• Disease-free three-year survival [ Designated as safety issue: No ]
  

• Extent of residual disease [ Designated as safety issue: No ]
  
• Change in the maximum dimension of the tumor over time [ Designated as safety issue: No ]
  
• Change in the tumor volume over time [ Designated as safety issue: No ]
  
• Maximum dimension of tumor size measure by MRI, mammography, and pathology [ Designated as safety issue: No ]
  
• MRI volume [ Designated as safety issue: No ]
  
• MRI peak signal enhancement ratio (SER) [ Designated as safety issue: No ]
  
• SER distribution (percent of tumor in highest SER category) [ Designated as safety issue: No ]
  
• Morphological pattern [ Designated as safety issue: No ]
  
• Change in tumor size by clinical exam [ Designated as safety issue: No ]
  

OBJECTIVES:

• Identify surrogate markers of response to neoadjuvant chemotherapy by contrast-enhanced magnetic resonance imaging (MRI) that are predictive of pathologic remissions and survival in women with stage III breast cancer.
• Identify two groups of patients who have statistically different 3-year disease-free survival using MRI measurements of tumor response to neoadjuvant chemotherapy.
• Determine whether MRI measurements of tumor response after the first course of neoadjuvant chemotherapy can predict which of these patients will ultimately have poor clinical response to chemotherapy.
• Compare the accuracy of MRI vs mammography in predicting the extent of residual disease as determined by histopathology in these patients.
• Determine whether initial MRI tumor characteristics (morphologic and vascular patterns) predict pathological response and/or survival in these patients.
• Estimate the conditional probability of response to paclitaxel based on MRI measurements of response to doxorubicin and cyclophosphamide in these patients.

OUTLINE: This is a multicenter study.

Patients receive an injection of gadopentetate dimeglumine and undergo magnetic resonance imaging (MRI) and magnetic resonance spectroscopy of the breast within 4 weeks before beginning neoadjuvant chemotherapy, 20-28 hours or 48-96 hours after the first course of doxorubicin and cyclophosphamide (Type 1 chemotherapy), between Type 1 chemotherapy and paclitaxel chemotherapy regimens (Type 2 chemotherapy) (MRI only) if the patient continues to Type 2 chemotherapy, and 3-4 weeks after final neoadjuvant chemotherapy treatment (1-2 weeks before surgery).

Patients also undergo mammograms and possibly ultrasounds that coincide with the first and last MRI. Core or needle biopsy is performed after the first MRI but before the first course of Type 1 chemotherapy and between Type 1 chemotherapy and Type 2 chemotherapy (if the patient continues to Type 2 chemotherapy).

Patients are followed every 6 months for 7-10 years.

PROJECTED ACCRUAL: A total of 244 patients will be accrued for this study within 3 years.