Introduction

Stroke is the second leading cause of death globally and ranks as the primary cause of mortality and the third leading cause of disability in the Chinese population1. Cerebral ischemic stroke (CIS) accounts for 81.9% of all stroke cases, with a recurrence risk of approximately 5–10% within three months following the initial event2,3. Previous studies4,5,6 had demonstrated that elevated levels of low-density lipoprotein cholesterol (LDL-C) serve as an independent risk factor for both the pathogenesis and recurrence of CIS. Reduction of LDL-C levels can mitigate atherosclerosis progression, thereby decreasing the incidence, recurrence, disability, and mortality associated with CIS. Atorvastatin is widely utilized in clinical practice for lipid-lowering therapy in CIS patients. In the context of secondary prevention of CIS, statin therapy that reduces LDL-C by 1 mmol L−1 is associated with a 12% reduction in the risk of CIS recurrence7. However, not all patients derive equal benefit from this treatment, as significant interindividual variability in lipid-lowering efficacy exists. Genetic polymorphisms have been identified as a critical factor contributing to these variations in therapeutic response8.

The ABCB1 (2677T > G) polymorphism has emerged as one of the most frequently investigated genetic loci influencing statin therapeutic efficacy, though conflicting evidence exists across clinical studies. Investigations by Kadam et al.9 demonstrated that GG genotype carriers exhibited more pronounced LDL-C reduction following atorvastatin administration. However, contradictory findings from Jemaa et al.10 suggested no significant association between ABCB1(2677T > G) polymorphism and lipid-lowering outcomes. While clinical practitioners and patients predominantly focus on serum lipid profile alterations (particularly LDL-C and HDL-C levels) during therapeutic monitoring, emerging evidence suggests the LDL-C/HDL-C ratio demonstrates superior predictive value for atherosclerosis progression compared to isolated lipoprotein measurements11,12. Notably, the potential impact of ABCB1(2677T > G) polymorphism on lipid profile variations (including LDL-C, HDL-C, and their dynamic ratios) and long-term clinical outcomes in ischemic stroke patients remains underexplored.To address this knowledge gap, we systematically evaluated six lipid-related parameters: ΔHDL-C, ΔLDL-C, ΔHDL-C/HDL-C, ΔLDL-C/LDL-C, Δ(LDL-C/HDL-C) and Δ(LDL-C/HDL-C)/(LDL-C/HDL-C). This investigation aimed to elucidate the association between the ABCB1(2677T > G) gene polymorphism and the lipid-lowering efficacy of atorvastatin in patients with CIS under different models, and compared 3-year cerebrovascular event rates across genetic subgroups. Our analytical framework provided a novel biomarker-driven strategy for optimizing precision lipid management in CIS patients.

Materials and methods

Study design and participants

This prospective cohort study consecutively enrolled 152 Han Chinese CIS patients admitted to the Department of Neurology between January 2021 and December 2021. All participants met diagnostic criteria for CIS through clinical evaluation and neuroimaging confirmation (MRI and/or CT). During follow-up, 25 patients were excluded (14 due to protocol non-compliance, 8 withdrew consent, and 3 lost to follow-up), resulting in 127 analyzable cases. Participants were stratified into TT, GT, and GG groups based on ABCB1(2677T > G) genotyping. The study protocol was approved by the Ethics Committee of Xuchang Central Hospital (Approval No.2019-K-12-006), with written informed consent obtained from all participants or legal guardians.Inclusion criteria: (1) First-ever CIS meeting WHO diagnostic criteria; (2) Radiologically confirmed ischemic lesion; (3) Han Chinese ancestry in Henan Province; (4) Age ≥ 18 years; (5) Comorbid hyperlipidemia per 2016 Chinese Guidelines for Dyslipidemia Management13; (6) Stable residency (> 10 years) in the study region.Exclusion criteria: (1) Fatal stroke presentation; (2) Severe systemic comorbidities (cardiac/hepatic/renal failure, malignancies, autoimmune/hematologic disorders); (3) Prior use of statins or lipid-lowering agents within 30 days pre-admission; (4) Atorvastatin intolerance.

Experimental protocols

Genotyping and biochemical profiling

On the morning of the second day of hospitalization, approximately 2 mL of venous blood was collected from patients using EDTA anticoagulant tubes. DNA extraction was performed according to the manufacturer’s protocol of the commercial kit from Sino Era Genotech(China).

employed. Genetic testing for the ABCB1 (2677T > G) polymorphism was conducted using the TL-998 fluorescence detection system(Sino Era Genotech, China) with digital fluorescence molecular hybridization technology, a well-established detection method widely adopted in clinical laboratories across China.Serum lipid profiles (HDL-C, LDL-C) were quantified using commercially available enzymatic assay kits (Roche Diagnostics, Germany) on a cobas 8000 automated analyzer (Roche Diagnostics GmbH, Germany).

Therapeutic regimen and outcome measures

All patients received standardized care including anticoagulation, antiplatelet therapy, and neuroprotection. Lipid-lowering therapy comprised atorvastatin calcium tablets (Lipitor®, Pfizer; 20 mg once daily for two consecutive months). The primary outcome measure comprised longitudinal changes in six lipid profile parameters between baseline and post-treatment measurements at 2 months, including ΔHDL-C, ΔLDL-C, ΔHDL-C/HDL-C, ΔLDL-C/LDL-C, Δ(LDL-C/HDL-C), and Δ(LDL-C/HDL-C)/(LDL-C/HDL-C), with the calculation methods shown in Table 1. The secondary outcome measure comprised the cumulative incidence of cerebrovascular events, including radiologically-confirmed ischemic stroke recurrence and transient ischemic attacks (TIAs), which was meticulously assessed through either telephone or outpatient follow-up within 36 months subsequent to the patients’ discharge from the hospital. Therapeutic efficacy was defined as achieving LDL-C < 1.8 mmol L−12. Baseline demographics (age, sex, BMI), clinical parameters (blood pressure, glucose), and adverse health behaviors (tobacco use history, alcohol consumption patterns, nighttime sleep deprivation) were extracted from the Hospital Information System (HIS).

Table 1 Calculation formula for blood lipid changes.
Full size table

Statistical analysis

All analyses were conducted using SPSS 25.0 (IBM Corp., Armonk, NY). The continuous variables conforming to normal distribution are presented as mean ± standard deviation, and one-way ANOVA was used for comparisons among three groups, with post-hoc testing conducted via the Bonferroni multiple comparison procedure. Non-normally distributed continuous variables are expressed as median (interquartile range), and the Kruskal-Wallis H test was applied. The Hardy-Weinberg equilibrium, categorical data, and prognostic differences between two groups were analyzed using the chi-square test. Analysis of covariance (ANCOVA) was performed to explore the differences in lipid-lowering efficacy of atorvastatin among different groups. Binary logistic regression analysis was employed to assess the association between ABCB1 (2677T > G) and the lipid-lowering efficacy of atorvastatin in CIS patients under additive, recessive, and dominant genetic models. A P-value < 0.05 was considered statistically significant.

Results

Demographic characteristics across genotype groups

The study population comprised 127 analyzable cases stratified by ABCB1(2677T > G) genotypes: GG (n = 48), GT (n = 49), and TT (n = 30). Baseline characteristics revealed comparable demographic profiles across groups (Table 2). The GG cohort included 27 males (56.3%) with mean age 63.17 ± 11.04 years and BMI 24.65 ± 2.99 kg/m2. The GT group contained 32 males (65.3%) aged 61.22 ± 13.53 years (BMI 24.46 ± 3.02 kg/m2), while the TT group consisted of 20 males (66.7%) with mean age 68.73 ± 12.02 years (BMI 23.73 ± 2.88 kg/m2). Intergroup comparisons demonstrated no statistically significant differences in sex distribution, BMI, or other baseline parameters (all P > 0.05), except for age variation (P = 0.031 by ANOVA).

Table 2 Baseline data comparison table between the three groups. Comparisons were made by one-way ANOVA followed by bonferroni multiple comparison test. Compared with the TT group, *P = 0.161, **P = 0.028.
Full size table

Hardy-Weinberg equilibrium analysis of ABCB1(2677T > G) genotypes

Genetic distribution analysis revealed the following genotype frequencies among 127 participants: GG (homozygous variant, 48 cases, 37.80%), GT (heterozygous variant, 49 cases, 38.58%), and TT (wild-type, 30 cases, 23.62%). Allelic frequencies were 57.09% for G and 42.91% for T. Chi-square test demonstrated concordance between observed and expected genotype frequencies (χ2 = 3.225, P = 0.199), confirming adherence to Hardy-Weinberg equilibrium (HWE). This genetic equilibrium validates the sample population as a representative Mendelian cohort without significant selection bias (Table 3).

Table 3 Hardy-Weinberg genetic balance of the ABCB1(2677T > G) gene type.
Full size table

ANCOVA analysis of atorvastatin therapeutic efficacy across genotype groups

In both the additive and recessive models with age as a covariate, ABCB1(2677T > G) gene polymorphism significantly affected ΔLDL-C, ΔLDL-C/LDL-C, Δ(LDL-C/HDL-C), and Δ(LDL-C/HDL-C)/(LDL-C/HDL-C). In the additive model: FΔLDL−C = 4.198, P = 0.001; FΔLDL−C/LDL−C = 3.042, P = 0.013; FΔ(LDL−C/HDL−C) = 3.870, P = 0.003; FΔ(LDL−C/HDL−C)/(LDL−C/HDL−C) = 2.551, P = 0.031. In the recessive model: FΔLDL−C = 5.142, P = 0.002; FΔLDL−C/LDL−C = 3.539, P = 0.017; FΔ(LDL−C/HDL−C) = 5.938, P = 0.001; FΔ(LDL−C/HDL−C)/(LDL−C/HDL−C) = 4.312, P = 0.006. In the dominant model with age as a covariate, ABCB1(2677T > G) gene polymorphism significantly affected Δ(LDL-C/HDL-C) (F = 3.571, P = 0.016), but had no significant effect on ΔLDL-C, ΔLDL-C/LDL-C, or Δ(LDL-C/HDL-C)/(LDL-C/HDL-C) (all P > 0.05) (Table 4). Comparative analysis across genetic models consistently indicated superior lipid-lowering efficacy in GG homozygous carriers.

Table 4 Covariance analysis of the lipid-lowering effect of atorvastatin in three groups.
Full size table

Univariate analysis of factors influencing atorvastatin treatment efficacy

Patients were stratified into responders (serum LDL-C level < 1.8 mmol L−1) and non-responders based on therapeutic efficacy criteria. Univariate analysis revealed significant differences (P < 0.1) between the two groups in nighttime sleep deprivation, and genetic inheritance models (additive and recessive models). These findings suggested potential confounding effects of lifestyle factors and genetic polymorphisms on lipid-lowering outcomes (Table 5).

Table 5 Univariate analysis of the influencing factors of atorvastatin’s lipid-lowering efficacy.
Full size table

Binary logistic regression analysis of factors influencing atorvastatin therapeutic efficacy

Using lipid-lowering efficacy (responder vs. non-responder) as the dependent variable, covariates with P < 0.1 from the univariate analysis including nighttime sleep deprivation, and genetic models (additive and recessive models for ABCB1(2677T > G) were incorporated into separate binary logistic regression analyses. The analysis identified nighttime sleep deprivation and ABCB1 (2677T > G) genotype as significant predictors of atorvastatin efficacy. After adjusting for nighttime sleep deprivation, the association between ABCB1 genotype and therapeutic response remained statistically robust. In the additive model, patients with the GG genotype exhibited a 3.181-fold higher likelihood of achieving lipid-lowering targets compared to wild-type TT carriers (OR = 3.181, 95%CI:1.159 ~ 8.730, P = 0.025). Similarly, in the recessive model, GG genotype patients demonstrated a 3.141-fold increased efficacy relative to T-allele carriers (OR = 3.141, 95%CI:1.397 ~ 7.061, P = 0.006). Nighttime sleep deprivation independently predicted reduced treatment efficacy across both genetic models, with comparable effect magnitudes in the additive (OR = 0.237, 95%CI:0.108 ~ 0.522, P < 0.001) and recessive models (OR = 0.238, 95%CI:0.108 ~ 0.522, P < 0.001) (Table 6).

Table 6 Binary logistic regression analysis of the influencing factors of atorvastatin’s lipid-lowering efficacy.
Full size table

Long-term outcomes assessment based on cerebral infarction recurrence

Cerebral infarction recurrence was evaluated as the primary endpoint to assess long-term therapeutic efficacy. Over a 36-month follow-up period, no statistically significant differences in the cumulative incidence of cerebrovascular events were observed among patients stratified by ABCB1 G2677T genotypes(χ2 = 1.465, P = 0.481) (Table 7).

Table 7 Comparative analysis of long-term clinical outcomes among the three groups.
Full size table

Discussion

This study investigated the association between ABCB1 (2677T > G) genetic polymorphism and lipid-lowering efficacy of atorvastatin in cerebral ischemic stroke (CIS). Our findings demonstrate superior therapeutic responses in GG genotype carriers compared to wild-type T allele carriers under both additive and recessive genetic models. It should be noted that the consistency of these results across distinct inheritance patterns strengthens the pharmacogenetic relevance of ABCB1 2677T > G in statin metabolism. Our findings carry direct implications for precision medicine approaches in stroke secondary prevention. Incorporating ABCB1 genotyping into clinical decision-making may enable stratified therapeutic strategies, particularly for patients exhibiting inadequate statin responses. Based on our findings, we propose the following clinical recommendations: For patients with the ABCB1 (2677T > G) GG genotype, standard-dose atorvastatin is recommended, along with regular lipid profile monitoring. For those with TT/GT genotypes (particularly TT), strategies such as dose escalation, combination therapy with ezetimibe, therapeutic drug monitoring of plasma concentrations, or switching to rosuvastatin are suggested to optimize lipid control.

Lipid-lowering therapy, endorsed by international guidelines as a cornerstone of secondary prevention in cerebral infarction survivors (CIS)14, mitigates atherosclerotic progression through LDL-C reduction, thereby improving long-term clinical outcomes. Atorvastatin, widely utilized in CIS management due to its favorable efficacy-safety profile15, demonstrates dose-dependent protective effects. Amarenco et al.16 reported a 16% reduction in stroke recurrence risk (median follow-up: 4.9 years) with high-dose atorvastatin (80 mg nightly). However, substantial interindividual variability in therapeutic response persists, with a proportion of patients exhibiting suboptimal lipid control17, underscoring the need to elucidate mechanisms underlying statin resistance.Our study identifies ABCB1 (2677T > G) polymorphism as a novel pharmacogenetic determinant of atorvastatin efficacy.

In this study, we assessed the magnitude of LDL-C reduction (ΔLDL-C/LDL-C) across the three genotype groups before and after treatment. Notably, the GG group exhibited a significantly greater LDL-C reduction rate compared to the GT and TT groups (Table 4).To further evaluate the absolute therapeutic efficacy, we defined treatment success as achieving an LDL-C level below 1.8 mmol/L. Strikingly, the proportion of responders was significantly higher in the GG group than in the GT and TT groups (Table 5). To the best of our knowledge, this was the first study to integrate both relative LDL-C reduction and absolute treatment success rate in assessing the impact of ABCB1 G2677T polymorphism on statin response. This dual-metric approach provided a more comprehensive evaluation of genotype-dependent lipid-lowering effects across patients with varying baseline LDL-C levels, thereby better reflecting real-world clinical outcomes.Our findings demonstrate a significant association between ABCB1 (2677T > G) polymorphism and the magnitude of LDL-C reduction in CIS receiving atorvastatin therapy, with GG allele carriers exhibiting enhanced therapeutic sensitivity. This genotype-dependent efficacy variation may stem from two interrelated pharmacokinetic mechanisms: (1) P-glycoprotein-Mediated Drug Disposition: The ABCB1-encoded P-glycoprotein functions as a critical efflux transporter, displaying interindividual expression variations associated with the ABCB1 2677T > G polymorphism. GG genotype carriers exhibit enhanced P-glycoprotein activity, which reduces first-pass hepatic clearance through elevated transporter functionality. This mechanistic cascade increases systemic drug exposure, evidenced by elevated area under the curve (AUC) values and prolonged elimination half-life. This had been confirmed in the study by Qu et al.18 (2) Structural Determinants of Transporter Affinity: The T → G substitution at position 2677 induces a serine-to-alanine amino acid transition (hydrophilic → hydrophobic) in transmembrane domain 11 of P-glycoprotein. This structural modification reduces transporter affinity for atorvastatin’s open-ring conformation, impeding its efflux from hepatocytes and intestinal epithelia. Consequently, GG genotype carriers retain higher intracellular drug concentrations, potentiating HMG-CoA reductase inhibition and LDL-C catabolism19,20.These mechanistic insights align with our clinical observations of genotype-stratified responses. The dual pharmacokinetic advantages prolonged circulation time and enhanced cellular retention collectively explain the superior efficacy in GG carriers. These findings underscore the importance of ABCB1 pharmacogenetics in optimizing statin therapy, particularly for CIS patients with suboptimal lipid control.

Previous studies had established that genetic polymorphisms in SLCO1B1 521T > C (rs4149056)21, ABCB1 C1236T (rs1128503), ABCB1 C3435T (rs1045642)9,19, UGT1A3(rs3821242), and CYP3A4 (rs55785340, rs35599367)22 influenced atorvastatin plasma concentrations, thereby modulating both drug efficacy and adverse effects.Specifically, Bosco et al.21 demonstrated that the SLCO1B1 521T > C (rs4149056) variant increased systemic statin exposure, contributing to a higher risk of adverse drug reactions. Kadam et al.9 showed in a study on Indian populations that patients with ABCB1 C3435T mutation genotypes had better lipid-lowering efficacy with atorvastatin than wild-type patients. A study by Wang et al.17 on the Chinese Uyghur population demonstrated that coronary heart disease patients with the ABCB1 C3435T TT genotype showed more significant triglyceride (TG) reduction after atorvastatin treatment. However, a study by Qu et al.18 on hyperlipidemic patients in southeastern China showed that the ABCB1 C3435T gene polymorphism was not associated with the lipid-lowering efficacy of atorvastatin.Nevertheless, due to limitations in detection conditions, this study did not examine the impact of the above-mentioned genetic loci on lipid-lowering therapy with atorvastatin in cerebral infarction patients. This represents a limitation of our research, which warrants further investigation in future studies.Our genetic models revealed substantial efficacy differences in atorvastatin response stratified by ABCB1 (2677T > G) genotypes. In the additive model, GG homozygotes demonstrated a 3.181-fold increased probability of achieving lipid targets compared to wild-type TT carriers. Similarly, under the recessive model, GG genotype patients exhibited a 3.141-fold higher therapeutic efficacy relative to T-allele carriers. These findings align with Jiang et al.23, who reported a 3% greater LDL-C reduction in GG versus TT genotype patients, .However, contradictory evidence exists. Prado et al.24 observed no ABCB1-statin association in Chilean populations, while a meta-analysis of 3, 088 participants across 10 studies25 found no global pharmacogenetic correlation. We hypothesize that ethnic-specific allele frequency variations may underlie these discrepancies. Supporting this, Wang et al.26 demonstrated divergent ABCB1 (2677T > G) genotype distributions between Han and Uyghur ethnic groups in China.Notably, our lipid profile analyses revealed genotype-dependent bidirectional effects: LDL-C reductions and HDL-C elevations were significantly amplified in GG carriers across both genetic models. This dual lipid-modifying advantage suggests preferential atorvastatin utilization for cerebral infarction survivors (CIS) harboring the GG genotype. Importantly, these findings underscore the necessity for clinicians to consider pharmacogenomic profiles when implementing statin-based secondary prevention strategies, extending beyond conventional lipid monitoring.

The dominant model analysis in this study revealed no significant differences in lipid-lowering efficacy across covariates or regression models, with the exception of Δ(LDL-C/HDL-C). Three potential mechanisms could underlie this observation: First, the relatively small cohort size inherently restricted statistical power for identifying subtle genotype-phenotype associations. Second, our exclusive focus on ABCB1 (2677T > G) polymorphism overlooks the pharmacokinetic complexity of atorvastatin metabolism, which likely involves polygenic interactions across multiple transporters and metabolic enzymes27,28. Additionally, prior pharmacogenomic investigations have implicated other ABCB1 variants (e.g., C3435T) in statin response heterogeneity29. Furthermore, the observed 2677T > G effects may reflect linkage disequilibrium with functional SNPs in regulatory regions rather than direct causality. Haplotype analyses demonstrate that ABCB1 polymorphisms frequently co-segregate, with the 2677T > G locus potentially serving as a proxy marker for clinically relevant variants affecting P-glycoprotein expression18.

Interestingly, our findings revealed that the efficacy rate of lipid-lowering therapy in CIS patients with nighttime sleep deprivation was approximately 24% of that observed in patients maintaining regular sleep patterns. This discovery highlights two critical implications. First, it underscores the clinical importance of prioritizing healthy sleep hygiene when administering atorvastatin-based lipid management to ischemic stroke patients. Second, the results suggest that the ABCB1 (2677T > G) gene polymorphism does not singularly account for the observed variability in therapeutic outcomes among CIS patients. Instead, modifiable lifestyle factors—including environmental influences, dietary habits, and physical activity levels—may synergistically modulate lipid metabolism and treatment responsiveness30.

In this study, no statistically significant differences in clinical outcomes were observed among the three patient groups, suggesting that the ABCB1 (2677T > G) polymorphism may not serve as an independent prognostic predictor for cerebral infarction patients. A plausible explanation lied in the multifactorial nature of stroke recurrence, which involves a complex interplay of platelet hyperreactivity following antiplatelet therapy31, adverse lifestyle habits32, hypertension33, diabetes mellitus34, and other contributing factors.To the best of our knowledge, this study was the first systematic investigation of the association between ABCB1 (2677T > G) (rs2032582) polymorphism and long-term stroke outcomes. Although the results were negative, our findings aligned with the previously reported observations by Liang et al.35 regarding the ABCB1 rs1045642 genetic variant.

This study demonstrated that the lipid-lowering efficacy of atorvastatin in CIS patients was significantly influenced by the ABCB1 (2677T > G) gene polymorphism, providing critical evidence for precision statin therapy in this population. However, several limitations warranted consideration. First, the single-center design with a limited sample size could introduce selection bias. Second, this study was conducted in a Han Chinese population from central China. Due to ethnic variations, the findings may not be generalizable to other racial or ethnic groups. Third, this study only monitored the lipid profile changes for two months in the enrolled patients. Fourth, this study did not experimentally validate the impact of ABCB1 (2677T > G) gene polymorphism on atorvastatin drug concentration, which warrants further investigation in our future research.Therefore, the results may not fully represent the long-term outcomes of lipid-lowering therapy. In light of the above, these conclusions warrant further validation through rigorously designed multicenter studies that should incorporate larger sample sizes, cohorts with broader ethnic representation, extended observation periods, and deeper investigations into the mechanistic impacts of genetic polymorphisms on drug actions.