Breast cancer and non-malignant breast data: Quantification by FPT - Signal Processing in MRS with Biomedical Applications - page 305

chemotherapy. MRI is more effective than mammography in finding breast
cancer at an earlier stage among women at high risk [354, 366]. Yet, false neg
ative findings using MRI have been reported. This can occur, e.g., with small
tumors, if they do not selectively take up the contrast agent. Moreover, MRI
cannot reveal microcalcifications [370] and occasionally misses invasive ductal
and lobular carcinomas, although nondetection with MRI is more common
with in situ ductal carcinoma [371]. Recent data suggest that MRI is more
effective for women at high family risk of breast cancer compared to those
with low family risk [372]. Based upon the data from an ongoing surveillance
trial centered in Bonn, Germany, recommendations have been made to discon
tinue mammographic screening in young women with BRCA1 mutations and
instead to use MRI, in view of the vulnerability to ionizing radiation together
with the fact that in the mentioned series, none of the BRCA1associated
breast cancers were found to be calcified [373].
The major problem with MRI, however, is that despite excellent spatial res
olution and generally superior sensitivity, it has poor specificity, consistently
lower than mammography [354]. Thus, while women at high risk undergoing
intensive surveillance programs appear to be relieved by the greater sensitiv
ity of MRS [374], a large number of false positive findings may unfavorably
affect quality of life [22, 375]. Moreover, a missed cancer, albeit relatively less
likely with MRI than mammography, can adversely affect prognosis [354]. In
contradistinction to prophylactic mastectomy which while drastic, does effec
tively provide primary prevention for women with BRCA1 or BRCA2 or other
high risk gene mutations, a significantly lowered mortality has not yet been
demonstrated for surveillance with MRI among women at high breast cancer
risk [376]. Thus, concerns remain about MRI as a breast cancer screening
tool, with the need to improve specificity particularly underscored [363, 376].
10.2.2 Studies to date using in vivo MRS for distinguishing
between benign and malignant breast lesions
In vivo MRS has been shown to increase the specificity of MRI with respect
to the diagnosis of breast cancer [116, 117], [377]-[381]. Most of the studies
using in vivo proton MRS have been based upon the composite choline signal,
assessed either qualitatively or quantitatively. In a metaanalysis [116] of five
clinical studies using in vivo singlevoxel MRS to assess 100 cancerous and
53 nonmalignant breast lesions, specificity was computed to be 85% (95%
Confidence Intervals (CI) = 71 - 93%), and sensitivity was 83% (95% CI =
73% - 89%). In that metaanalysis [116] the diagnostic accuracy was higher for
the twenty women who were age 40 or younger. Among those twenty younger
women, all eleven breast cancers were accurately diagnosed. A subsequent
publication describing singlevoxel MRS among thirty women with a positive
mammogram and rapid contrast enhancement on breast MRI similarly showed
that the presence of choline (with SNR≥2) improved the specificity for breast
cancer to 87.5%, with three false positive results [382]. Another subsequent