Enhanced high-performance liquid chromatography method for the determination of retinoic acid in plasma. Development, optimization and validation

Highlights

• Full validation for an endogenous compound.

• Holistic approach that included not only the classical parameters.

• Good results are achieved in less time than other HPLC methods for these compounds.

• Excellent analytical figures of merit.

Abstract

When determining endogenous compounds in biological samples, the lack of blank or analyte-free matrix samples involves the use of alternative strategies for calibration and quantitation. This article deals with the development, optimization and validation of a high performance liquid chromatography method for the determination of retinoic acid in plasma, obtaining at the same time information about its isomers, taking into account the basal concentration of these endobiotica. An experimental design was used for the optimization of three variables: mobile phase composition, flow rate and column temperature through a central composite design. Four responses were selected for optimization purposes (area under the peaks, quantity of peaks, analysis time and resolution between the first principal peak and the following one). The optimum conditions resulted in a mobile phase consisting of methanol 83.4% (v/v), acetonitrile 0.6% (v/v) and acid aqueous solution 16.0% (v/v); flow rate of 0.68 mL min⁻¹ and an column temperature of 37.10 °C. Detection was performed at 350 nm by a diode array detector. The method was validated following a holistic approach that included not only the classical parameters related to method performance but also the robustness and the expected proportion of acceptable results lying inside predefined acceptability intervals, i.e., the uncertainty of measurements. The method validation results indicated a high selectivity and good precision characteristics that were studied at four concentration levels, with RSD less than 5.0% for retinoic acid (less than 7.5% for the LOQ concentration level), in intra and inter-assay precision studies. Linearity was proved for a range from 0.00489 to 15.109 ng mL⁻¹ of retinoic acid and the recovery, which was studied at four different fortification levels in phuman plasma samples, varied from 99.5% to 106.5% for retinoic acid. The applicability of the method was demonstrated by determining retinoic acid and obtaining information about its isomers in human and frog plasma samples from different origins.

Introduction

The term ‘retinoid’ refers to a group of endogenous and synthetic molecules structurally similar to retinol (ROH), which is the principal form of Vitamin A, transported in the blood of most vertebrate organisms. ROH is metabolized to retinoic acid (RA) and related geometric isomers which have been detected in blood and embryonic target cells of mammals and birds [1]. These highly potent signaling molecules bind nuclear receptors evoking pleiotropic effects observed at the tissue-organism level. Retinoids have been extensively studied in birds and mammals where their imbalances are associated with multiple dysfunctions including various dermal lesions, immunosuppression, susceptibility to disease (including cancer and parasitic infections), changes in secondary sexual characteristics, inhibition of spermatogenesis, decreased embryo survival, deformities, embryonic development, and numerous other effects on reproduction [1]. There are a number of endogenous geometric isomers of RA (each of them with unique function) [2]. Many tissues and plasma have been reported containing all-trans-RA; 9,13-cis-RA; 13-cis-RA; 9-cis-RA and 11-cis-retinoids [3].

In the pharmaceutical field, both RA and 13-cis-RA are widely used in the treatment of various dermatological diseases such as acne, psoriasis, skin cancer and photoaging, regulating growth and differentiation of epithelial cell, sebum production, and collagen synthesis [2], [4].

Because RA isomers are isobaric and have overlapping ultraviolet (UV) spectral profiles, mass detection and/or single wavelength UV detection cannot distinguish the identities of geometric isomers that co-elute. Therefore, analysis of RA requires the chromatographic separation of endogenous isomers before detection [3]. A literature search reveals that techniques like gas chromatography (GC), high performance liquid chromatography (HPLC) column-switching with and without direct injection of plasma, ultra high performance liquid chromatography (UHPLC) and capillary electrophoresis (CE), using ultraviolet, fluorescence and mass spectrometer detectors, have been mostly used in developing methods for determining RA and its isomers [5], [6], [7], [8], [9]. Different methods have been presented to determine these compounds, in most of them the analytes being separated with retention times close to 30 min by using C18 columns (4.6 mm × 250 mm, 5.0 μm particle size) [6], [7], [8], [9].

Measurement of drug concentrations in biological matrices (such as serum, plasma, blood, urine, and saliva) and in pharmaceuticals is an important aspect of medicinal product development. Such data may be required to support applications for new actives substances and generics as well as variations to authorized drug products. The results of animal toxicokinetic studies and of clinical trials, including bioequivalence studies are used to make critical decisions supporting the safety and efficacy of a medicinal drug substance or product.

It is therefore paramount that the applied bioanalytical methods used are well characterized and fully validated in order to yield reliable results. Acceptance criteria wider than those defined in Guideline on Bioanalytical Method Validation of European Medicines Agency may be used in special situations, which should be prospectively defined based on the intended use of the method [10].

In this work a novel HPLC-UV method was developed, optimized and fully validated for its application in the simultaneous determination of RA and their isomers in plasma. For carrying out the objectives, three variables of the chromatographic system were studied through a central composite design to optimize four responses simultaneously in order to obtain the optimum parameters to decrease the retention time, the solvent expense and the cost of the analysis. Variables and design selection, as well as models fitting and optimization criteria to reach the global desirability are discussed.

Validation of bioanalytical methods is straightforward when analyte-free matrix and well-characterized reference standard of the analyte are available. However, the quantitative determination of endogenous (i.e. naturally occurring) compounds is more complicated because the lack of analyte-free samples of the authentic biological matrix or samples with accurately-know analyte concentrations. Therefore, the preparation of reference samples has to be addressed in a different way and, as a consequence, validation also becomes less straightforward [11]. In these cases, quantitation can be carried out by two strategies: the use of surrogate analyte in the authentic matrix or the use of authentic analyte in a surrogate matrix [11].

On the other hand, to the date, there are not official guidelines dealing with the validation of chromatographic methods for endogenous analytes and, usually, the ones existing for pharmaceuticals and xenobiotic compounds have been adapted for endogenous compounds. Thus, some authors applied method validation principles for drug assays, in particular those issued by the US FDA [12]. Whereas most authors define analytical figures of merit in the same way for xenobiotic and for endogenous compounds, Tsikas [13] pointed out that these parameters should be determined differently in both samples, as the basal concentration of endogenous (C_0,Ln) varies among biological samples, and defined the relative lower limit of quantitation (rLLOQ), which is corrected by the C_0,Ln. In addition, Schmidt et al. [14] proposed to subtract the peak areas from the corresponding unspiked blank sample to the peak area of each spiked sample, thus avoiding errors related to the calculation of C_0,Ln.

In the present work, a chromatographic method has been developed to determine the endogenous compound (RA) and obtaining information about its isomers in human and frog plasma samples in less than twelve min. Validation has been carried out using a holistic approach which considers the most relevant procedures for checking the quality parameters, as well as the estimation of robustness and measurement uncertainty. As in this case makes no sense the use of surrogate analyte in the authentic matrix because the DAD detector cannot discriminate it, the alternative approach would be the use of authentic analyte in a surrogate matrix. However, we think that to subtract the basal concentration of analyte from its signal is a more realistic option and thus, the signal corresponding to the unspiked sample was subtracted when it was suitable.