ABSTRACT: While computerised tomography (CT) is used for diagnosis and assessing response to treatment little work has been performed on the prognostic significance of the CT findings in women with liver metastases. The aim of this study was to assess if the CT findings in women diagnosed with liver metastases at the time of first presentation with metastatic breast cancer have any prognostic significance.
The staging CT scans of 78 consecutive women diagnosed with liver metastases at the time of first presentation of metastatic breast cancer were reviewed independently by two radiologists who were blinded to survival and the histological features of the tumour. The number and enhancement characteristics of liver metastases, whether metastases were solitary, multiple or diffuse, the diameter of the largest and the sum of the diameter of the five largest lesions, an estimate of % involvement (<10%, 10-50%, >50%), and the presence of metastases at other sites were assessed. HER-2 and ER status and histological grade of the patient's primary breast cancer were also recorded. Survival was ascertained from hospital records. The prognostic significance of these factors was assessed in a univariate and multivariate fashion.
At univariate analysis, number of liver metastases, sum of the diameter of the five largest lesions, percentage estimated involvement, presence of ascites, chest metastases and HER-2 status were significantly associated with reduced survival. Liver metastasis pattern (i.e. whether discrete or multiple), enhancement characteristics, ER status and histological grade were not associated with a significant outcome. At multivariate analysis estimated percentage liver involvement and the presence of chest metastases retained prognostic significance. Estimated percentage involvement was reproducible with 90% concordance between the two observers.
The CT appearances of patients with liver metastases at first presentation with metastatic breast cancer provide prognostic information which may be clinically useful.
Breast (Edinburgh, Scotland) 05/2011; 20(5):455-9. · 2.09 Impact Factor