Edward
A. Sickles,
MD
The Subtle of Invasive
I
and Atypical Mammographic Lobular Carcinoma’
lobular carcinoma (ILC) of breast is a relatively uncommon tumor that characteristically spreads by diffuse infiltration of single rows of malignant cells, in a manner that does not destroy underlying anatomic structures or incite substantial connective NVASIVE
the
tissue stages
reaction. As a result, in its early of development, ILC often does
not form a discrete mass, thereby escaping detection at both physical exarnination and mammography. The rarity of calcification within ILC makes “early” mamrnographic detection even more difficult. The study by Hilleren et a! in this issue of Radiology is a carefully compiled, well-documented, and largescale report on the mammographic features with which ILC eventually does become visible (1).
The
findings
of this
study
strongly
suggest that the insidious manner by which ILC spreads probably accounts for its unusual and often subtle mammographic findings. Unlike most breast cancers, which have a similar appearance regardless of mammographic projection, fewer than half of ILC cases demonstrate a principal mamrnographic abnormality on all three standard views. In almost 20% of ILC cases an abnormality can be found on only one of
these
views,
resulting
in an incomplete
imaging evaluation that does not even permit accurate lesion localization. Another atypical feature of ILC is that it is almost twice as likely to be seen on the craniocaudal view than either the oblique or lateral projections, which is different from what has been reported previously (2). In addition, at least half of ILC cases identified at mammogra-
phy
appear
opaque whereas
teristically
less opaque
than
as normal fibroglandular other invasive cancers
appear
to be more
or equally tissue, charac-
opaque
than equal volumes of parenchyma (3). Finally, an unusually high number (32%) of ILC cases identified at mammography can be detected only by the presence of indirect radiographic signs of malignancy (architectural distortion, asymmetric opacity, etc), and these relatively nonspecific findings are even more likely than other features of ILC to be seen on just one projection and to display less opacity than expected. The great majority of patients with ILC that demonstrate atypical mammographic features undergo biopsy because concurrent findings at physical examination are suspicious for malignancy. However, the presentation of ILC at physical examination can be atypical as well. In almost one-fourth of the cases reported by Hilleren et al,
the only
palpable
abnormality
was
an
area of thickening interpreted as representing “fibrocystic disease.” For these cases there had to be sufficient mammographic suspicion for malignancy to
prompt biopsy of the area of thickening because physical findings were too
Index
terms:
00.327
Breast neoplasms,
#{149} Breast,
parenchymal
diagnosis,
pattern,
00.92.
Editorials
Radiology
1991;
‘From 0628, cine,
and
the Department
University San Francisco,
accepted
requests ©
178:25-26
October
in
22, 1990.
Box of Medi-
Received Address reprint
to the author.
RSNA,
1991
See also the article 154)
of Radiology,
of California School CA 94143-0628.
this
issue.
by Hilleren
et at (pp 149-
nonspecific. Clearly, at a somewhat earlier stage in the development of ILC, some (perhaps many) cases will display subtle and atypical features at both physical examination and mammography, confounding timely detection and diagnosis. This difficulty in early diagnosis may explain my own anecdotal observation that ILC accounts for a disproportionately large percentage of poten-
tial malpractice
suits
investigated
for
failure to diagnose breast cancer. Approximately one-third of the cases sent to me by lawyers for review involve ILC, whereas only about 10% of breast cancers are ILC. In most of these mal-
Features
practice review cases, nonspecific palpable and mammographic findings have preceded the eventual diagnosis of cancer by several months to years; however, none of the mammographic lesions were sufficiently suspicious for malignancy to have prompted a tissue diagnosis.
Hilleren et al also cases display a much
indicate higher
that than
ILC usual
frequency of mammographic signs of locally advanced malignancy (skin retraction and thickening, nipple retraction, etc). This observation, coupled with the tendency of ILC to be multifocal and to demonstrate palpable findings at the time of mammographic de-
tection, nosis
such
suggests for
a relatively
patients
with
a conclusion
poor
ILC.
prog-
However,
is not supported
by
the major prognostic indicators reported in the study: The mean and median diameters of ILC tumors and the frequency of axillary metastasis were no different than those for other invasive breast cancers diagnosed during the study period. These findings are
both
unexpected
and
unexplained.
It
would be interesting to compare survival rates for ILC patients with rates for patients with all invasive breast cancers; these statistics should be available now because study cases have been followed for 5-15 years after diagnosis.
Finally,
the often
subtle
and
atypical
mammographic appearance of ILC raises an important question about our established interpretive criteria for the
diagnosis
of breast
cancer.
ILC more readily, should these criteria to consider
for malignancy parent
lesions
on oblique
and/or
To identify we expand as suspicious
not readily lateral
approjec-
tions or lesions less opaque than normal fibroglandular tissue? In my opinion, such an approach would be counterproductive for several reasons: (a) ILC missed on the basis of current interpretive criteria amounts to only 1.6% of all breast cancers (10% of breast cancer cases are ILC, and 16% of ILC is
either
mammographically
terpreted missed means
as benign). cancers are of abnormal
examination,
with
occult
or in-
(b) These few being detected by findings at physical
major
prognostic
in-
25
dicators that are similar to those of all other invasive breast cancers. (c) Revision of established interpretive criteria, especially to include lesions as subtle as those found in ILC, would substantially increase the number of false-positive mammographic readings. This revi-
sion,
in turn,
would
result
morbidity and costs, thereby ing two factors that already cant barriers to the widespread mammography.
26
#{149} Radiology
in increased worsenare signifiuse of
The most important conclusions to draw from the study of Hilleren et a! are those proposed by the authors themselves: the need for an increased awareness that ILC can be very difficult to detect mammographically, the need to correlate subtle and nonspecific mammographic findings with those of physical examination, and the need to proceed with the investigation of suspicious physical findings even in the absence of mammographic abnormalities. U
References 1.
Hitleren ell F. year 1991;
2.
3.
DJ, Andersson Invasive lobular spectrum
I, Lindholm carcinoma:
at mammography.
K, Lina 10Radiology
178:149-154.
Andersson I, Hildelt J, Muhtow A, Pettersson H. Number of projections in mammography: influence on detection of breast disease. AJR 1978; 130:349-351. Sickles EA. Breast masses: mammographic evaluation. Radiology 1989; 173:297-303.
January
1991