论文部分内容阅读
目的建立有限取样法估算霉酚酸(MPA)在自身免疫性疾病(AID)患者的体内暴露药量(AUC)。方法24例AID患者服用霉酚酸醋(MMF)0.75 g,q12 h,达稳态后,检测服药前即刻、服药后0.5,1,1.5,2,4,6,8和12h活性代谢物MPA的血药浓度梯形法计算各例实测AUC_(0-12 h);将24例患者数据资料随机平分为模型资料组与验证资料组,对模型资料组的数据采用多元线性回归法分别拟合1~4个采血点估算AUC_(0-12 h)的数学模型对初选模型用验证资料组数据检验预测偏差和精密度,并结合临床实际情况选择最佳模型。结果 12h谷浓度建立的单点预测模型r~2值为0.957平均预测误差及绝对预测误差分别为-3.93与11.93,均<15%;0.5,1.5,6,8 h或1.5,6,8,12h四点拟合的模型与AUC_(0-12 h)的相关性均达0.996,验证后预测误差在±15%以内的例数为9例;另有2点模型与3点模型,预测效果介于上述两种模型之间。结论结合临床实际情况,精确度较高的四点模型AUC=3.19+0.49c_(0.5h)、+1.76c_(1.5h)+2.95c_(6h)+5.46c_(8h)适用于晨起服药后取血监测的住院患者;相对简便的单点模型AUC=10.82+13.37c_(12 h)适用于夜间服药次日晨起入院监测的门诊患者。Bland-Altman分析显示,两种模型均能较好预测使用MMF+甲泼尼龙二联免疫抑制方案的中国成年AID患者的MPA AUC。
Objective To establish a finite sampling method to estimate the in vivo exposure dose (AUC) of mycophenolic acid (MPA) in patients with autoimmune diseases (AID). Methods Twenty-four patients with AID were treated with MMF 0.75 g for 12 h. After reaching steady state, the active metabolites MPA (0.5, 1, 1.5, 2, 4, 6, 8 and 12 h) (0-12 h). The data of 24 patients were randomly divided into model data group and verification data group, and the data of model data group were fitted by 1 ~ 4 mathematical models for estimation of AUC_ (0-12 h) at the blood collection point. The prediction bias and precision of the primary data model were verified by the verification data set data, and the best model was selected according to the clinical situation. Results The single point prediction model established at 12h of valley concentration had a r ~ 2 value of 0.957. The average prediction error and absolute prediction error were -3.93 and 11.93, respectively, <15%; 0.5, 1.5, 6 and 8 h or 1.5, 12h four-point fitting model and AUC_ (0-12 h) were 0.996, the prediction error is less than ± 15% of the number of cases in 9 cases; another 2 points model and 3 points model, the predictive effect Between the above two models. Conclusion AUC = 3.19 + 0.49c (0.5h), + 1.76c (1.5h) + 2.95c (6h) + 5.46c (8h) are suitable for early morning after taking medicine Hospitalized patients for blood monitoring; a relatively simple single-point model AUC = 10.82 +13.37c_ (12 h) for night-time medication the next morning hospitalization monitoring outpatients. Bland-Altman analysis showed that both models were better predictors of MPA AUC in Chinese adults with AID using MMF + methylprednisolone dual immunosuppression regimen.