Constance D. Lehman MD, PhD, Presenter: Instructor, General Electric Company

Helga Marques MS, Abstract Co-Author: Nothing to Disclose

Wanda Knight Bernreuter MD, Abstract Co-Author: Nothing to Disclose

Etta D. Pisano MD, Abstract Co-Author: Institutional research agreement, General Electric Company Institutional contract, Konica Minolta Group Institutional contract, VuCOMP, Inc Institutional contract, Real Imaging Institutional contract, Sectra AB

Mark Alan Rosen MD, PhD, Abstract Co-Author: Nothing to Disclose

Nola Michiko Hylton PhD, Abstract Co-Author: Research collaboration, Sentinelle Medical, Inc

To determine the most accurate method to assess residual disease post chemotherapy, prior to surgery in patients with locally advanced breast cancer. ACRIN 6657 is a multi-center study of MRI for measurement of breast tumor response to neoadjuvant chemotherapy.

This IRB approved study enrolled women between May 2002 and March 2006 who had 3cm or greater invasive breast cancer receiving anthracycline-cyclophosphamide neoadjuvant chemotherapy followed by a taxane. Contrast-enhanced MRI was performed after neoadjuvant therapy completion but prior to surgery. Size measurements of residual disease included clinical (C), mammographic longest diameter (MGLD), MRI longest diameter (MRLD) and MRI volume (MRV). Linear dimension was measured by the radiologist for MGLD and MRLD; MRV was calculated by computer using signal enhancement ratio thresholds. Pathologic residual disease size (PS) was measured by the pathologist. Lesion type was characterized by mass versus regional and single versus multiple. Pearson correlation coefficients were used to assess correlations between PS and pre-surgery tumor measurements.

237 patients were enrolled at 9 institutions. 216 patients with complete imaging formed the analysis set. Overall, PS was most highly correlated with MRLD (r=0.40, p<0.0001), followed by MRV and C (r=0.37, p=<0.0001 for both), and not significantly correlated with MGLD (r=0.14, ns). For single masses (N=127), PS was most strongly correlated with MRV (r=0.39, p=0.0013) and C was slightly correlated (r=0.27, p=0.0238). For single regional enhancements, PS was most highly correlated with MRLD (r=0.50, p=0.0028) followed by MRV (r=0.40, p=0.0139), and C (r=0.38, p=0.0178). MGLD was not significantly correlated with PS for single masses (r=0.22, ns) or for regional enhancements (r=-0.02, ns).

Across all lesions, longest diameter by MRI was the best correlate to PS, while MRI volume was the best correlate to PS for the subgroup of single masses. Mammography diameter did not correlate with PS overall, or for subgroups of single mass lesions or single regional enhancements. Work funded by NIH/ACRIN Grant U01CA79778 UO1, CA080098, CALGB Grants CA031946, CA033601 and BCRF.  

MRI is the best method to assess tumor size at surgery after chemotherapy. Mammography size does not correlate with true residual disease in patients treated with neoadjuvant chemotherapy.

Lehman, C, Marques, H, Bernreuter, W, Pisano, E, Rosen, M, Hylton, N, Assessing Residual Disease in Breast Cancer Patients Post Neoadjuvant Chemotherapy prior to Surgery: Findings of the American College of Radiology Imaging Network (ACRIN) Trial 6657.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8009188.html