A new diagnostic development in MRI screening for breast cancer

The worldwide incidence of carcinoma of the breast is higher than other
malignancies among women.
Breast cancer screening is one of the possibilities to reduce breast cancer
mortality.While mammography
is the standard breast cancer screening exam, screening with MRI is more
effective for genetic high risk
patients or for women with dense breast tissue. MRI is able to detect early
stage breast malignancy
which is occult to clinical examination and conventional imaging. The overall
sensitivity of MRI usually
exceeds 90%. The overall specificity of MRI in discriminating malignant from
benign has been variable
ranging from 37% to 97%, resulting in a considerable number of benign biopsy
outcomes. Therefore,
it is very important to develop new techniques to increase the specificity of
MRI aiming at a reduction of
the number of invasive diagnostic procedures on MRI. A promising approach to
clarify the precise nature
of a lesion is MR spectroscopy.MR spectroscopy can demonstrate metabolic
differences between tumors
and normal tissues. The diagnostic value of 1H MR spectroscopy is typically
based on the detection of
elevated levels of choline (Cho) compounds, which are a marker of an active
malignant tumor. In prior
studies of single-voxel 1H MR spectroscopy performed with 1.5 T MR units
specificity of 85% (range,
67%-100%) has been reported. The single-voxel technique has limitations in
terms of lesion coverage,
which may effect the sensitivity of Cho detection due tumor heterogeneity.
Multiple-voxel technique or
chemical-shift imaging (CSI) may be used to acquire spectroscopic information
from multiple voxels over
a large volume of interest in single measurement, and hence is suitable for
analyzing the regional
distribution of tumor metabolites. Although it is commonly used in the brain
and prostate, only two
studies with breast lesions have been reported. Beak et al. showed that the
metabolic information
measured by CSI may be used for improving specificity in the diagnosis of
breast cancer in 36 patients.
Also CSI may provide advantages over the single-voxel technique for
characterizations of breast lesions.
Further investigations in larger studies are needed.

Primary objective:To compare the specificity of multiple voxel MR spectroscopy
with MRI.
Secondary objective: To decrease the number of false positive diagnostic
outcomes by
using multiple-voxel MR-spectroscopy and therefore to reduce benign biopsies.

Clinical, observational study: to assess whether multiple-voxel MR spectroscopy
increases the specificity
of the MRI. Pathology is used as gold standard. Thirty patients with a solid
breast lesion >= 1cm and
BIRADS classification 3 of 4 on the mammogram will undergo MRI and MR
spectroscopy (in one setting).
Appoximately half of the lesions will be malignant if fifteen breast lesions
with BIRADS 3 and fifteen breast
lesions with BIRADS 4 will be included. This is based on the outcome of all
breast MR scans which have been
made in UMCG in 2006.
The patients will undergo a standard MRI with contrast agent (gadolinium),
which takes 25 minutes.
MR spectroscopy will be performed after the breast lesion is detected by MRI,
which will also take
15 minutes. The acquired raw MR spectroscopy data will be transferred to a
computer workstation
and elevated levels of choline compounds will be measured.

Until now, no hazardous effects of the MRI are documented. MRI is
contra-indicated in patients with
non-MR compatible implant, disorder of the kidney function and claustrophobia.
The contrast agent
used with MR examinations, Gadoterate megluine, is considered very safe, and
has been used in
clinical practice for many years. In some cases nausea (1-2%) or hives (<1%)
can occur. Very rarely an
anaphylactoid reaction can occur. The burden for the patient is one extra visit
to the hospital for the
MRI scan and the possibility of finding extra suspect lesions with MRI.