Trial Information

Evaluation of the Diagnostic Performance of MRI±Biopsy to Optimize Resection of Ductal Carcinoma In Situ (DCIS) Breast Cancer

Ductal carcinoma in situ (DCIS) is a frequent disease, concerning more than 15% of all
breast carcinomas in France. Frequency increases due to breast screening programs. Breast
conservative treatment can be done for limited lesions. Complete pathological excision with
safe surgical margins is the main factor of success rate for conservative treatment. But
safe surgical margins are not always obtained in the first surgery because preoperative or
peroperative evaluation of the extent of the lesions is difficult. A 20 to 38% rate of
involved margins is reported in the literature, implying a second surgery. MRI is not
commonly performed today in the preoperative staging of DCIS, despite interesting results of
breast MRI in DCIS evaluation published recently in a German study (Kuhl, Lancet 2007).
Nevertheless, this retrospective study focused on a population of high-risk women (BRCA1-2
mutation carriers, family history of breast cancer…).

Additional suspicious lesions detected by MRI require MRI-guided or CT-guided biopsy as they
are not detectable with mammography or ultrasound. Moreover, intraoperative pathological
analysis cannot be performed in DCIS and surgical strategy will be different for a unique
lesion (lumpectomy) or multiple/large lesions (mastectomy). If surgery is adapted according
to MRI and biopsy results, we can assume that the entire lesion could be removed in a single
operation, avoiding re-excisions, anxiety and reducing the costs.

The main objective of this protocol is to evaluate the value of MRI+biopsy in the staging of
mammography-detected DCIS in an unselected population of women. It is a multicentric,
randomised study including patients from France presenting a unique, limited (< 30 mm) DCIS
detected by mammography or breast ultrasound. The two compared arms are: preoperative breast
MRI+biopsy versus no additional diagnostic examination. High resolution dynamic 3D images
will be obtained in order to detect very small suspicious lesions, typical of DCIS. The main
endpoint is the re-excision rate for involved margins. In order to show a reduction of 50%
of the rate of second surgery (from 25% to 12.5% in the MRI arm), it is necessary to include
360 patients (bilateral test, alpha=5%, power=80%).

A cost-effective study will be performed. The costs likely to differ between the two arms
(MRI, biopsy, outpatient visits, hospitalisations, transportation, and sick leaves) will be
assessed using data collected in the trial. The endpoint of the cost-effectiveness study
will be the cost by success (cost-effectiveness ratio). A success is defined as achieving
margin-negative resection after initial surgery (no need for re-excision). The other
secondary endpoints will be to determine the rate of initial mastectomy and to describe MRI
morphologic features in the MRI+biopsy arm.