Effects of CYP1A2 genetic polymorphisms on the pharmacokinetics of pentoxifylline and its active metabolites

Abstract

Purpose

To investigate whether the CYP1A2 genetic polymorphisms could affect the pharmacokinetics of pentoxifylline and its two active metabolites in Chinese healthy participants.

Methods

Forty-six healthy Chinese volunteers were administered a single 400 mg oral dose of pentoxifylline. The plasma concentrations of pentoxifylline and its active metabolites were quantified using LC-MS/MS. The CYP1A2 genotypes for the following loci were determined by the SnapShot technique: -5347T > C (rs2470890), -3860G > A (rs2069514), -3594T > G (rs2069520), -3113 C > A (rs2069521), -2467delT (rs35694136), -739T > G (rs2069526), -163 C > A (rs762551), and 2159G > A (rs2472304).

Results

Participants heterozygous for the CYP1A2 -3860G/A genotype exhibited a 22.3% and 17.6% reduction in the Cmax of metabolite M5 compared to homozygous − 3860G/G and − 3860 A/A carriers, respectively. Participants carrying the CYP1A2 -3860G/A genotype exhibited reduced AUC_0–t and AUC_0–∞ values of metabolite M5 compared to both homozygous − 3860G/G and − 3860 A/A carriers. Additionally, participants with the CYP1A2 -163 C/A genotype had significantly lower pentoxifylline AUC_0–t and AUC_0–∞ than those with the − 163 C/C genotype (reduced by 29.14% and 28.62%, respectively). In contrast, no significant differences were observed in the pharmacokinetic parameters of the primary metabolites M1 and M5 across the different CYP1A2 genotype groups.

Conclusions

The two CYP1A2 polymorphisms significantly affected pharmacokinetics: the − 3860G > A variant was associated with a significant reduction in systemic exposure to metabolite M5, while potentially increasing exposure to pentoxifylline and M1. Conversely, the − 163 C > A variant significantly reduced the plasma exposure of the parent drug pentoxifylline, with a trend toward decreased exposure for metabolites M1 and M5.

Trial registration

This clinical trial has been registered in the Chinese Clinical Trial Register (accessible at http://www.chinadrugtrials.org.cn/index.htmL) with the registration number CTR20233180 on October 08, 2023.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40360-026-01106-2.

Introduction

Pentoxifylline (1-(5-oxohexyl)-3,7-dimethylxanthine), a methylxanthine derivative of theobromine, is a vasoactive agent characterized by potent hemorheological properties [1]. It is approved by the FDA for the symptomatic management of intermittent claudication associated with peripheral vascular disease [2]. Pentoxifylline, a non-selective phosphodiesterase (PDE) inhibitor, exerts its therapeutic effects through multimodal mechanisms. By elevating intracellular adenosine triphosphate (ATP) and cyclic nucleotide levels, pentoxifylline enhances erythrocyte flexibility, thereby improving blood filterability and ameliorating peripheral microcirculation [3]. In patients with chronic kidney disease (CKD), pentoxifylline attenuates renal functional decline, as evidenced by increased estimated glomerular filtration rate (eGFR) and reduced urinary albumin excretion [4, 5]. Furthermore, its anti-inflammatory action is mediated through suppression of pro-inflammatory cytokine secretion, which disrupts inflammatory signaling cascades [6, 7]. These combined pharmacological properties underpin its clinical application as adjuvant therapy in diverse conditions, including cardiovascular disorders, alcoholic hepatitis, osteoradionecrosis of the jaw, type 2 diabetes mellitus (T2DM)-related complications, and other pathologies associated with microvascular impairment or chronic inflammation [8–11].

Following oral administration, pentoxifylline is rapidly absorbed, and its extensive metabolism is primarily mediated by the hepatic cytochrome P450 isoenzyme CYP1A2. The biotransformation of pentoxifylline involves rapid and reversible reduction to M1 in erythrocytes and the liver, coupled with hepatic oxidation leading to the generation of M4 and M5 metabolites [12, 13]. Consequently, following oral administration of pentoxifylline in healthy participants, the plasma concentration-time curve area under the curve (AUC) values of metabolites M1 and M5 exceeded those of the parent drug, whereas M4 exhibited a lower AUC [14]. Pharmacological studies have demonstrated that M1 and M5 have significant hemorheologic effects [13]. Metabolite M1 plays a pivotal role in anti-fibrotic mechanisms and TNF-α modulation, and M5 exhibits superior efficacy to pentoxifylline in suppressing neutrophil activation, including the inhibition of superoxide anion generation, degranulation and surface expression of β2-integrin CD11b/CD18 (Mac-1) [15, 16].

Given the pivotal role of CYP1A2 in its metabolism, the activity of this enzyme significantly influences the pharmacokinetics of pentoxifylline. Pharmacokinetic studies indicate that coadministration of pentoxifylline with CYP1A2 inhibitors results in elevated AUC values for both pentoxifylline and its metabolite M1, whereas concomitant use with CYP1A2 inducers reduces the AUC of metabolite M5 [17–19]. Moreover, substantial interindividual variability in CYP1A2 activity exists, primarily attributable to single nucleotide polymorphisms (SNPs) within the CYP1A2 gene, which influence its transcriptional regulation and enzymatic efficiency [20, 21]. This genetic influence is well-established, polymorphisms such as CYP1A2 *1 C and *1F significantly alter the clearance of probe drugs such as theophylline and caffeine. Studies have shown that different CYP1A2 activities can lead to more than a threefold difference in the pharmacokinetic parameters of caffeine [22].

In previous literature, eight high-prevalence CYP1A2 polymorphisms were identified in Chinese populations, such as: CYP1A2 -5347T > C (*1B) (rs2470890), CYP1A2 -3860G > A (*1 C) (rs2069514), CYP1A2 -3594T > G (rs2069520), CYP1A2 -3113 C > A (rs2069521), CYP1A2 -2467delT (*1D) (rs35694136), CYP1A2 -739T > G (*1E) (rs2069526), CYP1A2 -163 C > A (*1F) (rs762551), and CYP1A2 2159G > A (rs2472304) [23, 24]. Reported allele frequencies for these variants in Asian populations range from approximately 5% to over 30%, with most exceeding 17%. A comparative analysis was performed using the 1000 Genomes Project database (https://www.ncbi.nlm.nih.gov/snp) to investigate these significant frequency differences relative to data reported in Caucasian populations. The analysis revealed distinct patterns for key variants. For instance, the frequency of the CYP1A2 -163 C > A (*1F) allele was higher in individuals of European ancestry (71.30%) than in those of Asian ancestry (66.12%). In contrast, the − 3860G > A (*1 C) allele, associated with reduced enzyme activity, was found almost exclusively in Asian populations (with a frequency of 17.90%) and was extremely rare (< 1%) in Europeans. Similarly, the − 5347T > C (*1B) variant demonstrated a notable population skew, being more prevalent in Asians. These pronounced disparities underscore the critical importance of conducting population-specific pharmacogenetic studies [25].

To the best of our knowledge, no previous studies have evaluated influences of the CYP1A2 genetic polymorphism on pentoxifylline pharmacokinetic parameters. Thus, we evaluated the effects of the CYP1A2 alleles on the pharmacokinetics of pentoxifylline and its metabolites in healthy participants, in order to elucidate if there is any relevant pharmacogenetic factor affecting the disposition of pentoxifylline.

Materials and methods

Ethics

This study included 46 healthy Chinese volunteers from a pentoxifylline bioequivalence trial conducted at the Clinical Trial Unit of Hebei General Hospital. The study protocol was approved by the Research Ethics Committee of Hebei General Hospital (Approval No. 2023-33) and registered on the Chinese Clinical Trial Registry (http://www.chinadrugtrials.org.cn/index.html, #CTR20233180).

The trial adhered to the International Conference on Harmonization’s Good Clinical Practice Guidelines and the ethical principles of the Declaration of Helsinki [26, 27].

Participants

The study employed specific inclusion criteria, which included non-smoking and no history of drug or alcohol abuse, male and female adult participants (aged 18 or older) with a body mass index (BMI) ranging from 19.0 to 26.0 kg/m². Enrolled participants exhibited no clinically significant abnormalities in vital signs, physical examination, clinical laboratory tests, 12-lead electrocardiogram (ECG), and chest X-ray. The exclusion criteria were as follows: a history of any drug hypersensitivity; the use of known inhibitors or inducers of hepatic drug metabolism, or medications interacting with pentoxifylline (e.g., ciprofloxacin, barbiturates, and cimetidine), within the four weeks prior to screening, as verified by participant interview and drug screening; a history of clinically significant bleeding; or participation in another investigational drug study within the three months prior to enrollment. Written informed consent was obtained from all participants prior to enrollment, with explicit acknowledgment of their right to withdraw from the study at any time without prejudice to future medical care or legal entitlements.

Study design

All participants were admitted to the clinical trial center on the afternoon prior to drug administration. After a 10-hour fasting period, each participant received a single 400 mg dose of pentoxifylline sustained-release tablet (Sanofi S.r.l., Paris, France, batch number: 1U004) with 240 mL water. Approximately 4 mL of venous blood was collected from each participant into EDTA-K2 anticoagulant vacuum tubes at 0 (pre-dose), 0.17, 0.33, 0.50, 1.00, 1.50, 2.00, 2.50, 3.00, 3.50, 4.00, 4.50, 5.00, 6.00, 8.00, 10.00, 12.00, 16.00, 20.00, 24.00, 27.00, 30.00, and 36.00 h post-dose. All blood samples were centrifuged at 1700 g at 4 °C for 10 min. The plasma samples were immediately transferred into polyethylene tubes, and stored at − 80 °C until analysis.

Determination of pentoxifylline, M1, and M5 in plasma

The concentrations of pentoxifylline and its active metabolites in plasma were measured using a validated liquid chromatography-tandem mass spectrometry (LC-MS/MS) method. Chromatographic analysis was conducted using Waters^® Acquity I-Class PLUS UPLC^® system (Milford, MA, USA) coupled with Waters^® Xevo TQ-XS triple quadrupole mass spectrometer equipped with an electrospray ionization (ESI) source. Briefly, pentoxifylline-d6, M1-d6, and M5-d6 were added to plasma samples as internal standards (IS), followed by protein precipitation. A 10 µL aliquot of the supernatant was injected for analysis. Chromatographic separation was achieved using a Shim-pack GIST-HP C18-AQ column (3.0 × 50 mm, 3.0 μm). Quantification was performed via multiple reaction monitoring (MRM) in positive ion mode on the mass spectrometer. Calibration curves for pentoxifylline, M1, and M5 demonstrated linearity over the ranges of 2.00-1000 ng/mL (R²= 0.999), 5.00-2500 ng/mL (R² = 0.999), and 5.00-2500 ng/mL (R²= 0.998), respectively. Intra-day and inter-day accuracy (%bias) and precision (%CV) met the validated acceptance criteria, with all values within ± 10% for pentoxifylline, M1, and M5.

Pharmacokinetic analysis

The pharmacokinetic parameters of pentoxifylline and its active metabolites were estimated using non-compartmental methods with the Phoenix WinNonlin 8.2 (Certara, Princeton, NJ, USA). All parameters were calculated based on the actual blood sampling times recorded for each individual. The peak concentrations (Cmax) and the time of the maximum observed concentration (Tmax) were estimated directly from the observed plasma concentration-time data. The AUC_0−t was calculated using the log-linear trapezoidal rule. The elimination rate constant (λz) was estimated from the least-squares regression slope of the terminal plasma concentration. The AUC_0–∞ was calculated as AUC_0–∞ = AUC_0−t + Ct/λz, where Ct is the most recently measured plasma concentration. The t_1/2 was calculated as ln2/λz and the CL/F of pentoxifylline was calculated as Dose/AUC_0–∞. To assess the in vivo metabolic conversion of pentoxifylline, the area under the curve ratios of its active metabolites to the parent drug were calculated as follows: AUC_M1:AUC_PTX and AUC_M5:AUC_PTX.

Safety assessment

The safety and tolerability of pentoxifylline were assessed by monitoring adverse events (AEs), vital signs, clinical laboratory tests, physical examinations, and 12-lead electrocardiograms (ECGs). Vital signs included blood pressure (BP), heart rate, and axillary temperature.Throughout the study, the participants were asked about any AEs they encountered, and all AEs were promptly reported by the participants and documented immediately. Karch and Lasagna criteria were used to determine the causal relationship of adverse events, which were categorized as definite, probable, possible, unlikely, or unrelated [28]. Only those AEs that were categorized as definite, probable or possible were considered as adverse drug reactions (ADRs) and considered for statistical analysis.

Genotyping

DNA was extracted from peripheral blood samples using an automatic DNA extractor (Lab-Aid 824s; Xiamen Zhishan Biological Technology Co.,Ltd., Xiamen, China). Polymerase chain reaction (PCR) was used to amplify the following eight fragments: CYP1A2 -5347T > C, CYP1A2 -3860G > A, CYP1A2 -3594T > G, CYP1A2 -3113 C > A, CYP1A2 -2467delT, CYP1A2 -739T > G, CYP1A2 -163 C > A, and CYP1A2 2159G > A. The primers were designed with the Premier 5 software (Premier Biosoft International, Palo Alto, CA, USA) and primers used in this study are shown in Table 1.

Table 1.

Primers used in this study

Polymorphism	Upper Primer (5’ to 3’)	Lower Primer (5’ to 3’)	Tm(℃)
rs2470890	GCAACTGGAGTTCAGCGTG	GCTCAAATGATCCTCCAACC	60
rs2069514	ACGGGACTTCTTGGATGCTTAT	TAATTCTAGCACTTTGGGAGGC	60
rs2069520	AACTCCTGGCCTCACTCAAG	CCTTGGAGGTTAGGTGCCAT	60
rs2069521	GTCTTCCCACCAACAAACCATA	GGCCATGAATGGGAGAAGAG	60
rs35694136	GCCCAGAAGTTCAAGACCAA	GGACAAGCCTTAAATTGGATG	60
rs2069526	CAACCCTGCCAATCTCAAGC	CTCCCCAGGGCATTCTTTAT	60
rs762551	GGTCACTTGCCTCTACTCCA	ATGCGTGTTCTGTGCTTGG	60
rs2472304	CACAGTCACCACAGCCATCT	CCACTAACCTCCCACATCTTCT	60

PCR thermal cycling conditions consisted of denaturation at 98 ℃ for 2 min, then 35 cycles of denaturation at 98 ℃ for 10 s, annealing at 60 ℃ for 10 s, and extension at 72 ℃ for 10 s, followed by the final extension at 72 ℃ for 5 min. Then the amplification reaction product was purified. Purified amplification reaction product was added into the extension reaction system, then the DNA sequences were analyzed by ABI 3730xl DNA sequencing machine (Applied Biosystems, Foster City, CA, USA) and 46 healthy Chinese participants’ genotypes were analyzed by GeneMapper 4.1 (Applied Biosystems, Inc.).

Statistical analysis

All statistical analyses were performed using SPSS^® software (version 22.0; IBM Corp., Armonk, NY, USA). Data in the text and tables are expressed as mean values ± SD, except for Tmax, which is presented as the median (range). For graphical clarity, the values are shown as means ± SD. Differences in genotype frequencies attributed to sex were determined using the corrected Pearson’s chi-square test (χ²). The Hardy-Weinberg equilibrium of each chosen SNP was verified through chi-squared test. The Shapiro-Wilk test was employed to assess data normality. All pharmacokinetic parameters underwent logarithmic transformation to normalize their distributions. For normally distributed pharmacokinetic variables, t-tests or analysis of variance (ANOVA) were conducted, with Bonferroni corrections applied to multiple comparisons to minimize Type I error risks. With a type I error rate set at 0.05, the study had more than 80% power to detect a 30% difference in AUC_0–∞ between the target genotypes [29, 30]. Non-parametric tests were used when parametric methods were inapplicable: Mann-Whitney U-tests for two-category variables and Kruskal-Wallis tests for three or more categories. Statistical significance was determined by p < 0.05.

The incidence of adverse drug reactions in different genotypes was compared using the chi-square test, and linear regression was used to analyze the correlation of AUC_0−t and Cmax with changes in BP, heart rate and QTc of the 12-lead ECG. The main analysis consisted of a multiple regression analysis (logistic regression for ADRs), including factors in the univariate analysis with p < 0.05. Therefore, dummy variables were used to analyze classified variables (such as polymorphism) with more than two categories, and statistical significance was set at p < 0.05.

Results

Demographic and genotypic characteristics

Table 2 summarizes the basic demographic characteristics of the participants. Blood samples were collected from all 46 participants for analysis, and the pharmacokinetic parameters of pentoxifylline and its metabolites M1 and M5 were determined. The genotypes and corresponding allele frequencies among participants are detailed in Table 3. The genotype distributions were in Hardy-Weinberg equilibrium in each group. Two SNPs (CYP1A2 -2467delT and CYP1A2 3594T > G) were omitted because they deviated from the values expected under Hardy-Weinberg equilibrium (p = 0.003, and p < 0.001, respectively). No statistically significant differences in genotypic frequencies were detected between male and female participants.

Table 2.

Demographic and clinical characteristics of the participants

Characteristics	Median (Range) or N(%)	Mean ± SD
Weight (kg)	63.50 (56.02,72.00)	63.97 ± 9.69
Hight (cm)	164.00 (158.38,175.63)	166.18 ± 9.55
BMI (kg/m2)	23.15 (21.63,24.40)	23.04 ± 1.68
Age (y)	29.00 (22.75,40.00)	31.22 ± 9.27
Sex (Male/Female)	26 (56.5%)/20(43.5%)

Table 3.

Analyzed single nucleotide polymorphisms (SNPs) characteristics in 46 healthy Chinese participants

Gene	Nucleotide change (A > B*)	Genotypic frequencies n (%)			MAF	HWE
		Wild%	Hetero%	Variant%
CYP1A2	-5347 T > C	1(2.17)	9(19.57)	36(78.26)	0.12	0.63
	-3860 G > A	25(54.35)	17(36.96)	4(8.70)	0.27	0.65
	-3594T > G	37(80.43)	5(10.87)	4(8.70)	0.14	0.000
	-3113 G > A	40(86.96)	6(13.04)	0	0.07	0.64
	-2467delT	18(39.13)	28(60.87)	0	0.30	0.003
	-739 T > G	40(86.96)	5(10.87)	1(2.17)	0.08	0.12
	-163 C > A	9(19.57)	21(45.65)	16(34.78)	0.31	0.66
	2159 G > A	36(78.26)	9(19.57)	1(2.17)	0.12	0.63

^* A, wild allele; B, mutant allele

Wild, wildtype; hetero, heterozygous mutant; variant, homozygous mutant

MAF, minor allele frequency; HWE, Hardy-Weinberg equilibrium. The P values were verified through chi-squared test (χ²)

Pharmacokinetic analysis

The mean and standard deviation (± SD) of the main pharmacokinetic parameters of 46 healthy participants were shown in Table 4. Tmax is presented as median values. Significant associations were observed between specific pharmacokinetic parameters and the CYP1A2 -3860G > A (*1 C) and − 163 C > A (*1F) polymorphisms. A proposed metabolic scheme, integrating the findings of this study, is presented and discussed later in Fig. 3.

Table 4.

Comparison of Pharmacokinetic parameters of Pentoxifylline and M1 and M5 metabolites after a single 400-mg oral dose of Pharmacokinetic according to the CYP1A2 -3860 G > A and CYP1A2 -163 C > A genotypes

	CYP1A2 -3860 G > A (*1 C) (rs2069514)				CYP1A2 -163 C> A (*1F) (rs762551)
	GG	GA	AA	GA + AA	CC	CA	AA
	n=25	n=17	n=4	n=21	n=9	n=21	n=16
Pentoxifylline
Cmax (ng/mL)	216.51 ± 84.69	353.811 ± 505.84	188.13 ± 105.27	322.25 ± 459.13	264.30 ± 90.38	217.42 ± 122.67	327.22 ± 518.97
AUC0−t(h·ng/mL)	1631.57 ± 658.70	1824.36 ± 1688.82	1291.54 ± 601.60	1722.87 ± 1543.35	1988.24 ± 750.12	1408.86 ± 681.04*	1843.08 ± 1664.27
AUC0−∞(h·ng/mL)	1655.33 ± 650.67	1841.74 ± 1685.46	1316.79 ± 605.45	1741.75 ± 1540.20	2003.18 ± 751.57	1429.95 ± 676.56*	1868.90 ± 1657.37
Tmax(h)	1.49(0.32,4.99)	0.99(0.16,4.99)	0.41(0.32,1.49)	0.49(0.16,4.99)	1.49(0.16,2.99)	0.99(0.32,4.99)	1.24(0.32,4.99)
t1/2 (h)	2.72 ± 1.41	2.52 ± 1.53	3.63 ± 2.52	2.73 ± 1.74	2.37 ± 0.69	2.99 ± 1.75	2.59 ± 1.63
CL/F(mL/h)	278.39 ± 102.53	323.41 ± 173.60	341.79 ± 111.36	326.91 ± 161.32	232.22 ± 102.27	339.14 ± 149.68	288.30 ± 113.19
M1
Cmax (ng/mL)	479.44 ± 227.56	620.35 ± 620.47	346.00 ± 113.83	568.10 ± 567.55	559.44 ± 286.64	456.86 ± 162.47	580.44 ± 654.90
AUC0−t(h·ng/mL)	4365.48 ± 1814.99	5345.51 ± 4999.77	3225.27 ± 751.36	4941.65 ± 4561.87	5076.99 ± 2246.31	4103.54 ± 1579.75	5065.28 ± 5151.91
AUC0−∞(h·ng/mL)	4409.97 ± 1801.93	5381.18 ± 4997.60	3271.62 ± 741.66	4979.36 ± 4558.93	5106.73 ± 2239.18	4145.07 ± 1578.38	5113.05 ± 5142.78
Tmax(h)	2.49(0.99,3.49)	1.99(0.99,4.99)	1.99(1.49,2.49)	1.99(0.99,4.99)	2.49(1.49,3.99)	1.99(0.99,3.49)	2.49(1.49,4.99)
t1/2 (h)	2.67 ± 1.32	2.28 ± 1.26	3.15 ± 1.60	2.44 ± 1.34	2.14 ± 0.32	2.56 ± 1.24	2.82 ± 1.71
AUCM1:AUCPTX	2.71 ± 0.70	3.14 ± 1.09	2.69 ± 0.80	3.05 ± 1.04	2.52 ± 0.63	3.10 ± 0.97	2.75 ± 0.82
M5
Cmax (ng/mL)	756.64 ± 181.99	587.65 ± 96.05**	713.50 ± 213.62	611.62 ± 129.58*	778.56 ± 210.06	665.48 ± 169.27	673.63 ± 154.31
AUC0−t(h·ng/mL)	7210.60 ± 1338.04	5229.88 ± 1460.39**	6911.08 ± 2063.39	5550.11 ± 1674.05**	7560.52 ± 1619.75	6204.07 ± 1639.59	6155.45 ± 1672.80
AUC0−∞(h·ng/mL)	7250.01 ± 1344.67	5264.54 ± 1467.09**	6942.12 ± 2075.45	5584.08 ± 1680.37**	7595.91 ± 1631.51	6244.32 ± 1648.03	6188.86 ± 1677.35
Tmax(h)	1.99(0.99,4.49)	1.99(0.99,4.99)	2.24(1.49,2.49)	1.99(0.99,4.99)	2.49(1.49,3.99)	1.99(0.99,4.49)	2.24(0.99,4.99)
t1/2 (h)	2.39 ± 1.03	1.98 ± 0.73	2.02 ± 0.34	1.99 ± 0.67	2.38 ± 0.76	2.24 ± 1.05	2.08 ± 0.77
AUCM5:AUCPTX	4.98 ± 1.94	4.01 ± 2.25	6.04 ± 2.90	4.40 ± 2.45	4.44 ± 2.25	5.04 ± 2.23	4.44 ± 2.16

^{*p<0.05 after ANOVA and Bonferroni post−Hoc}

^{**p<0.01 after ANOVA and Bonferroni post−Hoc}

Fig. 3.

Structures of pentoxifylline, metabolite 1 and metabolite 5

Effects of polymorphic CYP1A2 -3860 C > A(*1 C) (rs2069514) genotypes on pentoxifylline pharmacokinetics

The CYP1A2 -3860G > A(*1 C) (rs2069514) genetic polymorphism significantly influenced the pharmacokinetic parameters of M5. participants carrying the heterozygous − 3860G/A genotype exhibited a 22.3% and 17.6% reduction in Cmax compared to homozygous − 3860G/G and − 3860 A/A carriers, respectively (Bonferroni adjusted P < 0.01). The AUC_0–t decreased by 27.5% and 24.3%, and the AUC_0–∞ declined by 27.4% and 24.2% in heterozygous individuals relative to their homozygous counterparts (Bonferroni adjusted P < 0.01 for all comparisons).

Analysis under a co-dominant model initially suggested that only the heterozygous allele altered the metabolism of pentoxifylline. However, the limited sample size of only four participants carrying the CYP1A2 -3860 A/A genotype raised concerns about the statistical reliability of these findings. Therefore, a dominant model was applied by grouping A-allele carriers together (G/A + A/A) and comparing them against the G/G group to re-examine the statistical significance of Cmax and AUC. The results under this dominant model revealed statistically significant differences in the Cmax (P < 0.05) and AUC (P < 0.01) of the active metabolite M5 between the combined G/A + A/A group and the G/G group.

These findings underscore the critical role of CYP1A2 -3860G > A(*1 C) (rs2069514) genotype in modulating oxidative metabolism efficiency and systemic exposure to M5. Figure 1(A, B,C) displays the plasma concentration-time profiles of pentoxifylline, M1 and M5 stratified by the CYP1A2 -3860G > A(*1 C) (rs2069514) genotype.

Fig. 1.

Plasma concentration-time profiles of pentoxifylline (A), M1 (B), and M5 (C) in CYP1A2 -3860G > A genotype groups following a single 400 mg oral dose of pentoxifylline. Squares: CYP1A2 -3860G/G (n = 25); circles: CYP1A2 -3860G/A (n = 17); triangles: CYP1A2 -3860 A/A genotype (n = 4). Data are presented as mean ± SD

Effects of polymorphic CYP1A2 -163 C > A(*1F) (rs762551) genotypes on pentoxifylline pharmacokinetics

For the CYP1A2 -163 C > A(*1F) (rs762551) polymorphism, compared to -163 C/C homozygous participants, carriers of the − 163 C/A genotype demonstrated significant reductions of 29.1% in AUC_0–t (P < 0.05) and 28.6% in AUC_0–∞ (P < 0.05) for pentoxifylline. Plasma exposures of metabolites M1 and M5 also showed non-significant decreasing trends. Figure 2 depicts the mean plasma concentration-time profiles of pentoxifylline and the active metabolites M1 and M5, stratified by the CYP1A2 -163 C > A (*1F) (rs762551) genotype.

Fig. 2.

Plasma concentration-time profiles of pentoxifylline (A), M1 (B), and M5 (C) in CYP1A2 -163 C > A genotype groups following a single 400 mg oral dose of pentoxifylline. Squares: CYP1A2 -163 C/C (n = 9); circles: CYP1A2 -163 C/A (n = 21); triangles: CYP1A2 -163 A/A genotype (n = 16). Data are presented as mean ± SD

Effects of polymorphic CYP1A2 -3113 C > A (rs2069521), CYP1A2 -739T > G(*1E) (rs2069526), CYP1A2 -5347T > C (*1B) (rs2470890) and CYP1A2 2159G > A (rs2472304) genotypes on pentoxifylline pharmacokinetics

Statistical analysis revealed no significant differences in the pharmacokinetic parameters of pentoxifylline or its M1 and M5 metabolites between the CYP1A2 -3113 C > A (rs2069521), CYP1A2 -739T > G(*1E) (rs2069526), CYP1A2 -5347T > C (*1B) (rs2470890) and CYP1A2 2159G > A (rs2472304) genotypes. The mean ± SD of the main pharmacokinetic parameters for the 46 healthy participants are presented in Supplementary Tables.

Adverse drug reactions

No serious or life-threatening adverse events (AEs) were observed during the study. Investigators assessed and recorded safety in accordance with the International Conference on Harmonisation Good Clinical Practice guidelines, in accordance with industry standards and investigator guidance on safety reporting for Investigational New Drug (IND) and Bioavailability/Bioequivalence (BA/BE) studies [31].

Adverse events and adverse drug reactions recorded during the trial are enumerated in Table 5. A total of 10 participants (21.7%) experienced 22 adverse events. Of these, 2 participants (4.3%) heterozygous for the CYP1A2 -3860G/A variant experienced adverse drug reactions (ADRs). The most frequently reported AEs were abnormal white blood cell count and arrhythmia/tachycardia, which were considered potentially related to the study drug. Specifically, one participant (2.2%) exhibited decreased white blood cell count, and one participant (2.2%) exhibited elevated heart rate.

Table 5.

Summary of adverse events in healthy participants

Adverse Event	No. of Events	No. of Participants (%)	Severity Grade	Relationship to Investigational Product
All adverse events	22	10(21.7)	Grade I-Ⅱ	Possibly related/ Unrelated
Urinary leukocytes positive	8	4(8.7)	Grade I	Possibly Unrelated
Neutrophil count decreased	1	1(2.2)	Grade Ⅱ	Possibly related
Urinary sediment detected	2	2(4.3)	Grade I	Possibly Unrelated
Blood triglycerides increased	2	2(4.3)	Grade I	Possibly Unrelated
Urinary erythrocytes positive	1	1(2.2)	Grade I	Possibly Unrelated
Urinary occult blood positive	1	1(2.2)	Grade I	Possibly Unrelated
Hemoglobin decreased	1	1(2.2)	Grade I	Possibly Unrelated
Lymphocyte percentage decreased	1	1(2.2)	Grade I	Possibly Unrelated
Mean corpuscular hemoglobin concentration decreased	1	1(2.2)	Grade I	Possibly Unrelated
Heart rate increased	1	1(2.2)	Grade I	Possibly related
Blood cholesterol increased	1	1(2.2)	Grade I	Possibly Unrelated
Serum creatinine increased	1	1(2.2)	Grade I	Possibly Unrelated
Upper respiratory tract infection	1	1(2.2)	Grade Ⅱ	Possibly Unrelated

Given the limited number of ADR cases (n = 2), no statistically significant association could be established between genotype and specific ADRs. Studies with larger sample sizes are required to validate this observation.

Discussion

In a bioequivalence study conducted at our center, significant interindividual variability was observed in the pharmacokinetics of pentoxifylline among healthy Chinese volunteers. For the reference formulation, the pharmacokinetic variability was substantial: the CV% for pentoxifylline Cmax approached 119.64% and for its AUC was 67.97%. Similarly, high variability was noted for the active metabolite M1, with CV% values of 79.95% for Cmax and 71.95% for AUC. Genetic factors are considered potential contributors to the differential response to pentoxifylline. This study aimed to further elucidate the interactions between genetic polymorphisms and the pharmacokinetics as well as adverse reactions associated with pentoxifylline. To the best of our knowledge, this represents the first pharmacogenetic investigation examining the association between CYP1A2 polymorphisms and pentoxifylline metabolism.

The role of CYP1A2 in pentoxifylline metabolism is well established, as evidenced by mechanistic drug-drug interaction studies involving CYP1A2 inhibition and by the observed accumulation of pentoxifylline in CYP1A2-knockout mice. It is widely recognized that CYP1A2 expression and activity exhibit substantial interindividual variability (up to 40-fold differences in protein expression), with genetic polymorphisms accounting for 35%-75% of this variation [32–34].

CYP1A2 allelic variants comprise one or more SNPs, and a given SNP can reside within multiple haplotypes, which complicates the identification of these variants. In our study, we investigated a total of 8 SNPs and identified that the CYP1A2 -3860G > A (*1 C) (rs2069514) and CYP1A2 -163 C > A (*1F) (rs762551) polymorphisms exhibited significant effects on the pharmacokinetic parameters of pentoxifylline, M1 and M5. The two common CYP1A2 polymorphisms, -3860G > A (rs2069514) and − 163 C > A (rs762551), result in differential alterations in enzymatic activity. The − 3860G > A polymorphism (rs2069514), located in the 5′-flanking region of the CYP1A2 gene, tags the CYP1A2*1 C haplotype associated with diminished enzyme inducibility and reduced overall catalytic activity [35, 36]. The CYP1A2 -163 C > A polymorphism has been implicated in the increased inducibility observed in smokers and individuals taking inducing agents [37, 38]. However, evidence regarding its contribution to CYP1A2 activity is inconsistent in published reports [39–42].

Carriers of the − 3860G > A variant had significantly lower Cmax and AUC of the metabolite M5 compared to wild-type homozygotes. This finding persisted under a dominant genetic model, suggesting that the − 3860G > A variant impairs the hepatic oxidative conversion of pentoxifylline to M5. Bonferroni correction to control for type I error inflation in multiple comparisons and thoroughly investigated potential confounding factors specific to this association.

Furthermore, participants carrying the − 163 C > A genotype showed significantly reduced pentoxifylline AUC compared to wild-type homozygotes, with heterozygous carriers also demonstrating a trend toward decreased plasma exposure to both M1 and M5 metabolites. This pattern may be explained by enhanced CYP1A2-mediated metabolism of M1. Notably, while pentoxifylline exposure was significantly lower in -163 C/A heterozygotes than in -163 C/C individuals, the reduction in -163 A/A homozygotes did not reach statistical significance. It is known that the − 163 C > A polymorphism influences the enzyme’s inducibility rather than its basal activity. The heterozygous state may represent an intermediate phenotype with altered regulatory kinetics sufficient to influence drug metabolism [43]. In contrast, the homozygous mutant may engage different compensatory metabolic pathways or exhibit non-linear changes in enzyme function that do not translate into a statistically significant difference in this specific metabolic pathway for pentoxifylline.

On the other hand, CYP1A2 is among the most environmentally sensitive drug-metabolizing enzymes. Although this study benefited from the inclusion of non-smokers and the exclusion of known CYP1A2 inhibitors or inducers, it is virtually impossible in practice to fully control all environmental factors that modulate CYP1A2 activity. Numerous dietary components, medications, and lifestyle factors can exert significant effects. For instance, habitual caffeine intake is a well-characterized inducer [44], whereas cruciferous vegetables and grilled meats can also modulate activity [45]. Common medications, including oral contraceptives and certain fluoroquinolone antibiotics [46], are potent inhibitors. Residual variation in CYP1A2 activity due to environmental influences may have obscured potential gene-dose effects.

Given the reversible nature of the metabolic conversion between pentoxifylline and M1, the depletion of M1 consequently reduces the AUC of the parent drug. As illustrated in the metabolic profile (Fig. 3), the same CYP1A2 enzyme is likely responsible for the xanthine 7-demethylation of pentoxifylline to M6 and of M1 to M7. The findings of this study suggest that the CYP1A2 -163 C > A polymorphism contributes at least partially to the variability in pentoxifylline plasma exposure levels.

While previous in vitro studies have established that CYP1A2 inhibition increases systemic exposure to both pentoxifylline and metabolite M1, specific pharmacokinetic data for metabolite M5 remain limited. Pentoxifylline undergoes conversion to M5 through sequential metabolic steps. Although the complete spectrum of cytochrome P450 enzymes involved has not been fully characterized, our results suggest that CYP1A2 activity plays a significant role in this metabolic pathway. The − 3860G > A variant is predicted to reduce enzymatic function, thereby suppressing the oxidative transformation of pentoxifylline to M5. The lack of statistically significant effects of this genotype on pentoxifylline and M1 exposure warrants careful consideration. Importantly, in vitro evidence indicates that CYP1A2 inhibitors predominantly affect the active enantiomer (S)-M1. As our analytical method quantified total M1 content rather than the individual enantiomers, this methodological difference likely accounts for the apparent discrepancy between our findings and previously reported in vitro data.

The carboxylated metabolite M5 is considered a key active metabolite, responsible for pentoxifylline’s potent inhibition of neutrophil functions such as degranulation and superoxide anion generation—effects significantly more potent than those of the parent drug.²¹ Therefore, the therapeutic efficacy of pentoxifylline in neutrophil-mediated pathologies (e.g., myocardial infarction, venous leg ulcers, and ARDS) is critically dependent on the adequate systemic exposure to M5 [47]. Consequently, standard pentoxifylline dosing may yield subtherapeutic M5 concentrations at the site of inflammation, leading to attenuated efficacy in these individuals. This pharmacogenetic mechanism may explain a portion of the interindividual variability in treatment response observed clinically.

Interestingly, the plasma concentration–time curve of pentoxifylline exhibited a typical biphasic absorption profile.This phenomenon may be largely attributable to enterohepatic recirculation, a process in which the drug is metabolized in the liver, excreted into the intestine via bile, and subsequently reabsorbed. In this study, a standard lunch was provided at 4 h post-dose. Food intake can influence drug absorption through multiple pathways, which may further enhance the recirculation process by stimulating bile secretion, thereby potentially aiding the absorption of the drug.

This study has several limitations. First, the relatively small sample size may limit the statistical power of the analyses, underscoring the need for larger cohort studies to provide more robust comparative evidence. Second, the genetic analysis was confined to single variants of the CYP1A2 gene; haplotype-based analyses combining multiple polymorphisms were not performed. Third, the influence of environmental factors on CYP1A2 activity can introduce residual variation, thereby potentially masking a clear gene-dose response. Fourth, this was a single‑dose study conducted in healthy Chinese volunteers, and future investigations should evaluate multiple‑dose, long‑term administration in patient populations. Finally, further pharmacokinetic–pharmacodynamic studies in Chinese participants are warranted to better guide the clinical dosing of pentoxifylline in this patient population.

Conclusion

This study suggests that the CYP1A2 − 3860G > A (*1 C) (rs2069514) and CYP1A2 − 163 C > A (*1F) (rs762551) polymorphisms may be associated with interindividual variability in the pharmacokinetics of pentoxifylline and its active metabolite. Specifically, the − 3860G > A variant significantly reduced the systemic exposure of the key active metabolite M5, while the − 163 C > A variant significantly decreased the exposure of the parent drug and showed a non-significant trend toward reducing metabolite M5 exposure. These reductions may lead to subtherapeutic concentrations at standard doses, explaining part of the observed variability in clinical response and providing a foundation for future genotype-guided therapy. Furthermore, this study offers valuable insights for advancing precision medicine in this field.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ATP

Adenosine triphosphate

AUC_0 − t

Area under the curve from time 0 to t‌

AUC_0−∞

Area under the curve from time 0 to infinity

AE

Adverse Events

BMI

Body mass index

Cmax

Maximum concentration

CKD

Chronic kidney disease

CV

Coefficient of variation

CYP 450

Cytochrome P450

eGFR

Estimated glomerular filtration rate

ESI

Electrospray ionization

IS

Internal standard

Mean ± SD

Mean ± standard deviation

MRM

Multiple reaction monitoring

P

Probability

PDE

Phosphodiesterase

PK

Pharmacokinetics

PTX

Pentoxifylline

SNPs

Single nucleotide polymorphisms

Tmax

Time to maximum concentration

UPLC-MS/MS

Ultra Performance liquid chromatography-tandem mass spectrometry

Author contributions

Lingfang Guo: Conceptualization, Investigation, Formal analysis, Writing – original draft, Writing – review and editing. Xue Sun: Investigation, Formal Analysis, Writing – review and editing. Bo Qiu: Conceptualization, Formal Analysis, Writing – review and editing. Wanjun Bai: Conceptualization, Formal Analysis, Writing – review and editing. Yabin Du: Validation, Conceptualization, Writing – review and editing. Haojing Song: Validation, Conceptualization, Project administration, Writing – review and editing, Funding acquisition.

Funding

This study was supported by the Medical Applicable Technology Tracking Project of Hebei (No. GZ20260036) and Medical Science Research of Hebei (No. 20260018).

Data availability

The authors declare that all the data supporting the findings of this study are contained within the paper.

Declarations

Ethical approval

The study protocol was approved by the Research Ethics Committee of Hebei General Hospital (Approval No. 2023-33) and registered on the Chinese Clinical Trial Registry (http://www.chinadrugtrials.org.cn/index.html, #CTR20233180).

Informed consent

Informed consent was written by all individual participants included in the study.

Competing interests

The authors declare no competing interests.