Perfusion parameters predict pathology and mutation status in lung cancer brain metastases based on CT imaging: a pilot study

Abstract

Abstract Objectives To define the pathological subtype and mutational status of metastatic disease in patients with non-small cell lung cancer (NSCLC) using CT imaging-based perfusion parameters analysis of brain metastatic lesions. Methods We retrospectively identified NSCLC patients with brain metastases who got treatment in our hospital during 2019 to 2021 and had their underlying lung cancer pathologically and genotypically tested. CT perfusion images of the brain were utilized to segment enhancing tumors and peritumoral edema, as well as extract CT perfusion parameters. The most relevant perfusion parameters were identified to classify the pathological subtype and mutation status. Squamous cell carcinoma was found in 15 of the 45 patients in the research cohort (mean age 66.29 ± 9.98 years; M: F = 26:19), while adenocarcinoma was found in 30. In addition, 19 had an epidermal growth factor receptor (EGFR) mutation, and 11 had an EGFR wild-type. After admission, all patients were subjected to a CT perfusion imaging examination. The differences in CT perfusion parameters between different pathological subtypes and mutational status were analyzed. The receiver-operating characteristic (ROC) curve was used to predict the pathological subtype and mutational status of the metastasis. Results Among the quantitative parameters, CBF and MTT were significantly different between the two lung cancers, as well as the two mutational statuses (adenocarcinoma vs. squamous cell carcinoma: P < 0.001, P < 0.001.EGFR mutation vs. EGFR wild-type: P < 0.016, P < 0.046.). For classification of pathological subtype, EGFR mutation status, the model developed with both CBF and MTT resulted in area-under-the-curve (AUC) values of 0.849 and 0.790, respectively. Conclusions Perfusion parameters analysis of brain metastases using CT imaging in patients with primary lung cancer could be used to classify pathological subtype and mutational status. This method might be beneficial for developing treatment plans and determining prognosis.