Optimizing GVHD Prevention with Systems Pharmacology Models

项目摘要 定量系统药理学（QSP）是一个快速扩展的领域，它整合了体外， 临床前和临床数据代表现有知识，以实现反向翻译。反向 翻译使用实时人类临床数据直接告知新的发现，现有的治疗方法， 疾病进展的属性。在这里，我们寻求成为第一个建立QSP模型的人，先验地预测 使用群体先验（群体药代动力学或 popPK建模）和生理学预测（使用基于生理学的药代动力学或PBPK建模） 结合基于体外和临床前数据的药效学模型。我们将把这种新的混合动力 popPK-PBPK-PD QSP模型用于同种异体造血细胞移植（HCT），因为其成功需要 细胞从一个个体（供体）移植到另一个个体（宿主，HCT受体）时的微妙平衡。指导 根据我们的初步数据，我们的工作假设是，QSP模型可以最大限度地减少个体间的变异性， 这些免疫抑制剂，同时也优化了新的移植物抗宿主病（GVHD）治疗方案， 移植环磷酰胺（PTCy）。目的1寻求使用popPK和PBPK确定最佳PTCy剂量 模型我们的临床前数据显示，PTCy具有窄的剂量窗口，中间剂量具有 最低的GVHD率。为了在每个患者中实现最佳的PTCy剂量，我们寻求开发一种有效的popPK- PBPK模型建立在我们在定量4-羟基环磷酰胺（4 HCY）方面的独特专业知识基础上， CY的细胞毒性代谢物的主要前体，并使用popPK指导的给药使CY个性化。这 将使用我们的回顾性和前瞻性（n=150）队列开发混合popPK-PBPK CY模型。的 前瞻性队列将在国家癌症研究所（NCI）和希望之城（COH）招募。NCI队列 将确定PTCy剂量和时间表是否可以减少（减少75%）而不影响GVHD率; COH队列将使用传统的PTCy给药。在目标2中，我们将描述药代动力学特征， 霉酚酸活性代谢产物霉酚酸（MPA）与其靶酶肌苷的药效学 单磷酸脱氢酶（IMPDH）。与CY一样，MPA具有显著的药代动力学变异性， 不同的代谢和转运途径，因此需要单独的药代动力学模型。我们寻求创造 popPK-PBPK-PD模型，以确定MPA和IMPDH活性的最佳血浆暴露。在目标3中，我们 建立T细胞反应和急性GVHD的定量系统药理学（QSP）模型。我们的临床前 数据显示PTCy预防急性GVHD与CD 4 + CD 25-Foxp 3- 在第+7天常规T细胞（Tcon）增殖，随后是CD 4 + CD 25 + Foxp 3+的优先扩增 在第+21天的调节性T细胞（TcB）。基于完全集成免疫反应模型（FIRM），我们寻求 整合体外、临床前和临床数据以构建QSP模型。

PROJECT SUMMARY Quantitative systems pharmacology (QSP) is a rapidly expanding area that integrates available in vitro, preclinical, and clinical data representing existing knowledge to achieve a reverse translation. Reverse translation uses real-time human clinical data to directly inform new discoveries, of existing therapies and attributes of disease progression. Here, we seek to be the first to build a QSP model to, a priori, predict interpatient pharmacokinetics and pharmacodynamics using population priors (population pharmacokinetic or popPK modeling) and physiologic predictions (using physiologically based pharmacokinetic or PBPK modeling) combined with pharmacodynamic models based on in vitro and preclinical data. We will apply this novel hybrid popPK-PBPK-PD QSP model to allogeneic hematopoietic cell transplant (HCT) because its success requires a delicate balance as the grafting of cells from one individual (donor) to another (host, the HCT recipient). Guided by our preliminary data, our working hypothesis is that QSP modeling can minimize interindividual variability of these immunosuppressants, while also optimizing the novel graft versus host disease (GVHD) regimen of post- transplant cyclophosphamide (PTCy). Aim 1 seeks to identify the optimal PTCy dose using popPK and PBPK models. Our preclinical data shows that PTCy has a narrow dose window, with intermediate doses having the lowest GVHD rates. To achieve the optimal PTCy dose in each patient, we seek to develop a validate a popPK- PBPK model building upon our unique expertise in quantitating 4-hydroxycyclophosphamide (4HCY), the primary precursor to the cytotoxic metabolite of CY and personalizing CY using popPK-guided dosing. This hybrid popPK-PBPK CY model will be developed using our retrospective and prospective (n=150) cohort. The prospective cohort will be enrolled at National Cancer Institute (NCI) and City of Hope (COH). The NCI cohort will determine if the PTCy dose and schedule can be reduced (by 75%) without compromising GVHD rates; the COH cohort will use the traditional PTCy dosing. In Aim 2, we will characterize the pharmacokinetics and pharmacodynamics of mycophenolic acid (MPA), the active metabolite of MMF, with its target enzyme inosine monophosphate dehydrogenase (IMPDH). Like CY, MPA has substantive pharmacokinetic variability but different metabolic and transport pathways so separate pharmacokinetic models are needed. We seek to create a popPK-PBPK-PD model to identify the optimal plasma exposure of MPA and IMPDH activity. In Aim 3, we will create a quantitative systems pharmacology (QSP) model of T-cell response and acute GVHD. Our preclinical data show that acute GVHD prevention with PTCy is associated with reduction of CD4+CD25-Foxp3- conventional T-cell (Tcon) proliferation at day +7 followed by the preferential expansion of CD4+CD25+Foxp3+ regulatory T cells (Tregs) at day +21. Building upon fully-integrated immune response model (FIRM), we seek integrate in vitro, preclinical, and clinical data to build a QSP model.