Prognostic Analysis of Butylphthalide Therapy Guided by CYP2C19 Genotype in Patients with Acute Cerebral Infarction

Summary

This article describes a study evaluating the prognosis of CYP2C19 genotype-guided butylphthalide therapy in acute cerebral infarction, showing that poor metabolizers have worse outcomes, and that Hcy, NIHSS, and CYP2C19 are independent predictors.

Abstract

To investigate the prognosis of patients with acute-phase cerebral infarction treated with butylphthalide based on cytochrome P450 2C19 (CYP2C19) genotype. A total of 175 patients with acute cerebral infarction treated with butylphthalide at our hospital from January 2022 to January 2025 were retrospectively enrolled. Based on modified Rankin Scale (mRS) scores at 3 months post-treatment, patients were categorized into a good prognosis group (mRS ≤2) and a poor prognosis group (mRS ≥3). CYP2C19 genotypes and metabolic phenotypes were analyzed. Clinical data and genotype distributions were compared between groups, and logistic regression analysis was performed to identify prognostic factors. Intermediate metabolizers predominated (46.29%), while poor metabolizers were the least common (12.00%). Compared to the good prognosis group, the poor prognosis group exhibited a significantly higher prevalence of hypertension (P <0.05), elevated serum homocysteine (Hcy) levels (P <0.05), and higher baseline National Institutes of Health Stroke Scale (NIHSS) scores (P <0.05). Poor metabolizers (CYP2C19*2/*2, *2/*3, *3/*3) were overrepresented in the poor prognosis group, whereas rapid metabolizers (CYP2C19*1/*1) were more frequent in the good prognosis group (P <0.05). Logistic regression analysis identified three independent predictors of poor prognosis after controlling for covariates: elevated homocysteine levels (odds ratio (OR) = 2.255; 95% CI: 1.404–3.622; P < 0.001), higher baseline NIHSS score (OR = 3.127; 95% CI: 1.508–6.482; P = 0.002), and CYP2C19 poor metabolizer status (OR = 4.559; 95% CI: 2.392–8.688; P < 0.001). Elevated Hcy levels, increased NIHSS scores, and CYP2C19 poor metabolizer phenotypes are significant prognostic factors in butylphthalide-treated acute cerebral infarction. Patients with poor metabolizer genotypes exhibit worse outcomes, highlighting the need for personalized treatment strategies to optimize outcomes.

Introduction

Acute cerebral infarction, accounting for over 80% of all stroke cases, represents a critical global health burden characterized by high mortality, disability, and recurrence rates. As the second leading cause of death worldwide, it imposes severe socioeconomic impacts through long-term neurological impairment and functional dependency¹. The urgency of therapeutic intervention is underscored by the concept of "time-is-brain" – where approximately 1.9 million neurons perish per minute during untreated ischemia².

In this clinical landscape, butylphthalide (dl-3-n-butylphthalide) has emerged as a first-line neuroprotective agent in China, demonstrating multimodal efficacy in improving cerebral microcirculation, reducing oxidative stress, and mitigating neuronal apoptosis³. However, substantial interindividual variability in therapeutic response persists, with 12%–18% of patients experiencing limited clinical benefits or adverse events, including hepatotoxicity and neurological deterioration⁴. This heterogeneity highlights the critical need for personalized treatment approaches. Pharmacogenomic advances have elucidated cytochrome P450 enzymes, particularly CYP2C19, as pivotal determinants of drug metabolism and efficacy⁵. CYP2C19 polymorphisms significantly influence the pharmacokinetics of neuroactive agents, with well-established impacts on clopidogrel responsiveness in stroke prophylaxis. Over 40% of Asian populations carry loss-of-function alleles (*2, *3) that confer poor metabolizer phenotypes, leading to altered drug exposure and clinical outcomes. Notably, while CYP2C19 genotyping has revolutionized antiplatelet therapy⁶, its role in guiding neuroprotectant regimens remains unexplored. Butylphthalide undergoes complex hepatic metabolism involving CYP450 isoforms, yet the prognostic implications of CYP2C19 variants on its therapeutic efficacy are unknown. Unlike clopidogrel, a prodrug requiring CYP2C19-mediated activation, butylphthalide's metabolic pathways and the precise role of CYP2C19 in its disposition and effectiveness remain to be fully characterized. This distinction highlights the novel aspect of our investigation: moving beyond the established paradigm of antiplatelet pharmacogenomics into the emerging field of neuroprotective agent personalized therapy. This knowledge gap impedes the implementation of precision medicine in acute stroke care. Therefore, this study investigates the association between CYP2C19 genetic polymorphisms and 3-month functional outcomes in patients with acute cerebral infarction receiving butylphthalide. By integrating genotype data with established prognostic markers, we aim to develop a pharmacogenomic framework to optimize neuroprotective therapy in this population, thereby extending the applications of stroke pharmacogenomics beyond antiplatelet agents.

This study is among the first to investigate the impact of CYP2C19 pharmacogenomics on outcomes of butylphthalide monotherapy. The findings reveal a significant association between poor metabolizer phenotypes and worse functional recovery, providing a rationale for personalized neuroprotective strategies.

Protocol

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of Huizhou Central People's Hospital (Approval No. 2022-015). This study was not registered as a clinical trial due to its retrospective design. Informed consent was waived due to the retrospective nature of the analysis, and all patient data were de-identified. The regents and the equipment used are listed in the Table of Materials.

1. Study design and patient selection

As a retrospective multicenter study, we acknowledge the potential for selection bias. To mitigate this, patient identification followed a consecutive enrollment protocol from prospective stroke registries at each participating center. All patients presenting with acute cerebral infarction during the study period (January 2022 to January 2025) were screened for eligibility, and those meeting the inclusion criteria were enrolled sequentially. Data were pooled from three institutions: Shaoxin Shangyu People's Hospital, Run Run Shaw Hospital, and Huizhou Central People's Hospital. Data sharing agreements were established between all participating centers to ensure proper oversight and data access.

To handle missing data, a multiple imputation approach was implemented using chained equations for variables with <5% missingness. For variables with >5% missing data, a complete-case analysis was employed, with transparent reporting of missingness patterns. Critical variables, including CYP2C19 genotype, baseline NIHSS, and 3-month mRS, had complete data for all participants. The potential confounding were addressed through multivariate adjustment in our statistical models, including known prognostic factors such as age, baseline stroke severity, vascular risk factors, and concomitant medications. Specifically, data on concomitant medications (antiplatelets, statins, antihypertensives) were collected and adjusted for in analyses.

1. Inclusion criteria
   The inclusion criteria were as follows: (1) Diagnosis of acute cerebral infarction confirmed by imaging and clinical criteria per the Chinese Guidelines for the Diagnosis and Treatment of Acute Ischemic Stroke; (2) First-ever stroke with onset-to-admission time <72 h; (3) Treatment with butylphthalide (0.1 g per capsule) at a dose of 0.2 g three times daily on an empty stomach, initiated within 24 h of admission; (4) No prior thrombolytic or anticoagulant therapy.
2. Exclusion criteria
   The exclusion criteria were as follows: (1) Comorbid psychiatric disorders or cognitive impairment; (2) Severe hepatic, renal, or cardiac dysfunction; (3) Hypersensitivity to the study drug; (4) Systemic inflammatory or autoimmune diseases. Patients were stratified into good prognosis (mRS ≤2) and poor prognosis (mRS ≥3) groups based on 3-month modified Rankin Scale scores⁷. The patient inclusion process is shown in Figure 1.
   ​Patient adherence to butylphthalide therapy was monitored through direct observation of medication administration during hospitalization, pill counts, and medication diary reviews during follow-up visits, structured telephone interviews at 2-week intervals to assess compliance and adverse effects, and review of pharmacy refill records for outpatients. Patients with <80% adherence were excluded from analysis (n=7). The 3-month follow-up endpoint was selected based on established stroke outcome research standards, as mRS at 90 days post-stroke is widely accepted as a primary endpoint capturing the critical recovery period.

2. Clinical data collection

Data were systematically collected using electronic case report forms covering five domains: demographic characteristics (age, gender, body mass index, Table 1), vascular risk factors (hypertension, diabetes mellitus, smoking history, drinking history), clinical presentation (infarct site, baseline NIHSS score), laboratory parameters, and treatment details. Blood biomarkers were measured as follows:

1. Lipid profile (TC, TG, LDL-C, HDL-C)
   Analyzed using an automated biochemical analyzer, with assay range 0–20 mmol/L for TC, 0–10 mmol/L for TG, 0–15 mmol/L for LDL-C, and 0.5–2.0 mmol/L for HDL-C.
2. Serum homocysteine (Hcy)
   ​Measured by ELISA (detection range 5–50 µmol/L, coefficient of variation <5%). Samples were stored at -80 °C within 2 h of collection to avoid degradation.

3. CYP2C19 genotyping

1. Sample collection and DNA extraction
   Fasting venous blood (5 mL) was collected from the median cubital vein in the morning. After 30 min of clotting at room temperature, samples were centrifuged at 1,600 × g for 10 min at 4 °C using a high-speed centrifuge to isolate the serum. Genomic DNA was extracted from 300 µL of serum following the manufacturer's protocol. Briefly, samples were incubated with 300 µL lysis buffer at 56 °C for 10 min, followed by the addition of 100 µL absolute ethanol. The mixture was transferred to purification columns and centrifuged at 12,000 × g for 1 min at room temperature. DNA was eluted with 50 µL pre-warmed (70 °C) elution buffer after two wash steps with 500 µL wash buffer each, with centrifugation at 12,000 × g for 1 min after each wash. Final DNA concentration was adjusted to 50–100 ng/µL using a spectrophotometer and stored at -20 °C.
2. Real-time PCR genotyping
   CYP2C19 genotyping was performed via real-time PCR. The 20 µL reaction mixture consisted of 10 µL 2× TaqMan Master Mix, 1 µL primer-probe mix, 2 µL DNA template, and 7 µL nuclease-free water. Thermal cycling conditions were set as follows: initial denaturation at 95 °C for 30 s, followed by 40 cycles of denaturation at 95 °C for 5 s and combined annealing/extension at 60 °C for 34 s. Fluorescence data were collected during the annealing/extension step.
3. Genotype confirmation and quality control
   Genotype calls were confirmed using second-generation Hi-SNP sequencing with 100× coverage. For discordant results (n=3), restriction fragment length polymorphism analysis was employed as a backup method: 20 µL PCR product was incubated with 1 µL HpyCH4V restriction enzyme and 4 µL corresponding buffer at 37 °C for 1 h, followed by separation via 2% agarose gel electrophoresis at 100 V for 45 min.
   Rigorous quality control measures were implemented. All samples were processed in duplicate with blinded replicate testing; internal controls (known genotypes) were included in each PCR run; concordance between real-time PCR and Hi-SNP sequencing results was 100%; randomly selected samples (10%) were sent to an independent laboratory for validation, achieving 100% concordance; call rates exceeded 99.5% with minimum genotype quality score of Q30.
4. Metabolic phenotype classification
   ​Metabolic phenotypes were categorized based on CYP2C19 genotype as follows: Poor metabolizers (*2/*2, *2/*3, *3/*3); Intermediate metabolizers (*3/*17, *2/*17, *1/*2, *1/*3); Rapid metabolizers (*1/*1). This three-tier classification schema was employed for consistency with regional clinical laboratory practices. The frequency of the *17 allele and the distribution of genotypes within the IM group are detailed in the Results section.

4. Statistical analysis

Statistical analyses were performed using SPSS version 25.0. Categorical data were expressed as [n (%)] and compared via χ² tests (with Yates' correction for small samples). Continuous variables were tested for normality using the Shapiro-Wilk test; normally distributed data were presented as mean ± standard deviation and analyzed with independent t-tests, while non-normal data were analyzed with Mann-Whitney U tests.
Multicollinearity was assessed using tolerance (≥0.1) and variance inflation factor (VIF <10). Logistic regression analysis (using the Analyze → Regression → Binary Logistic menu path in SPSS) was used to identify independent prognostic factors, with entry and exit criteria set at P <0.05 and P >0.10, respectively. Variables entered into the model included Hcy levels (continuous), baseline NIHSS score (continuous), hypertension (categorical), and CYP2C19 metabolic phenotype (categorical, with rapid metabolizer as reference). The Hosmer-Lemeshow goodness-of-fit test was performed to assess model calibration. A two-tailed P <0.05 was considered statistically significant.
Effect sizes were calculated as follows: for continuous variables, Cohen's d was computed using the Analyze < Compare MeansIndependent-Samples T Test function with additional effect size calculation enabled; for categorical variables, the phi coefficient was obtained from the χ² test output under Analyze Descriptive StatisticsCrosstabs with Phi and Cramer's V selected in the Statistics options.

Results

Distribution of CYP2C19 genotypes and metabolic phenotypes

The analysis of CYP2C19 genotypes and metabolic phenotypes in 175 patients with acute cerebral infarction revealed that intermediate metabolizers were the most prevalent (81/175, 46.29%), while poor metabolizers were the least common (21/175, 12.00%). Genotype distribution showed that CYP2C19 * 1/*1 (rapid metabolizer) was the most frequent genotype (73/175, 41.71%), followed by *1/*2 (intermediate metabolizer, 72/175,...

Discussion

Acute cerebral infarction, a syndrome caused by cerebrovascular insufficiency, represents a critical neurological emergency^8,9,10. Butylphthalide has emerged as a principal therapeutic agent for this condition. Studies have demonstrated that its therapeutic mechanisms involve suppression of intracellular free radical generation, enhancement of antioxidant enzyme activity, and preservation of mitochondrial integrity, collectively...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
2× TaqMan Master Mix	Thermo Fisher Scientific	4371355	Used for CYP2C19 genotyping real-time PCR reaction
Agarose	N/A	N/A	Used to prepare 2% gel for agarose gel electrophoresis
Butylphthalide capsules	NBP Pharmaceutical Co., Ltd.	H20050299	0.1 g per capsule; administered at 0.2 g three times daily on an empty stomach
DNA extraction kit	Tiangen Biotech	DP304	Used for genomic DNA extraction from serum samples; includes lysis buffer, wash buffer and elution buffer
Electronic case report forms (e-CRF)	N/A	N/A	Systematic clinical data collection tool covering five clinical data domains
ELISA kit for homocysteine (Hcy)	Shanghai Enzyme-linked Biotech	N/A	Detection range 5–50 μmol/L; coefficient of variation <5%; for serum Hcy measurement
Elution buffer	Tiangen Biotech	N/A	Pre-warmed to 70°C; component of DP304 DNA extraction kit, for DNA elution
Eppendorf 5425R Refrigerated High-Speed Centrifuge	Eppendorf	5425R	Used for serum isolation and genomic DNA purification centrifugation steps
Hi-SNP sequencing platform/analysis system	BGI Genomics	N/A	Second-generation sequencing; used for CYP2C19 genotype confirmation with 100× coverage
HpyCH4V restriction enzyme	New England Biolabs	N/A	Used for restriction fragment length polymorphism analysis of discordant genotyping results
Lysis buffer	Tiangen Biotech	N/A	Component of DP304 DNA extraction kit; incubated with samples at 56°C for 10 min
NanoDrop spectrophotometer	Thermo Fisher Scientific	N/A	Used to adjust final genomic DNA concentration to 50-100 ng/μL
Nuclease-free water	N/A	N/A	Used to prepare CYP2C19 genotyping real-time PCR reaction mixture
Primer-probe mix (CYP2C19*2)	Thermo Fisher Scientific	C__25986767_30	Specific for CYP2C19*2 genotyping in real-time PCR
Primer-probe mix (CYP2C19*3)	Thermo Fisher Scientific	C__27861809_10	Specific for CYP2C19*3 genotyping in real-time PCR
Roche LightCycler 480 II system	Roche Diagnostics	LightCycler 480 II	Real-time PCR system for CYP2C19 genotyping
Roche/Hitachi 7600 analyzer	Roche Diagnostics/Hitachi	7600	For lipid profile (TC, TG, LDL-C, HDL-C) measurement; assay ranges for lipids specified in the study
SPSS	IBM Corp.	25	Statistical analysis software; used for all study data analysis (χ² tests, t-tests, logistic regression, etc.)
Wash buffer	Tiangen Biotech	N/A	Component of DP304 DNA extraction kit; two wash steps with 500 μL each in DNA purification