in humans [5].

Assessing vitamin A and E status

Requests for the determination of vitamin A and vitamin E status are most commonly made so that patients with a deficient state can be identified. Deficiency is almost exclusively seen in patients with an impaired ability to absorb fat-soluble vitamins, for example in patients with cystic fibrosis and short bowel syndrome, or those that have impairment of fat absorption or metabolism, such as pancreatic insufficient cystic fibrosis patients. Other requests for vitamin A and E status evaluation include patients that are on supplementation, such as enteral and parenteral nutrition patients [6].

As interest in vitamins A and E has increased in recent years, so too has the need for rapid and reliable methods by which these compounds can be assayed. Determination of fat-soluble vitamins in serum typically involves lengthy sample preparation (such as saponification and extraction), followed by HPLC coupled to ultraviolet (UV) or fluorescence detection. Vitamins A and E are often determined co-currently in the same analytical procedure because of their similar sample preparation requirements [6]. The authors’ laboratory previously used reversed phase chromatography with UV detection which featured long chromatographic run times requiring considerable time and solvent usage during sample analyses. In order to meet the increasing demand for analysis, we validated and implemented a SFC-MS/MS-based method for the assessment of retinol and α-tocopherol.

HPLC-UV method limitations

  • Elution of hydrophobic phospholipids from the column, such as forms of vitamin E, requires a high percentage of organic solvent. An estimated 22 ml of methanol mobile phase was used per sample during the HPLC method of analysis.
  • Phospholipids render the method prone to matrix effect, resulting in co-elution and interfering peaks.
  • Although HPLC methods of analysis for vitamins are well established, they require considerable consumption of solvents and time. Approximately 15 minutes are required to elute and quantify serum retinol and α-tocopherol per sample.
  • UV detection is a less sensitive and less selective method of detection when compared to mass spectrometry.

Supercritical fluid chromatography

Instrumental and technological advances have facilitated the adoption of supercritical fluid chromatography (SFC) by clinical laboratories. The application for vitamins A and E is based on the principle that under pressure, CO2 occurs as a liquid that has unique properties as a solvent for chromatography and demonstrates high diffusivity and low viscosities. This allows for enhanced chromato-graphic efficiency and resolution [7]. While conventional HPLC systems require high flow rates to achieve reduced run times, the high flow rates can reduce desolvation and suppress ionization and therefore reduce the sensitivity of the mass spectrometer. An important advantage of SFC is the possibility of using sub-2-µm particles in the analytical column which allows for higher flow rates, thus offering much shorter separation times. For detection, SFC can be coupled to the same panel of detectors as HPLC, including UV/fluorescence detectors or mass spectrophotometers.

The ACQUITY Ultra Performance Convergence Chromatography (UPC2; Waters) is an example of a commercially available SFC solution which works through the combination of multiple solvent systems and pressure regulators. The system makes use of liquid organic modifiers of different polarities, which when combined with the supercritical CO2 and wide range of stationary phases results in exceptional versatility and selectivity. An automated back-pressure regulator is a crucial component of UPC2 systems as this prevents the decompression of the supercritical CO2 during chromatographic analysis. Figure 3 shows the schematic diagram of the SFC (UPC2)-MS/MS system. After the post-column regulator, the CO2 decompresses and loses its solvating power, and therefore the decompression must be managed without sacrificing efficient analyte transport into the ionization source. This is achieved by employment of a split-flow interface with a make-up pump. In addition to preventing precipitation the make-up pump is used to aid ionization when chromatography operates under about 5% modifier [7].

Material and methods

A bioanalytical method for the quantification of retinol and α-tocopherol in serum and plasma samples using SFC-MS/MS in positive electrospray ionization mode (ESI+) was developed and validated in accordance with local guidelines, based on the U.S. Food and Drug Administration (FDA) Guidance for Industry Bioanalytical Method Validation.

Samples were prepared for analysis using a MultiPurpose Sampler (MPS; Gerstel). The MPS enables automated and highly reproducible internal standard addition, protein precipitation capability, automated centrifugation and supernatant handling. This approach enables a high number of samples to be processed. Moreover, each sample extract is treated in exactly the same way and prepared just before the analysis, improving sample-to-sample reproducibility when compared to a manual batch process. Sequential automated sample preparation was as follows: 100 µl of serum or plasma is subjected to protein precipitation using 1% acetic acid in acetonitrile containing deuterated retinol-d8 and deuterated α-tocopherol-d6 as internal standards; this is followed by extraction of the lipid phase into heptane. Each sample is subsequently vortexed and centrifuged for 30 seconds to assist protein precipitation and separation. Samples are transferred to the UPC2 sample manager and 1 µl of extracts are further separated in an ACQUITY UPC2 HSS C18 column maintained at 35°C. A flow rate of 1.5 ml/min is applied to the column, using a gradient elution of supercritical carbon dioxide and methanol (2–10% methanol), and 1% formic acid (Table 1) in methanol solution for the make-up pump allowing for efficient ionization. Detection and quantification is carried out by operating the mass spectrometer in ESI+ multiple-reaction-monitoring (MRM) mode on a Xevo TQ-S micro (Waters; Table 2).


Chromatographic baseline separation of retinol and α-tocopherol was achieved in 2.5 minutes, with a total injection-to-injection cycle time of 4 minutes. This compares favourably to the 15-minute chromatographic run time in the HPLC-UV assay used for the method comparison and to that of LC-MS/MS methods [8]. Representative chromatograms are shown in Figure 4. The method proved to be linear in the calibration range for retinol up to 14 µmol/L and for α -tocopherol up to 111 µmol/L, with the mean linear correlation coefficients r2 >0.997 for both calibration curves. The lower limits of quantification (LLOQ) for retinol and α-tocopherol were 0.09 µmol/L and 0.78 µmol/L respectively. The LLOQ was defined with a coefficient of variation (CV) of <20% while maintaining an accuracy of 80–120%. Intraand inter-assay imprecision in all quality control materials tested yielded CV values ≤8% for both vitamins. Desirable specification for imprecision for retinol and α-tocopherol is 6.8%, and 6.9% respectively [7]. No carry-over was observed. Selectivity and dilution integrity were validated using spiked recovery experiments that yielded mean recoveries of 104.5% (range, 93.7–116.9%) for retinol and 110.0% (range, 101.8–116.2%) for α-tocopherol. Method comparison between HPLC-UV and SFC-MS/MS demonstrated an excellent agreement. Bland–Altman analysis demonstrated mean negative bias of 0.15 µmol/L (95% confidence interval (CI) −0.90 to 1.19) and 0.2 µmol/L (95% CI −12.4 to 12.7) for retinol and α-tocopherol, respectively, for the SFC-MS/MS method. Performance of the method has been assessed by United Kingdom National External Quality Assurance Scheme (UK NEQAS) for the past three years. Reports show consistent good accuracy scores for retinol and α-tocopherol when compared to the all laboratory trimmed mean (ALTM). Analysis of UK NEQAS samples yielded mean bias from the target value of −1.9% for retinol and −3.45% for α-tocopherol (Fig. 5).


SFC-MS/MS provides improved specificity and significantly faster retinol and α-tocopherol analysis when compared with HPLC-based methodologies. The Nutristasis laboratory have used this methodology for the analysis of >16 000 samples since 2019. The combination of optimized SFC separation coupled to MS detection provides a reliable method for the analysis that has proven to be fast, easy and robust compared to HPLC-based methods. The method shows enhanced safety owing to reduced use of harmful solvents, as well as a reduction in cost compared with previously published metho-dologies. The full procedure has been applied in our diagnostic laboratory increasing sample throughput and reducing solvent consumption. The method demonstrates a good agreement with UK NEQAS samples. In addition, the use of the MPS for pre-analytical sample preparation further reduces time and labour.

Highlights of the SFC-MS/MS method are as follows. • SFC-MS/MS provides improved specificity and significantly faster analysis compared with HPLC-based methodologies. • EQA assessment reports the analysis to be performing consistently well. • The MPS reduces sample preparation time and complexity, limiting the potential for human errors and freeing up time of laboratory staff. • Other benefits of this method include reduced use of solvents; approximately 85% less, using less than 350 µl of methanol mobile phase per sample.

Figure 1. Structure of retinol: unsaturated long chain alcohol attached to a β-ionone ring

Figure 2. Structure of α-tocopherol: chromanol ring with a saturated side chain

Figure 3. Components and flow path of the SFC (UPC2)-MS/MS system
ABPR, automatic back-pressure regulator.

Figure 4. Typical chromatograms of retinol and α-tocopherol after SFC-MS/MS analysis
The patient plasma sample contained 2.23 µmol/L retinol and 26.62 µmol/L α-tocopherol, which eluted at retention times of 1.62 min and 1.48 min, respectively.

Figure 5. SFC-MS/MS performance of plasma retinol and α-tocopherol analysis assessed by UK NEQAS
Diagrams compare consecutive SFC-MS/MS results of EQA samples to the all laboratory trimmed mean (ALTM). Diamonds represent ALTM, bars represent the +/− 2 standard deviation range from the ALTM, squares represent the SFC-MS/MS results.

Table 1. SFC (UPC2)-MS/MS method summary

Table 2. Mass spectrometer detector settings for quantifier, qualifier and internal standards MRM transitions

The authors

Ella Freke* BSc, Renata Gorska MSc and Dominic Harrington PhD

The Nutristasis Laboratory, Viapath, St Thomas’ Hospital,
London SE1 7EH, UK

*Corresponding author

E-mail: Ella.Freke@viapath.co.uk


  1. Harrington. D. Laboratory assessment of vitamin status, 1st Edition. Elsevier 2018.
  2. Zabetian-Targhi F, Mahmoudi MJ, Rezaei N, Mahmoudi M. Retinol binding protein 4 in relation to diet, inflammation, immunity, and cardiovascular diseases. Adv Nutr 2015;6(6):748–762 (https://academic.oup.com/advances/article/6/6/748/4555140?login=false).
  3. Tanumihardjo SA, Russell RM, Stephensen CB et al. Biomarkers of Nutrition for Development (BOND)-Vitamin A Review. J Nutr 2016;146(9):1816S–1848S (https://academic.oup.com/jn/article/146/9/1816S/4584876?login=false).
  4. Rizvi S, Raza ST, Ahmed F et al. The role of vitamin e in human health and some diseases. Sultan Qaboos Univ Med J 2014;14(2):e157–e165 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3997530/).
  5. Shamim AA, Schulze K, Merrill RD et al. First-trimester plasma tocopherols are associated with risk of miscarriage in rural Bangladesh. Am J Clin Nutr 2015;101(2):294–301 (https://academic.oup.com/ajcn/article/101/2/294/4494385?login=false).
  6. Greaves RF, Woollard GA, Hoad KE et al. Laboratory medicine best practice guideline: vitamins A, E and the carotenoids in blood. Clin Biochem Rev 2014;35(2):81–113 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4159783/).
  7. Tarafder A. Designs and methods for interfacing SFC with MS. J Chromatogr B Analyt Technol Biomed Life Sci 2018;1091:1–13.
  8. Hinchliffe E, Rudge J, Reed P. A novel high-throughput method for supported liquid extraction of retinol and alpha-tocopherol from human serum and simultaneous quantitation by liquid chromatography tandem mass spectrometry. Ann Clin Biochem 2016;53(4):434–445 (https://journals.sagepub.com/doi/10.1177/0004563215596024?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed).