Next Article in Journal
Feasibility Study of Nose-to-Brain Delivery of Galantamine for Alzheimer’s Disease: Enhancing Olfactory-Region Deposition to Improve Therapeutic Efficacy
Previous Article in Journal
Plant Extracts and Bee Products as Bioactive Compounds: From Mechanism to Delivery
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Pharmaceutics
Volume 18
Issue 7
10.3390/pharmaceutics18070884
pharmaceutics-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Pharmacogenomics of Rivaroxaban: Association of CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 Variants with Bleeding and Thrombotic Outcomes in Real-World Clinical Practice †

by
Ana Marija Slišković
1,*,
Vladimir Trkulja
2,
Lana Ganoci
2,3,
Tamara Božina
4,
Vedran Pašara
1,
Majda Vrkić Kirhmajer
1,5,
Jozefina Palić
3,
Dominik Strikić
6,
Marino Narančić
7,
Ivana Sopek Merkaš
8,
Iveta Merćep
5,6,
Joško Bulum
1,5 and
Livija Šimičević
3,4
1
Department of Cardiovascular Diseases, University Hospital Centre Zagreb, 10000 Zagreb, Croatia
2
Department of Basic and Clinical Pharmacology, University of Zagreb School of Medicine, 10000 Zagreb, Croatia
3
Division of Pharmacogenomics and Therapy Individualization, Department of Laboratory Diagnostics, University Hospital Centre Zagreb, 10000 Zagreb, Croatia
4
Department of Medical Chemistry, Biochemistry and Clinical Chemistry, University of Zagreb School of Medicine, 10000 Zagreb, Croatia
5
Department of Internal Medicine, University of Zagreb School of Medicine, 10000 Zagreb, Croatia
6
Division of Clinical Pharmacology, Department of Internal Medicine, University Hospital Centre Zagreb, 10000 Zagreb, Croatia
7
Division of Hematology, Department of Internal Medicine, University Hospital Centre Zagreb, 10000 Zagreb, Croatia
8
Department of Cardiology, Special Hospital for Medical Rehabilitation Krapinske Toplice, 49217 Krapinske Toplice, Croatia
*
Author to whom correspondence should be addressed.
†
A nested case–control study was conducted at the Department of Cardiovascular Diseases and the Department of Laboratory Diagnostics, University Hospital Centre Zagreb (UHC Zagreb), Croatia, between September 2021 and September 2024, as part of the Croatian Science Foundation project PGxCardioDrug (UIP-2020-02-8189).
Pharmaceutics 2026, 18(7), 884; https://doi.org/10.3390/pharmaceutics18070884
Submission received: 6 June 2026 / Revised: 6 July 2026 / Accepted: 17 July 2026 / Published: 20 July 2026
(This article belongs to the Section Clinical Pharmaceutics)
Browse Figures
Versions Notes

Abstract

Aim: To evaluate associations between polymorphisms in CYP3A4 (*1B, *22), CYP3A5 (*3), CYP2J2 (*7, rs11572325), ABCB1 (c.1236C>T, c.2677G>T/A, c.3435C>T, rs4148738) and ABCG2 (c.421C>A) and the occurrence of bleeding or occlusive events in patients receiving rivaroxaban in real-world clinical practice. Methods: A nested case-control study, divided into two substudies (bleeding and thromboembolic events), was conducted within a prospective cohort of 385 adults receiving rivaroxaban at University Hospital Centre Zagreb (September 2021–September 2024). Bleeding events were classified per ISTH criteria, and genotyping was performed using TaqMan real-time PCR. Cases and controls were balanced using entropy balancing, and associations were estimated with Bayesian logistic regression under a skeptical prior N(0, 0.355); LASSO regression was used to identify clinical and genetic predictors of outcomes. Results: In total, 71 patients (18.4%) experienced bleeding events, most frequently gastrointestinal (47.9%), while 314 patients served as controls. No pharmacogenomic variant showed a clear association with bleeding risk (raw and balanced odds ratios 0.80–1.35; 95% credible intervals crossing 1.0). LASSO regression identified age (OR 2.00 per decade), gastrointestinal comorbidity (OR 8.77), and eGFR as the dominant predictors of bleeding. Twenty-one patients experienced occlusive events (15 venous, 6 arterial); however, the low event count precluded meaningful pharmacogenomic analysis. Conclusions: Individual pharmacogenomic variants in CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 together with pharmacogenetic-based phenotypes were not associated with clinically relevant bleeding in rivaroxaban-treated patients. Traditional clinical risk factors, particularly advanced age and gastrointestinal comorbidity, remain the dominant determinants of adverse outcomes. Routine pharmacogenomic testing to guide rivaroxaban dosing is not currently supported.

1. Introduction

Direct oral anticoagulants (DOACs)—factor IIa and factor Xa inhibitors—have fundamentally transformed the management of thromboembolic diseases over the past two decades. By offering predictable pharmacokinetics, fixed dosing regimens, and freedom from routine coagulation monitoring, DOACs have largely supplanted vitamin K antagonists (VKAs) for both stroke prevention in non-valvular atrial fibrillation (AF) and the prevention and treatment of venous thromboembolism (VTE) [1,2,3].
Rivaroxaban, the first of the direct Xa inhibitors, is metabolized primarily by hepatic cytochrome P450 enzymes CYP3A4 (~18%) and CYP2J2 (~14%), with additional contributions from CYP3A5, while its intestinal absorption and biliary/renal elimination are regulated by efflux transporters P-glycoprotein (P-gp, encoded by ABCB1) and breast cancer resistance protein (BCRP, encoded by ABCG2) [4,5].
The goal of pharmacogenetics is to identify gene variants, most commonly single nucleotide polymorphisms (SNPs), that underlie interindividual differences in drug response, and to implement pharmacogenetic testing in clinical practice in order to predict therapeutic response, reduce the risk of treatment inefficacy, and minimize the occurrence of adverse drug reactions (ADRs) [6,7].
Single nucleotide polymorphisms (SNPs) in these ADME (absorption, distribution, metabolism, excretion) genes can alter enzyme activity or transporter function, potentially modifying systemic drug exposure and thereby influencing clinical outcomes [8,9].
While the genotype represents an individual’s unique genetic constitution, the phenotype reflects their actual metabolic response. The phenotype results from interactions between the genotype and environmental factors (e.g., diet, smoking, co-medication), and therefore does not always fully correspond to the genotype [10].
Based on the genotype, five phenotypes are distinguished according to the rate and capacity of drug metabolism: poor, intermediate, normal, rapid and ultrarapid metabolizers (PM; IM; NM; RM; UM) [11].
Genome-wide association studies (GWAS) have led to the discovery of numerous variants associated with the development of adverse drug reactions, further increasing the interest of the pharmaceutical industry, regulatory bodies, and the academic community in the genetic basis of variable therapeutic response.
Despite a plausible biological rationale, evidence linking specific pharmacogenomic variants to clinical outcomes in rivaroxaban-treated patients remains heterogeneous and inconsistent [12,13,14,15,16,17].
A 2025 systematic review identified 25 polymorphic loci across ABCB1, ABCG2, CYP3A4, CYP3A5, CYP2J2, and other pharmacogenes associated with altered rivaroxaban pharmacokinetics, yet concluded that evidence is insufficient for routine clinical implementation due to significant inter-study heterogeneity [18]. Furthermore, pharmacokinetic signals do not necessarily translate into measurable clinical effects, particularly for drugs with a relatively wide therapeutic window such as rivaroxaban [19].
Although direct oral anticoagulants (DOACs), including rivaroxaban, demonstrate a more favorable pharmacokinetic profile relative to vitamin K antagonists (VKAs), considerable interindividual variability in plasma drug concentrations remains, the underlying determinants of which have not been fully elucidated. Given the established roles of cytochrome P450 enzymes and membrane transport proteins in the metabolism and elimination of rivaroxaban, it is postulated that genetic variants in CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 may contribute to this variability, as well as to the susceptibility to adverse drug reactions. The objective of the present study was to examine the association between the aforementioned pharmacogenetic variants and the occurrence of adverse drug reactions in patients receiving rivaroxaban therapy, with the aim of advancing the understanding of factors governing individual therapeutic response.
The present study aimed to assess the association of clinically relevant polymorphisms in CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 with bleeding and occlusive (thrombotic) events in a real-world cohort of patients receiving rivaroxaban according to approved indications at a tertiary cardiovascular center in Croatia.
To the best of our knowledge, the present nested case-control study constitutes one of the more comprehensive real-world assessments of multiple pharmacogenomic variants across the key CYP enzymes and ABC transporters implicated in rivaroxaban pharmacokinetics, evaluated against clinically defined bleeding and thromboembolic outcomes.

2. Materials and Methods

2.1. Study Outline

This nested case-control study was conducted at the Department of Cardiovascular Diseases and the Clinical Department of Laboratory Diagnostics, University Hospital Centre Zagreb (UHC Zagreb), Croatia, between September 2021 and September 2024, as part of the Croatian Science Foundation project PGxCardioDrug (UIP-2020-02-8189) [20]. The study was approved by the Ethics Committees of the University of Zagreb School of Medicine (380-59-10106-20-111/125) and UHC Zagreb (02/21 AG). The pharmacogenes of interest were CYP3A4 *1B (rs2740574) and *22 (rs35599367); CYP3A5 *3 (rs776746); CYP2J2 *7 (rs890293) and rs11572325; ABCB1 c.1236C>T (rs1128503), c.2677G>T/A (rs2032582), c.3435C>T (rs1045642), and rs4148738; and ABCG2 c.421C>A (rs2231142). Combined CYP3A phenotype was assigned based on both CYP3A4*22 and CYP3A5*3 allelic status [21], and these phenotypes were further defined according to Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines [11].
The base cohort comprised adult patients treated with rivaroxaban (incident, prevalent) for an indication requiring ≥3 months of treatment (Figure 1). Two substudies/analyses were performed on a case-control principle (Figure 1): (i) Substudy 1 explored association between polymorphisms and bleeding events. Cases were patients who experienced bleeding after at least 1 month of treatment had elapsed. Potential controls were patients who had been on treatment ≥3 months with no bleeding events over the previous period and over a 12-month follow-up. If, during the follow-up, they developed bleeding—they were counted as “cases”, if not, they were considered “controls” (Figure 1); (ii) Substudy 2 explored associations between polymorphisms and occlusive events: cases had experienced such events, controls were free of such events at the time of inclusion and over the subsequent 12-month follow-up.

2.2. Patients, Patient Management and Follow-Up

Eligible were adults with documented indication for rivaroxaban therapy of at least 3 months who provided written informed consent for genotyping of the pharmacogenes of interest, and for the use of anonymized data for research purposes and publication. Patients with significant thrombocytopenia (<100 × 109/L), hepatic dysfunction (Child–Pugh B or C), severe renal impairment (eGFR < 15 mL/min/1.73 m2), and indications in which rivaroxaban is not routinely recommended (mechanical heart valves, rheumatic mitral stenosis, antiphospholipid syndrome) were not included.
For some of the “cases” (either bleeding or occlusive event), the time of enrollment was the time of the index event, whereas for others, as well as for the controls, time of enrollment was the start of the 12-month follow-up. For cases, data were collected at the index event/enrollment, or at the start of the follow-up and at a subsequent index event (and these were considered in the analysis). For controls, data were collected at the start of the follow-up. Collected were data on demographics, indication for and dose of rivaroxaban, comorbidities, concomitant treatments, and the estimate of the renal function: at the start of the rivaroxaban treatment (to guide dose selection), estimated glomerular filtration rate (eGFR) was recorded based on the CKD-EPI creatinine equation [22].
Patients also provided blood samples for genotyping the pharmacogenes of interest. Concomitant medication was evaluated for potential drug–drug interactions at the level of CYP enzymes and transporters (CYP3A4, CYP3A5, CYP2J2, ABCB1, ABCG2) and classified as inducers, inhibitors, substrates based on the LexiComp® [23] and DrugBank databases [24].
Patients were managed as outpatients with visits scheduled in agreement with their main diagnosis and comorbidities, and no procedure beyond the standard of care procedures was undertaken for the purpose of this study. At each visit, detailed and targeted medical histories and—when indicated—clinical/laboratory investigations were taken in order to identify the outcomes of interest. Since practically all included patients originated from the Hospital’s administrative catchment population, the institutional electronic medical records were periodically searched for the same purpose. After the 12-month follow-up had elapsed, all patients were re-checked to confirm their case or control status, either in person or through a telephone contact.
All events were identified descriptively and characterized regarding severity. Bleeding events were classified per International Society on Thrombosis and Haemostasis (ISTH) criteria [25,26] as major (fatal, in critical organ, or causing Hb drop ≥ 20 g/L or requiring ≥2 units of packed red cells) or clinically relevant non-major (CRNM).

2.3. Genotyping

Genomic DNA was extracted from peripheral blood using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). Genotyping of ten SNPs was performed using validated TaqMan SNP/DME Genotyping Assays on a Real-Time PCR System (Applied Biosystems, Foster City, CA, USA): CYP3A4 *1B (rs2740574) and *22 (rs35599367); CYP3A5 *3 (rs776746); CYP2J2 *7 (rs890293) and rs11572325; ABCB1 c.1236C>T (rs1128503), c.2677G>T/A (rs2032582), c.3435C>T (rs1045642), and rs4148738; ABCG2 c.421C>A (rs2231142).

2.4. Statistical Analysis

The main analysis intended to assess associations between individual polymorphisms or genotype-predicted phenotypes and the case status (separately in each substudy): in a series of consecutive analyses (one per polymorphism/phenotype), cases and controls were balanced on a range of covariates except for the polymorphisms/phenotype of interest using entropy balancing, and then (weighted) logistic regression models were fitted to the case status. Entropy balancing balances distributional moments (mean, variance, skewness, kurtosis) between cases and controls without excluding participants. It was performed using package WeightIt [27] in R language and environment for statistical computing (Version 4.5.0) (R Foundation for Statistical Computing). Bayesian logistic models [package rstanarm in R [28]] were fitted to the data with a moderately informed skeptical prior on ln(OR) [N(0, 0.355)] which assigns 95% probability to odds ratios (ORs) between 0.50 and 2.00, in line with the general views on contributions of individual pharmacogenetic polymorphisms to clinical outcomes, and in line with predominantly null findings in previous research. Raw and adjusted (weighted) ORs were generated (95% credible intervals). As a supplementary analysis, all collected clinical and genetic variables were considered as potential predictors of the case status. Least absolute shrinkage and selection operator (LASSO) logistic regression models were used (proc glmselect in SAS 9.4 for Windows, SAS Inc., Cary, NC, USA) to identify variables associated with the outcomes (Akaike’s information criterion). Hardy–Weinberg equilibrium was verified for all loci prior to the analysis [package genetics in R [28]].

3. Results

3.1. Cohort Characteristics

A total of 385 patients (median age 67 years, IQR 58–74); 238 (61.8%) males and 147 (38.2%) females were enrolled. Some of them suffered from several conditions that required rivaroxaban treatment, the most common being nonvalvular atrial fibrillation (AF)/atrial flutter (AFL) (284, 73.8%), followed by chronic coronary syndrome (95, 24.7%), peripheral arterial disease (68, 17.7%), ischaemic stroke/transient ischaemic attack (TIA) (57, 14.8%), and VTE (52, 13.5%). The most common rivaroxaban dose was 20 mg once daily (273, 70.9%). Median eGFR was 75 mL/min/1.73 m2 (IQR 58–89). The most frequent comorbidities were hypertension (82.3%), dyslipidaemia (69.6%), and diabetes mellitus (22.1%).
The majority of participants carried the CYP3A4 wild-type genotype (*1/*1: 96.4% for *1B, 93.8% for *22). CYP3A5*3/*3 (non-expressor) genotype predominated (89.6%) (Table 1). The combined CYP3A phenotype was intermediate metabolizer (IM) in 83.6%, rapid metabolizer (RM) in 10.1%, and poor metabolizer (PM) in 6.2% (Table 1). CYP2J2*7 was observed in 11.9%. For ABCG2 c.421C>A and CYP2J2 c.A>T (rs11572325), 18.5% and 21.5%, respectively, carried at least one variant allele. The ABCB1 variant allele frequencies were substantially higher (40.1–53.9%) (Table 1).

3.2. Bleeding Events

Seventy-one patients (18.4%) experienced at least one bleeding event (cases), with 74 total episodes (three patients had two concurrent manifestations). Six major bleeds were recorded (5 intracranial hemorrhages; 1 hemopericardium), whereas the most frequent non-major bleeding sites were gastrointestinal (34, 47.9%) (Figure 2).
Duration of rivaroxaban treatment was longer in controls than in cases (median 24 months, range 15–108 vs. median 15 months, range 3–96) (see also Supplementary Material Figure S1). Cases and controls somewhat differed in a range of demographic and comorbidity characteristics (Table 2), but only a few differences appeared marked: cases were older (74 vs. 66 years) and had lower eGFR (69 vs. 76 mL/min/1.73 m2) and higher prevalence of gastrointestinal comorbidities (45.1 vs. 7.6%) and malignancy (28.1 vs. 16.3%), more commonly used the 15 mg/day dose (28.2 vs. 10.5%), and less commonly used the 20 mg/day dose (57.8 vs. 73.2%) (Table 2).
The use of relevant comedication appeared similar in cases and controls (Table 3).
Also, prevalence of the investigated polymorphisms and genotype-predicted phenotypes appeared similar in cases and controls (Table 4).
Since the number of patients with a poor metabolizer CYP3A phenotype and with reduced-function phenotype based on CYP2J*7 was low, relationship between these polymorphisms and bleeding could not be meaningfully evaluated. For the remaining six polymorphisms, dichotomized to variant allele carriage and wild type genotype, entropy balancing enabled perfect balance between cases and controls on a number of covariates (Supplementary Material Tables S1–S6). Differences in age, eGFR and prevalence of gastrointestinal comorbidities were such that balance could not be achieved; hence, these three variables were used as covariates in the analysis of raw and balanced (weighted) data: neither raw nor the fully adjusted analyses indicated association of any of these polymorphisms with the risk of bleeding (Table 5).
In the supplementary analysis, among the range of demographic, clinical and genetic variables, the LASSO regression selected only three variables associated with the outcome: age, where older age was associated with a higher odds of bleeding (OR 2.00, 95%CI 1.34–2.99 per 10 years); gastrointestinal comorbidity, strongly associated with the risk of bleeding (OR = 8.77, 95%CI 4.49–17.1); and eGFR, where higher eGFR tended to be associated with a lower risk of bleeding. Figure 3 illustrates the discrepancy between the strong associations of age and gastrointestinal comorbidity with the risk of bleeding, and the lack of association between the investigated polymorphisms and the outcome.

3.3. Occlusive Events

A total of 21 patients experienced occlusive events: 15 venous thromboembolic events (10 DVT, 5 PE) and 6 arterial occlusions (3 ischemic strokes, 2 TIA, 1 peripheral arterial occlusion). The number of cases was too low for a meaningful analysis of associations between the polymorphisms of interest and this outcome.

4. Discussion

In this real-world study, no clear or consistent association was observed between the investigated pharmacogenetic variants and bleeding or thromboembolic outcomes in rivaroxaban-treated patients. In contrast, advanced age, gastrointestinal comorbidity, and renal function emerged as the main predictors of bleeding risk, suggesting that clinical factors may be more important determinants of adverse outcomes than individual genetic variants. Prevalence of genotypes and genotype-predicted phenotypes in this study cohort corresponds to those reported for European reference populations, including the marked predominance of CYP3A5 non-expressors (*3/*3; 89.6%) [29,30], low variant allele frequency of CYP2J2*7 (6.0%) concordant with 5.49% observed in Caucasians [31].
Prevalence of ABCG2 c.421 C>A (9.6%) correlates with previous study of this polymorphism in the Croatian population [32] and aligns with European reference data (∼5–10%) [33] together with the common ABCB1 variants (40.1–53.9%) [34]. The distribution of the analyzed polymorphisms was consistent with European populations.
Due to the marked predominance of the CYP3A rapid/intermediate metabolizer phenotype and the CYP2J2 normal phenotype according to the *7 allele among cases and controls, it was not possible to reliably assess their association with bleeding risk, and for other evaluated variants neither raw nor the fully adjusted analyses indicated association of any of these polymorphisms with the risk of bleeding reflecting that no polymorphism showed a significant independent association with the outcomes.
Our findings thus suggest that, in patients treated with rivaroxaban at standard guideline-recommended dosing, the risk of adverse events is determined primarily by clinical predictors, whereas the contribution of individual pharmacogenetic variants is likely limited.
Results of this study showed that age, gastrointestinal comorbidity, and eGFR were the only variables identified by the LASSO model as relevant predictors, with older age and gastrointestinal comorbidity being clearly associated with a higher risk of adverse events during rivaroxaban treatment.
In Substudy 1 (Bleeding) cases were older (median 74 vs. 66 years) and had a substantially higher prevalence of gastrointestinal comorbidities (45.1% vs. 7.6%) and malignancy (28.1% vs. 16.3%). The 15 mg/day dose was more frequent among cases (28.2% vs. 10.5%), reflecting confounding by indication. Pharmacogenomic variant frequencies did not differ materially between cases and controls. This finding is expected and consistent with known biological mechanisms. Older age is associated with multiple comorbidities, including the increasing prevalence of cancer in the elderly, alterations in drug pharmacokinetics and pharmacodynamics, increased propensity for bleeding [35,36], whereas gastrointestinal disorders provide a direct anatomical and pathophysiological substrate for bleeding [37].
In Substudy 2 (Occlusive Events), the low event count, just 21 thromboembolic cases, precluded reliable pharmacogenomic analysis. In addition, no pharmacogenomic variant showed a clear signal (all OR 0.93–1.19; 95% CrIs inclusive of 1.0). Nevertheless, the low number of thromboembolic events—only 21 in the cohort (10 DVT, 5 PE, and 6 arterial occlusions: 3 ischemic strokes, 2 TIAs, and 1 peripheral arterial occlusion) among 385 patients—suggests the efficacy and effectiveness of rivaroxaban for its most common indication, namely the prevention of venous and systemic thromboembolism, even at the cost of six major bleeding events (five ICH, one hemopericardium).
Existing studies evaluating the pharmacogenetics of rivaroxaban have produced inconsistent findings. In the present study, neither CYP3A4 nor CYP2J2 variants were associated with clinical outcomes.
Although CYP3A4*22 variant has well-established functional consequences for several drugs, including tacrolimus and statins [38], its impact on rivaroxaban remains unclear. Previous studies have reported conflicting results: Li et al. identified associations between CYP3A4 variants and minor bleeding in an AF cohort [13], whereas Wu et al. [12] found no such effect. Likewise, CYP2J2*7 was not associated with clinical outcomes in our cohort, despite the recognized contribution to rivaroxaban metabolism [4]. CYP2J2 additionally metabolizes arachidonic acid to cardioprotective epoxyeicosatrienoic acids, introducing potential pleiotropy that may complicate genotype–outcome relationships [39].
Studies investigating ABCB1 polymorphisms in DOAC-treated patients have yielded inconsistent results [15,18,40,41,42]. Wang et al. identified an association between the ABCB1 c.3435T/T with increased risk of VTE [15], while a Korean study reported an association between ABCB1 rs1045642 and bleeding events [14]. In contrast, neither individual ABCB1 variants nor dominant haplotype were associated with bleeding or thromboembolic outcomes in our cohort. The two most prevalent haplotypes both predicted net-normal transporter function, reducing the likelihood of detecting a haplotype-mediated pharmacokinetic effect.
The ABCG2 c.421C>A variant is known to reduce BCRP-mediated efflux and is associated with higher rivaroxaban exposure, with higher AUC and Cmax values in pharmacokinetic studies [43,44], yet consistent translation to clinical bleeding risk has not been demonstrated [12,18,45], which is in line with the findings of the present study.
Our findings are broadly consistent with several recent studies. The retrospective cohort by Campos-Staffico et al., comprising 2364 patients and evaluating eight pharmacokinetic variants across ABCB1, ABCG2, CYP2J2, and CYP3A4/5, likewise found no significant association between pharmacogenomic profile and DOAC-related bleeding [19], while the prospective multicenter analysis of Wu et al. reached the same conclusion for the same gene panel [12]. Furthermore, the recent systematic review concluded that the current evidence remains insufficient to support routine pharmacogenomic-guided rivaroxaban dosing [18], an interpretation consistent with our nested case-control cohort. The inconsistencies among published studies likely reflect heterogeneity in study design, outcome definition, ethnic allele-frequency distribution, and sample size, and highlight the difficulty of detecting modest pharmacogenomic effects in the presence of dominant clinical determinants. In our cohort, advanced age and gastrointestinal comorbidity emerged as the principal predictors of bleeding, reinforcing the view that clinical risk stratification, not genotype, currently provides the more actionable basis for individualizing rivaroxaban therapy. Comparison of key pharmacogenomic studies evaluating rivaroxaban clinical outcomes is shown in Table 6.
The absence of a detectable pharmacogenomic effect may be explained by several biological and methodological factors. First, rivaroxaban undergoes metabolism via several parallel pathways, so, reduced activity of a single enzyme or transporter may be compensated by alternative routes, attenuating the net pharmacokinetic impact of any single genetic variant [19]. Second, unlike warfarin, whose narrow therapeutic index renders pharmacogenomic variability clinically decisive, rivaroxaban has a relatively broad therapeutic window, meaning that modest concentration changes driven by individual SNPs are unlikely to generate a clinically relevant association [19]. Third, the number of participants carrying rare genotypes was small, particularly for CYP3A4 and CYP2J2 variants, limiting the statistical power to detect associations.
Our study has several limitations. First, the research was conducted at a single center. Although the study had sufficient statistical power for the primary endpoint, bleeding (71 cases), the number of occlusive events (N = 21) was too small to allow reliable assessment of pharmacogenetic associations. Consequently, substantially larger multicenter studies will be required to adequately evaluate pharmacogenomic determinants of occlusive outcomes.
Additionally, plasma rivaroxaban concentrations and anti-Xa activity were not measured, precluding direct assessment of genotype–pharmacokinetics relationships, while treatment adherence was not directly quantified. Measurement of plasma rivaroxaban concentrations was not included in the original study protocol, which was designed to evaluate pharmacogenetic associations with clinical outcomes rather than pharmacokinetic parameters. Furthermore, accurate assessment of rivaroxaban plasma concentrations requires strictly time-controlled blood sampling (e.g., at peak [Cmax] or trough [Ctrough] concentrations), which is logistically challenging in routine clinical practice.
Consequently, the lack of pharmacokinetic data limits mechanistic interpretation of the observed genotype–outcome relationships. Previous pharmacokinetic studies have demonstrated that certain variants, particularly ABCG2 c.421C>A, are associated with increased rivaroxaban exposure [43,44], whereas consistent associations between these pharmacokinetic differences and bleeding or thrombotic outcomes have not been demonstrated [18]. Similarly, recent studies evaluating CYP3A4/5, ABCB1, ABCG2 and CYP2J2 variants have reported modest effects on rivaroxaban pharmacokinetics but absent effects on clinical outcomes, particularly bleeding risk [12]. In line with these observations, Nakagawa et al. concluded that the dose-adjusted plasma trough concentration ratio (C0h/D) of rivaroxaban did not differ significantly among ABCB1 c.3435C>T, c. 2677 G>A/T, c.1236 C>T, ABCG2 c.421C>A, CYP3A5*3 and CYP2J2*7 genotypes [17].
However, to our knowledge, this nested case–control study represents one of the larger real-world evaluations of multiple pharmacogenomic variants in the principal CYP enzymes and ABC transporters involved in rivaroxaban disposition, assessed against clinically defined hemorrhagic and thrombotic outcomes. Despite a biologically plausible rationale and the application of advanced statistical methods, no clear or consistent pharmacogenomic signal was identified. Clinical factors, particularly advanced age (OR 2.00 per decade) and gastrointestinal comorbidity (OR 8.77 for bleeding), far outweighed any genetic contribution in determining outcome risk.
Our results align with extensive real-world evidence showing that rivaroxaban’s GI bleeding risk is driven primarily by patient characteristics [37], thereby supporting current guideline recommendations. On the other hand, the recent COBRRA randomized trial demonstrated that the choice of DOAC itself is a powerful modifiable determinant of bleeding: among patients with acute VTE, apixaban approximately halved clinically relevant bleeding compared with rivaroxaban (3.3% vs. 7.1%; RR 0.46, 95% CI 0.33–0.65) without compromising efficacy [46].
Additional interesting findings come from network meta-analysis comparing real-world effectiveness and safety of all DOACs (apixaban, dabigatran, edoxaban, rivaroxaban) vs. VKAs in European NVAF patients. DOACs showed benefit over VKAs for most outcomes, of which major bleeding and all-cause mortality were most commonly reported. Edoxaban demonstrated a comparable effectiveness/safety profile to other DOACs and significantly reduced risk of major bleeding (hazard ratio [95% credible interval]: 0.67 [0.54, 0.84]) and intracranial hemorrhage (0.69 [0.51, 0.94]) versus rivaroxaban [47]. In light of these findings, apixaban is associated with a more favorable safety profile than rivaroxaban in patients with VTE, whereas edoxaban demonstrates a more favorable safety profile than rivaroxaban in patients with AF, as both DOACs are associated with a lower risk of bleeding.
Validated clinical risk scores, such as HAS-BLED [48] and ORBIT [49], which incorporate age, renal function, and GI disease status should guide rivaroxaban prescribing and monitoring, independent of pharmacogenomic testing. In the present study, variant carrier status was not associated with bleeding or occlusive events. This discrepancy between pharmacogenetic and clinical findings highlights that drug exposure alone does not determine outcome risk within standard dosing regimens. Consequently, our results do not support routine testing of the analyzed variants to guide rivaroxaban dosing or patient selection in current clinical practice.
Nevertheless, pharmacogenetic testing may still prove helpful in selected, high-risk situations (e.g., with unexpectedly severe bleeding, atypical therapeutic response or complex polypharmacy). However, further large, prospective studies integrating pharmacokinetic biomarkers, polygenic approaches, and drug–drug–gene interactions are needed before implementation in routine clinical practice can be considered.

5. Conclusions

In conclusion, our results suggest that routine testing of the analyzed variants is unlikely to improve risk stratification in rivaroxaban-treated patients. In contrast, traditional clinical risk factors (older age and gastrointestinal comorbidity) remain the dominant determinants of adverse outcomes, primarily bleeding risk. None of the investigated polymorphisms showed a significant independent association with clinical outcomes.
Future research should integrate pharmacokinetic biomarkers with polygenic risk approaches and drug–drug–gene interaction analyses in larger, multicenter prospective cohorts to improve predictive performance compared with the single-variant approach used in the present study.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/pharmaceutics18070884/s1, Figure S1. Distribution of time on rivaroxaban for cases (experienced bleeding) and controls (no bleeding). Table S1. Major characteristics of cases (bleeding) and controls (no bleeding) after covariate balancing for the purpose of assessment of the relationship between polymoprhism CYP2J2A>T and bleeding. Data are weighted percentages. Standardized mean differences (d) < 0.1 indicates adequate balance. Table S2. Major characteristics of cases (bleeding) and controls (no bleeding) after covariate balancing for the purpose of assessment of the relationship between polymoprhism ABCG2 421C>A and bleeding. Data are weighted percentages. Standardized mean differences (d) < 0.1 indicates adequate balance. Table S3. Major characteristics of cases (bleeding) and controls (no bleeding) after covariate balancing for the purpose of assessment of the relationship between polymoprhism ABCB1 1236C>T and bleeding. Data are weighted percentages. Standardized mean differences (d) < 0.1 indicates adequate balance. Table S4. Major characteristics of cases (bleeding) and controls (no bleeding) after covariate balancing for the purpose of assessment of the relationship between polymoprhism ABCB1 2677G>T/A and bleeding. Data are weighted percentages. Standardized mean differences (d) < 0.1 indicates adequate balance. Table S5. Major characteristics of cases (bleeding) and controls (no bleeding) after covariate balancing for the purpose of assessment of the relationship between polymoprhism ABCB1 3435C>T and bleeding. Data are weighted percentages. Standardized mean differences (d) < 0.1 indicates adequate balance. Table S6. Major characteristics of cases (bleeding) and controls (no bleeding) after covariate balancing for the purpose of assessment of the relationship between polymoprhism ABCB1 2482-2236G>A and bleeding. Data are weighted percentages. Standardized mean differences (d) < 0.1 indicates adequate balance.

Author Contributions

Conceptualization: V.T., T.B., J.B. and L.Š. Methodology: A.M.S., V.T., L.G., T.B., J.B. and L.Š. Formal analysis: V.T. Funding acquisition: T.B. and I.M. Investigation: A.M.S., L.G., V.P., M.V.K., J.P., D.S., M.N., I.S.M., I.M., J.B. and L.Š. Data curation: A.M.S., L.G., J.P. and L.Š. Resources: A.M.S., L.G., V.P., M.V.K., I.M., J.B. and L.Š. Visualization: A.M.S. Writing—original draft preparation: A.M.S. Writing—review and editing: V.T., L.G., T.B., V.P., M.V.K., J.P., D.S., M.N., I.S.M., I.M., J.B. and L.Š. Supervision: J.B. and L.Š. Project administration: T.B. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Croatian Science Foundation, project PGxCardioDrug (UIP-2020-02-8189).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board; Ethics Committee of the University of Zagreb School of Medicine (380-59-10106-20-111/125) and UHC Zagreb (02/21 AG).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The original contributions presented in this study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Acknowledgments

The authors would like to acknowledge the support provided by Vladimir Trkulja for comprehensive statistical analysis.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Heestermans, M.; Poenou, G.; Hamzeh-Cognasse, H.; Cognasse, F.; Bertoletti, L. Anticoagulants: A Short History, Their Mechanism of Action, Pharmacology, and Indications. Cells 2022, 11, 3214. [Google Scholar] [CrossRef] [PubMed]
  2. Van Gelder, I.C.; Rienstra, M.; Bunting, K.V.; Casado-Arroyo, R.; Caso, V.; Crijns, H.J.G.M.; De Potter, T.J.R.; Dwight, J.; Guasti, L.; Hanke, T.; et al. 2024 ESC Guidelines for the management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur. Heart J. 2024, 45, 3314–3414. [Google Scholar] [CrossRef] [PubMed]
  3. Konstantinides, S.V.; Meyer, G.; Becattini, C.; Bueno, H.; Geersing, G.J.; Harjola, V.P.; Huisman, M.V.; Humbert, M.; Jennings, C.S.; Jiménez, D.; et al. 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur. Respir. J. 2019, 54, 1901647. [Google Scholar] [CrossRef] [PubMed]
  4. Mueck, W.; Stampfuss, J.; Kubitza, D.; Becka, M. Clinical Pharmacokinetic and Pharmacodynamic Profile of Rivaroxaban. Clin. Pharmacokinet. 2014, 53, 1–16. [Google Scholar] [CrossRef] [PubMed]
  5. Kreutz, R. Pharmacodynamic and pharmacokinetic basics of rivaroxaban. Fundam. Clin. Pharmacol. 2012, 26, 27–32. [Google Scholar] [CrossRef] [PubMed]
  6. Spear, B.B.; Heath-Chiozzi, M.; Huff, J. Clinical application of pharmacogenetics. Trends Mol. Med. 2001, 7, 201–204. [Google Scholar] [CrossRef] [PubMed]
  7. Ayalew, M.B.; Tegegn, H.G.; Abdela, O.A. Drug Related Hospital Admissions; A Systematic Review of the Recent Literatures. Bull. Emerg. Trauma 2019, 7, 339–346. [Google Scholar] [CrossRef] [PubMed]
  8. Ahmed, S.; Zhou, Z.; Zhou, J.; Chen, S.Q. Pharmacogenomics of Drug Metabolizing Enzymes and Transporters: Relevance to Precision Medicine. Genom. Proteom. Bioinform. 2016, 14, 298–313. [Google Scholar] [CrossRef] [PubMed]
  9. Zhou, Y.; Lauschke, V.M. The genetic landscape of major drug metabolizing cytochrome P450 genes-an updated analysis of population-scale sequencing data. Pharmacogenom. J. 2022, 22, 284–293. [Google Scholar] [CrossRef] [PubMed]
  10. Shenfield, G.M. Genetic polymorphisms, drug metabolism and drug concentrations. Clin. Biochem. Rev. 2004, 25, 203–206. [Google Scholar] [PubMed]
  11. Caudle, K.E.; Dunnenberger, H.M.; Freimuth, R.R.; Peterson, J.F.; Burlison, J.D.; Whirl-Carrillo, M.; Scott, S.A.; Rehm, H.L.; Williams, M.S.; Klein, T.E.; et al. Standardizing terms for clinical pharmacogenetic test results: Consensus terms from the Clinical Pharmacogenetics Implementation Consortium (CPIC). Genet. Med. 2017, 19, 215–223. [Google Scholar] [CrossRef] [PubMed]
  12. Wu, T.; Wu, S.; Li, L.; Xiang, J.; Wang, N.; Chen, W.; Zhang, J. The impact of ABCB1, CYP3A4/5 and ABCG2 gene polymorphisms on rivaroxaban trough concentrations and bleeding events in patients with non-valvular atrial fibrillation. Hum. Genom. 2023, 17, 59. [Google Scholar] [CrossRef] [PubMed]
  13. Li, X.; Gu, Z.; Wang, Z.; Xu, Q.; Ma, C.; Lv, Q. Mutant CYP3A4/5 Correlated with Clinical Outcomes by Affecting Rivaroxaban Pharmacokinetics and Pharmacodynamics in Patients with Atrial Fibrillation. Cardiovasc. Drugs Ther. 2024, 38, 1315–1325. [Google Scholar] [CrossRef] [PubMed]
  14. Kim, H.; Song, T.J.; Yee, J.; Kim, D.H.; Park, J.; Gwak, H.S. ABCG2 Gene Polymorphisms May Affect the Bleeding Risk in Patients on Apixaban and Rivaroxaban. Drug Des. Dev. Ther. 2023, 17, 2513–2522. [Google Scholar] [CrossRef] [PubMed]
  15. Wang, F.; Li, Z.; Huang, Y.; Liu, Q.; Zhao, L.; Wang, H.; Gao, H.; Chen, M.; Lin, Y.; Li, X.; et al. Effect of ABCB1 SNP polymorphisms on the plasma concentrations and clinical outcomes of rivaroxaban in Chinese NVAF patients: A population pharmacokinetic-based study. Front. Pharmacol. 2025, 16, 1574949. [Google Scholar] [CrossRef] [PubMed]
  16. Ma, Y.; Song, Z.; Li, X.; Jiang, D.; Zhao, R.; Yi, Z. Toward Genetic Testing of Rivaroxaban? Insights from a Systematic Review on the Role of Genetic Polymorphism in Rivaroxaban Therapy. Clin. Pharmacokinet. 2024, 63, 279–291. [Google Scholar] [CrossRef] [PubMed]
  17. Nakagawa, J.; Kinjo, T.; Iizuka, M.; Ueno, K.; Tomita, H.; Niioka, T. Impact of gene polymorphisms in drug-metabolizing enzymes and transporters on trough concentrations of rivaroxaban in patients with atrial fibrillation. Basic Clin. Pharmacol. Toxicol. 2021, 128, 297–304. [Google Scholar] [CrossRef] [PubMed]
  18. Wang, L.; Chen, G.; Hu, W.; Chen, J.; He, Y. Effect of Gene Polymorphism on the Pharmacokinetics and Clinical Outcomes of Rivaroxaban: State-of-the-Art Review. Drug Des. Dev. Ther. 2025, 19, 7321–7331. [Google Scholar] [CrossRef] [PubMed]
  19. Campos-Staffico, A.M.; Dorsch, M.P.; Barnes, G.D.; Zhu, H.J.; Limdi, N.A.; Luzum, J.A. Eight pharmacokinetic genetic variants are not associated with the risk of bleeding from direct oral anticoagulants in non-valvular atrial fibrillation patients. Front. Pharmacol. 2022, 13, 1007113. [Google Scholar] [CrossRef] [PubMed]
  20. Božina, T.; Ganoci, L.; Šimičević, L.; Kirhmajer, M.V.; Mucalo, I.; Samardžić, J.; Palić, J.; Slišković, A.M.; Trkulja, V. Pharmacogenomics in the prediction of adverse effects of cardiovascular drugs: The PGx-CardioDrug project. Croat. Med. J. 2025, 66, 446–455. [Google Scholar] [CrossRef] [PubMed]
  21. Lloberas, N.; Elens, L.; Llaudó, I.; Padullés, A.; van Gelder, T.; Hesselink, D.A.; Colom, H.; Andreu, F.; Torras, J.; Bestard, O.; et al. The combination of CYP3A4*22 and CYP3A5*3 single-nucleotide polymorphisms determines tacrolimus dose requirement after kidney transplantation. Pharmacogenet. Genom. 2017, 27, 313–322. [Google Scholar] [CrossRef] [PubMed]
  22. Levey, A.S.; Stevens, L.A.; Schmid, C.H.; Zhang, Y.L.; Castro, A.F., 3rd; Feldman, H.I.; Kusek, J.W.; Eggers, P.; Van Lente, F.; Greene, T.; et al. A new equation to estimate glomerular filtration rate. Ann. Intern Med. 2009, 150, 604–612. [Google Scholar] [CrossRef] [PubMed]
  23. Elovic, A.; Pourmand, A. Lexicomp App Review. J. Digit. Imaging 2020, 33, 17–20. [Google Scholar] [CrossRef] [PubMed]
  24. Wishart, D.S.; Feunang, Y.D.; Guo, A.C.; Lo, E.J.; Marcu, A.; Grant, J.R.; Sajed, T.; Johnson, D.; Li, C.; Sayeeda, Z.; et al. DrugBank 5.0: A major update to the DrugBank database for 2018. Nucleic Acids Res. 2018, 46, D1074–D1082. [Google Scholar] [CrossRef] [PubMed]
  25. Schulman, S.; Kearon, C.; The Subcommittee on Control of Anticoagulation of the Scientific and Standardization Committee of the International Society on Thrombosis and Haemostasis. Definition of major bleeding in clinical investigations of antihemostatic medicinal products in non-surgical patients. J. Thromb. Haemost. 2005, 3, 692–694. [Google Scholar] [CrossRef] [PubMed]
  26. Kaatz, S.; Ahmad, D.; Spyropoulos, A.C.; Schulman, S. Definition of clinically relevant non-major bleeding in studies of anticoagulants in atrial fibrillation and venous thromboembolic disease in non-surgical patients: Communication from the SSC of the ISTH. J. Thromb. Haemost. 2015, 13, 2119–2126. [Google Scholar] [CrossRef] [PubMed]
  27. Gerifer, N. WeightIt: Weighting for Covariate Balance in Observational Studies Package Version 1.5.2026. Available online: https://ngreifer.github.io/WeightIt/ (accessed on 5 March 2026).
  28. Goodrich, B.; Gnaby, J.; Ali, I.; Brilleman, S. Rstanarm. Bayesian Applied Regression Modeling via Stan. R Package 2.32.2 2025. Available online: https://mc-stan.org/rstanarm/ (accessed on 5 March 2026).
  29. van Schaik, R.H.; van der Heiden, I.P.; van den Anker, J.N.; Lindemans, J. CYP3A5 Variant Allele Frequencies in Dutch Caucasians. Clin. Chem. 2002, 48, 1668–1671. [Google Scholar] [CrossRef]
  30. Birdwell, K.; Decker, B.; Barbarino, J.; Peterson, J.; Stein, C.; Sadee, W.; Wang, D.; Vinks, A.; He, Y.; Swen, J.; et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) Guidelines for CYP3A5 Genotype and Tacrolimus Dosing. Clin. Pharmacol. Ther. 2015, 98, 19–24. [Google Scholar] [CrossRef] [PubMed]
  31. Wang, H.; Jiang, Y.; Liu, Y.; Lin, C.; Cheng, G.; Chen, X.; Hao, B.; Tan, W.; Lin, D.; He, F. CYP2J2*7 single nucleotide polymorphism in a Chinese population. Clin. Chim. Acta 2006, 365, 125–128. [Google Scholar] [CrossRef] [PubMed]
  32. Božina, T.; Ganoci, L.; Karačić, E.; Šimičević, L.; Vrkić-Kirhmajer, M.; Klarica-Domjanović, I.; Križ, T.; Sertić, Z.; Božina, N. ABCG2 and SLCO1B1 gene polymorphisms in the Croatian population. Ann. Hum. Biol. 2022, 49, 323–331. [Google Scholar] [CrossRef] [PubMed]
  33. Hira, D.; Terada, T. BCRP/ABCG2 and high-alert medications: Biochemical, pharmacokinetic, pharmacogenetic, and clinical implications. Biochem. Pharmacol. 2018, 147, 201–210. [Google Scholar] [CrossRef] [PubMed]
  34. Jeannesson, E.; Albertini, L.; Siest, G.; Gomes, A.; Ribeiro, V.; Aslanidis, C.; Schmitz, G.; Visvikis-Siest, S. Determination of ABCB1 polymorphisms and haplotypes frequencies in a French population. Fundam. Clin. Pharmacol. 2007, 21, 411–418. [Google Scholar] [CrossRef] [PubMed]
  35. Prathap, R.; Kirubha, S.; Rajan, A.T.; Manoharan, S.; Elumalai, K. The increasing prevalence of cancer in the elderly: An investigation of epidemiological trends. AGING Med. 2024, 7, 516–527. [Google Scholar] [CrossRef] [PubMed]
  36. Edwina, A.E.; Dia, N.; Dreesen, E.; Vanassche, T.; Verhamme, P.; Spriet, I.; Van der Linden, L.; Tournoy, J. Insights into the Pharmacokinetics and Pharmacodynamics of Direct Oral Anticoagulants in Older Adults with Atrial Fibrillation: A Structured Narrative Review. Clin. Pharmacokinet. 2023, 62, 351–373. [Google Scholar] [CrossRef] [PubMed]
  37. Benamouzig, R.; Guenoun, M.; Deutsch, D.; Fauchier, L. Review Article: Gastrointestinal Bleeding Risk with Direct Oral Anticoagulants. Cardiovasc. Drugs Ther. 2022, 36, 973–989. [Google Scholar] [CrossRef] [PubMed]
  38. Mulder, T.A.M.; van Eerden, R.A.G.; de With, M.; Elens, L.; Hesselink, D.A.; Matic, M.; Bins, S.; Mathijssen, R.H.J.; van Schaik, R.H.N. CYP3A4∗22 Genotyping in Clinical Practice: Ready for Implementation? Front. Genet. 2021, 12, 711943. [Google Scholar] [CrossRef] [PubMed]
  39. Berlin, D.S.; Sangkuhl, K.; Klein, T.E.; Altman, R.B. PharmGKB summary: Cytochrome P450, family 2, subfamily J, polypeptide 2: CYP2J2. Pharmacogenet. Genom. 2011, 21, 308–311. [Google Scholar] [CrossRef] [PubMed]
  40. Sychev, D.A.; Sokolov, A.V.; Reshetko, O.V.; Fisenko, V.P.; Sychev, I.N.; Grishina, E.A.; Bochkov, P.O.; Shevchenko, R.V.; Abdullaev, S.P.; Denisenko, N.P.; et al. Influence of ABCB1, CYP3A5 and CYP3A4 gene polymorphisms on prothrombin time and the residual equilibrium concentration of rivaroxaban in patients with non-valvular atrial fibrillation in real clinical practice. Pharmacogenet. Genom. 2022, 32, 301–307. [Google Scholar] [CrossRef] [PubMed]
  41. Lähteenmäki, J.; Vuorinen, A.; Pajula, J.; Harno, K.; Lehto, M.; Niemi, M.; van Gils, M. Pharmacogenetics of Bleeding and Thromboembolic Events in Direct Oral Anticoagulant Users. Clin. Pharmacol. Ther. 2021, 110, 768–776. [Google Scholar] [CrossRef] [PubMed]
  42. Gouin-Thibault, I.; Delavenne, X.; Blanchard, A.; Siguret, V.; Salem, J.E.; Narjoz, C.; Gaussem, P.; Beaune, P.; Funck-Brentano, C.; Azizi, M.; et al. Interindividual variability in dabigatran and rivaroxaban exposure: Contribution of ABCB1 genetic polymorphisms and interaction with clarithromycin. J. Thromb. Haemost. 2017, 15, 273–283. [Google Scholar] [CrossRef] [PubMed]
  43. Guo, L.; Sun, X.; Bo, Q.; Bai, W.; Du, Y.; Song, H. Effects of ABCG2 > A genetic polymorphism on the pharmacokinetics of rivaroxaban in healthy Chinese subjects. Xenobiotica 2025, 55, 445–453. [Google Scholar] [CrossRef] [PubMed]
  44. dos Santos, A.F.; Francisco, Q.A.S.; Nunes, J.B.; Colombo, F.A.; Boralli, V.B. ABCG2 polymorphism and rivaroxaban pharmacokinetics in healthy individuals after a single dose. Braz. J. Med. Biol. Res. 2024, 57, e13257. [Google Scholar] [CrossRef] [PubMed]
  45. Li, H.; He, X.; Guan, Y.; Lu, Q.; Fang, J.; Bian, X. Association of Genetic Variants with Rivaroxaban Pharmacokinetics and Pharmacodynamics and Hemorrhage Risk Factors in Patients with Venous Thromboembolism. J. Am. Heart Assoc. 2025, 14, e040698. [Google Scholar] [CrossRef] [PubMed]
  46. Castellucci, L.A.; Chen, V.M.; Kovacs, M.J.; Lazo-Langner, A.; Greenstreet, P.; Kahn, S.; Côté, B.; Schulman, S.; de Wit, K.; Douketis, J.; et al. Bleeding Risk with Apixaban vs. Rivaroxaban in Acute Venous Thromboembolism. N. Engl. J. Med. 2026, 394, 1051–1060. [Google Scholar] [CrossRef] [PubMed]
  47. Cichewicz, A.; Jawla, S.; Zuchinali, P.; Akin-Fajiye, M.; Massierer, D.; Porto, I.; Garcia-Moll, X. Real-world evidence comparing oral anticoagulants for NVAF in Europe: A systematic review and network meta-analysis. Future Cardiol. 2025, 21, 371–390. [Google Scholar] [CrossRef] [PubMed]
  48. Pisters, R.; Lane, D.A.; Nieuwlaat, R.; de Vos, C.B.; Crijns, H.J.G.M.; Lip, G.Y.H. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: The Euro Heart Survey. Chest 2010, 138, 1093–1100. [Google Scholar] [CrossRef] [PubMed]
  49. Liu, X.; Wang, S.; He, W.; Guo, L. HAS-BLED vs. ORBIT scores in anticoagulated patients with atrial fibrillation: A systematic review and meta-analysis. Front. Cardiovasc. Med. 2022, 9, 1042763. [Google Scholar] [CrossRef] [PubMed]
Pharmaceutics 18 00884 g001
Figure 1. Study outline. Adults requiring ≥3 months of rivaroxaban treatment were followed up to occurrence of bleeding or occlusive events, and over a subsequent 12-month follow-up (a patient could have experienced more than one of either or both events). Data were used in two substudies based on the case-control principle: (i) Substudy 1 explored polymorphisms of interest with respect to bleeding events. Cases were patients treated for at least 1 month who developed bleeding. Controls were selected among patients with at least 3 months of a previous treatment without bleeding based on developments over a subsequent 12-month follow-up: if bleeding occurred, they were considered cases, if not, they were considered controls; (ii) Substudy 2 explored polymorphisms with respect to occlusive events. Cases and controls were identified on the same principles as in Substudy 1.
Figure 1. Study outline. Adults requiring ≥3 months of rivaroxaban treatment were followed up to occurrence of bleeding or occlusive events, and over a subsequent 12-month follow-up (a patient could have experienced more than one of either or both events). Data were used in two substudies based on the case-control principle: (i) Substudy 1 explored polymorphisms of interest with respect to bleeding events. Cases were patients treated for at least 1 month who developed bleeding. Controls were selected among patients with at least 3 months of a previous treatment without bleeding based on developments over a subsequent 12-month follow-up: if bleeding occurred, they were considered cases, if not, they were considered controls; (ii) Substudy 2 explored polymorphisms with respect to occlusive events. Cases and controls were identified on the same principles as in Substudy 1.
Pharmaceutics 18 00884 g001
Pharmaceutics 18 00884 g002
Figure 2. Bleeding events by anatomical site and severity (n  =  71 patients, 74 events). Bars show clinically relevant non-major (CRNM) (orange) and major (dark red) bleeding by site; event counts (n) and site percentages annotated above each bar. GI—gastrointestinal.
Figure 2. Bleeding events by anatomical site and severity (n  =  71 patients, 74 events). Bars show clinically relevant non-major (CRNM) (orange) and major (dark red) bleeding by site; event counts (n) and site percentages annotated above each bar. GI—gastrointestinal.
Pharmaceutics 18 00884 g002
Pharmaceutics 18 00884 g003
Figure 3. Comparative effect sizes of clinical predictors versus pharmacogenomic variants for bleeding outcome (log-scale). Squares (clinical factors) versus diamonds (pharmacogenomic variants). GI comorbidity (OR 8.77) and age per decade (OR 2.00) dominate outcome prediction. All pharmacogenomic variants cluster tightly around the null (OR 0.85–1.35) with overlapping credible intervals. GI—gastrointestinal.
Figure 3. Comparative effect sizes of clinical predictors versus pharmacogenomic variants for bleeding outcome (log-scale). Squares (clinical factors) versus diamonds (pharmacogenomic variants). GI comorbidity (OR 8.77) and age per decade (OR 2.00) dominate outcome prediction. All pharmacogenomic variants cluster tightly around the null (OR 0.85–1.35) with overlapping credible intervals. GI—gastrointestinal.
Pharmaceutics 18 00884 g003
Table 1. Prevalence of genotypes and genotype-predicted phenotypes in the study cohort (N = 385).
Table 1. Prevalence of genotypes and genotype-predicted phenotypes in the study cohort (N = 385).
CYP Enzymesn (%)Transportersn (%)
CYP3A4*1B (rs2740574) genotype ABCG2 c.421C>A (rs2231142)
*1/*1371 (96.4)C/C314 (81.5)
*1/*313 (3.4)C/A68 (17.7)
*3/*31 (0.2)A/A3 (0.8)
Variant allele frequency1.9%Variant allele frequency9.6%
CYP3A4*22 (rs35599367) genotype ABCG2 phenotype per rs2231142
*1/*1361 (93.8)Normal activity314 (81.5)
*1/*2223 (6.0)Reduced activity71 (18.5)
*22/*221 (0.2)ABCB1 c.1236C>T (rs1128503)
Variant allele frequency3.2%C/C124 (32.2)
CYP3A4 phenotype C/T183 (47.5)
Normal metabolizer (NM)361 (93.8)T/T78 (20.3)
Intermediate metabolizer (IM)23 (6.0)Variant allele frequency44.0%
Poor metabolizer (PM)1 (0.2)ABCB1 c.2677G>T/A (rs2032582)
CYP3A5*3 (rs776746) genotype G/G158 (41.0)
*1/*12 (0.5)G/T or G/A145 (37.7)
*1/*338 (9.9)T/T or T/A or A/A82 (21.3)
*3/*3345 (89.6)Variant allele frequency40.1%
Variant allele frequency94.5%ABCB1 c.3435C>T (rs1045642)
CYP3A5*3 phenotype C/C89 (23.1)
Enzyme expressors40 (10.4)C/T177 (46.0)
Non-expressors345 (89.6)T/T119 (30.9)
CYP3A combined phenotype Variant allele frequency53.9%
Poor metabolizer (PM)24 (6.2)ABCB1 phenotype per rs1045642
Intermediate metabolizer (IM)322 (83.6)Normal function266 (69.1)
Rapid metabolizer (RM)39 (10.1)Reduced function119 (30.9)
CYP2J2*7 (rs890293) genotype ABCB1 c.2482-2236G>A (rs4148738)
*1/*1339 (88.1)G/G109 (28.3)
*1/*746 (11.9)G/A184 (47.8)
*7/*70A/A92 (23.9)
Variant allele frequency6.0%Variant allele frequency47.8%
CYP2J2*7 phenotype ABCB1 phenotype per rs4148738
Normal activity339 (88.1)Normal activity109 (28.3)
Reduced activity46 (11.9)Increased/extensive activity276 (71.7)
CYP2J2 c.A>T (rs11572325)
A/A302 (78.4)
A/T79 (20.5)
T/T4 (1.0)
Variant allele frequency11.3%
Table 2. Main demographic and (co)morbidity characteristics of cases (experienced bleeding events) and controls (no bleeding events). Data are counts (percentages) or medians (quartiles, minimum-maximum).
Table 2. Main demographic and (co)morbidity characteristics of cases (experienced bleeding events) and controls (no bleeding events). Data are counts (percentages) or medians (quartiles, minimum-maximum).
Cases (Bleeding)Controls (No Bleeding)
N71314
Demographics and renal function
Age (years)74 (67–80; 30–88)66 (57–72; 26–98)
Males41 (57.8)197 (62.7)
eGFR (mL/min/1.73 m2)69 (47–85; 10–104)76 (60–90; 15–125)
Indication of rivaroxaban
Nonvalvular atrial fibrillation/atrial flutter57 (80.3)227 (72.3)
Peripheral artery disease12 (16.9)56 (17.8)
Venous thromboembolism12 (16.9)40 (12.7)
Ischemic stroke/TIA7 (9.9)50 (15.9)
Coronary heart disease/acute myocardial infarction23 (32.4)72 (22.9)
Thrombus in the left atrial apex/appendage2 (2.8)7 (2.2)
Adult congenital heart disease06 (1.9)
Thrombophilia2 (2.8)6 (1.9)
Daily rivaroxaban dose
1 × 20 mg41 (57.8)232 (73.9)
1 × 15 mg20 (28.2)33 (10.5)
1 × 10 mg7 (9.9)5 (1.6)
2 × 2.5 mg (total 5 mg)3 (4.2)44 (14.0)
Comorbidities
Hypertension64 (90.1)253 (80.6)
Dyslipidemia46 (64.8)222 (70.7)
GI disease (peptic ulcer/diverticulosis/polyps/IBD)32 (45.1)24 (7.6)
Malignancy (solid organ/hematological)20 (28.1) (16 + 4)51 (16.3) (41 + 10)
Diabetes mellitus14 (19.7)71 (22.6)
Anxious or depressive disorder9 (12.7)30 (9.6)
Systemic inflammatory disease4 (4.6)7 (2.2)
Musculoskeletal pain syndromes3 (4.2)18 (5.7)
eGFR—estimated glomerular filtration rate; GI—gastrointestinal; IBD—inflammatory bowel disease; TIA—transient ischemic attack.
Table 3. Concomitant treatments in cases and controls. Data are count (percent).
Table 3. Concomitant treatments in cases and controls. Data are count (percent).
Cases (Bleeding)Controls (No Bleeding)
N71314
Proton pump inhibitors49 (69.0)226 (72.0)
Platelet aggregation inhibitors15 (21.1)67 (21.1)
Dual antiplatelet therapy5 (7.0)10 (3.2)
CYP3A4 inhibitors
047 (66.7)186 (59.2)
At least 1 (1 to 3)24 (33.8)128 (40.8)
CYP3A4 substrates
03 (4.2)21 (6.7)
1–230 (42.3)125 (39.8)
3 or more38 (53.5)168 (53.5)
CYP3A5 inhibitors
064 (90.1)269 (85.7)
At least 1 (1–2)7 (9.9)45 (14.3)
CYP3A5 substrates
026 (36.6)123 (39.2)
1–240 (56.3)159 (50.6)
3 or more5 (7.1)32 (10.2)
CYP2J2 inhibitors
055 (77.5)241 (76.8)
1–216 (22.5)73 (23.2)
ABCB1 inhibitors
04 (5.6)30 (9.5)
1–240 (56.3)164 (52.3)
3 or more27 (38.1)120 (38.2)
ABCB1 inducers
064 (90.1)279 (88.8)
1–27 (9.9)35 (11.2)
ABCB1 substrates
04 (5.6)24 (7.6)
1–243 (60.6)182 (58.0)
3 or more24 (33.7)108 (34.4)
ABCG2 inhibitors
024 (33.8)92 (29.3)
147 (66.2)222 (70.7)
ABCG2 substrates
013 (18.3)28 (8.9)
1–253 (74.6)254 (80.9)
3 or more5 (7.1)32 (10.2)
Table 4. Prevalence of the investigated polymorphisms and genotype-predicted phenotypes in cases and controls. Data are count (percent).
Table 4. Prevalence of the investigated polymorphisms and genotype-predicted phenotypes in cases and controls. Data are count (percent).
Cases (Bleeding)Controls (No Bleeding)
N71314
CYP3A phenotype
Rapid/intermediate metabolizer68 (95.8)294 (93.6)
Poor metabolizer4 (4.2)20 (6.4)
CYP2J2 phenotype (according to CYP2J*7)
Normal function66 (93.0)273 (86.9)
Reduced function5 (7.0)41 (13.1)
CYP2J2 A>T (rs11572325)
Wild type60 (84.5)242 (77.1)
Variant allele carrier11 (15.5)72 (22.9)
ABCG2 c.421C>A
Wild type55 (77.5)259 (82.5)
Variant allele carrier 16 (22.5)55 (17.5)
ABCB1 c.1236C>T
Wild type24 (33.8)100 (31.8)
Variant allele carrier 47 (66.2)214 (68.2)
ABCB1 2677G>T/A
Wild type28 (39.4)129 (41.1)
Variant allele carrier 43 (60.6)185 (58.9)
ABCB1 3435C>T
Wild type 19 (26.8)70 (22.3)
Variant allele carrier52 (73.2)244 (77.7)
ABCB1 2482-2236G>A
Wild type 17 (23.9)92 (29.3)
Variant allele carrier54 (76.1)222 (70.7)
Table 5. Raw and adjusted estimates of association (odds ratio, OR) between the investigated polymorphisms and the risk of bleeding.
Table 5. Raw and adjusted estimates of association (odds ratio, OR) between the investigated polymorphisms and the risk of bleeding.
VariantRaw Data OR (95% CrI)Balanced Data OR (95% CrI)
CYP2J2 c.A>T variant allele vs. wild type0.80 (0.50–1.38)0.94 (0.55–1.65)
ABCG2 c.421C>A variant allele vs. wild type1.16 (0.70–1.88)1.25 (0.76–2.12)
ABCB1 c.1236C>T variant allele vs. wild type1.12 (0.70–1.75)1.05 (0.59–1.90)
ABCB1 c.2677G>T/A variant allele vs. wild type1.20 (0.76–1.95)1.35 (0.81–2.07)
ABCB1 c.3435C>T variant allele vs. wild type1.04 (0.76–1.88)0.85 (0.49–1.51)
ABCB1 c.rs4148738 variant allele vs. wild type1.20 (0.76–1.95)1.12 (0.66–2.01)
Estimates for “raw” and “balanced” data were adjusted for age, eGFR, and gastrointestinal comorbidity.
Table 6. Comparison of key pharmacogenomic studies evaluating rivaroxaban clinical outcomes.
Table 6. Comparison of key pharmacogenomic studies evaluating rivaroxaban clinical outcomes.
Study (Year)NDesignMain VariantsKey FindingConsistency with Present Study
Wu et al. 2023 [12]95Prospective multicenterCYP3A4/5, ABCB1, ABCG2No bleeding association Yes
Wang et al. 2025 [18]SRSystematic review 25 lociInsufficient evidence for implementationYes
Li et al. 2024 [13]165Prospective CYP3A4/5CYP3A4 variants associated with bleedingPartial
Wang et al. 2025 [15]228Prospective PK researchABCB1 c.3435C>T, c.1236C>T, c.2677G>T/Ac.3435 T/T: higher VTE riskNo
Kim et al. 2023 [14]293Case-controlABCG2, ABCB1ABCG2 rs3114018, ABCB1 rs1045642 assoc. with bleedingNo
Campos-Staffico 2022 [19]2364Retrospective cohort8 PK variants
ABCB1, ABCG2, CYP2J2, CYP3A4/5		No PGx–DOAC bleeding associationYes
Present study (2026)385Nested case-control10 SNPs
(5 genes)		No PGx signal; age + GI comorbidity dominate—
PGx, pharmacogenomics; PK, pharmacokinetics; SR, systematic review; VTE, venous thromboembolism.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Slišković, A.M.; Trkulja, V.; Ganoci, L.; Božina, T.; Pašara, V.; Vrkić Kirhmajer, M.; Palić, J.; Strikić, D.; Narančić, M.; Sopek Merkaš, I.; et al. Pharmacogenomics of Rivaroxaban: Association of CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 Variants with Bleeding and Thrombotic Outcomes in Real-World Clinical Practice. Pharmaceutics 2026, 18, 884. https://doi.org/10.3390/pharmaceutics18070884

AMA Style

Slišković AM, Trkulja V, Ganoci L, Božina T, Pašara V, Vrkić Kirhmajer M, Palić J, Strikić D, Narančić M, Sopek Merkaš I, et al. Pharmacogenomics of Rivaroxaban: Association of CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 Variants with Bleeding and Thrombotic Outcomes in Real-World Clinical Practice. Pharmaceutics. 2026; 18(7):884. https://doi.org/10.3390/pharmaceutics18070884

Chicago/Turabian Style

Slišković, Ana Marija, Vladimir Trkulja, Lana Ganoci, Tamara Božina, Vedran Pašara, Majda Vrkić Kirhmajer, Jozefina Palić, Dominik Strikić, Marino Narančić, Ivana Sopek Merkaš, and et al. 2026. "Pharmacogenomics of Rivaroxaban: Association of CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 Variants with Bleeding and Thrombotic Outcomes in Real-World Clinical Practice" Pharmaceutics 18, no. 7: 884. https://doi.org/10.3390/pharmaceutics18070884

APA Style

Slišković, A. M., Trkulja, V., Ganoci, L., Božina, T., Pašara, V., Vrkić Kirhmajer, M., Palić, J., Strikić, D., Narančić, M., Sopek Merkaš, I., Merćep, I., Bulum, J., & Šimičević, L. (2026). Pharmacogenomics of Rivaroxaban: Association of CYP3A4, CYP3A5, CYP2J2, ABCB1, and ABCG2 Variants with Bleeding and Thrombotic Outcomes in Real-World Clinical Practice. Pharmaceutics, 18(7), 884. https://doi.org/10.3390/pharmaceutics18070884

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop