Cancer Imaging (Mar 2023)

Early evaluation of liver metastasis using spectral CT to predict outcome in patients with colorectal cancer treated with FOLFOXIRI and bevacizumab

  • Shenglin Li,
  • Long Yuan,
  • Mengying Yue,
  • Yuan Xu,
  • Suwei Liu,
  • Feng Wang,
  • Xiaoqin Liu,
  • Fengyan Wang,
  • Juan Deng,
  • Qiu Sun,
  • Xianwang Liu,
  • Caiqiang Xue,
  • Ting Lu,
  • Wenjuan Zhang,
  • Junlin Zhou

DOI
https://doi.org/10.1186/s40644-023-00547-w
Journal volume & issue
Vol. 23, no. 1
pp. 1 – 15

Abstract

Read online

Abstract Purpose Early evaluation of the efficacy of first-line chemotherapy combined with bevacizumab in patients with colorectal cancer liver metastasis (CRLM) remains challenging. This study used 2-month post-chemotherapy spectral computed tomography (CT) to predict the overall survival (OS) and response of CRLM patients with bevacizumab-containing therapy. Method This retrospective analysis was performed in 104 patients with pathologically confirmed CRLM between April 2017 and October 2021. Patients were treated with 5-fluorouracil, leucovorin, oxaliplatin or irinotecan with bevacizumab. Portal venous phase spectral CT was performed on the target liver lesion within 2 months of commencing chemotherapy to demonstrate the iodine concentration (IoD) of the target liver lesion. The patients were classified as responders (R +) or non-responders (R −) according to the Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 at 6 months. Multivariate analysis was performed to determine the relationships of the spectral CT parameters, tumor markers, morphology of target lesions with OS and response. The differences in portal venous phase spectral CT parameters between the R + and R − groups were analyzed. Receiver operating characteristic (ROC) curves were used to evaluate the predictive power of spectral CT parameters. Results Of the 104 patients (mean age ± standard deviation: 57.73 years ± 12.56; 60 men) evaluated, 28 (26.9%) were classified as R + . Cox multivariate analysis identified the iodine concentration (hazard ratio [HR]: 1.238; 95% confidence interval [95% CI]: 1.089–1.408; P < 0.001), baseline tumor longest diameter (BLD) (HR: 1.022; 95% CI: 1.005–1.038, P = 0.010), higher baseline CEA (HR: 1.670; 95% CI: 1.016–2.745, P = 0.043), K-RAS mutation (HR: 2.027; 95% CI: 1.192–3.449; P = 0.009), and metachronous liver metastasis (HR: 1.877; 95% CI: 1.179–2.988; P = 0.008) as independent risk factors for patient OS. Logistic multivariate analysis identified the IoD (Odds Ratio [OR]: 2.243; 95% CI: 1.405–4.098; P = 0.002) and clinical N stage of the primary tumor (OR: 4.998; 95% CI: 1.210–25.345; P = 0.035) as independent predictor of R + . Using IoD cutoff values of 4.75 (100ug/cm3) the area under the ROC curve was 0.916, sensitivity and specificity were 80.3% and 96.4%, respectively. Conclusions Spectral CT IoD can predict the OS and response of patients with CRLM after 2 months of treatment with bevacizumab-containing therapy.

Keywords