HangAmDan-B1 (HAD-B1) is a blended herbal extract of four critical herbs (Panax notoginseng Burk radix, Panax ginseng C.A. Meyer, Cordyceps militaris L., and Boswellia carterii Birdwood) that has been modified from the HangAmDan-B (HAD-B). It has been used as an anticancer herbal medicine at the East-West Cancer Center (EWCC; Dunsan Korean Medicine Hospital, Daejeon University, Korea) (Kang et al. 2018; Kang et al. 2019).

The chemical constituents of each ingredient in HAD-B1 are complex. Saponins are one of the main active ingredients of both Panax notoginseng and Panax ginseng. Some of the same saponins are contained in both notoginseng and ginseng, such as ginsenoside Rb1, Rg1, Re, Rc, and Rd. Among these, ginsenoside Rg1 and Rb1 are the most abundant, and they are included in more than 20% of the total saponins (Wei et al. 2011; Kim 2012; Lee et al. 2015; Liu et al. 2020b). Some saponins are unique to notoginseng; among these, notoginsenoside R1 is the most abundant (Wei et al. 2011; Kim 2012; Liu et al. 2020b). Various bioactive compounds are found in Cordyceps militaris, and they are divided into amino acids, fatty acids, and nucleosides (Yang et al. 2007; Hur 2008). Nucleosides and their metabolic compounds play important roles in biochemical processes related to several diseases and metabolic disorders (Yang et al. 2007). Among these, cordycepin, a derivative of the nucleoside adenosine, and adenosine are the most abundant nucleosides isolated. They participate in various molecular processes in cells (Tuli et al. 2013; Huang et al. 2014; Ballesteros-Yáñez et al. 2018). For Boswellia carterii, the isolated active constituents in the extracted frankincense oleogum resin are composed of boswellic acids (BAs) (Badria et al. 2003).

The 3-dimensional HPLC analysis of the HAD-B1 extract showed six major components, including notoginsenoside R1, ginsenoside Rg1, ginsenoside Rb1, cordycepin, α-boswellic acid, and β-boswellic acid (Kang et al. 2019). Each of the ingredients and detected major components of HAD-B1 are listed in Table 1 (in-house data).

Table 1 . Ingredients and major components of HAD-B1.

Ingredients of the HAD-B1 herbal mixture (Kang et al. 2019)
Scientific name	Amount (relative amounta)
Panax notoginseng Burk (radix)	25.2 g (32.3%)
Panax ginseng C.A. Meyer (radix)	19.2 g (24.6%)
Cordyceps militaris L.	19.2 g (24.6%)
Boswellia carterii Birdwood	14.4 g (18.5%)
Total amount	78.0 g (100%)
Major components of the HAD-B1 extract pill (in-house data)
Compound	Amount (relative amountb)
Notoginsenoside R1	1.166 mg (0.181%)
Ginsenoside Rg1	0.630 mg (0.0977%)
Ginsenoside Rb1	1.348 mg (0.209%)
Cordycepin	0.376 mg (0.0582%)
α-Boswellic acid	0.0353 mg (0.00548%)
β-Boswellic acid	0.0660 mg (0.0102%)
Total amount (one pill)	645 mg (100%)

aRelative amount=Amount of each ingredientgAmount of the total HAD−B1 ingredientsg×100%.

bRelative amount=Amount of each ingredientmgAmount of one HAD−B1 pillmg×100%.

Until now, numerous studies have been conducted to investigate the therapeutic effects of HAD-B and HAD-B1 (Bang et al. 2011; Choi et al. 2011; Kim et al. 2012; Li et al. 2015; Park et al. 2017; Kang et al. 2018; Kang et al. 2019). These studies have shown that HAD-B and HAD-B1 have anticancer effects in various cancer cells, such as human non-small-cell lung carcinoma cell lines, H460 and A549, human H1975 lung cancer cells, and Lewis lung carcinoma cells, etc. (Choi et al. 2011; Park et al. 2017; Kang et al. 2018; Kang et al. 2019). Numbers of action mechanisms for the anticancer effects of HAD-B and HAD-B1 have been reported, such as Her2 downregulation in NIH: OVCAR-3 human ovarian cancer cells, galectin-3-independent downregulation of GABABR1, downregulation of STAT3 in A549CR cells, etc. For detail, HAD-B1 exhibited an anti-cancer effect against lung-cancer cells through downregulation effects on STAT3 in A549CR cells, consequently leading to downregulation of Mcl-1 gene expression in cancer cells (Kang et al. 2018), which is known to induce caspase-dependent cell death (Aoki et al. 2003). In addition, another study showed that the combined treatment of HAD-B1 with afatinib on H1975 (L858R/T790M double mutation) lung cancer cells significantly induced early apoptosis and cell cycle arrest of the cells compared with the afatinib control group (Kang et al. 2019). In the study, downregulation of pERK1/2 and upregulation of p16 in the cells were reported. Furthermore, the therapeutic effects of each major component (for instance, ginseng and ginsenosides, boswellic acids, cordycepin, etc.) have been studied thoroughly (Kim and Park 2011; Kim et al. 2013; Tuli et al. 2014; Iram et al. 2017; He et al. 2020; Liu et al. 2020a).

Pharmacokinetics study

Animal experiments were managed under the protocol approved by the Animal Ethics Committee of Chungnam National University in Daejeon, South Korea (NO. 2019012A-CNU-193, approved on December 27th, 2019). All procedures were conducted in accordance with the assurance statement and guidelines in the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals. Male Sprague-Dawley rats (6 weeks, 190-220 g) were purchased from the Koatech Experimental Animal Center (Pyeongtaek, Korea). The animals were maintained under specific pathogen-free conditions in a controlled environment (temperature, 22 ± 2℃; humidity, 50 ± 10%). Rats were provided free access to food and water and fasted overnight before the PK studies.

Twenty-four rats were weighed and randomly divided into two groups (n = 8/group), and each received HAD-B1 extracted powder at a dose of either 600 or 1000 mg/kg. HAD-B1 was provided by the Kyungbang Pharmaceutical Company (Incheon, Korea). Before the PK studies, HAD-B1 was thoroughly dissolved in distilled water to make a final concentration of 150 or 250 mg/mL, respectively. The solution was then orally administered to rats via a gavage needle.

Blood samples (approximate 0.2 mL) were collected from the tail vein pre-dose (0 h) and post-dose at pre-determined time points (0.083, 0.25, 0.5, 1, 2, 4, 8, 24, and 32 h). The whole blood samples were stayed at 25℃ for 30 min then centrifuged at 6000 rpm for 10 min at 4℃ to remove the clot. After centrifugation, the supernatant (serum) was immediately separated and stored at −80℃ until analysis.

Prediction of ability to detect major components of HAD-B1

An extensive literature search for studies related to the determination of notoginsenoside R1, ginsenoside Rg1 and Rb1, cordycepin, and α- and β-boswellic acids following oral administration of these components in rats were performed. Information about administered dosage (Dose_prev), maximum concentration (C_{max_prev}), and a lower limit of quantification (LLOQ) of the analysis system was extracted from that research. We assumed that all these compounds showed linear PK profiles. C_max is proportional to the dose administered. Accordingly, the predicted C_max for each component (C_{max_pred}) after administration of 600 mg/kg HAD-B1 was predicted based on C_{max_prev} as described in the below equation.

$$C_{\max \_pred}=\frac{Dos{e}_{real}}{Dos{e}_{prev}}\times C_{\max \_prev}$$

where Dose_real was the amount of the component con-tained in the HAD-B1 administered dose in this study.

If the C_{max_pred} was not higher than the reported LLOQ, the possibility to detect that compound in this study would be “Low.” Otherwise, the possibility would be “High”. That prediction is listed in Table 2, showing that only notoginsenoside R1 and ginsenoside Rg1 and Rb1 were supposed to be detected following oral administra-tion of HAD-B1 600 mg/kg in rats.

Table 2 . Prediction of ability to detect the major compounds of HAD-B1 extracted powder.

Compound	Doseprev (mg/kg)	Cmax_prev (ng/mL)	LLOQ (ng/mL)	Reference	Dosereal(mg/kg)	Cmax_pred (ng/mL)	Ability to detect
Notoginsenoside R1	22.1	2940	3.03	(Li et al. 2007)	1.085	244	High
Ginsenoside Rg1	79.0	6420	4.00	(Li et al. 2007)	0.586	47.6	High
	1.44	225	10	(Han et al. 2018)		91.6
Ginsenoside Rb1	104.0	5080	2.77	(Li et al. 2007)	1.254	61.3	High
	7.74	210	10	(Han et al. 2018)		34.0
Cordycepin	80	Undetectable	2	(Lee et al. 2019)	0.349	Undetectable	Low
α-Boswellic acid	13.44	511	5	(Hüsch et al. 2013)	0.033	1.25	Low
β-Boswellic acid	28.13	994	5	(Hüsch et al. 2013)	0.061	2.16	Low

C_max, maximum concentration; Dose_prev, administered dose extracted from previous studies; C_{max_prev} and C_{max_pred}, C_max collected from previous studies and predicted following administration of HAD-B1 in this study, respectively; Dose_real, the real administered dose of the component included in HAD-B1 extract powder at a dose of 600 mg/kg..

Preparation of standard samples and real samples

Since notoginsenoside R1 and ginsenoside Rg1 and Rb1 were predicted to be detected following oral administration of HAD-B1 600 mg/kg in rats, investigating PK profiles of these compounds was the initial aim of this study. However, notoginsenoside R1 was finally excluded from this study due to our mistakes in the step for the preparation of the stock solution. In addition, due to the availability of cordy-cepin, adenosine, and ginsenoside Rc (standard samples) in our laboratory, PK profiles of these compounds were addi-tionally performed. Finally, the objective of this study was amended to investigate the PK of ginsenoside Rg1, Rb1, and Rc, cordycepin, and adenosine in rat serum after oral ad-ministration of HAD-B1 (600 or 1000 mg/kg) in rats.

Analytes and gliclazide (internal standard (IS) for the analysis) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). HPLC-grade methanol was obtained from Thermo Fisher Scientific (Waltham, MA, USA). Water was obtained using an option-Q purification system (Elga Ltd., High Wycombe Bucks, UK). Other chemicals were of the analytical grades.

An approximate amount of each analyte was accurately weighed and then dissolved in methanol (except for ade-nosine dissolved in methanol containing 0.1% DMSO) to make a standard stock solution at a 1 mg/mL strength. These solutions were diluted at appropriate ratios in methanol to prepare standard solutions at a concentra-tion range of 97.5-100,000 ng/mL. To prepare standard samples, an aliquot of 0.6 μL of each standard (total of 3.0 μL) was mixed with 27 μL of blank serum to yield a series of samples with final concentrations ranging from 1.95 to 2000 ng/mL. These samples were then treated the same as the real serum samples from the PK experiment.

Analytes were extracted from the serum samples using a simple protein precipitation method. 270 μL of IS solu-tion were mixed with 30 μL of a serum sample for ex-traction using a vortex mixer (Vortex-Genie 2 Mixer, Scien-tific Industries, Inc., Bohemia, NY, USA). After centrifugation at 15,000 rpm for 10 min at 4℃, 5 μL of the supernatant was injected into the LC-MS/MS system for analysis.

Analytical conditions for determination of analyte concentrations

The plasma concentration of analytes was determined using an HPLC system (1260 series HPLC; Agilent, Santa Clara, CA, USA) coupled with a triple quadrupole mass spectrometer system (API 6500; SCIEX, Concord, Ontario, Canada). Analytes were separated using a Zorbax Eclipse Plus C18 column (50 mm × 2.1 mm, 1.8 μm; Agilent, Santa Clara, CA, USA) and an isocratic flow of mobile phase, which consists of distilled water and methanol at a ratio of 10:90 (v/v). The column was maintained at a temperature of 24 ± 0.5℃. The flow rate was 0.3 mL/min, and the total run time was 2 minutes. The injection volume was 5 μL.

The mass detection was operated in the positive electro-spray ionization mode with multiple reaction monitoring transitions. The conditions of the mass spectrometry for the detection of the analytes and IS were as described in Table 3.

Table 3 . Analytical conditions of the MS/MS system for the detection of analytes.

Compound	Cordycepin	Adenosine	Rg1	Rb1	Rc	Gliclazide (IS)
MW	251.24	267.24	801	1109	1079	323.41
Q1 → Q3 transition (m/z)	252.1 → 136.0	268.1 → 136.0	823.3 → 643.4	1131.3 → 365.3	1101.3 → 334.9	324.1 → 127.1
DP (V)	51	106	41	271	296	61
CE (eV)	25	23	51	73	75	25
CXP (eV)	18	14	42	28	12	14
Retention time (min)	0.62	0.62	0.59	0.61	0.61	0.65
LLOQ (ng/mL)	31.3	31.3	1.95	7.8	7.8

MW, molecular weight; MRM, multiple reaction monitoring; DP, declustering potential; CE, collision energy; CXP, collision cell exit potential..

Non-compartment pharmacokinetic analysis

The PK parameters were calculated using Phoenix Win-Nonlin (version 8.2.0.4383, Certara L.P, Princeton, NJ, USA). For the PK parameters, C_max and time to reach C_max (T_max) were obtained directly from observations. The area under the curve (AUC) from zero to the last measurement (AUC_0-t) was calculated using a linear log trapezoidal method. AUC from zero to infinity (AUC_0-∞) was the sum of AUC_0-t and extrapolated AUC from the last time point to infinity (AUC_0-∞ = AUC_0-t + C _last/k_el), where C_last is the last plasma concentration and k_el is the terminal rate constant deriving from the slope of linear regression log-transformed of plasma concentration. The apparent clearance (CL/F) was derived by dividing the administered dose by AUC_0-∞ CL/F = dose/AUC_0-∞), and the apparent volume of distribution (V/F) was calculated by dividing CL/F by the terminal rate constant (V/F = CL/F/k_el), where F is bioavailability.

The concentrations of adenosine and cordycepin in rat serum after oral administration of HAD-B1 at doses of 600 or 1000 mg/kg in rats were undetectable in our analysis system. Only ginsenoside Rg1, Rb1, and Rc were detected. PK profiles of these detected compounds are presented in Fig. 1. PK parameters obtained by NCA for each analyte are listed in Table 4.

Table 4 . The main PK parameters of Ginsenoside Rg1, Rb1, and Rc in rat serum following oral administration of HAD-B1 extract powder at a dose of either 600 mg or 1000 mg/kg in rats (n = 8/group).

Parameter (unit)	600 mg/kg				1000 mg/kg
	Rg1	Rb1	Rc		Rg1	Rb1	Rc
Cmax (ng/mL)	21.3 (11.2)	102 (32.2)	38.7 (11.3)		31.1 (12.5)	136 (45.9)	45.7 (13.2)
Tmax (h)	0.385 (0.285)	4.00 (0)	4.00 (0)		0.333 (0.299)	4.25 (1.67)	4.031 (2.07)
t1/2 (h)	4.39 (1.83)	18.0 (3.21)	28.6 (14.6)		9.36 (18.6)	19.6 (3.52)	27.7 (4.48)
AUC0-t (ng/mL*h)	31.6 (9.55)	1947 (572)	809 (174)		73.6 (26.8)	2399 (507)	924 (194)
AUC0-∞ (ng/mL*h)	57.9 (10.4)	2798 (757)	1530 (524)		240 (461)	3436 (877)	1685 (460)
CL/F (L/kg/h)	10.4 (1.72)	0.482 (0.145)			11.7 (5.35)	0.656 (0.229)
Vz/F (L/kg)	65.7 (32.8)	12.6 (4.68)			52.4 (10.7)	18.0 (4.59)

Value of PK parameters is presented as mean (standard deviation); amounts of Rg1 and Rb1 are 0.586 and 1.254 mg in 600 mg HAD-B1 extract powder and 0.977 and 2.090 mg in 1000 mg HAD-B1 extract powder; CL/F and Vz/F for Rc were not reported due to amount of Rc in the HAD-B1 powder was unknown..

Figure 1. The serum concentration-time curves of Rg1, Rb1, and Rc after oral administration of HAD-B1 extract powder at the dose of either 600 or 1000 mg in rats. Circles, individual concentrations; solid lines, the average concentration of the group (n = 8/group).

As seen in Fig. 1 and Table 4, after oral administration of HAD-B1 extract powder in rats, Ginsenoside Rg1 was absorbed rapidly into the systemic circulation. The peak concentration was reached within 0.5 h after the dose administration. Rg1 was eliminated with an average t_1/2 of 4.39 h following the dose of 600 mg/kg and maintained longer with an average t_1/2 of 9.36 h following the dose of 1000 mg/kg. These observed parameters were similar to those reported in previous studies (Li et al. 2007; Zhou et al. 2015; Han et al. 2018). For ginsenoside Rb1 and Rc, these compounds reached the peak concentration at around 4 h following the doses. The calculated average t_1/2 was approximately 18-20 h for Rb1 and 28-29 h for Rc, which were also consistent with those reported previously (Li et al. 2007; Chu et al. 2013; Zhou et al. 2015; Han et al. 2018).

HAD-B1 is a blended herbal extract that has been modified from the HAD-B. Until now, the therapeutic effects of HAD-B and HAD-B1 have been investigated in numerous studies (Bang et al. 2011; Choi et al. 2011; Kim et al. 2012; Li et al. 2015; Park et al. 2017; Kang et al. 2018; Kang et al. 2019). Furthermore, the therapeutic effects of each major component have been studied thoroughly (Kim and Park 2011; Kim et al. 2013; Tuli et al. 2014; Iram et al. 2017; He et al. 2020; Liu et al. 2020a). However, when performing a literature search, we found that PK profiles of major components of HAD-B1 varied considerably between studies. For example, average t1/2 values for Rg1 were from 5.0-6.0 h (Li et al. 2007; Zhou et al. 2015) to 42.0 h (So et al. 1996). The time to reach peak concentration of Rb1 was reported to be from 0.83 h (Li et al. 2007) to 8 h (Zhou et al. 2015; Han et al. 2018). This considerable variability suggested that the major components of HAD-B1 might follow non-linearity PK profiles as the administered dose in the above studies was not the same. Interactions between complex components of each herbal medicine also could be one factor that affects non-linear PK profiles. Consequently, the PK profile of the same compound in each herb could be diverse, even at an equal dose. For the above reasons, even if therapeutic effects have been studied thoroughly, the determination of full PK profiles of HAD-B1 is imperative.

In the present study, we performed a fast simultaneous determination of three (ginsenoside Rg1, Rb1, and Rc) of the major components of the HAD-B1 extract after oral administration of HAD-B1 600 or 1000 mg/kg in rats using an HPLC-MS/MS system. Accordingly, the PK parameters of these components were successfully reported, showing a non-linear kinetics of Rg1, Rb1, and Rc. In detail, when the dose of HAD-B1 increased approximately 1.67 folds (from 600 to 1000 mg/kg), the AUC_0-t and C_max of Rg1, Rb1, and Rc increased by 2.33- and 1.46-fold (Rg1), 1.23- and 1.33-fold (Rb1), and 1.14- and 1.18-fold (Rc), respectively. In the cases of Rb1 and Rc, the increases in the exposures of the drugs are less than dose proportional. The elimination half-life of the compound was not significantly different between the two doses. They were 18.0 h and 19.6 h for Rb1; 28.6 h and 27.7 h for Rc, after the dose of 600 and 1000 mg HAD-B1, respectively. Meanwhile, the apparent clearance (CL/F) and volume of distribution (Vd/F) of the compounds were significantly increased (Table 4). These results suggest that the non-linear PK profiles of Rb1 and Rc might come from their reduced amount of absorption (F) (due to their limited solubility or limited permeability). This assumption was also supported by the fact that the time to reach concentrations of both these two compounds was not significantly different between the two doses (approximately 4h for Rb1 and Rc). In the case of Rg1, the increase in the drug exposure was higher than dose-proportional. As seen, the elimination half-life of Rb1 increased significantly from around 4.4 h to 9.4 h. This suggested that the non-linear PK of Rg1 might come from the prolongation in the elimination process of the compound. Further studies need to be performed to fully understand the PK of these compounds. This analysis method could be applied to clinical trials after the administration of HAD-1.

Adenosine and cordycepin were not detected in rat serum after the administration of HAD-B1 at the doses of 600 or 1000 mg/kg. These results were similar to the previous report. (Lee et al. 2019). In this study, cordycepin was not detected in plasma at any time point, even at the high dose of 80 mg/kg. Instead, its metabolite, 3’-deoxyinosine, was found systemically at a high concentration. Importantly, 3’-deoxyinosine can be converted to cordycepin 5’-triphosphate, which shows therapeutic effects (Lee et al. 2019). Therefore, elucidation of PK profiles of 3’-deoxyinosine and cordycepin 5’-triphosphate, instead of cordycepin, after oral administration of HAD-B1 would be necessary for the next study.

The non-linear PK profiles of ginsenoside Rg1, Rb1, and Rc after the administration of 600 or 1000 mg HAD-B1 were confirmed in rats by this study research. These results suggested that the non-linear PK profiles of the components might also occur in humans after the administration of HAD-B1. Therefore, in the cases where the adjustment for HAD-B1 dosage is necessary, one should consider not using the simple dose proportional calculation. For the dosing regiments purpose, PK profiles of major components of HAD-B1 in rats can be applied to extrapolate from rats into humans using allometric scaling methods. In addition, the PK data could be used to develop a population PK model, which performs a compartmental analysis with covariate effects to confirm the explanation for the non-linear PK profiles of these compounds. These are the objectives of our next study of HAD-B1.

There were some limitations in our present study. First, the content of Rc in the extract powder of HAD-B1 was not measured although the compound was detected in rat plasma samples. Second, due to our mistake in the preparation step for the stock solution of notoginsenoside R1, the PK of this compound was not determined, although it is one of the main components of HAD-B1 and is predicted to be detectable in rat serum before the study. Third, PKs of α- and β-boswellic acids were not determined. Furthermore, we didn’t consider the active metabolite of major components associated with therapeutic effects. For the next research project, the content of Rc as well as all the major components in the extract powder of HAD-B1 need to be measured. An analysis method must be developed for the determination of PK profiles of the remaining major components (notoginsenoside R1, α- and β-boswellic), as well as for the major active metabolites of each major component (e.g., 3’-deoxyinosine, cordycepin 5’-triphosphate) included in HAD-B1.

No potential conflict of interest relevant to this article was reported.

This research was supported by the National Research Foundation of Korea funded by the Korean Government (NRF-2018R1C1B6007898 and 2018R1C1B5085278) and the Institute of Information & communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (No.2020-0-01441 and 2022-00155857, Artificial Intelligence Convergence Research Center (Chungnam National University)).