Mycophenolate及mTOR inhibitors對臺灣腎臟與肝臟移植病人tacrolimus藥動學之影響

Abstract

研究背景與目的： Tacrolimus（TAC）是腎臟及肝臟移植後免疫抑制維持治療的基石，在個體間存在著高度變異性，主要源自影響TAC代謝的細胞色素P450（cytochrome P450，CYP450）基因多型性。其他相關的基因變異包含ATP-Binding Cassette, Sub-Family B1（ABCB1）以及P450 oxidoreductase（POR）。除了基因的因素之外，許多臨床因素也會影響到TAC的藥動學。 TAC長期使用會有發生腎毒性的疑慮，現今建議TAC搭配沒有腎毒性的免疫抑制劑來降低腎毒性發生的風險，常見併用藥品有mycophenolate和mTOR inhibitors。然而，mycophenolate和mTOR inhibitors與TAC有部分重疊的代謝途徑和運輸蛋白，併用可能發生藥動學交互作用，造成TAC血中濃度過低或過高的疑慮。目前為止，有部分的研究探討mycophenolate以及mTOR inhibitors和TAC之間的藥動學交互作用，但是結論不一致；本研究將以自我對照研究方式（self-control study），納入影響TAC藥動學之基因及非基因因素，探討移植手術後至少6個月進入穩定期的病人，長期使用以TAC為主幹的免疫抑制療法中，在控制其他因素之下，sirolimus（SRL）、everolimus（EVL）和mycophenolate對TAC劑量調整谷濃度（dose-normalized TAC C0，dnC0）的影響。 方法： 本研究為回溯性研究（retrospective study），利用臺大醫療體系醫療整合資料庫（Integrated Medical Database, National Taiwan University Hospital, NTUH-iMD），分析本團隊過去研究中已納入的98位腎臟移植病人和89位肝臟移植病人，使用自我對照研究方式，分析移植術後至少6個月後進入穩定期，在增加或減少併用MMF或mTOR inhibitors；或調整併用之mycophenolate或mTOR inhibitors劑量前後6個月，並且控制基因型（CYP3A5*3、CYP3A4*1G、CYP3A4*18B、POR*28、ABCB1 1236C>T、ABCB1 2677G>T/A、ABCB1 3435C>T）和臨床因素後，mycophenolate或mTOR inhibitors對TAC dnC0的影響；再利用迴歸分析找出會顯著影響TAC dnC0的因素。統計分析在成對樣本將使用paired sample t-test，非常態分布會使用Wilcoxon signed rank test。連續性資料將使用Student's t-test，非常態分布會使用Mann-Whitney U test。類別性資料將使用Chi-squared test或Fisher's exact test。迴歸分析中若依變項為二元類別變項將使用羅吉斯迴歸（logistic regression），若依變項為連續變項則使用線性迴歸（linear regression）。 結果： 腎臟移植病人在移植手術後6個月到研究觀察結束前，用藥組合種類以及藥品轉換次數都多於肝臟移植病人。分析發現在腎臟移植病人中新加入或調升mycophenolate劑量有增加TAC dnC0和dnC0/DW的趨勢，但未達顯著差異；肝臟移植病人MMF對TAC影響則有不一致的趨勢，可能是因為在移植穩定期mycophenolate和TAC使用的劑量較低且樣本數少而造成這樣的分析結果。EVL對TAC藥動學影響分析中，用藥轉變的組別較多，但是因為每組人數都小於20人加上用藥劑量和濃度較低而導致EVL對TAC藥動學影響分析結果不一致。SRL對TAC藥動學影響分析也出現不一致的結果，然而在樣本達到統計檢定力為0.80所需樣本組數34組的用藥轉換組別中，SRL有降低TAC的dnC0和dnC0/DW的趨勢，但未達顯著差異。 由於納入分析病人的臨床檢驗值都在正常範圍或是雖然異常但用藥改變前後數值沒有顯著差異，所以沒有進入羅吉斯迴歸或是多元線性迴歸分析。排除生化檢驗值後，將其他因素和基因多型性納入羅吉斯迴歸分析，發現POR*28基因多型性在由TAC + mycophenolate轉變成TAC + mycophenolate + SRL的組別會顯著影響TAC dnC0，加入SRL後POR*28病人TAC dnC0會顯著下降。 此外，將其他因素和基因多型性納入線性迴歸變項分析，發現CYP3A4*18B和CYP3A4*1G基因多型性會影響在SRL加入TAC + mycophenolate的用藥組別中，SRL降低TAC dnC0和dnC0/DW的程度。CYP3A4*1G和CYP3A4*18B分別可以解釋14%和7%的TAC dnC0相對變化量；在TAC dnC0絕對變化量的模型中，CYP3A4*1G和CYP3A4*18B分別有11%和7%的解釋力，但是CYP3A4*1G的p值大於0.05。 最後，將CYP3A4*1G、CYP3A4*18B和POR*28進行次分組分析，發現SRL在CYP3A4*1 (20230 G>A)、CYP3A4*1 (878 T>C)病人和POR*28病人中SRL會顯著降低TAC dnC0和dnC0/DW。此外，雖然CYP3A5和CYP3A4*1G存在連鎖不平衡的關係，但是分組分析後未發現SRL會顯著降低CYP3A5表現型或CYP3A5非表現型病人的TAC dnC0和dnC0/DW。 結論： 本研究是第一個分析SRL對TAC藥動學交互作用影響並以7個基因型做次分組分析的研究，也是臺灣第一個探討移植6個月後MMF和mTOR inhibitors對TAC藥動學交互作用的研究。主要結果包含腎臟和肝臟移植病人加上或改變mycophenolate和mTOR inhibitors劑量後，TAC dnC0和dnC0/DW的變化。次要結果為分析基因型在mycophenolate和mTOR inhibitors對TAC藥動學交互作用中的影響。 過去研究對於mTOR inhibitors和MMF對TAC藥動學的影響沒有一致的結果，本研究礙於用藥轉換人數較少且用藥劑量和血中濃度較低，不能確定EVL對TAC dnC0和dnC0/DW的影響；不過分析發現mycophenolate有增加TAC dnC0和dnC0/DW的趨勢，而SRL則有降低TAC dnC0和dnC0/DW的趨勢，但都未達顯著差異。但透過迴歸分析及基因型次分組分析發現CYP3A4*1 (20230 G>A)、CYP3A4*1 (878 T>C)和POR*28病人，SRL會顯著降低TAC dnC0，這是過去研究未曾發現者。 關鍵字：tacrolimus、mycophenolate、sirolimus、everolimus、藥動學交互作用

Background Tacrolimus (TAC) is the cornerstone of maintenance immunosuppressive therapy after kidney and liver transplant. It has high interpatient variability and cytochrome P450 (CYP450) gene polymorphisms are important factors. Other related genetic factors include ATP-Binding Cassette, Sub-Family B1 (ABCB1) and P450 oxidoreductase (POR) polymorphisms. Many clinical factors also influenced TAC pharmacokinetics (PK). Long-term use of TAC could cause nephrotoxicity. Combination of TAC with non-nephrotoxic agents are recommended such as mycophenolate and mTOR inhibitors. Because of common metabolism and transport pathways of these drugs, there is a potential for PK drug interaction, leading to TAC under- or over-exposure. Some studies tried to characterize the PK interaction of mycophenolate or mTOR inhibitors with TAC, but conflicting data existed. Our study used self-control study design and took genetic and clinical factors into consideration to investigate the influence of SRL, EVL and MMF on TAC dose-normalized C0 (dnC0) in stable transplant patients with long-term TAC-based regimen use. Method Our study is a self-control, retrospective study, using Integrated Medical Database, National Taiwan University Hospital (NTUH-iMD). Eligible patients were from previous research project, which included 98 kidney and 89 liver transplant patients. After controlling genetic factors (CYP3A5*3, CYP3A4*1G, CYP3A4*18B, POR*28, ABCB1 1236C>T, ABCB1 2677G>T/A, ABCB1 3435C>T) and clinical factors, we analyze TAC dnC0 after dose adjustment, addition or discontinuation of MMF and mTOR inhibitors in stable patients (at least 6 months after transplant). Moreover, regression analysis is used to explore factors significantly influencing TAC dnC0. This study used Chi-squared test or Fisher's exact test to analyze categorical data and use Student's t-test or Mann–Whitney U test to analyze continuous data. For paired samples, we use Paired sample t-test or Wilcoxon signed rank test, and for regression analysis. We use logistic regression when dependent variables are binary data and linear regression when dependent variables are continuous data. Variables with the p-value＜0.2 were then incorporated into stepwise multivariate regression to explore significant factors. Results There were much more regimens and frequency of regimen change in kidney transplant patients than in liver transplant patients during the study period. In kidney transplant patients, increased MMF dose tended to increase TAC dnC0 and dnC0/DW, but not statistically significant. The PK effect of MMF on TAC in liver transplant patients and EVL on TAC in kidney transplant patients were inconsistent, probably due to low dose of EVL and MMF, and small sample size. There were also inconsistent results in the analysis of influence of SRL on TAC PK. However, in the regimen with sample size＞34 patients (to reach statistical power of 0.8), SRL tended to reduce TAC dnC0 and dnC0/DW, but not statistically significant. In logistic regression, we found TAC dnC0 in patients with POR*28 tended not increase TAC dnC0 when changing regimen from TAC + MMF to TAC + MMF + SRL (odds ratio, OR = 0.32, p＜0.05). In linear regression, we found that CYP3A4*18B and CYP3A4*1G significantly influenced SRL’s suppressed effect on TAC dnC0 and dnC0/DW. CYP3A4*1G and CYP3A4*18B could explain 14% and 7% of the relative change in TAC dnC0 respectively and explain 11% and 7% of the absolute change in TAC dnC0 respectively. However, the p-value of CYP3A4*1G is greater than 0.05 in the simple regression model of TAC dnC0 absolute change. Finally, subgroup analysis according to CYP3A4*1G, CYP3A4*18B and POR*28 found that SRL significantly reduced TAC dnC0 and dnC0 in patients with CYP3A4*1 (20230 G>A)、CYP3A4*1 (878 T>C) or POR*28. In addition, though there is a linkage disequilibrium between CYP3A5 and CYP3A4*1G, we found no significant decrease in TAC dnC0 in CYP3A5 expressers or non-expressers when SRL was added to the regimen. Conclusion This is the first study to analyze the effect of SRL on TAC PK interactions with subgroup analysis by 7 genotypes. It is also the first TAC PK interaction study in patients at least 6 months after transplant in Taiwan. The primary results included changes in TAC dnC0 and dnC0/DW after dose adjustment, addition or discontinuation of MMF and mTOR inhibitors in kidney or liver transplant patients. The secondary results were the effect of genotypes on the interaction between MMF or mTOR inhibitors on TAC PK. Previous studies have inconsistent results on the effect of mTOR inhibitors and MMF on TAC PK. Due to small sample size and low drug dose and concentration, our study was unable to determine the effect of EVL on TAC dnC0 and dnC0/DW, either. However, though not statistically significant, we found that MMF tended to increase TAC dnC0 and dnC0/DW, while SRL tended to decrease TAC dnC0 and dnC0/DW. The novel find in this study is patients with CYP3A4*1 (20230 G>A)、CYP3A4*1 (878 T>C) or POR*28, SRL significantly decrease TAC dnC0. Keywords: tacrolimus、mycophenolate、sirolimus、everolimus、PK interaction