Purpose

This study evaluates the potential of within-individual Metabolic Connectivity (wi-MC), from dynamic [¹⁸F]FDG PET data, based on the Euclidean Similarity method. This approach leverages the biological information of the tracer’s full temporal dynamics, enabling the direct extraction of individual metabolic connectomes. Specifically, the proposed framework, applied to glioma pathology, seeks to assess sensitivity to metabolic dysfunctions in the whole brain, while simultaneously providing further insights into the pathophysiological mechanisms regulating glioma progression.

Methods

We designed an index (Distance from Healthy Group, DfHG) based on the alteration of wi-MC in each patient (n = 44) compared to a healthy reference (from 57 healthy controls), to individually quantify metabolic connectivity abnormalities, resulting in an Impairment Map highlighting significantly compromised areas. We then assessed whether our measure of metabolic network alteration is associated with well-established markers of disease severity (tumor grade and volume, with and without edema). Subsequently, we investigated disruptions in wi-MC homotopic connectivity, assessing both affected and seemingly healthy tissue to deepen the pathology’s impact on neural communication. Finally, we compared network impairments with local metabolic alterations determined from SUVR, a validated diagnostic tool in clinical practice.

Results

Our framework revealed how gliomas cause extensive alterations in the topography of brain networks, even in structurally unaffected regions outside the lesion area, with a significant reduction in connectivity between contralateral homologous regions. High-grade gliomas have a stronger impact on brain networks, and edema plays a mediating role in global metabolic alterations. As compared to the conventional SUVR-based analysis, our approach offers a more holistic view of the disease burden in individual patients, providing interesting additional insights into glioma-related alterations.

Conclusion

Considering our results, individual PET connectivity estimates could hold significant clinical value, potentially allowing the identification of new prognostic factors and personalized treatment in gliomas or other focal pathologies.

中文翻译：

---

来自动态 [18F]FDG PET 的个体水平代谢连接揭示了神经胶质瘤诱导的大脑结构损伤，并提供了超越 SUVR 临床标准的新见解

 目的

本研究基于欧几里得相似性方法，从动态 [¹⁸F]FDG PET 数据中评估个体内代谢连接 （wi-MC） 的潜力。这种方法利用示踪剂完整时间动态的生物学信息，能够直接提取单个代谢连接组。具体来说，应用于神经胶质瘤病理学的拟议框架旨在评估对整个大脑代谢功能障碍的敏感性，同时提供对调节神经胶质瘤进展的病理生理机制的进一步见解。

 方法

我们根据每位患者 （n = 44） 的 wi-MC 变化与健康参考（来自 57 名健康对照者）相比设计了一个指数 （Distance from Healthy Group， DfHG），以单独量化代谢连接异常，从而产生一个损伤图突出显着受损的区域。然后，我们评估了我们对代谢网络改变的测量是否与疾病严重程度的公认标志物 （肿瘤分级和体积，伴或不伴水肿） 相关。随后，我们研究了 wi-MC 同位连接的中断，评估了受影响和看似健康的组织，以加深病理对神经通讯的影响。最后，我们将网络损伤与由 SUVR 确定的局部代谢改变进行了比较，SUVR 是临床实践中经过验证的诊断工具。

 结果

我们的框架揭示了神经胶质瘤如何导致脑网络地形的广泛改变，即使在病变区域以外的结构未受影响的区域，对侧同源区域之间的连接显着减少。高级别胶质瘤对脑网络的影响更强，水肿在整体代谢改变中起中介作用。与传统的基于 SUVR 的分析相比，我们的方法提供了更全面的个体患者疾病负担视图，为神经胶质瘤相关改变提供了有趣的额外见解。

 结论

考虑到我们的结果，个体 PET 连接估计可能具有重要的临床价值，可能允许识别新的预后因素和神经胶质瘤或其他局灶性病变的个性化治疗。