Introduction

As currently available irreversible P2Y12 receptor inhibitors, thienopyridine antiplatelet agents are widely used for the treatment of myocardial infarction, cerebral infarction, and arterial or venous thrombosis, as well as diseases related to platelet aggregation [1,2,3]. The thienopyridine P2Y12 receptor inhibitors used in clinical settings are clopidogrel and prasugrel [4,5,6].

However, the use of clopidogrel as a first-line P2Y12 receptor inhibitor has a certain disadvantage known as “clopidogrel resistance”, in which some patients show low response, no response, or resistance to treatment for cardiovascular diseases [7]. The United States Food and Drug Administration issued a black box warning cautioning that CYP2C19 poor metabolizers are at a higher risk of treatment failure with clopidogrel [8, 9]. Moreover, a high oral dose of clopidogrel is required because it is inefficiently metabolized into active metabolite in vivo. Clopidogrel undergoes a two-step CYP450-dependent oxidation to form the active thiol metabolite in the liver. However, during the process, ~85% of clopidogrel is hydrolyzed by human carboxylesterase 1 to form inactive clopidogrel carboxylic acid, and <15% of clopidogrel is transformed into 2-oxo-clopidogrel in the first step (Fig. 1) [10,11,12]. The active metabolite M15-2 is generated from 2-oxo-clopidogrel, and the inefficient biotransformation to 2-oxo-clopidogrel in the first step restricts the use of clopidogrel.

Fig. 1: Main pathway in forming active metabolite of clopidogrel and vicagrel (* indicates 14C labeled position).
figure 1

AADAC arylacetamide deacetylase, CES carboxylesterase.

Full size image

Prasugrel, another P2Y12 receptor inhibitor commonly used in clinical settings, shows more consistent platelet inhibition and higher efficiency in producing active metabolite than clopidogrel; however, its use leads to a higher risk of bleeding than clopidogrel [13, 14]. Therefore, new thienopyridine antiplatelet agents must be developed to improve the current antiplatelet therapy.

Vicagrel, a novel thienopydine P2Y12 antagonist, is undergoing phase III trials in China [9]. Vicagrel, which has a structure similar to that of clopidogrel, also requires a two-step metabolic process to form an active metabolite, and the second step of both agents is exactly the same [15, 16]. In the first step, almost all vicagrel is hydrolyzed by human carboxylesterase 2 and arylacetamide deacetylase to form 2-oxo-clopidogrel [15, 17]; hence, the exposure of 2-oxo-clopidogrel to the two drugs is substantially different (Fig. 1). Thus, a considerably lower dose of vicagrel may yield a plasma exposure of the active metabolite M15-2 comparable to that of clopidogrel. Previous studies demonstrated that the same exposure level of M15-2 can be achieved after oral administration of 5 mg vicagrel versus 75 mg clopidogrel, a finding that is consistent with the proposed drug design hypothesis [18]. In view of these advantages, vicagrel is expected to induce less drug resistance than clopidogrel, as well as to improve drug safety by decreasing the risk of bleeding.

Although vicagrel shows potential as a first-line medicine and has been investigated in numerous studies, its use still has issues [7, 8]. Thus far, no data are available to evaluate its overall in vivo metabolism. Preliminary vicagrel mass balance studies using nonradiolabeled vicagrel demonstrated that ~18% of the dose is excreted, mainly in human urine (unpublished data). However, the fate of the remaining metabolites is still unclear. Owing to the low recovery rate, determining the major metabolites and the metabolism mechanism is difficult. In addition, no data are available concerning the metabolism of [14C]vicagrel in animals or humans. A better understanding of vicagrel metabolism is important to identify its metabolic profile, assess the safety of its metabolites, understand its pharmacokinetics, and determine its mechanisms of clearance, which in turn are necessary to identify potential drug–drug interactions (DDIs) [19,20,21]. This information is necessary in guiding future clinical evaluations of vicagrel and ultimately determining appropriate dosages for patients [22]. The use of radioactive tracers, such as 14C-labeled compounds, in absorption, distribution, metabolism, and excretion studies enables the determination of the mass balance and major metabolic routes of a drug candidate [23,24,25,26,27]. In the present study, [14C]vicagrel suspension solution was administered to healthy male Chinese subjects, and its pharmacokinetics, mass balance, and biotransformation pathway were investigated.

Materials and methods

Chemical and reagents

Vicagrel (99.20% purity) and the reference substances M3, M6, d6-M6, M9-2, MP-M15-1, and MP-M15-2 were kindly supplied by Jiangsu Vcare PharmaTech Co., Ltd. (Nanjing, China). Both [14C]vicagrel (120 µCi, 98.68% purity) and vicagrel (20 mg) were dissolved in a 95% ethanol solution (~3 mL, pharmaceutical grade) and preserved at ~–80 °C. The oral suspension was prepared by diluting the ethanol solution with 80 mL pure water on the day of administration. The 14C-label was located on the ester carbonyl group (Fig. 1). The derivatization reagent 2-bromo-3′-methoxyacetophenone (MPB) and high-performance liquid chromatography (HPLC)-grade acetonitrile and methanol were purchased from Sigma (St. Louis, MO, USA). Ammonium acetate was obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Formic acid was purchased from Rhawn Chemical Reagent Co., Ltd. (Shanghai, China). Ultrapure water was obtained from a Milli-Q integral water purification system (Millipore, Molsheim, France). The other commercially available reagents used herein were of analytical grade.

Instruments and conditions

With the help of mass defect filters and background subtraction techniques, high-resolution mass spectrometry (HR-MS) and HR-MS2 acquisition are widely used for metabolite identification [28, 29]. In the present study, HR-MS and HR-MS2 acquisition were performed on a Vanquish ultra-high-performance liquid chromatography (UHPLC) system coupled with a Q Exactive Plus mass spectrometer (Thermo, Waltham, MA, USA) equipped with an electrospray ionization (ESI) source. Chromatographic separation was achieved on an Agilent Extend C18 column (150 mm × 4.6 mm, 5 µm; Agilent, Santa Clara, CA, USA) at 25 °C with a flow rate of 0.6 mL/min. The mobile phase consisted of 5 mM ammonium acetate in water with 0.05% formic acid (A) and acetonitrile (B). The gradient elution program was as follows: 0 min, 5% B; 3 min, 5% B; 35 min, 35% B; 45 min, 75% B; 46 min, 95% B; 48 min, 95% B; 49 min, 5% B; 60 min, 5% B. The eluent was monitored by UV detection at 254 nm, and (+)ESI mode was applied in MS detection at a scan range of m/z 100–800 Da. The optimized MS parameters were as follows: sheath gas, 45 L/min; aux gas, 10 L/min; capillary temperature, 320 °C; aux gas heater temperature, 400 °C; capillary voltage, 3.5 kV. Data acquisition was performed via Xcalibur software (Thermo), and data analysis was accomplished using Compound Discoverer software (Thermo).

Study design, subjects, and sample collection

This open-label, single-center, and single-dose study was conducted at the First Affiliated Hospital of Soochow University (Suzhou, China). Six healthy male Chinese subjects aged between 19 and 36 years with a body mass index between 21 and 26 kg/m2 were recruited. The study was conducted in compliance with the ethical principles required in the Declaration of Helsinki and approved by the Hospital Ethics Committee. All participants provided written informed consent prior to the commencement of the study. Plasma, urine, and feces samples were collected 0–168 h after oral administration, and the subjects were discharged from the clinic when the following criteria were satisfied: the accumulated excreted radioactivity was over 80% of the administered radioactivity, the excreted radioactivity was lower than 1% of the administered radioactivity over 2 consecutive days, and the measured radioactivity in plasma collected was less than twofold that of predose over 2 consecutive days [30]. The subjects were instructed to stay in the clinical facility for follow-up until they reached normal health conditions if they had major clinically abnormal conditions. Each of the six subjects received a single oral dose of 20 mg [14C]vicagrel (120 µCi) as a suspension solution after fasting for at least 10 h. The bottle used for drug administration was rinsed with pure water, and the volunteers consumed the rinsing fluids (the total volume was ~240 mL) within 5 min. The subjects were prohibited from drinking water for 1 h and fasted for 4 h. Subsequently, 18 mL of whole blood was collected at predose and 0.25, 0.5, 1, 2, 4, 8, and 24 h postdose. From these samples, 7 mL was immediately mixed with esterase inhibitor in an ice water bath, and another 7 mL was immediately mixed with MPB and allowed to react for at least 10 min before centrifugation to yield the respective plasma. Approximately 1 mL of the remaining blood was aliquoted into a blood tube, and the leftover blood (3 mL) was centrifuged to yield plasma. Furthermore, 8 mL of whole blood was collected at 1.5, 2.5, 3, 6, 12, 36, 48, 72, 96, 144, and 168 h. Approximately 2 mL of whole blood was immediately mixed with esterase inhibitor in an ice water bath, and another 2 mL was immediately mixed with MPB and allowed to react for at least 10 min before centrifugation to yield the respective plasma. Approximately 1 mL of the remaining blood was aliquoted into a blood tube, and the leftover blood (3 mL) was centrifuged to yield plasma. Urine samples were collected at predose and 0–4, 4–8, 8–12, 12–24, 24–48, 48–72, 72–96, 96–120, 120–144, and 144–168 h postdose. Fecal samples were collected at predose and 0–24, 24–48, 48–72, 72–96, 96–120, 120–144, and 144–168 h postdose. All samples were preserved at ~−80 °C until analysis.

Safety assessment

Drug safety was evaluated via physical examination, assessment of vital signs, and laboratory tests, and the occurrence of any adverse event was recorded. The vital signs of the subjects were assessed at least once during each study day. Physical examinations were performed before and after each study period. Adverse events were recorded throughout the study period.

Analysis of total radioactivity

The total radioactivity in plasma and urine was measured using a liquid scintillation counter (LSC) (Tri-Carb 3110 TR, Perkin Elmer, Waltham, MA, USA). Each fecal sample was mixed with twice the weight of ethanol-water (1:1, v:v) and then homogenized. Blood and fecal homogenates were weighed and combusted using an OX-501 Biological Oxidizer (Harvey, Tappan, NY, USA). The 14CO2 generated was then trapped in alkaline RDC liquid scintillation cocktail (RDC, Hilladale, NJ, USA) and measured using LSC.

Sample preparation and metabolic radioprofiling

Plasma

The 0–8 h plasma pretreated with MPB from the six subjects was individually pooled according to the Hamilton pooling principle, and then an equal volume from each subject was pooled [31]. An aliquot of pooled plasma sample (4.5 mL) was extracted by adding 13.5 mL methanol-acetonitrile (50:50, v:v). The mixture was then vortexed for 1 min and centrifuged at 5000 rpm for 10 min at room temperature. The supernatant was transferred into another clean tube, and then the postextraction solid was added to 2.25 mL water and extracted again with 13.5 mL methanol-acetonitrile (50:50, v:v). The two supernatants were combined and dried under nitrogen at 25 °C. The residues were reconstituted in 400 μL acetonitrile-water (20:80, v:v), 90 μl of which was injected into the UHPLC-Fraction collector system (Thermo). The UHPLC eluent was collected in Deepwell LumaPlate 96 plates (Perkin Elmer, USA) for 0–48 min with every 8 s eluent collected in one well. The plates were dried at room temperature, and the CPM values were read by a Sense Beta plate reader (Hidex, Turku, Finland). Data were reconstructed into a radiochromatogram by using Laura software (Lablogic, Broomhill, Sheffield, UK). Another 15 μl of the reconstituted solution was injected for UHPLC-Q Exactive Plus MS analysis.

Urine

An equal percentage by volume of each urine sample was pooled for 0–24 h across the six subjects. Then, 9 mL of the pooled urine sample was dried under nitrogen at 25 °C. The residues were reconstituted in 700 μL acetonitrile-water (20:80, v:v), 35 μL of which was injected for UHPLC-β-RAM (Lablogic, UK) online radioactivity detection. Another 7 μL of the reconstituted solution was injected for UHPLC-Q Exactive Plus MS analysis.

Feces

An equal percentage by weight of each fecal homogenate sample was pooled for 0–48 h across the six subjects. An aliquot of pooled fecal homogenate (5 g) was extracted by adding 15 g methanol-acetonitrile (50:50, v:v). The mixture was then vortexed for 1 min and centrifuged at 5000 rpm for 10 min at room temperature. The supernatant was transferred into another clean tube, and then the postextraction solid was added to 2.5 mL water and extracted again with 15 g methanol-acetonitrile (50:50, v:v). The two supernatants were combined and dried under nitrogen at 25 °C. The residues were reconstituted in 550 μL acetonitrile-water (20:80, v:v), and 35 μL of the reconstituted solution was used for radiochromatogram analysis following the same method applied in plasma samples. Another 15 μL of the reconstituted solution was injected for UHPLC-Q Exactive Plus MS analysis.

Metabolite identification

The structures of metabolites were identified by UHPLC-Q Exactive Plus MS by comparing their mass spectral fragmentations with those produced by the parent compound and other reference substances available. The chlorine isotopic distribution pattern (ratio of 3:1) was also used for metabolite identification.

Determination of M3, M9-2, M15-1, and M15-2 in human plasma

The plasma sample pretreated with esterase inhibitor was used to determine the M3 concentration. The analyte M3 and the internal standard (d6-M6) were separated on a BEH C18 column (50 mm × 2.1 mm, 1.7 µm; Waters, Milford, MA, USA). The column temperature was maintained at 45 °C. The mobile phase consisted of water with 0.05% formic acid and acetonitrile at a flow rate of 0.55 mL/min. Detection was performed in (+)ESI mode by using Qtrap 5500 LC-MS/MS (Sciex, Framingham, MA, USA). The quantitative transition ion pairs were m/z 324.0 → 112.9 (M3) and m/z 348.2 → 159.0 (d6-M6).

The plasma sample pretreated with MPB was analyzed to quantify M9-2, M15-1, and M15-2 by using a validated LC-MS/MS method [18, 32].

Pharmacokinetic analysis

Plasma concentration-time data for vicagrel and its metabolites and the total radioactivity concentration-time data for [14C]vicagrel-related components in plasma and blood were analyzed using noncompartmental methods in Phoenix WinNonlin (version 7.0; Pharsight Corporation, Mountain View, CA, USA). The following parameters were calculated: area under the curve (AUC) for plasma concentration versus time up to the last measurable time point (AUClast), AUC curve to infinity (AUCinf), maximum plasma concentration (Cmax), time to reach Cmax (Tmax), and half-life (t1/2). The AUCinf and t1/2 values were not calculated unless at least three time points (of which the first time point must be greater than Tmax) with quantifiable concentrations were obtained. The blood-to-plasma AUCinf ratio was calculated by dividing the AUCinf of total radioactivity in blood by the AUCinf of total radioactivity in plasma. The AUC ratio of metabolites to total radioactivity was calculated by dividing the AUCinf of metabolites by the AUCinf of total radioactivity. The excretion of radioactivity in urine and feces (percentage of dose) was calculated by dividing the excreted radioactivity by the administered radioactivity.

Results

Safety assessment

During the clinical trial, two cases of adverse events occurred in one of six subjects; i.e., the patient experienced bleeding under the left foot nail and fever. Fever was considered to be possibly related to the study drug, and adverse events were recorded and followed up until recovery/cure. The other case (bleeding under the left foot nail) was regarded as possibly unrelated to the study drug. No serious adverse events were reported. According to the Common Terminology Criteria for Adverse Events v. 5.0, the severity of these two adverse events was defined as grade I. Therefore, vicagrel was considered safe in clinical practice.

HR-MS analysis of vicagrel and reference standards of five possible metabolites

The chromatographic and HR-MS fragmentation patterns of vicagrel and five possible metabolic reference standards were examined. The fragmentation patterns of the major product ions were characterized on the basis of accurate mass measurements and used to determine potential metabolites.

Vicagrel, C18H18O4NClS, with [M + H]+ at m/z 380.0717, eluted at 47.88 min and showed product ions at m/z 152.0260, 184.0521, 212.0469, and 338.0608 (Fig. 2a, b). The base peak ion at m/z 212.0469 was generated by N-C cleavage; further neutral loss of CO and a C2H4O2 chain led to m/z 184.0521 and 152.0260, respectively.

Fig. 2: Extracted ion chromatograms, product ion spectra, and proposed fragmentaion patterns.
figure 2

Vicagrel (a, b), M3 (c, d), M6 (e, f), M9-2 (g, h), MP-M15-1/MP-M15 (i, j).

Full size image

M3, a pair of isomers, possessed protonated ions at m/z 324.0458 and 324.0459 with retention times of 19.54 and 20.61 min, respectively. They yielded almost the same product ions at m/z 141.0100, 156.0477, 169.0051, and 278.0402 (Fig. 2c, d). M3 was generated from vicagrel by ester hydrolysis at two sites.

M6, also a mixture of isomers, eluted at 21.30 and 21.50 min and displayed the protonated molecular ion m/z 342.1107. M6 produced characteristic product ions at m/z 141.0783, 155.0258, 183.0206, 282.0889, and 324.0992 (Fig. 2e, f).

M9-2 with [M + H]+ at m/z 370.0874 eluted at 41.66 min and showed product ions at m/z 155.0258, 183.0205, 212.0470, and 322.0836 (Fig. 2g, h). The ion at m/z 322.0836 was generated from neutral loss of CH3SH and further loss of hydrogen. The other product ions were the same as those of M6.

MP-M15-1 and MP-M15-2 displayed protonated ions at m/z 504.1251 and 504.1250 with retention times of 44.92 and 45.42 min, respectively. They yielded similar product ions at m/z 155.0259, 183.0207, 212.0471, and 354.0558 (Fig. 2i, j). The product ion at m/z 354.0558 was generated by cleavage of MPB and further loss of hydrogen. The other product ions were similar to those of M9-2.

Pharmacokinetics

The radioactivity concentration-time profiles in blood and plasma after a single oral administration of 20 mg [14C]vicagrel (120 µCi) to healthy male Chinese subjects are shown in Fig. 3a, and the related pharmacokinetic parameters for radioactivity are summarized in Table 1. In plasma and blood, the mean Cmax values of radioactivity were 1430 and 694 ng eq./mL, the mean AUClast values were 11,200 and 6020 ng eq./mL ·h, and the mean AUCinf values were 13,700 and 7590 ng eq./mL ·h, respectively. The mean Tmax and t1/2 were ~0.625 and 38 h in blood and plasma, respectively. The blood-to-plasma AUCinf ratio of the radioactivity was 0.55. The mean plasma concentrations of the metabolites are shown in Fig. 3b. The pharmacokinetic parameters of vicagrel and its metabolites are presented in Table 1. The mean Tmax of M3, M9-2, M15-1, and M15-2 was 0.42, 0.83, 0.42, and 0.33 h, respectively. The mean t1/2 of M3, M9-2, M15-1, and M15-2 was 1.97, 13.40, 2.54, and 2.05 h, respectively. The mean Cmax of M3 was 405, 288, 76.4 and 60.1 ng/mL, respectively. The mean exposure (AUCinf) of M9-2, M3, M15-1, and M15-2 was 2890, 461, 59.0, and 41.8 ng/mL ·h, respectively. According to the AUCinf total radioactivity ratios, M9-2 was the most abundant moiety in plasma (21.09%), followed by M3 (3.36%), M15-1 (0.43%), and M15-2 (0.31%).

Fig. 3: Radioactive pharmacokinetics, LC-MS/MS analysis, Mass balance.
figure 3

(a), mean concentration of radioactivity in blood and plasma; (b), mean concentration of metabolites M3, M9-2, M15-1 and (c), M15-2 in plasma and mean cumulative excretion of total radioactivity in urine and feces following a single oral administration of [14C]vicagrel . Each point represents the mean ± SD of six subjects.

Full size image
Table 1 Pharmacokinetic parameters of radioactivity in blood, plasma, and metabolites in plasma after a single oral administration of [14C]vicagrel to healthy volunteers (mean ± SD) using noncompartmental methods in Phoenix WinNonlin.
Full size table

Mass balance

In the six healthy male Chinese subjects after oral administration of 20 mg [14C]vicagrel (120 µCi), the mean total radioactivity recovery was 96.71% (93.14%–100.90%). Urine excretion was the predominant route of elimination, accounting for a mean of 68.03% of the administered dose, whereas the mean fecal excretion was 28.67%. The total recovery of radioactivity in urine and feces at 48 and 72 h after dosing was 92.00% and 94.41%, respectively, as shown in Fig. 3c.

Quantitative metabolite profiling

The UHPLC eluents of 0–8 h plasma and 0–48 h fecal samples were collected by a UHPLC-Fraction collector into Deepwell LumaPlate 96 plates. The radioactivity in each well was counted by Sense Beta and then used to reconstruct the respective radiochromatograms by using Laura software. The 0–24 h urine sample was injected for UHPLC separation, followed by β-RAM online radiodetection. Plasma, urine, and fecal samples were also injected for UHPLC-Q Exactive Plus MS analysis.

The radiochromatograms of 0–8 h plasma, 0–24 h urine, and 0–48 h fecal samples are shown in Fig. 4. The identities of 22 proposed metabolites, including their biotransformation, elemental composition, protonated m/z ([M + H]+) values, mass error (ppm), and characteristic fragment ions, are summarized in Table 2. The nomenclature of the metabolites is consistent with that from Dr. Cai Liu’s doctoral dissertation for the already identified metabolites; for the newly identified metabolites, the assigned name was the molecular weight prefixed with “M”.

Fig. 4: Representative radio-chromatograms of metabolites.
figure 4

(a), Plasma (0–8 h); (b), Urine (0–24 h); (c), Feces (0–48 h).

Full size image
Table 2 Information of vicagrel metabolites detected in human plasma, urine, and feces by using UHPLC-Q Exactive Plus MS.
Full size table

Plasma

In AUC-pooled 0–8 h plasma, unchanged vicagrel was not detected, and a total of 12 radiochromatographic peaks were identified (Fig. 4). Among them, M9-2 was the predominant drug-related component in the AUC-pooled 0–8 h plasma pretreated with MPB. This result was consistent with that of a previous study [18].

M9-2

The protonated molecular ion [M + H]+ of M9-2 was detected at m/z 370.0879 with a formula of C17H20O4NClS according to exact mass measurement. The major product ions of M9-2 were at m/z 155.0258, 183.0205, 212.0469, and 322.0836 (Fig. 5a, b). The structure of M9-2 was confirmed by comparing the retention time and LC-MS/MS spectrum with those of the reference standard.

Fig. 5: Extracted ion chromatograms, product ion spectra, and proposed fragmentaion patterns.
figure 5

M9-2 (a, b), M3-1 (c, d), M4 (e, f), M5 (g, h), M2 (i, j), M352 (k, l) and M21 (m, n).

Full size image

Several metabolites (M3-1, M4, M8, M352, M11-4, MP-M15-1, and MP-M15-2) were identified as minor metabolites of vicagrel. The structures of MP-M15-1 and MP-M15-2 were confirmed by comparing them with reference standards. M9-2, M3-1, M8, M352, and M4 accounted for 39.43%, 8.71%, 5.70%, 5.23%, and 3.48% of the total plasma radioactivity, respectively.

Urine

In the 0–24 h pooled urine sample, a total of 13 radiochromatographic peaks were identified, but the parent vicagrel was not detected (Fig. 4). Four abundant metabolites were assigned as M3-1, M4, and M5/M2 (coeluting), accounting for 12.59%, 16.25%, and 11.28% of the dose, respectively.

M3-1

The [M + H]+ of M3-1, m/z 324.0459, was 56.0261 Da lower than that of vicagrel and possessed the same molecular weight as the reference standard M3. The formula obtained from exact mass measurement implied the loss of C3H4O, corresponding to the loss of CH2CO (42.0106 Da) and CH2 (14.0157 Da) from the parent drug. The main ions at m/z 141.0101, 152.0260, 169.0051, and 198.0314 also supported the structure of M3-1 (Fig. 5c, d). The double bond position was proposed by comparing the LC-MS/MS spectrum with that of the reference substance of M3.

M4

The [M + H]+ of M4, m/z 340.0411, was 15.9952 Da larger than that of M3. The calculated elemental composition was C15H14O4NClS, indicating mono-oxidation of M3. The same product ions at m/z 141.0100 and 169.0052 as those of M3 suggested that the oxidation was located at the thienopyridine moiety (Fig. 5e, f). The double bond could not be confirmed by MS data alone.

M5

The [M + H]+ of M5, m/z 340.0951, was 2.0163 Da lower than that of M6. The calculated elemental composition was C16H18O5NCl, indicating dehydrogenation of M6. The main ions at m/z 139.0627, 155.0258, 183.0206, and 212.0470 supported the structure of M5 (Fig. 5g, h). The stereochemistry of the hydroxyl moiety could not be confirmed by the MS data.

M2

The [M + H]+ of M2, m/z 324.1003, was 18.0111 Da lower than that of M6. The calculated elemental composition was C16H18O4NCl, indicating a water molecule less than that of M6. The accurate mass product ions at m/z 155.0258, 184.0521, and 212.0470 also supported the structure of M2 (Fig. 5i, j).

Several phase II glucuronide acid conjugates (M14-1, M14-2, M18-1, and M18-2) were also detected as minor metabolites in urine.

Feces

A total of 12 radiochromatographic peaks were identified in the pooled 0–48 h fecal samples (Fig. 4). Among them, two abundant metabolites, M352 and M21, accounted for 6.81% and 4.76% of the dose, respectively.

M352

The [M + H]+ of M352 was detected at m/z 353.1269. Its calculated elemental composition was C17H23O4N2Cl according to exact mass measurement. The main product ions were at m/z 152.0261, 169.0052, 198.0314, 264.0783, and 307.1203. The structure of M352 was deduced by comparing its ion fragments with those of M3 and M6 (Fig. 5k, l).

M21

The [M + H]+ of M21, m/z 310.0850, was 14.0153 Da lower than that of M2, implying the loss of CH2 (14.0157 Da) from M2. Its calculated elemental composition was C15H16O4NCl, indicating that M21 was generated by ester hydrolysis from M2. The accurate mass product ions at m/z 152.0261, 169.0051, and 198.0314 also supported the structure of M21 (Fig. 5m, n).

M7-1, M11-4, and M1 were also identified in the fecal sample as minor metabolites.

Discussion

This study reported the pharmacokinetics, mass balance, and metabolism of [14C]vicagrel in humans. The metabolic profile of vicagrel was determined in plasma, urine, and fecal samples after oral administration of 20 mg [14C]vicagrel (120 µCi) suspension solution in six healthy male Chinese subjects. In expected real-world clinical practice after marketing, the recommended regimen is 20 mg as the starting dose and 5 mg as the subsequent maintenance dose. After oral administration, 96.71% of the dosed radioactivity was recovered in urine (68.03%) and feces (28.67%) at 168 h postdose, indicating complete excretion. Mass balance analysis indicated that the predominant route of elimination for drug-related material was renal excretion, a result consistent with that of a previous study (unpublished, Dr. Cai Liu’s doctoral dissertation).

On the basis of high radioactivity recovery, metabolite profiling in plasma, urine, and fecal samples was conducted. A total of 22 metabolites were identified, and unchanged vicagrel was not detected in the plasma, urine, or fecal samples following oral administration of [14C]vicagrel. Actually, the result of “the parent compound has not been detected in any matrix” was what we expected. Vicagrel, as a structural analog of clopidogrel, was designed as a prodrug on the basis of clopidogrel metabolism. As shown in Fig. 1, 85% of clopidogrel underwent ester hydrolysis (C-7) to clopidogrel carboxylic acid, which is an inactive metabolite. Only 15% of clopidogrel was oxidized (C-17) to 2-oxo-clopidogrel, the precursor of the active metabolite M15-2. For vicagrel, which possesses a second ester moiety, the ester at C-17 was preferably hydrolyzed rather than the ester at C-7 to form 2-oxo-clopidogrel; hydrolysis efficiency was almost 100%, and thus, no parent drug was detected in any matrix. The subsequent metabolic pathway from 2-oxo-clopidogrel was the same as that of clopidogrel. As evidence of the success of this drug design hypothesis, the exposure of M9-2 and MP-M15-2 was detected at almost the same concentration in the 5 mg vicagrel dose group and the 75 mg clopidogrel dose group [18].

In plasma pretreated with the derivatization reagent MPB, M9-2 was the predominant drug-related component, accounting for 39.43% of radioactivity in the pooled AUC0–8 h plasma. The concentration of M9-2, the most abundant metabolite in plasma, was similar to that from previous clinical trials [18]. The concentrations of the other metabolites, such as M3, M15-1, and M15-2, were also at the same level. In urine, M4 was the most abundant metabolite, accounting for 16.25% of the dose, followed by M3-1, accounting for 12.59% of the dose. In feces, M21 was the major metabolite, accounting for 6.81% of the dose.

The biotransformation pathway proposed herein is shown in Fig. 6. Phase I metabolism, such as deacetylation, demethylation, hydrolysis, and reduction, was the major metabolic pathway, whereas phase II metabolism, such as glucuronidation and methylation, was the minor metabolic pathway. As shown in Fig. 6, the most susceptible metabolic site of vicagrel was the thiophene ring, resulting in S- or R-isomerization at C-4 and E- or Z-isomerization at the double bond between C-3 and C-16 [33, 34]. Although several reference standards were synthesized to characterize the structures of the metabolites, determining the exact structures of multiple isomers was challenging. On the basis of the MS2 spectra of available reference standards and possible metabolites, the product ions at m/z 212.0470, 198.0314 (212.0470–CH2), 183.0208, and 169.0051 (183.0208–CH2) were used as diagnostic ions to determine the position of the double bond. However, this rule is not absolutely applicable and has some exceptions.

Fig. 6: Metabolic pathway of vicagrel in healthy Chinese male subjects.
figure 6

GluA glucuronic acid, MPB derivation reagent (2-bromo-3′-methoxyacetophenone).

Full size image

Among all metabolites, M352 showed a distinctive structure that has never been reported before. Its calculated elemental composition, C17H23O4N2Cl, had one more nitrogen atom than the parent drug, indicating that M352 might be generated from a metabolic intermediate through conjugation with amino acids. Alanine, one of the most abundant amino acids in healthy humans, was considered the most likely because of its low molecular weight. The proposed formation mechanism of M352 includes the following two pathways. First, alanine, as a nucleophilic reagent, attacks the mercaptomethyl group at the C-4 position to form a metabolic intermediate in which the alanine is substituted as a new group. The reaction is considered an SN2 reaction, and during the process, mercaptomethyl is considered a good leaving group [35]. The other possible mechanism is that the ketone in the C-4 position is prone to amination reactions and forms enamine with alanine, which is then transformed to a metabolic intermediate by reductase [36]. Why did we not observe ketones from the metabolic pathway? The metabolites with ketones at the C-4 position tended to form compounds possessing hydroxyl groups, such as M5, M6, and M6-3. Finally, M352 is generated from the metabolic intermediate via decarboxylation [37]. On the basis of this premise, we attempted to isolate M352 to identify its exact structure via nuclear magnetic resonance but failed to obtain a sufficient amount because M352 was a minor metabolite in feces.

According to the 2020 FDA guidance Safety Testing of Drug Metabolites Guidance for Industry, metabolites constituting over 10% of the total drug-related exposure in plasma should be evaluated for safety concerns. In this study, we collected 0–168 h plasma from six subjects and measured the radioactivity for each time point. We also evaluated AUC pooling at 0–12 h or 0–24 h to analyze the metabolite profiles in addition to 0–8 h pooling. However, the concentration in the pooled sample was insufficient for metabolite identification. After balancing the pooling time interval and sample concentration in pooled plasma, we finally chose 0–8 h. In this study, only M9-2 accounted for over 10% of the total drug-related components in human plasma. Several in vitro and in vivo drug metabolism and pharmacokinetics studies were performed in-house or outsourced by Vcare Company. In an in vitro study, the percentages of metabolites detected in human S9 were similar to those in monkey, rat, and dog S9 (unpublished data). In a drug metabolism study after intragastric administration of [14C]vicagrel in rats, no parent compound was detected in plasma, urine, feces or bile. In rat plasma, M9-2 and MP-M15-2 were the major metabolites. The AUC of M9-2 was 2862 ng/mL ·h in male rats and 5264 ng/mL ·h in female rats after oral administration of 18 mg/kg vicagrel. In the current study, M9-2 exposure in humans (20 mg) was 2840 ng/mL ·h. Thus, no further testing was required to evaluate its safety.

A previous study reported that M15-2 is the main active metabolite of vicagrel [38]. M15-2, which possesses a thioenol group, is unstable in vivo. Thus, a derivatization reagent, namely, MPB, was used to pretreat blood to trap M15-2 to MP-M15-2 for further quantitation. However, comparing the AUC of M9-2/MP-M15-1/MP-M15-2 with that of plasma not pretreated with MPB is inaccurate, while we would not obtain any information about the active metabolite M15-2 if we had identified the metabolites in the unpretreated plasma sample. Moreover, the results of the quantitative study and radiochromatograms of M3 showed remarkable differences. In the quantitative study, M3 was found to be a major drug-related component, but it was almost undetectable in the AUC0–8 h pooled plasma radiochromatogram. The inconsistency was due to different plasma sample types. The plasma sample used for quantitative analysis of M3 was from blood pretreated with esterase inhibitor, whereas that for radiochromatograms was from blood pretreated with MPB rather than esterase inhibitor.

Recently, our laboratory reported the effects of CYP2C19 genetic polymorphisms on vicagrel and clopidogrel [7]. The results demonstrated that CYP2C19 genetic polymorphisms have a considerable effect on clopidogrel in both PK and PD but only a slight influence on vicagrel. We attributed this difference to their different mechanisms in their first step in producing 2-oxo-clopidogrel from clopidogrel or vicagrel. For clopidogrel, the first step is mediated by CYP2C19, whereas carboxylesterase 2 and arylacetamide deacetylase play an important role for vicagrel. Such differences imply that CYP2C19 genetic polymorphisms have a lesser influence on vicagrel than on clopidogrel. The subsequent metabolism from 2-oxo-clopidogrel in vicagrel is the same as that in clopidogrel. Therefore, the possibility of DDI mediated by the CYP2C19 polymorphism in vicagrel is markedly lower than that in clopidogrel. A prior study investigated the DDI between vicagrel and aspirin in rodents [39]. The results showed that vicagrel can increase platelet responses to aspirin and enhance the inhibition of thrombus formation by aspirin, whereas aspirin does not enhance thrombus inhibition by vicagrel.

Vicagrel, as a promising P2Y12 receptor inhibitor, addresses clopidogrel resistance in the clinical use of clopidogrel by changing the metabolic pathway as described in this article. The results of preclinical and phase I studies have demonstrated that vicagrel has a definite antiplatelet aggregation effect, rapid onset, and good safety.

In conclusion, this study expands our understanding of the absorption, metabolism, and excretion of vicagrel in humans. Mass balance analysis revealed that 96.71% of the dose was recovered in urine (68.03%) and feces (28.67%) after oral administration. In plasma pretreated with MPB, M9-2 was the predominant drug-related component. M4 and M3-1 were the most abundant metabolites in urine, accounting for 16.25% and 12.59% of the dose, respectively; meanwhile M21 was the major metabolite in feces, accounting for 6.81% of the dose.