Introduction

In recent years, the roles of vitamins in human health have received extensive attention, including in patients with Hashimoto’s thyroiditis (HT). Many patients with HT have nutritional deficiencies, excess body weight, metabolic disorders, and reports problems in daily life [1]. Most of the studies to date have focused on the role of vitamin D in this population [2,3,4]. Water-soluble vitamins and their derivatives, including vitamin B2 and vitamin B6, are essential cofactors and coenzymes in many biochemical reactions, being involved in energy metabolism. Their deficiencies negatively affect the metabolism of glucose, amino acids and fatty acids [5]. Some studies have shown that vitamin B2, B12, B9 and D shortages impact the thyroid gland function [6]. However, there are no studies on the concentration of vitamin B2 and B6 in patients with HT.

Riboflavin (vitamin B2) is a water-soluble vitamin necessary for metabolism, especially lipid metabolism and energy production [7,8,9,10,11]. As a crucial enzymatic cofactor, riboflavin takes part in oxidation–reduction reactions and immune homeostasis [5, 10,11,12,13,14,15]. Vitamin B2 is converted into two active forms, flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN), which are cofactors in redox reactions and fatty acid oxidation. Moreover, this vitamin is involved in T-cell activation, differentiation and proliferation [10, 11, 14]. Riboflavin also participates in the metabolism of homocysteine, which high level is associated with a risk of developing cardiovascular diseases (CVD). Homocysteine can be metabolised in two ways. First, by transsulfuration, depending on vitamin B6 (pyridoxine), and second by remethylation, depending on vitamin B2, B9 and B12 [10].

Mazur-Bialy and Pocheć [16] showed that a deficiency of vitamin B2 increases the pro-inflammatory activity of adipocyte cells, which intensifies the severity of chronic inflammation associated with obesity, insulin resistance and metabolic syndrome. Studies with mouse adipocytes demonstrated that vitamin B2 deficiency resulted in decreased levels of interleukin-1 (IL-1), tumor necrosis factor-α (TNF-α), reactive oxygen species (ROS) and increased levels of anti-inflammatory adiponectin and interleukin-10 (IL-10) [16]. The clinical trial “the RISE-UP study” demonstrated that after riboflavin supplementation for 3 weeks in the daily dose of 100 mg in 70 patients with Crohn’s disease, the levels of inflammatory markers significantly decreased. Moreover, oxidative stress and clinical disease activity decreased. These data indicated that vitamin B2 has anti-inflammatory and anti-oxidant properties [10, 11, 17]. Moreover, FAD coenzyme is necessary for the activity of glutathione reductase, which converts oxidised glutathione to the reduced form, which is an endogenous antioxidant and deactivates ROS. Vitamin B2 may also affect the activity of antioxidant enzymes, including glutathione peroxidase (GPx), superoxide dismutase (SOD) and catalase [11]. The sources of vitamin B2 include milk and other dairy products, meats, fish, eggs, legumes, cereals, grain products, bread, nuts, mushrooms and green leafy vegetables [7, 8, 13, 15].

Pyridoxine (vitamin B6) is a coenzyme in over 150 biochemical reactions involved in the metabolism of carbohydrates, lipids, amino acids and neurotransmitters. The processes in which vitamin B6 participates are, among others, glycogenolysis, transamination, decarboxylation, metabolism of sphingosine phosphate and catabolism of tryptophan. As mentioned previously, vitamin B6 is a cofactor of homocysteine metabolism; therefore, its deficiency may result in an increased level of homocysteine and, consequently, the development of atherosclerosis. This vitamin exhibits antioxidant properties and has the ability to lower the level of advanced glycation end products (AGE). Vitamin B6 is also involved in the metabolism of folate [18]. The biologically active form of vitamin B6 is pyridoxal-5′-phosphate (PLP). Therefore, PLP is a useful biomarker for evaluating vitamin B6 status [5, 18, 19].

Studies indicated that vitamin B6 deficiency is associated with chronic diseases, including CVD, like high-risk of atherosclerosis, stroke, and thrombosis [18, 20]. A low concentration of PLP is also related to hyperhomocysteinemia [21]. Furthermore, this vitamin protects against diabetes by reducing inflammation, inflammasomes and oxidative stress [22].

Kumrungsee et al. [23] indicated that vitamin B6 helps to treat severe of COVID-19, possibly through reducing complications of chronic diseases such as CVD, hypertension, and diabetes. Vitamin B6 also reduces IL-1β and ROS production [24]. Framingham Offspring Study by Sakakeeny et al. [25] with 2229 adults showed that the lowest levels of PLP were in people who had chronic inflammation, whereas those with high levels had the lowest inflammation scores. There was also an inverse relationship between plasma PLP levels and inflammatory/immune-related conditions [25]. Some studies demonstrated that vitamin B6 deficiency has an impact on the immune system [20, 26, 27]. Qian et al. indicated that deficiency of this vitamin inhibits the proliferation of lymphocytes, interferes with its differentiation, and influences their maturation [27, 28]. Gombart et al. [27] showed that vitamin B6 plays an important role in maintaining the Th1- and NK cell-mediated immune response and additionally enhancing the cytotoxic activity of the latter. Grains, nuts, seeds, potatoes, meat and meat products (e.g. liver) and fish are the main sources of vitamin B6 [7, 29].

Various analytical methods were used for the analysis of vitamins, including chromatographic, electrochemical, microbiological, immunological and others [7]. Many of these methods use aggressive reagents, require long execution times and multiple stages of sample processing. However, during the last decades, high‐performance liquid chromatography (LC) has become the most common methods for the determination of vitamins in food matrices, drinks and supplements. Only a relatively small number of experimental studies focused on biological fluids, including serum and plasma. Among these methods, LC coupled with MS/MS is the main analytical technique used to identify vitamins because of its high selectivity and sensitivity. This method also simplifies the sample preparation [7, 30,31,32,33,34,35]. There are many methods for measuring the concentration of fat-soluble vitamins, mainly vitamin D, in serum or plasma, but much less for water-soluble vitamins.

The aim of this study was to develop and validate a novel, accurate, rapid, reliable and efficient LC–MS/MS method for the simultaneous determination of vitamin B2 and vitamin B6 in human serum. The method was applied for the analysis of these compounds in clinical samples obtained from patients with Hashimoto’s thyroiditis.

Materials and Methods

Chemicals and Reagents

Riboflavin (vitamin B2), pyridoxal 5′-phosphate (vitamin B6), riboflavin-(dioxopyrimidine-13C4,15N2) and formic acid were purchased from Sigma-Aldrich (Steinheim, Germany). Trichloroacetic acid (TCA) and hydrochloric acid (HCl) were obtained from Merck (Darmstadt, Germany). Methanol and water of LC–MS grade were used for a mobile phase and purchased from Merck (Darmstadt, Germany). The same water of LC–MS grade and TCA were used for the preparation of samples and solutions. The human serum of healthy volunteers was obtained from the Regional Centre of Blood Donation (Poznan, Poland) and was aliquoted and stored at  − 80 °C.

LC–MS/MS Chromatographic Conditions

The study was performed using a Shimadzu UPLC Nexera set (Shimadzu, Kyoto, Japan) equipped with a degasser (DGU-20A5) and thermostated autosampler (SIL-30AC). UPLC was coupled with a triple-quadrupole tandem mass spectrometer detector (LC–MS-8030). Data processing was performed using LabSolutions Series Workstation software (Shimadzu, Kyoto, Japan). FCV-20AH2 unit (Shimadzu, Kyoto, Japan) was used to divert the flow to waste during the first 5 min and between each injection.

The separation of two analytes, vitamin B2 and vitamin B6, in addition to the internal standard (IS) riboflavin-(dioxopyrimidine-13C4,15N2) was performed in ZORBAX SB-C8 5 μm analytical column (4.6 mm × 150 mm) connected to a security guard cartridge (both from Agilent Technologies, Santa Clara, California, USA). The temperature of the analytical column was kept constant at 25 °C by column oven model CTO-2AC (Shimadzu, Kyoto, Japan). The mobile phase consisted of methanol–water (50:50, v/v) both containing 0.1% (v/v) formic acid and the flow rate was set to 0.3 mL/min. Electrospray ionization in the positive electrospray ionization mode (ESI +) was used for eluent introduction from LC column straight to the MS interface. The electrospray needle voltage was maintained at 4.5 kV. The MS interface has been adapted with the following parameters: DL temperature, 220 °C, heat block temperature 400 °C, interface temperature 350 °C, nitrogen as the drying gas and as the nebulizing gas with flow rates of 10 and 2 L/min, respectively. To minimize the carry-over effect, a mixture of methanol–water (50:50, v/v) was used to rinse the needle of the autosampler before and after aspiration of the sample. MS conditions were examined using automated optimization of LabSolution software.

The multiple reaction monitoring (MRM) mode was used for observing the analyte’s specific transitions. The primary mass transitions (m/z) chosen for quantification of the analytes were 377.2 → 243.0 for vitamin B2, 247.9 → 150.0 for vitamin B6 and 383.0 → 249.0 for IS. The primary mass transitions were selected according to the highest intensity, which allowed the greatest sensitivity and the highest specificity. The secondary mass transitions for qualitative confirmation were 377.2 → 172.0 for vitamin B2, 247.9 → 94.0 for vitamin B6 and 383.0 → 201.9 for IS.

Standard Solutions

Stock solutions of vitamin B2 and vitamin B6 (as PLP) at a concentration of 100 μg/mL were prepared in 0.1% TCA and 0.1 mol/L HCl, respectively. The stock solution of IS which was riboflavin-(dioxopyrimidine-13C4,15N2) was prepared at a concentration of 10 µg/mL in 5% TCA. All stock solutions were aliquoted and stored at  − 80 °C in darkness until needed. The working standard solutions of analytes were prepared fresh daily from stock solutions in ultrapure water at the following concentrations: 25, 50, 100, 200, 400 and 600 ng/mL of vitamin B2 and 25, 50, 200, 400, 800 and 1000 ng/mL of vitamin B6. All solutions were refrigerated in dark tubes to protect vitamins from light-induced oxidation and used immediately after preparation.

The calibration standards were freshly prepared from the working standard solutions for each validation and assay run. Calibration standards were prepared by spiking 125 μL of pooled serum of healthy volunteers with 12.5 μL of the working standard solutions to give nominal concentrations of 2.5, 5, 10, 20, 40 and 60 ng/mL for vitamin B2 and 2.5, 5, 20, 40, 80 and 100 ng/mL for vitamin B6. The IS concentration amounted was 10 ng/mL.

Sample Preparation for Method Validation and Patients

For method validation, 125 µL of serum samples and 12.5 µL aliquots of the analytes standard solution were added to 1.5 mL microfuge tubes. Patients’ samples were prepared by adding 12.5 µL of ultrapure water to 125 µL of patients’ serum. A volume of 12.5 µL of the IS solution was added to the aforementioned samples. Protein precipitation was performed by adding 700 µL 5% aqueous solution of TCA. Next, the mixture was vortexed for 2 min. After vortexing, the samples were left at room temperature in the dark for 30 min after which they were shaken vigorously for 2 min. Following a 1 h incubation at room temperature, the samples were vortexed-mixed and centrifuged at 16,000 × g for 10 min to precipitate proteins. The supernatants were transferred to new glass vials and placed in the autosampler of the LC–MS/MS system. A 20 μL aliquot of the supernatant was directly injected for analysis. Samples with analytes at all stages of preparation were protected from light. For each subject sample, the analysis was performed in duplicate.

Method Validation

The method was validated in accordance with the European Medicines Agency (EMA) guidelines [36]. Method validation involved determinations of the linearity over the calibration range, the sensitivity (lower limits of quantification, LLOQ), the precision (intra‐ and inter‐day), the accuracy (intra‐ and inter‐day), the recovery and the stability.

The chromatographic conditions were optimized for peak resolution and retention time. The mobile phase composition was evaluated using methanol, acetonitrile, and water as solvents, as well as ammonium hydroxide, formic acid, and acetic acid as additives. After several tests, we obtained chromatographic conditions that gave adequate efficiency with respect to the standards of quality and sensitivity.

Linearity

Calibration curves were prepared in the range of 2.5–60 ng/mL for vitamin B2 and 2.5–100 ng/mL for vitamin B6 in pooled serum of healthy volunteers. The linearity of the method was evaluated at six concentration levels for each compound and estimated for the ratio of the difference between analyte peak areas from the spiked and nonspiked (endogenous) serum samples to that of the IS, as a function of the analyte concentration added to the sample. The retention time of vitamin B2 was about 8.5 min and vitamin B6 was 7 min.

The resulting peak areas were processed and then calibration curves were created in Microsoft Excel. The correlation coefficient (r) was evaluated. To check the significance of the intercept Student’s t test was performed. The equations of calibration curves were used to calculate the concentration of analytes in patients’ serum.

Precision, Accuracy and LLOQ

The method intra- and inter-day precision have been expressed as the relative standard deviation (RSD) and was estimated for quality control samples (QCS) spiked with low (5 ng/mL vitamin B2 and 5 ng/mL vitamin B6), medium (20 ng/mL vitamin B2 and 40 ng/mL vitamin B6) and high (60 ng/mL vitamin B2 and 100 ng/mL vitamin B6) concentrations, prepared in five replicates in serum and analyzed during the same day (intra-day precision) and over 5 consecutive days (inter-day precision). Intra-day and inter-day precision expressed by the RSD and was calculated as (SD/Cmeasured)·100%.

Accuracy was estimated over the same analyte concentration ranges as for precision of the method. The accuracy was estimated as ((Cmeasured − Cendogenous)/Cadded) × 100%.

The determination of Limit of Detection (LOD) and LLOQ was based on the known concentration of each metabolite and observing the peak signal with the corresponding signal to noise ratio (S/N). LOD was defined as the smallest concentration of the analytes that could be detected with the corresponding S/N greater than 3:1. LLOQ was determined as the lowest concentration of the analytes determined by the method with the precision  ≤ 20% and accuracy in the range of 80–120%.

Recovery and Matrix Effect

Recovery of vitamin B2 and vitamin B6 from the serum samples was evaluated by analyzing five replicates of low (5 ng/mL for vitamin B2 and 5 ng/mL for vitamin B6) and high (60 ng/mL for vitamin B2 and 100 ng/mL for vitamin B6) concentrations for each analyte. Recovery was studied by dividing a pooled serum into two portions. Series I consisted of spiked serum with the analytes and internal standards. Samples of series II were supplemented only with internal standards. Then the sample preparation of Series I and II have been performed in accordance with sample preparation for method validation. To series II, analytes were added at the end of sample preparation. The recovery of the analytes was calculated as (CI-series)/(CII-series)·100%.

The effect of serum components on ionization of vitamin B2, vitamin B6 and IS was evaluated using six individual serum samples spiked with QCS at low and high concentrations of the analytes. Matrix factor (MF) was calculated by dividing the peak measured in a blank matrix spiked with low (5 ng/mL of vitamin B2 and 5 ng/mL of vitamin B6) and high (60 ng/mL of vitamin B2 and 100 ng/mL of vitamin B6) concentrations after protein precipitation by the peak area of the analytes at equivalent concentrations in the absence of the matrix. Moreover, the IS normalized MF was calculated by dividing the analyte’s MF by the IS’s MF. The RSD of the IS-normalized MF should not be greater than 15%.

Stability

The stability of vitamin B2 and vitamin B6 in serum samples was evaluated at two concentrations, low and high levels (5 and 60 ng/mL of vitamin B2 and 5 and 100 ng/mL of vitamin B6) in three replications for each concentration. Stability was checked after three freeze–thaw cycles at  − 80 °C, long-term (1 month at  − 80 °C) and after storage of the prepared samples in autosampler for 24 h at 10 °C. Moreover, the short‐term stability of the analytes was analyzed (25 °C for 3 h). Based on the EMA guidelines for bioanalytical method validation, the stability of the analytes is confirmed if the deviation from the nominal concentration is within  ± 15% [36].

In Vivo Application

The validated LC–MS/MS method has been applied for the quantitative determination of vitamin B2 and vitamin B6 in the serum of patients with Hashimoto’s thyroiditis recruited from the Department of Endocrinology, Metabolism and Internal Medicine at Poznan University of Medical Sciences. The study included 107 Caucasian females. Seventy-four female patients with HT (study group) and thirty-three female healthy controls (control group) were involved in the study.

The inclusion criteria for the study group were based on biochemical evaluation (presence of circulating thyroid autoantibodies) and ultrasound assessment (inhomogeneous hypoechogenic structure of the thyroid gland) [37]. In the study, some patients were euthyroid, others hypothyroid, and some were treated with levothyroxine. Inclusion criteria for both groups were age from 18 to 65 and lack of vitamin B2 and B6 supplementation. The exclusion criteria were other thyroid diseases, a positive history of cancer, pregnancy and breastfeeding. The control group comprised healthy females without HT [TSH 0.27–2.5 mIU/L, no autoantibodies to thyroperoxidase (TPOAb) and thyroglobulin (TgAb), no levothyroxine (LT4) use]. Moreover, women who had poor health according to physical examination and laboratory analyses were excluded from the control group.

Each participant in the study gave written informed consent. The study has been conducted in alignment with outlined ethical principles in the Declaration of Helsinki [38]. The conducted study was approved by the Ethical Committee at Poznan University of Medical Sciences (no. 873/19 and 201/21).

The blood sample was collected under standard conditions after overnight fasting. Serum was separated by centrifugation and an aliquot was stored at  − 80 °C until the LC–MS/MS analysis. The serum samples were also subjected to biochemical analysis immediately after collection for the thyroid and lipid profile. Serum thyroid stimulating hormone (TSH), free tetraiodothyronine (fT4), free triiodothyronine (fT3) and TPOAb and TgAb concentrations were determined by commercial kits using electrochemiluminescence (ECLIA) by Hitachi and Roche Diagnostics on a Cobas e601 analyzer (Indianapolis, IN, USA). Lipid profile, such as total cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL) and triglycerides (TG) were measured using the enzymatic method with standardized commercial tests performed at the central laboratory of the University Clinical Hospital in Poznan.

The anthropometric measurements were taken after overnight fasting, and participants were dressed only in underwear. Body weight and height were analysed with accuracy to 0.1 kg (using a certified weight) and to 0.1 cm (stadiometer). The obtained results allowed for the calculation of the body mass index (BMI).

Data Processing and Statistical Analysis

Statistical analysis of the results was performed using Statistica 13 program with the Medical (StatSoft, Tulsa, OK, USA). The normality of the data distribution was assessed using the Shapiro–Wilk test [39]. The obtained results are presented in the article as means ± standard deviations (SD). The results were analyzed statistically, using elements of descriptive statistics and statistical procedures, such as correlation analysis (Pearson’s for parametric distributions and Spearman test for non-normal distributions). Comparisons between the study and the control group were made using a Student’s t test for normal data distribution and a Mann–Whitney U test for non-normal data distribution. The level of statistical significance was taken as p < 0.05.

Results and Discussion

LC–MS/MS Analysis

The MS conditions were optimized for obtaining a good signal and high sensitivity for vitamin B2, vitamin B6 and IS in serum samples. The highest intensity transitions m/z 377.2 → 243.0 for vitamin B2, 247.9 → 150.0 for vitamin B6 and 383.0 → 249.0 for IS were chosen for quantification of the analytes with the highest sensitivity. The selection of a column and mobile phase components was optimized to obtain separation of all the analytes in a total time of analysis of 11 min. Some methods for the determination of vitamin B2 and vitamin B6 have a shorter retention time, among others Geng et al. (6 min) [32] or Zhang et al. (5 min) [40]. Analysis time in the method developed by Kushnir et al. [41] and Ghassabian et al. [42] was also shorter, but they analyzed only PLP (3 min and 5 min, respectively). However, most of them are longer than the described method, among others Khaksari et al. which was 16 min [31, 43] or Cabo et al. with 13 min run time [44].

The representative chromatograms of standard and clinical samples are displayed in Fig. 1.

Fig. 1
figure 1

Representative chromatograms of the studied compounds: (a) calibration standard at the LLOQ level; (b) calibration standard at concentration 10 ng/mL of vitamin B2 and 20 ng/mL of vitamin B6; (c) serum sample of HT patient (determined concentrations—6.85 ng/mL of vitamin B2 and 14.7 ng/mL of vitamin B6)

Full size image

Method Validation

Standard curves estimated for the analytes were linear in the ranges of concentrations 2.5–60 ng/mL for vitamin B2 and 2.5–100 ng/mL for vitamin B6. Similar ranges of concentrations were obtained by Geng et al. [32] (5–40 and 1–80 ng/mL for vitamin B2 and B6, respectively) and M. Zhang et al. [40] (0.2–40 ng/mL for riboflavin and 0.5–100 ng/mL for vitamin B6). In LC–MS/MS method by Diniz et al. [45] vitamin B2 was linear in the range from 0.5 to 50.0 ng/mL.

Calibration curves were tested using six standard concentrations of two water-soluble vitamins and they were run in triplicate. The method was linear across the whole range of concentration. The correlation coefficient r values were  ≥ 0.991 indicating the method’s decent linearity for all compounds.

In the worked out conditions, the LLOQs were 2.5 ng/mL for vitamin B2 and 2.5 ng/mL for vitamin B6. The value of LLOQ for vitamin B2 proved to be lower than the respective value of 5 ng/mL reported by Geng et al. [32], but higher for vitamin B6 which was 1 ng/mL. However, the method required 500 μL of serum sample, which is much more than in our method, where only 125 µL of serum was used. Lower LLOQs were obtained by Kahoun et al. [35] for vitamin B2 (0.27 ng/mL) and vitamin B6 (0.91 ng/mL) using larger aliquots of whole blood (500 μL) and analysis time (25 min). M. Zhang et al. [40] demonstrated slightly lower LLOQ of 0.2 ng/mL for vitamin B2 and 0.5 ng/mL for vitamin B6 in LC–MS/MS assay using dried blood spots. Moreover, Diniz et al. [45] validated LC–MS/MS method for the determination of vitamin B2 in human plasma and LLOQ was 0.5 ng/mL.

Precision and accuracy were calculated at four concentration levels (LLOQ, low, medium and high QCS). The validation results of the inter-day and intra-day accuracy and precision for each metabolite are shown in Table 1. The intra- and inter-day precisions of the method expressed as RSD fitted the range required for testing analyte content in body fluids and were  < 11% for all quality control standards of the compounds mentioned above. The accuracy obtained in our study was in the range 89.1–107.16%. Similar values of precision and accuracy were reported by Geng et al. [32] in the LC–MS/MS method for the determination of vitamins B1, B2, B6 and B9 in human serum with the precision of less than 10% and accuracy in the range of 90–110%. Other results were obtained by Armah et al. [46] where the precision was ≤ 20% for vitamin B6 and very high for vitamin B2 (60%). The same results for vitamin B2 as in our study were obtained by Khaksari et al. [31] and Diniz et al. [45]. The intra- and inter-day precision for vitamin B2 was  < 10% and  < 11%, respectively. Better precision for vitamin B6 than in described study was obtained by Roelofsen-de Bee et al. [47] in LC–ESI–MS/MS method. The mean intra- and inter-day precision for PLP were 3.4% and 6.1%, respectively.

Table 1 Intra- and inter-day precision and accuracy for vitamin B2 and vitamin B6
Full size table

Recoveries of the analytes from serum samples obtained during our study were 96.79% (low concentration) and 99.18% (high concentration) for vitamin B2 and 92.21% (low concentration) and 109.79% (high concentration) for vitamin B6 (Table 2).

Table 2 Recovery of vitamin B2 and B6 reported in described study and other methods
Full size table

The values of recoveries of vitamin B2 and B6 obtained in our study are similar to the literature data demonstrated by M. Zhang et al. [40], Ghassabian et al. [42], Khaksari et al. (2017a), Cabo et al. [44], Diniz et al. [45] Roelofsen-de Bee [47] and Footitt et al. [48]. The recoveries of the analytes in described method and other methods are presented in Table 2. The current recoveries values were also higher than those previously reported by Geng et al. [32] and Armah et al. [46].

As shown in Table 3, matrix components did not significantly suppress or enhance the MS signal of vitamin B2, as proved by the IS-normalized MF ranging from 0.99 to 1.02. However, for vitamin B6, a significant matrix effect was observed, indicating the suppression of the vitamin signal by serum components. For both vitamins, RSD of the IS-normalized MF was below 15% which is consistent with EMA recommendations.

Table 3 Matrix effect expressed by IS-normalized MF
Full size table

Stability results showed that vitamin B2 and vitamin B6 were stable in serum samples for at least 1 month at  − 80 °C and after three freeze–thaw cycles. The stabilities were confirmed by the RE which estimated ranges of 5.11–13.84% and 3.80–14.48%, respectively. The stability test results for the analytes while stored in the autosampler for 24 h were within the RE range of 1.58–4.23%. Vitamin B2 and B6 also proved to be stable in serum samples in a short-term stability test (25 °C for 3 h), as demonstrated by the RE of estimates in the range of 4.73–9.44%. The results of stability assessment under various conditions including long-term, short-term, autosampler and freeze–thaw stability are illustrated in Table 4. The stability data obtained for the studied compounds are similar to those reported previously. The good stability of vitamin B2 and vitamin B6 in plasma or serum were confirmed by Geng et al. [32]. They tested stabilities at  − 20 °C for 7 days, freezing (− 20 °C) and thawing (24 ± 2 °C) for one cycle, and storing on the bench for 6 h or in the auto-sampler for 20 h at room temperature. Moreover, in the method for quantification of PLP concentrations by Ghassabian et al. [42], the samples were stable for four freeze–thaw cycles, at room temperature for 17 h, at  − 80 °C for 10 months and in the autosampler at 10 °C for at least 84 h.

Table 4 The stability of vitamins in serum samples under different conditions
Full size table

In Vivo Studies

The validated LC–MS/MS method was successfully applied for the analysis of vitamin B2 and vitamin B6 concentrations in patients with HT. In vivo studies have confirmed that the validated method may be a useful tool for the determination of vitamin deficiency in this group of patients. According to our knowledge, this is the first study to establish vitamin B2 and vitamin B6 status in HT population. The results of serum concentrations of vitamin B2 and B6 in patients with HT and healthy volunteers are shown in Table 5. In the present study, the mean concentrations of vitamin B2 and B6 in all serum individuals were 8.69 ± 8.69 ng/mL and 32.31 ± 20.59 ng/mL, respectively. The concentration of vitamin B2 was significantly lower in patients with HT compared to the control group. The concentrations of vitamin B6 was slightly lower in the study group compared to controls, but without a statistically significant difference (p > 0.05). The concentration of vitamin B2 is similar to those reported by Geng et al. [32] and the concentration of vitamin B6 is higher. The mean concentrations of vitamin B2 and B6 in the studied patient’s serum were 5.86 ± 0.85 ng/mL and 11.90 ± 3.34 ng/mL, respectively.

Table 5 Serum concentrations of vitamin B2 and B6 in patients with HT and healthy volunteers
Full size table

No differences were noted in the lipid profile between patients with HT and control subjects. Statistically significant differences between the groups occurred in TPOAb, TgAb and fT4 levels. Table 6 presents the characteristics of the studied groups.

Table 6 Characteristics of the studied populations
Full size table

In the entire studied population, the concentration of vitamin B2 was significantly negatively correlated with TSH level (R =  − 0.254; p < 0.005). In the control group, there was a positive correlation between vitamin B2 concentration and fT4 level (R = 0.378; p < 0.005). Moreover, in the HT group, there was a negative correlation between vitamin B2 concentration and TSH level (R =  − 0.250; p < 0.005) and positive with age (R = 0.262; p < 0.005). Our findings indicated a relationship between vitamin B2 and vitamin B6 levels and thyroid function; therefore, their concentration could be relevant to patients with Hashimoto’s thyroiditis. Although, further studies are needed.

Currently, no data are available on the correlation between vitamin B2 or vitamin B6 and thyroid profile to compare our results. This study is the first evaluation of these associations.

Conclusion

We developed and validated a new LC–MS/MS method for the measurement of vitamin B2 and vitamin B6 (as PLP) in human serum. The method is rapid, sensitive, specific, repeatable, reproducible, adequately accurate, and precise and fulfils the EMA validation requirements for the bioanalytical method. Moreover, it has simple sample preparation. The method was successfully applied for the analysis of these compounds in 74 serum samples of patients with HT and 33 healthy controls. This allowed the identification of patients with deficient levels of vitamin B2 and vitamin B6. The concentration of vitamin B2 was significantly lower in patients with HT compared to the control group (7.20 ± 7.86 ng/mL vs. 12.03 ± 9.62 ng/mL). Moreover, the concentration of vitamin B6 was slightly lower in the HT group compared to controls, but without a statistically significant difference (30.09 ± 19.16 ng/mL vs. 37.29 ± 23.02 ng/mL). Monitoring of vitamin B2 and B6 concentrations may be helpful in the management of HT and supplementation of deficient vitamins. Further research using the above method on a larger population of HT patients should be performed to evaluate levels of vitamin B2 and B6 in this population.