Med Phys. 2025 Dec;52(12): e70147
Haiyan Wang,
Zilong Guan,
Jiaxiang Qu,
Xuetong Tao,
Yaping Wu,
Ziwei Liu,
Zixiang Chen,
Fanting Luo,
Na Zhang,
Yanhua Duan,
Zhaoping Cheng,
Dong Liang,
Hairong Zheng,
Meiyun Wang,
Greta S P Mok,
Zhanli Hu.
BACKGROUND: Lung adenocarcinoma (ADC) and squamous cell carcinoma (SCC) are major subtypes of lung cancer (LC) with distinct clinical features and pathological mechanisms, necessitating tailored treatments. However, early noninvasive differentiation remains challenging.
PURPOSE: With the development of total-body PET/CT, this study attempts to provide new insights into this challenge by analyzing and comparing inter-organ and inter-system metabolic connections across the total body in ADC and SCC patients.
METHODS: Based on static and dynamic total-body 18F-fluorodeoxyglucose (18F-FDG) PET/CT data from 12 ADC patients, 12 SCC patients, and 20 healthy controls (HCs), abnormal metabolic connection networks of ADC and SCC were constructed relative to HCs for individual and group analysis. An external patient cohort consisting of 7 ADC patients and 7 SCC patients was introduced to evaluate the generalizability of the analysis results by topological correlation. The metabolic correlations among organs/systems were observed by Pearson correlation coefficient between standardized uptake values or Ki values of 36 regions of interest (ROIs) across the total body. ROIs were also classified into 18 organs and 8 major systems. Specifically, for the individual analysis, the alteration in metabolic connections between the lungs (respiratory system), and other systems were investigated, as well as among the seven nonrespiratory systems, relative to HCs. For the group analysis, the abnormal metabolic connections across the total body of the ADC group and the SCC group were analyzed at the organ and system levels to explore the disease abnormal patterns of different LC subtypes.
RESULTS: At the individual level, ADC and SCC patients exhibited distinct patterns of abnormal metabolic connectivity between the lungs (respiratory system) and other systems, highlighting subtype-specific metabolic reorganization. Both subtypes consistently showed significant abnormalities in nervous-motor and nervous-digestive system connectivity among the seven nonrespiratory systems. At the group level, ADC patients showed significant metabolic connectivity abnormalities mainly between specific organs and systems, with most changes reflecting reduced connectivity as compared to HCs. In contrast, SCC patients demonstrated markedly enhanced connectivity across multiple organ and system pairs, consistent with their broader diffusion patterns in clinical practice. Notably, dynamic Ki-based analysis was more sensitive than static SUV-based analysis in capturing these subtype-specific systemic metabolic alterations. These findings were highly consistent across internal and external cohorts, as evidenced by the generally strong topological correlations (Pearson correlation coefficient > 0.9) at both the organ and system levels. Moreover, the raw SUV and Ki analyses of organs among HC, ADC, and SCC groups provided complementary evidence to the metabolic network findings.
CONCLUSIONS: The proposed abnormal metabolic networks are feasible to characterize the metabolic alterations between organs or systems in ADC and SCC patients based on static and dynamic total-body PET/CT relative to HCs. Dynamic Ki-based analysis sensitively captures the complex and heterogeneous metabolic network alterations in LC subtypes, with SCC showing more extensive systemic abnormalities than ADC. The findings have the potential to enhance understanding of LC's physiological mechanisms and inform precision medicine strategies for early diagnosis and tailored therapies.
Keywords: lung adenocarcinoma; lung squamous cell carcinoma; metabolic network analysis; total‐body PET/CT