Association of a CCB-Related Pharmacogenomics-Guided Strategy with Blood Pressure Outcomes and Short-Term Safety in Chinese Patients with Hypertension: A Secondary Exploratory Analysis of Two Randomized Cohorts

Abstract

Introduction

Calcium channel blockers (CCBs) show notable interindividual variability in antihypertensive response. This study evaluated the association of a pharmacogenomics (PGx)-guided antihypertensive strategy incorporating CACNA1C rs2238032 and CYP3A5 rs776746 with short-term blood pressure (BP) outcomes among Chinese patients with hypertension receiving CCB-containing regimens.

Methods

This secondary exploratory analysis included patients receiving CCB-containing regimens from two previously completed multicenter randomized cohorts conducted in Hunan and Fujian provinces, China. The Hunan cohort assessed a 4-week PGx-guided strategy combined with structured management, while the Fujian cohort evaluated genotype-guided treatment during 4–8 weeks after an initial standardized management phase. The primary endpoint was BP control at the last follow-up; secondary endpoints included changes in systolic BP (SBP) and diastolic BP (DBP), medication use, and short-term safety outcomes. Multivariable regression models were used to adjust for prespecified clinical covariates.

Results

A total of 4608 patients in Hunan and 418 patients in Fujian were included. In the Hunan cohort, the genotype-guided group had higher 4‑week BP control (81.83% vs 25.07%; adjusted OR = 14.56, P < 0.001) and greater SBP/DBP decreases. In the Fujian cohort, genotype-guided therapy was associated with higher 8‑week BP control (81.74% vs 77.39%; adjusted OR = 2.43, P = 0.003) and greater DBP decrease. Associations were stronger among patients whose BP remained uncontrolled after standardized management. No short-term safety concerns were identified.

Conclusion

A genotype-guided antihypertensive strategy incorporating CCB-related PGx information, including CACNA1C rs2238032 and CYP3A5 rs776746, was associated with improved short-term BP control among Chinese patients with hypertension receiving CCB-containing regimens, without an apparent short-term safety disadvantage. However, these findings should be interpreted as exploratory associations rather than proof of causal benefit. Further prospective randomized studies are needed to confirm these findings.

Graphical Abstract

Introduction

Hypertension remains a leading cause of cardiovascular morbidity and mortality globally, serving as a major modifiable risk factor for adverse outcomes such as stroke, coronary heart disease, heart failure, and chronic kidney disease,1 and imposes a heavy economic and medical burden on society.2 Its pathogenesis is multifactorial, involving genetic predisposition, aging, dyslipidemia, obesity, impaired glucose tolerance, psychological stress, and unhealthy lifestyle factors.3 Currently, an estimated 1.4 billion adults aged 30 to 79 years are affected by hypertension worldwide, with two-thirds living in low- and middle-income countries.4 However, nearly half of these individuals remain undiagnosed, and only about one-fifth achieve adequate blood pressure (BP) control.4 In China, the burden of hypertension is particularly significant due to the large population size, rapid demographic aging, and evolving lifestyle patterns, indicating a critical need for improvement in disease awareness, treatment, and control rates.5 Consequently, identifying more effective and individualized antihypertensive strategies within real-world clinical settings is of urgent public health importance.

Calcium channel blockers (CCBs) are widely regarded as a first-line antihypertensive class due to their efficacy in lowering both systolic blood pressure (SBP) and diastolic blood pressure (DBP), thereby reducing the risk of major adverse cardiovascular events, including stroke.6,7 Their primary mechanism involves blocking L-type calcium channels on the cell membranes of cardiomyocytes and vascular smooth muscle cells (VSMCs).4 By inhibiting the intracellular influx of calcium ions, CCBs prevent VSMC contraction, leading to decreased peripheral vascular resistance.6 They also affect sinoatrial node depolarization and myocardial contractility, ultimately achieving their antihypertensive effects.8 Based on their diverse chemical structures and pharmacological properties, CCBs are broadly classified into dihydropyridine and non-dihydropyridine subclasses. Dihydropyridine CCBs (eg, amlodipine and nifedipine) exert a more pronounced selective effect on VSMCs. In contrast, among the non-dihydropyridines, verapamil acts predominantly on cardiomyocytes, whereas diltiazem exerts balanced effects on both the heart and vascular smooth muscle.9

Despite the well-documented broad efficacy of CCBs, substantial interindividual variability exists in the clinical BP control, and patients receiving identical medications and dosages often achieve inconsistent therapeutic outcomes.10,11 This heterogeneity suggests that beyond traditional clinical factors (eg, age, sex, and baseline BP), intrinsic genetic variations play a critical role in determining treatment response.12 Emerging evidence indicates that genetic polymorphisms affecting drug metabolism, transmembrane transport, and receptor sensitivity are key determinants of both the efficacy and safety profiles of antihypertensive agents.13

Specifically, CACNA1C encodes the α1C subunit of the L-type voltage-dependent calcium channel, which is the principal therapeutic target for CCBs.4 The rs2238032 polymorphism within CACNA1C has been associated with variable therapeutic responses, and carriers of the G allele may exhibit reduced sensitivity to CCBs, contributing to suboptimal BP control.14 In addition to pharmacodynamics, genetic variations in drug-metabolizing enzymes significantly influence systemic drug exposure. The primary metabolic pathway for most CCBs involves the hepatic cytochrome P450 system including CYP3A5.15,16 The CYP3A5*3 (rs776746) allele is a highly prevalent loss-of-function variant and the presence of its G allele induces aberrant mRNA splicing, resulting in a truncated, non-functional protein.17 Compared to carriers of the functional wild-type allele (CYP3A5*1), individuals with the CYP3A5*3/*3 genotype exhibit significantly impaired metabolic clearance, potentially leading to elevated plasma drug concentrations and enhanced antihypertensive efficacy.18,19

Pharmacogenomics (PGx) provides the molecular foundation for precision medicine by elucidating how genetic variants modulating metabolizing enzymes, transporters, and targets influence drug responses.20 Integrating PGx profiling into hypertension management has the potential to optimize efficacy and minimize adverse events (AEs).21,22 Recent studies have increasingly explored the clinical implementation of PGx-guided antihypertensive therapy, moving beyond the identification of genetic determinants toward individualized medication selection in real-world clinical settings. A multicenter randomized trial demonstrated that PGx-guided antihypertensive treatment was associated with improved BP control and medication optimization compared with conventional management, providing emerging clinical evidence supporting the potential utility of genotype-guided therapy in hypertension management.23 Additionally, a retrospective study by Khan et al suggested that integrating pharmacogenetic testing into antihypertensive decision-making may improve BP outcomes and medication tolerability.24 Current precision medicine approaches increasingly emphasize that the clinical value of genomic information depends on the availability of evidence-based interpretation strategies and the ability to incorporate genetic results into routine healthcare decision-making.25,26 The Clinical Pharmacogenetics Implementation Consortium (CPIC) continues to develop and update genotype-based prescribing guidelines to facilitate the clinical translation of PGx evidence.27,28

However, despite these advances, several practical challenges remain, including the genetic complexity, interindividual differences in antihypertensive treatment response, uncertainty regarding the clinical utility of some drug–gene pairs, and implementation barriers such as the integration of genomic information into clinical workflows, clinician resistance to consider pharmacogenetic information, reimbursement considerations, and equitable access to PGx testing resources.25,29 These challenges should be considered when evaluating the feasibility of implementing PGx-guided antihypertensive strategies in routine clinical care.

Although previous studies have independently explored the genetic basis of CCB responses from either metabolic or target perspectives,14,18,19 and databases such as PharmGKB have annotated the associations between CACNA1C/CYP3A5 variants and CCB efficacy, the current level of supporting evidence remains relatively weak.30 Existing literature is largely confined to single-locus candidate gene association studies or retrospective cohorts with restricted sample sizes, which often lack reproducibility across diverse ethnic populations.31 Moreover, prospective clinical evidence evaluating the real-world utility of a combined PGx-guided strategy integrating both metabolic and target-gene information remains insufficient. Given the substantial ethnic differences in genetic background and antihypertensive drug response,32,33 further evidence from Chinese populations with hypertension is needed to facilitate the clinical translation of PGx-guided therapy.

Therefore, in the present study, we conducted an exploratory observational analysis of two multicenter prospective PGx-guided antihypertensive cohorts in Hunan and Fujian provinces, China, focusing on patients with hypertension who received CCB-containing regimens. In the Hunan cohort, the genotype-guided group received a comprehensive PGx-informed management strategy that incorporated both genotype-based medication recommendations and standardization management. This design allowed us to evaluate the association of a real-world PGx-informed management approach with short-term BP outcomes, but the independent contribution of genotype-guided prescribing could not be fully distinguished from that of intensified clinical management. To provide complementary evidence, we also analyzed the Fujian cohort, in which genotype-guided treatment adjustment was implemented after an initial 4-week standardized hypertension management phase. This design reduced background variability in clinical management and provided an opportunity to examine whether a strategy incorporating CCB-related genetic information remained associated with subsequent BP outcomes after the influence of standardized management had been minimized. By jointly evaluating these two complementary cohorts, we sought to examine the association between a PGx-guided antihypertensive strategy incorporating CACNA1C rs2238032 and CYP3A5 rs776746 information and short-term BP control, including its short-term safety profile, among Chinese patients receiving CCB-containing regimens, and to explore its potential clinical relevance in precision hypertension management.

Methods

Data Sources and Study Population

This study was an exploratory observational analysis of two previously completed multicenter, prospective, randomized controlled cohorts evaluating PGx-guided individualized antihypertensive therapy. The two parent cohorts were conducted in Hunan and Fujian provinces, China, respectively. The data used in this study were obtained from the “Hypertension Precision Medicine Database” of Xi’an Times Gene Medicine Technology Co., Ltd. The database was established from provincially supported key research and development projects after standardized case screening and data quality control, and has obtained data intellectual property registration certificates from the Shaanxi Intellectual Property Protection Center and the Shanghai Intellectual Property Administration (Certificate Nos. SZ2025220002120.7 and SZ2025520002384.0, respectively). The dataset used in the present study was obtained through data-sharing agreements.

The Hunan cohort was conducted from 2021 to 2024 across 17 hospitals in Hunan Province. The Fujian cohort was conducted from March 2022 to July 2024 across 8 hospitals in Fujian Province. The Hunan cohort was approved by the Clinical Research Ethics Committee of the Second Xiangya Hospital, Central South University (approval No. LYG2021041), and was registered in the Medical Research Registration and Filing Information System (https://www.medicalresearch.org.cn/index, registration No. MR-43-24-024233). The Fujian cohort was approved by the Ethics Committee of the First Affiliated Hospital of Fujian Medical University (approval No. [2021]213), and was registered at the Chinese Clinical Trial Registry (https://www.chictr.org.cn, registration No. ChiCTR2200057507). The present analysis was additionally approved by the Ethics Committee of the Third Xiangya Hospital, Central South University (approval No. kuai 26216). Both parent studies were conducted in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice. Written informed consent was obtained from all participants before enrollment. To protect participant privacy and ensure data security, all clinical data were de-identified and stored confidentially throughout the study.

The Hunan parent cohort enrolled patients with hypertension who met the following criteria: age >18 years; a clinical diagnosis of hypertension; and SBP ≥140 mmHg and/or DBP ≥90 mmHg measured at least three separate time points, according to the 2010 Chinese Guidelines for the Management of Hypertension.34 The main exclusion criteria were secondary hypertension; a definite diagnosis of stroke, poorly controlled diabetes, or severe heart failure, defined as New York Heart Association functional class III–IV, within the preceding 3 months; severe hepatic or renal dysfunction, defined as transaminase levels more than three times the upper limit of normal or chronic kidney disease stage 4–5; severe trauma or major surgery; pregnancy or lactation; and women of childbearing potential who were unable to use effective contraception.

The Fujian parent cohort enrolled patients aged ≥18 years with grade 1–2 primary hypertension, defined as SBP of 140–179 mmHg and/or DBP of 90–109 mmHg. The main exclusion criteria were use of three or more classes of antihypertensive drugs at enrollment; resting heart rate ≤55 beats/min; malignant tumors; hepatic or renal dysfunction; severe heart failure, defined as an ejection fraction <40%; coronary artery disease or stroke; major trauma or recent major surgery; pregnancy or lactation; and women of childbearing potential who were not using effective contraception.

Participants in the present analysis were selected from the two prospective PGx-guided individualized antihypertensive treatment cohorts described above. From the parent cohorts, we further included patients who were receiving CCB monotherapy or CCB-based combination antihypertensive therapy at the last follow-up, which was 4 weeks in the Hunan cohort and 8 weeks in the Fujian cohort. This selection was intended to examine the association between a genotype-guided antihypertensive strategy incorporating CCB-related PGx information and BP outcomes among patients receiving CCB-containing regimens.

Clinical Management Strategies in Parent Cohorts

In the Hunan parent cohort, eligible patients were assigned in a 1:1 ratio to either the genotype-guided group or the standard-treatment group in an open-label manner and were followed for 4 weeks. At baseline, patients in the genotype-guided group underwent genotyping and the genotyping results were used to assist clinicians in selecting and adjusting antihypertensive treatment. In addition, patients in the genotype-guided group received standardized management during follow-up. Patients in the standard-treatment group received conventional empirical antihypertensive therapy without reference to genetic information. BP levels, antihypertensive medication regimens, and safety outcomes were assessed at 4 weeks.

In the Fujian parent cohort, patients were randomly assigned in a 1:1 ratio to either the genotype-guided group or the standard-treatment group at baseline. Patients were blinded to treatment allocation, whereas investigators responsible for treatment and follow-up were aware of group assignment. All patients first received 4 weeks of standardized hypertension management, including lifestyle guidance, standardized BP measurement training, and guideline-based adjustment of antihypertensive therapy when necessary. This phase was intended to reduce variability across centers in treatment and establish a relatively consistent therapeutic background. After the 4-week visit, both groups continued to receive standardized management. In the genotype-guided group, antihypertensive regimens were further optimized according to PGx testing results, whereas the standard-treatment group continued to receive conventional guideline-based treatment without reference to genetic information. Both groups were then followed for an additional 4 weeks, and BP levels, medication use, and AEs were assessed at 8 weeks.

Because the Hunan and Fujian cohorts differed in study design, management procedures, timing of PGx intervention, and follow-up structure, the two cohorts were analyzed separately and were not pooled in the present study. The two cohorts were considered to provide complementary evidence regarding the association between genotype-guided antihypertensive management and BP outcomes in patients receiving CCB-containing regimens.

Blood Pressure Measurement in Parent Cohorts

BP was measured according to a standardized protocol throughout the study in parent cohorts. All participating centers followed a uniform procedure at each follow-up visit. Before measurement, participants were required to rest in a seated position for at least 5 minutes and to avoid smoking and caffeinated beverages for at least 30 minutes. BP was measured in the seated position three consecutive times, and the average of the three readings was used as the final SBP and DBP. Measurements were performed using validated upper-arm electronic BP monitors or calibrated standard mercury sphygmomanometers. All measurements were conducted by uniformly trained investigators.

Genotyping and Genotype-Guided Treatment Strategy in Parent Cohorts

The genetic variants analyzed in the present study were predefined in the parent cohorts based on their potential relevance to CCB pharmacological response. The selected variants included CACNA1C rs2238032, representing pharmacodynamic variability related to the CCB target, and CYP3A5 rs776746, representing pharmacokinetic variability related to CCB metabolism. Therefore, integrating these two genetic factors may provide complementary information regarding variability in CCB treatment response.

In the parent cohorts, peripheral venous blood samples were collected from participants in the genotype-guided group at baseline for genotyping. A commercial fluorescence polymerase chain reaction (PCR) detection kit (Guangyin Biotechnology Co., Jinan, China) was used to detect the target single-nucleotide polymorphisms (SNPs). Genotyping was performed using the TaqMan MGB probe method for a prespecified antihypertensive PGx panel. Briefly, 10 μL of whole blood was processed with a nucleic acid release reagent, and 1 μL of the processed sample was added to the PCR reaction mixture. PCR amplification was performed using the FQD-48A real-time fluorescence quantitative PCR system (Bioer Technology Co., Hangzhou, China) according to the manufacturer’s instructions. Genotypes were determined based on fluorescence signals and amplification curves. Positive and negative controls were included in each run to monitor the validity of the TaqMan MGB genotyping system, minimize experimental interference, and ensure the accuracy of genotype calls.

In the genotype-guided group, antihypertensive regimens for CCB were individualized according to prespecified PGx interpretation rules (Table S1). Investigators assessed the predicted metabolic capacity and potential drug response based on each patient’s genotype. The final treatment regimen was determined by clinicians after integrating genotype results, BP levels, comorbidities, potential adverse reaction risks, and previous medication history. Patients in the standard-treatment group received empirical antihypertensive therapy according to hypertension management guidelines and routine clinical practice, without using genotype information during medication adjustment.

Endpoints in the Present Study

The primary endpoint in the present study was the BP control rate at the last follow-up. BP control was defined as SBP <140 mmHg and DBP <90 mmHg. Secondary endpoints included changes in SBP and DBP, antihypertensive medication use, and short-term safety outcomes during the study period.

Statistical Analysis in the Present Study

All statistical analyses were performed using SPSS version 25.0 (SPSS Inc., Chicago, IL, USA). All statistical tests were two-sided, and a P value <0.05 was considered statistically significant. Continuous variables were expressed as the mean ± standard deviation (SD) or mean ± standard error (SE), as appropriate, and categorical variables were presented as numbers and percentages. The normality of continuous variables was assessed using the Kolmogorov–Smirnov test. For normally distributed continuous variables, between-group comparisons were performed using the independent-samples t-test, and within-group comparisons were performed using the paired-samples t-test. For non-normally distributed continuous variables, between-group comparisons were performed using the Mann–Whitney U-test, and within-group comparisons were performed using the Wilcoxon signed-rank test. Categorical variables were compared between groups using the chi-square test, and within-group paired categorical data were compared using the McNemar test.

Patients with missing key medication information, BP outcome data, or genotype information were excluded from the corresponding analyses. As the proportion of missing data for the majority of covariates in the final analysis population was low, no imputation was carried out, and all analyses were conducted using the available observed data. As this was an analysis of existing multicenter prospective cohorts, the sample size was determined by the number of eligible participants from the parent cohorts, and no separate a priori sample size calculation was performed. All subgroup and exploratory analyses were hypothesis-generating. No formal adjustment for multiple comparisons was performed.

In addition to descriptive analyses, univariable and multivariable logistic regression models were used to evaluate the association between the genotype-guided treatment strategy and BP control at the last follow-up. In these models, treatment group was included as the main independent variable, and BP control status at the last follow-up was included as the dependent variable. Because the present study was a secondary analysis restricted to patients receiving CCB-containing regimens within previously randomized cohorts, the baseline balance achieved by the original randomization may have been partially disrupted, thus potentially introducing selection bias. Therefore, the associations observed in the present secondary analysis should not be interpreted as equivalent to the treatment effects estimated from the original randomized cohorts. Multivariable models were further used to adjust for potential confounders. Covariates were selected a priori based on clinical relevance and potential associations with BP control. In the Hunan cohort, the multivariable model adjusted for demographic characteristics, including age and sex; baseline BP levels, including SBP and DBP; lifestyle-related factors, including smoking, drinking, high-salt diet, physical inactivity, and psychological stress; and comorbidities, including diabetes and dyslipidemia. Because the proportion of missing body mass index (BMI) data in the Hunan cohort was relatively high (23.57%), BMI was not included in the primary multivariable model for this cohort. In the Fujian cohort, the multivariable model adjusted for age, sex, BMI, 4-week BP levels (SBP and DBP), smoking, drinking, high-salt diet, physical inactivity, psychological stress, diabetes, and dyslipidemia. In addition, we performed a sensitivity analysis restricted to patients with available BMI data in the Hunan cohort. In this analysis, we compared a model adjusted for the same covariates as the primary Hunan model with a second model that additionally included BMI, to assess the influence of the missing BMI data.

Univariable and multivariable linear regression models were further used to assess the association between the genotype-guided treatment strategy and the magnitude of BP decrease. In the Hunan cohort, decreases in SBP and DBP from baseline to 4 weeks were analyzed. In the Fujian cohort, decreases in SBP and DBP from 4 weeks to 8 weeks were analyzed. The covariates included in the linear regression models were consistent with those used in the corresponding logistic regression models.

Because of the relatively small sample size of the Fujian cohort, prespecified subgroup analyses were conducted only in the Hunan cohort. Stratified multivariable logistic regression models were used to evaluate the association between the genotype-guided treatment strategy and 4-week BP control across subgroups defined by sex, age, and baseline hypertension grade, after adjustment for the covariates described above. Odds ratios (ORs) and 95% confidence intervals (CIs) were reported. Age was categorized according to the median value of the overall analysis population as <62 years and ≥62 years. Baseline hypertension grade was classified as grade 1, grade 2, or grade 3 according to the Chinese hypertension management guidelines,35 as shown in Table S2. Potential heterogeneity of treatment associations across subgroups was assessed by adding multiplicative interaction terms between treatment group and subgroup variables to the multivariable logistic regression models.

In the Fujian cohort, exploratory analyses were performed according to BP control status after the initial 4-week standardized management phase. Regression models were used to separately examine the association between the genotype-guided strategy and 8-week BP outcomes among patients with controlled BP and those with uncontrolled BP at 4 weeks. This analysis was conducted to explore whether the genotype-guided strategy was more strongly associated with subsequent outcomes among patients with different early responses to standardized management.

Inverse probability of treatment weighting (IPTW) based on propensity scores was performed separately within each cohort to reduce measured confounding. Propensity scores were estimated using prespecified covariates measured before the evaluated treatment period. In the Hunan cohort, these covariates included age, sex, baseline SBP, baseline DBP, smoking, drinking, high-salt diet, physical inactivity, psychological stress, diabetes, and dyslipidemia. In the Fujian cohort, these covariates included age, sex, BMI, 4-week SBP, 4-week DBP, smoking, drinking, high-salt diet, physical inactivity, psychological stress, diabetes, and dyslipidemia. Stabilized weights were applied to generate weighted pseudo-populations. Covariate balance before and after weighting was assessed using absolute standardized mean differences (SMDs), with values <0.10 considered indicative of adequate balance. Weighted logistic regression models were then used to estimate IPTW-adjusted ORs and 95% CIs for target BP control.

In addition, the distribution of specific CCB medications and patterns of antihypertensive regimen use were summarized descriptively by cohort, treatment group, and follow-up time point. The number of antihypertensive drugs used at different follow-up time points was compared between treatment groups to characterize medication burden. AEs during the study period were summarized descriptively by treatment group to evaluate short-term safety signals.

Results

Baseline Characteristics of the Study Population

In the Hunan parent cohort, 9260 patients with primary hypertension were screened, and 8691 completed the 4-week follow-up, including 4396 patients in the genotype-guided group and 4295 in the standard-treatment group. Among them, 1659 patients were excluded because medication information at the last follow-up was unavailable and CCB use could not be determined. An additional 2424 patients were excluded because they were not receiving CCB monotherapy or CCB-based combination therapy at the last follow-up. Finally, 4608 patients receiving CCB-containing regimens at the 4-week follow-up were included in the present analysis, including 1904 in the genotype-guided group and 2704 in the standard-treatment group (Figure 1).

Figure 1.

Flow diagram of participant selection from the parent cohort to the present analysis. (A) In the Hunan parent cohort, patients received either pharmacogenomics-guided treatment or conventional treatment and were followed for 4 weeks. (B) In the Fujian parent cohort, all patients underwent a 4-week standardized management period first. From 4–8 weeks, participants in the genotype-guided group received pharmacogenomics-guided treatment, whereas those in the standard-treatment group received conventional treatment. All patients were followed for 8 weeks in total.

In the Fujian parent cohort, 1027 patients were screened, and 754 patients with primary hypertension completed the 8-week follow-up, including 433 patients in the genotype-guided group and 321 in the standard-treatment group. At the last follow-up, 335 patients were excluded because they were not receiving CCB-containing regimens, and 1 patient was excluded because of incomplete genotype data. Finally, 418 patients were included in the present analysis, including 219 in the genotype-guided group and 199 in the standard-treatment group (Figure 1).

Baseline characteristics of the Hunan cohort are shown in Table 1. Because the present study was an observational secondary analysis restricted to patients receiving CCB-containing regimens at follow-up, the original balance achieved by randomization in the parent cohort may have been partially disrupted. Accordingly, several baseline characteristics differed between groups. Compared with the standard-treatment group, the genotype-guided group comprised younger patients, had a higher proportion of men, and displayed higher BMI and baseline DBP values. Baseline SBP did not differ significantly between groups. Regarding lifestyle-related factors, the genotype-guided group had higher proportions of smoking, drinking, high-salt diet, and psychological stress, whereas the proportion of physical inactivity was similar between groups. For comorbidities, the prevalence of diabetes did not differ significantly, whereas dyslipidemia was more frequent in the genotype-guided group. Baseline characteristics of the Fujian cohort are shown in Table 2. Patients in the genotype-guided group had higher baseline SBP than those in the standard-treatment group, whereas no statistically significant between-group differences were observed for the other baseline characteristics. Although several variables differed statistically between groups, these variables were further adjusted for in subsequent multivariable analyses.

Table 1.

Demographic and Baseline Clinical Characteristics of the Study Participants in the Hunan Cohort

Characteristics	Total (n=4608)	Genotype-Guided Group (n=1904)	Standard-Treatment Group (n=2704)	P value
Age (years)	61.78±12.35	59.89±13.03	63.11±11.68	<0.001
Gender				0.017
Male	2527 (54.84)	1084 (56.93)	1443 (53.37)
Female	2081 (45.16)	820 (43.07)	1261 (46.63)
BMI (kg/m2)	25.00±3.53	25.27±3.46	24.80±3.56	<0.001
Baseline SBP (mmHg)	156.72±12.66	156.63±13.36	156.79±12.14	0.228
Baseline DBP (mmHg)	92.23±11.06	93.23±11.62	91.53±10.59	<0.001
Smoking	524 (11.37)	271 (14.23)	253 (9.36)	<0.001
Drinking	377 (8.18)	193 (10.14)	184 (6.80)	<0.001
High-salt diet	262 (5.69)	150 (7.88)	112 (4.14)	<0.001
Physical inactivity	535 (11.61)	218 (11.45)	317 (11.72)	0.775
Psychological stress	119 (2.58)	68 (3.57)	51 (1.89)	<0.001
Diabetes	119 (2.58)	42 (2.21)	77 (2.85)	0.176
Dyslipidemia	63 (1.37)	37 (1.94)	26 (0.96)	0.005
CACNA1C (rs2238032)
TT	/	1686 (88.55)	/	/
GT	/	205 (10.77)	/	/
GG	/	13 (0.68)	/	/
CYP3A5 (rs776746)
AA	/	153 (8.04)	/	/
GA	/	683 (35.87)	/	/
GG	/	1068 (56.09)	/	/

Notes: Continuous variables are presented as mean ± standard deviation (SD), and categorical variables are presented as numbers (percentages).

Abbreviations: BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Table 2.

Demographic and Baseline Clinical Characteristics of the Study Participants in the Fujian Cohort

Characteristics	Total (n=418)	Genotype-Guided Group (n=219)	Standard-Treatment Group (n=199)	P value
Age (years)	54.75±12.28	53.85±12.80	55.73±11.64	0.117
Gender				0.832
Male	248 (59.33)	131 (59.82)	117 (58.79)
Female	170 (40.67)	88 (40.18)	82 (41.21)
BMI (kg/m2)	25.37±3.80	25.58±4.04	25.14±3.51	0.139
Baseline SBP (mmHg)	141.99±16.91	144.14±17.50	139.62±15.95	0.006
Baseline DBP (mmHg)	87.32±12.55	88.22±13.05	86.32±11.92	0.190
Smoking	96 (22.97)	43 (19.63)	53 (26.63)	0.089
Drinking	86 (20.57)	45 (20.55)	41 (20.60)	0.989
High-salt diet	86 (20.57)	51 (23.29)	35 (17.59)	0.150
Physical inactivity	135 (32.30)	71 (32.42)	64 (32.16)	0.955
Psychological stress	71 (16.99)	33 (15.07)	38 (19.10)	0.274
Diabetes	58 (13.88)	37 (16.89)	21 (10.55)	0.222
Dyslipidemia	53 (12.68)	28 (12.79)	25 (12.56)	0.562
CACNA1C (rs2238032)
TT	/	206 (94.06)	/	/
GT	/	13 (5.94)	/	/
GG	/	0 (0)	/	/
CYP3A5 (rs776746)
AA	/	10 (4.57)	/	/
GA	/	73 (33.33)	/	/
GG	/	136 (62.10)	/	/

Notes: Continuous variables are presented as mean ± standard deviation (SD), and categorical variables are presented as numbers (percentages).

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; SBP, systolic blood pressure.

Among patients in the genotype-guided group, we further summarized the distributions of the CCB-related genotypes, including CACNA1C rs2238032 and CYP3A5 rs776746. In the Hunan cohort, the CACNA1C rs2238032 TT genotype predominated, accounting for 1686 patients (88.55%), whereas the GT and GG genotypes were identified in 205 patients (10.77%) and 13 patients (0.68%), respectively. For CYP3A5 rs776746, the GG genotype was the most common genotype in the Hunan cohort, observed in 1068 patients (56.09%), followed by GA in 683 patients (35.87%) and AA in 153 patients (8.04%). In the Fujian cohort, a similar predominance of the CACNA1C rs2238032 TT genotype was observed, with 206 patients (94.06%) carrying TT and 13 patients (5.94%) carrying GT; no GG genotype was detected. The distribution of CYP3A5 rs776746 in the Fujian cohort was also comparable to that in the Hunan cohort, with GG being the most frequent genotype (136 patients, 62.10%), followed by GA (73 patients, 33.33%) and AA (10 patients, 4.57%).

Blood Pressure Outcomes

BP outcomes in the genotype-guided and standard-treatment groups are shown in Figure 2. In the Hunan cohort, after 4 weeks of follow-up, 1558 patients in the genotype-guided group achieved BP control, corresponding to a control rate of 81.83%. In the standard-treatment group, 678 patients achieved BP control, corresponding to a control rate of 25.07%. The 4-week BP control rate was significantly higher in the genotype-guided group than in the standard-treatment group (P < 0.001). The absolute difference in BP control rate between the genotype-guided and standard-treatment groups was 56.76% in the Hunan cohort.

Figure 2.

Blood pressure outcomes in the genotype-guided and standard-treatment groups. (A) Blood pressure control rate at 4 weeks in the Hunan cohort. (B) Blood pressure control rates at 4 and 8 weeks in the Fujian cohort. (C) Decreases (mean±standard error) in SBP and DBP from baseline to 4 weeks in Hunan cohort. (D) Decreases (mean±standard error) in SBP and DBP from baseline to 4 weeks and from 4 to 8 weeks in the Fujian cohort. **p < 0.01, ***p < 0.001.

Abbreviations: ns, not significant; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Regarding BP decrease, within-group comparisons showed that both SBP and DBP decreased significantly from baseline to 4 weeks in both treatment groups (P < 0.001). BP levels at different follow-up time points are shown in Table S3. In the Hunan cohort, the decrease in SBP from baseline to 4 weeks was greater in the genotype-guided group than in the standard-treatment group (25.98 ± 29.18 mmHg vs 12.34 ± 36.77 mmHg; P < 0.001). Similarly, the decrease in DBP was greater in the genotype-guided group than in the standard-treatment group (15.19 ± 12.48 mmHg vs 8.58 ± 11.15 mmHg; P < 0.001).

BP levels at different time points in the Fujian cohort are shown in Table S4. At 4 weeks, after the initial standardized management phase, the BP control rate was higher in the standard-treatment group than in the genotype-guided group (72.36% vs 54.34%; P < 0.001). At the 8-week last follow-up, the BP control rate increased to 81.74% in the genotype-guided group and was 77.39% in the standard-treatment group (P = 0.270). The corresponding absolute difference in BP control rate was 4.35% in the Fujian cohort. Within-group comparisons showed that the BP control rate increased significantly from 4 to 8 weeks in the genotype-guided group (P < 0.001), whereas no significant change was observed in the standard-treatment group (P = 0.193).

Changes in BP levels in the Fujian cohort were consistent with this pattern. From 4 to 8 weeks, the decrease in SBP was greater in the genotype-guided group than in the standard-treatment group (4.99 ± 14.03 mmHg vs 1.02 ± 11.57 mmHg; P = 0.003). The decrease in DBP was also greater in the genotype-guided group than in the standard-treatment group (2.37 ± 8.85 mmHg vs 0.31 ± 8.11 mmHg; P = 0.006). Within-group comparisons showed that both SBP and DBP decreased significantly from 4 to 8 weeks in the genotype-guided group (P < 0.001), whereas no statistically significant changes in SBP or DBP were observed in the standard-treatment group during the same period (P = 0.140 and P = 0.286, respectively).

Multivariable Regression Analyses

To further evaluate the association between the genotype-guided strategy and BP control at the last follow-up, univariable and multivariable logistic regression analyses were performed separately in the Hunan and Fujian cohorts. The results are shown in Table 3.

Table 3.

Logistic Regression Analysis of Blood Pressure Control at the Last Follow-up

Analysis	Hunan Cohort (4 Weeks)			Fujian Cohort (8 Weeks)
	OR	95% CI	P value	OR	95% CI	P value
Univariable analysis	13.46	11.64, 15.56	<0.001	1.31	0.81, 2.11	0.271
Multivariable analysis	14.56	12.49, 16.96	<0.001	2.43	1.34, 4.39	0.003

Abbreviations: OR, odds ratio; CI, confidence interval.

In the Hunan cohort, logistic regression analysis showed that the genotype-guided strategy was significantly associated with 4-week BP control after adjustment for prespecified covariates (adjusted OR = 14.56, 95% CI: 12.49–16.96; P < 0.001). Sensitivity analyses supported the robustness of the main findings. In the Hunan cohort, among participants with available BMI data, the association between the genotype-guided strategy and 4-week BP control remained stable after additional adjustment for BMI (Table S5). The adjusted OR was 19.53 (95% CI: 16.25–23.47; P < 0.001) in the model without BMI and 19.64 (95% CI: 16.33–23.61; P < 0.001) after further adjustment for BMI.

In the Fujian cohort, univariable logistic regression did not show a statistically significant association between the genotype-guided strategy and 8-week BP control (OR = 1.31, 95% CI: 0.81–2.11; P = 0.271). However, after further adjustment for 4-week BP levels and other covariates, the genotype-guided strategy was significantly associated with higher odds of 8-week BP control (adjusted OR = 2.43, 95% CI: 1.34–4.39; P = 0.003). These findings indicate that the genotype-guided strategy was associated with higher 4-week BP control in the Hunan cohort and with higher adjusted odds of 8-week BP control in the Fujian cohort.

Linear regression models were further used to assess the association between the genotype-guided strategy and the magnitude of BP decrease. The results are shown in Table 4. In the Hunan cohort, the genotype-guided strategy was significantly associated with greater decreases in both SBP (adjusted β =13.05, 95% CI:11.90–14.20, P<0.001) and DBP (adjusted β =5.62, 95% CI:5.13–6.11, P<0.001) from baseline to 4 weeks. In the Fujian cohort, from 4 to 8 weeks, the genotype-guided strategy was significantly associated with greater DBP decrease after multivariable adjustment (adjusted β = 1.71, 95% CI: 0.18–3.25; P = 0.029). For SBP decrease, the genotype-guided strategy was associated with a greater decrease in the univariable analysis (β = 3.97, 95% CI: 1.48–6.45; P = 0.002), but this association did not reach statistical significance after multivariable adjustment (adjusted β = 1.86, 95% CI: −0.39 to 4.11; P = 0.105). These results suggest that the association between the genotype-guided strategy and BP decrease was consistent in the Hunan cohort, whereas in the Fujian cohort the adjusted association was mainly observed for DBP decrease.

Table 4.

Linear Regression Analysis of the Decrease in Blood Pressure at the Last Follow-up

Analysis	Hunan Cohort (Baseline-4 Weeks)			Fujian Cohort (4 Weeks-8 Weeks)
	β	95% CI	P value	β	95% CI	P value
Decrease in SBP
Univariable analysis	12.34	11.07, 13.62	<0.001	3.97	1.48, 6.45	0.002
Multivariable analysis	13.05	11.90, 14.20	<0.001	1.86	−0.39, 4.11	0.105
Decrease in DBP
Univariable analysis	8.58	8.14, 9.03	<0.001	2.06	0.43, 3.70	0.014
Multivariable analysis	5.62	5.13, 6.11	<0.001	1.71	0.18, 3.25	0.029

Abbreviations: CI, confidence interval; SBP, systolic blood pressure; DBP, diastolic blood pressure.

In addition, IPTW analyses were performed to further account for measured baseline differences between groups. After weighting, covariate balance improved substantially in both cohorts, with all SMDs below 0.10 (Table 5). The IPTW-weighted results were consistent with the primary multivariable analyses (Table S6). The genotype-guided strategy remained associated with higher odds of target BP control in both the Hunan cohort (OR = 14.74, 95% CI: 12.68–17.13; P < 0.001) and the Fujian cohort (OR = 2.63, 95% CI: 1.48–4.68; P = 0.001).

Table 5.

Standardized Mean Differences Before and After Inverse Probability of Treatment Weighting

Characteristics	Hunan Before IPTW	Hunan After IPTW	Fujian Before IPTW	Fujian After IPTW
Age	0.260	0.006	0.154	0
Pre-intervention SBP	0.013	0.003	0.290	0.004
Pre-intervention DBP	0.153	0.003	0.206	0.001
BMI	/	/	0.119	0.011
Male sex	0.072	0.004	0.020	0.004
Smoking	0.151	0.003	0.082	0
Drinking	0.120	0	0.030	0.005
High-salt diet	0.158	0.004	0.054	0.027
Physical inactivity	0.008	0.006	0.061	0.023
Psychological stress	0.103	0.006	0.009	0
Diabetes	0.041	0	0.131	0.005
Dyslipidemia	0.082	0	0.060	0.008

Notes: Standardized mean difference values < 0.10 were considered indicative of adequate covariate balance between treatment groups. Pre-intervention blood pressure refers to baseline blood pressure in the Hunan cohort and 4-week blood pressure in the Fujian cohort.

Abbreviations: IPTW, inverse probability of treatment weighting; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Subgroup Analyses in the Hunan Cohort

Stratified multivariable logistic regression models were used to examine the association between the genotype-guided strategy and 4-week BP control across prespecified clinical subgroups. The results are shown in Figure 3.

Figure 3.

Subgroup analyses of blood pressure control at 4 weeks in the Hunan cohort. Detailed patient distribution, ORs, 95% CIs, P value and the corresponding forest plot of ORs (95% CI) showing the association between the genotype-guided strategy and blood pressure control across prespecified subgroups defined by sex, age, and baseline hypertension grade. P values for interaction are provided to assess heterogeneity in the observed associations across subgroups.

Abbreviations: OR, odds ratio; CI, confidence interval.

After stratification by sex, the genotype-guided strategy was significantly associated with higher odds of BP control in both male and female patients. No statistically significant interaction was observed between sex and treatment group (P for interaction = 0.821), suggesting that the association was generally consistent across sexes.

In age-stratified analyses, a statistically significant interaction was observed between age group and treatment group (P for interaction = 0.035). The association between the genotype-guided strategy and BP control appeared stronger among patients aged ≥62 years (adjusted OR = 17.73, 95% CI: 14.14–22.22; P < 0.001) than among those aged <62 years (adjusted OR = 12.41, 95% CI: 10.01–15.38; P < 0.001).

In analyses stratified by baseline hypertension grade, a statistically significant interaction was observed between hypertension grade and treatment group (P for interaction < 0.001). The genotype-guided strategy was associated with higher odds of BP control across all hypertension grades, but the magnitude of the association decreased as baseline hypertension grade increased.

Exploratory Analysis According to 4-Week Blood Pressure Control Status in the Fujian Cohort

Because all patients in the Fujian cohort received standardized hypertension management during the first 4 weeks, exploratory analyses were further performed according to BP control status at 4 weeks. This analysis aimed to explore whether the association between the genotype-guided strategy and subsequent 8-week outcomes differed between patients with controlled and uncontrolled BP after the initial standardized management phase. The results are shown in Tables S7 and S8.

Among patients whose BP was controlled at 4 weeks, the 8-week BP control rate was similar between the genotype-guided and standard-treatment groups (88.24% vs 86.81%; P = 0.728). Logistic regression analysis showed no statistically significant association between the genotype-guided strategy and 8-week BP control in this subgroup after multivariable adjustment (adjusted OR = 1.77, 95% CI: 0.71–4.44; P = 0.220). Similarly, linear regression analyses showed no statistically significant associations between the genotype-guided strategy and decreases in SBP or DBP from 4 to 8 weeks. The adjusted β was 0.63 for SBP decrease (95% CI: −1.87 to 3.13; P = 0.619) and 0.74 for DBP decrease (95% CI: −1.18 to 2.66; P = 0.446).

In contrast, among patients whose BP remained uncontrolled at 4 weeks, the 8-week BP control rate was higher in the genotype-guided group than in the standard-treatment group (74.00% vs 52.73%; P = 0.007). The genotype-guided strategy was significantly associated with higher odds of 8-week BP control after multivariable adjustment (adjusted OR = 4.84, 95% CI: 1.99–11.77; P < 0.001). In linear regression analyses, the genotype-guided strategy was significantly associated with greater DBP decrease from 4 to 8 weeks after adjustment (adjusted β = 4.11, 95% CI: 1.29–6.93; P = 0.005). The association with SBP decrease was directionally favorable but did not reach conventional statistical significance after adjustment (adjusted β = 4.49, 95% CI: −0.04 to 9.02; P = 0.052).

Overall, this exploratory analysis suggests that the genotype-guided strategy may be more strongly associated with subsequent BP improvement among patients whose BP remained uncontrolled after the initial 4-week standardized management phase, whereas no clear additional association was observed among patients whose BP had already been controlled at 4 weeks.

Safety and Medication Use

AEs recorded during the study period are summarized in Tables S9 and S10. Overall, the incidence of AEs was low in both cohorts. In the Hunan cohort, no AEs were recorded in the genotype-guided group at 4 weeks, whereas 2 AEs were recorded in the standard-treatment group, corresponding to an incidence of 0.07%. In the Fujian cohort, 18 AEs were recorded at 4 weeks, including 6 events in the genotype-guided group (2.74%) and 12 events in the standard-treatment group (6.03%). At 8 weeks, no AEs were recorded in the genotype-guided group, whereas 3 AEs were recorded in the standard-treatment group (1.51%). The reported AEs were mostly mild symptoms commonly observed during antihypertensive therapy, most commonly comprising dizziness, headache, cough, gastrointestinal symptoms, bradycardia, and peripheral edema. Given the small number of events and the short follow-up duration, no obvious short-term safety signal suggesting increased AE risk with the genotype-guided strategy was observed.

The number of antihypertensive drugs used at different follow-up time points is shown in Table S11. In the Hunan cohort, the number of antihypertensive drugs used at baseline did not differ significantly between the genotype-guided and standard-treatment groups (P = 0.689). At 4 weeks, however, patients in the genotype-guided group used a greater number of antihypertensive drugs than those in the standard-treatment group (P < 0.001). This difference may be related to the standardized management provided in the genotype-guided group, as well as more active treatment adjustment or optimization of combination therapy during follow-up. In the Fujian cohort, although the number of antihypertensive drugs differed between groups at baseline (P = 0.007), no significant between-group differences were observed after the initial standardized management phase at 4 weeks (P = 0.118) or at the 8-week follow-up (P = 0.124). These findings suggest that no obvious difference in medication count was observed in the Fujian cohort during the evaluated genotype-guided optimization period, and that the observed association with BP outcomes was not primarily reflected by an increase in the number of antihypertensive drugs at the last follow-up.

Antihypertensive regimen patterns were further summarized descriptively (Table S12). In the Hunan cohort, the proportion of patients receiving combination therapy increased in both groups from baseline to 4 weeks, with a more pronounced increase in the genotype-guided group. At 4 weeks, two-drug or more complex combination therapy was used in 65.65% of patients in the genotype-guided group and 43.57% in the standard-treatment group. In the Fujian cohort, combination therapy was common in both groups after standardized management. From 4 to 8 weeks, the genotype-guided group showed a modest decrease in the proportion of monotherapy and corresponding increases in the proportions of two-drug and three-drug combination therapy, whereas the distribution of regimen complexity remained relatively stable in the standard-treatment group.

The distribution of specific CCB medications was generally consistent across treatment groups and follow-up time points (Table S13). Amlodipine and nifedipine were the most frequently used CCBs in both cohorts. In the Hunan cohort, amlodipine accounted for approximately 60% of CCB use at 4 weeks in both groups, followed by nifedipine. Similar patterns were observed in the Fujian cohort, where amlodipine and nifedipine remained the predominant CCBs at both 4 and 8 weeks. These descriptive findings suggest that the observed associations were not driven by a marked shift toward a single specific CCB medication.

Discussion

In this exploratory analysis of two multicenter prospective randomized parent cohorts, we examined the association between a genotype-guided antihypertensive strategy incorporating CCB-related PGx information and BP outcomes among Chinese patients with hypertension receiving CCB-containing regimens. Overall, the genotype-guided strategy was associated with a higher 4-week BP control rate and greater decreases in both SBP and DBP in the Hunan cohort. In the Fujian cohort, the strategy was associated with higher adjusted odds of 8-week BP control and greater DBP decrease, with a more evident association among patients whose BP remained uncontrolled after the initial 4-week standardized management phase. No obvious short-term safety signal was observed. These findings suggest that PGx-informed treatment optimization may be clinically relevant for individualized hypertension management in patients receiving CCB-containing regimens, although the results should be interpreted as exploratory associations rather than definitive causal effects.

Interindividual variability in antihypertensive drug response has long been recognized as a major challenge in hypertension management, and inherited genetic variation may contribute to this heterogeneity.36 Identifying clinically relevant polymorphisms may help improve the prediction of individual BP response to specific antihypertensive agents. From a pharmacological perspective, CACNA1C and CYP3A5 are biologically relevant to CCB therapy. CACNA1C encodes the α1C subunit of the L-type voltage-dependent calcium channel, which is an important pharmacological target through which CCBs exert vasodilatory and antihypertensive effects.37 CYP3A5, a member of the CYP3A enzyme family, contributes to the metabolism of multiple drugs, and the CYP3A5*3 allele is associated with reduced expression of functional CYP3A5 protein.15,16 Therefore, incorporating both a pharmacodynamic target-gene locus and a pharmacokinetic metabolism-related locus may provide a more comprehensive framework for understanding variability in CCB-related treatment response than single-locus evaluation alone.

Previous studies have explored associations between individual genetic loci and CCB response. For example, Bremer et al reported an association between the CACNA1C rs2238032 variant and uncontrolled hypertension in Caucasian patients.14 Studies by Zhao et al19 and Huang et al18 suggested that CYP3A5 genetic variation may influence CCB metabolism and BP response in Chinese populations. However, most existing evidence has been derived from single-locus candidate gene studies, retrospective cohorts, or studies with limited sample sizes, and the current level of evidence supporting routine PGx-guided CCB therapy remains limited.30 By evaluating patients receiving CCB-containing regimens within two prospective clinical cohorts and considering both metabolism-related and target-gene information, the present study adds clinical evidence regarding the potential relevance of PGx-informed treatment optimization in real-world hypertension management.

The differences in study design and management procedures between the Hunan and Fujian cohorts are important for interpreting the observed findings. The Hunan cohort reflected a more comprehensive real-world management model, in which genotype-guided treatment was accompanied by structured follow-up and more active treatment optimization. Therefore, the large difference in 4-week BP control between the genotype-guided and standard-treatment groups, as well as the large adjusted OR observed in the Hunan cohort, likely reflects the combined influence of PGx-informed prescribing, clinical management, and treatment adjustment rather than the isolated effect of genotyping alone. However, this design also reflects the real-world implementation process of PGx-guided care, in which genetic information is typically translated into clinical actions through interpretation frameworks, decision-support systems, and treatment optimization workflows rather than being applied as an isolated intervention.38–40 Previous implementation studies have emphasized that the clinical utility of PGx depends not only on the availability of genetic information but also on its integration into clinical decision-making processes.39,41 Nevertheless, the contribution of PGx information and intensified clinical management cannot be quantitatively separated in the Hunan cohort. Future randomized trials with designs that isolate the independent effect of genotype-guided prescribing from treatment intensification are warranted. To provide complementary evidence, we further examined the Fujian cohort, in which all participants received an initial standardized management phase before genotype-guided optimization. This design may have reduced background variability in clinical care and allowed a more focused evaluation of the additional association of PGx-guided adjustment with subsequent BP outcomes. The relatively high 8-week BP control rates in both groups of the Fujian cohort also highlight the importance of standardized hypertension management as a foundation for improving short-term BP control.42–45

It is also noteworthy that baseline or pre-intervention BP levels differed between groups in some analyses. In the Hunan cohort, baseline DBP was higher in the genotype-guided group than in the standard-treatment group. In the Fujian cohort, 4-week SBP and DBP before the PGx-guided optimization phase were also higher in the genotype-guided group. Higher initial BP is generally associated with greater difficulty in achieving target control and may require more intensive or longer treatment.46 Despite these imbalances, the genotype-guided group showed greater BP decreases and favorable adjusted associations with BP control. In the Fujian cohort, the unadjusted difference in 8-week BP control between groups did not reach statistical significance, whereas the adjusted model showed a significant association after accounting for 4-week BP levels and other covariates. This finding is clinically plausible because 4-week BP status was strongly related to subsequent 8-week control. Nevertheless, this adjusted association should be interpreted cautiously, as residual confounding and regression to the mean cannot be fully excluded.

Several sensitivity analyses further supported the robustness of the main findings. In the Hunan cohort, among participants with available BMI data, the association between the genotype-guided strategy and 4-week BP control remained stable after additional adjustment for BMI. Moreover, IPTW analyses substantially improved measured covariate balance in both cohorts, with all SMDs below 0.10 after weighting. The IPTW-weighted results were consistent with the primary multivariable analyses, showing that the genotype-guided strategy remained associated with BP control in both the Hunan and Fujian cohorts. However, IPTW can only account for measured covariates included in the propensity score model, and unmeasured confounding, treatment-selection processes, and post-randomization selection related to follow-up CCB use may still have influenced the results.

Subgroup analyses in the Hunan cohort suggested that the association between the genotype-guided strategy and BP control was generally consistent across prespecified clinical subgroups. Both male and female patients showed favorable associations, and no significant interaction by sex was observed. A significant interaction was observed by age, with a stronger association among patients aged ≥62 years. Age-related changes in vascular stiffness, baroreflex sensitivity, renal function, drug metabolism, and polypharmacy may increase variability in antihypertensive response, potentially making individualized treatment strategies more clinically relevant in older patients.47 In analyses stratified by baseline hypertension grade, the genotype-guided strategy was associated with higher odds of BP control across all grades, but the magnitude of association decreased as hypertension grade increased. This pattern may reflect the greater absolute BP decrease required for patients with more severe hypertension to reach the target of <140/90 mmHg.48,49 Higher-grade hypertension may also reflect more complex pathophysiological mechanisms, including increased volume load, sodium retention, renin–angiotensin–aldosterone system activation, sympathetic overactivity, vascular remodeling, arterial stiffness, and endothelial dysfunction.50–52 Therefore, short-term PGx-informed optimization alone may be insufficient to fully overcome the treatment challenges in patients with more severe hypertension. Because these subgroup analyses were exploratory and some subgroups had limited sample sizes, the findings should be considered hypothesis-generating.

The exploratory analysis in the Fujian cohort according to 4-week BP control status provides additional clinical insight. Among patients whose BP remained uncontrolled after the initial standardized management phase, the genotype-guided strategy was associated with a higher 8-week control rate and greater DBP decrease. In contrast, among patients whose BP was already controlled at 4 weeks, no clear additional association was observed. This finding suggests that PGx-guided treatment optimization may be more clinically useful in patients with inadequate response to standardized management. In such patients, suboptimal BP control may reflect mismatched drug selection, variability in drug metabolism, insufficient regimen optimization, or reduced pharmacodynamic responsiveness.51 PGx information may help identify potential differences in drug response and support subsequent individualized treatment adjustment.10 Conversely, for patients who have already achieved target BP after standardized care, the incremental value of further genotype-guided adjustment may be limited.

Medication-use analyses provide important context for interpreting the observed associations. In the Hunan cohort, the number of antihypertensive drugs did not differ between groups at baseline, but was higher in the genotype-guided group at 4 weeks. This finding suggests that the genotype-guided strategy in the Hunan cohort may have been accompanied by more active treatment adjustment or optimization of combination therapy. Therefore, the favorable BP outcomes in this cohort should not be attributed solely to genotype information, but rather to the broader PGx-informed management strategy. In the Fujian cohort, although the number of antihypertensive drugs differed at baseline, this difference was no longer significant after the initial standardized management phase and remained nonsignificant at 8 weeks. It should also be noted that the number of antihypertensive drugs is only a crude measure of medication burden and does not fully capture drug dose, use of fixed-dose combinations, dosing frequency, regimen complexity, adherence, economic burden, or long-term tolerability. Therefore, findings related to medication count should be interpreted cautiously, and future studies should incorporate more detailed assessments of dose intensity and specific combination regimens.

Further analyses of regimen patterns and CCB medication types also support a cautious interpretation. In the Hunan cohort, combination therapy became more common from baseline to 4 weeks, particularly in the genotype-guided group. In the Fujian cohort, from 4 to 8 weeks, the genotype-guided group showed a modest shift from monotherapy toward two-drug or three-drug combination therapy, whereas the distribution of regimen complexity remained relatively stable in the standard-treatment group. These findings suggest that regimen optimization may be an important pathway through which PGx-informed management is associated with BP outcomes. At the same time, the distribution of specific CCB medications was broadly similar across groups and follow-up time points, with amlodipine and nifedipine being the predominant CCBs in both cohorts. Therefore, the observed associations are unlikely to be explained by a marked shift toward a single specific CCB medication. Nevertheless, treatment adjustment itself may act as an intermediate factor between PGx information and BP outcomes, and the present study cannot fully separate the relative contributions of genotype information, clinician decision-making, treatment intensification, and combination-regimen optimization.

The safety findings should also be interpreted with caution. AEs were infrequent in both cohorts, and no obvious short-term safety signal suggesting increased AE risk with the genotype-guided strategy was observed. However, the short follow-up duration and small number of AEs limit the ability to evaluate rare adverse reactions, long-term tolerability, and cardiovascular safety outcomes. Therefore, the present findings should not be interpreted as definitive evidence of long-term safety.

From a translational perspective, several issues require consideration before widespread implementation of PGx-guided antihypertensive therapy. Although PGx-guided antihypertensive therapy has shown potential clinical utility, the routine implementation of genotyping in hypertension care remains challenging due to the genetic complexity, variability in drug responses, and practical implementation barriers.25 Beyond clinical and operational feasibility, the economic sustainability of PGx-guided strategies represents another critical consideration for their broader adoption. Although PGx testing has shown potential cost-effectiveness in selected clinical settings, its economic value depends on testing costs, clinical effectiveness, healthcare systems, and the prevalence of actionable genetic variants.53–55 Therefore, dedicated economic evaluations specific to hypertension management are needed. Furthermore, identifying patients most likely to benefit remains critical. In the present study, patients with uncontrolled BP appeared to derive greater benefit from the PGx-informed strategy, suggesting that individuals with inadequate response to conventional treatment may represent a potential target population. A multicenter randomized trial indicated that certain subgroups, including female patients, may experience greater benefit.23 However, these findings require further validation in larger and more diverse populations.

In addition, polygenic risk scores (PRS) aggregate information from multiple genetic variants and may provide additional predictive value beyond individual candidate variants.56 Artificial intelligence (AI) has recently emerged as a promising tool for improving the accuracy and efficiency of precision cardiovascular medicine by integrating complex multimodal biomedical datasets to generate multimodal insights and facilitate precision medicine.57,58 Future precision hypertension strategies may integrate PRS with AI-based predictive algorithms, which may provide more comprehensive individualized treatment recommendations.

Several limitations should be acknowledged. First, the present analysis was restricted to patients receiving CCB-containing regimens, and the original balance achieved by randomization may have been partially disrupted, thus potentially introducing selection bias. Although multivariable adjustment and IPTW analyses were performed, residual confounding cannot be fully excluded. In addition, multiple subgroup and exploratory analyses were conducted without formal adjustment for multiple comparisons. Therefore, these findings should be interpreted as exploratory and hypothesis-generating rather than definitive evidence of clinical benefit. Second, in the Hunan cohort, genotype-guided treatment was accompanied by more intensive treatment optimization and structured management, making it difficult to distinguish the independent contribution of PGx guidance from that of enhanced clinical management. Therefore, the findings from the Hunan and Fujian cohorts should be viewed as complementary evidence. Third, the follow-up durations in both cohorts were relatively short, which precluded any assessment of long-term cardiovascular outcomes. Fourth, this analysis focused on only two variants related to CCB response and did not evaluate other potentially relevant loci that may also modulate CCB responses, such as CACNB2 and KCNMB1. Furthermore, medication adherence, which is an important determinant of BP control, was not systematically collected in the parent cohorts. Therefore, the potential influence of adherence-related factors could not be fully evaluated in the present study. Finally, although the analysis focused on patients receiving CCB-containing regimens, the potential influence of concomitant antihypertensive agents and treatment intensification cannot be fully excluded. Future prospective studies with stricter control of combination therapy, more detailed medication-dose information, longer follow-up, or designs focusing on CCB monotherapy are needed to better clarify the independent clinical value of CCB-related PGx guidance.

Conclusions

In this exploratory secondary analysis of two prospective antihypertensive cohorts, a genotype-guided antihypertensive strategy incorporating CCB-related PGx information, including CACNA1C rs2238032 and CYP3A5 rs776746, was associated with better short-term BP control and greater BP decrease among Chinese patients with hypertension receiving CCB-containing regimens. The association appeared particularly evident among patients whose BP remained uncontrolled after standardized management and among certain clinical subgroups, including older patients and those with lower baseline hypertension grades. No obvious short-term safety signal was observed. These findings should be interpreted as exploratory and hypothesis-generating. Further prospective randomized studies specifically designed to evaluate the independent clinical effect of CACNA1C/CYP3A5-guided CCB therapy are warranted before routine implementation in clinical practice to confirm its clinical effectiveness, safety, and long-term value in hypertension management.

Funding Statement

This work was supported by the National Natural Science Foundation of China (No. 82574507), the Scientific Research Program of FuRong Laboratory (No. 2025PT5029), and the Major Research Project for High-Level Health and Wellness Talents of Hunan Province (R2023042).

Data Sharing Statement

Due to ethical, privacy, and data ownership restrictions involving patient-level clinical and genetic information, the data are not publicly available. De-identified data may be available from the corresponding author upon reasonable request and subject to approval by the data owner and relevant institutions.

Ethical Approval

The Hunan cohort was approved by the Clinical Research Ethics Committee of the Second Xiangya Hospital, Central South University (approval No. LYG2021041), and was registered in the Medical Research Registration and Filing Information System (https://www.medicalresearch.org.cn/index, registration No. MR-43-24-024233). The Fujian cohort was approved by the Ethics Committee of the First Affiliated Hospital of Fujian Medical University (approval No. [2021]213), and was registered at the Chinese Clinical Trial Registry (https://www.chictr.org.cn, registration No. ChiCTR2200057507). The present analysis was additionally approved by the Ethics Committee of the Third Xiangya Hospital, Central South University (approval No. kuai 26216). Both parent studies were conducted in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice. Written informed consent was obtained from all participants before enrollment. To protect participant privacy and ensure data security, all clinical data were de-identified and stored confidentially throughout the study.

Author Contributions

Kunhong Deng and Yan Guo contributed equally to this work and share first authorship. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

Yu Wang, Di Meng, and Meng Zhang are employees of Xi’an Times Gene Medicine Technology Co., Ltd., which provided access to the Hypertension Precision Medicine Database used in this study. The remaining authors declare that they have no conflicts of interest.