摘要
大約10∼30％的癲癇患者無法藉抗癲癇藥品得到良好的控制，屬於難治型癲癇。為了瞭解難治型癲癇症之機轉，本研究從藥物基因學及基礎研究兩方面進行探討。 在藥物基因學方面，本研究共收納了331位癲癇病人，其中108位屬於難治型癲癇，223位屬於藥物可治療型癲癇與287位正常人 (Normal controls)。以PCR-RFLP與real-time PCR genotyping 分析ABCB1、PXR基因多型性位點與癲癇症治療效果相關性分析，結果顯示難治型癲癇病人較傾向於在此三位點帶有 CGC/CGC、CGC/TGC、CGC/TTT 及 TGC/TTT 之雙套子基因型；而藥物可治療型癲癇病人及正常人則是較傾向於在此三位點帶有 TTC/TTC、TTC/CGT、TTC/TGT、CGT/CGT 及 TGT/CGT 之雙套子基因型。進一步探討調控ABCB1表現量之NR1I2基因，其基因多型性位點是否與ABCB1基因多型性存在有加成性作用，以邏輯式回歸模型分析之結果顯示，主要影響癲癇症療效之基因仍為ABCB1基因，在校正了癲癇症之分類與病因之模型中，ABCB1 2677G>T、3435C>T及2677G>T與3435C>T之交互作用項與難治型癲癇症顯著相關，但NR1I2 11156A>C、11193T>C並不參與其中。另外，在藥效學方面，本研究亦證實SCN1A IVS5-91 G>A基因型顯著影響phenytoin之維持劑量與血中濃度，顯示在癲癇症之治療上，基因扮演著重要的角色。 在基礎研究方面，本研究更進一步探討抗癲癇藥與Pgp之交互作用及其機轉，發現phenyotin、carbamazepine、phenobarbital、lamotrigine都有抑制Pgp排除受質之作用，其中phenytoin為potency最強者，且其有效抑制濃度與臨床治療濃度較接近，所以進一步探討其作用機轉，進而發現phenytoin透過競爭性抑制作用於rhodamine 123結合位置，且1236T-2677T/A-3435T基因型之Pgp與抑制劑結合效果較差，導致較弱之抑制效果；此一研究結果與臨床相關性研究之結果可互相呼應，帶有1236T-2677T-3435T基因型之Pgp其與抑制劑之結合較差，導致較弱的抑制效果，細胞中之Pgp受質依然持續被排除至細胞外，所以細胞中藥物濃度較低，可能因此而導致難治型癲癇症。 本研究針對難治型癲癇症之藥物基因學做了全面性之研究，從基因體到基礎蛋白質功能，從藥動學到藥效學之探討，有助於醫師從藥物基因層面做進一步的了解，對病人的癲癇控制與治療效果評估極為重要。
並列摘要
About 10% to 30% of patients with epilepsy are refractory to antiepileptic drugs (AEDs). To investigate the underlying mechanisms of intractable epilepsy, the present study explore this issue from genetics to mechanisms. In the aspect of pharmacogenetics, the present study included 108 drug-resistant epileptic patients, 223 seizure-free patients and 287 normal controls. The results of ABCB1 and NR1I2 genotyping revealed that drug-resistant patients tended to have the CGC/CGC, CGC/TGC, CGC/TTT and TGC/TTT haplotype combinations over the seizure-free patients and controls (all p-values<0.0001). In contrast, patients with the TTC/TTC, TTC/CGT, TTC/TGT, CGT/CGT and TGT/CGT haplotype combinations were more likely to be seizure-free (all p-values<0.0001 except CGT/CGT (p=0.0063)). The polymorphisms in NR1I2 gene were not significantly associated with epilepsy treatment responses in the present study. The SCN1A IVS5-91 G>A polymorphism was associated with phenytoin serum concentration at maintenance dose. To investigate whether ABCB1 haplotype combinations would affect P-glycoprotein (Pgp) function and impact the clinical responses of antiepileptic drugs, the present study also established cell lines harbor all combinations of the three polymorphisms. Transport of substrate rhodamine123 and calcein-AM by human Pgp carrying twelve haplotype combinations of 1236C>T, 2677G>T/A, and 3435C>T were assayed in the absence and presence of known inhibitors and antiepileptic drugs. The inhibitory potency of the tested drugs from the dose-response relations was cyclosporin A > verapamil > phenytoin > carbamazepine > lamotrigine > phenobarbital > valproic acid, levetiracetam, gabapentin. The silent polymorphisms combination (1236T-3435T) and triple haplotypes (1236T-2677A/T-3435T) resulted in profoundly less effective inhibition against substrates with significantly lower intracellular substrate concentration. These results confirmed that ABCB1 polymorphisms were associated with clinical responses of antiepileptic drugs. Our findings demonstrated that human ABCB1 polymorphisms may alter the interactions between Pgp and substrates and provided functional evidence of ABCB1 haplotypes-associated epilepsy treatment responses.