Computer-aided assessment of DCE MR images predicts breast cancer survival

Volume and perfusion changes of breast cancers before and after neoadjuvant chemotherapy (NAC) assessed with computer-aided evaluation of dynamic contrast agent-enhanced (DCE) MR images offer practical imaging parameters associated with survival, according to a study published in the September issue of Radiology.

The study was the first to use commercialized evaluation software in the NAC setting of breast cancer patients, according to Ann Yi, MD, of Seoul National University College of Medicine, Seoul, South Korea, and colleagues.

To validate whether computer-assessed DCE MR parameters are associated with recurrence-free and overall survival in breast cancer patients who received NAC, Yi and colleagues retrospectively reviewed data from 187 consecutive women who had undergone NAC between January 2007 and December 2009. All patients had DCE MR imaging before and after NAC and surgery. A computer-aided evaluation program measured tumor size, volume and kinetic parameters, such as washout component.

There were 50 events at a mean follow-up of 47.4 months. This included 12 deaths and 38 recurrences—29 distant, six local and three both distant and local. Multivariate analysis revealed that a smaller reduction in tumor volume and a smaller reduction in washout component after NAC were independent significant variables for worse recurrence-free survival and overall survival, according to the authors.

Yi and colleagues used a 91.3 percent washout component reduction as a cutoff point for worse overall survival. Patients with tumors that had less than 91.3 percent reduction in the washout component had a three-year overall survival probability of 84.8 percent compared with 95.7 percent for those with a greater reduction in washout component.

“Thus, when a breast cancer patient showed a smaller reduction (<91.3%) in the washout component within a tumor at DCE MR imaging after NAC, targeted therapy for angiogenesis may be considered,” wrote the authors.

Yi and colleagues added that the differences in survival with the DCE MR imaging parameters used in the study persisted after adjusting for multiple clinical-pathologic factors. These included menopausal status, histologic grade, clinical and pathologic response and pathologic tumor stage. The study also adds to previous research that has shown volumetric measurement was superior to diameter measurement or clinical assessment when predicting response to NAC.

“On the basis of our observation, DCE MR imaging with computer-aided evaluation has the potential to serve as an additional tool for the risk stratification of patients with breast cancers who receive NAC,” wrote Yi and colleagues.