Erythropoiesis Stimulating Agents (ESAs) were developed for therapeutic purposes to stimulate red blood cell (RBC) production. Consequently, tissue oxygenation is enhanced as athlete's endurance and ESAs misuse now benefits doping. Our hypothesis is that most of ESAs should have similar mechanisms and thus have the same effects on metabolism. Studying the metabolome variations could allow suspecting the use of any ESAs with a single method by targeting their effects. In this objective, a metabolomic study was carried out on 3 thoroughbred horses with a single administration of 4.2 μg/kg of Mircera^®, also called Continuous Erythropoiesis Receptor Activator (CERA). Blood and urine samples were collected from D_-17 to D₊₇₄ and haematological parameters were followed throughout the study as plasmatic CERA concentration (ELISA). Urine and plasma metabolic fingerprints were recorded by Liquid Chromatography coupled to High Resolution Mass Spectrometry (LC-HRMS) in positive and negative mode. After preprocessing steps, normalized data were analyzed by multivariate statistics to build OPLS models. Hemoglobin concentration and hematocrit showed a significant increase after CERA administration unlike reticulocytes. CERA concentration showed a high intensity peak and then a slow decrease until becoming undetectable after D₊₃₁. Models built with multivariate statistics allow a discrimination between pre and post-administration plasma and urine samples until 74 days after administration, i.e. 43 days longer than ELISA method. By reducing and studying variables (ions), some potential candidate biomarkers were found.

红细胞生成刺激剂（ESA）被开发用于治疗目的，以刺激红细胞（RBC）的产生。因此，由于运动员的耐力和ESA滥用现在有益于兴奋剂，因此组织的氧合作用得到了增强。我们的假设是，大多数ESA应该具有相似的机制，因此对新陈代谢具有相同的作用。研究代谢组的变异可能会针对其效应，从而怀疑以单一方法使用任何ESA。在此目标中，代谢物组进行了研究，在3名纯种马为4.2微克/千克的Mircera的单次给药^®，也称为连续红血球生成受体活化剂（CERA）。从D _-17至D ₊₇₄采集血液和尿液样本在整个研究过程中，均采用血浆CERA浓度（ELISA）跟踪血液学参数。尿液和血浆代谢指纹图谱通过液相色谱与高分辨率质谱法（LC-HRMS）结合以正负两种方式记录。在预处理步骤之后，通过多元统计分析归一化的数据以建立OPLS模型。与网织红细胞不同，CERA给药后血红蛋白浓度和血细胞比容显着增加。CERA浓度显示高强度峰，然后缓慢下降，直到在D ₊₃₁之后变得无法检测到。利用多元统计数据建立的模型可以区分给药前和给药后血浆和尿液样品，直到给药后74天，即比ELISA方法长43天。通过减少和研究变量（离子），发现了一些潜在的候选生物标记。