基因多型性及臨床因素對使用tacrolimus肝臟移植病人臨床結果之影響

Abstract

研究背景與目的： Tacrolimus（TAC）為肝臟移植病人免疫抑制的基石。過去的研究顯示基因多型性和許多臨床因素皆會影響TAC藥動學，導致個體間血中濃度和臨床反應變異性大。其中CYP3A5*3為影響TAC藥動學最主要的基因型。其他在亞洲人較廣泛研究的基因型還包含CYP3A4、ABCB1、POR等。截至目前，基因型及臨床因素和肝臟移植的臨床結果相關性研究並不一致，且較少研究同時考量多種基因型及病人其他臨床因素，如併用藥品以及檢驗數據等影響。本團隊過去進行之研究計畫「ABCB1、CYP3A4、CYP3A5、POR基因多型性及其他因素對腎臟與肝臟移植病人對tacrolimus血中濃度之影響(案號:201512005RINC)」，及「肝臟捐贈者ABCB1、CYP3A4、CYP3A5、POR基因多型性對肝臟移植病人tacrolimus血中濃度之影響(案號:201612023RIND)」，已完成捐贈者及受贈者基因多型性及其他臨床因素對TAC血中濃度影響的收案。本研究持續追蹤並分析已納入上述之研究計畫並已有基因型資料之肝臟移植病人，分析影響臨床結果的因素，主要結果包含病人存活率、移植體存活率及切片證實之急性排斥反應；次要結果包含TAC相關不良反應如移植後新發糖尿病、移植後新發惡性腫瘤、移植後新發血脂異常、感染及腎毒性。 方法： 本研究為回溯性研究，將分析本團隊過去研究計畫中已簽署知情同意書，並已有基因資料之肝臟移植病人。該計畫之納入條件為2008年1月1日至2015年12月31日進行活體肝臟移植的病人，移植時的年齡介於20 ~ 65歲，在移植後持續使用TAC做為免疫抑制劑至少12個月；排除條件包含再移植或多重器官移植、非臺灣人以及人類免疫缺乏病毒（human immunodeficiency virus，HIV）反應呈陽性的病人。本研究蒐集上述計畫於研究期間持續追蹤病人常規醫療檢查之臨床數值、併用藥品、療劑監測等，至2017年12月31日。利用Kaplan Meier analysis分析病人存活率、移植器官存活率和移植後一年內切片證實之急性排斥反應發生率，利用迴歸分析了解多基因型和病人臨床因素對肝臟移植臨床結果像是移植後糖尿病、移植後血脂異常、移植後新發惡性腫瘤、感染以及腎功能的影響。類別性資料將使用Chi-square test以及Fisher’s exact test，連續性資料將檢定是否為常態分布，若為常態分布則使用Student`s t-test，非常態分布則使用Mann-Whitney U test。迴歸分析使用logistic regression、cox regression以及linear regression，單變相分析時p＜0.2的自變數將納入多變項分析， p＜0.05將視為達到統計學上的顯著差異。 結果： 本研究共納入89位肝臟移植病人及其捐贈者，在追蹤期間結束時沒有死亡事件或再移植事件的發生。BPAR發生人數共19人（21.3 %），平均發生時間約為63.74 ± 66.68天，中位數為32天（7, 203）。所有肝臟移植病人皆使用誘導藥品，包含basiliximab以及daclizumab。使用basiliximab誘導的病人，發生風險較使用daclizumab的病人低（p = 0.032），在多變項分析中誘導藥品是唯一會影響BPAR的因素。 共8人（12.5 %）發生PTDM，平均發生時間為2.95 ± 1.60年，中位數為3.15年（0.84, 4.76）。移植前為C型肝炎帶原的病人PTDM發生的風險顯著較高（p = 0.025），而第3年平均TAC C0較高也增加PTDM發生的機率，這兩個變項的整體解釋能力可達17 %。移植後發生感染事件的共有32 人（36.0 %），平均發生時間為0.75 ± 0.99年，第1年有併用與TAC有交互作用藥品（如calcium channel blockers、proton pump inhibitors）有顯著較高的感染風險（p＝0.01）。在是否有併用TAC相關DDI之藥品分組分析顯示，有併用DDI之藥品的病人相較於沒有併用的病人，移植前HCV帶原的比例較高（p＝0.03），高血壓（p＜0.001）及血脂異常（p＝0.023）的比例也較高。 共9人（10.1 %）發生de novo dyslipidemia，平均發生時間為2.72 ± 2.39年，中位數為1.91（0.50, 8.17）年。捐贈者及受贈者皆為CYP3A5*3變異的病人相較於捐贈者或受贈者帶有CYP3A5*1的病人風險較高（p＝0.015），以及移植後第1年有使用everolimus的病人風險顯著較高（p＜0.001），兩個變項整體解釋能力可達18 %。發生de novo malignancy共3人，癌症種類分別為子宮頸癌、攝護腺癌以及皮膚鱗狀細胞癌（Bowen's disease）。使用basiliximab誘導的病人發生風險顯著較使用daclizumab的病人低（p＝0.026），由於使用daclizumab的病人僅5人，因此有待更多的研究證實。移植後腎功能與多個臨床因素有關：包含捐贈者及受贈者年齡、移植前糖尿病有無、BMI等。基因型的影響上，第3年平均eGFR變化量會顯著下降，捐贈者及受贈者ABCB1（C3435T）變異則分別和第5年eGFR上升及第7年eGFR下降有顯著相關，詳細機轉有待更多研究。 討論： 使用basiliximab的病人發生BPAR風險顯著較使用daclizumab的病人低，然而因daclizumab於2009年時下市，且本研究使用daclizumab的病人僅5位，皆為2008年至2009年間執行肝臟移植手術的病人，因此除了樣本數的限制外，也不排除年代所造成的影響。在PTDM的結果顯示，移植前為HCV帶原的病人會增加PTDM的發生風險，目前有關HCV造成PTDM風險增加的機轉還不清楚，但HCV的治療對於PTDM是否有保護效果也是未來可進一步研究的方向。TAC相關交互作用藥品的使用和感染風險增加有關，在DDI有無的分組分析中發現病人共病症的多寡可能是影響移植後感染風險的主要因素。捐贈者及受贈者皆為CYP3A5*3 non expresser的病人，de novo dyslipidemia發生的風險較高，過去並沒有基因型及de novo dyslipidemia在肝臟移植的相關結果，因此本研究為第一篇顯示CYP3A5*3基因型和使用TAC的肝臟移植病人血脂異常有關的研究。受贈者POR*28和移植後第3年的腎功能變化量有顯著相關、捐贈者及受贈者ABCB1（C3435T）則分別和移植後第5年及第7年的腎功能變化量有關，但本研究未能確立是否因長期較高的TAC血中濃度造成。

Background： Genetic polymorphism and numerous clinical factors could influence tacrolimus (TAC) pharmacokinetics (PK), which led to large inter-and intra-individual variability and even clinical outcomes of liver transplantation. TAC is mainly metabolized by CYP3A5. Other genetic factors that are widely investigated in Asians include CYP3A4, ABCB1 and POR. Previous studies on the correlation between genetic factors and clinical outcome were controversial. Furthermore, very limited studies took multiple genetic factors and clinical factors into account. Our previous study “The influence of ABCB1, CYP3A4, CYP3A5, POR genetic polymorphism and other factors on tacrolimus blood concentration among renal and liver transplant patient” and “The influence of ABCB1, CYP3A4, CYP3A5, POR genetic polymorphism of donor on tacrolimus blood concentration in liver transplant patients” have completed patient recruitment. The present study recruited liver transplant patients to see their clinical outcome. Primary outcome included patient survival, graft survival, biopsy-proven acute rejection. Secondary outcomes included TAC-related side effect such as post-transplant diabetes mellitus, de novo dyslipidemia, de novo malignancy, infection and renal function. Methods： This is a retrospective study. Eligible patients were those who had signed informed consent for genetic study in previous research projects. Inclusion criteria of the previous study were patients who underwent living-donor liver transplantation at the age of 20-65, during January 2008 to December 2015, with TAC-based immunosuppression after liver transplant for at least 12 months. Exclusion criteria were retransplantation, multi-organ transplantation, non-Taiwanese and human immunodeficiency virus positive. The present study collected laboratory data, concurrent medications, and TDM data until December 2017. Patient survival, graft survival and acute rejection will be assessed by Kaplan Meier analysis. Multiple regression will be used to evaluate the influence of genetic and clinical factors on clinical outcome such as post transplantation diabetes mellitus（PTDM）、de novo dyslipidemia、de novo malignancy、infection and renal function. Significance of categorical data will be determined by Chi-square test and Fisher’s exact test. Continuous variables will use student t-test. Result： 89 liver transplant recipients and their coupled donors were enrolled in the study. All patients are alive through the end of the study and no retransplantation incidence. 19 (21.3 %）patients had BPAR in our study. Multivariate analysis showed that induction therapy was the only influential variable. Patients using basiliximab as induction therapy has significantly lower BPAR risk compared to daclizumab use. 8 (12.5 %) patients had PTDM in our study. Patients with HCV at the baseline and higher TAC C0 at the 3rd time point were associated with higher PTDM risk, with a total R2 of 17 %. 32 (36.0 %) patients had infectious disease in our study.TAC-related DDI use at the 1st time point was shown to be associated with higher risk of infection. In the basiliximab-use subgroup analysis, DDI use at the 1st time point and patients with hypertension at the baseline had higher incidence of infection, which might be due to pre-transplant comorbidities. 9 (10.1 %) patients had de novo dyslipidemia. Donor and recipient matched CYP3A5*3 mutant type and everolimus use at the 1st time point were associated with higher risk of de novo dyslipidemia, with a total R2 of 18 %. Four cases of de novo malignancy were noted. Among the four cases, one was cervical cancer, one prostate cancer, one metastatic lung cancer and one Bowen's disease. Multiple clinical factors significantly influenced renal function after liver transplantation including age of donor and recipient, pre-transplant diabetes mellitus, BMI etc. Recipient POR*28 variate was associated with more reduction in eGFR at the 3rd time point. Donor ABCB1（C3435T）was associated with eGFR increase at the 5th time point. Whereas recipient ABCB1（C3435T）was associated with eGFR decrease at the 7th time point. The mechanism need to be further studied in the future. Discussion： Patients using basiliximab has lower BPAR risk compared to daclizumab group. However, with limited patient number, this finding need to be further confirmed. Patients with HCV at the baseline were associated with more incidence of PTDM. Whether treatment of HCV has protective impact on PTDM is still unknown but of great interest in the future. TAC-related DDI use significantly influenced the incidence of infection. Among subgroup analysis, pre-transplant comorbidities might be the main cause of infection. Patients with everolimus use at the 1st time point and CYP3A5*3 matched non expresser had higher risk of de novo dyslipidemia. This is the first study that showed possible association between CYP3A5*3 and de novo dyslipidemia in liver transplant patients. POR*28 variate was associated with more reduction in eGFR at the 3rd time point. Donor and recipient ABCB1（C3435T）was associated with change in eGFR at the 5th and 7th time point separately. However, whether it was resulted from higher TAC C0 need to be studied in the future.