An adverse drug reaction (ADR) is defined as a noxious and unintended reaction to a drug at doses normally used in humans (World Health Organization, 2002). ADRs are a burden to the healthcare system, as patients with ADRs are found to have a longer duration of hospitalization and higher hospitalization costs (Suh et al., 2000). However, studies related to ADRs are mainly conducted among hospitalized patients rather than patients seen in outpatient settings (Laville et al., 2020). Among patients with chronic illness, chronic kidney disease (CKD) patients frequently report ADR. CKD patients might be more susceptible to ADRs given the need for multiple therapeutic agents to manage the various comorbidities of CKD (Laville et al., 2020).

ADRs could be potentially difficult to be predicted, given the multifactorial nature of ADRs, especially among CKD patients. Due to different physiological factors as a result of kidney function decline, it is difficult to extrapolate findings on the propensity of ADRs from existing studies among the general population to CKD patients. In addition, the unpredictable interindividual drug responses in the form of ADR might be driven by genetic polymorphisms that affect drug metabolism pathways and drug metabolism activity (Zanger and Schwab, 2013). Genetic polymorphisms that affect the drug metabolism pathways, such as the cytochrome P (CYP) 450 system, are of clinical prominence, as CYP450 metabolizes more than 80% of drugs (Zanger and Schwab, 2013). Therefore, CYP450 pharmacogenomics might be a promising approach to mitigate ADRs, especially in CKD patients.

The CYP3A family is the most abundantly expressed isoform of CYP450 enzymes, especially the CYP3A4 and CYP3A5 enzymes (Dorji et al., 2019). The presence of single-nucleotide polymorphism (SNP) in genes encoding these enzymes may result in variations in expression and activity of these enzymes (Lolodi et al., 2017). The consequential change in CYP enzyme activity causes alterations to the pharmacokinetic properties of the affected drugs, which then causes variation in the drug effects (Dorji et al., 2019). While SNPs to genes-encoding CYP3A4 enzyme are rare in East Asians, SNPs of the CYP3A5 gene, especially CYP3A5*3 (rs776746), are more common in Asian populations with an estimate of 65.7–71.3% (Dorji et al., 2019; Liang et al., 2021). The CYP3A5*3 polymorphism, in which guanine (G) replaces adenine (A) at position 6,986 of the gene, causes alternative splicing that affects the quantity of the functioning CYP3A5 enzyme, which reduces the metabolic capacities of CYP3A5-substrate drugs (Kuehl et al., 2001; Zhang et al., 2014). The wild-type allele of the CYP3A5 gene is CYP3A5*1, in which individuals with this allele express the CYP3A5 protein (Kuehl et al., 2001).

CYP3A5 polymorphism is potentially associated with hypertension and blood pressure regulation (Zhang et al., 2014). Furthermore, drug responses might differ according to the status of CYP3A5*3 polymorphism. The blood pressure response to the angiotensin-converting enzyme inhibitor (ACEI) was previously found to be significantly blunted in CYP3A5*1 carriers, which might be contributed by sodium-retaining effects and/or elevated activity of the renin-angiotensin-aldosterone system (RAAS) (Eap et al., 2007). The opposite might occur in the absence of the CYP3A5*1 allele, in which the adverse effect of hyperkalemia commonly seen with RAAS blockade by angiotensin-converting enzyme inhibitor (ACEI), angiotensin II blocker (ARB), or spironolactone might be potentiated. However, there is limited evidence linking CYP3A5 polymorphism with these agents. In addition, CKD patients with RAAS blockers might be more predisposed to clinically significant hyperkalemia, especially those with advanced CKD (Weir and Rolfe, 2010).

CYP3A5*3 polymorphism was previously reported to be associated with peripheral edema associated with amlodipine in general population (Liang et al., 2021). However, the generalizability of ADR studies among the general population to the CKD population might be limited, given the possibility of CKD influencing drug disposition (Yeung et al., 2014). In addition to reduction of renal clearance, CKD might attenuate a number of CYP-mediated metabolic pathways through several possible mechanisms, ranging from direct competitive inhibition by uremic constituents to alterations in gene transcription and translation (Yeung et al., 2014).

Hypertension is closely associated with CKD as a decline in kidney function precipitates the increase in blood pressure, but hypertension accelerates the progression of CKD (Judd and Calhoun, 2015). Optimization of antihypertensive therapy is therefore important for CKD patients. However, antihypertensive agents were found to be commonly related to ADRs (Laville et al., 2020). The identification of the potential contributing factors to ADRs related to antihypertensive agents is therefore important for preventive measures to minimize the suboptimal effects associated with antihypertensive agents in CKD patients. This study aimed to describe the relationship between CYP3A5*3 polymorphism and ADR to antihypertensive drugs in CKD patients.

This study provides a novel, pharmacogenomics-driven approach to assess ADRs related to antihypertensives in CKD. CKD patients might be more susceptible to ADRs given the need for multiple medications, in addition to physiological differences contributed by kidney function decline. In addition, the possibility of genetic susceptibility to ADRs has to be taken into consideration. The study findings improved the understanding of the relationship between genetic polymorphism and the development of ADRs related to antihypertensive agents in CKD patients.

In view of the common occurrence of CYP3A5*3 polymorphism among Southeast Asian populations, it is important to understand the association of the gene polymorphism with antihypertensives, which are commonly used in CKD populations and play an important role in managing CKD (KDIGO, 2012). Furthermore, CKD patients have different physiological states given the changes to their excretory functions in contrast with healthy patients, which might affect the propensity of ADRs among these patients. The differences between the general population and CKD patients reflected the limited applicability of the current literature to CKD patients.

ADRs with antihypertensives were found in 16.5% of patients, which was similar to the findings from previous studies of hospitalized cohorts of 10–20% (Danial et al., 2018; Laville et al., 2020). From our study findings, RAAS blockers were found to be the most common suspected agents, which was also similar to that previously reported (Laville et al., 2020). The study findings corresponded with the prescribing pattern of the antihypertensives in the study population, with more than half reported to use ACEI or ARB, corresponding to the RAAS blockade as an important foundation of pharmacotherapy in CKD (KDIGO, 2012). Of note, calcium channel blockers, another frequently implicated agent, were also commonly used among the study cohort. Contrary to the findings of Laville et al. in which acute kidney injury was most commonly reported across several antihypertensive classes (Laville et al., 2020), the present study recorded hyperkalemia as the most common ADR.

The current work demonstrated that anemia was a factor of antihypertensive-related ADR, which concurred with previous reports in CKD patients (Laville et al., 2020). While causality could not be established from this study, the mechanism linking anemia and antihypertensive-related ADR could be explored in future studies. Blood pressure control is likely suboptimal in CKD patients with fluid overload, requiring the use of antihypertensives. A potential mechanism to be examined is pharmacokinetic changes driven by increased fluid status in CKD patients that may have led to a diluted hemoglobin concentration due to volume overload, as supported by the findings that more than half of anemic CKD patients were found to have volume overload (Hung et al., 2015), which led to dilutional anemia (Hildegard Stancu et al., 2018). Nevertheless, the findings improved the understanding of the potential presentation of CKD patients possibly affected by ADRs, in which these patients might have other underlying concurrent medical issues that need to be corrected.

Poor medication adherence was found to be associated with antihypertensive-related ADR in CKD patients, which corroborates the outcomes from previous studies (Laville et al., 2020; Seng et al., 2020). While poor adherence might be multifactorial, the association between poor medication adherence and antihypertensive-related ADR might reflect underlying issues with medication use (Laville et al., 2020). Of these, the most obvious link between ADR and adherence has been reported in studies that document poor adherence to medications that were suspected to be associated with previous ADR episodes (McKillop and Joy, 2013). It is quite possible that the uncomfortable nature of certain ADRs may lead patients to forgo their medication. Given the dynamic nature of medication adherence that might change over time (Unni et al., 2015), this study highlighted the importance of assessing medication adherence regularly among CKD patients through optimization of pharmaceutical care. In addition, current findings that many antihypertensive regimens were adjusted post-ADRs also reflect the need for closer monitoring among these patients, given the close association of antihypertensive drug adjustments with poorer disease outcomes (Lee et al., 2021b).

Although there was a lack of association between CYP3A5*3 and ADR occurrence, further work is recommended. A challenge of studying CYP3A5*3 polymorphism is the potential compensatory functions by CYP3A4 enzyme, owing to the structural similarity with the CYP3A5 enzyme (Lolodi et al., 2017). The low activity of the CYP3A5 enzyme, as a result of CYP3A5*3 polymorphism, might be compensated by the CYP3A4 enzyme, which conceals the phenotype normally expected of the CYP3A5*3 polymorphism. Hence, future work should take CYP3A4 activity into account to evaluate the relationship between CYP3A5*3 polymorphism and susceptibility to ADRs related to antihypertensives in CKD.

Another challenge in utilizing pharmacogenetics in clinical practice is the potential confounding factor of TCM use. TCM may affect the drugs’ metabolism and the propensity of ADRs. An example of this is the use of CYP3A4/5 inhibitors, such as Allium sativum, which could potentially inhibit CYP3A5 activity, regardless of the polymorphism status. Of note, TCM use was not found to be a factor of ADRs from our findings. While TCM use is in an increasing trend in this country (Siti et al., 2009; Abdullah et al., 2018), the low number of TCM reported among our study population may reflect underreporting, which limits the ability to detect an association between TCM use and antihypertensive-associated ADRs.

This study’s major strength is the ascertainment of ADR cases from intensive medical record review, which addressed the possible limitation of underreporting found in many ADR studies using data from spontaneous pharmacovigilance reporting systems. In addition, CYP3A5*3 genotyping was performed using the gold standard direct sequencing method. However, a few limitations were noted due to the retrospective study design, which could not imply causality. As such, some ADRs, such as dizziness, might be caused by anemia rather than the implied drugs. On the other hand, there is a possibility of underestimation in the ADRs related to antihypertensives, as some ADRs might not be recorded in the medical records. The small sample size of the study cohort is another limitation in which the study might be underpowered to detect the association between CYP3A5*3 polymorphism and specific antihypertensives or specific doses. In addition, the retrospective nature of the study renders limitations in detailed analysis pertaining to drug–TCM interactions. It was challenging to verify the detailed composition of each TCM, not only due to incomplete data but also due to the fact that many TCMs are not registered with the Ministry of Health, which renders the complete contents of unregistered TCMs unknown (Lee et al., 2020).

In conclusion, the findings of the study indicate that ADRs related with antihypertensives among CKD patients were not associated with CYP3A5*3 polymorphism alone among our study cohort, which requires verification through further studies with greater sample size. These ADRs might be propagated by pathways other than CYP3A5-related metabolism. Additional prospective studies with multiple genetic polymorphisms, and assessments of fluid status could be conducted to verify the findings.