降血脂藥引發橫紋肌溶解與基因多型性之研究

Abstract

研究背景 史達汀 (statins)，全名為3-hydroxy-3-methylglutaryl CoA (HMG CoA) reductase inhibitors，是近年來被認為有效治療心血管疾病的降血脂藥物之一類。目前上市的六種史達汀降血脂效果明顯，適用病人範圍也很廣。纖維酸衍生物(fibrates)是降血脂的第二線用藥，主要治療伴隨有代謝性症候群或第二型糖尿病的心血管疾病患者。纖維酸衍生物治療效果顯著，病人也有很好的耐受性。雖然史達汀與纖維酸衍生物都是降脂效果明顯，安全性很高的藥物，但發生機率小危險性卻極高的肌肉性不良反應，如：橫紋肌溶解等嚴重副作用，兩種藥物都有持續的案例發生。因為個體間的發病比率有很大的差異性，隨著基因體醫學及實驗技術的發展，陸續有文獻探討與脂質合成、藥物代謝相關的基因，其變異性與降血脂藥所引發嚴重副作用之間的關聯性。然而這些研究多著重於高加索人種，對於亞洲人種的相關資料仍不足。綜合以上所述，本研究將探討個體間嚴重藥物副作用的發生率與那些基因的多型性有關？這些基因型的變異與其他人種是否有不同？人種間臨床表現的差異是否能以這些基因型變異來解釋？ 研究方法 本研究利用回溯性方法收集民國86年8月至民國96年8月期間，在台大醫院曾被診斷因服用單一史達汀或單一纖維酸衍生物或合併此兩種藥物而引發橫紋肌溶解等藥物不良反應之紀錄者，符合此納入條件的有34個案例，為發病組；另外收集在台大醫院曾經服用過史達汀但沒有肌肉性藥物不良反應的病患作為控制組。收集受試者之性別、年齡、用藥資料、檢驗生化值，並採集受試者靜脈血以進行特定SNP的基因型鑑定。本實驗所研究的基因包括COQ1、COQ2、COQ3、cytochrome P450 system (CYP)︰CYP3A、CYP2D6、CYP2C9、organic anion transporting polypeptide、bile acid transporter、multidrug resistance transporter、multidrug resistance-associated protein 2 (MRP2)、UDP-glucuronosyltransferase 1A、UGT2B7等21個基因，共有46個SNP，並進一步研究受試者因服用降血脂藥物而引發橫紋肌溶解之發生率與藥物基因體學之相關性。 研究結果 在符合發病組納入條件的34個案例中，有17位受試者同意參與本試驗；另外收集75位為控制組，共有92位受試者參與本試驗。全部受試者皆被成功檢測出特定SNP的基因型型別。受試者發病時的CK與AST、CKMB與AST、CKMB與ALT具有顯著的相關性，推測肝細胞損壞應是發生橫紋肌溶解而連帶引起的病理現象。本實驗發現3個SNP (rs6925344、rs10829053、rs4149080)與橫紋肌溶解發生有相關性, 各別在特定的遺傳模式下，其勝數比分別為3.56、4.73、3.05。實驗中也發現一組單套型與八個危險組合與橫紋肌溶解發生率具有顯著的相關性並可進ㄧ步求出羅吉斯機率函數。利用羅吉斯迴歸分析納入以上九個危險因子所得到的預測機率函數，其整體正確預測分類百分比為89.1%。 結論 本實驗提出ㄧ些臨床上可能有意義的基因多型性、不同SNP組成的單套型或組合，提供臨床醫師作為決定史達汀用藥種類與劑量上的依據。並藉由機率函數的建立，計算病人發生肌肉性藥物不良反應的機率，增加史達汀的用藥安全性。

Background 3-hydroxy 3-methylglutaryl CoA (HMG CoA) reductase inhibitors﹐known as statins﹐are considered to be one of the most effective drug groups in reduction of cardiovascular risks in recent years. Six marketed statins﹐compared with other antihyperlipidemic drugs﹐are relatively effective in reduction of low-density lipoprotein. Statins are now recommended for a wide range of people at cardiovascular risk﹐including those with average and below-average lipid levels. Fibrates are the second-line medication for reduction in cardiovascular events in patients with metabolic syndrome or type 2 diabetes. The efficacy of fibrates is good and patients are generally well tolerated. Unfortunately﹐rare but severe muscle toxicity by usage of statins or fibrates﹐for example rhabdomyolysis, is occasionally reported. Risks caused by statins or fibrates vary depending on the individuals. As pharmacogenomics and experimental techniques develop﹐there are more and more studies concerning about the association between incidence of lipid-lowering agents-induced adverse reactions and genetic polymorphisms in enzymes related with cholesterol synthesis and drug metabolism. However﹐ most studies are focused on Caucasians﹐ data from Asian populations are insufficient. This study is aimed to search for genes possibly involved in severe adverse drug reactions and their genotype-phenotype relations. Methods The subjects with records of rhabdomyolysis upon statins or fibrates monotherapy or combined therapy were retrospectively collected in National Taiwan University Hospital from August 1997 to August 2007. Patients receiving statins without diagnosis of rhabdomyolysis were as controls. Data were collected, including sex, age, records of drug use and serological testing values. 10-ml EDTA-containing blood samples were drawn and DNAs were extracted for genotyping of SNPs. Genes studied in this study included COQ1, COQ2, COQ3﹐ cytochrome P450 system (CYP)︰CYP3A﹐ CYP2D6﹐ CYP2C9﹐ organic anion transporting polypeptide﹐ bile acid transporter﹐ multidrug resistance transporter﹐multidrug resistance-associated protein 2 (MRP2)﹐ UDP – glucuronosyltransferase (UGT) 1A and UGT2B7. There are 46 SNPs in 21 genes in total in this study. Analysis was focused on the association between incidence of rhabdomyolysis and genetic variations in patients receiving statins or fibrates. Results Thirty-four cases were collected with diagnosis of rhabdomyolysis from the database. Among them, 17 patients were enrolled as the disease group in this study. Seventy-five patients without diagnosis of rhabdomyolysis were as the control group. There were 92 patients in total enrolled in this study. All of them have been genotyped successfully. Levels of CK-AST, CKMB-AST and CKMB-ALT are highly associated with patients suffering from rhabdomyolysis, indicating hepatocyte damage coexists with rhabdomyolysis. Three SNPs (rs6925344, rs10829053, rs4149080) are correlated with the incidence of rhabdomyolysis. The odds ratios are 3.56, 4.73 and 3.05, respectively, in specific genetic models. One haplotype and eight risk combinations have association with incidence of rhabdomyolysis in logistic regression models. Nine risk predictors are subjected to logistic regression analysis to receive an accuracy rate of 89.1%. Conclusions Results could provide some genetic markers in use of statins or fibrates to prevent from rhabdomyolysis. By constructing a regression model, we can predict the incidence of statin- or fibrate-induced myopathy and thereby choose a right lipid-lowering agent to the right patient to improve drug safety.