Effects of CYP3A5 Polymorphism on Rapid Progression of Chronic Kidney Disease: A Prospective, Multicentre Study

Personalised medicine is potentially useful to delay the progression of chronic kidney disease (CKD). The aim of this study was to determine the effects of CYP3A5 polymorphism in rapid CKD progression. This multicentre, observational, prospective cohort study was performed among adult CKD patients (≥18 years) with estimated glomerular filtration rate (eGFR) ≥30 mL/min/1.73 m2, who had ≥4 outpatient, non-emergency eGFR values during the three-year study period. The blood samples collected were analysed for CYP3A5*3 polymorphism. Rapid CKD progression was defined as eGFR decline of >5 mL/min/1.73 m2/year. Multiple logistic regression was then performed to identify the factors associated with rapid CKD progression. A total of 124 subjects consented to participate. The distribution of the genotypes adhered to the Hardy–Weinberg equilibrium (X2 = 0.237, p = 0.626). After adjusting for potential confounding factors via multiple logistic regression, the factors associated with rapid CKD progression were CYP3A5*3/*3 polymorphism (adjusted Odds Ratio [aOR] 4.190, 95% confidence interval [CI]: 1.268, 13.852), adjustments to antihypertensives, young age, dyslipidaemia, smoking and use of traditional/complementary medicine. CKD patients should be monitored closely for possible factors associated with rapid CKD progression to optimise clinical outcomes. The CYP3A5*3/*3 genotype could potentially be screened among CKD patients to offer more individualised management among these patients.


Introduction
Chronic kidney disease (CKD) is a rising public health problem with an alarming increasing trend [1]. During management of CKD patients, optimal control is important to delay disease progression [2]. The control of progression among CKD patients is very often reliant on various pharmacological treatments, such as antihypertensives and antidiabetic drugs, to manage complications associated with kidney failure. However, pharmacotherapy requires close monitoring in order to delay rapid progression of CKD. Rapid progression of CKD is associated with poorer clinical outcomes, including cardiovascular to determine the effects of CYP3A5*3 genetic polymorphisms in rapid CKD progression among an Asian population of CKD patients with routine nephrology care.

Study Design
This multicentre, observational, prospective cohort study was performed among adult CKD patients (aged ≥ 18 years), with an estimated glomerular filtration rate (eGFR) of 30 mL/min/1.73 m 2 and above [3], in three tertiary hospitals with specialist nephrology clinics in Malaysia. The study was approved by the Medical Research Ethics Committee, Malaysia (KKM.NIHSEC.P19-2320 (11))and the Universiti Kebangsaan Malaysia Research Ethic Committee (UKM PPI/111/8/JEP-2020-048). This study was conducted in compliance with ethical principles outlined in the Declaration of Helsinki and the Malaysian Good Clinical Practice Guidelines. The study report follows the Strengthening the Reporting Of Pharmacogenetic Studies (STROPS) guidelines [17].
Potentially eligible patients were identified via pre-screening from patient clinic lists and data from medical records. Patients were then recruited by the investigators during clinic visits from March 2020 until September 2020. Written informed consent was obtained from every subject prior to participation in this study. Patients with at least four outpatient visits and non-emergency eGFR values during the three-year study period were recruited, to ensure sufficient information was available to estimate the risk of CKD progression [18]. Patients who were pregnant, lactating, had incomplete medication regimen or without routine nephrology care were excluded.

Data Collection
After informed consent was obtained from subjects, each participant was assigned a unique subject identification number. Subjects' names were kept on a password-protected database. Demographic data, clinical information, laboratory data and medication characteristics for each subject from January 2018 to December 2020 were collected from the medical records from the respective institutions. Demographic data that were collected were age, sex and ethnicity.
The clinical information included the primary cause of CKD, co-morbidities, obesity (defined as body mass index (BMI) of 30 kg/m 2 and above), smoking status and blood pressure level, measured during clinic visits. The laboratory data collected were serum creatinine, albuminuria/proteinuria status and haemoglobin level during the study period.
Medication profiles of the subjects, adherence to medications and the use of traditional or complementary medicine (TCM) were compiled from the electronic medical record and prescription data, as well as from a structured interview with the subjects on their medication-taking behaviour, conducted by two investigators using a standardised questionnaire [2]. The data on the number of adjustments to antihypertensives (changes in dosage, frequency, timing or cessation, or commencement of new antihypertensives) were collected.

Study Definitions
Patients' renal function were quantified via eGFR calculated using the CKD Epidemiology Collaboration (CKD-EPI) equation. The CKD and albuminuria classification was based on the Kidney Disease: Improving Global Outcomes Workgroup (KDIGO) 2012 guidelines [18]. The definitions for each category are detailed in Table 1.
Patients' renal function were quantified via eGFR calculated using the CKD-EPI equation, using outpatient, non-emergency serum creatinine values [3]. Emergency serum creatinine values were defined as serum creatinine values obtained during visits to an emergency department. These values were excluded to avoid interference of transiently elevated serum creatinine values, due to acute illness or acute kidney injury (AKI) rather than actual CKD progression. The rapid progression of CKD was defined as a sustained decline in eGFR of more than 5 mL/min/1.73 m 2 /year, in line with the KDIGO CKD guidelines [18] (Table 1). AKI was detected using the KDIGO guideline criteria from the serum creatinine values, as well as information about AKI episodes that occurred in other healthcare institutions and were documented in the subjects' medical records. Protein-to-creatinine ratio (PCR) of less than 15 mg/mmol and below or negative to trace from urine protein reagent strip A2 PCR of 15-50 mg/mmol or trace to + from urine protein reagent strip A3 PCR of more than 50 mg/mmol, or greater than + from urine protein reagent strip

Rapid CKD progression
Sustained decline in eGFR of more than 5 mL/min/1.73 m 2 /year [18], based on the rate of annual eGFR change using linear regression model to identify the eGFR slope using the eGFR collected during the study period [3] Types of non-adherence [19] Initiation phase Medication is not taken by patient at all Implementation phase A dose is missed, omitted or an extra dose taken Persistence phase The medication is ceased without the instruction of prescriber Others TCM consumption The use of therapies not included in the treatment and medicines prescribed by hospitals or health clinics, such as the use of herbs (or botanicals), as well as over-the-counter nutritional and dietary supplements, based on patient recall [20] Routine nephrology care was defined as documentation of ambulatory nephrology care by a nephrologist in the medical records for at least 5 years [21].
Adjustments to antihypertensives included changes in dosage, frequency and timing, as well as cessation or commencement of new antihypertensives [20].
Adherence to medications was considered to be poor if there was a discrepancy between the prescribers' order and actual medication taken [19], based on documentations in medical records and patient recall ( Table 1).
Consumption of TCM was defined as the use of treatment and medicines not prescribed from hospitals or health clinics, such as the use of herbs (or botanicals), as well as over-the-counter nutritional and dietary supplements, based on patient recall [20].

Sample Size
Prior data indicate that the proportion of the CYP3A5*3/*3 genetic polymorphism status is 0.437 [6] and the population size of eligible patients was 180. If the Type I error probability and precision are 0.05 and 0.05, the sample size is 124 samples [22].

Detection of CYP3A5*3 Gene Polymorphism
Genomic DNA was extracted from blood samples using a DNeasy ® Blood and Tissue extraction kit (Qiagen, Hilden, Germany). The extracted genomic DNA was then analysed for purity and measured for concentration through OPTIZEN NanoQ spectrophotometer (Kaia Bio-Ingenieria, Daejeon, Korea), in which the 260/280 absorption ratio between 1.70 to 1.99 was considered to be DNA with sufficient purity without contamination during the extraction process [23]. The extracted genomic DNA was stored at −80 • C until use.
The intron 3 of the CYP3A5 gene encompassing the rs776746 (RefSeq NG_007938.2: g.12083A>G) polymorphism was amplified by the primers 5 -CAGCAAGAGTCTCACA CAGG-3 (Forward) and 5 -TACCACCCAGCTTAACGAAT-3 (Reverse) (IDT DNA, Singapore) and TopTaq Mastermix Kit (Qiagen, Hilden, Germany) using the Arktik TM Thermal Cycler (Thermo Fisher Scientific, Finland). The polymerase chain reaction (PCR) products were examined by gel electrophoresis to ensure the quality of PCR products through In-vitrogen™ 2% E-Gel™ Agarose Gels with SYBR Safe™ (Thermo Fisher Scientific, Kiryat Shmona, Israel) for 26 min, in which the PCR products were segregated by size and captured using the E-Gel ® Safe Imager TM Realtime Transilluminator (Life Technologies, Kiryat Shmona, Israel) [24]. The PCR product was then purified by a commercialised PCR purification kit (Applied Biosystems, UK) as a precondition for DNA Sanger sequencing. Purified DNA fragments were analysed using the BigDye ® Terminator version 3.1 cycle sequencing kit, which were run on a 96-capillary 3730xl DNA Analyzer at First BASE Laboratories Sdn. Bhd., Malaysia (developed by Applied Biosystem, USA, and produced by Thermo Fisher Scientific). The laboratory personnel were blinded, such that they were unable to distinguish samples with or without rapid CKD progression. The DNA sequences of the SNP results were transcribed using Sequence Scanner version 2.0 software (Applied Biosystems) and checked with the reference sequences in the Basic Local Alignment Search Tool (BLAST) program to confirm the presence of the polymorphism [25].

Statistical Analysis
All statistics were performed using IBM Statistical Package for Social Science for Windows version 23 (IBM Corp, Armonk, NY, USA). The results are presented as frequencies and percentages for categorical data, mean ± standard deviation (SD) for normally distributed numerical data, or as median (range) for non-normally distributed numerical data, based on the inspection of histograms. Adherence of the genotype groups to the Hardy-Weinberg equilibrium (HWE) assumption was examined. Expected percentages for each genotype group were calculated based on the Hardy-Weinberg equation using the allele frequencies (p 2 + 2pq + q 2 = 1). Chi-square test was then used to compare the allele and genotype distribution found with the predicted distribution. The observed genotype distribution was considered to be consistent with the assumptions of HWE if the p-value > 0.05 [24]. An independent T-test was used to compare the normally distributed numerical data between two groups, while the Mann-Whitney U test was used to compare the nonnormally distributed numerical data between two groups. One-way ANOVA test was used to analyse normally distributed numerical data for comparison of more than two groups. Pearson's Chi-square test for independence was used to study the association between categorical data and categorical data, while Fisher's exact test was used if assumptions of Pearson's Chi-square test for independence were not met.
To investigate the relationship between CYP3A5 polymorphism and rapid CKD progression among the study population, linear regression was first performed using outpatient, non-emergency serum creatinine values over 3 years, to quantify the eGFR slope of each subject to identify subjects with rapid CKD progression [3]. Multiple logistic regression was then applied to identify the factors associated with rapid CKD progression, as the assumptions to perform a linear regression were not met. A simple logistic regression was performed with each independent variable, to determine factors at a level of significance of p ≤ 0.05 [26]. A multiple stepwise logistic regression was then performed with all factors with p < 0.25, in which variables with p ≤ 0.05 were considered as factors associated with rapid progression of CKD, followed by an examination of multicollinearity and interaction between these factors, by a Variance Inflation Factor (VIF) of 5 and above defined as presence of multicollinearity [27]. The Hosmer-Lemeshow goodness-of-fit test, classification tables and area under the receiving operator characteristic (ROC) curve were used to investigate any misrepresentation of data [26].
The baseline eGFR did not differ significantly with variants of CYP3A5 allele (p = 0.731) nor genotypes (p = 0.438) ( Table 3). By the end of the study period, the average eGFR among subjects with the CYP3A5*3/*3 genotype of 45.7 ± 20.9 mL/min/1.73 m 2 was significantly lower than subjects with CYP3A5*1/*1 genotype (58.2 ± 32.6 mL/min/1.73 m 2 ) or CYP3A5*1/*3 genotype (63.3 ± 34.7 mL/min/1.73 m 2 ) (p = 0.030); while there was no statistically significant difference between subjects in terms of the allelic frequency (p = 0.862). Baseline albuminuria status did not differ significantly with variants of CYP3A5 allele (p = 1.000) nor genotype (p = 0.487). By the end of the study period, the distribution of albuminuria status was significantly different by allele (p = 0.007), as well as by genotype (p = 0.029) ( Table 3). From the perspective of genotype, subjects with CYP3A5*1/*1 genotype had a decline in the number of A3 albuminuria status by the end of the study period, from 28 (38.4%) to 22 (30.1%). Subjects with CYP3A5*1/*3 genotype had more A3 albuminuria status by the end of the study period, from 15 (34.9%) to 21 (48.8%), while subjects with CYP3A5*3/*3 genotype had the highest proportion of patients with A3 albuminuria category at baseline and at the end of the study period (n = 4, 50%). The number of patients with CYP3A5*1 allele in A1 category declined from 74 (39.2%) to 62 (32.8%) by the end of the study period. For subjects with CYP3A5*3 allele, A3 category patients increased from 23 (39.0%) to 29 (49.2%) by the end of the study period.
Twenty-nine (23.4%) subjects had rapid CKD progression, with 4 (13.8%) having the CYP3A5*3/*3 genotype. From the remaining 95 patients without rapid CKD progression, 4 (4.2%) patients had CYP3A5*3/*3 genotype, while 4 (13.8%) patients with rapid CKD progression had CYP3A5*3/*3 genotype. Table 4 shows the factors associated with rapid CKD progression. The simple logistic regression showed that adjustments to antihypertensives, CYP3A5*3/*3 polymorphism, previous episode of AKI, smoking and use of TCM were factors associated with rapid CKD progression, with a significance level of p < 0.05. After adjusting for potential confounding factors with p < 0.25 using multiple logistic regression, the factors associated with rapid CKD progression were adjustments to antihypertensives (

Discussion
Rapid progression of CKD occurred in approximately a fifth of our study population with mild-to-moderate CKD within a span of three years. To the best of our ability and knowledge, this is the first study describing the link between CYP3A5 polymorphism and rapid CKD progression. As the HWE assumptions were not violated, systematic error was likely absent in the genotyping assays. After adjusting for potential confounding factors, CYP3A5*3/*3 genotype, adjustments to antihypertensives, young age, dyslipidaemia, smoking and use of TCM were found to be factors associated with rapid CKD progression.
Our study supports previous reports [15,28] that have demonstrated that CYP3A5 polymorphism may be associated with factors of CKD progression. CYP3A5 is not only a drug metabolising enzyme present in the liver, it is also present in the kidneys and might have important physiological functions [7]. The CYP3A5*3 allele is associated with less expression of the CYP3A5 enzyme in the kidneys of healthy human adults compared to the wildtype, CYP3A5*1 [28]. The diminished CYP3A5 activity from the CYP3A5*3/*3 polymorphism reduces the protection against aldosterone-induced active sodium transport in the kidneys, as less intrarenal conversion of the corticosterone into 6βhydroxycorticosterone occurs [8]. Less inhibition of RAAS, which is related to glomerular hyperfiltration, exacerbates damage to the kidneys. This is supported by our finding that a greater proportion of subjects with CYP3A5*3/*3 genotype had category A3 albuminuria by the end of the study period, than those with CYP3A5*1/*1 or CYP3A5*3/*3 genotype, in line with findings from a previous study in which proteinuria was found to be an indicator of structural kidney damage [3]. Furthermore, maximal doses of ACEI or ARB did not improve renal function among patients with aldosterone excess [29], while CKD patients with aldosterone excess were reported to have accelerated progression of CKD [29].
Another possible mechanism for the effect of CYP3A5*3/*3 genotype on rapid CKD progression is the elevation of 20-HETE production. The arachidonic acid-derived metabolites of CYP3A5 enzyme are 19-HETE and 6β-hydroxycortisol [30]. The CYP3A5*3 polymorphism results in a reduced expression of CYP3A5 enzymes, which in turn reduces the formation of 19-HETE [30]. This may lead to an increased availability of the arachidonic acid precursor for greater production of 20-HETE. The increase of 20-HETE has been shown to increase renal vasoconstriction and peripheral vascular resistance, as 20-HETE is a potent vasoconstrictor that mediates angiotensin II-related renal effects in the proximal tubule and thick ascending limb of the loop of Henle [30]. In addition, a higher level of 20-HETE was recently identified as an independent predictor of CKD progression [15]. This demonstrates the need for more studies to be conducted to elucidate the mechanism of association between CYP3A5*3 polymorphism and rapid CKD progression. Genotyping may be beneficial to identify CKD patients with CYP3A5*3/*3 genotype for closer monitoring, given the association found between CYP3A5*3/*3 genotype and accelerated CKD progression.
There was a significant association between antihypertensive adjustments and rapid CKD progression found in the current work. During early progression, first-line RAAS inhibitors may have been stopped once patients presented with a rapidly deteriorating eGFR, which could account for part of the medication adjustments [31]. On the other hand, genetic predisposition may also account for frequent medication adjustments. Pharmacokinetic properties of CYP3A5 substrate drugs are known to differ according to CYP3A5 polymorphism [8]. In particular, calcium channel blocker antihypertensives, such as amlodipine [8], felodipine [9], diltiazem [10] and verapamil [11,12] have been reported to be affected by CYP3A5 polymorphism, with blood pressure responses varying according to their CYP3A5 polymorphism status [8,11,12]. However, the complexity of genetic effects are evident as metabolism of CYP3A5 substrates among individuals who are CYP3A5 non-expressors (those expressing CYP3A5*3/*3 genotype) have been reported to be carried out by the CYP3A4 enzyme, which is more prone to inductions and inhibitions by concurrent drugs [5]. These genetic effects may have led to the need for frequent medication adjustments for optimum outcome, as observed in the current work, supporting the need for closer monitoring of antihypertensive management.
Younger age and dyslipidaemia were also found to be associated with rapid CKD progression, in line with findings from previous studies on rapid CKD progression [32,33]. It is believed that the CKD aetiology is different in older patients, with more aggressive disease found among younger patients, while in adults some decline of eGFR is believed to occur as part of aging, rather than from deteriorating CKD [34]. On the other hand, in CKD patients, lipoproteins are oxidised, especially the small dense high-density lipoprotein cholesterol (LDL) particles, intermediate-density lipoproteins and chylomicron remnants [35]. Accumulation of these oxidised LDL, intermediate-density lipoproteins and chylomicron remnants accelerates systemic inflammation, through stimulating release of proinflammatory cytokines and chemokines from monocytes and macrophages [35]. The subsequent systemic inflammation and oxidative stress is believed to cause eGFR decline and CKD progression [35].
Smoking was also found to be associated with a higher risk of CKD progression, similar to previous work [36]. The potential mechanisms of smoking-associated CKD progression include smoking-induced hypoxic injury [36], myointimal hyperplasia of intrarenal arterioles [37] and adverse effects on intrarenal hemodynamics, through nicotineinduced release of angiotensin II [36]. This leads to increased activation of RAAS, as well as increased glomerular hypertension, which could potentially accelerate the progression of CKD. The study finding suggests that smoking cessation is an important component to preserve the kidney function of CKD patients who are currently smoking.
The association between TCM and CKD progression in Asian countries has been inconsistent to date [1,38]. A few studies have reported no association between TCM and CKD progression [38,39]. This may be due to the relatively short follow-up period that might not capture renal damage in the long term [38]. On the other hand, the lack of association has also been attributed to TCM that was prescribed by board-certified physicians and produced by pharmaceutical companies, which had certified manufacturing practices [39]. In the current work, we found a significant association between rapid CKD progression and TCM consumption, of which the TCM was used without the supervision of a registered practitioner or pharmacist. It was noted that many reported the use of TCM with unknown ingredients and quality. Most worrying is that some TCM, most often involving the use of herbal remedies popular among Asians, have been shown to contain nephrotoxic ingredients [40]. However, very often CKD patients report the use of TCM due to the lack of conventional medications that cure CKD, as well as the desire to see immediate improvement in their disease condition [40]. As there is no cure for CKD in conventional medicine, some patients might be inclined to use TCM, owing to their cultural beliefs and social influences [40]. Therefore, TCM use for CKD should be monitored closely for mitigation efforts in preventing rapid CKD progression.
There were a few limitations to our findings. Firstly, the findings of the study have limited applicability to advanced CKD patients with Stage 4 and above, in which the majority of these patients were shown to exhibit non-linear CKD progression [32]. In addition, the report of TCM use might be subject to recall bias at the time of interview. The potential effects of such bias were reduced by incorporating the report of TCM use from medical records. The lack of identification of TCM also provides fewer specific details of which moieties were nephrotoxic to the patients. Furthermore, the effects of other potential genes were not studied. As CYP3A4 and CYP3A5 enzymes have some overlapping in substrate specificity, the complete loss of metabolic activity with the CYP3A5*3 allele might pronounce the impact of genetic variation in CYP3A4 expression among these patients [5]. Therefore, further work involving genetic variants of CYP3A4 polymorphism could possibly improve the current findings. The study might be limited by the absence of direct measurement in the expression level or activities of CYP3A5. Future studies could be designed to investigate the activity of CYP3A5 with renal function through usage of endogenous markers, such as 4β-hydroxycholesterol. More studies could be conducted to investigate other outcomes, such as initiation of renal replacement therapy, heart failure and mortality, as well as CKD progression over a longer period of time.

Conclusions
In conclusion, CKD must be monitored closely to reduce the risk of rapid progression. This could potentially mean monitoring of patient genetics, as the CYP3A5*3/*3 genotype was found to be associated with accelerated CKD progression after adjusting for possible confounding factors. CKD patients with such characteristics, as well as those requiring antihypertensive adjustments, young age, dyslipidaemia, smokers and TCM users, may also benefit from intensified monitoring and care to reduce the propensity of developing adverse clinical outcomes. Most importantly, our findings suggest a potentially important role of CYP3A5 polymorphism in the pathogenesis of accelerated CKD progression. A personalised management approach could therefore be potentially useful for CKD patients based on genotyping data.