Ductal Carcinoma
In Situ
373
FIG. 11.55.
DCIS, estrogen receptor.
A,B:
Strong and diffuse nuclear immunoreactivity for ER
(B)
in
solid DCIS with intermediate-grade nuclei and central necrosis. Nuclear reactivity has been lost in
the dying cells around the necrotic center.
A
B
Samples of ER-negative comedo DCIS had a high prolifera-
tive rate prior to transplantation, and this was maintained
but did not increase when xenografts were exposed to es-
trogen. On the other hand, noncomedo, ER-positive DCIS
exhibited increased proliferation after exposure to estrogens,
although the proliferative levels did not reach those of the
comedo DCIS. These results suggest that ER-negative DCIS
is estrogen independent and that antiestrogen therapy may
not be beneficial for patients with these lesions.
Hormone receptor expression in DCIS may be a signifi-
cant factor for local recurrence after breast conservation
treatment. Roka et al.
181
reported that DCIS with high nuclear
grade and absence of ER had a significantly higher recurrence
rate. However, not all high-grade DCIS lack ER expression.
Collins and Schnitt
182
found that 14 (12%) of 114 ER-positive
DCIS had high nuclear grade and overexpressed HER2 pro-
tein. Concurrent ER and HER2 expression was not found in
any of the 74 examples of low- and intermediate-grade DCIS,
all of which expressed ER and lacked HER2 overexpression.
In general, approximately 75% of DCIS cases show posi-
tive nuclear immunoreactivity for ER (ranging from 1% to
100% cells), and a somewhat lower frequency are positive for
PR. However, determination of “positivity” of ER and PR in
DCIS can be problematic.
183
There are three main scoring
methods for ER and PR immunoreactivity. One is the
simple
percentage method
, that is, simple proportion of cells staining
without regard to intensity. A second is the
histochemical “H”
method
, that is, the sum of the percentage of cells with three
degrees of staining of cells resulting in a range of 0 to 300. The
third,
Allred method
, produces an intensity score between 1
and 3, and a percentage of cells staining compartmentalized
in a nonlinear manner using scores between 1 and 5, to yield
a final score of between 2 and 8. The Allred and “H” scor-
ing methodologies include assessment of staining intensity,
whereas the simple percentage method does not. Optimally,
a scoring method utilizing both proportion and intensity of
ER staining should be used for DCIS, if only because there is
evidence for the value of such assessment in invasive breast
cancer. Excellent interobserver agreement has been reported
in the use of “H” score for the assessment of ER and PR in in-
vasive breast carcinoma since it provides a “continuous mea-
sure of tumor hormone-receptor content” as opposed to the
Allred system that has a “limited dynamic range.”
184
The UK breast-screening guidelines recommend the use
of less than 5% cells staining for DCIS and the use of the
Allred score for invasive cancers.
185
A cutoff point of Allred
score 3 for positivity has been used in a study of various phe-
notypes of DCIS. Baqai and Shousha
186
reviewed 56 cases
of pure DCIS. ER positivity was defined as more than 10%
of cells showing dark brown nuclear staining (Allred score
greater than 4, “H” score of greater than 20). Eighty-eight
percent of high-grade DCIS were ER negative. Data from the
Sloane Project (named after the late Professor John Sloane
of the Royal Liverpool Infirmaray who had a special inter-
est in the pathology of DCIS) show that intermediate and
low-grade DCIS are almost invariably ER positive, whereas
high-grade DCIS is positive in 69% of cases.
185
Oncogenes
HER2:
Immunohistochemical studies have demonstrated
membrane immunoreactivity for the HER2 (HER2) onco-
gene in 42% to 61% of DCIS.
187–191
(Fig. 11.56).
HER2
gene
amplification has been reported in 40% to 48% of DCIS
specimens isolated by microdissection and studied by the
polymerase chain reaction (PCR).
192,193
Ho et al.
193
found
significantly higher frequencies of HER2 amplification in
comedo than in noncomedo DCIS (69% vs. 18%) and in le-
sions with high rather than low nuclear grade (63% vs. 14%).
Expression of HER2 occurs in 85% to 100% of comedocar-
cinomas and is associated with the pleomorphic nuclear
cytology in these lesions.
187,191,194
Most investigators have not detected HER2 in small
cell micropapillary and cribriform DCIS.
187,191,195
Using
