Dynamic Breast MRI in Assessing Locally Advanced Breast Cancer

We will conduct DCE-MRI prior and after the completion of neoadjuvant chemotherapy to monitor response to therapy. This will be done in a pilot study of 5 patients enrolled at Sunnybrook Health Sciences Centre recruited through the project investigation. The patients will be scanned on the research General Electric, 1.5 Tesla imaging system (version 12). We will use specially designed coils which we have developed and optimized for breast MR imaging. These allow modest compression of the breast in a medial-lateral direction and serves to bring surface coils as close the breast as possible to both immobilize the breast and achieve maximum coil couple and image signal/noise.

The dynamic data needs to cover the entire breast, so we will use a fast T1 weighted, dynamic breast MRI sequence [18, 19]. This will allow us to collect dynamic MRI data to cover the tumour volume with a temporal resolution of at least 20 seconds per data set and adequate spatial resolution (~1mm) to ensure adequate definition of the tumour boundary. The pulse sequence will be a spoiled gradient recalled sequence (SPGR) with imaging parameters TR/TE=8.2ms and min-TE. The image will be collected with a field of view of 16-20 cm depending on the patient breast size. Imaging resolution will be approximately 1 mm in-plane and 4-5 mm in slice thickness. The specific sequence of the imaging study is as follows:

1. A set of localizer images to position the breast in the MRI system.
2. A series of 2D fast spin-echo T2 weighted images to cover the breast.
3. An SPGR imaging sequence will be used to measure the T1 distribution over the breast for subsequent pharmacokinetic analysis as described by Chen[20].
4. We will administer intravenous Gd-DTPA (0.1 mmol/kg) followed by a saline flush.
5. Immediate prior to the Gd injection, we will start a continuous application of the dynamic MRI sequence to the involved breast collecting data for 10 minutes post injection of Gd-DTPA. The imaging will precede the Gd injection by approximately 2 minutes to provide adequate imaging data to serve as the non contrast-enhanced baseline data.

3.2 - Data Analysis:

These images will be used to assess the tumour as follows:

1. The simplest measurement will be that of tumour size as determined by the region of enhancement seen between prior to contrast enhancement versus that seen throughout the 10 minutes period post injection.
2. We will calculate the T1 distribution throughout the breast based on the multi-angle T1 weighted SPGR sequence as per Cheng[20].
3. This data will then be used to calculate the [Gd] for each voxel in the image sets.
4. These will be used to estimate Ktrans and Ve from a two compartment tumour model[15].
5. We will calculate these parameters over a region of interest defined by the enhancing tumour border.
6. We will estimate these parameters on a pixel basis to study tumour heterogeneity over the tumour as characterized by the statistics of the parameters Ktrans and Ve.

The tumour volume measurements will be done by Dr. P. Causer of the Department of Medical Imaging of Sunnybrook Health Sciences Centre. The analysis of the dynamic data will be done by Dr. Plewes' group and Dr. A Martel of Imaging Research in conjunction with Dr. Causer. The final result, will be measures of the changes in tumour volume before and after neoadjuvant chemotherapy. In addition, we will have information regarding the distribution of Ktrans and Ve over the regions of the tumour and their variation before and after therapy.