在个体化用药的大背景下， 随着高特异性、高通量及高灵敏度的基因检测新技术的快速发展， 具体的分析技术问题已经不再是难题。然而， 证据级别高的或公认的药物基因组学检测项目在临床应用还很少， 其瓶颈在于检测项目尚未纳入医保范围， 使得患者不愿意支付额外的医疗费用而接受测试。其次， 药物基因检测的一些试剂还没有获得国家食品药品监督管理局注册证， 这两大原因导致不少医院的药物精准治疗仅仅停留在科研水平， 没有真正投入临床应用。建议有关部门出台相应的法规， 制定标准的实验室检测体系， 规范个体化治疗检测市场， 理顺价格支付机制， 切实为药物精准治疗的推广而保驾护航。
相对来说， 为避免严重不良反应的药物基因组学测试项目更受临床医生欢迎， 从表1和表2也可见，基于安全性的药物基因组学检测项目比基于疗效预测的项目要多， 因此， 围绕治疗安全性的生物标志物挖掘与应用， 是药物精准治疗在临床推广的突破口。
合理用药强调安全、有效、经济、适当。因此， 药物经济学也是开展药物精准治疗所必须考虑的实际因素。系统评价显示， 氯吡格雷用药前的 CYP2C19基因检测、阿巴卡韦用药前的人类白细胞抗原HLA-B*5701 检测、卡马西平用药前的 HLA-B*1502和HLA-A*3101检测、以及别嘌醇用药前HLA-B*5801检测被证实具有成本效益， 而对于硫嘌呤药物用药前 TPMT 检测、香豆素类药物用药前 CYP2C9 和VKORC1 检测、甲氨蝶呤用药前 MTHFR 检测是否具有经济学意义， 结论尚未统一[33].Donnan 等[34]也证实， 与基于体重的给药方案相比， 急性淋巴细胞性白血病儿童患者中基于TPMT基因型检测或基于TPMT酶活性检测的给药方案并不具有成本效益。更有研究表明， 香豆素类药物用药前药物基因组学检测的临床执行率依赖于检测的成本以及新型口服抗凝药的可及性。与华法林相比， 一些新颖的口服抗凝药无需在用药前进行药物基因组学检测[35].也有人提出对需要使用口服抗凝药的房颤患者进行用药前药物基因组学检测， 根据基因检测结果对患者进行分诊。假如患者具有对华法林敏感性正常的基因型 （CYP2C9*1/*1 和 VKORC1 G/A）， 则选择使用华法林； 假如患者具有对华法林敏感性高 （CYP2C9*2、CYP2C9*3或 VKORC1 A/A） 或敏感性低 （VKORC1 G/G 且CYP2C9*1/*1） 的基因型 , 则选择新颖口服抗凝药（如达比加群、利伐沙班、阿哌沙班）。与常规接受华法林抗凝的治疗模式相比， 基于药物基因组学的分诊模式具有更好的成本？效果比[36].随着药物精准治疗理念的推行， 未来将会有更多药物经济学研究来为政府和临床决策提供证据。目前而言， 担心成本效益问题仍是影响开展基于药物代谢酶和转运体基因组学的药物精准治疗的一个制约因素。美国一家基因检测技术公司评价了药物基因组学检测用于接受 5种或 5 种以上药物的长期照护患者精准治疗的经济获益， 发现平均每个患者每年节约成本 621 美元[37].
该研究提供了一个重要启示， 即基于药物基因组学的精准治疗可在特殊群体中优先应用。
另外， 一些生物标志物在临床证据总体级别不高， 还需要通过开展进一步的基础研究和大样本量的数据分析来论证。荟萃分析通过合并资料而增大样本量， 可增加结论的可信度， 解决研究结果的不一致性。因此， 可尝试开展药物基因组学检测临床价值的荟萃分析研究。
4 结语。
精准医学开启了个体化治疗新时代。基于药物代谢酶和转运体的药物精准治疗， 具有重要的临床意义和应用前景[38].应重点关注多基因和非遗传因素对药物效应的影响， 重视 TDM 与药物基因组学检测的整合， 同时需解决临床应用中面临的瓶颈问题。临床上的药物精准治疗涉及药物表观基因组学、药物基因组学、药物代谢学、生物统计学、临床药理学、药物治疗学和药物经济学等， 因此， 需要多学科交叉， 基础与临床协同研究， 加强研发和挖掘证据级别高的新的生物标志物， 积极推动临床转化应用， 为临床药物精准治疗提供依据。
References:
[1] Zhou Q, Zeng S. Drug metabolism, transport and clinicalimplications [M] // Jiang XH. Clinical Pharmacokinetics （临床药动学）。 Beijing: Higher Education Press, 2007, 62-105.
[2] Clinical Pharmacogenetics Implementation Consortium.Dosing Guidelines-CPIC [S/OL]. [2016-05-06].
[3] FDA. Table of Pharmacogenomic Biomarkers in Drug Labeling[EB/OL]. 2015 [2016-05-06].
[4] Huang FH, Xiao HL. FDA's requirements for includingpharmacogenomic information in durg labeling [J]. DrugEval Res （药物评价研究）， 2014, 37: 201-206.
[5] Wang B, Canestaro WJ, Choudhry NK. Clinical evidencesupporting pharmacogenomic biomarker testing provided inUS Food and Drug Administration drug labels [J]. JAMAIntern Med, 2014, 174: 1938-1944.
[6] Rojas L, Neumann I, Herrero MJ, et al. Effect of CYP3A5*3on kidney transplant recipients treated with tacrolimus: asystematic review and meta-analysis of observational studies[J]. Pharmacogenomics J, 2015, 15: 38-48.
[7] Zhu HJ, Yuan SH, Fang Y, et al. The effect of CYP3A5polymorphism on dose-adjusted cyclosporine concentration inrenal transplant recipients: a meta-analysis [J]. Pharmaco-genomics J, 2011, 11: 237-246. .
[8] Jung JA, Lim HS. Association between CYP2D6 genotypesand the clinical outcomes of adjuvant tamoxifen for breastcancer: a meta-analysis [J]. Pharmacogenomics, 2014, 15:49-60.
[9] De Iuliis F, Salerno G, Taglieri L, et al. Are pharmacogenomicbiomarkers an effective tool to predict taxane toxicity andoutcome in breast cancer patients- Literature review [J].Cancer Chemother Pharmacol, 2015, 76: 679-690.
[10] Yin JY, Huang Q, Zhao YC, et al. Meta-analysis on pharma-cogenetics of platinum-based chemotherapy in non small celllung cancer （NSCLC） patients [J]. PLoS One, 2012, 7: e38150.
[11] He HR, Chen SY, You HS, et al. Association betweenmethylenetetrahydrofolate reductase polymorphisms and therelapse of acute lymphoblastic leukemia: a meta-analysis [J].Pharmacogenomics J, 2014, 14: 432-438.
[12] Meulendijks D, Henricks LM, Sonke GS, et al. Clinical rele-vance of DPYD variants c.1679T>G, c.1236G>A/HapB3, andc.1601G>A as predictors of severe fluoropyrimidine-associatedtoxicity: a systematic review and meta-analysis of individualpatient data [J]. Lancet Oncol, 2015, 16: 1639-1650.
[13] Luo JQ, He FZ, Wang ZM, et al. SLCO1B1 variants andangiotensin converting enzyme inhibitor （enalapril）-inducedcough: a pharmacogenetic study [J]. Sci Rep, 2015, 5: 17253.
[14] de Keyser CE, Peters BJ, Becker ML, et al. The SLCO1B1c.521T>C polymorphism is associated with dose decrease orswitching during statin therapy in the Rotterdam Study [J].Pharmacogenet Genomics, 2014, 24: 43-51.
[15] Zheng Q, Wu H, Yu Q, et al. ABCB1 polymorphisms predictimatinib response in chronic myeloid leukemia patients: asystematic review and meta-analysis [J]. PharmacogenomicsJ, 2015, 15: 127-134.
[16] Wen CC, Yee SW, Liang X, et al. Genome-wide associationstudy identifies ABCG2 （BCRP） as an allopurinol transporterand a determinant of drug response [J]. Clin Pharmacol Ther,2015, 97: 518-525.
[17] Zhou Q, Ruan ZR, Yuan H, et al. CYP2C9*3 （1075A>C），MDR1 G2677T/A and MDR1 C3435T are determinants ofinter-subject variability in fluvastatin pharmacokinetics inhealthy Chinese volunteers [J]. Arzneimittelforschung, 2012,62: 519-524.
[18] Zhou Q, Ruan ZR, Yuan H, et al. ABCB1 gene polymorphisms,ABCB1 haplotypes and ABCG2 c.421c>A are determinants ofinter-subject variability in rosuvastatin pharmacokinetics [J].Pharmazie, 2013, 68: 129-134.
[19] Altar CA, Carhart JM, Allen JD, et al. Clinical validity:combinatorial pharmacogenomics predicts antidepressantresponses and healthcare utilizations better than single genephenotypes [J]. Pharmacogenomics J, 2015, 15: 443-451.
[20] Kim HS, Chang K, Koh YS, et al. CYP2C19 poor metabolizeris associated with clinical outcome of clopidogrel therapy inacute myocardial infarction but not stable angina [J]. CircCardiovasc Genet, 2013, 6: 514-521.
[21] Zhou Q, Chen QX, Ruan ZR, et al. CYP2C9*3 （1075A>C），ABCB1 and SLCO1B1 genetic polymorphisms and genderare determinants of inter-subject variability in pitavastatinpharmacokinetics [J]. Pharmazie, 2013, 68: 187-194.
[22] van Schie RM, Wessels JA, le Cessie S, et al. Loadingand maintenance dose algorithms for phenprocoumon andacenocoumarol using patient characteristics and pharmacogeneticdata [J]. Eur Heart J, 2011, 32: 1909-1917.
[23] Tan SL, Li Z, Song GB, et al. Development and comparisonof a new personalized warfarin stable dose prediction algorithmin Chinese patients undergoing heart valve replacement [J].Pharmazie, 2012, 67: 930-937.
[24] Shah RR, Smith RL. Addressing phenoconversion: the Achilles'heel of personalized medicine [J]. Br J Clin Pharmacol, 2015,79: 222-240.
[25] Zhou Q, Yao TW, Zeng S. The advancement in phenotypingstudy of drug metabolism [J]. Chin J Mod Appl Pharm （中国现代应用药学）， 2000, 17: 423-429.
[26] Preskorn SH, Kane CP, Lobello K, et al. Cytochrome P4502D6 phenoconversion is common in patients being treated fordepression: implications for personalized medicine [J]. J ClinPsychiatry, 2013, 74: 614-621.
[27] Klieber M, Oberacher H, Hofstaetter S, et al. CYP2C19phenoconversion by routinely prescribed proton pumpinhibitors omeprazole and esomeprazole: clinical implicationsfor personalized medicine [J]. J Pharmacol Exp Ther, 2015,354: 426-430.
[28] Li W, Zeng S, Yu LS, et al Pharmacokinetic drug interactionprofile of omeprazole with adverse consequences and clinicalrisk management [J]. Ther Clin Risk Manag, 2013, 9: 259-271.
[29] Owusu Obeng A, Egelund EF, Alsultan A, et al. CYP2C19polymorphisms and therapeutic drug monitoring of voriconazole:are we ready for clinical implementation of pharmacogenomics-[J]. Pharmacotherapy, 2014, 34: 703-718.
[30] Chouchana L, Narjoz C, Roche D, et al. Interindividualvariability in TPMT enzyme activity: 10 years of experiencewith thiopurine pharmacogenetics and therapeutic drugmonitoring [J]. Pharmacogenomics, 2014, 15: 745-757.
[31] Jannetto PJ, Bratanow NC. Pain management in the 21stcentury: utilization of pharmacogenomics and therapeutic drugmonitoring [J]. Expert Opin Drug Metab Toxicol, 2011, 7:745-752.
[32] Manvizhi S, Mathew BS, Fleming DH, et al. Combinedapproach with therapeutic drug monitoring and pharmacoge-nomics in renal transplant recipients [J]. Indian J Nephrol,2013, 23: 71-73.
[33] Plumpton CO, Roberts D, Pirmohamed M, et al. A Systematicreview of economic evaluations of pharmacogenetic testing forprevention of adverse drug reactions [J]. Pharmacoeconomics,2016, 34: 771-793.
[34] Donnan JR, Ungar WJ, Mathews M, et al. A cost effectivenessanalysis of thiopurine methyltransferase testing for guiding6-mercaptopurine dosing in children with acute lymphoblasticleukemia [J]. Pediatr Blood Cancer, 2011, 57: 231-239.
[35] Baker WL, Chamberlin KW. New oral anticoagulants vs.warfarin treatment: no need for pharmacogenomics- [J]. ClinPharmacol Ther, 2014, 96: 17-19.
[36] You JH. Pharmacogenetic-guided selection of warfarin versusnovel oral anticoagulants for stroke prevention in patientswith atrial fibrillation: a cost-effectiveness analysis [J].Pharmacogenet Genomics, 2014, 24: 6-14.
[37] Saldivar JS, Taylor D, Sugarman EA, et al. Initial assessmentof the benefits of implementing pharmacogenetics into themedical management of patients in a long-term care facility [J].Pharmgenomics Pers Med, 2016, 9: 1-6.
[38] Liu YQ, Zheng XL, Yu QQ, et al. Epigenetic activation ofthe drug transporter OCT2 sensitizes renal cell carcinoma tooxaliplatin [J]. Sci Transl Med, 2016, 348: 348ra97.
