Development and validation of a liquid chromatography coupled with tandem mass spectrometry method for determining total and unbound pamiparib in human plasma and brain tumors

Abstract

Pamiparib (BGB-290) is an orally bioavailable, small molecule inhibitor of poly (ADP-ribose) polymerase 1 (PARP1) and PARP2. A reversed-phase LC with tandem mass spectrometry method was developed and fully validated for determining total and unbound pamiparib concentrations in human plasma and brain tumor tissue. Plasma and tissue homogenate samples were prepared by methanol protein precipitation. Pamiparib and the internal standard [¹³C₂,¹⁵N₂]pamiparib were separated on a Waters BEH C₁₈ (50 × 2.1 mm, 1.7 μm) column, with a gradient elution consisting of mobile phases A (0.1% formic acid in water) and B (0.1% formic acid in acetonitrile) at a flow rate of 0.25 mL/min. The analytes were monitored with multiple reaction monitoring mode under positive electrospray ionization. The method was fully validated for specificity, linearity, accuracy and precision, matrix effect and recovery, and short- and long-term stability. The lower limit of quantitation was 0.5 nM of pamiparib in plasma or tissue homogenate. The calibration curve was linear over the pamiparib concentration range of 0.5–1000 nM in plasma. The intra- and inter-day precision and accuracy were within the generally accepted criteria for bioanalytical method. Pamiparib was stable in plasma at −80°C for at least 6 months. The method was successfully applied to assess the plasma and tumor pharmacokinetics of total and unbound pamiparib in patients with glioma.

1 |. INTRODUCTION

Poly (ADP-ribose) polymerases (PARPs) are a family of enzymes that catalyze the transfer of ADP-ribose to target proteins. PARPs play a fundamental role in DNA repair and many other cellular processes including transcription and modulation of chromatin structure (Morales et al., 2014). In light of the central role of PARPs in nucleotide excision repair and base excision repair that enable repair of DNA damages caused by radiation, alkylating agents, or chemotherapeutic drugs, PARP inhibition may sensitize cancer cells to radiation or DNA damaging agents (Jannetti et al., 2020). Certain tumors defective in homologous recombination pathway (e.g., tumors with BRCA1/2 mutations) are more dependent on PARP-mediated DNA repair for survival and thus are particularly sensitive to PARP inhibitor-rendered “synthetic lethality” (Jannetti et al., 2020; Morales et al., 2014). Several orally bioavailable, small molecule PARP inhibitors, such as olaparib (AstraZeneca; Tutt et al., 2010), rucaparib (Clovis; Thomas et al., 2007), and talazoparib (Pfizer; Wang et al., 2016) have been approved by the United States Food and Drug Administration (FDA) for the treatment of ovarian or breast cancer harboring BRCA1/2 mutations. In addition, PARP inhibitors act as chemo- or radio-sensitizers in combination with DNA-damaging agents or radiation therapy for the treatment of a broad range of malignancies (Jannetti et al., 2020; Morales et al., 2014).

Pamiparib (BGB-290) is a potent, selective, orally bioavailable, small molecule PARP1/2 inhibitor. Similar to other PARP inhibitors (e.g., olaparib), pamiparib has demonstrated single-agent antiproliferative activity against a variety of in vitro and in vivo tumor models with known BRCA1/2 mutations or other homologous recombination deficiency (Xiong et al., 2020). In particular, pamiparib has shown strong synergistic antitumor activity in combination with temozolomide, a DNA alkylating agent, in intracranial brain tumor xenografts models, and additional preclinical data indicate a good penetration of pamiparib across the rodent blood–brain barrier (Xiong et al., 2020). These data collectively support further clinical development of pamiparib for the treatment of brain tumors. Several clinical trials are currently ongoing to evaluate pamiparib with or without temozolomide or radiation in newly diagnosed or recurrent gliomas (NCT03914742, NCT03749187, NCT03150862, and NCT04614909).

The knowledge of the pharmacokinetics of a drug in human plasma and brain (or brain tumor) is critical to the selection of appropriate drug candidates and design of optimal dosing regimens for brain cancer drug development and therapy. Based on the free drug theory, unbound (free) drug concentration drives pharmacological activity. Thus, it is important to develop a sensitive and reliable method for the determination of not only total drug but also unbound drug concentrations in human plasma and brain (or brain tumors). In this study, we developed and fully validated an LC coupled with tandem mass spectrometry (LC–MS/MS) method to determine pamiparib concentrations in human plasma and brain tumor tissue. In addition, we optimized an equilibrium dialysis method to determine the fraction unbound of pamiparib in human plasma and brain tumor tissue, based on which the unbound drug concentration could be determined. Our validated method was successfully applied in a Phase 0 neuro-oncology clinical trial to evaluate the plasma and brain tumor pharmacokinetics of total and unbound pamiparib in adult patients with recurrent IDH-mutant gliomas.

2 |. EXPERIMENTAL

2.1 |. Chemicals and reagents

Pamiparib (BGB-290) and its stable isotope-labeled internal standard (IS), [¹³C₂,¹⁵N₂]pamiparib, were provided by BeiGene (Cambridge, MA). All other chemicals and reagents (LC–MS grade) were purchased from Fisher Scientific (Waltham, MA). Water was processed with a US Filter PureLab Plus UV/UF system (Siemens, Detroit, MI) and used in all aqueous solutions. Drug-free pooled human plasma and plasma from six individual healthy donors (with sodium–ethylenediaminetetraacetic acid anticoagulant) were purchased from Innovative Research Inc (Novi, MI).

2.2 |. Chromatographic and mass spectrometric conditions

2.2.1 |. Instrumentation

All LC–MS/MS analyses were carried out on a Waters Xevo TQ-XS LC–MS/MS mass spectrometer coupled with a Waters ACQUITY H-class UPLC system (Milford, MA). The instrument operation and data acquisition were controlled by MassLynx 4.2 software, and data processing and quantification were carried out with TargetLynx XS software.

2.2.2 |. Liquid chromatography

Chromatographic separation was achieved on a Waters BEH C₁₈ UPLC column (2.1 × 50 mm, 1.7 μm) with a gradient elution consisting of mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) at a flow rate of 0.25 mL/min. The elution gradient program was as follows [shown as the time (min), (mobile phase B)]: 0, 10%; 0.5, 10%; 1.5, 100%; 3.1, 100%; 3.2, 10%; 5.5, 10%. The column was maintained at 35°C. A mixture of methanol and water (50/50, v/v) was used as needle wash to minimize carryover.

2.2.3 |. Mass spectrometry

The Waters Xevo TQ-XS mass spectrometer was operated in electrospray positive ionization mode using multiple reaction monitoring mode. The capillary voltage was set at 3.00 kV, and desolvation temperature was set at 500°C. Gas flow parameters were optimized to 1000 L/h for desolvation, 150 L/h for cone, and 7.0 bar for nebulizer. The dwell time was set at 50 ms. Cone voltage and collision energy were optimized at 60 V and 38 eV for pamiparib, and 70 V and 42 eV for [¹³C₂,¹⁵N₂]pamiparib. The most sensitive MS transitions of m/z 299.0 → 133.0 and 303.0 → 134.9 were selected for monitoring pamiparib and [¹³C₂, ¹⁵N₂]pamiparib, respectively (Figure 1).

FIGURE 1.

Mass spectra of the parent ion (left) and product ion of (a) pamiparib and (b) the internal standard [¹³C₂,¹⁵N₂]pamiparib. The most sensitive mass transitions of m/z 299.0 → 133.0 and m/z 303.0 → 134.8 were selected for monitoring pamiparib and [¹³C₂,¹⁵N₂]pamiparib, respectively.

2.3 |. Sample preparation

2.3.1 |. Stock solutions, calibration standards, and quality control samples

Pamiparib and [¹³C₂,¹⁵N₂]pamiparib stock solutions were prepared in dimethyl sulfoxide (DMSO) at a final concentration of 10 mM and stored in brown glass vials at −80°C. Pamiparib working solution was freshly prepared by serial dilution of the stock solution with 50% acetonitrile in water on each day of analysis. [¹³C₂,¹⁵N₂]pamiparib working solution was prepared with 50% acetonitrile in water at the concentration of 500 nM. To determine pamiparib concentrations in plasma, tumor homogenates, or post-dialysis phosphate buffer solution (PBS), the calibration standards were prepared by spiking 2 μL of pamiparib working solution into 38 μL of blank human plasma or PBS to make the final concentrations at 0.5, 1, 2, 5, 10, 20, 50, 100, 200, 500, and 1000 nM. The quality control (QC) samples were prepared in human plasma at pamiparib concentrations of 0.5 nM (lower limit of quantitation [LLOQ]), 1.5 nM (low QC), 60 nM (medium QC), and 600 nM (high QC). All calibrator and QC samples were prepared freshly daily. Low and high QC samples for long-term and freeze–thaw stability were prepared as a batch and stored at −80°C.

2.3.2 |. Plasma and brain tumor tissue sample preparation

Frozen plasma or brain tumor tissue samples were thawed at room temperature. Tissue homogenate was prepared by adding three volumes of PBS (e.g., 300 μL PBS for 100 mg tissue) and homogenizing in a Precellys homogenizer at 2000g for two 10-s sessions (with a 10-s pause). Blank pooled human plasma was used to dilute plasma or tissue homogenate when necessary. Because total pamiparib concentrations in patient brain tumor tissue were high, tissue homogenate was diluted by 10-fold with human blank plasma to ensure the measured tissue homogenate concentrations fall within the calibration curve range (0.5–1000 nM). A calibration curve in plasma was prepared for determining pamiparib concentrations in both plasma and tissue samples. This was acceptable for three reasons: (1) human brain tissue was not readily available for preparing a calibration curve; (2) plasma was the predominant matrix in 10-fold diluted tissue homogenate; and (3) the stable isotope-labeled internal standard effectively corrected for matrix effect difference (if any).

An aliquot of 40 μL plasma or diluted tissue homogenate was transferred into a microcentrifuge tube, spiked with 4 μL of 500 nM [¹³C₂,¹⁵N₂]pamiparib, and incubated at room temperature for 5 min. The sample was then extracted by protein precipitation with 120 μL ice-cold methanol, followed by vortex mixing (10 s) and centrifugation (at 21,952g, 4°C, 10 min). The supernatant was transferred into an autosampler vial and 5 μL was injected into the LC–MS/MS system.

2.3.3 |. Equilibrium dialysis

The fraction unbound of pamiparib in plasma and brain tissue was determined by equilibrium dialysis on a 96-well Equilibrium DIALYZER with 10-kDa molecular weight cut-off regenerated cellulose membrane (Harvard Apparatus, Holliston, MA), as previously described with modifications (Li et al., 2006). Briefly, equilibrium dialysis was performed with 180 μL of plasma or tissue homogenate against an equal volume of PBS (pH 7.4) on a rotator (Harvard Apparatus) at 37°C. The optimal time to equilibrium was assessed with pooled human plasma at pamiparib concentration of 1 μM for dialysis of 2, 6, 16, and 24 h. At the end of dialysis, 180 μL methanol was added into the PBS compartment to reduce non-specific drug binding (if any), and the mixture was transferred into a microcentrifuge tube; the post-dialysis plasma or homogenate sample was directly transferred into a microcentrifuge tube. An aliquot of 80 μL post-dialysis PBS mixture was extracted by adding 4 μL of 500 nM [¹³C₂,¹⁵N₂]pamiparib and 80 μL ice-cold methanol, followed by vortex mixing and centrifugation (at 21,952g at 4°C for 10 min). The supernatant was transferred into an autosampler vial and 5 μL was subjected to LC–MS/MS analysis to determine the post-dialysis PBS drug concentration (C_u).The post-dialysis plasma or tissue homogenate sample was processed and subjected to LC–MS/MS analysis, as described in Section 2.3.2, to determine the post-dialysis plasma (C_p) or tissue homogenate (C_hom) drug concentration.

Fraction unbound in plasma (f_u, plasma) and in tissue homogenate (f_u, hom) was calculated using Equations (1) and (2), respectively. Fraction unbound in original (undiluted) brain tissue was estimated based on f_{u, hom} and dilution factor (D_f) using Equation (3) (Wan et al., 2007). The total drug recovery was calculated as the ratio of the post-dialysis total drug concentration recovered from the PBS and plasma or homogenate compartments to the total drug concentration in the original plasma or homogenate samples.

$${\text{f}}_{\text{u},\text{ plasma}}=\frac{{\text{C}}_{\text{u}}}{{\text{C}}_{\text{p}}}$$ (1)

$${\text{f}}_{\text{u},\text{ hom}}=\frac{{\text{C}}_{\text{u}}}{{\text{C}}_{\text{hom}}}$$ (2)

$${\text{f}}_{\text{u},\text{ tissue}}=\frac{\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{${\text{D}}_{\text{f}}$}\right.}{\left(\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{${\text{f}}_{\text{u},\text{home}}-1$}\right.\right)+\raisebox{1ex}{$1$}\!\left/ \!\raisebox{-1ex}{${\text{D}}_{\text{f}}$}\right.}$$ (3)

2.4 |. LC–MS/MS method validation

The method was fully validated for the specificity, linearity, accuracy and precision, matrix effect and recovery, as well as stability of short-and long-term storage, according to the US FDA Guidance on Bioanalytical Method Validation (FDA Guidance on Bioanalytical Method Validation, n.d.).

Specificity was evaluated by assessing potential endogenous interfering peaks from blank matrix (i.e., blank plasma from six donors). The interfering peak area should be less than 10% of the peak area of pamiparib at the LLOQ (0.5 nM) and less than 5% of the peak area of the internal standard [¹³C₂,¹⁵N₂]pamiparib.

Linearity was evaluated at a pamiparib concentration range of 0.5–1000 nM in pooled human plasma. The calibration curve was established by fitting the calibrator concentrations versus pamiparib-to-internal standard peak area ratios using least-square linear regression with a weighing function of 1/x² (where x is the analyte concentration).

The intra- and inter-day accuracy and precision were assessed for the calibrator standards (each in duplicate) and QCs (including LLOQ, low, medium, and high QCs, each in quintuplicate) on 3 days. The accuracy was evaluated as the percentage of the determined concentration to nominal concentration. The intra- and inter-day precisions were evaluated by one-way analysis of variance using SPSS (IBM, Armonk, NY), as described previously (Bao et al., 2018; Wu et al., 2016). The analyte response at the lower limit of detection should have a signal-to-noise ratio of at least 5. The analyte response at the LLOQ should have a signal-to-noise ratio of at least 10, and has an accuracy of 80–120% and precision (i.e., coefficient variation) within 20%.

Matrix effect and recovery were evaluated with plasma samples from six individual donors, as described previously (Bao et al., 2018; Wu et al., 2016). Briefly, three sets of QC samples were prepared. Set 1 QC samples were prepared in 40 μL of individual donor human plasma at pamiparib concentrations of 1.5, 60, and 600 nM and [¹³C₂,¹⁵N₂]pamiparib at 50 nM. Set 1 samples were processed by adding 120 μL ice-cold methanol, followed by centrifugation (21,952g, 4°C, 10 min), and 5 μL of the supernatant was injected into the LC–MS/MS system. Set 2 and 3 QC samples were prepared by spiking the same amount of pamiparib and [¹³C₂,¹⁵N₂]pamiparib as those in Set 1 into 160 μL neat solvent (75% MeOH in water, for Set 2) and 160 μL blank matrix extract (i.e., post-precipitation supernatant of the corresponding individual donor blank plasma, for Set 3). Set 2 and 3 samples (5 μL each) were directly injected into the LC–MS/MS system. The matrix factor was estimated as pamiparib peak area ratio of the post-precipitation (Set 3) to neat solvent (Set 2). The recovery was estimated as pamiparib peak area ratio of plasma (Set 1) to the post-precipitation matrix (Set 3). The relative (or internal standard normalized) matrix factor and recovery were calculated based on pamiparib-to-[¹³C₂,¹⁵N₂]pamiparib peak area ratios.

The bench-top stability of pamiparib in 50% acetonitrile (at 1 and 100 μM) and in plasma (at 1.5 and 600 nM) was evaluated at room temperature for 1, 2, 4, and 6 h. The autosampler stability of pamiparib in post-extracted plasma samples (at 1.5 and 600 nM) was assessed in the autosampler (at 5°C) for up to 9 h. Freeze–thaw stability of pamiparib in plasma (at 1.5 and 600 nM) was assessed for three freeze–thaw cycles with at least 1-day storage at −80°C between each thawing. The long-term stability of pamiparib in plasma (at 1.5 and 600 nM) was assessed up to 6 months. All stability tests were run in triplicates.

2.5 |. Applications

The developed method was used to evaluate the plasma and tumor concentrations of total and unbound pamiparib in patients with recurrent IDH-mutant gliomas in a multi-center neuro-oncology clinical trial (ABTC 1801). The protocol was approved by the Institutional Review Board at all study sites. All patients provided a written informed consent. Eligible patients were treated with pamiparib orally twice a day (60 mg, b.i.d.). Paired blood and tumor tissue samples were collected from three patients in the surgical arm. Briefly, the patients were treated with pamiparib for 7–10 days before the surgery to reach the drug steady state. On the day of surgical resection of tumor, blood and tumor tissue (including gadolinium contrast–enhancing region and non-enhancing region as shown on magnetic resonance imaging) samples were collected at 4–8 h following the administration of the last dose. Plasma was separated from the whole blood by centrifugation (at 4°C, 1500g for 10 min), and plasma samples were stored at −80°C until analysis. Tumor samples were immediately rinsed with ice-cold PBS to remove residual blood, dried on tissue paper, and snap frozen in liquid nitrogen. All samples were stored at −80°C until analysis.

Plasma and brain tumor tissue samples were prepared, as described in Section 2.3.2. Pamiparib total concentrations in plasma and tumor tissue homogenate samples were determined using the validated LC–MS/MS method. The unbound fraction of pamiparib in plasma and tumor tissue was determined as described in Section 2.3.3. Unbound pamiparib concentration in plasma or tumor was calculated by multiplying the total drug concentration and respective fraction unbound.

3 |. RESULTS AND DISCUSSION

3.1 |. Method development

Pamiparib is soluble in DMSO (59 mg/mL) and ethanol (49 mg/mL). Thus, the stock solution of pamiparib and [¹³C₂,¹⁵N₂]pamiparib was prepared in DMSO at the concentration of 10 mM. Pamiparib is a weak base with a non-conjugated tertiary amine in the pentacyclic tetrahydrotetraazacycloheptafluorenones moiety, which is protonated in a relatively acidic pH, for example, solvents acidified with 0.1% formic acid. The MS settings of pamiparib and [¹³C₂,¹⁵N₂]pamiparib were optimized using the IntelliStart function of Waters mass spectrometer controlled by MassLynx version 4.2. The sensitive and specific mass transitions of m/z 299.0 → 133.0 and m/z 303.0 → 134.8 were selected to monitor pamiparib and [¹³C₂,¹⁵N₂]pamiparib, respectively (Figure 1).

Pamiparib was well resolved by reversed-phase LC on a Waters AQUITY BEH C₁₈ column with a gradient elution consisting of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). Pamiparib was eluted at 2.34 min, with a sharp symmetrical peak (Figure 2).

FIGURE 2.

Extracted ion chromatography monitored at m/z 299.0 → 133.0 (pamiparib) and m/z 303.0 → 134.8 ([¹³C₂,¹⁵N₂]pamiparib) for (a) blank human plasma; (b) human plasma spiked with pamiparib at the LLOQ (0.5 nM); and (c) a patient plasma collected at 2 h following oral administration of pamiparib (60 mg), which was diluted 20-fold for analysis.

3.2 |. Method validation

Full method validation demonstrated the specificity, sensitivity, reliability, and reproducibility of the developed method. The selectivity was shown by symmetrical resolution of the chromatographic peaks, with no apparent interference in six different donors of plasma. In addition, in analyses of patient plasma and tumor samples, no endogenous interferences were observed. Figure 2 shows the representative extracted ion chromatograms of blank human plasma, plasma spiked with pamiparib at the LLOQ, and patient plasma collected at 2 h after the oral administration of pamiparib (60 mg). Their corresponding IS signal is at the bottom.

Linear calibration curves were constructed at a pamiparib concentration range of 0.5–1000 nM in human plasma, with a linear correlation coefficient (R²) of over 0.99 in all analytical runs. The LLOQ was established at 0.5 nM of pamiparib in human plasma. For all calibrators, the average accuracy in terms of the percentage of the determined concentration to nominal concentrations ranged from 94.3% to 102.4% and the intra- and inter-day precision (variation) was less than 10.2% (Table 1). For all QC samples (including LLOQ; low, medium, and high QCs; and 50-fold dilution QC), the average accuracy ranged from 93.3% to 103.9%, and the intra- and inter-day precision (variation) was less than 10.3% (Table 2).

TABLE 1.

Intra- and inter-day precision and accuracy of pamiparib calibrator standards in human plasma^a and PBS.

Nominal concentration (nM)	Determined concentration (nM)	Average accuracy (%)	Intra-day precision (%)	Inter-day precision (%)
In plasma
0.5	0.5 ± 0.0	101.0	4.0	2.1
1	1.0 ± 0.1	102.0	7.3	–b
2	1.9 ± 0.1	94.3	2.2	7.6
5	5.1 ± 0.5	102.4	5.3	10.2
10	10.0 ± 0.3	100.3	2.4	1.8
20	19.8 ± 0.7	98.8	0.7	3.5
50	49.5 ± 1.5	98.9	3.5	– b
100	101.7 ± 2.9	101.7	2.9	0.0
200	204.2 ± 11.2	102.1	3.6	4.7
500	500.6 ± 21.5	100.1	4.0	1.8
1000	1005.9 ± 34.5	100.6	2.8	2.2
In PBS
0.5	0.5 ± 0.0	99.3	2.0	0.6
1	1.0 ± 0.0	100.2	2.9	2.9
2	2.0 ± 0.0	101.9	0.7	2.1
5	5.1 ± 0.1	101.4	1.6	1.5
10	10.1 ± 0.2	101.4	1.8	–b
20	20.4 ± 0.4	102.0	1.4	1.8
50	50.3 ± 1.4	100.6	2.8	–b
100	101.9 ± 1.0	101.9	1.2	–b
200	196.9 ± 1.7	98.4	0.8	0.4
500	489.0 ± 18.2	97.8	3.7	0.7
1000	976.0 ± 18.9	97.6	2.1	–b
2000	1963.8 ± 50.1	98.2	1.7	2.2

^aEach calibrator standard was evaluated in duplicate on 3 different days.

^bNo additional variation was observed as a result of performing assay on different days.

TABLE 2.

Intra- and inter-day precision and accuracy of pamiparib QC samples in human plasma^a and PBS.

Nominal concentration (nM)	Determined concentration (nM)	Average accuracy (%)	Intra-day precision (%)	Inter-day precision (%)
In plasma
0.5 (LLOQ)	0.5 ± 0.0	98.0	9.8	–b
1.5 (low QC)	1.5 ± 0.1	98.9	4.3	–b
60 (medium QC)	56.0 ± 3.6	93.3	5.4	4.5
600 (high QC)	623.2 ± 64.6	103.9	5.7	10.3
35,000 (50-fold dilution QC)	33,030 ± 1955	94.4	4.6	4.3
In PBS
0.5 (LLOQ)	0.5 ± 0.0	98.1	4.2	3.9
1.5 (low QC)	1.5 ± 0.1	97.9	4.4	3.8
60 (medium QC)	59.1 ± 1.9	98.5	2.7	2.1
600 (high QC)	589.8 ± 24.4	98.3	2.6	3.8

^aEach QC was performed in quintuplicate on 3 different days.

^bNo additional variation was observed from performing the assay on different days.

The matrix effect of human plasma from six different donors was assessed for potential ionization suppression or enhancement on pamiparib and [¹³C₂,¹⁵N₂]pamiparib, as summarized in Table 3. The average matrix factor of pamiparib from six donors’ plasma ranged from 0.87 to 0.99 across the three QC concentrations, with the inter-individual variability (i.e., coefficient of variation of 6 donors) less than 10.7%. The average matrix factor of [¹³C₂,¹⁵N₂] pamiparib (at 50 nM) from six donors’ plasma was 1.06, with the inter-individual variability of 7.6%. The relative (or internal standard normalized) matrix factor of pamiparib ranged from 0.88 to 0.91 across three QC concentrations, with the inter-individual variability less than 3.3% (Table 3). Collectively, these data suggested that while human plasma had no apparent matrix effect on the ionization of pamiparib and [¹³C₂,¹⁵N₂]pamiparib, the use of isotope-labeled internal standard could effectively reduce the inter-individual variability of different source of matrices.

TABLE 3.

Matrix effect and recovery of pamiparib and [¹³C₂,¹⁵N₂]pamiparib from plasma of six different individual donors.

		Absolute		Internal standard normalizedd
Analyte	Nominal concentration (nM)a	Matrix factorb	Recovery (%)c	Matrix factor	Recovery (%)
Pamiparib	1.5	0.99 (10.6)	86 (10.0)	0.89 (2.5)	102 (4.4)
	60	0.97 (10.7)	84 (9.6)	0.91 (2.2)	99 (2.5)
	600	0.87 (3.5)	91 (8.7)	0.88 (3.3)	100 (3.5)
[13C2,15N2]pamiparib	50	1.06 (7.6)	86 (7.3)	–	–

^aNominal concentrations of the analyte were spiked in 40 μL of plasma before precipitation with three volumes of methanol (Set 1). The same amount of the analyte as in Set 1 was spiked in 160 μL of neat solvent (75% MeOH in water) and 160 μL of blank plasma extract for Sets 2 and 3, respectively.

^bAbsolute matrix factor is calculated as the mean peak area ratio of Set 3 to Set 2. Data are shown as the mean (%CV) from six donors’ plasma extract, where CV is the coefficient variation.

^cAbsolute recovery is calculated as the mean peak area ratio of Set 1 to Set 3. Data are shown as the mean (%CV) from six donors’ plasma.

^dInternal standard-normalized (or relative) matrix factor and recovery were calculated based on pamiparib-to-[¹³C₂,¹⁵N₂]pamiparib peak area ratios. Data are shown as the mean (%CV) from six donors’ plasma.

The average recovery of pamiparib from six donors’ plasma across three QC concentrations ranged from 84% to 91%, with the inter-individual variability less than 10% (Table 3). The average recovery of [¹³C₂,¹⁵N₂]pamiparib from six donors’ plasma was 86%, with the inter-individual variability less than 7.3% (Table 3). The relative (or internal standard normalized) recovery of pamiparib was consistently 99%–102% across all three QC concentrations, with the inter-individual variability less than 4.4% (Table 3).

The short- and long-term stability of pamiparib under relevant conditions for routine sample processing and analysis is summarized in Table 4. Bench-top stability test showed that pamiparib was stable in both 50% acetonitrile (at 1 and 100 μM) and human plasma (at 1.5 and 600 nM) at ambient temperature (~25°C) for at least 6 h, suggesting that pamiparib was stable during the routine preparation of calibrators, QCs, and plasma samples at ambient temperature. Autosampler stability test suggested that pamiparib was stable in post-extracted plasma samples at 5°C for at least 9 h, thus allowing the assay to be performed continuously overnight for a large number of samples. Freeze–thaw stability test demonstrated that pamiparib was stable after three freeze–thaw cycles, thus allowing sample reassay if needed. Long-term stability test showed that pamiparib was stable in human plasma (at low and high QC concentrations) at −80°C for at least 6 months, suggesting that clinical samples can be stored at −80°C for at least 6 months before the analysis.

TABLE 4.

Assessment of pamiparib stability.

Condition	Pamiparibc
Bench-top stability (in 50% acetonitrile, 25°C)a	1 μM	100 μM
1 h	96.9	99.7
2h	98.8	101.0
4h	97.1	101.8
6h	97.3	101.7
Bench-top stability (in plasma, 25°C)a	1.5 nM	600 nM
1 h	100.4	102.4
2 h	100.4	102.4
4 h	100.0	103.2
6 h	99.2	102.2
Autosampler stability (in post-precipitation plasma, 5°C)b	1.5 nM	600 nM
1h	100.2	100.7
3h	101.6	100.8
6h	102.5	101.5
9h	103.2	101.7
Freeze-thaw stability (in plasma, −80°C)b	1.5 nM	600 nM
Cycle 1	105.5	110.4
Cycle 2	110.7	111.0
Cycle 3	103.4	106.7
Long-term stability (in plasma, −80°C)b	1.5 nM	600 nM
1 month	106.6	108.5
3 months	98.9	101.7
6 months	99.1	98.3

^aStability data are expressed as the mean percentage of the peak area determined at a certain time relative to that at time zero.

^bStability data are expressed as the mean percentage of the analyte concentration determined at certain time point relative to the nominal concentration (%).

^cEach concentration at each time point was assessed in triplicate.

3.3 |. Optimization of equilibrium dialysis time

The optimal time to equilibrium was assessed with pooled human plasma spiked with pamiparib at 1 μM for dialysis of 2, 6, 16, and 24 h. As shown in Figure 3, the dialysis reached equilibrium at 16 h, at which the mean fraction unbound of pamiparib in pooled human plasma was measured as 12%. The dialysis recovery, calculated as the ratio of total amount of pamiparib recovered from the post-dialysis plasma and PBS samples to the amount spiked in pre-dialysis plasma sample, ranged from 98% to 107% for all QC samples at different dialysis time points. Thus, 16 h was selected as the optimal equilibrium time for determining pamiparib fraction unbound in patient plasma and tumor samples.

FIGURE 3.

Optimization of equilibrium dialysis time. Fraction unbound of pamiparib (spiked at 1000 nM) in pooled human plasma determined after 2, 6, 16, and 24 h of equilibrium dialysis. All data are presented as the mean and standard deviation from triplicate measurement.

3.4 |. Application

The validated method is being applied to assess the tumor penetration of pamiparib in patients with IDH-mutant recurrent gliomas in an ongoing clinical trial (ABTC 1801). The total and unbound pamiparib concentrations in paired plasma and tumor (including contrast-enhancing and non-enhancing regions) samples from three patients are summarized in Table 5. These data have been presented as an oral presentation at the 2021 Annual Meeting of Society of Neuro-oncology (Schiff et al., 2021).

TABLE 5.

Total and unbound concentrations of pamiparib in plasma and brain tumor samples in three patients with IDH-mutant recurrent gliomas^a.

Patient number	Total drug in plasma (μM)	Unbound drug in plasma (μM)	Total drug in tumor (μM)		Unbound drug in tumor (μM)		Kp		Kp,uu
			EN	NEN	EN	NEN	EN	NEN	EN	NEN
0245	7.134	0.243	2.767	NA	0.209	NA	0.39	NA	0.86	NA
0006	11.294	0.420	2.130	1.918	0.188	0.160	0.19	0.17	0.45	0.38
0301	23.243	0.744	8.130	7.022	1.133	0.731	0.35	0.30	1.52	0.98
Mean	13.890	0.469	4.342	4.470	0.510	0.446	0.31	0.24	0.94	0.68
Median	11.294	0.420	2.767	4.470	0.209	0.446	0.35	0.24	0.86	0.68

^aThe patients were treated with pamiparib for 7–10 days before the surgery to reach the drug steady state. On the day of surgical resection of tumor, blood and contrast-enhancing and non-enhancing tumor tissue samples were collected at 4–8 h following the administration of the last dose.

Abbreviations: EN, contrast-enhancing tumor; K_p, total drug tumor-to-plasma concentration ratio; K_p,uu, unbound drug tumor-to-plasma concentration ratio; NA, not available; NEN, non-enhancing tumor.

The median unbound pamiparib concentrations in contrast-enhancing and non-enhancing tumor regions were 209 and 446 nmol/kg (or nmol/L) in three patients (Schiff et al., 2021), which were over 100-fold of the in vitro IC₅₀ for the inhibition of PARP1 (1.6 nM) and PARP2 (0.92 nM) determined from cell-free assay (Xiong et al., 2020). The median pamiparib unbound tumor-to-plasma ratio (K_p,uu) was 0.86 and 0.68 in the enhancing and non-enhancing tumor regions, respectively (Schiff et al., 2021). These data suggested that pamiparib had a good tumor penetration even in areas with a largely intact blood–brain barrier, and it likely achieved sufficient pharmacologically active drug concentrations for target engagement in both enhancing and non-enhancing tumors in patients with glioma.

4 |. CONCLUSION

A simple, sensitive, and reliable LC–MS/MS method based on reversed-phase chromatography was developed for quantitatively determining total and unbound pamiparib concentrations in patient plasma and brain tumors. The method was fully validated with human plasma. Sample preparation involved one-step protein precipitation with methanol. The LLOQ was established at 0.5 nM of pamiparib in plasma, and the linear calibration curve range was 0.5–1000 nM in plasma. The validated method was successfully applied to assess the plasma pharmacokinetics and tumor penetration of pamiparib in patients with recurrent gliomas.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.” cd_value_code=“text.