Background/Objectives:This study investigates the utility of multiparametric PET/MRI in delineating changes in physiologically distinct intratumoral habitats during trastuzumab-induced alterations in a preclinical HER2+ breast cancer model.Methods:By integrating diffusion-weighted MRI, dynamic contrast-enhanced MRI, [18F]Fluorodeoxyglucose- and [18F]Fluorothymidine-PET, voxel-wise parametric maps were generated capturing cellular density, vascularity, metabolism, and proliferation. BT-474 tumor-bearing mice have high expression of HER2 and, in response to trastuzumab, an anti-HER2 antibody, effectively show changes in proliferation and tumor microenvironment alterations that result in decreases in tumor volume through time.Results:Single imaging metrics and changes in metrics were incapable of identifying treatment-induced alterations early in the course of therapy (day 4) prior to changes in tumor volume. Hierarchical clustering identified five distinct tumor habitats, which enabled longitudinal assessment of early treatment response. Tumor habitats were defined based on imaging metrics related to biology and categorized as highly vascular (HV), hypoxic responding (HRSP), transitional zone (TZ), active tumor (ATMR) and responding (RSP). The HRSP cluster volume significantly decreased in trastuzumab-treated tumors compared to controls by day 4 (p= 0.015). The volume of ATMR cluster was significantly different at baseline between cohorts (p= 0.03). The TZ cluster, indicative of regions transitioning more to necrosis, significantly decreased in treated tumors (p= 0.031), suggesting regions had already transitioned. Multiparametric image clustering showed a significant positive linear correlation with histological multiparametric mapping, with R2values of 0.56 (HRSP,p= 0.013, 0.64 (ATMR,p= 0.0055), and 0.49 (responding cluster,p= 0.024), confirming the biological relevance of imaging-derived clusters.Conclusions:These findings highlight the potential utility of multiparametric PET/MRI to capture biological alterations prior to any single imaging metric which has potential for better understanding longitudinal changes in biology, stratifying tumors based on those changes, optimizing therapeutic monitoring and advancing precision oncology.
背景/目的：本研究旨在探讨多参数PET/MRI在临床前HER2阳性乳腺癌模型中，用于描绘曲妥珠单抗治疗期间肿瘤内部不同生理特征区域变化的效用。 方法：通过整合扩散加权MRI、动态对比增强MRI、[18F]氟代脱氧葡萄糖-PET及[18F]氟代胸苷-PET，生成体素级参数图以捕捉肿瘤的细胞密度、血管分布、代谢活性及增殖状态。BT-474荷瘤小鼠具有高表达的HER2受体，在抗HER2抗体曲妥珠单抗作用下，其增殖活性和肿瘤微环境发生显著改变，导致肿瘤体积随时间逐渐缩小。 结果：单一影像学指标及其变化无法在治疗早期（第4天）肿瘤体积尚未改变时识别出治疗诱导的生物学变化。通过层次聚类分析识别出五种不同的肿瘤区域，实现了对早期治疗反应的纵向评估。这些区域基于与生物学相关的影像学指标定义，分别为高血管区（HV）、低氧反应区（HRSP）、过渡区（TZ）、活性肿瘤区（ATMR）及反应区（RSP）。与对照组相比，曲妥珠单抗治疗组肿瘤的HRSP区域体积在第4天显著减少（p=0.015）。ATMR区域体积在基线时两组间已存在显著差异（p=0.03）。TZ区域（提示向坏死过渡的区域）在治疗组中显著减少（p=0.031），表明该区域已发生转变。多参数影像聚类与组织学多参数映射呈显著正线性相关，R²值分别为0.56（HRSP，p=0.013）、0.64（ATMR，p=0.0055）和0.49（反应区，p=0.024），证实了影像学聚类结果的生物学相关性。 结论：这些发现凸显了多参数PET/MRI在捕捉单一影像指标无法反映的早期生物学改变方面的潜在价值，有助于深入理解肿瘤生物学的纵向变化、基于这些变化进行肿瘤分层、优化治疗监测并推动精准肿瘤学的发展。
肿瘤(癌症)患者之家