1. Immunology Letters 156 (2013) 123–126
Contents lists available at ScienceDirect
Immunology Letters
journal homepage: www.elsevier.com/locate/immlet
CD8+ lymphocytes and apoptosis in typical and atypical medullary
carcinomas of the breast
Ika Nurlailaa
, Premasiri Upali Telisingheb
, Ranjan Ramasamya,∗
a
Institute of Health Sciences, Universiti Brunei Darussalam, Gadong, Brunei Darussalam
b
Department of Histopathology, Raja Isteri Pengiran Anak Saleha Hospital, Bandar Seri Begawan, Brunei Darussalam
a r t i c l e i n f o
Article history:
Received 4 August 2013
Received in revised form
20 September 2013
Accepted 1 October 2013
Available online xxx
Keywords:
Apoptosis
CD8+ cytotoxic lymphocytes
Medullary breast carcinoma
Tumour immunity
a b s t r a c t
Medullary breast carcinoma (MBC) is a form of ductal invasive carcinoma (DIC) characterized by an
abundant infiltration of the tumour by lymphocytes. MBC has been classified histologically into typical
medullary carcinoma (TMC) and atypical medullary carcinoma (AMC), with TMC having a better prognosis
than AMC and other DIC. The distribution of CD8+ lymphocytes within tumour nests and lymphocyte
tracts, and apoptosis in lymphocytes and tumour cells within tumour nests, were studied in archived
formalin fixed and paraffin embedded tissues of TMC and AMC. CD8+ lymphocytes tend to accumulate
along the margins of lymphocyte tracts that adjoin tumour nests. There were significantly more CD8+
lymphocytes within tumour nests of TMC than AMC. TMC also tended to have more CD8+ lymphocytes
within lymphocyte tracts than AMC. Apoptosis of lymphocytes in contact with tumour cells and of tumour
cells in contact with lymphocytes was observed in both AMC and TMC within tumour nests but differences
in the proportions of apoptotic tumour cells and lymphocytes between the two tumour types could
not be established. The findings are consistent with CD8+ cytotoxic lymphocyte-mediated immunity
contributing to the more favourable prognosis for TMC compared to AMC.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction
Medullary breast carcinoma (MBC) is a type of ductal invasive
breast carcinoma (DIC) constituting 3–6% of all invasive breast can-
cers. According to the Ridolfi classification [1], typical medullary
breast carcinoma (TMC) has morphological characteristics of pre-
dominantly syncytial growth, macroscopically and microscopically
circumscribed tumour margins, marked or moderate lymphoplas-
mocytic infiltration of the stroma, pleomorphic nuclei with a high
mitotic index, and the absence of tubule formation and an in situ
component. Atypical medullary breast carcinoma (AMC) is a closely
related breast carcinoma that also has a predominantly syncytial
growth pattern but lack up to two of the other features of TMCs
[1]. AMCs may therefore show up to two of the following features:
focal areas of tumour infiltration at the margins, absent or mild
lymphoplasmocytic infiltration of the tumour or lymphoplasmo-
cytic infiltration only at the tumour margins, uniform nuclei with
infrequent mitosis, focal tubule formation and the presence of an
in situ component [1]. Alternative classifications for MBC based on
histological criteria have also been proposed [2,3]. Although TMCs
∗ Corresponding author. Tel.: +44 777 9073537.
E-mail address: ranjanramasamy@yahoo.co.uk (R. Ramasamy).
are poorly differentiated and have a high mitotic index, their better
prognosis compared to other types of DIC including AMCs poses
a clinical paradox. This may be partly due to immunity associated
with the characteristic mononuclear cell infiltrate [1–8].
Tumour-infiltrating lymphocytes (TILs) in TMC and AMC were
previously characterized using the markers CD4, CD8, CD20, CD25,
CD45RO and CD56 and significantly higher proportions of CD45RO+
and CD8+ cells and lower proportion of CD20+ cells were observed
overall in the lymphocytic infiltrates of TMC compared to AMC [5].
This and other pertinent observations [8] suggest that CD8+ cyto-
toxic lymphocytes(CTLs) rather than antibodies may be responsible
for the better tumour control in TMC relative to AMC. However, the
relative distribution of CD8+ cells within tumour nests and lym-
phocytic tracts of TMC and AMC has previously not been reported.
Tumour size, lymph node status, histological grade and recep-
tor status are presently used to predict outcome and treatment
response in breast cancer patients. The rate of apoptosis of tumour
cells is assumed to play a key role in tumour progression as well
as response to treatment [6,7]. Escape from immunosurveillance
in tumours through immunosubversion may involve many mech-
anisms that can cause apoptosis in TILs and the development of
resistance in tumour cells to immune effector mechanisms [9]. We
therefore investigated the differential localization of CD8+ lympho-
cytes in lymphocyte tracts and tumour nests, as well as apoptosis
characterized by the cleavage of caspase-3 in lymphocytes and
0165-2478/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.imlet.2013.10.001

2. 124 I. Nurlaila et al. / Immunology Letters 156 (2013) 123–126
tumour cells within tumour nests, in TMC and AMC as a paired
model for studying tumour immunity.
2. Materials and methods
2.1. Tumour specimens
TMC and AMC samples were obtained from the Department of
Pathology’s archive at the Raja Isteri Pengiran Anak Saleha (RIPAS)
Hospital Brunei Darussalam in the form of formalin-fixed paraffin
embedded (FFPE) blocks from surgical resections performed in the
period 2004–2011 [10]. The FFPE blocks were cut into 4 ␮m thick
sections using a microtome, affixed to silanized slides (Electron
Microscopy Science, USA) and prepared for immunohistochemical
staining as described [10].
2.2. Detection of CD8+ cells
Haematoxylin and eosin was used as a counter-stain for
immunohistochemical staining and to examine the morphological
features of TMC and AMC as previously described [10]. CD8+ cells
were identified by immunohistochemical staining using a mon-
oclonal mouse anti-human CD8 antibody clone C8/144B (Dako,
Germany) followed by a secondary goat anti-mouse IgG conju-
gated to peroxidase-labelled polymer solution (Dako, Denmark).
Diaminobenzidine was then used as a peroxidase substrate. The
TMC and AMC specimens were derived from six and eight dif-
ferent tumours respectively. Between five and seven fields were
examined per tumour under a light microscope (Olympus DP25,
Japan) at 600× magnification. Positively stained lymphocytes were
enumerated as a proportion of all lymphocytes in the same fields.
2.3. Detection of apoptotic cells
Apoptotic cells were detected by immunohistochemical stain-
ing with a rabbit monoclonal antibody to cleaved caspase-3
(Cell Signaling Technology, UK), followed by a secondary F(ab )2
peroxidase-conjugated goat antibody to rabbit IgG (Santa Cruz
Biotechnology, USA). The primary antibody detected the large frag-
ment of activated caspase-3 resulting from cleavage adjacent to
Asp175. A blocking peptide of cleaved caspase-3 (Cell Signaling
Technology, UK) was pre-mixed with the primary antibody before
application in order to establish the specificity of the primary anti-
body. Etoposide-treated Jurkat cell lines (Cell Signaling Technology,
UK) were employed as positive controls.
Apoptotic cells were enumerated by light microscopy (Olympus
DP25, Japan) at 600× magnification in six different TMC and five
different AMC samples. Between five to seven randomly selected
fields containing tumour nests were examined in each sample and
the percentage of apoptotic lymphocytes in contact with tumour
cells or apoptotic tumour cells in contact with lymphocytes in the
respective total lymphocyte or tumour cell populations in the fields
determined.
2.4. Statistical analysis
A two-tailed non-parametric Mann Whitney U-tailed test was
performed to determine the significance of differences between the
proportions of CD8+ and apoptotic cells in TMC and AMC.
3. Results
3.1. CD8+ lymphocytes in tumour nests and lymphocyte tracts
The CD8+ cells were abundant in both TMC and AMC (Fig. 1).
However, instead of being evenly distributed within the lympho-
cyte tract, CD8+ cells tend to accumulate in the periphery of the
tracts adjoining tumour nests (Fig. 1).
The mean percentages of CD8+ cells in the lymphocyte tracts
of TMC and AMC were 33.4% ± 6.6 (mean ± standard deviation)
and 25.1% ± 11.3 respectively. The results were not significantly
different (z = −1.48, p = 0.14). However the percentage of CD8+ lym-
phocytes within tumour nests of TMC (66.4% ± 14.3) was higher
than AMC (45.0% ± 16.2) with the difference being statistically sig-
nificant (z = −2.26, p = 0.02).
3.2. Apoptosis in tumour nests
Etoposide-treated Jurkat cell lines show prominent apoptotic
cells, indicated by dense brown staining (Fig. 2A). Fig. 2C and D
show apoptotic cells in representative experimental samples of
TMC and AMC respectively. In an AMC sample used as a negative
control, where cleaved caspase-3 blocking peptide was preincu-
bated with the primary antibody, some cells show characteristic
hyperchromicity, nuclear shrinkage and the presence of apoptotic
bodies (arrowed in Fig. 2B) but no brown staining (Fig. 2B). This
result confirms the specificity of detection of activated caspase-3.
The results also provide evidence to suggest that direct con-
tact between tumour cells and lymphocytes within the tumour
nest may lead to apoptosis. Fig. 3A shows an apoptotic tumour
cell (arrowed) in direct contact with an adjacent lymphocyte. Its
nucleus is condensed and hyperchromatic with a shrunken shape.
Fig. 3B shows apoptosis in two adjacent lymphocytes (arrowed)
that are in contact with a tumour cell. The upper lymphocyte has
Fig. 1. Localization of CD8+ lymphocytes in TMC and AMC. Panels show the location of CD8+ lymphocytes stained brown in (A) TMC and (B) AMC (100×). The full and dotted
lines delineate lymphocyte tracts (lt) and tumour nests (tn).

3. I. Nurlaila et al. / Immunology Letters 156 (2013) 123–126 125
Fig. 2. Apoptotic cells in TMC and AMC. Apoptotic cells stained brown were detected using a primary rabbit monoclonal antibody to cleaved caspase-3 in (A) Etoposide-
treated Jurkat cells as a positive control (200×), (B) AMC sample reacted with primary antibody that was pre-treated with cleaved caspase-3 blocking peptide as a negative
control (200×), (C) TMC (400×) and (D) AMC (400×). In the AMC sample used as a negative control (B), apoptotic tumour cells are morphologically observed (black arrows)
but no staining for cleaved caspase-3 is seen in them.
Fig. 3. Contact-mediated apoptosis of tumour cells and lymphocytes in TMC and AMC. The panels show apoptosis in a tumour cell in contact with a lymphocyte (A) and two
lymphocytes in contact with a tumour cell (B) at 600× magnification. The apoptotic cells are indicated with white arrowheads.
undergone nuclear shrinkage and condensation and a decrease in
size. The other lymphocyte has already formed an apoptotic body
since its organelles, especially the nucleus, has degraded into small
granules.
Apoptotic tumour cells within tumour nests that were in direct
contact with lymphocytes as a percentage of all tumour cells in
the fields examined were 0.8% ± 0.6 (mean ± standard deviation)
and 0.6% ± 0.5 in TMC and AMC respectively. The difference was
not statistically significant (p > 0.05). Apoptotic lymphocytes that
were in direct contact with tumour cells as a percentage of all lym-
phocytes in the fields examined were 7.4% ± 5.9 and 12.5% ± 21.7
in TMC and AMC respectively. The difference was not statistically
significant (p > 0.05).
4. Discussion
Despite the anaplastic cytological feature and high mitotic rate,
the prognosis for patients diagnosed with MBC, particularly TMC,
is better than for other types of DIC [1–8]. A positive correlation
between the intensity of lymphoid infiltration and patient survival
suggests that the immune system may be involved in restraining
the spread of this type of breast cancer [4].

4. 126 I. Nurlaila et al. / Immunology Letters 156 (2013) 123–126
CD8+ CTLs are a component of the adaptive immune system.
The capacity for rapid expansion and the ability of a single CD8+
CTL to destroy more than one target cell, while sparing ‘innocent’
bystanders, make CD8+ CTLs efficient antigen-specific effector cells.
Destruction of target cells by CD8+ CTLs typically requires cell con-
tact mediated by the recognition of peptide antigen presented on
MHC Class 1 molecules on the target cell by the T cell antigen recep-
tor, followed by target cell killing through the release of cytotoxic
granules or Fas/FasL interaction [11]. In a previous study on MBC in
Brunei, a higher proportion of CD8+ lymphocytes and a lower pro-
portion of CD20+ B-lineage cells were observed to be characteristic
of TMC in comparison to AMC [5]. Observations in the present and
previous studies [5,7] are consistent with close contact between
CD8+ lymphocytes and tumour cells in TMC. They suggest that the
better tumour control in TMC might be mediated by CD8+ CTLs.
However, the comparative distribution of CD8+ cells within tumour
nests and lymphocytic tracts of TMC and AMC had previously not
been reported. The present findings suggests that CD8+ lympho-
cytes in lymphocyte tracts tend to localize close to tumour nests
in MBC and that tumour nests of TMC had a significantly greater
percentage of CD8+ lymphocytes among the infiltrating lympho-
cytes than AMC. The lymphocyte tracts in TMC also tended to have
a greater proportion of CD8+ lymphocytes than AMC but statisti-
cal significance could not be established possibly due to the small
number of samples studied. These findings are however consistent
with CD8+ CTLs being responsible for the better tumour control in
TMC compared to AMC.
Apoptosis is a form of programmed cell death that plays an
essential role in many biological processes including normal cell
turnover, immune response, embryonic development and hor-
mone dependent atrophy [12,13]. Apoptosis in tumour cells has
been associated with the better prognosis in TMC [6,7]. A study
of 50 cases of MBC and 50 cases of non-medullary DIC utilizing
the terminal deoxynucleotidyl transferase-mediated dUTP-biotin
nick end-labelling (TUNEL) method for detecting apoptotic tumour
cells and immunohistochemistry for detecting p53, bcl-2, and Ki-
67 showed that tumour cells in MBC had higher apoptotic and
proliferative indices and significantly lower positivity for the anti-
apoptotic protein Bcl-2 than matched DIC controls [6]. A similar
independent study confirmed that tumour cells in both TMC and
AMC had higher apoptotic indices than matched cases of DIC [7].
No study to date has definitively established the cause of increased
apoptosis in MBC.
Taking into account previous observations of close contacts
between CD8+ lymphocytes and tumour cells in TMC [5,7], it was
expected that the better prognosis for TMC compared with AMC,
might be reflected in more tumour cells and the fewer lymphocytes
in TMC than AMC undergoing contact mediated-apoptosis within
tumour nests. However, while apoptosis in lymphocytes in contact
with tumour cells and in tumour cells in contact with lymphocytes
was observed within the tumour nests in both TMC and AMC, the
proportions of apoptotic cells were small and significant differences
could not be established between the two MBC types.
At least two strategies used by tumours to evade rejection by the
immune system are related to apoptosis [9,14]. Firstly, malignant
cells can alter the expression of molecules involved in apopto-
sis signalling, resulting in resistance to immune cell mediated
killing mechanisms. Secondly, tumours may adopt a mechanism
to delete attacking anti-tumour lymphocytes for example through
the expression of Fas ligand (CD95L) [9,14]. Tumour cells can
resist apoptosis at the membrane receptor and intracellular levels
[9,14]. Tumour cells down-regulate membrane Fas receptor (CD95)
expression thus inhibiting engagement with Fas ligand on immune
effector cells that can cause apoptosis in Fas-expressing tumour
cells. At the intracellular level, apoptosis resistance can be caused
by up-regulation of anti-apoptotic molecules or down-regulation
or loss of pro-apoptotic molecules. Higher levels of anti-apoptotic
Bcl-2 for example protects tumour cells against an immune attack
and promotes tumour survival and proliferation [6,13,14]. Details
of such possible mechanisms need to be investigated in parallel in
TMC and AMC. Additional roles for CD4+ T cells [15], other effector
cells such as NK cells [5], and antibodies in tumour immunity also
warrant further comparative investigation in TMC and AMC.
The present study underscores the previously proposed impor-
tance of TMC and AMC as a paired model system for studying
immune mechanisms in cancer [5]. Because Brunei has a popula-
tion of approximately 400,000 persons [10], and therefore only a
limited number of MBC cases, a larger number of samples from a
multi-centre investigation will be helpful in improving the statis-
tical power of such studies.
Conflict of interest statement
The authors declare no conflict of interest.
Acknowledgements
This study was supported by a graduate research scholarship to
IN from the Universiti Brunei Darussalam.
Author contributions: IN performed experiments; PUT provided
tumour specimens; RR and PUT conceived, designed and coordi-
nated the study; IN, RR and PUT drafted the manuscript. All authors
read and approved the final manuscript.
References
[1] Ridolfi RL, Rosen PP, Port A, Kinne D, Mike V. Medullary carcinoma of the breast:
a clinicopathology study with a 10-year follow up. Cancer 1977;40:1365–85.
[2] Pedersen L, Zedeler K, Holck S, Schiodt T, Mouridsen HT. Medullary carcinoma
of the breast: prevalence and prognostic importance of classical risk factors in
breast cancer. Eur J Cancer 1995;31:2289–95. A(13-14).
[3] Wargotz ES, Silverberg SG. Medullary carcinoma of the breast: a clinico-
pathology study with appraisal of current diagnostic criteria. Hum Pathol
1988;19(11):1340–6.
[4] Malyuchik SS, Kiyamova RG. Medullary breast carcinoma. Exp Oncol
2008;30(2):96–101.
[5] Lim KHJ, Telisinghe PU, Abdullah MS, Ramasamy R. Possible significance of
differences in proportions of cytotoxic T cells and B-lineage cells in the tumour-
infiltrating lymphocytes of typical and atypical medullary carcinomas of the
breast. Cancer Immunity 2010;10:3.
[6] Kajiwara M, Toyoshima S, Yao T, Tanaka M, Tsuneyoshi M. Apoptosis and
cell proliferation in medullary carcinoma of the breast: a comparative study
between medullary and non-medullary carcinoma using the TUNEL method
and immunohistochemistry. J Surg Oncol 1999;70(4):209–16.
[7] Yakirevich E, Maroun L, Cohen O, Ben Izhak O, Rennert G, Resnick MB. Apo-
ptosis, proliferation, and Fas (APO-1, CD95)/Fas ligand expression in medullary
carcinoma of the breast. J Pathol 2000;192:166–73.
[8] Yakirevich E, Izhak OB, Rennert G, Kovacs ZG, Resnick MB. Cytotoxic phenotype
of tumour infiltrating lymphocytes in medullary carcinoma of the breast. Mod
Pathol 1999;12(11):1050–6.
[9] Zitvogel L, Tesniere A, Kroemer G. Cancer despite immunosurveil-
lance: immunoselection and immunosubvertion. Nature Rev Immunol
2006;6:715–27.
[10] Tan S, Abdullah MS, Telisinghe PU, Ramasamy R. Breast cancer in Brunei
Darussalam—incidence and the role of evaluation of molecular markers. Brunei
Int Med J 2011;7(5):250–9.
[11] Broere F, Apasov SG, Sitkovsky MV, van Eden W. Principles of immunophar-
macology, 3rd
edition. T cell subsets and T cell-mediated immunity. Basel,
Switzerland: Springer; 2011.
[12] Cohen GM. Caspase: the executioners of apoptosis. Biochem J 1997;326:1–16.
[13] Strasser A, Cory S, Adams JM. Deciphering the rules of programmed cell death
to improve therapy of cancer and other diseases. EMBO J 2011;30:3667–83.
[14] Igney FH, Krammer PH. Immune escape of tumours: apoptosis resistance and
tumour counterattack. J Leukoc Biol 2002;71:907–20.
[15] Perez-Diez A, Joncker NT, Choi K, Chan WFN, Anderson CC, Lantz O, et al.
CD4 cells can be more efficient at tumour rejection than CD8 cells. Blood
2007;109(12):5346–53.