Anatobreast : First Step to Virtual Breast Cancer Localization

Breast cancer is the most commonly cancer diagnosed among women, with approximately 42000 women diagnosed with breast cancer annually in France. Incidence rates of breast cancer rose steadily from 1980 to 2005 due to the increasing use of screening mammography and longer life expectancy.

With advances in breast cancer screening techniques, many women now have malignant tumors detected before they become clinically palpable. For these women, breast conserving surgery (BCS) has become the preferred treatment option, comprising more than 75 % of surgeries in most breast cancer centers. The goal of BCS is to resect the tumor with a surrounding margin of tissue free of cancer while simultaneously preserving the overall shape and appearance of the breast. For non palpable lesions, localization of the tumor prior to surgery is needed. The standard technique for such patients is wire localization. Wire localization suffers from several limitations. It adds another procedure prior to surgery, complicating and lengthening the process. It can be uncomfortable for the patient and entails additional cost. This technique requires the surgeon to estimate the 3-dimensional (3D) position of the cancer from 2-dimensional mammography images. Wire localized excision results in positive margins approximately 30-50 % of the time. These cases require additional surgery to remove the remaining cancer, which is both costly and emotionally difficult for patients.

Alternatives to wire localization have been tested. Radioguided occult lesion localization (ROLL) has been shown to be superior to wire localization with shorter risk having positive resection margins and shorter duration of localization and surgical excision.

Magnetic resonance imaging of the breast has been shown to be more sensitive than ultrasound or mammography for the detection of cancer. Despite this evidence, the use of MRI has not been demonstrated to produce better short-term outcomes: the prospective randomized COMICE trial failed to demonstrate a decrease in the positive margin rate and need for re-excision with the use of preoperative prone MRI. The COMICE study authors recognized that the inability of prone MRI to result in decreased positive margin rates might be due to the way the imaging data is presented to the surgeon, stating that ''techniques to ensure surgical precision is at an optimum need further examination.'' The shape of the breast during prone MR imaging (with the breast pendant in a bilateral imaging coil) is radically different from the supine, arm-extended position in the operating room (OR). Supine MRI more accurately replicates the surgical position. Indeed, a study by Sakakibara et al. randomized patients with DCIS to undergo wire localized or supine MRI- guided BCS and found the positive margin rate and resected tissue volume were lower in patients undergoing supine MRI-guided excision.

The aim is to develop a novel technique of MRI- guided breast tumor localization that incorporates preoperative prone MRI and uses a preoperative and intraoperative 3D optical scans to adjust the MRI image to the exact position in the operating room. This novel technique will provide the surgeons with 3D views and measurements of tumor shape and position within the breast as it appears during surgery.