Histopathology 2019, 74, 578–586. DOI: 10.1111/his.

13801

HER2 immunohistochemical scores provide prognostic

information for patients with HER2-type invasive breast
cancer
Yi-Fang Tsai,1,2 Ling-Ming Tseng,1,2 Pei-Ju Lien,1 Chih-Yi Hsu,2,3,4 Yen-Shu Lin,1
Kuang-Liang King,1 Yu-Ling Wang,1 Ta-Chung Chao,2,5 Chun-Yu Liu,2,5 Jen-Hwey Chiu1,2
& Muh-Hwa Yang2,5
1
Comprehensive Breast Health Center and Division of General Surgery, Department of Surgery, Taipei Veterans General
Hospital, 2Institute of Clinical Medicine and Faculty of Medicine, School of Medicine, National Yang-Ming University,
3
Department of Pathology and Laboratory Medicine, Taipei Veterans General Hospital, 4College of Nursing, National
Taipei University of Nursing and Health Sciences, and 5Division of Medical Oncology, Department of Oncology, Taipei
Veterans General Hospital, Taipei, Taiwan

Date of submission 20 July 2018

Accepted for publication 1 December 2018
Published online Article Accepted 4 December 2018

Tsai Y-F, Tseng L-M, Lien P-J, Hsu C-Y, Lin Y-S, King K-L, Wang Y-L, Chao T-C, Liu C-Y, Chiu J-H & Yang M-H
(2019) Histopathology 74, 578–586. https://doi.org/10.1111/his.13801
HER2 immunohistochemical scores provide prognostic information for patients with HER2-
type invasive breast cancer

Aims: Invasive breast cancer patients with human
epidermal growth factor receptor 2 (HER2)
immunohistochemical (IHC) scores of 3+ or 2+ with
reflex in-situ hybridisation (ISH) positivity are suit-
able for anti-HER2 therapies. The aim of this study
is to investigate whether the prognoses between
IHC 3+ patients and IHC 2+/ISH+ patients are dif-
ferent.
Methods and results: We analysed the clinicopatho-
logical information of 886 consecutive cases of
HER2-positive early breast cancer. The influences of
the patients’ age, cancer stage, hormone receptor sta-
tus and anti-HER2 treatment were adjusted using a
multivariate Cox regression model. Both HER2 copy
numbers and HER2 ISH ratios of the IHC 3+ group
Keywords: breast neoplasms, erbB-2, immunohistochemistry, in-situ hybridisation, prognosis
were significantly higher than those of the IHC 2+/
ISH+ group. The outcomes of IHC 3+ patients were
significantly better than those of IHC 2+/ISH+
patients in the univariate and multivariate analyses.
HER2 copy numbers of ≥8 represented the best prog-
nostic value, and it was chosen to be the cut-off
value. The reflex ISH for IHC 2+ patients with high
HER2 copy numbers (≥8) predicted a better overall
survival than that for those with low HER2 copy
numbers.
Conclusion: HER2 IHC scores and HER2 copy num-
bers can provide prognostic information for patients
with HER2-positive invasive breast cancer. Both IHC
3+ and IHC 2+ patients with high HER2 copy num-
bers had a better prognosis.

Introduction targeted therapy has improved the prognosis of

Human epidermal growth factor receptor 2 (HER2)
is a key marker for breast cancer, and HER2-
Address for correspondence: C-Y Hsu, Department of Pathology and
Laboratory Medicine, Taipei Veterans General Hospital, No. 201,
Section 2, Shipai Road, Taipei 112, Taiwan. e-mail:
cyhsu@vghtpe.gov.tw
© 2018 John Wiley & Sons Ltd.
patients with HER2-positive breast cancer. Current
guidelines recommend that HER2 testing should be
performed for all newly diagnosed and recurrent
invasive carcinomas of the breast to aid treatment
selection and provide prognostic information.1,2
HER2 positivity is defined as HER2 protein overex-
pression as measured using immunohistochemical

(IHC) stain or HER2 gene amplification, which was
evaluated using in-situ hybridisation (ISH).1 The
assessment of HER2 protein expression by IHC is
semiquantitative, and is scored as 0 (negative), 1+
(negative), 2+ (equivocal) and 3+ (positive). Elevated
HER2 protein expression is strongly correlated with
HER2 gene amplification.3 HER2-targeted therapies
are recommended for patients with invasive breast
cancer that is either HER2 ISH-positive or HER2
IHC-positive (score 3+), while HER2-targeted thera-
pies are not recommended if the HER2 test result is
negative.1 When a HER2 test result is equivocal
(score 2+), reflex testing should be performed either
with the same specimen by an alternative test or on
an alternative specimen.1 In Taiwan, HER2 testing
is generally first performed by IHC. Reflex ISH test-
ing for an equivocal IHC result (score 2+) is reim-
bursed by the National Health Insurance of Taiwan.
Patients of invasive breast cancer with either
HER2 IHC score of 3+ or 2+ but positive for reflex
ISH are indicated for anti-HER2 target therapies;
however, their prognoses may differ. A previous
study suggests that patients with high HER2 protein
expression levels quantified by the HERmark assay
may respond better to the combined treatment with
lapatinib and trastuzumab.4 As the protein expres-
sion of HER2 IHC 3+ is higher than that of HER2
IHC 2+, we predicted that patients with HER2 IHC
3+ tumours would have a better response to anti-
HER2 therapy, and their prognosis would be better
than those with HER2 IHC 2+ tumours. Other prog-
nostic factors that may have impacts on the survival
of breast cancer patients, such as age,5 cancer stage,6
hormone receptor status,7 intermediate HER2 copy
number (6–12),8 HER2 low ISH ratio (≥2 to <5)7
and dual blockade of anti-HER2 therapy,9 are all
included in the analysis.
This study aimed to evaluate whether HER2 IHC
scores indicate the outcome of HER2-type early breast
cancer. Moreover, the clinicopathological features
and survival rates among cases with IHC 3+ and IHC
2+/ISH were compared.
Materials and methods
The study protocol was approved by the Institutional
Review Board of Taipei Veterans General Hospital.
Clinicopathological information of 886 consecutive
patients with HER2+ early breast cancer who under-
went surgery at Taipei Veterans General Hospital,
Taiwan, ROC from October 2007 to December 2016
were retrieved from their medical records. The med-
ian follow-up time was 53.4 months. Recurrences
© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.
Prognostic value of HER2 IHC score 579
and disease-related deaths were observed in 114
(12.8%) and 68 (7.6%) of cases, respectively.
The original histopathological slides, including IHC
stains for oestrogen receptor (ER) (clone 6F11; Leica
Biosystems, Newcastle, UK; 1:100), progesterone
receptor (PR) (clone 16; Leica Biosystems, Newcastle,
UK; 1:150) and HER2 (A0485; Dako, Glostrup, Den-
mark; 1:900), were evaluated by author C.Y.H. by
following previously reported instructions.1,10 In
addition, ≥1% of tumour cells exhibiting nuclear
staining were regarded as positive for ER and PR.10
HER2 IHC positivity (score 3+) was defined by com-
plete intense membrane staining in >10% of tumour
cells.1 Reflex ISH testing by fluorescence ISH
(PathVysion HER2 DNA Probe Kit; Abbott Laborato-
ries, Des Plaines, IL, USA) was performed for cases
with equivocal HER2 IHC results (score 2+). Patients
with average HER2 copy numbers of ≥6 signals/cell
or a HER2 ISH ratio (HER2 gene signals to chromo-
some 17 centromere signals) of ≥2 were regarded as
ISH-positive by 2013 American Society of Clinical
Oncology (ASCO)/College of American Pathologists
(CAP) guidelines.1 Additional ISH tests for evaluat-
ing the HER2 gene status were performed for 464
patients with available HER2 IHC-positive (score 3+)
tumour samples.
Fisher’s exact test was used to compare the distri-
butions of categorical variables. Differences among
continuous variables were compared using the
Kruskal–Wallis test. Recurrence-free survival (RFS)
was measured from the date of surgery to the date
of first occurrence of local, regional or distant recur-
rence. Contralateral disease and other second pri-
mary cancers were considered censoring events.
Overall survival (OS) was measured from the date of
surgery to the date of death or last follow-up. Sur-
vival curves were plotted using the Kaplan–Meier
method, and their differences were calculated by the
log-rank test. The Cox regression model was used to
evaluate the risk of recurrence and death, and the
influence of patients’ ages, cancer stages, hormone
receptor status and anti-HER2 treatment was
adjusted accordingly in multivariate analyses. The
prognostic values of different models were compared
using Harrell’s C-index, which is a rank parameter
that measures the ordinal predictive power of a sur-
vival model by determining the probability of con-
cordance between the predicted and observed
survival.11 Harrell’s C-index can range from 0.5 (no
predictive discrimination) to 1.0 (perfect separation
of patients with different outcomes). P-values were
derived from two-tailed tests, and P < 0.05 were
considered statistically significant.
580 Y-F Tsai et al.

pathological complete response was obtained in 43.4%

Results
CORRELATION BETWEEN HER2 IHC AND
CLINICOPATHOLOGICAL FEATURES
Clinicopathological features of the study cohort are listed
in Table 1. Hormone receptor negativity and treatment
with dual anti-HER2 agents were observed more fre-
quently in the IHC 3+ group than the IHC 2+ group. For
patients receiving neoadjuvant anti-HER2 therapy,
(49 of 113) and 28.0% (7 of 25) of those with IHC 3+
and IHC 2+ tumour, respectively. The survival between
patients receiving adjuvant or neoadjuvant anti-HER2
therapy was not significantly different (Figure S1).
CORRELATION BETWEEN HER2 IHC AND ISH
Both HER2 copy numbers and HER2/CEP17 ratios of
the IHC 3+ group were significantly higher than

Table 1. Clinicopathological features by HER2 immunohistochemical scores

Total IHC score of 2+ IHC score of 3+

Characteristics (n = 886) (n = 202) (n = 684) P

Median age, years (IQR) 54 (47, 63) 53 (46, 63) 55 (47, 62) 0.329

Stage, n (%)

I 328 (37.0%) 76 (37.6%) 252 (36.8%) 0.291

II 380 (42.9%) 93 (46.0%) 287 (42.0%)

III 178 (20.1%) 33 (16.3%) 145 (21.2%)

Hormonal receptor, n (%)

Negative 431 (48.7%) 47 (23.3%) 384 (56.1%) <0.001

Positive 455 (51.4%) 155 (76.7%) 300 (43.9%)

Anti-HER2 therapy, n (%)

None 246 (27.8%) 56 (27.7%) 190 (27.8%) 0.006

Single 510 (57.6%) 130 (64.4%) 380 (55.6%)

H 495 (55.9%) 129 (63.9%) 366 (53.5%)

L 15 (1.7%) 1 (0.5%) 14 (2.1%)

Dual 130 (14.7%) 16 (7.9%) 114 (16.7%)

HP 96 (10.8%) 10 (5.0%) 86 (12.6%)

HL 21 (2.4%) 5 (2.5%) 16 (2.3%)

TP 11 (1.2%) 1 (0.5%) 10 (1.5%)

TL 1 (0.1%) 0 (0%) 1 (0.2%)

LP 1 (0.1%) 0 (0%) 1 (0.2%)

Neoadjuvant anti-HER2 therapy, n (%) 138 (15.6%) 25 (12.4%) 113 (16.5%) 0.185

Pathological complete response, n (%) 56 (40.6%) 7 (28.0%) 49 (43.4%) 0.182

Median follow-up time, months (IQR) 56.2 (31.3, 81.2) 55.8 (31.3, 79.0) 56.3 (31.3, 82.0) 0.460

Cases with recurrence, n (%) 122 (13.8%) 37 (18.3%) 85 (12.4%) 0.037

Disease-related deaths, n (%) 71 (8.0%) 24 (11.9%) 47 (6.9%) 0.027

5-year survival rate, % (95% CI) 92.9 (90.7–94.7) 90.2 (84.1–94.1) 93.7 (91.2–95.6) 0.018

CI, Confidence interval; H, Trastuzumab; HER2, Human epidermal growth factor receptor 2; IHC, Immunohistochemistry; IQR, Interquartile
range; L, Lapatinib; P, Pertuzumab; T, Trastuzumab emtansine.

© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.

 Prognostic value of HER2 IHC score 581

Table 2. Results of HER2 in-situ hybridisation by HER2 immunohistochemical scores

Total IHC score of 2+ IHC score of 3+

(n = 666) (n = 202) (n = 464) P

HER2 signals, median (IQR) 11.2 (6.4, 17.2) 6.0 (4.4, 9.2) 13.8 (9.2, 18.6) <0.001

≥4, n (%) 627 (94.1%) 169 (83.7%) 458 (98.7%) <0.001

≥6, n (%) 532 (79.9%) 104 (51.5%) 428 (92.2%) <0.001

≥8, n (%) 441 (66.2%) 63 (31.2%) 378 (81.5%) <0.001

≥10, n (%) 368 (55.3%) 45 (22.3%) 323 (69.6%) <0.001

≥12, n (%) 318 (47.8%) 35 (17.3%) 283 (61.0%) <0.001

HER2 ISH ratio, median (IQR) 5.5 (3.4, 8.1) 3.2 (2.4, 4.9) 6.4 (4.6, 9.4) <0.001

≥3, n (%) 540 (81.1%) 112 (55.5%) 428 (92.2%) <0.001

≥4, n (%) 458 (68.8%) 67 (33.2%) 391 (84.3%) <0.001

≥5, n (%) 367 (55.1%) 46 (22.8%) 321 (69.2%) <0.001

≥6, n (%) 302 (45.4%) 39 (19.3%) 263 (56.7%) <0.001

HER2, Human epidermal growth factor receptor 2; IHC, Immunohistochemistry; IQR, Interquartile range; ISH, In-situ hybridisation.

those of the IHC 2+ group (P < 0.001) (Table 2). RFS of patients with hormone-receptor-negative dis-
Moreover, the frequency of increased HER2 copy ease (Tables S1 and S2).
numbers (≥4, ≥6, ≥8, ≥10 or ≥12) and HER2/CEP17
ratios (≥3, ≥4, ≥5 or ≥6) in the IHC 3+ group were
significantly higher than those in the IHC 2+ group. HER2 ISH CORRELATED WITH SURVIVAL

Cases with higher HER2 copy numbers or HER2 ISH

HER2 IHC CORRELATED WITH SURVIVAL
Survival curves stratified by HER2 IHC results are
plotted in Figure 1. Both the RFS and OS of patients
with IHC 3+ tumours were significantly better than
those of patients with IHC 2+ tumours (P = 0.028
and P = 0.014, respectively).
In subgroup analysis for patients receiving anti-
HER2 therapy, those with IHC 3+ tumours displayed
a significantly better OS than those with IHC 2+
tumours (P = 0.011) (Figure S2D).
For patients with hormone-receptor-positive dis-
ease, both the RFS and OS were significantly better
(P = 0.011 and P = 0.011, respectively) in patients
with IHC 3+ tumours than those with IHC 2+
tumours (Figure S3C, D).
The multivariate Cox regression model revealed
that the hazard ratios of the IHC 3+ group were sig-
nificantly lower than those of the IHC 2+ group for
both RFS (P = 0.001) and OS (P = 0.001) (Table 3).
In subgroup multivariate analysis, the outcomes of
IHC 3+ cases were significantly better than those
of IHC 2+ cases in every subgroup except the OS of
patients not receiving anti-HER2 therapy and the
© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.
ratios had better survival than did those with lower
HER2 copy numbers or HER2 ratios in both univariate
and multivariate analyses (Tables S3 and 4). Also, HER2
copy numbers of ≥8 represented the best prognostic
value in univariate and multivariate analyses, and it was
chosen to be the prognostic cut-off value for HER2 ISH.
In multivariate analysis subgrouped by anti-HER2
therapy and hormone receptor status, the RFS and
OS of HER2 copy numbers of ≥8 cases were better
than those of HER2 copy numbers of <8 cases in
most subgroups. However, the difference of RFS in
the subgroup not receiving anti-HER2 therapy was
not significant (Tables S4 and S5).
Of the 202 IHC 2+ tumours, cases with HER2 copy
numbers of ≥8 had a significantly lower risk for death
(hazard ratio = 0.323, P = 0.032) compared to those
with HER2 copy numbers of <8 in multivariate anal-
ysis (Table S6).
Discussion
In this study, we confirmed that HER2 IHC scores
provided prognostic information for patients with
582 Y-F Tsai et al.

A 1.0

0.8

Recurrence-free survival 0.6

0.4

0.2 P = 0.028
IHC 3+
IHC 2+ ISH+
0.0
0 12 24 36 48 60 72 84 96 108 120
# at risk Survival time (months)
IHC 3+ 684 643 547 448 367 310 221 151 102 62 22
IHC 2+ ISH+ 202 188 166 127 103 80 55 41 21 6 2

B 1.0

0.8
Overall survival

0.6

0.4

0.2 P = 0.014
IHC 3+
IHC 2+ISH+
0.0
0 12 24 36 48 60 72 84 96 108
# at risk Survival time (months)
IHC 3+ 684 653 575 476 388 327 230 155 105 64 23
IHC 2+ ISH+ 202 192 173 137 115 89 62 43 23 9 2

Figure 1. Survival curves of the study cohort. A, Recurrence-free survival. B, Overall survival.

HER2-positive breast cancer. The outcome of IHC 2+
patients was significantly inferior to that of IHC 3+
patients in the multivariate analysis with adjustments
for age, cancer stage, hormone receptor status and
anti-HER2 treatment. HER2 copy numbers of ≥8 rep-
resented the best prognostic value, and this could be
the prognostic marker. The reflex ISH results provided
further information for IHC 2+ cases, and high HER2
copy numbers (≥8) predicted a better OS.
As a retrospective study, it may be limited by a
selection bias which may distort the estimate of the
association between risk factors and disease. More-
over, HER2 IHC is a semiquantitative measure of pro-
tein levels; thus, the potential for interlaboratory
variation in interpretation of IHC scores is a justifi-
able concern. Although the ASCO/CAP guideline rec-
ommendations have improved the accuracy of HER2
IHC testing,12 in this study the staining intensity of

© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.

 Prognostic value of HER2 IHC score 583

Table 3. Multivariate Cox regression analysis of prognostic 2005 revealed that the benefit from trastuzumab
factors in patients with HER2-positive breast cancer mainly relies upon IHC or FISH HER2 testing, irre-
spective of the degree of amplification.15 However, in
Recurrence-free survival Overall survival
Factors n a secondary analysis of the HERA trial with a median
Hazard Hazard follow-up time of 8 years, ER-positive patients with a
ratio P ratio P low ISH ratio (≥2 to <5) derived significantly less ben-
Age (years) 886 1.023 0.003 1.026 0.009 efit from adjuvant trastuzumab therapy.7 This indi-
cates that a longer follow-up time may be necessary
Stage 886
to accurately assess the benefits from anti-HER2
I 328 1 1 treatment in such a group. In our hormone-receptor-
positive subgroup, patients with a low ISH ratio
II 380 3.273 <0.001 5.255 <0.001
tumour displayed a worse outcome (Table S5), which
III 178 7.829 <0.001 12.308 <0.001 is consistent with the findings from the HERA trial.
Hormonal 886
Similarly, a worse outcome was also observed in our
receptor cases with HER2 IHC 2+ or HER2 copy numbers of
<8. The known bidirectional cross-talk between hor-
Negative 431 1 1 mone receptors and the HER2 signalling pathway
Positive 455 0.659 0.030 0.575 0.032 may stand for the difference in the outcome of this
subgroup.16 Simultaneously blocking both signalling
Anti-HER2 886
pathways is mandatory by combined targeted ther-
therapy
apy. Furthermore, approximately 20% of HER2 test-
None 246 1 1 ing might be inaccurate if it was performed before
Single 510 0.457 <0.001 0.481 0.006 implementation of the ASCO/CAP guidelines in
2007,17,18 which provided substantial improvement
Dual 130 0.378 0.003 0.198 0.003 in testing accuracy.12 As the patients in the HERA
HER2 IHC 886 trial were enrolled from 39 countries during Decem-
score ber 2001 to June 2005, and in addition patients cate-
gorised as IHC2+ by local identification were not
2+ 202 1 1
rechecked centrally, those IHC 2+ might not reflect
3+ 684 0.506 0.001 0.399 0.001 the true HER2 status if using current standards. In
contrast, patients in our study were collected from
HER2, Human epidermal growth factor receptor 2; IHC, Immuno-
histochemistry.
October 2007 to December 2016, so that we can
assess HER2 status following the ASCO/CAP guideline
recommendations and have a longer median follow-
immunostains may be affected by the thickness of tis- up time of 56.2 months. These differences make our
sue sections and other factors. Because there was no findings inconsistent with previous results from HERA
internal positive control as a reference, interpretation or N9831 trials.
of the staining intensity of IHC 2+ and IHC 3+ was Sample size is also an important issue. Compared
subjective. Because an essential HER2 IHC stain and with the present study including 886 cases, one
a reflex ISH were used to provide prognostic informa- smaller-scale study conducted in patients presenting
tion for patients with HER2-positive breast cancer, it between 2005 and 2008 did not reveal a significant
is unneccesary to overemphasise these limitations. survival difference between 163 HER2 IHC 3+
Retrospective analyses from the HERA trial data set patients and 59 IHC 2+/ISH+ patients receiving adju-
with a median follow-up time of 2 years suggested vant trastuzumab therapy.8 However, similar to our
that benefit from adjuvant trastuzumab treatment results, IHC 2+/ISH+ patients with high HER2 copy
was not reduced in HER2 IHC 2+/ISH+ cases.13 Also, numbers (>12) had better survival rates than did
there was no association between staining intensity, those with intermediate HER2 copy numbers (6–12)
as assessed by image analysis, and benefits derived (it was not possible to analyse P-values).8 As the
from 1 year of trastuzumab treatment.14 Although study cohorts were different, the cut-off value of
the intensity of HER2 staining was correlated with HER2 signals to define a high HER2 copy number for
HER2 gene copy numbers,14 the level of HER2 ampli- the other study was different from that used in our
fication did not influence prognosis.13 Another study (≥8). In a study of metastatic breast cancer,
(N9831) trial conducted during March 2002 to April high HER2 copy numbers (>13) were associated with
© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.
584 Y-F Tsai et al.

Table 4. Prognostic values (Harrell’s C-index) of the HER2 ISH results by different cut-off values using multivariate Cox
model

Recurrence-free survival Overall survival

N (%)
(n = 666) Hazard ratio (P) Harrell’s C index Hazard ratio (P) Harrell’s C index

HER2 signals 0.958 (0.012) 0.736 0.945 (0.010) 0.776

≥4 627 (94.1%) 0.394 (0.005) 0.727 0.306 (0.003) 0.763

≥5 565 (84.8%) 0.404 (<0.001) 0.739 0.287 (<0.001) 0.782

≥6 532 (79.9%) 0.411 (<0.001) 0.742 0.326 (<0.001) 0.779

≥7 482 (72.4%) 0.441 (<0.001) 0.741 0.335 (<0.001) 0.776

≥8 441 (66.2%) 0.453 (<0.001) 0.746 0.300 (<0.001) 0.792

≥9 412 (61.9%) 0.473 (<0.001) 0.746 0.353 (<0.001) 0.791

≥10 368 (55.3%) 0.553 (0.004) 0.738 0.432 (0.001) 0.783

≥11 337 (50.6%) 0.592 (0.011) 0.735 0.537 (0.015) 0.776

≥12 318 (47.8%) 0.584 (0.010) 0.737 0.567 (0.028) 0.774

≥13 283 (42.5%) 0.678 (0.064) 0.731 0.642 (0.092) 0.769

≥14 255 (38.3%) 0.689 (0.082) 0.729 0.657 (0.118) 0.766

HER2 ISH ratio 0.900 (0.002) 0.743 0.878 (0.003) 0.781

≥2 663 (99.6%) 1.129 (0.905) 0.724 0.648 (0.676) 0.763

≥3 540 (81.1%) 0.410 (<0.001) 0.739 0.308 (<0.001) 0.775

≥4 458 (68.8%) 0.449 (<0.001) 0.743 0.386 (<0.001) 0.780

≥5 367 (55.1%) 0.503 (0.001) 0.740 0.484 (0.005) 0.777

≥6 302 (45.4%) 0.575 (0.009) 0.733 0.535 (0.018) 0.773

≥7 232 (34.8%) 0.586 (0.017) 0.736 0.545 (0.030) 0.774

The influences of age, cancer stage, hormone receptor status and anti-HER2 treatment were adjusted. Bold value indicates the best value.
HER2, Human epidermal growth factor receptor 2, ISH, In-situ hybridisation.

a short time-to-metastasis; however, patients
responded well to trastuzumab-based therapy.19
Moreover, although high HER2 copy numbers were
correlated with better survival for HER2-positive
breast cancer, benefit from anti-HER2 therapy did not
correlate linearly with HER2 copy numbers.19 Our
cohort used HER2 signals of ≥8 as the cut-off value,
and could reach a better statistical significance.
According to the 2013 ASCO/CAP guidelines, cases
with HER2 copy numbers ≥4 to <6 and ISH ratio <2
were considered as ISH-equivocal.1 If the HER2 test
result is ultimately deemed to be equivocal, the oncol-
ogists may consider anti-HER2 target therapy.1 The
2018 updated ASCO/CAP guideline recommendations
suggest that the formerly ISH-equivocal cases with
concurrent IHC scored 2+ are defined as HER2-
negative.20 The benefit from anti-HER2 target ther-
apy for these cases is uncertain.20 HER2 IHC 2+ cases
with HER2 copy numbers of <4 and ISH ratios ≥2 in
the reflex ISH test were regarded as HER2-negative
by the 2018 updated ASCO/CAP guidelines.20 The
present study including 202 IHC 2+ cases did not dis-
close significant survival differences between HER2
copy numbers of ≥4 and HER2 copy numbers of <4
cases (Table S6). Instead, we identified high HER2
copy numbers (≥8) more powerful to predict longer
survival. Although this finding is different from previ-
ous concepts, it may provide prognostic information.
In conclusion, HER2 IHC scores can provide prog-
nostic information for patients diagnosed as HER2-
type invasive breast cancer. Although anti-HER2
therapy is currently indicated for both HER2 IHC 3+
© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.

and IHC 2+/ISH+ patients, the prognosis of IHC 3+
patients is significantly better than that of IHC 2+/
ISH+ patients. Of the IHC 2+/ISH+ patients we exam-
ined, those with high HER2 copy numbers in reflex
ISH had better predicted OS than those with low
HER2 copy numbers.
Acknowledgements
The authors wish to thank Ya-Lin Yang, Ming-Jhu
Pan, Li-Rung Liao, Yi-Wen Chang, Shu-Ying Wang,
Hsiu-Hsun Ma, Chiu-Yuan Ma and Huei-Mei Chen for
their technical assistance. This work was supported
by grants from Taipei Veterans General Hospital
(V106C-037).
Conflicts of interest
The authors declare no conflicts of interest.
References
1. Wolff AC, Hammond ME, Hicks DG et al. Recommendations for
human epidermal growth factor receptor 2 testing in breast
cancer: American Society of Clinical Oncology/College of Amer-
ican Pathologists clinical practice guideline update. Arch.
Pathol. Lab. Med. 2014; 138; 241–256.
2. Rakha EA, Pinder SE, Bartlett JMS et al. Updated UK recom-
mendations for HER2 assessment in breast cancer. J. Clin.
Pathol. 2015; 68; 93–99.
3. Allred DC. Issues and updates: evaluating estrogen receptor-
alpha, progesterone receptor, and HER2 in breast cancer. Mod.
Pathol. 2010; 23(Suppl. 2); S52–S59.
4. Scaltriti M, Nuciforo P, Bradbury I et al. High HER2 expression
correlates with response to the combination of lapatinib and
trastuzumab. Clin. Cancer Res. 2015; 21; 569–576.
5. Partridge AH, Gelber S, Piccart-Gebhart MJ et al. Effect of age
on breast cancer outcomes in women with human epidermal
growth factor receptor 2-positive breast cancer: results from a
herceptin adjuvant trial. J. Clin. Oncol. 2013; 31; 2692–2698.
6. Zurawska U, Baribeau DA, Giilck S et al. Outcomes of HER2-
positive early-stage breast cancer in the trastuzumab era: a
population-based study of Canadian patients. Curr. Oncol.
2013; 20; e539–e545.
7. Loi S, Dafni U, Karlis D et al. Effects of estrogen receptor and
human epidermal growth factor receptor-2 levels on the effi-
cacy of trastuzumab: a secondary analysis of the HERA trial.
JAMA Oncol. 2016; 2; 1040–1047.
8. Borley A, Mercer T, Morgan M et al. Impact of HER2 copy
number in IHC2+/FISH-amplified breast cancer on outcome of
adjuvant trastuzumab treatment in a large UK cancer network.
Br. J. Cancer 2014; 110; 2139–2143.
9. Swain SM, Baselga J, Kim SB et al. Pertuzumab, trastuzumab,
and docetaxel in HER2-positive metastatic breast cancer. N.
Engl. J. Med. 2015; 372; 724–734.
10. Hammond ME, Hayes DF, Dowsett M et al. American Society of
Clinical Oncology/College of American Pathologists guideline
© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.
Prognostic value of HER2 IHC score 585
recommendations for immunohistochemical testing of estrogen
and progesterone receptors in breast cancer. Arch. Pathol. Lab.
Med. 2010; 134; 907–922.
11. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic mod-
els: issues in developing models, evaluating assumptions and
adequacy, and measuring and reducing errors. Stat. Med.
1996; 15; 361–387.
12. Middleton LP, Price KM, Puig P et al. Implementation of Amer-
ican Society of Clinical Oncology/College of American Patholo-
gists HER2 guideline recommendations in a tertiary care
facility increases HER2 immunohistochemistry and fluores-
cence in situ hybridization concordance and decreases the
number of inconclusive cases. Arch. Pathol. Lab. Med. 2009;
133; 775–780.
13. Dowsett M, Procter M, McCaskill-Stevens W et al. Disease-free
survival according to degree of HER2 amplification for patients
treated with adjuvant chemotherapy with or without 1 year of
trastuzumab: the HERA Trial. J. Clin. Oncol. 2009; 27; 2962–
2969.
14. Zabaglo L, Stoss O, Ruschoff J et al. HER2 staining intensity in
HER2-positive disease: relationship with FISH amplification and
clinical outcome in the HERA trial of adjuvant trastuzumab.
Ann. Oncol. 2013; 24; 2761–2766.
15. Perez EA, Reinholz MM, Hillman DW et al. HER2 and chromo-
some 17 effect on patient outcome in the N9831 adjuvant
trastuzumab trial. J. Clin. Oncol. 2010; 28; 4307–4315.
16. Veeraraghavan J, De Angelis C, Reis-Filho JS et al. De-escala-
tion of treatment in HER2-positive breast cancer: determinants
of response and mechanisms of resistance. Breast 2017; 34
(Suppl. 1); S19–S26.
17. Perez EA, Suman VJ, Davidson NE et al. HER2 testing by local,
central, and reference laboratories in specimens from the North
Central Cancer Treatment Group N9831 intergroup adjuvant
trial. J. Clin. Oncol. 2006; 24; 3032–3038.
18. Wolff AC, Hammond ME, Schwartz JN et al. American Society
of Clinical Oncology/College of American Pathologists guideline
recommendations for human epidermal growth factor receptor
2 testing in breast cancer. Arch. Pathol. Lab. Med. 2007; 131;
18–43.
19. Fuchs E-M, K€ostler WJ, Horvat R et al. High-level ERBB2 gene
amplification is associated with a particularly short time-to-
metastasis, but results in a high rate of complete response once
trastuzumab-based therapy is offered in the metastatic setting.
Int. J. Cancer 2014; 135; 224–231.
20. Wolff AC, Hammond MEH, Allison KH et al. Human epidermal
growth factor receptor 2 testing in breast cancer: American
Society of Clinical Oncology/College of American Pathologists
clinical practice guideline focused update. Arch. Pathol. Lab.
Med. 2018; 142; 1364–1382. https://doi.org/10.5858/arpa.
2018-0902-SA.
Supporting Information
Additional supporting information may be found
online in the Supporting Information section at the
end of the article:
Table S1. Subgroup multivariate survival analysis
stratified by anti-HER2 therapy and HER2 IHC scores.
Table S2. Subgroup multivariate survival analysis
stratified by hormone receptor status and HER2 IHC
scores.
586 Y-F Tsai et al.
Table S3. Prognostic values of the HER2 ISH
results by different cutoff values using univariate Cox
model.
Table S4. Subgroup multivariate survival analysis
stratified by anti-HER2 therapy and HER2 ISH
results.
Table S5. Subgroup multivariate survival analysis
stratified by hormone receptor status and HER2 ISH
results.
Table S6. Prognostic values (Harrell’s C-index) of
the reflex HER2 ISH testing in 202 HER2 IHC 2+
cases.
Figure S1. Survival curves of patients receiving
neoadjuvant and adjuvant anti-HER2 therapy.
Figure S2. Subgroup survival analyses stratified by
anti-HER2 therapy and HER2 IHC scores.
Figure S3. Subgroup survival analyses stratified by
hormone receptor status and HER2 IHC scores.
© 2018 John Wiley & Sons Ltd, Histopathology, 74, 578–586.