JOURNAL OF PRACTICAL HEPATOLOGY ›› 2012, Vol. 15 ›› Issue (5): 421-423.doi: 10.3969/j.issn.1672-5069.2012.05.0016

Previous Articles     Next Articles

Application of apparent diffusion coefficient for predicting response of hepatic metastases to chemotherapy

Xu Yonghua, Xu Jingci, Yang Lixia, at al.   

  1. Department of Medical Imaging,Xuhui Central Hospital Affiliated to Shanghai Medical Center,China Academy of Sciences,Shanghai 200031
  • Received:2012-01-26 Online:2012-10-10 Published:2017-03-09

Abstract: Objective To investigate the feasibility of apparent diffusion coefficients(ADC) in diffusion weighted MR imaging in predicting response to chemotherapy in patients with hepatic metastatic tumors. Methods Forty-three consecutive patients with hepatic metastases from gastric and colorectal cancer,breast cancer and other tumors underwent MRI diffusion-weighted imaging(DWI) sequences before and 2 to 3 weeks following chemotherapy for assessing the ADC value of lesions. RECIST was utilized to evaluate the efficacy of the chemotherapy. Results A total of 68 lesions in the 43 patients were observed,and the sizes of hepatic metastases decreased in 30 ones(the respond group) and increased or did not change in 38 ones (the non-respond group) after chemotherapy. The mean ADC values of lesions were 1.22±0.39×10-3 mm2/s in the respond group and 1.54±0.55×10-3 mm2/s in the non-respond group(P=0.008),respectively;After treatment,the ADC values increased to 1.43×10-3 mm2/s(P=0.005) in the responders,whereas they decreased to 1.23×10-3 mm2/s(P=0.13) in the non-responders;We supposed ADC equal to 1.50×10-3 mm2/s as a threshold value for responders and non-responders,the positive predictive value,negative predictive value and accuracy were 94%,92% and 93%,respectively. Conclusion ADC value of diffusion weighted MR imaging was potential to be an imaging marker for predicting and assessing early chemotherapeutic responses of hepatic metastases.

Key words: Liver neoplasms, Chemotherapy, MR imaging, Apparent diffusion coefficient