Structural and functional connectivity coupling as an imaging marker for bone metastasis pain in lung cancer patients.

Background: Cancer pain is a common symptom in patients with malignant tumors and associated with poor prognosis and a high risk of death. Structural connectivity (SC) and functional connectivity (FC) couplings have not yet been explored in lung cancer patients with bone metastasis pain.

Methods: In total, 51 patients with lung cancer without bone metastasis pain (BMP-), 52 patients with lung cancer with bone metastasis pain (BMP+), and 28 healthy controls (HC) were prospectively enrolled in our study. Firstly, SC-FC couplings were measured and analyzed at global, regional, and modular levels. Subsequently, individualized SC-FC coupling networks were constructed based on the Euclidean distance metric. In addition, the convolutional neural network (CNN) model was selected to analyze and classify three groups based on individualized networks.

Results: The coupling analysis demonstrated that weaker SC-FC couplings related to lung cancer itself were present at various levels, including global, regional, inter-network, and intra-network couplings. Notably, hyper-couplings related to bone metastasis pain were present in several brain regions, mainly involving the default mode network, frontoparietal network, salience network, and limbic system. Significant positive correlations were observed between regional coupling in the right amygdala and the numeric rating scale scores in BMP+. Moreover, CNN model built on individualized networks exhibited relatively great classification performance.

Conclusion: Alterations in SC-FC coupling patterns may play a crucial role in the development and modulation of bone metastasis pain. Understanding these changes could provide valuable insights into the neural mechanisms underlying cancer pain.