Analytical Validation of Aptamer-Based Serum Vancomycin Monitoring Relative to Automated Immunoassays
收藏资源简介：
***ABSTRACT*** _________________________ The practice of monitoring therapeutic drug concentrations in patient biofluids can significantly improve clinical outcomes while simultaneously minimizing adverse side effects. A model example of this practice is vancomycin dosing in intensive care units. If dosed correctly, vancomycin can effectively treat methicillin-resistant staphylococcus aureus (MRSA) infections. However, it can also induce nephrotoxicity or fail to kill the bacteria if dosed too high or low, respectively. Although undeniably important to achieve effectiveness, therapeutic drug monitoring remains inconvenient in practice, due primarily to the lengthy process of sample collection, transport to a centralized facility, and analysis using costly instrumentation. Adding to this workflow is the possibility of backlogs at centralized clinical laboratories, which is not uncommon and may result in additional delays between biofluid sampling and concentration measurement, which can negatively affect clinical outcomes. Here we explore the possibility of using point-of-care, electrochemical aptamer-based (E-AB) sensors to minimize the time delay between biofluid sampling and drug measurement. Specifically, we conducted a clinical agreement study comparing the measurement outcomes of E-AB sensors to the benchmark automated competitive immunoassays for vancomycin monitoring in serum. Our results demonstrate that E-ABs are selective for free vancomycin—the active form of the drug—over total vancomycin. In contrast, competitive immunoassays measure total vancomycin, including both protein-bound and free drug. Accounting for these differences in a pilot study consisting of 85 clinical samples, we demonstrate that the E-AB vancomycin measurement achieved a 95% positive correlation rate with the benchmark immunoassays. Therefore, we conclude that E-AB sensors could provide clinically useful stratification of patient samples at trough sampling to guide effective vancomycin dose recommendations. ***README FILE *** _________________________ Project: Analytical Validation of Aptamer-Based Serum Vancomycin Monitoring Relative to Automated Immunoassays Date Accepted: December 5th, 2023 DOI: Contact: Netz Arroyo, netzarroyo@jhmi.edu ORGANIZATION All files in this repository are sorted based on their corresponding figure in the associated manuscript and supplementary information document (both found at DOI above). Organization and files found in each figure folder are as follows: - 'Figure_1': - Final .svg image for Figure 1. - 'Figure_2': -"Figure_2.png" - PNG image of Figure 2. Figure panels were generated as detailed in the following entries. Figure was compiled in Adobe Illustrator. -"swv trace.svg" - SVG file for Panel C. Generated using Igor Pro v8 from data in file "SWV Traces.pxp". -"ON&OFF.svg" - SVG file for Panel D . Generated using Igor Pro v8 from data in file "Signal ON&OFF.pxp". -"ratio box whisker plot.svg" - SVG file for Panel E. Generated using Igor Pro v8. -"calibration.svg" - SVG file for Panel F. Generated using Igor Pro v8. -Data: "SWV Traces.xlsx" "SWV Traces.csv" "SWV Traces.pxp" Data was obtained from sheet "SWV curves" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. Data in all files are identical, except for file format. -Data: "Signal ON&OFF.xlsx" "Signal ON&OFF.csv" "Signal ON&OFF.pxp" Data was obtained from sheet "signal-on&signal-off" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. Data in all files are identical, except for file format. Regression to the Hill equation was performed in Igor Pro v8, constraining rate=1. -Data: "Dose response curve box plot.xlsx" "Dose response curve box plot.csv" "Dose response curve box plot.pxp" Data was obtained from sheet "Curve fitting_Ratio(SUM_MEA)" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. Data in all files are identical, except for file format. Box and whisker plot was generated using Igor Pro v8. -Data: "Day1-Day6 Dose Response Curve.xlsx" "Day1-Day6 Dose Response Curve.csv" "Day1-Day6 Dose Response Curve.pxp" Data was obtained from sheet "Curve fitting_Ratio(SUM_MEA)" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. Data in all files are identical, except for file format. Nonlinear regression to the Hill equation was generated using Igor Pro v8. No variables were constrained. -'Figure_3': - DATA FOR FIGURE 3 WAS OBTAINED FROM THE "Ratiometric response" SHEET OF THE FILE "Confidential_ZiO Health Stability Test Data" RECIEVED 12:08, NOVEMBER 8 2023, FROM FRANCESCA NAPOLI. - DATA FROM FIELDS "Average response to 12.5 mg/L for chips tested in specific week" AND "SD response to 12.5 mg/L for chips tested in specific week" (ROWS 39 AND 40) , WITH ASSOCIATED TIMEPOINT IDENTIFIERS, WERE EXTRACTED AND USED AS RECIEVED - IN ROWS 41 AND 42, DATA FROM THE 0.1 MG/L CONCENTRATION POINT (ROW 13) WAS EXCTRACTED AND PROCESSED IN THE SAME WAY AS "Average response to 12.5 mg/L for chips tested in specific week" AND "SD response to 12.5 mg/L for chips tested in specific week" (I.E, AVERAGE AND STANDARD DEVIATION OF THREE OR FOUR CHIPS PER WEEK) - TIMEPOINT IDENTIFIERS WERE CONVERTED TO NUMERIC AND GIVEN THE TITLE "Time" and UNIT "Weeks" - FIELD "Average response to 12.5 mg/L for chips tested in specific week" WAS TITLED "12.5 mg/L Sensor response ratio" AND GIVEN UNIT "360Hz/20Hz" - FIELD "SD response to 12.5 mg/L for chips tested in specific week" WAS TITLED "12.5 mg/L SD" AND GIVEN UNIT "360Hz/20Hz" - THE SAME WAS DONE FOR THE 0.1 MG/l CONCENTRATION DATA - DATA FILE USED FOR MAKING THE FIGURE IS "2023_11_08_Chip_Stability.pxp" - BOTH O.1 MG/L AND 12.5 MG/L WERE SUBJECTED TO LINEAR REGRESSION ANALYSIS IN IGOR V8 -'Figure_4': - Final PNG image for Figure 4. -'Figure_5': - "2023_11_07_Fig5.png" - Final .PNG file used for Figure 5. Panels were arranged and Mean/StDev numbers were added in Adobe Illustrator. - "EAB_diff.svg" -.svg file used for panel A of Figure 5. Generated in Igor v8. - "EAB_pct_aveplot.svg" -.svg file used for panel C of Figure 5. Generated in Igor v8. - "ImA_diff.svg" -.svg file used for panel B of Figure 5. Generated in Igor v8. - "ImA_pct_aveplot.svg" -.svg file used for panel D of Figure 5. Generated in Igor v8. - Data: "2023_07_10_Differentials.xlsx" "2023_07_10_Differentials.csv" "2023_07_10_Differentials.pxp" Data was obtained from sheet "Predict con.c" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. -'Figure_6': - "2023_07_12_Fig6.png" - Final .PNG file used for Figure 6. Panels were arranged and statistics valued added in Adobe Illustrator. - "serum.svg" -.svg file used for panel A of Figure 5. Generated in Igor v8 from Table 0. - "Filtrate smaller points.svg" -.svg file used for panel C of Figure 5. Generated in Igor v8 from Graph 0. - "Filtrate_insert.svg" -.svg file used for panel B of Figure 5. Generated in Igor v8 from Graph 5. - "Bland-Altman.svg" -.svg file used for panel D of Figure 5. Generated in Igor v8 from Graph 1. - "Bland-Altman insert.svg" -.svg file used for panel D of Figure 5. Generated in Igor v8 from Graph 1. - "Rplot0_png.png" -.png file of Bangdiwala plot. Generated in Rstudio by code mentioned below. - "Agreement plot code.txt" -Code used to generate Bangdiwala agreement plot using Rstudio. - Data: "Concentrations_all.xlsx" "Concentrations_all.csv" Data was obtained from sheet "Predict con.c" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. - Data: "Bland-Altman.xlsx" "Bland-Altman.csv" Data was obtained from sheet "Bland–Altman Plot" of Excel file "ZiO-JHU_March 2023_Data processing Updates" recieved 10:37 EST, June 2 2023, from Yu Liu. Data was used as recieved, and imported to Igor Pro v8 for plotting. - Data: "ROC-input data for agreement plot.xlsx" "ROC-input data for agreement plot.csv" Data was recieved 12:30 EST, December 7 2023, from Francesca Napoli. - Data: "Fig6.pxp" Table 0: all samples from "Concentrations_all.xlsx" Table 1: average and difference values for EA-B filtrate samples vs immunoassay filtrate samples from "Bland-Altman.xlsx", as well as upper and lower 95% CI boundaries Table 2: samples sorted by concentration and filtered by immunoassay values <15mg/L and <10mg/L. Graph 0: plot of immunoassay vs E-AB serum concentration values from Table 0. Graph 1: plot of immunoassay vs E-AB filtrate concentration values from Table 0. Graph 2: plot of average vs difference values from Tablel 1. Graph 4: plot of filtered immunoassay vs E-AB filtrate concentration values <10mg/L from Table 3. Graph 5: plot of filtered immunoassay vs E-AB filtrate concentration values <15mg/L from Table 3. NOTES (1) Raw voltammetry data is the property of ZiO Health, and is therefore not available for public release. All data presented here were extracted from raw voltammetric traces by ZiO Health, and transfered to the Arroyo lab for further analysis. (2) RStudio is available at: https://posit.co/products/open-source/rstudio/ (3) Igor Pro is available at: https://www.wavemetrics.com/
***摘要*** _________________________ 监测患者生物体液中的治疗药物浓度，可显著改善临床结局，同时最大限度降低不良反应。万古霉素（vancomycin）在重症监护病房的给药方案便是该实践的典型范例。若给药剂量恰当，万古霉素可有效治疗甲氧西林耐药金黄色葡萄球菌（methicillin-resistant staphylococcus aureus, MRSA）感染；但剂量过高或过低时，分别可能引发肾毒性（nephrotoxicity）或无法杀灭病原菌。尽管实现治疗有效性至关重要，但治疗药物监测（therapeutic drug monitoring）在临床实践中仍存在诸多不便，其核心原因在于样本采集、转运至集中化检测机构以及使用昂贵仪器进行分析的流程耗时冗长。此外，集中化临床实验室可能出现检测积压，这一情况并不少见，可能进一步延长生物体液采样至浓度检测之间的时间间隔，进而对临床结局产生负面影响。本研究探讨了使用床旁（point-of-care）电化学适体传感器（electrochemical aptamer-based, E-AB）缩短采样与药物浓度检测之间时间延迟的可能性。具体而言，我们开展了一项临床一致性研究，将E-AB传感器的检测结果与用于血清万古霉素监测的基准自动化竞争性免疫分析法（competitive immunoassays）进行对比。我们的研究结果表明，相较于总万古霉素，E-AB传感器对游离万古霉素——即药物的活性形式——具有选择性；与之相对，竞争性免疫分析法检测的是总万古霉素，包括蛋白结合型与游离型药物。在一项纳入85份临床样本的预试验中，我们对上述差异进行校正后证实，E-AB万古霉素检测与基准免疫分析法的阳性相关率可达95%。因此，我们认为E-AB传感器可在谷浓度采样时对患者样本进行具有临床实用价值的分层，以辅助制定有效的万古霉素给药方案。 ***README文件*** _________________________ 项目：基于适体的血清万古霉素检测相对于自动化免疫分析法的分析验证 接收日期：2023年12月5日 DOI： 联系方式：Netz Arroyo，netzarroyo@jhmi.edu ## 组织机构说明 本仓库中的所有文件均按照相关手稿与补充材料文档（均可通过上述DOI获取）中的对应图表进行分类整理。每个图表文件夹内的组织形式与文件清单如下： - 'Figure_1': - 图1的最终.svg格式图像文件。 - 'Figure_2': - "Figure_2.png"：图2的PNG格式图像。该图的子图按照下述说明生成，整体使用Adobe Illustrator进行排版。 - "swv trace.svg"：图2子图C的SVG文件，使用Igor Pro v8基于文件"SWV Traces.pxp"中的数据生成。 - "ON&OFF.svg"：图2子图D的SVG文件，使用Igor Pro v8基于文件"Signal ON&OFF.pxp"中的数据生成。 - "ratio box whisker plot.svg"：图2子图E的SVG文件，使用Igor Pro v8生成。 - "calibration.svg"：图2子图F的SVG文件，使用Igor Pro v8生成。 - 数据文件： - "SWV Traces.xlsx"、"SWV Traces.csv"、"SWV Traces.pxp"：数据来源于2023年6月2日美国东部时间10:37由Yu Liu提供的Excel文件"ZiO-JHU_March 2023_Data processing Updates"中的"SWV curves"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。所有数据文件内容完全一致，仅文件格式不同。 - "Signal ON&OFF.xlsx"、"Signal ON&OFF.csv"、"Signal ON&OFF.pxp"：数据来源于上述同一Excel文件中的"signal-on&signal-off"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。所有数据文件内容完全一致，仅文件格式不同。本研究在Igor Pro v8中进行了希尔方程（Hill equation）回归分析，将速率参数约束为1。 - "Dose response curve box plot.xlsx"、"Dose response curve box plot.csv"、"Dose response curve box plot.pxp"：数据来源于上述同一Excel文件中的"Curve fitting_Ratio(SUM_MEA)"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。所有数据文件内容完全一致，仅文件格式不同。箱线图使用Igor Pro v8生成。 - "Day1-Day6 Dose Response Curve.xlsx"、"Day1-Day6 Dose Response Curve.csv"、"Day1-Day6 Dose Response Curve.pxp"：数据来源于上述同一Excel文件中的"Curve fitting_Ratio(SUM_MEA)"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。所有数据文件内容完全一致，仅文件格式不同。希尔方程非线性回归分析使用Igor Pro v8生成，未对变量进行约束。 - 'Figure_3': - 图3的数据来源于2023年11月8日美国东部时间12:08由Francesca Napoli提供的文件"Confidential_ZiO Health Stability Test Data"中的"Ratiometric response"工作表。 - 提取并使用了行39与行40中"针对特定周测试的芯片的12.5 mg/L平均响应值"与"针对特定周测试的芯片的12.5 mg/L响应标准差"及其对应的时间点标识。 - 在第41与42行，提取了0.1 mg/L浓度点（第13行）的数据，并采用与上述12.5 mg/L数据相同的方式进行处理（即每周3~4片芯片的平均值与标准差）。 - 将时间点标识转换为数值，并命名为"Time"，单位为"Weeks"。 - 将"针对特定周测试的芯片的12.5 mg/L平均响应值"命名为"12.5 mg/L Sensor response ratio"，单位为"360Hz/20Hz"。 - 将"针对特定周测试的芯片的12.5 mg/L响应标准差"命名为"12.5 mg/L SD"，单位为"360Hz/20Hz"。 - 0.1 mg/L浓度的数据采用相同方式命名与处理。 - 用于生成该图的数据文件为"2023_11_08_Chip_Stability.pxp"。 - 对0.1 mg/L与12.5 mg/L两组数据均在Igor v8中进行了线性回归分析。 - 'Figure_4': - 图4的最终PNG格式图像文件。 - 'Figure_5': - "2023_11_07_Fig5.png"：图5的最终PNG格式文件。子图排版与均值/标准差数值使用Adobe Illustrator添加。 - "EAB_diff.svg"：图5子图A的SVG文件，使用Igor v8生成。 - "EAB_pct_aveplot.svg"：图5子图C的SVG文件，使用Igor v8生成。 - "ImA_diff.svg"：图5子图B的SVG文件，使用Igor v8生成。 - "ImA_pct_aveplot.svg"：图5子图D的SVG文件，使用Igor v8生成。 - 数据文件： - "2023_07_10_Differentials.xlsx"、"2023_07_10_Differentials.csv"、"2023_07_10_Differentials.pxp"：数据来源于前述Excel文件"ZiO-JHU_March 2023_Data processing Updates"中的"Predict con.c"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。 - 'Figure_6': - "2023_07_12_Fig6.png"：图6的最终PNG格式文件。子图排版与统计数值使用Adobe Illustrator添加。 - "serum.svg"：图6子图A的SVG文件，使用Igor v8基于Table 0生成。 - "Filtrate smaller points.svg"：图6子图C的SVG文件，使用Igor v8基于Graph 0生成。 - "Filtrate_insert.svg"：图6子图B的SVG文件，使用Igor v8基于Graph 5生成。 - "Bland-Altman.svg"：图6子图D的SVG文件，使用Igor v8基于Graph 1生成。 - "Bland-Altman insert.svg"：图6子图D的补充SVG文件，使用Igor v8基于Graph 1生成。 - "Rplot0_png.png"：Bangdiwala图（Bangdiwala plot）的PNG文件，通过下述代码在Rstudio中生成。 - "Agreement plot code.txt"：用于在Rstudio中生成Bangdiwala一致性图的代码。 - 数据文件： - "Concentrations_all.xlsx"、"Concentrations_all.csv"：数据来源于前述Excel文件中的"Predict con.c"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。 - "Bland-Altman.xlsx"、"Bland-Altman.csv"：数据来源于前述Excel文件中的"Bland–Altman Plot"工作表。数据按原始格式使用，并导入Igor Pro v8进行绘图。 - "ROC-input data for agreement plot.xlsx"、"ROC-input data for agreement plot.csv"：2023年12月7日美国东部时间12:30由Francesca Napoli提供的数据文件。 - "Fig6.pxp"： - Table 0：来自"Concentrations_all.xlsx"的所有样本数据。 - Table 1：来自"Bland-Altman.xlsx"的E-AB滤液样本与免疫分析法滤液样本的均值与差值，以及95%置信区间的上下边界。 - Table 2：按浓度排序，且免疫分析法检测值<15mg/L与<10mg/L的筛选后样本。 - Graph 0：基于Table 0绘制的免疫分析法与E-AB血清浓度值对比图。 - Graph 1：基于Table 0绘制的免疫分析法与E-AB滤液浓度值对比图。 - Graph 2：基于Table 1绘制的均值与差值对比图。 - Graph 4：基于Table 3绘制的免疫分析法与E-AB滤液浓度值（<10mg/L）筛选后对比图。 - Graph 5：基于Table 3绘制的免疫分析法与E-AB滤液浓度值（<15mg/L）筛选后对比图。 ## 备注 (1) 原始伏安法数据为ZiO Health所有，因此不对外公开。本研究中所有呈现的数据均由ZiO Health从原始伏安曲线中提取，并移交至Arroyo实验室进行后续分析。 (2) RStudio可通过以下网址获取：https://posit.co/products/open-source/rstudio/ (3) Igor Pro可通过以下网址获取：https://www.wavemetrics.com/
