LAS VEGAS, NV -- April 19, 2005 -- Magnetic resonance imaging (MRI) alone may not be accurate in predicting a complete therapeutic response to chemotherapy in breast cancer patients, a Stanford University study shows.

"We still believe that MRI is sensitive on invasive disease and can assist clinical examination, mammography and ultrasound," said Dr. Kathleen C. Horst, clinical instructor, Department of Radiation Oncology, Stanford University. "But it's not advisable to just go by MRI alone."

Dr. Horst said that this may be especially true in patients with invasive lobular cancer where MRI is known to have a higher rate of false negative findings.

She presented the study results here on April 18^th at the American Society of Breast Disease annual meeting.

The study analyzed 32 women, median age 49 years, who had been diagnosed as having clinical Stage IIB – IV breast cancer based on clinical examination, mammography and ultrasound.

The women underwent MRI before and after neoadjuvant chemotherapy to treat early stage or locally advanced breast tumors. MRI findings were reviewed retrospectively and compared with final pathology after surgical resection.

The women received neoadjuvant adriamycin-based chemotherapy during a 10-year period ending in 2002. The chemotherapy was given to downstage invasive ductal carcinoma or invasive lobular carcinoma.

At initial presentation the women's tumors were clinically staged and about half had T3 tumors and seven had T2 and T3 tumors, respectively.

Two women (6%) were determined to have a complete response by clinical breast examination, mammography, ultrasound and MRI. After whole-breast radiotherapy with no surgical staging, the women remained free of disease at a median follow up of 53 months. "They had a good response with no evidence of disease," Dr. Horst said.

The remaining 30 women had surgical staging after it was determined that they did not achieve a complete response to chemotherapy, based on clinical examination, mammography, ultrasound and MRI.

Two women (6%) with a complete response by MRI alone had false negative readings with residual invasive lobular carcinoma on final pathology. Of the two women with non complete responses by MRI, two were false positive readings with pathology demonstrating no residual invasive tumor.

MRI sensitivity was 92%, specificity was 67%, positive predictive value was 92% and negative predictive value was 67%.

The findings suggest that MRI findings alone may not be an accurate prediction of complete responses after neoadjuvant chemotherapy, particularly in cases of invasive lobular carcinoma, Dr. Horst said.

"Normal MRI findings after neoadjuvant chemotherapy should not preclude diagnostic surgical intervention in the setting of suspicious clinical examination or imaging by mammography or ultrasound."