Development of innovative methodology for determination of 6-thioguanine in whole blood erythrocytes by HPLC–PDA-based technique for medical diagnostics purposes

6-Thioguanine is an immunosuppressive drug, an analogue of guanine, applied to treat acute leukemia and inflammatory bowel disease. Excessive use of 6-thioguanine during clinical treatment may cause side effects. Moreover, providing a dose too low will be ineffective. Therefore, there is a critical need for a rapid, selective and routine approach to quantifying 6-thioguanine in body fluids to support a clinical application. A fully validated HPLC method has been developed to determine 6-thioguanine in whole blood samples using 5-bromouracil as an internal standard. 6-Thioguanine nucleotides were released from erythrocytes by perchloric acid, and then hydrolysed at 100 °C to the parent thiopurine, 6-thioguanine. The following validation parameters of the method were determined: specificity/selectivity, linearity range (479–17,118 ng/mL, R > 0.992), limits of detection (150 ng/mL) and quantification (479 ng/mL), accuracy (− 5.6 < Bias < 14.7), repeatability (CV 1.30–3.24%), intermediate precision (CV 4.19–5.78%), extraction recovery (79.1–103.6%) and carryover. Furthermore, the stability of the drug in whole blood samples under various storage conditions was investigated. The suggested method is suitable for determining 6-thioguanine in whole blood erythrocyte samples for drug level monitoring, thus correct dosing.

activity together with altered thiopurine S-methyltransferase (TPMT) activity stimulate sensitivity and resistance to this compounds 12 .Furthermore, defects in the DNA mismatch repair system have been shown to stimulate acquired resistance to numerous anticancer pharmaceutical compounds, including 6-thioguanine 13 .Measurement of 6-TG erythrocyte metabolite levels helps determine the adequacy of AZA dosage and can be used to optimise the dose of antimetabolite therapy dose to achieve an improved clinical response without inducing leucopenia 14 .Combining infliximab and thiopurines (such as 6-thioguanine) in patients with inflammatory bowel diseases is more effective than monotherapy.Measurement of TPMT activity is encouraged before starting treatment of patients with thiopurine drugs such as 6-thioguanine, allowing adjustment of thiopurine dose or complete avoidance of the drug entirely 15 .In many ethnicities, TPMT polymorphisms result in decreased (10% prevalence) or absent (0.3%) TPMT activity 16 .Individuals who are homozygous or heterozygous for these types of genetic variation may have increased levels of TGN metabolites, increasing the risk of drug-induced bone marrow toxicity due to accumulation of the unmetabolised drug 17 .6-TGN levels greater than 230 pmol/8 × 10 8 RBC have been associated with improved outcomes in patients on monotherapy.A 6-thioguanine level of 125 pmol/8 × 10 8 RBCs or greater may be adequate to achieve therapeutic infliximab levels 18 .Generally, 6-TGN levels greater than 235 pmol/8 × 10 8 RBC are recommended as a therapeutic cut-off level for patients on AZA or 6-MP 19 .A safe and therapeutic level is between 500 and 800 pmol/8 × 10 8 RBCs 20 .For 6-TG therapy in patients with IBD, no precise therapeutic or toxic 6-TGN levels have been described yet.However, hepatotoxicity, particularly nodular regenerative hyperplasia (NRH) and venoocclusive disease (VOD), is dose-dependent and related to 6-TGN levels higher than 1000 pmol/8 × 10 8 RBCs 21 .Overuse of 6-TG during clinical treatment may have extra-toxic side effects 6 .Furthermore, underdosing will be ineffective.Therefore, a quick and fully validated approach to quantitatively determining 6-TG in body fluids is critical to aid in clinical use.Currently, immunoassays and high-performance liquid chromatography (HPLC) are widespread, practical, and sensitive techniques in therapeutic drug monitoring 22,23 .
Until now, several analytical approaches have been described for determining thiopurine metabolites applying UPLC-MS/MS 24 and LC-MS/MS [25][26][27] .The defined methods vary in biological matrix, sample preparation procedure, and chemical form of the analytes.However, none of the 6-TGs allowed above can be quantified directly from red blood cells by HPLC alone.Numerous laboratories worldwide observe the metabolites of thiopurine to personalise the treatment.6-TG is currently measured in red blood cells 11 and whole blood 28 .Leukocyte 6-TG levels ranged from 100 to 2305 pmol/mg DNA, while erythrocyte 6-TG levels ranged from 64 to 1038 pmol/8 × 10(8) red blood cells; leukocyte DNA 6T-G levels correlated directly with erythrocyte 6-TG levels 29 .Most of the effects of thiopurine, such as DNA incorporation and immunosuppression, occur in white blood cells rather than red blood cells.However, these cells exist abundantly and are effortlessly separated, making them a rational alternative matrix 30 .
The purpose of our work was to develop a fully validated HPLC method for the rapid and selective quantitative determination of 6-TG in whole blood erythrocytes for diagnostic purposes.The method has been rigorously validated.Furthermore, specific validation parameters and clinical evaluation have been determined and demonstrated that the method suits the routine determination of 6-TG in patients.Our goal was to develop a routine method in the first place so that any laboratory could repeat the entire procedure for clinical studies.To our knowledge, this article is the only simple, inexpensive, sensitive, well-characterised, robust, and fully validated HPLC method for monitoring 6-TG in whole blood.

Results and discussion
The HPLC method was developed for the determination of 6-thioguanine (6-TG) in whole blood samples containing red blood cells (RBCs) using 5-bromouracil (5-BU) as internal standard (IS) was developed.A mobile phase consisting of ACN/buffer KH 2 PO 4 (3/97, v/v) with 50 µl H 3 PO 4 in isocratic flow was used.6-TG was detected at wavelength 341 nm and 6-BU at 280 nm.Under the chromatographic conditions, both the compounds and the biological background were well separated.A simple and rapid method of releasing 6-thioguanine nucleotides from erythrocytes by perchloric acid and then hydrolysed (100 °C) to the parent thiopurine (6-TG) was used to prepare the samples and then hydrolysed (100 °C) to the parent thiopurine (6-TG).By converting the released 6-TG content to 6-TGN, the amount of 6-TGN present in the erythrocytes can be determined.The method has been fully validated according to EMA 22 , ICH 23 and ISO 17025 24 .Such analytical parameters are described as specificity, selectivity, linearity range, extraction recovery, and repeatability, precision (intraday and intra-day precision), accuracy, LOD, and LOQ.

Specificity.
The method turned out to be specific in the assumed experimental conditions.A comparison of the chromatogram of the sample of the aqueous solution of the 6-thioguanine standard and the internal standard (Fig. 1) shows the lack of the peaks of substances that are eluting, respectively, during the retention of the internal standard and the retention of the analyte standards.Similarly, a comparison of the chromatogram showing the separation of the sample without analytes and the IS (Fig. 2) reveals the absence of substance peaks that could negatively affect the result of the determination of the test substance.
The spectrum of 6-TG and the internal standard taken in samples containing red blood cells are shown in Fig. 3, respectively.The purity of the peaks of the substance in the samples fortified with 6-TG and IS was analysed based on the compounds absorption spectrum of the tested and the internal standard.The results of this series of tests are shown in Fig. 4. The obtained maximum purity index (PPI, expressed in [%]) for 6-TG is 99.9% and for the internal standard, 99.9%.The PPI values (criterion: PPI > 95%) demonstrates the high specificity of the method.
In conclusion, the lack of peaks of substances that elute at and around the retention times of the test substances (6-TG) and the internal standard (6-BU) in the excipients proves the accurate specificity of the determination of both test compounds.The validity of this conclusion is confirmed by the data presented in the section on the verification of the method selectivity and the analysis of the ratio of signals obtained during the retention of analytes and the internal standard in the blank samples of various samples concerning those characterising the 6-TG fortified sample at the LLOQ (lower limit of quantification) level.

Selectivity.
According to the guidelines, the absence of interfering substances is determined by comparing the signals obtained during the analyte retention time and the internal standard for the blank sample and the sample at the LLOQ level (Figs. 5, 6).The detector response during analyte retention for the blank sample should be less than 20% of the signal found in the LLOQ sample and less than 5% of the signal during IS retention.In this case, it is 0% for 6-TG and 0% for IS, respectively.Thus, the validated procedure has good selectivity.Furthermore, analysis of the peaks of retention times of the interfering substances in all tested blank samples confirms a high degree of peaks of their separation from the peaks of both analytes and IS (resolution RS > 2).The acceptability criterion was met.www.nature.com/scientificreports/Linearity range.Analysis of the linearity of the 6-thioguanine chromatographic method was carried out on blood enriched with test substances in the range that covers the therapeutic level of 6-TG concentration (100-450 pmol/8 × 10 8 RBC).However, the levels of 6-thioguaninelevels vary widely,, and the range of the method has been extended and validated from 479 to 17,118 ng/mL (48 pmol/8 × 10 8 RBC and 1650 pmol/8 × 10 8 RBC, respectively).For this purpose, a series of working calibration solutions were analysed from the 6-TG standard at a concentration of 10 mg/mL with the addition of a precisely known and equal amount of the internal standard (IS standard in a solution of 100 µg/mL).Four calibration solutions were prepared independently for each concentration level and dosed onto the chromatographic system.To verify the precision of the calculation of the 6-TG concentration based on the equations of the calibration curves (calculated based on the average values of the 6-TG and internal standard surface areas), the results are summarised in Table 1.To facilitate the comparison, the "experimental" concentrations of the calibration solutions used for calibration are presented for  ).In addition, the tables contain concentration difference (SC), measurement bias (Bias), and CV.
From the data presented above, for calibration solutions 1 to 6, Bias is in the range of − 16.3 < Bias < 10.1 (for 6-TG).Considering that the criterion of acceptability of calibration solutions is to obtain a concentration calculated from the equation of the calibration curve not differing by more than 15% from the nominal concentration (in the case of the LLOQ level, the estimated concentration should not exceed 20% of the nominal concentration), the obtained results confirm the fulfilment of the criterion of acceptability for the tested compound.
LOD and LOQ.Based on the mathematical relationships mentioned in the Materials and Methods section, the LOD (Fig. 7) and LOQ (Fig. 8) values for 6-TG were 371 ng/mL and 1123 ng/mL, respectively.However, after  www.nature.com/scientificreports/obtaining the calculation results, it was checked that the EP s/n (signal-to-noise ratio) for the concentrations of 371 ng/ml and 1123 ng/ml, respectively, are much higher than the requirements assumed; it was decided to find a concentration that would meet these requirements.LOD of 150 ng/mL (equivalent to 14 pmol/8*10 8 RBC) and EP s/n = 8 and LOQ of 479 ng/mL (equivalent to 46 pmol/8*10 8 RBC) where EP s/n = 21.The results confirm the fulfilment of the acceptability criterion for the tested compound.www.nature.com/scientificreports/analysts using different standard working solutions, and the results are presented in Table 3.The values obtained for the coefficient of variation meet the acceptability criterion.

Accuracy of the method (trueness).
Recovery.The surface areas of analytes and the internal standard obtained for five different EDTA blood samples spiked with 6-TG and IS at three concentration levels (750 ng/mL; 4389 ng/mL, 12,540 ng/mL) are presented along with the recoveries calculated for each level in Table 4.
The comparison of the average values of 6-TG and IS recovery obtained at individual concentration levels of both analytes shows that for all compounds, it is comparable and independent of the concentration of the analyte.The average recovery of 6-TG calculated for three concentration levels of this compound is 79.1-103.6%,respectively.The acceptability criterion was fulfilled.
Carryover.The signals read during analyte retention time and IS in the blank sample dosed after the calibration solution at the highest concentration level were collected.The analysis of the results revealed the lack of peaks in the chromatogram corresponding to the blank sample.The calibration solution with the highest analyte concentration was dosed immediately after dosing.It proves the absence of sample transfer, which confirms that the acceptance criterion is fulfilled.

Stability.
Stock solutions have been stable for 30 days.Standard to calibration and control-whole blood samples spiked with 6-TG at low (650 ng/mL) and high (8906 ng/mL) IS levels were used for stability checks and stored for 24 h in an autosampler at 5 °C.The choice of using concentrations of 650 ng/mL and 8906 ng/mL for stability evaluation was based on several considerations.First, these concentrations fall within the linear range of the analytical method used for the evaluation.It is essential to assess stability at concentrations within this range to ensure accurate and reliable results during routine analysis.Second, the decision to evaluate the stability  www.nature.com/scientificreports/ at these specific concentrations was influenced by the observation of such concentrations in samples from several patients.In clinical practice, it is not uncommon to encounter a diverse range of analyte concentrations in patient samples.Therefore, evaluating the stability at concentrations commonly found in real-world samples provides valuable information on the performance of the assay under realistic conditions.The detector response obtained for the 6-TG and IS peaks was compared in the sample exposed to the verified parameter.Recovery [%] was adopted as the acceptance criterion for the tested solutions about the freshly prepared solution in the range indicated above in the recovery tests.The results are presented in Table 5 as the value of the independently prepared samples.The analysis of the data presented shows that: (a) samples of the internal standard, as well as samples of basic and working standards, should not be frozen; precipitated substances were observed in the tested samples, which did not dissolve after leaving at room temperature, as well as after using vortex; and (b) the whole blood sample fortified with low and high level EDTA is very stable.Stored in an autosampler feeder (5 °C), it is stable for up to 24 h.In conclusion, whole blood samples fortified at two concentration levels of 6-TG are stable for up to 24 h at 5 °C.Patient samples were tested for 10 days, performing daily tests (samples were stored at room temperature, in the refrigerator, and frozen).Analysis revealed that samples could not be frozen (recovery below 50%).Samples stored at room temperature for 2 days recovered above 50% (for samples stored for more than 2 days).Samples stored for 7 days in the refrigerator showed appropriate recovery.The stability of the research standards (30 days-freezer) and the stability of the prepared test samples for analysis within 24 h were assessed.The study allowed for determining the stability of the tested sample (7 days at room temperature).Before developing the method described in this paper for the rapid and sensitive determination of 6-TG in red blood cells, we tested many research protocols in scientific articles on sample preparation, knowledge of sample stability, and analytical methods.Modified, validated methods from all these articles were used, adapting, and extending our approach.These studies inspired us and gave us a broad picture of method optimization.
The developed analytical method is faster compared to that contained in the work by Cangemi et al. 31 The total time of the test takes about 40 min and allows the use of a smaller volume of sample injected into the chromatographic column, which in turn can extend the life of the column, resulting in savings during routine maintenance.
Compared to the publication by Franzin et al. 32 , our method allows for a broader range of tested samples, lower limits of detection (LOD) and quantification (LOQ), as well as the ability to determine 6-TG with a resolution greater than 2.0.
Stefan et al. 33 focused on different aspects of sample stability and reference range in various study groups; our approach minimized the number of chemicals used in test samples and prompted efforts to reduce sample Preparation of calibration and control samples.A series of working calibration solutions were obtained from the 6-thioguanine standard at a concentration of 0.05 mg/ml (C r 6-TG ) with the addition of a precisely known and equal amount of the internal standard (IS standard concentration in a solution of 1 mg/ml) and red blood cells without study substances and internal standard.The red cells added to the calibration serve as a template to better replicate the conditions of blood samples.This affects the solubility and interaction of the analysis with other blood components, which is essential for calibrating the analytical method.In this case, red blood cells were added as a matrix element to adapt the samples to more realistic laboratory conditions.The volume ratios used to prepare the working calibration solutions are summarised in Table 6.Control samples were prepared in whole blood containing neither study substances nor internal standard by fortification with 6-TG standard solutions at a concentration of 0.05 ng/mL (C r 6-TG) with the addition of a precisely known and equal amount of the internal standard concentration (concentration of the IS standard in a solution of 1 mg/mL).The volume ratios used to prepare the control samples are summarised in Table 7.
Sample preparation procedure for chromatographic analysis.Pipette exactly 200 µl of whole blood, 140 µl of saline (for calibration and control samples, add the appropriate amount of working solution and a correspondingly smaller amount of saline), 10 µl of internal standard solution (C r IS), 25 µl of DTT (ditriethiol-1.1 M solution) and 50 µl of perchloric acid (65%) in an eppendorf tube.The tube contents were thoroughly mixed using a vortex (30 s) and centrifuged (12,500 rpm, 10 min).The supernatant was then transferred to Eppendorf and heated at 100 °C for 40 min.After this time, the sample was cooled and transferred to the HPLC plate.The vial contents were thoroughly mixed using a shaker (10 s) and then placed in a vial of the autosampler feeder.Essential to ensure that the liquid in the insert does not contain air bubbles.

Chromatographic conditions.
All analyses were performed on the Nexera X2 liquid chromatographic system model (Shimadzu, Japan), equipped with a SIL-30ACXR autosampler, LC-30 AD pump, and Separations were carried out using a Luna™, Phenomenex analytical column (250 × 4.6 mm, 5 μm) analytical column.A mobile phase consisting of 6.8 g of ACN/buffer KH2PO4 was used in 1L 3/97% v/v with the addition of 50ul H3PO4 at a constant flow rate of 1.0 mL/min.The stationary phase was acetonitrile.The column was kept at wavelengths of 30 °C, and the eluents were monitored at 280 nm (5-BU analysis) and 341 nm (6-TG analysis).The injection volume was 50 μl.

Assay validation.
Method validation was carried out under the guidelines EMA 34 , ICH 35 and ISO 17025 36 in selectivity/specificity, linearity range, limits of detection and quantification, accuracy, precision, extraction recovery, carryover, and stability.Selectivity/specificity.A particular/specific analytical method should allow the resolution and detection of the drug of interest and the internal standard in the occurrence of potential coadministered drugs and coeluting endogenous compounds in the matrix 37 .
The specificity of the determination of the 6-thioguanine method was evaluated by analysing serum samples without test substances and internal standard (negative test) and the resolution between the peaks of aqueous solutions of test substance standards (positive test) and aqueous solutions of the internal standard (positive test), as well as between these peaks and whole blood samples with the addition of test substances (samples fortified with 6-thioguanine) and serum samples fortified with an internal standard.
Five blank samples from different patients were analysed to verify the method's selectivity.In each chromatogram, the signals obtained during the retention time of 6-TG and IS were integrated and compared with those obtained in the sample enriched with the level of test substance at the LLOQ (lower limit of quantification).

Linearity range.
The linearity of an analytical method means the direct proportional response of the results obtained to the analyte concentrations in an appropriate calibration set.The linearity study of the 6-thioguanine chromatographic method was carried out on blood enriched with test substances in the range covering the therapeutic level of 6-TG concentration (100-450 pmol/8 × 10 8 RBC), according to the CLSI guidelines 38 .However, 6-thioguanine levels varied significantly, and the range of the method was extended and validated from 479 to 17,118 ng/mL (48 pmol/8 × 10 8 RBC and 1650 pmol/8 × 10 8 RBC, respectively).In converting the concentration range in pmol/8 × 10 8 RBC to ng/mL, the molar mass of 6-thioguanine should be considered.Convert pmol to ng: divide the concentration range in pmol/8 × 108 RBC by the number of red blood cells (8 × 10 8 ) to get the concentration in pmol/RBC.The concentration in pmol/RBC was then multiplied by the molar mass of 6-thioguanine to convert to ng/RBC.Convert ng/RBC to ng/mL: concentration in ng/RBC was divided by the blood volume (mL) used for the analysis, and the results were adjusted to the patient's red blood cell (RBC) count.
For this purpose, a series of working calibration solutions obtained from the 6-TG standard at a concentration of 10 mg/mL with the addition of a precisely known and equal amount of an internal standard (concentration of Table 6.Volume ratios are used to prepare working calibration solutions.

Figure 1 .
Figure 1.Chromatograms of an aqueous solution of 6-TG and IS.

Figure 2 .Figure 3 .
Figure 2. Chromatograms of a sample devoid of 6-TG and IS.

Table 1 .
Comparison of experimental concentrations and those calculated using the 6-TG calibration curve equation.

Table 2 .
Repeatability of 6-TG analysis in a whole blood sample, free of a test substance, five times fortified with 6-TG at three concentration levels.

Table 3 .
Intermediate precision of 6-TG analysis in various fortified serum samples (n = 10) at three analyte concentration levels.

Table 4 .
6-TG/IS surface areas were obtained for blood samples fortified at three levels.

Table 5 .
The stability of the whole blood sample in the autosampler feeder (temp.5 °C) was fortified at 650 ng/mL and 8906 ng/mL concentrations of 6-TG.

Table 7 .
Volume ratios were used to prepare control samples.