Effect of donor GSTM3 rs7483 genetic variant on tacrolimus elimination in the early period after liver transplantation

Patients

From Jan 2017 and Dec 2020, we meticulously screened patients who underwent orthotopic liver transplantation at The Third Affiliated Hospital of Sun Yat-sen University, China (Cohort A = 110, Cohort B = 93.) The criteria were: (1) age ≥18 years, (2) LT from DCD/DBD, (3) immunosuppressive regimen was triple therapy incorporating tacrolimus, mycophenolate, and steroid, (4) Signed informed consent. The exclusion criteria were: (1) multiorgan transplant patients, (2) follow-up time less than 1 month, (3) Immunosuppressive regimen altered (e.g., to cyclosporin). The patient characteristics were summarized in Table 1.

Ethics statement

Human participants written informed consent was obtained from the subjects or the next-of-kin. The retrospective research was approved by the Ethics Committee of The Third Affiliated Hospital of Sun Yat-sen University (A2023-232-01). All procedures were performed in accordance with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Assurances were made to ensure that no livers were obtained from the executed prisoners (1) Voluntary organ donation (DCD/DBD), (2) human participants written informed consent was obtained from the next-of-kin of donors, (3) the quality of donor’s liver meets the transplantation standard, (4) China made posthumous voluntary donation the only legitimate source of organs in 2015).

Data collection

Therapeutic drug monitoring (TDM) was performed routinely after LT. The C₀ was measured in laboratories by the Pro-TracTMII tacrolimus ELISA kit (Diasorin) with a microparticle enzyme immunoassay (ELx 800NB analyzer; BioTek), using the whole blood collected before the morning administration. We used the Tac C/D ratio (ug/L per mg/kg) as a measure of Tac pharmacokinetics, calculating it by dividing the trough concentration (ug/L) by the dosage adjusted for body weight (mg/kg). Clinical data encompassed demographic insights (age, gender, child-pugh score, primary liver disease), liver function index (ALT, AST, TB, DB, Alb), and renal function index (Cr, Urea).

Genomic DNA isolation & genotyping

Genomic DNA obtained from donor and recipient hepatic samples was purified using the AllPrep DNA/RNA mini-kit (Qiagen, Hilden, Germany), previously secured at −80 °C. In the training set (Cohort A), genomic DNA from 110 patients was genotyped by Affymetrix DMET Plus array according to the molecular inversion probe (MIP) technology as previously described (Dumaual et al., 2007; Di Martino et al., 2011a; Di Martino et al., 2011b). And in the validated set (Cohort B), genotyping of CYP3A5 rs776746 and GSTM3 rs7483 was conducted using the Sequenom MassARRAY SNP genotyping platform (Sequenom, San Diego, CA, USA) (Gabriel, Ziaugra & Tabbaa, 2009). The sequencing primers for rs776746 and rs7483 were as follows: rs776746: forward 5′-AGGAAGCCAGACTTTGATCATTATGTT-3′, reverse 5′-GAGAGTGGCATAGGAGATACCCA-3′; rs7483: forward 5′-CCAGTATCGCAGCGATTC AATT-3′, reverse 5′-GCCTACTTACAGTCTGATCAGTTCTG-3′.

Statistical analysis

SPSS version 19.0 (SPSS, Chicago, IL, USA) was used for statistical analysis. Genetic equilibrium and allele distribution were analyzed using the SHEsis software platform (Shi & He, 2005). Tac C/D ratios were assessed for normality of distribution and logarithmically transformed if non-normal. Mean substitution where we substituted the missing values. Tac C/D ratios between genotype groups were analyzed using Student independent t-tests or Mann–Whitney test. Given the null hypothesis that the group means are equal, the ANOVAs were conducted. We utilized stepwise multiple linear regression to evaluate the impact of the CYP3A5 rs776746 and GSTM3 rs7483 genetic variant on Tac C/D ratios, along with clinical characteristics such as ALT, AST, TB, DB, Alb, Cr, Urea. Variables with a univariate p < 0.10 were included in multivariate analysis. The enter method was used to confirm the results derived from the training dataset. We assessed the impact of genotype cluster on the risk of Tac blood concentrations >15 ug/L using logistic regression analysis. Statistical significance was determined by two-tailed p-values of less than 0.05.

Gene distribution

The distribution of genotypes of CYP3A5 rs776746 and GSTM1-5 SNPs within the DEMT microarray is presented in Table S1. Fourteen SNPs with no variants in our study population were excluded for further analysis. The remaining seven SNPs conformed to Hardy–Weinberg equilibrium (P > 0.05). No significant linkage disequilibrium was observed between the CYP3A5 rs776746 and the six GSTM2-5 SNPs.

Influence of CYP3A5 and GSTM2-5 genetic vatiants on Tac elimination (Cohort A)

The association between donor CYP3A5 and GSTM2-5 SNPs and Tac C/D ratios in the early period after LT was shown in Table 2. Among CYP3A5 rs776746 carriers, those with AA/AG genotypes have been observed to have lower Tac C/D ratios than GG genotype carriers at weeks 1, 2, 3, 4 (p = 0.005, 0.007, 0.002, <0.001, respectively). Tac C/D ratios of donor GSTM3 rs7483 AA genotype were 231.0 ± 164.9, 127.3 ± 73.6, 120.4 ± 82.4 and 116.1 ± 71.1 at weeks 1, 2, 3 and 4 respectively. For AG and GG genotype carriers, the corresponding Tac C/D ratios at each time point were 328.2 ± 243.6, 195.1 ± 146.6, 213.4 ± 219.6 and 235.20 ± 180.3. The differences were significant (p = 0.035, 0.010, 0.035, 0.002, respectively).

The effects of recipient CYP3A5 and GSTM2-5 SNPs on Tac C/D ratios in the early period after LT was shown in Table 3. Tac C/D ratios of recipient CYP3A5 rs776746 AA/AG carriers were 220.7 ± 190.0, 126.2 ± 77.1, 120.5 ± 107.3 and 141.7 ± 222.3 at week 1, 2, 3 and 4 respectively, and 328.8 ± 216.7, 188.7 ± 139.9, 202.0 ± 199.6 and 195.2 ± 190.5 for GG genotype carriers. Tac C/D ratios of recipient CYP3A5 rs776746 AA/AG carriers were significantly lower than GG carriers at all investigated time points (p = 0.001, 0.006, 0.005, 0.003, respectively). However, there was not significant association between the recipient GSTM2-5 genotype groups in the early post-transplantation period.

Multivariate analysis for factors influencing Tac metabolism (Cohort A and B)

We investigated the effect of the genetic and clinical factors on Tac metabolism in the early post-transplantation period through multiple linear regression analysis. The cofactors that were incorporated into the analysis included CYP3A5 rs776746 and donor GSTM3 rs7483, liver function indices (ALT, AST, TB, DB, Alb), and renal function indices (Cr and Urea). In the training set (Cohort A: Table 4), donor and recipient CYP3A5 rs776746, donor GSTM3 rs7483 and the liver function indices (TB, DB) were identified as independent predictors of Tac metabolism in the early period after LT.

To confirm the effect of significant factors on Tac metabolism, we proceeded with further analysis in the validating set (Cohort B: Table 5). Donor and recipient CYP3A5 rs776746 and donor GSTM3 rs7483 were significantly associated with Tac elimination in the early period after LT. In addition, total bilirubin was also a predictor of Tac elimination for week 1.

Combined effects of CYP3A5 rs776746 and GSTM3 rs7483 genotypes

Donor and recipient CYP3A5 rs776746 allele A and donor GSTM3 rs7483 allele A were associated with fast Tac metabolism as stated above, we combined CYP3A5 rs776746 and GSTM3 rs7483 genotypes and investigated the effects of the number of alleles associated with fast metabolism on Tac C/D ratios (Table 6). Group 1 consisted of 0-1 alleles (poor metabolizers); group 2 contained 2-3 alleles (intermediate metabolizers); group 3 contained 4-6 alleles (extensive metabolizers). With increasing numbers of alleles associated with fast metabolism, Tac C/D ratios were increasingly lower at each time points within the first month (group 1 >group 2 >group 3; p = 0.001, 0.004, <0.001, <0.001, respectively).

The geometric mean of Tac concentrations were 13.6 ug/L, 9.2 ug/L, 5.7 ug/L at week 1 for patients from group 1, group 2 and group 3, respectively. Logistic regression analysis showed that the risk of presenting a supratherapeutic C₀ (Tac >15 ug/L) at week 1 was significantly higher for group 1 (Fig. 1), compared with group 2 (odds ratio: 4.143; 95% CI [1.305–13.175]; p = 0.016) and group 3 (odds ratio: 3.295 ; 95% CI [1.356–8.005]; p = 0.008). However, no significant differences were observed between the different groups regarding to the risk of a C₀ <8 ug/L (p = 0.742, 0.163, respectively). These results indicated that poor metabolizers require lower Tac doses to reach the target blood concentrations and genotype classification demonstrated a better predictive ability for the initial Tac doses after LT.