Novel technique grades breast cancer to predict its response to chemotherapy

March 5, 2005

CONTEXT: Duke University researchers have shown that a test assessing functional and morphologic information gathered during contrast-enhanced MRI renders a score that accurately predicts the response of nine of 10 breast cancer cases to neoadjuvant chemotherapy.

RESULTS: T1-weighted imaging was performed in 20 women with locally advanced breast cancer. Principal investigator Oana Craciunescu, Ph.D., an assistant clinical professor of radiation oncology at Duke, studied the characteristics of tumor perfusion, permeability, and morphology/cellularity. Highly vascular tumors with efficient vascular systems carried more tracer and chemotherapy than tumors fed by less efficient vessels. Homogeneous tumors with an even distribution of blood vessels responded best to therapy. Tumors with densely packed cancer cells did not effectively retain the tracer and responded poorly.

Tumors featuring an equatorial ring of blood vessels also tended to resist chemotherapy because of collapsed vessels.

IMAGE: Morphophysiological tumor scores based on a five-point scale were calculated for a responder (top) and a nonresponder (bottom) to chemotherapy performed before surgery. Enhancement curves revealed the tumor's washin (WiP), washout (WoP), and centripetal (CP) and centrifugal (CF) characteristics. WiP measured the mass's vascularity, permeability, and homogeneity. Tumors with an inhomogeneous pattern or ring enhancement were classified as CP. Tumors with a homogeneous pattern were classified as CF. WoP correlated with tumor cellularity.

IMPLICATIONS: Poor responders identified by this technique can be referred directly to surgery, thereby avoiding the discomfort of chemotherapy. The method will be tested at Duke in a larger trial of patients with locally advanced breast cancer.