- 學位論文
於接受Cisplatin或Carboplatin治療之肺癌病人運用基因及非基因因子進行腎毒性之多元分析
Multiple Analytical Approaches Demonstrate a Complex Relationship of Genetic and Nongenetic Factors with Cisplatin- or Carboplatin-induced Nephrotoxicity in Lung Cancer Patients
摘要
Cisplatin及carboplatin是目前使用最廣泛的化學治療藥物，為許多固態腫瘤之第一線用藥。然而，腎毒性限制了這類藥物的使用，約有三分之一病人即使給予適當的輸液還是會發生。Cisplatin或carboplatin所引起腎毒性之機轉推測可能為多條路經交互作用之結果，這些路徑包括增加攝入細胞、腎毒性物質之形成、ROS過度累積、DNA修復能力受損及細胞凋亡。本研究一共納入116位接受cisplatin或carboplatin治療之肺癌病人，擬探討台灣族群的7 個基因共11個基因多型性(GSTPi A313G, TP53 G215C, UGT1A7 T387G, UGT1A7 T622C, NAT2 T481C, NAT2 G590A, NAT2 G857A, ERCC1 C118T, ERCC1 C8092A, ERCC4 T2505C and NQO1 C609T)與使用cisplatin及carboplatin引起腎毒性之關聯性，並建立腎毒性預測模型，尋找腎毒性之易感族群。本研究採Risk, Injury, Failure, Loss of function, End-stage renal disease(RIFLE)準則評估腎毒性之發生有無。並以classification and regression tree(CART)及Framingham study risk score建立腎毒性預測模型，探討基因與非基因因子，及其之間交互作用對腎毒性之重要性。結果發現在男性研究對象中，帶有TP53對偶基因者出現腎毒性風險較野生型顯著下降(OR =0.12, 95%CI 0.02-0.81)。此外，主要使用cisplatin治療之病人，帶有ERCC1 118T基因型亦有同樣趨勢(OR =0.14, 95%CI 0.03-0.72)。CART所建立之預測模型發現，主要使用cisplatin且Scr≦1mg/dL之病人，帶有TP53野生型或TP53 G對偶基因－ERCC1 C118T 野生型者為腎毒性之易感族群。此模型可預測63.0%腎毒性組、89.6%無腎毒性組。另一預測模型，腎毒性危險因子評估量表，顯示同時含有基因及非基因因子時，有最好預測性，評量表之最佳切點為12分，可預測64.3%腎毒性組、90.9%無腎毒性組。本研究顯示男性且帶有TP53 G對偶基因或使用cisplatin者帶有ERCC1 C118T C對偶基因可能為腎毒性之危險因子。累積治療次數及cisplatin累積給藥強度亦可能為腎毒性之危險因子。然而，基於臨床實用性考量，以CART預測模型較腎毒性危險因子評量表佳，以納入較少的因子達相近的預測力。
並列摘要
Introduction: The purpose of this study was to investigate the association between the cisplatin- or carboplatin-induced nephrotoxicity and the multiple single nucleotide polymorphisms (SNPs). Multiple statistical methods including genetic and nongenetic factors were applied to establish clinical useful model for predicting the cisplatin- or carboplatin-induced nephrotoxicity. Patients and Methods: The retrospective study investigated 11 polymorphisms in 7 genes (GSTPi A313G, TP53 G215C, UGT1A7 T387G/T622C, NAT2 T481C/G590A/G857A, ERCC1 C118T/C8092A, ERCC4 T2505C and NQO1 C609T) in 116 Taiwanese patients who had received cisplatin or carboplatin more than twice for lung cancer at Wanfang Hospital. The Risk, Injury, Failure, Loss of function, End-stage renal disease (RIFLE) criteria was use to evaluate the occurrence of nephrotoxicity. Multiple regression with classification and regression tree (CART) and Framingham study risk score were used to study the interactions between the genetic and nongenetic factors with cisplatin- or carboplatin-induced nephrotoxicity. Results: In logistic regression, the male patients carrying TP53 C allele had lower risk of nephrotoxicity (OR =0.12, 95%CI = 0.02-0.81). Subgroup analysis on cisplatin, patients revealed that ERCC1 T allele at 118 also had protective effects in nephrotoxicity (OR =0.14, 95%CI = 0.03-0.72). In CART analysis, individuals who were using cisplatin and baseline serum creatinie ≦1mg/dL with TP53 Arg/Arg or TP53 Pro allele－ERCC1 C/C were higher nephrotoxicity risk. The overall prediction rate of CART was 82.7%. The sensitivity and specificity are 0.630 and 0.896, respectively. Another prediction model, according to Framingham study risk score, was contained 7 factors. Its overall prediction rate was 84.5%. The sensitivity and specificity were 0.643 and 0.909, respectively. Conclusion: We found that genes, TP53 Arg72Pro and ERCC1 C118T, might be the risk factors of nephrotoxicity. Because of containing fewer factors and equal prediction rate, the results suggest that CART model was better than nephrotoxicity risk score. In addition to simultaneously containing the genetic and nongenetic factors, a favor model to predict the cisplatin- or carboplatin-induced nephrotoxicity also had to take interaction of these factors in consideration.