降血糖多组分动态加减临床设计与组方优选方法的研究

Clinical dynamic design based on symptoms change and its quantitative analysis for the optimal combination of multiple components in treatment of diabetics

目的建立多组分动态配伍、随症加减设计方法和定量分析方法并进行35味降血糖中药组方优选。方法设计采用"药味设限,辩证选药,随症加减,剂量可调",以及"随机、双盲、多组分动态配伍、安慰剂对照、多中心试验"。将120例患者分为中药配伍组(102例)和安慰剂组(18例)。筛选药味采用多元线性回归模型,建模和分析采用正交模拟和非线性混合效应模型法。分析指标为治疗4周的餐后2h血糖下降值(2hPG)。结果在35味中药中,天花粉(A)、知母(B)、葛根(C)、山药(D)、玄参(E)共5个组分的组方频率较高。D为主要药效组分(P<0.05);其次为C和B;E和A作用较小。ABCE、CD、AC与ACE有轻度协同作用;ABCDE组合有较强的拮抗作用;其他组合(ABCD、BCDE与BDE)有轻度拮抗作用。BCDE合用时,可达到2hPG最大效应(Emax=4.02mmol.L-1)。结论本研究方法对多组方筛选具有可行性,可获得大量组方定量设计信息。

Objective To set up an approach to design a dynamic combinations of multiple components based on symptoms change,and to make a quantitative analysis for the optimal combination in treatment of diabetics.Methods A randomized,double-blind,multi-component dynamic combination,placebo-controlled,multi-center trial was carried out using different combinations of the thirty-five components on a dynamic manner.One hundred and twenty patients were assigned into traditional Chinese medicine group（n=102）and placebo group（n=18）.The multiple linear regression was applied to screen the components,and the orthogonal simulation method and nonlinear mixed effect model（NONMEM）were used for modeling and analyzing.The decrease in 2-hour postprandial glucose（2hPG）at 4 week was the primary outcome.Results The five components,including Radix Trichosanthis（A）,Rhizoma Anemarrhenae（B）,Radix Puerariae（C）,Rhizoma Dioscoreae（D）and Radix Scrophulariae（E）,had a relevant higher combination frequency in the 35 components.The component D was the main component in these combinations（P0.05）,E and A had a minor effect,and the contribution to the combination of those components was ordered as D,C and B.There was a mild synergistic action in the combinations of ABCE,CD,AC and ACE.The strong antagonistic action was showed at the combination of ABCDE,and the best combination of BCDE could theoretically reduce 4.02 mmol·L-1 of 2hPG.Conclusion The approach has a feasibility of screening optimal combination from multiple components,and can provide more information about the quantitative design of combination drug therapy.