Prediction of Therapeutic Effect of Chemotherapy for NSCLC Using Dual-Input Perfusion CT Analysis: Comparison among Bevacizumab Treatment, Two-Agent Platinum-based Therapy without Bevacizumab, and Other Non-Bevacizumab Treatment Groups.

ABSTRACT

Purpose To determine whether dual-input perfusion computed tomography (CT) can predict therapeutic response and prognosis in patients who underwent chemotherapy for non-small cell lung cancer (NSCLC). Materials and Methods The institutional review board approved this study and informed consent was obtained. Sixty-six patients with stage III or IV NSCLC (42 men, 24 women; mean age, 63.4 years) who underwent chemotherapy were enrolled. Patients were separated into three groups those who received chemotherapy with bevacizumab (BV) (n = 20), those who received two-agent platinum-based therapy without BV (n = 25), and those who received other non-BV treatment (n = 21). Before treatment, pulmonary artery perfusion (PAP) and bronchial artery perfusion (BAP) of the tumors were calculated. Predictors of tumor reduction after two courses of chemotherapy and prognosis were identified by using univariate and multivariate analyses. Covariates included were age, sex, patient's performance status, baseline maximum diameter of the tumor, clinical stage, pretreatment PAP, and pretreatment BAP. For multivariate analyses, multiple linear regression analysis for tumor reduction rate and Cox proportional hazards model for prognosis were performed, respectively. Results Pretreatment BAP was independently correlated with tumor reduction rate after two courses of chemotherapy in the BV treatment group (P = .006). Pretreatment BAP was significantly associated with a highly cumulative risk of death (P = .006) and disease progression after chemotherapy (P = .015) in the BV treatment group. Pretreatment PAP and clinical parameters were not significant predictors of therapeutic effect or prognosis in three treatment groups. Conclusion Pretreatment BAP derived from dual-input perfusion CT seems to be a promising tool to help predict responses to chemotherapy with BV in patients with NSCLC. © RSNA, 2017.