Aims: Triple-negative breast cancer (TNBC) represents a specific group that lacks the expression of estrogen receptors, progesterone receptors, and human epidermal growth factor receptor-2 and might also lack the expression other breast markers like GATA3, mammaglobin (MG), GCDFP15 (growth cystic disease fluid protein 15), and NYBR1; when this occurs, proving the breast origin of a metastasis is a challenging task. In the present study, we assessed the added value of SOX10 immunohistochemistry to known GATA3, MG, GCDFP15, and NY-BR-1 statuses in a series of CK5-positive primary TNBCs. Methods: Tissue microarrays were made from the formalin-fixed and paraffin-embedded blocks of 120 TNBCs, and 3-4-mm-thick sections were immunostained for SOX10. The cut-off for a positive reaction was at least 10% of tumor cells staining. Results: In our cohort, SOX10 positivity was seen in 82/119 cases, 61, 74, 76, and 82 all of which were GATA3, MG, GCDFP15, and NY-BR-1 negative, respectively. Of the SOX10 negative cases, 12 stained with at least another breast marker. Nevertheless, 25/119 (21%) cases remained negative with all markers assessed. Discussion: SOX10 proved to be the most commonly positive breast marker in our CK5 expressing TNBCs, but the other markers also had some additive value to SOX10.

About 15% of breast carcinomas belong to the so called triple negative category (TNBC), lacking the expression of estrogen receptors (ER), progesterone receptors (PR), and human epidermal growth factor receptor-2 (HER2) [1]. Most (although not all) of these cancers are aggressive and may give rise to metastasis relatively early in the course of the disease [2, 3]. The expression of ER, PR, sometimes even HER2, or “breast markers” like GATA3, mammaglobin (MG), GCDFP15 (growth cystic disease fluid protein 15), and NY-BR-1 can point to the breast origin of a metastasis, but TNBCs – by definition – lack the first 3 and might also lack the others [4, 5]. In a previous work, we identified GATA3 as the most gratifying breast marker, which could still be complemented by MG and GCDFP15, with practically no added value of NY-BR-1. Acknowledging that neither of these markers are absolutely specific, we also suggested that only about half of cytokeratin 5 (CK5) expressing TNBCs could be proven to be of mammary origin with their help; therefore, better or alternative markers would be useful in clinical practice [5].

SOX10 is a transcription factor involved in neural crest differentiation [6]. Accordingly, SOX10 positivity can be seen in melanoma, nerve sheath tumors [7-10], and is also expressed in myoepithelial cells in the breast [11]. SOX10 positivity has been described in salivary gland and cutaneous adnexal gland tumors, as well [12, 13]. As many TNBCs are basal-like on the basis of gene expression profile [14], show CK5 and/or epidermal growth factor receptor positivity, and classify as basal-like on the basis of an immunohistochemistry (IHC) based surrogate classification proposed by Nielsen and Perou [15], it is not surprising that these cancers may also stain for another myoepithelial marker, SOX10. Indeed, SOX10 has been reported to be positive in 40–70% of TNBCs [4, 16-18] and may be positive even in GATA3 negative cases [18]. In the present study, we assessed the added value of SOX10 IHC to GATA3, MG, GCDFP15, and NY-BR-1 IHC in a series of CK5-positive TNBCs.

A series of CK5-positive TNBCs previously characterized by GATA3, MG, GCDFP15, and NY-BR-1 IHC in a previous analysis was used for SOX10 IHC [5]. The tumors were assessed in tumor microarrays (TMAs) being represented by dual 2-mm-diameter tissue cores; the details of TMA building were reported earlier [5].

Briefly, the tumors were derived randomly from patients operated on and diagnosed with TNBC at the Bács-Kiskun County Teaching Hospital, Kecskemét between August 2005 and August 2015. The surgical specimens were fixed in 10% neutral buffered formalin for at least 24 h. The TMAs had been constructed from archived paraffin-embedded blocks using a TMA builder device (Histopathology Ltd, Pécs, Hungary), with each TMA incorporating 20 tumor tissue cores. The TMA blocks were stored at room temperature, similarly to other paraffin blocks. 3–4-mm-thick sections were cut for SOX10 IHC using a monoclonal mouse antibody specific for an epitope mapping between amino acids 2–29 at the N terminus of SOX-10 of human origin (Santa Cruz Biotechnology, Inc., Dallas, TX, USA). The antibody was used with 1:500 dilution for 30 min incubation period and pretreatment was performed at pH 9.

TMAs were scanned, and the proportion of positive cells was independently evaluated on the digital slides by the authors; the few discrepant cases were reassessed by consensus on the original slides. Rate, localization, and intensity were registered in all cases. The data for GATA3, MG, GCDFP15, NY-BR-1 were taken from a previous analysis [5].

The institutional ethical committee of the Bács-Kiskun County Teaching Hospital of the University of Szeged was consulted and approved this non-interventional retrospective study. The institutional data safety manager also gave approval for this study not requiring patients’ identity related data. The study was finally approved by the Ethical Committee of the Albert Szent-Györgyi Medical Center of the University of Szeged.

Of the 120 TNBCs represented in the TMA cores, 119 could be assessed for SOX10 staining. SOX10 staining was generally a nuclear staining occurring in <1–100% of tumor cells (Fig. 1); therefore, 2 different cutoffs for positive staining were evaluated. With a positivity threshold of >1 and ≥10%, 93 and 82 were defined as positive. Because the proportion of cases with 1–10% staining was relatively low, and the cases are less easy to pick up, the greater threshold was used for further analysis.

Fig. 1.

Examples of SOX10 immunostaining. Diffuse strong nuclear staining (×20) (a); partial staining in >10% of the cells with moderate intensity (×20) (b); over 10% of nuclei staining weakly (×40) (c); less than 1% of nuclei staining (×40) (d).

Fig. 1.

Examples of SOX10 immunostaining. Diffuse strong nuclear staining (×20) (a); partial staining in >10% of the cells with moderate intensity (×20) (b); over 10% of nuclei staining weakly (×40) (c); less than 1% of nuclei staining (×40) (d).

Close modal

Of the 94 GATA3-negative cases, 61 cases were positive with SOX10. Similar results were observed with the other breast markers. Seventy four out of 104 MG negative cases, 76 out of 109 GCDFP15 negative cases, and 82 out of 117 NY-BR-1-negative cases stained positive with SOX10. Our series included 78 cases that were negative with all the previously tested markers, 53 of which were identified as positive with SOX10, still leaving 25 (21%) as breast marker negative. The sequential hierarchical staining for breast markers from the most commonly positive to the least commonly positive is shown in Fig-ure 2. Mutual staining figures of pairs of breast markers are shown in Table 1. Figure 3 and online suppl. Table 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000512006) illustrate the proportion of single and multiple marker expression in the series with all staining combinations experienced.

Table 1.

Pairs of breast markers and their expressions in the tumors investigated

Pairs of breast markers and their expressions in the tumors investigated
Pairs of breast markers and their expressions in the tumors investigated
Fig. 2.

Hierarchical labeling of the tumors with SOX10, GATA3, MG, GCDFP15, and NY-BR-1 as “breast specific” markers. MG, mammaglobin.

Fig. 2.

Hierarchical labeling of the tumors with SOX10, GATA3, MG, GCDFP15, and NY-BR-1 as “breast specific” markers. MG, mammaglobin.

Close modal
Fig. 3.

Number of cases demonstrating a given pattern of breast marker expression in the series investigated. MG, mammaglobin.

Fig. 3.

Number of cases demonstrating a given pattern of breast marker expression in the series investigated. MG, mammaglobin.

Close modal

The expected lifetime risk of developing cancer in women is around 1 out of 3 [19], making the possibility of developing multiple primary cancers a real possibility and giving emphasis to the distinction between metastasis of a known primary tumor and an independent second primary cancer. TNBCs represent a minority of breast cancers, but as triple negative no special type carcinomas are often of poor prognosis, metastases may occur at a relatively higher rate. Owing to their phenotype overlapping with myoepithelial differentiation, TNBCs may show variable histologies, including spindle cells (sarcomatoid appearance), squamous metaplasia, and other rarer (e.g., sebaceous, chondroid or osseous) metaplasias, making their recognition as metastatic breast carcinomas at a metastatic site more difficult. This is why IHC markers supporting the breast origin of TNBCs is important.

In our previous study, by using GATA3, MG, GCDFP15, and NYBR1, we came out with an algorithmic value of these “breast markers” in CK5 expressing TNBCs believed to represent basal-like breast carcinomas, where GATA3 was the mostly expressed marker. Our results suggested that only about half of these cancers could be classified as of mammary origin on the basis of these 4 markers [5].

Cimino-Mathews et al. [16] published the first study investigating the utility of SOX10 IHC labeling in TNBC and metaplastic breast carcinoma cases. TMA blocks from 168 primary breast cancers were investigated; 40% showed positivity with SOX10. These were primarily basal-like unclassified TNBCs and metaplastic carcinomas. There were 66% of TNBCs, but only 5% of luminal and HER2-positive cases were positive with SOX10 [16].

Nelson and coauthors reported about promising SOX10 results. TMA blocks were made from 26 patients’ samples; 38% stained positively with SOX10, while no cases were positive with ER and HER2. A retrospective study was also performed in metastatic carcinomas of possible breast origin, and 57% of cases were labeled with SOX10. All SOX10-positive cases were confirmed to be negative with ER. Nelson and coauthors recommended the use of SOX10 in metastatic cases from unknown primary tumors to prove their melanoma or TNBC origin [20].

Since 2017, several research groups started to investigate SOX10. Al-Zahrani and coauthors compared its application with androgen receptor (AR). AR staining was positive in 95% of cases, mostly along with HER2 or ER and PR positivity, but no special breast cancer subtype could be identified on the basis or AR staining alone. SOX10 proved to be positive in one-third of the cases that were triple-negative [21].

Tozbikian examined the IHC profile of 57 TNBC cases. Among them, 82% showed positivity with GATA3, 58% with SOX10, and 25% with AR; 95% proved to be positive with either GATA3 or SOX10 and 46% showed dual positivity; and 80% of GATA3-negative cases were SOX10 positive. Their study concluded that while GATA3 was a more sensitive marker for TNBC cases, it is useful to add SOX10 to the IHC panel [18].

Harbhajanka et al. [22] investigated 48 TNBC cases in TMAs, and SOX10 showed positivity in 37.5% of the cases. A negative correlation with AR-positive molecular subtypes of TNBCs was observed, and a positive correlation was proved with WT1. However, no correlation was seen with the breast markers GATA3, MG, and GCDFP15, and basal-like subtype markers epidermal growth factor receptor and CK5/6 [22].

The most comprehensive study is the one by Laurent et al. [4] who reported their results about SOX10, GATA3, GCDFP15, AR, and MG in 207 metastatic TNBC cases and compared them with 152 primary lung adenocarcinomas. SOX10 showed the best sensitivity (62.3%) and specificity (100%) in comparison with GATA3 (30.4% sensitivity and 98.7% specificity), GCDFP15 (20.8% sensitivity and 98% specificity), MG (38.2% sensitivity and 81.6% specificity), and AR (30% sensitivity and 86% specificity); 6.3% of TNBC cases had no reaction with any of the above mentioned markers [4].

This year, Qazi et al. [23] suggested the combined use of SOX10 and GATA3 in cases of low ER expression and reduced GATA3 intensity. The study included 246 patients’ samples as TMA blocks containing both ER-positive and ER-negative cases. Overall, it was 93%, and of TNBC cases, 63% showed positivity with GATA3; in parallel, 15% of all and 74% of TNBCs were positive with SOX10. Less than 1% had no reaction with either antibody and only 3% of ER-positive cases were SOX10 positive [23].

In our current study, by adding SOX10 IHC to our previous “breast marker” panel, there was an improvement in identifying CK5-expressing TNBCs as of mammary origin, and the algorithm could be changed substantially (Fig. 2). Using the 10% cutoff for SOX10 positivity, 68.9% (95% CI: 59.8–77.1) of the cases were found to be SOX10 positive and 9.3% of cases were positive only with the previously used markers, which proves the added value of GATA3, MG, GCDFP15, and NYBR1; 21% of the cases remained negative with all “breast markers,” suggesting that negativity of all the examined markers does not securely exclude mammary origin. These results are in keeping with former reports [4, 23] and suggest that SOX10 is probably the best “breast marker” of TNBCs, followed by GATA3. Minor discrepancies in the proportion of cases staining and the value of MG versus GCDPF15 may stem from our cohort being restricted to CK5 expressing (and most likely basal-like) TNBCs, whereas others also included apocrine TNBCs which are expected to be positive with GCDFP15 (also an apocrine marker) more commonly. In 41% of the cases in our study, the staining observed was focal; in these cases, the application of the TMA technique can be a limitation.

Based on our data, SOX10 proved to be the most sensitive breast marker in CK5-expressing TNBCs, likely to correspond to basal-like TNBCs on the basis of the IHC-based surrogate classification. With the additive value of GATA3, MG, GCDFP15, and NY-BR-1, more than 3 quarters of the investigated 119 cases could be identified as breast cancers. With the joint use of SOX10, GATA3, MG, and GCDFP15, 78.2% (95% CI: 69.7–82.2) sensitivity was achieved. We propose SOX10 as the first-line approach to identify TNBCs, with the addition of GATA3, MG, and GCDFP15 for the negative cases; NY-BR-1 has little added value in this context.

This study was funded by the National Research, Development and Innovation Office grant GINOP-2.3.2-15-2016-00020.

The institutional Ethical committee of the Bács-Kiskun County Teaching Hospital was consulted and approved this non-interventional retrospective study, which was also approved by the Regional Ethical Committee of the Albert Szent-Györgyi Medical Center of the University of Szeged.

No editorial or financial conflicts of interest exist for this submission.

Concept and design – Anita Sejben, András Vörös, and Gábor Cserni. Evaluation of immunostained slides - all authors. Consensus on near cut-off staining cases – all authors. Search and evaluation of references – all authors. Drafting the manuscript – Anita Sejben, András Vörös, and Gábor Cserni. Approval of final manuscript – all authors.

1.
Yao
H
,
He
G
,
Yan
S
,
Chen
C
,
Song
L
,
Rosol
TJ
, et al
Triple-negative breast cancer: is there a treatment on the horizon?
Oncotarget
.
2017
;
8
(
1
):
1913
24
. .
2.
Aysola
K
,
Desai
A
,
Welch
C
,
Xu
J
,
Qin
Y
,
Reddy
V
, et al
Triple negative breast cancer: an overview
.
Hereditary Genet
.
2013
;
2013
(
Suppl 2
):
001
. .
3.
Collignon
J
,
Lousberg
L
,
Schroeder
H
,
Jerusalem
G
.
Triple-negative breast cancer: treatment challenges and solutions
.
Breast Cancer
.
2016
;
8
:
93
107
. .
4.
Laurent
E
,
Begueret
H
,
Bonhomme
B
,
Veillon
R
,
Thumerel
M
,
Velasco
V
, et al
SOX10, GATA3, GCDFP15, androgen receptor, and mammaglobin for the differential diagnosis between triple-negative breast cancer and TTF1-negative lung adenocarcinoma
.
Am J Surg Pathol
.
2019
;
43
(
3
):
293
302
. .
5.
Zombori
T
,
Cserni
G
.
Immunohistochemical analysis of the expression of breast markers in basal-like breast carcinomas defined as triple negative cancers expressing keratin 5
.
Pathol Oncol Res
.
2018
;
24
(
2
):
259
67
. .
6.
Motohashi
T
,
Watanabe
N
,
Nishioka
M
,
Nakatake
Y
,
Yulan
P
,
Mochizuki
H
, et al
Gene array analysis of neural crest cells identifies transcription factors necessary for direct conversion of embryonic fibroblasts into neural crest cells
.
Biol Open
.
2016
;
5
(
3
):
311
22
. .
7.
Mohamed
A
,
Gonzalez
RS
,
Lawson
D
,
Wang
J
,
Cohen
C
.
SOX10 expression in malignant melanoma, carcinoma, and normal tissues
.
Appl Immunohistochem Mol Morphol
.
2013
;
21
(
6
):
506
10
. .
8.
Willis
BC
,
Johnson
G
,
Wang
J
,
Cohen
C
.
SOX10: a useful marker for identifying metastatic melanoma in sentinal lymph nodes
.
Appl Immunohistochem Mol Morphol
.
2015
;
23
:
109
12
.
9.
Kang
Y
,
Pekmezci
M
,
Folpe
AL
,
Ersen
A
,
Horvai
AE
.
Diagnostic utility of SOX10 to distinguish malignant peripheral nerve sheath tumor from synovial sarcoma, including intraneural synovial sarcoma
.
Mod Pathol
.
2013
;
27
(
1
):
55
61
. .
10.
Karamchandani
JR
,
Nielsen
TO
,
van de Rijn
M
,
West
RB
.
SOX10 and S100 in the diagnosis of soft-tissue neoplasms
.
Appl Immunohistochem Mol Morphol
.
2012
;
20
(
5
):
445
50
. .
11.
Chiu
K
,
Ionescu
DN
,
Hayes
M
.
SOX10 expression in mammary invasive ductal carcinomas and benign breast tissue
.
Virchows Arch
.
2019
;
474
(
6
):
667
72
. .
12.
Hsieh
M
,
Lee
YH
,
Chang
YL
.
SOX10-positive salivary gland tumors: a growing list, including mammary analogue secretory carcinoma of the salivary gland, sialoblastoma, low-grade salivary duct carcinoma, basal cell adenoma/adenocarcinoma, and a subgroup of mucoepidermoid carcinoma
.
Hum Pathol
.
2016
;
56
:
134
42
. .
13.
Lezcano
C
,
Ho
J
,
Seethala
RR
.
Sox10 and DOG1 expression in primary adnexal tumors of the skin
.
Am J Dermatopathol
.
2017
;
39
:
896
02
. .
14.
Prat
A
,
Adamo
B
,
Cheang
MC
,
Anders
CK
,
Carey
LA
,
Perou
CM
.
Molecular characterization of basal-like and non-basal-like triple-negative breast cancer
.
Oncologist
.
2013
;
18
(
2
):
123
33
. .
15.
Nielsen
TO
,
Perou
CM
.
CCR 20th anniversary commentary: the development of breast cancer molecular subtyping
.
Clin Cancer Res
.
2015
;
21
(
8
):
1779
81
. .
16.
Cimino-Mathews
A
,
Subhawong
AP
,
Elwood
H
,
Warzecha
HN
,
Sharma
R
,
Park
BH
, et al
Neural crest transcription factor Sox10 is preferentially expressed in triple-negative and metaplastic breast carcinomas
.
Hum Pathol
.
2013
;
44
(
6
):
959
65
. .
17.
Peevey
J
,
Sumpter
I
,
Paintal
A
,
Laskin
W
,
Sullivan
M
.
SOX10 is a useful marker for triple negative breast cancer
.
Am J Clin Pathol
.
2015
;
144
(
Suppl 2
):
A299
. .
18.
Tozbikian
GH
,
Zynger
DL
.
A combination of GATA3 and SOX10 is useful for the diagnosis of triple-negative breast cancer
.
Hum Pathol
.
2019
;
44
:
959
65
.
20.
Nelson
ER
,
Sharma
R
,
Argani
P
,
Cimino-Mathews
A
.
Utility of SOX10 labeling in metastatic breast carcinomas
.
Hum Pathol
.
2017
;
67
:
205
10
. .
21.
Al-Zahrani
KN
,
Cook
DP
,
Vanderhyden
BC
,
Sabourin
LA
.
Assessing the efficacy of androgen receptor and Sox10 as independent markers of the triple-negative breast cancer subtype by transcriptome profiling
.
Oncotarget
.
2018
;
9
(
70
):
33348
59
. .
22.
Harbhajanka
A
,
Chahar
S
,
Miskimen
K
,
Silverman
P
,
Harris
L
,
Williams
N
, et al
Clinicopathological, immunohistochemical and molecular correlation of neural crest transcription factor SOX10 expression in triple-negative breast carcinoma
.
Hum Pathol
.
2018
;
80
:
163
9
. .
23.
Qazi
MS
,
McGregor
SM
.
Combined use of SOX10 and GATA3 in mammary carcinoma
.
Pathol Res Pract
.
2020
;
216
(
2
):
152801
. .

Anita Sejben and András Vörös contributed equally to the present work and qualify as first authors.

Copyright / Drug Dosage / Disclaimer
Copyright: All rights reserved. No part of this publication may be translated into other languages, reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording, microcopying, or by any information storage and retrieval system, without permission in writing from the publisher.
Drug Dosage: The authors and the publisher have exerted every effort to ensure that drug selection and dosage set forth in this text are in accord with current recommendations and practice at the time of publication. However, in view of ongoing research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions, the reader is urged to check the package insert for each drug for any changes in indications and dosage and for added warnings and precautions. This is particularly important when the recommended agent is a new and/or infrequently employed drug.
Disclaimer: The statements, opinions and data contained in this publication are solely those of the individual authors and contributors and not of the publishers and the editor(s). The appearance of advertisements or/and product references in the publication is not a warranty, endorsement, or approval of the products or services advertised or of their effectiveness, quality or safety. The publisher and the editor(s) disclaim responsibility for any injury to persons or property resulting from any ideas, methods, instructions or products referred to in the content or advertisements.