藥物治療監測之臨床應用： methicillin-resistant Staphylococcus aureus感染個案teicoplanin濃度及住院天數之分析

Clinical practice in therapeutic drug monitoring: analysis of teicoplanin concentration and length of stay in hospital among methicillin-resistant Staphylococcus aureus patients

太古盤寧素 (teicoplanin) 為耐甲氧西林金黃色葡萄 球菌 (methicillin-resistant Staphylococcus aureus; MRSA) 感染之治療藥物，但在台灣 teicoplanin 藥物治療 監測非常規檢驗項目。本研究針對 12 名 MRSA 感染個案，在給予 teicoplanin 治療 5 日、8 日、12 日後，分別收集給藥前之血液檢體，進行血中抗生素 濃度之檢測，同時檢測 blood urea nitrogen 、creatini ne、及 cystatin C 等腎功能指標。並利用兩組多變 量線性迴歸模型，比較納入 teicoplanin 數據與否所得 之預測住院天數與實際住院天數之差距，以評估 teicoplanin 藥物治療監測的重要性。12名研究個案中，teicoplanin 平均濃度為19.4 mg/L (1.7 – 41.4 mg/L)，其中 9 名個案 teicoplanin 濃度均達到理想治療濃度 (10 – 60 mg/L)，2 名個案其 3 次採檢中有 1 次檢體中 teicoplanin 濃度低於 10 mg/L，1 名個案之 3 次 teicopla nin 濃度皆低於 10 mg/L。5 名個案 3 項腎功能指標皆正常，7 名個案至少一項指標異常升高。住院天數平 均為 21.3 天(9 – 32 天)，死亡率 16.7%。經迴歸 模型分析，納入 teicoplanin 數據之預測住院天數(擬合住 院天數)，較不納入者，接近實際住院天數 (1.75天 vs 2.85 天差距)。本研究提供了 12 名 MRSA 個案 teicoplanin 藥物治療之臨床監測結果，並由納入teicoplanin 數據才能正確預測住院天數之迴歸模型分析 結果，顯示藥物治療監測之於治療方案執行與管理的重要。

Teicoplanin is a therapeutic drug for methicillin-resistant Staphylococcus aureus (MRSA) infection. Therapeutic drug monitoring (TDM) of teicoplanin conc entration is not performed routinely in Taiwan. We prospectively collected blood samples from 12 MRSA patients prior to 5, 8, and 12 days of teicoplanin prescriptions. Teicoplanin concentrations were measur ed during therapy. Three additional biomarkers, blood urea nitrogen, creatinine, and cystatin C, were also determined to evaluate patient’s renal function. Medical charts of the 12 patients were reviewed and clinical information including the lengths of stay (LOS) results were collected. We performed two multivariate linear regression models with or without teicoplanin data inputs to generate two predict LOS (Fitted LOS). The difference between actual and predict LOS results of the two models were compared. Among the 12 patients, the average teicoplanin concentration was 19.4 mg/L (1.7 to 41.4 mg/L). Optimal therapeutic concentration (10-60 mg/L) was determined in all samples from 9 patients. Suboptimal concentration (<10 mg/L) was found in 3 patients. Of the 5 patients, all 3 biomarkers were in normal range. The average durat ion of hospitalization was 21.3 days (9 to 32 days). Mortality rate was 16.7%. In the model analysis, the predict LOS from the model with teicoplanin data inputs was close to the actual LOS relative to the model without teicoplanin data inputs (1.75 days vs 2.85 days). In this study, we provided practical monitoring results of teicoplanin among 12 MRSA patients. Data inputs of teicoplanin in test model is necessary to predict LOS indicating that teicoplanin TDM is essential in therapeutic regimen management.