Progress in pathology
Metallographic in situ hybridizationB
Richard D. Powella,*, James D. Pettayb,c, William C. Powelld, Patrick C. Roched,
Thomas M. Grogand, James F. Hainfelda, Raymond R. Tubbsb,c
aNanoprobes, Incorporated, 95 Horseblock Road, Unit 1, Yaphank, NY 11980, USA
bCleveland Clinic Foundation, Division of Pathology and Laboratory Medicine, Cleveland, OH 44195, USA
cCleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland, OH 44195, USA
dVentana Medical Systems, Incorporated, 1910 Innovation Park Drive, Tucson, AZ 85755, USA
Received 23 March 2007; revised 30 April 2007; accepted 1 May 2007
Summary Metallographic methods, in which a target is visualized using a probe or antibody that deposits
metal selectively at its binding site, offers many advantages for bright-field in situ hybridization (ISH)
detection as well as for other labeling and detection methods. Autometallographically enhanced gold
labeling procedures have demonstrated higher sensitivity than conventional enzyme chromogens.
Enzyme metallography, a novel procedure in which an enzymatic probe is used to deposit metal directly
from solution, has been used to develop bright-field ISH methods for HER2 gene determination in breast
cancer and other biopsy specimens. It provides the highest level of sensitivity and resolution, both for
visualizing endogenous gene copies in nonamplified tissues and for resolving multiple gene copies to
allow copy enumeration in amplified tissues without the need for oil immersion or fluorescence optics.
An automated enzyme metallography procedure, silver ISH, has been developed for use in slide-staining
instruments. Metallographic staining also provides excellent results for immunohistochemistry and may
be combined with other staining procedures for the simultaneous detection of more than one gene or
combinations of genes and proteins.
© 2007 Published by Elsevier Inc.
In situ hybridation;
Automated slide staining;
1.1. Historical development of in situ
hybridization detection methods
After the determination of the structure of DNA , it was
not until 15 years later that methods for localizing DNA and
RNA at the molecular level began to emerge, using
complementary sequences as probes. The development of
the method was driven by the discovery that chemical
denaturation procedures such as high temperature or sodium
hydroxide treatment could alter the staining properties of
different regions along chromosomes, leading to the
☆This project was supported by Small Business Technology Transfer
Program (STTR) grant 5R42CA83618-03 and Small Business Innovation
Research (SBIR) grant 1R43CA111182-01 from the National Cancer
Institute (National Institutes of Health) to Nanoprobes, Inc., Yaphank, NY,
and Cleveland Clinic Lerner College of Medicine, Cleveland, OH, and by
grant number 5R44 GM064257-03 from the National Institute of General
Medical Sciences (National Institutes of Health) to Nanoprobes, Inc., and
Cleveland Clinic Lerner College of Medicine. Its contents are solely the
responsibility of the authors and do not necessarily represent the official
views of the National Cancer Institute or the National Institute of General
E-mail address: email@example.com (R. D. Powell).
0046-8177/$ – see front matter © 2007 Published by Elsevier Inc.
Human Pathology (2007) 38, 1145–1159
development of banding profiles that could be used to
identify and localize chromosomal features . Differential
staining with fluorescent dyes was used to elucidate banding
structures within chromosomes . The genesis of modern in
situ hybridization (ISH) was the discovery that specific
sequences localized to these bands. Initially, detection used
radioisotopically labeled RNA probes to localize comple-
mentary DNA (cDNA) in tissue sections and cell spreads
using autoradiography: [4-6] in one of the three earliest
reports, Pardue and Gall  used thymidine-3H–labeled
RNA fractions extracted from cultured mouse and Xenopus
cellular extracts to localize cDNA in denatured tissue
sections. Although cumbersome and limited in resolution
by the scattering cone of radioactive emissions, these early
experiments effectively demonstrated the localization of
specific DNA sequences to chromosomal regions.
The technique of ISH achieved wider use as a number of
enabling technologies were developed. Improved probe
preparation and labeling methods, including random prime
labeling, nick translation reaction, polymerase chain reaction
(PCR)–based labeling, and improved methods for cloning
target sequences, made probe preparation more convenient
and affordable . The critical development, however, was
the introduction of fluorescent labels, which provide higher
resolution and simpler visualization and avoid the hazards of
radioactive probes and the long delay required to develop the
autoradiographic signal. These enabled the technique of
fluorescence ISH (FISH) . Labeling may be direct, in
which the fluorescent label is linked directly to the nucleic
acid hybridization probe , or indirect, in which this
hybridization probe is bound in turn by a labeled secondary
probe such as fluorescently labeled antibody or protein
targeted to a unique feature or hapten incorporated into the
probe . Another important advantage of FISH is the
capability of imaging multiple targets: the development of a
variety of spectrally distinct fluorescent labels enabled the
simultaneous detection of multiple genetic targets using
differently labeled probes . More recently, sensitivity has
been increased by the application of target or signal
amplification protocols  such as tyramide signal
amplification (TSA), in which either fluorescently labeled
tyramides  or biotin-tyramide, which is then detected
with fluorescently labeled avidin or streptavidin , is
deposited by an enzyme targeted to the gene of interest. FISH
has been used successfully for mapping repetitive and single-
copy DNA sequences on metaphase chromosomes, inter-
phase nuclei, chromatin fibers, and naked DNA molecules,
and for chromosome identification and karyotype analysis
through the localization of large repeated families, typically
the ribosomal DNAs and major tandem array families .
However, one of the most important applications has been
FISH of single-copy DNA sequences, in particular disease-
related genes in humans and other eukaryotic model species
, and the detection of infectious agents [15-17]. The
acceptance of FISH as a clinical molecular diagnostic
method  has established it as an important tool for the
diagnosis of cancer and other medical conditions. However,
despite these advantages, FISH is not always ideal for
clinical diagnostic practice, and in this article, we will
discuss alternative methods for ISH using enzyme metallo-
graphy (EnzMet), which can visualize single genes in a
bright field, requires less complex equipment, enables faster
throughput, and can provide more information for the
1.2. The HER2 gene: significance and status
The detection methods presented in this article are
applicable to the visualization of many genetic targets.
However, because of its medical significance, their initial
application has been directed to the evaluation of the HER2
gene in breast cancer. Breast cancer is the second leading
cause of cancer death in women: the American Cancer
Society predicts that 178,480 new cases of invasive breast
cancer and 62,030 new cases of carcinoma in situ will be
diagnosed in the United States in 2007, and 40,060 women
and 450 men will die from the disease . A woman's
lifetime chance of developing invasive breast cancer is about
1 in 8 . Tumors from individual patients differ in their
biologic behavior and present the pathologist with 2
important challenges: (a) to identify smaller tumors of low
stage, which, although they appear to have a relatively
favorable clinical prognosis, possess phenotypic and geno-
typic characteristics that portend aggressive malignant
behavior; and (b) to select the most efficacious therapeutic
Only about half of breast cancer patients are cured by
local treatment , and therefore, indicators that predict
aggressive malignant behavior are important. The HER2
(ERBB2) proto-oncogene encodes a transmembrane receptor
tyrosine kinase of Mr185 kDa (p185), which is structurally
and functionally analogous to the epidermal growth factor
receptor;  the oncogene is known to stimulate the
transcription of N-acetylglucosaminyltransferase V, resulting
in a 3-fold increase in enzyme activity in neu-transformed
NIH3T3 cells . Amplification of the HER2 oncogene
(also referred to as HER2/neu) and overexpression of the
corresponding HER2-encoded protein are found in about
20% to 30% of invasive breast carcinomas  and correlate
clinically with aggressive malignant behavior in tumors and
poor patient outcomes [25,26], including higher rates of
metastasis [27,28], higher mortality , more rapid
progression, and increased likelihood of recurrence .
Genomic amplification of the HER2 gene in lymph node–
negative invasive breast carcinoma has been associated with
a significant reduction in disease-free survival  and
poorer prognosis in metastatic breast cancer . Amplifica-
tion of the HER2 gene correlates with reduced levels of
estrogen and progesterone receptors, lower natural killer cell
activity , and significantly higher proliferative activity
. A recent study demonstrated that proliferation index
1146 R. D. Powell et al.
and HER2 overexpression were the biologic markers with
the highest predictive prognostic value in minimal breast
HER2 gene amplification is also associated with
resistance to chemotherapy , possibly due to the
increased proliferation . In these cases, humanized
monoclonal antibodies against p185 (Herceptin, trastuzu-
mab) have been shown to be highly effective in both the
metastatic and adjuvant settings [37,38] and even more so
in combination with the antitumor drugs cisplatin  or
paclitaxel . However, in the absence of HER2 over-
expression, antibody therapy is significantly less effective
 and can be hazardous. In combination with conven-
tional chemotherapy, it has been shown to cause cardio-
toxicity: in one study, 28% of patients receiving both
trastuzumab and an anthracycline developed heart failure
. Combination therapies have also been linked to
bleeding . Therefore, determination of HER2 status is
important both for prognosis and therapy selection.
HER2 status in single cells or cell extracts can be
determined by quantitative methods including dot  and
slot blots , Southern blotting with autoradiographic
detection , and competitive or differential PCR relative
to a reference gene that is not amplified . However, the
primary approach used by pathologists is to examine the
expression and distribution of cancer markers in paraffin-
embedded biopsy sections using either immunohistochem-
istry (IHC) or ISH with labeled probes visualized by light
or fluorescence microscopy. This allows the pathologist to
compare the distribution pattern of cancer markers with
other morphological features of the tissues and make a
At this time, EnzMet has been developed for application
in breast carcinoma. However, the opportunity for extension
to in situ detection of other clinically important genes, such
as the EGFR gene copy in lung cancer, is readily apparent.
1.3. IHC or ISH?
HER2 status may be evaluated by assessment of either the
HER2 gene or the degree of expression of the encoded protein.
HER2 protein overexpression is assessed by staining the
expressed protein (p185) in cells and tissues . Image
analysis has been used to quantitate overexpression, and the
results were found to correlate closely with those obtained by
Western blotting  and DNA analysis . An immunohis-
tochemical assay using a polyclonal antibody to the encoded
Administration (FDA) . Although useful, this approach
presents problems with interpretive and technical variability. In
a comparison between quantitative PCR, HercepTest, and
monoclonal antibody (CB11 and TAB250) staining, only 80%
concordance was found. Most of the errors were false-positive,
in which staining indicated apparent overexpression of the
encoded protein, but the gene was not amplified. Of 254 cases,
61 were found to be amplified by PCR; 23% (93%
concordance) were positive with CB11, 27% (90% concor-
dance) with TAB250, and 37% with HercepTest .
ISH offers the significant advantage of enabling quanti-
tative evaluation of HER2 gene amplification: each HER2
accomplished by counting the number of signals per nucleus
[51,52]. FISH studies showed that HER2 amplification was
women  and a stronger discriminator than tumor size in
predicting poor clinical outcome . Two manual FISH
diagnostic assays to detect amplification of the HER2 gene
(Ventana Inform, Ventana Medical Systems Inc, Tucson, AZ,
and Abbott Molecular/Vysis PathVysion, Abbott Molecular/
Vysis, Des Plaines, IL) have been approved by the FDA;
PathVysion has been FDA-approved as a stand-alone test to
select patients for trastuzumab (Herceptin) therapy. Fluores-
cently-labeled DNA gene probes are used to identify
individual HER2 genes in nuclei of formalin-fixed, paraf-
fin-embedded specimens; diagnosis is made by counting the
number of signals observed by fluorescence microscopy.
Comparison of gene amplification, measured by FISH,
with immunohistochemical staining yields significant dis-
cord, which has furthered debate on the significance of
HER2 gene amplification and protein overexpression. This
has important implications for therapy selection ,
especially given the approval of Herceptin for use in
adjuvant setting in combination with docetaxel [56-59]. A
recent study found that although CB11 monoclonal antibody
IHC results showed concordance with FISH (κ = 0.83),
polyclonal antibodies may not be sufficiently accurate and
consistent for reliable assessment of HER2 protein over-
expression . Currently, at least 10 different antibodies are
available for commercial purchase, with differing perfor-
mance characteristics in terms of cell conditioning require-
ments (antigen retrieval)  and reported interpretive
patterns (cytoplasmic versus cell membrane or both) :
a more reliable indicator must be included for full and
accurate diagnosis and informed therapeutic decisions
regarding the use of Herceptin. Two comparison studies
found that the most significant differences between IHC and
FISH were found for cases with no gene amplification but
significant positive staining for protein overexpression. In
the first, the difference was attributed to an alternative
mechanism for overexpression, such as transcriptional
activation . In a second study of 52 cases, the 2 FISH
assays gave 98% concordance, with 14 cases showing
amplification and 37 no amplification, whereas the IHC
assay detected HER2 overexpression in a larger proportion
of cases: 13 high-positive and 13 medium-positive. Of the 13
high-positive cases, 10 also showed gene amplification, but 9
of the 13 medium-positive cases did not , confirming that
IHC tends to produce false-positive results. The possibility
that HER2 protein was being overexpressed without gene
amplification was tested using messenger RNA (mRNA)
hybridization . Four hundred breast biopsies were
1147Metallographic in situ hybridization
evaluated using monoclonal (CB11) and polyclonal (Her-
cepTest) antibodies; 145 were subsequently evaluated by
FISH and 144 by mRNA. Expression of mRNA was highly
concordant with FISH and digoxigenin-based FISH ampli-
fication. Immunohistochemical false-positive cases (appar-
ent protein overexpression with no HER2 gene
amplification) occurred with both HercepTest (23%) and
CB11 (17%); most false-positive results (34 of 44) were
scored as 2+. All 2+ false-positive cases were mRNA-
negative, and only 17% had increased gene copy. Therefore,
it was concluded that the 2+ score as defined in the
guidelines for the FDA-approved HercepTest should not be
used as a criterion for trastuzumab therapy unless confirmed
The concordance between HER2 gene amplification and
protein overexpression and trastuzumab response has also
been investigated. Overall, both 2-color FISH and the more
accurate of the IHC assays have similar predictive value. In a
study on 88 patients receiving weekly trastuzumab and
paclitaxel therapy for metastatic breast cancer, both IHC (2+
or 3+ score was scored as amplified) and ISH results showed
concordance with response. Overall response rate in patients
with normal HER2 status ranged from 41% to 46%, whereas
responses in HER2-amplified (FISH) and overexpressing
(IHC) patients ranged from 67% to 81%. Tests using the
CB11 and TAB250 antibodies and FISH showed the
strongest significance. Good concordance of the results
between HercepTest and FISH was found for the 78 patients
for whom data from both assays were available, with a κ
value of 0.72 (95% CI, 0.56-0.87); a higher degree of
concordance, indicating excellent agreement, was found
between FISH and the CB11 antibody, with a κ of 0.80 (95%
CI, 0.66-0.93). Of note, 10 patients had tumors categorized
as HER2 overexpressing by the HercepTest and HER2
normal by the CB11 antibody, suggesting that the less
accurate HercepTest antibody tends to produce false-positive
results . In a second study on 26 women receiving
weekly docetaxel with trastuzumab therapy for metastatic
breast cancer, patients with tumors that were either HER2 3+
by HercepTest or FISH-positive had response rates in the
63% to 65% range; all HER2 3+ patients were also FISH-
positive. Patients with tumors that were HER2 2+ or FISH-
negative had response rates ranging from 0% to 25%. The
only HER2 2+ patient who achieved a positive response was
positive by FISH. These data suggest that tumors found to be
HER2 2+ by HercepTest should be tested for HER2
amplification by FISH: only if the tumor is positive by
FISH should the patient subsequently receive trastuzumab
because FISH-negative patients are unlikely to respond .
Determination of HER2 gene status by FISH may be the
more accurate and reproducible method for selecting patients
eligible for trastuzumab therapy. However, a panel recently
convened by the American Society of Clinical Oncology
(ASCO) and the College of American Pathologists (CAP)
affirmed that a firm conclusion as to the superiority of FISH
over IHC could not be drawn from the studies performed
thus far, primarily because most studies have not measured
clinical outcome, and the few clinical outcome–based
studies published did not use IHC reagents that are in
current common use and FDA-approved .
ISH assessment of HER2 status can yield 3 results:
nonamplified, with 2 or fewer gene copies per cell;
amplified, with more than 6 or 8 gene copies; and low-
level amplified, generally with 3 to 6 copies per cell.
Interpretation of low-level amplification requires considera-
tion of polysomy of chromosome 17, on which the HER2
gene resides, which is a significant cause of false-positive
results. Genuine amplification is therefore defined not in
terms of the number of copies of the HER2 gene but by the
ratio of HER2 to chromosome 17 centromere copies. This is
established by using either a reflexed second test or an
internal, visually distinct centromeric chromosome 17 probe
. The PathVysion system includes a second color
centromeric chromosome 17 probe as an internal control:
HER2 amplification is based on a ratio of HER2 to
chromosome 17 of equal to or greater than 2.0 . A
recent study compared HER2 status from IHC and both
FISH assays for 2279 cases and found discordant FISH
results in 89 tumors: this was due to chromosome 17
polysomy in 57 cases classified as positive by Inform but
negative by PathVysion, and chromosome 17 monosomy in
32 cases classified as positive by PathVysion but negative by
Inform . Although chromosome 17 polysomy in such
cases is usually regarded as a false-positive result, the full
clinical and prognostic significance of these conditions and
of genuine low-level amplification is not fully understood:
for example, it has been reported that a net increase in HER2
gene copy number consecutive to polysomy 17 in the
absence of specific gene amplification might lead to a strong
protein overexpression in a small subset of breast carcinomas
. Therefore, in cases with apparent low-level amplifica-
tion, this control is an important requirement [72-74].
1.4. Bright-field ISH methods
Although the fluorescence-based PathVysion FISH pro-
cedure is FDA-approved and reliable, it has disadvantages
for the pathologist. It requires fluorescent optics, which
require dark adaptation on the part of the observer and are not
immediately available in many laboratories. The need for
high magnification and, sometimes, oil immersion add to the
complexity of the procedure. Furthermore, morphological
changes and standard counterstained morphology cannot
readily be visualized simultaneously, making it difficult to
evaluate the amplification in a cellular context. Fluorescently
labeled specimens are subject to fading, hindering permanent
archival. A reliable, sensitive bright-field colorimetric or
chromogenic modification of this assay would be highly
desirable for most practicing pathologists.
In response to this need, chromogenic ISH (CISH)
procedures were developed, using enzymatic labeling with
conventional organic chromogens. Tanner et al  used a
1148 R. D. Powell et al.
digoxigenin-labeled HER2 probe detected with fluorescein-
labeled antidigoxigenin followed by peroxidase-labeled anti-
fluorescein developed with diaminobenzidine (DAB) for
HER2 gene status in a series of 157 breast carcinomas;
amplification of HER2 was identified with CISH in 27
(17.2%) cases. However, 10 tumors in that series were FISH-
positive but CISH-negative. Sensitivity was increased by the
(HRP) reagents [76-78] or peroxidase-avidin-biotin com-
plexes [79,80] for detection. A series of 119 cases was
analyzed using CISH with polymerized secondary detection:
low-level amplified (6-10 copies), or high-level amplified
(more than 10 copies). A repeated staining procedure using a
chromosome 17 centromeric probe was used to confirm
amplificationfor cases with3to10copies.Withtheexception
of 1 discordant case that was classified as low-level amplified
by FISH but amplified by CISH , the FISH and CISH
results agreed completely. A second study used a tissue
or borderline amplified cases where differences might be
significant . One hundred two cases were successfully
analyzed by both methods. A 100% concordance was found
between FISH and CISH in classifying tumors as HER2
amplified or nonamplified: 63 cases were scored as amplified
and 39 as not amplified. The amplified tumors were classified
further into 3 categories: borderline amplification (HER2-to-
CEP17 ratio, 2.0-2.5), low-level amplification (ratio, 2.6-3.9),
22 (35%) were scored as borderline (6 cases) or low-level (16
also borderline by CISH, whereas 2 were scored low-level
(ratios, 2.7 and 3.2), and 1 case was high-level (ratio, 4.1). Of
the 16 low-level amplification cases by FISH, 6 also were
considered low-level by CISH, whereas 1 case was borderline
(ratio, 2.5), and 9 cases were high-level (ratio range, 4.0-6.8)
by CISH. Overall, the FISH and CISH ratios were correlated
highly (r = 0.825). A procedure using TSA has been run on
an automated platform . However, the characteristics of
the staining are not ideal: diffusion and the continuous nature
of the stain can make the spots difficult to resolve and
obscure the underlying ultrastructure. In one recent compar-
ison, FISH was still occasionally able to detect low-level
amplification missed using a chromogenic method .
A recent report presents an efficient protocol for dual-
color CISH (dc-CISH), using cohybridization of a digox-
igenin-labeled HER2 probe and biotinylated chromosome 17
centromeric probe . The chromosome 17 centromeric
probe was detected first, by means of incubation with mouse
antibiotin antibody followed by alkaline phosphatase poly-
mer kit developed with new fuchsin to give a red color; after
washing with distilled water, detection of the digoxigenin-
labeled HER2 probe was performed with antidigoxigenin
antibody and HRP polymer secondary developed with
tetramethyl benzidine chromogen to give a blue-green signal.
In a comparison of the results for 4 cell lines and 40 tumor
samples with those obtained using FISH (PathVysion), the
results of FISH and dc-CISH showed high concordance
(91%, κ = 0.82), indicating that dc-CISH, a new alternative
to FISH, enables the assessment of copy number ratio
(HER2/17 centromere) in conjunction with proper histo-
pathologic evaluation and the ease of bright-field micro-
scopy. The features and performance characteristics of FISH
and CISH are compared with those of the metallographic
methods that are the subject of this article in Table 1.
2. Application of metallographic methods
2.1. Colloidal gold and Nanogold
Colloidal gold labeling, in which antibodies or other
proteins conjugated to gold nanoparticles 1 to 30 nm in
diameter are used to label targets of interest, was first
introduced as an immunoelectron microscopic method
[85,86] and later adapted for ISH electron microscopic
chromosome , satellite , and gene localization .
The subsequent discovery that gold nanoparticles catalyze
the highly selective deposition of silver from mixtures of
silver (I) salts and reducing agents to give specific, highly
resolved, and visually distinct optical staining , a process
called autometallography or silver enhancement, enabled a
new light microscopy method, immunogold-silver staining
. This proved effective for histopathology applications,
where it was found to give superior or at least equal results to
those obtained with the peroxidase antiperoxidase technique;
in some cases, staining was obtained with immunogold-
silver staining method when the peroxidase antiperoxidase
technique gave no result . Streptavidin-colloidal gold
with silver acetate autometallography proved to be a
sensitive and specific method for ISH detection .
Improved results were enabled by Nanogold, a highly
monodisperse 1.4-nm gold particle, exclusively manufac-
tured by Nanoprobes, Incorporated (Yaphank, NY). Nano-
goldmay be covalently linked to a wide variety of molecules,
including antibodies, antibody fragments, proteins, peptides,
oligonucleotides, and lipids . Developed as an improved
alternative to noncovalently linked colloidal gold ,
Nanogold is a stable compound that does not dissociate
, interact nonspecifically with biologic systems to cause
background signal , or aggregate to give poor sample
penetration  as colloidal gold has been found to do.
Nanogold-Fab' antibody conjugates penetrate up to 40 μm
into specimens; in comparison, 10-nm colloidal gold showed
only surface labeling . When used with silver enhance-
ment, Nanogold conjugates may be used to visualize as little
as 0.1 pg (approximately 10-18mol) of a target immunoglo-
bulin G;  this is one of the most sensitive detection
methods available, similar to chemiluminescence.
1149 Metallographic in situ hybridization
When used for ISH detection, silver-enhanced Nanogold
probes demonstrated higher sensitivity than enzymatic
probes run in parallel experiments, both by direct ISH
using silver-enhanced streptavidin-Nanogold to detect a
biotinylated hybridization probe and by an indirect method
in which silver-enhanced streptavidin-Nanogold was used
to detect a biotinylated secondary applied after target
amplification by PCR  or to detect deposited biotin-
tyramide after TSA, initially known as catalyzed reporter
deposition (CARD) [101,102]. Highly sensitive detection
was achieved by hybridization with a fluorescein-labeled
probe, then with a biotinylated antifluorescein antibody,
followed by 2-step signal amplification with avidin-biotin
complex followed by a TSA procedure in which strepta-
vidin-peroxidase was used to deposit biotinylated tyramide.
In the in situ detection of the HPV-16 virus in SiHa cells,
which in the SiHa cell line carries 1 to 2 integrated copies
of HPV 16 per cell, streptavidin-Nanogold combined with
silver acetate autometallography occasionally produced
speckled nuclear staining even without signal amplification
. A panel of 10 cervical cancer specimens, positive by
solution-phase PCR, was examined using both direct and
CARD-amplified Nanogold-streptavidin and streptavidin-
peroxidase. Without CARD, 5 were positive with silver-
enhanced Nanogold but only 2 with peroxidase-DAB. With
CARD, 9 were positive with silver-enhanced Nanogold but
only 5 with DAB (G.W.H. and C.H.-K., unpublished
results). An example of the results obtained with CARD-
amplified Nanogold-streptavidin is shown in Fig. 1.
2.2. Gold enhancement: gold-facilitated ISH
Although the Nanogold-silver enhancement procedure
produced increased sensitivity, the lengthy procedure and
requirement for multiple signal amplification steps made it
cumbersome for routine use. Similar sensitivities were
obtained using a simplified procedure in which gold
of FISH with CISH using organic chromogens and using
different metallographic methods
Comparison showing advantages and disadvantages
FISH• Clear signal• Requires dark
• Requires specialized
• Signal can fade;
must be archived
• Does not show tissue
• Adapted for
• Read with
• Underlying tissue
morphology can be
• Adaptable for
• 2-Color protocol
• Diffuse signals can
reduce ease of
• Signals can be
difficult to distinguish
• Some chromogens
• Higher sensitivity
• Background signal
can be problematic
• Complex protocol
protocols needed for
• Difficult to adapt for
• Read with
• Underlying tissue
• Multicolor protocols
• Highly sensitive• Gene copy
• Difficult to adapt for
• Multicolor protocols
• Easy, rapid visual
• Clean signal with
• Read with
• Underlying tissue
morphology can be
Table 1 (continued)
EnzMet• Highest sensitivity • Multicolor protocols
not yet available (may
be in future)
• Highest resolution
• Most accurate
• Best visualization
• Adaptable for
• Simple procedure
1150 R. D. Powell et al.
enhancement replaced silver enhancement. Gold enhance-
ment is a novel gold-based autometallographic process,
developed at Nanoprobes, in which gold, instead of silver, is
catalytically deposited onto gold particles . It demon-
strates several advantages over silver enhancement: it may be
used in physiologic buffers, including halides, which
precipitate silver; it is not light sensitive; and it is more
selective—for example, gold enhancement can be carried out
in cells cultured on metal substrates . Gold enhance-
ment used with Nanogold for ISH after TSA gave sensitivity
equivalent to silver enhancement, with lower background
. When the experiment was conducted using a
combined Cy3 and Nanogold-labeled streptavidin, signals
from HPV16/18 in SiHa cells were visualized by both
fluorescence and bright-field light microscopy. In addition,
the high atomic number of the gold allows localization at
higher resolution by electron microscopy .
A simplified gold-enhanced Nanogold-streptavidin
method, called Gold-facilitated ISH (GOLDFISH), was
developed to detect amplification of the HER2 gene in
paraffin-embedded breast carcinoma sections. Fig. 2 shows
a schematic for GOLDFISH, and a comparison of staining
in normal (nonamplified) tissue (2 copies of the HER2
gene) and amplified carcinoma tissue (10-20 copies) .
The GOLDFISH assay demonstrated high interobserver
reproducibility: for a series of 66 excisional biopsies or
resections, the mean κ among GOLDFISH observers was
0.84, and the mean among GOLDFISH observers versus
direct FISH was 0.83, which was at least concordant of
observers scoring nuclear grade (κ = 0.50) and the presence
of in situ carcinoma (κ = 0.57) by conventional histopathol-
ogy. Two specimens that scored 2+ or 3+ by IHC but
demonstrated chromosome 17 polysomy without amplifica-
tion by 2-color FISH were appropriately assigned as
nonamplified by GOLDFISH . Staining is opaque,
black, and punctated with sharply defined edges. The assay
was designed for qualitative interpretation without the need
for oil immersion: in amplified tissue, large, confluent
nuclear signals were generated, whereas endogenous HER2
copies in nonamplified tissue produced small, punctuate
signals. Preliminary results have verified that GOLDFISH
staining may be quantitated using automated image analysis
. Signal clarity, sensitivity, and reproducibility were
significantly better than equivalent detection methods based
on enzymatic deposition of organic chromogens .
The black signal is permanent, does not fade, and is
easily distinguished from overall section color and from
2.3. Enzyme metallography
The GOLDFISH assay was configured for simplified,
rapid manual interpretation without spot counting: the
autometallographic process was optimized to give large,
confluent signals in amplified tissuerather than discrete spots
that could be counted. However, with the elucidation of the
gene amplification signals, it became apparent that spot
counting was necessary to differentiate genuine low-level
amplification from chromosome 17 polysomy [70-74,81,82].
In addition, the growing significance of image analysis
(automated spot counting) rendered a quantitative interpreta-
tion method preferable to the more qualitative method used
Subsequently, it was discovered that HRP can selectively
deposit metal from solution even in the absence of a
particulate metal nucleating agent such as Nanogold. This
by antibody-peroxidase-ABC complex. This is then treated with biotinylated tyramide: the peroxidase then catalytically converts the tyramide
to a reactive free radical species, which quickly reacts with the aromatic ring of a nearby tyrosine residue and anchors the biotin there. When
sufficient biotin is deposited, it is detected with Nanogold-labeled streptavidin and silver acetate autometallography (Society for Applied
Immunohistochemistry; used with permission). Right, Single copies of HPV-16 in SiHa cells detected by CARD-amplified Nanogold-silver
ISH. Copies of HPV-16 appear as single spots (hematoxylin-eosin counterstain, original magnification ×560).
CARD–Nanogold-silver ISH. Left, A fluorescein-labeled hybridization probe is detected using an antifluorescein antibody, followed
1151 Metallographic in situ hybridization
resulted in highly sensitive and specific metal deposition on
targets. This new method, termed EnzMet [110,111], was
found to produce superior results both for detecting HER2
A hapten-labeled hybridization probe is bound by a primary
antibody against the hapten then visualized using a polymerized
peroxidase-labeled secondary reagent treated with the enzyme
metallographic substrate, which deposits metallic silver at the active
site. B, Endogenous HER2 signals in nonamplified tissue; 1 to 2
HER2 signals are clearly visualized per nucleus. C, HER2 in highly
amplified tissue, showing multiple, distinct signals corresponding
to individual gene copies. B and C both stained using Zymed
polymeric antibody-HRP reagents (Invitrogen, Carlsbad, CA;
hematoxylin counterstain, original magnification ×400).
EnzMet assay. A, Schematic showing reaction sequence.
matic of assay. A biotinylated HER2 hybridization probe is detected
using peroxidase-streptavidin followed by biotin-tyramide, which is
catalytically deposited at the target site; the biotin-tyramide is
detected with Nanogold-streptavidin followed by gold enhance-
ment. B, GOLDFISH staining of nonamplified tissue: infiltrating
ductal carcinoma with a normal HER2 gene copy, showing 1 to 2
spots per nucleus, corresponding to endogenous HER2 (NFR
counterstain, original magnification ×400). C, GOLDFISH proce-
dure in tissue with HER2 gene amplification, showing large,
confluent nuclear signals from multiple gene copies in close
proximity (NFR counterstain, original magnification ×400; Amer-
ican Society for Investigative Pathology; used with permission).
GOLDFISH HER2 gene amplification assay. A, Sche-
1152R. D. Powell et al.
gene amplification [112,113] and for immunohistochemi-
cally staining overexpressed HER2 protein in paraffin
sections of human breast carcinoma .
ISH using a biotin-labeled HER2 probe with mouse
antibiotin followed by polymerized goat antimouse–HRP,
developed with metallographic substrate, yielded very high
sensitivity and resolution. Individual spots corresponding to
gene copies were clearly visible in both amplified and
nonamplified tissues and could be enumerated in an identical
manner to the fluorescent signals obtained by FISH; single
genes developed into approximately 1-μm dense metal spots,
and background was almost nonexistent [112,113]. Fig. 3
shows the method with representative examples. A compar-
ison of the staining pattern obtained by EnzMet with that
observed for DAB is shown in Fig. 3. A series of 101 cases
was assessed using this method versus FISH. Three cases
were low-level amplified by absolute HER2 copy number but
found to be nonamplified by 2-color FISH because of chro-
mosome 17 polysomy;  similar cases are easily resolved
by conducting a second ISH procedure on a second slide.
2.4. Adaptation to automated staining systems:
EnzMet protocols share many steps with conventional
enzyme chromogenic detection methods, and therefore,
adaptation of this chemistry to automated platforms would
seem relatively straightforward. Differences in technology
betweensteps thatcanbe performedmanually andthose used
in automated assays can cause issues during assay develop-
ment. However, once the assay has been adapted to
automation, it yields improvements in efficiency and work-
flow advantages that make automated silver ISH (SISH) a
routine assay with increased throughput and high-quality
results that are easily interpreted. The following description
Systems, Inc, with the BenchMark Series of instruments.
3. Practical aspects of automated
The automation of any assay is primarily dependent on
the technology of the instrument used, and these can vary
significantly from manufacturer to manufacturer. The bulk
reagent catalog available for the instrument is also another
consideration. However, the key aspects of any ISH protocol
(independent of manufacturer) fall into 5 main categories:
sample preparation, tissue pretreatment, hybridization,
detection, and counterstaining.
3.1. Sample preparation
As with any tissue-based assay, the quality of tissue
collection, fixation, and processing can make the difference
between obtaining an appropriately stained and interpre-
table slide and one that cannot produce a result. Generally,
the time between tissue collection and fixation affects the
quality of the morphology: excessive times can have
deleterious effects on the quality of the DNA to be
hybridized. The type of fixative used can greatly affect the
results. For instance, metal-containing fixatives (B5,
Zenker's, and zinc formalin) have the potential to cause
background. Acidic fixatives (Bouin's) can have significant
effects on the intactness of DNA, depending on the length
of fixation . Formalin-based (non–metal-containing)
fixatives have produced the best results. It is also important
that tissue be completely fixed; in tissue that is too large to
fix in the appropriate amount of time, the center may
undergo autolysis, causing loss of morphology and a loss
After deparaffinization and rehydration of the tissue on
the slide, access to the DNA is the next critical
parameter. There are 2 primary methods to permeabilize
the tissue: heating the tissue in a retrieval solution and
protease digestion. Most manual commercial assays use
both methods to obtain optimal access to the DNA. The
correct amount of tissue conditioning (retrieval and
protease treatment) is important because undercondition-
ing can result in weak and patchy signal, whereas
overconditioning can actually strip the DNA from the
slide, resulting in little or no signal. The highly controlled
conditions of the automated assay remove most of the
variability from this process and result in a consistent
sample for probe hybridization.
Hybridization is almost entirely a function of the probe
used, size of the probe fragments, Tm, and hybridization
buffer. Probe design, labeling, and formulation constitute a
wide field in themselves; a detailed discussion would be
the subject of another review and will not be addressed
here. The primary drawback to the FDA-approved manual
assays is the need for overnight hybridization, which limits
the throughput to 1 run per day. The new automated HER2
SISH assay has decreased the overnight hybridization to
approximately 2 hours, enabling multiple runs per day on a
The adaptation of EnzMet to automated instrumenta-
tion was relatively straightforward. The most important
consideration was the specificity of the probe. Because of
the very high sensitivity of the SISH detection technol-
ogy, probe specificity and stringency washes become
critical to eliminate specific and nonspecific background.
1153 Metallographic in situ hybridization
Particularly with detection of single-copy genes, any
nonspecific precipitation of silver on the slide can create
interpretation issues for the pathologist. Thus, as we
discuss in the section on interpretation, it is very
important that nonspecific background be eliminated.
The final step in the automated ISH process is the
counterstain. The deposition of silver on the slide creates a
black signal, and the ease of interpretation is dependent on
the level of contrast between the silver and the counter-
stain. We have found that hematoxylin and nuclear fast
red (NFR) counterstains produce a slide with the needed
contrast for interpretation. Although it may be suspected
that the black signal with the light blue hematoxylin
nuclear stain might have insufficient contrast, as long as
the hematoxylin is configured as a counterstain and not at
the intensity of a hematoxylin-eosin stain, the results are
highly scorable. Hematoxylin also has the advantage of
providing nuclear detail for pathologic identification of the
neoplastic cells, which is not possible with FISH-based
assays. NFR provides a higher contrast pattern with the
SISH signal that allows for rapid identification of nuclei
and signal; however, NFR counterstain does lack some of
the discreetness of hematoxylin with regard to nuclear area
and the potential use of image analysis for automated
interpretation with an instrument such as the Ventana
Image Analysis System (VIAS; Ventana).
4.1. Low-level amplification: significance of CEP17
There is debate in the field of HER2 testing as to the
need for correction of HER2 copy number for chromo-
some 17 (CEP17) copy number. The recently published
ASCO/CAP guidelines for HER2 testing include in their
suggested algorithm protocols for interpretation with and
without CEP17 . The concern about including CEP17
testing is that HER2 testing of a polysomic 17 case can
look like amplified HER2 when, in fact, HER2 has not
been specifically amplified. This controversy has been
addressed in multiple publications [72-74]. The HER2
SISH assay was developed as a 2-slide assay, 1 for the
HER2 probe and a second for the CEP17 probe, and it
is the ratio of HER2 to CEP17 that determines the
amplification status of the case. It is very important that
cases in the area of the FDA-approved 2.0 cut-off are
quantitatively assessed. This analysis was addressed in the
ASCO/CAP guidelines, and an “equivocal” range was set
for ISH applications to indicate the uncertainty surround-
ing the actual cutoff. The ASCO/CAP guidelines direct
that these equivocal cases have additional cells counted or
that the assay be rerun or confirmed by a second
independent assay such as IHC. There is little information
on the response rates of individuals who fall within the
ISH equivocal and IHC equivocal ranges. Cases who
exhibit high-level amplification are easily scored by light
microscopy at magnifications as low as ×20, whereas
nonamplified cases are typically scored at either ×40 or
×60 without oil immersion.
4.2. Signal morphology
The HER2 SISH signal has several manifestations
depending on the presence or absence of amplification.
Nonamplified signal is typically represented by small,
discreet dots in the nucleus. There can be 1 to 2 per cell,
but there will not be signal in every single nucleus because in
many cases, only a portion of a nucleus falls within the
section, and this may not contain the target sequence. With
amplification, 2 patterns of hybridization are typically seen.
The first are homogeneously staining regions, which are
manifested by 2 large clusters of SISH signals that indicate
repeated copies of HER2 on a single chromosome .
These homogeneously staining region–type cases are usually
scored by an estimation of the copy number based on the
SISH signal size in the normal adjacent cells. The second
pattern seen is a more distributed pattern that lends itself to
counting of individual signals. This pattern is indicative of a
more random amplification throughout the genome or
restriction of the amplified genes to double-minute chromo-
5. Future developments
5.1. Application of metallographic methods to IHC
In addition to its utility for ISH, EnzMet also produces
excellent results when used for immunohistochemical
staining [110,118]. A comparison between conventional
DAB-peroxidase and EnzMet was conducted using high-
complexity tissue midiarrays; 88 solid tumors were evaluated
by automated EnzMet IHC (Nanoprobes and Ventana) .
Targets were chosen to assess the ability of EnzMet IHC to
specifically localize encoded antigens in the nucleus
(estrogen receptor), cytoplasm (cytokeratins), and cytoplas-
mic membrane (HER2). Full concordance was found
between the EnzMet IHC and conventional IHC; further-
more, the EnzMet staining was sharper and more highly
resolved, with many regions assuming a punctuate appear-
ance that defined the staining more clearly. In addition, the
black color of the EnzMet reaction product is easily
differentiated from other stains and counterstains, and the
sensitivity of the detection procedure allows additional
dilution of the primary antibodies.
1154 R. D. Powell et al.
5.2. Combining ISH with IHC: a concomitant gene
and protein assay
The feasibility of combining ISH and IHC in a single
protocol was previously demonstrated in the concomitant
oncoprotein detection with FISH method , a 3-color
fluorescence-based system that simultaneously profiles
HER2 oncogene copy by FISH and HER2 encoded protein
by the use of a versatile alkaline phosphatase chromogen,
fast red K, in either fluorescence or bright-field mode.
denoting HER2 gene copies are present in the nucleus, with associated 3+ cellular membrane expression of the encoded HER2 protein
(hematoxylin counterstain, original magnification ×1000). Arrow denotes endogenous HER2 copies in lymphocytes. B, Breast carcinoma case
nonamplified for HER2 by FISH. Tumor cells demonstrate 1 or 2 nuclear HER2 signals (black) and absence of HER2 encoded protein on the
cell surface (hematoxylin counterstain, original magnification ×400; inset, original magnification ×1000). C, Breast carcinoma case polysomic
for chromosome 17 but not amplified for HER2 (HER2-to-CEP17 ratio, 1.9). Tumor cells demonstrate 3 to 5 discrete black signals per nucleus,
denoting HER2 gene copy number, and the cellular membrane is devoid of HER2-encoded protein expression (hematoxylin counterstain,
original magnification ×400; inset, original magnification ×1000). D, Breast carcinoma case amplified for HER2 by FISH but demonstrating
only FDA 1+ immunostaining by IHC. Numerous confluent black metallic silver signals denoting HER2 gene copies are present in the
nucleus, but associated 1+ cellular membrane expression of the encoded HER2 protein is negative by FDA scoring criteria (hematoxylin
counterstain, original magnification ×1000; Lippincott Williams and Wilkins; used with permission).
EnzMet GenePro assay. A, Breast carcinoma case amplified for HER2 by FISH. Numerous confluent black metallic silver signals
1155 Metallographic in situ hybridization
Fast red K protein IHC may be readily combined with
enzyme metallographic ISH detection without either
component interfering with the other. For the concomitant
gene and protein assay (GenePro), HER2 protein was first
stained using CB11 polyclonal antibody followed by
alkaline phosphatase-labeled secondary and fast red K
development. HER2 gene copies were then visualized using
a DNP-labeled cDNA probe, detected with a polymerized
HRP-antibody conjugate, and developed using EnzMet
substrate. The enzyme metallographic procedure did not
affect the fast red K stain . With the exception of the
ISH and the enzyme metallographic development, all steps
were automated (Benchmark, Ventana). The method was
validated on tissue midiarrays , consisting of 94 cores
from 80 cases of invasive breast carcinoma, including 4
cases of infiltrating lobular carcinoma, 1 case of mixed
infiltrating ductal and lobular carcinoma, and 89 cases of
infiltrating ductal carcinoma.
Staining results are shown in Fig. 4. The mean HER2
copy number count, produced from the enzyme metallo-
graphic component of the assay when performed in concert
with the IHC component, was compared with the direct FISH
HER2 copy number and showed an excellent correlation,
with a Pearson coefficient of 0.96 (P b .001). The IHC
component of the GenePro assay showed a Pearson
correlation of 0.84 (P b .001) in comparison with CB11
antibody staining performed on whole tissue sections. The
combination of gene and protein detection (EnzMet
GenePro) gave a specificity of 100% and an accuracy of
92.6% (95% confidence interval, 85.3-97.0), using the
consensus diagnosis for each component of the assay.
Correlation between the IHC component of the assay and
the HER2 gene amplification status of the assay had
agreement in 98% (92/94) of the cases. Most HER2
amplified cases had CB11 staining that was 2+ or 3+,
whereas HER2 nonamplified or polysomic cases had CB11
staining that was 0 or 1+. The 2 discrepant cases included a
CB11-negative and EnzMet-positive case and a CB11-
positive and EnzMet-negative case. In a 5-person inter-
observer reproducibility study, interobserver κ for each
component was excellent (IHC, κ = 0.94; EnzMet, κ = 0.96).
5.3. Multiple targets and multiple colors
Until recently, the application of multiple staining
protocols, particularly ISH, to a single specimen has been
prohibitively complex for most applications because of the
complexity of the procedures involved. However, the
development of automated slide staining systems for
applications such as immunohistochemical slide staining
 and ISH  has relieved their demands on personnel
time and enabled the widespread adoption of increasingly
complex procedures in reliable, standardized formats. There-
fore, the incorporation of metallographic protocols into
research is directed toward this goal.
Our longer-term goal is to develop technologies for
multiplexed correlated molecular morphological labeling and
profiling of tumors. Evaluation of the relative intensity and
spatial distribution of a particular set of tumor markers with
independent predictive or prognostic value will provide a
“personalized medicine” and individualized therapy selection.
In addition, it may also provide tools for cancer research and
applications such as bright-field spectral karyotyping. Future
studies may identify additional predictive value, such as a
correlation between a specific result and outcome from a
combination therapy. Ultimately, this methodology will be
applicableto mayother types oftumors[124,125] and to other
medical conditions, including the recent work of Dr Ricardo
Lloyd on the in situ detection of infection agents [15-17],
where differences in biologic behavior or therapeutic response
are related to different markers or infection agents.
for deoxyribose nucleic acids. Nature (London) 1953;171:737-8.
 de Jong H. Visualizing DNA domains and sequences by
microscopy: a fifty-year history of molecular cytogenetics. Genome
 Caspersson T, Lindsten J, Lomakka G, Moller A, Zech L. The use of
fluorescence techniques for the recognition of mammalian chromo-
somes and chromosome regions. Int Rev Exp Pathol 1972;11:1-72.
 Pardue ML, Gall JG. Molecular hybridization of radioactive DNA to
the DNA of cytological preparations. Proc Natl Acad U S A Sci 1969;
 John HA, Birnstiel ML, Jones KW. RNA-DNA hybrids at the
cytological level. Nature 1969;223:582-7.
 Buongiorno-Nardelli M, Amaldi F. Autoradiographic detection of
molecular hybrids between RNA and DNA in tissue sections. Nature
 Trask BJ. Fluorescence in situ hybridization: applications in
cytogenetics and gene mapping. Trends Genet 1991;7:149-54.
 Bauman JG, Wiegant J, Borst P, van Duijn P. A new method for
fluorescence microscopical localization of specific DNA sequences
by in situ hybridization of fluorochromelabelled RNA. Exp Cell Res
 Rudkin GT, Stollar BD. High resolution detection of DNA-RNA
hybrids in situ by indirect immunofluorescence. Nature 1977;265:
 Nederlof PM, van der Flier S, Wiegant J, et al. Multiple fluorescence
in situ hybridization. Cytometry 1990;11:126-31.
 Qian X, Lloyd RV. Recent developments in signal amplification
methods for in situ hybridization. Diagn Mol Pathol 2003;12:1-13.
 van de Corput MP, Dirks RW, van Gijlswijk RP, van de Rijke FM,
Raap AK. Fluorescence in situ hybridization using horseradish
peroxidase–labeled oligodeoxynucleotides and tyramide signal
amplification for sensitive DNA and mRNA detection. Histochem
Cell Biol 1998;110:431-7.
 Kerstens HM, Poddighe PJ, Hanselaar AG. A novel in situ
hybridization signal amplification method based on the deposition
of biotinylated tyramine. J Histochem Cytochem 1995;43:347-52.
 Tsuda H. HER2 (c-erbB-2) test update: present status and problems.
Breast Cancer 2006;13:236-48.
 Qian X, Guerrero RB, Plummer TB, Alves VF, Lloyd RV. Detection
of hepatitis C virus RNA in formalin-fixed paraffin-embedded
1156R. D. Powell et al.
sections with digoxigenin-labeled cRNA probes. Diagn Mol Pathol
 Abdulkarim AS, Petrovic LM, Kim WR, Angulo P, Lloyd RV, Lindor
KD. Primary biliary cirrhosis: an infectious disease caused by
Chlamydia pneumoniae? J Hepatol 2004;40:380-4.
 Hayden RT, Isotalo PA, Parrett T, et al. In situ hybridization for the
differentiation of Aspergillus, Fusarium, and Pseudallescheria species
in tissue section. Diagn Mol Pathol 2003;12:21-6.
 Masood S, Bui MM, Yung JF, et al. Reproducibility of LSI HER-2/
neu SpectrumOrange and CEP 17 SpectrumGreen dual color
deoxyribonucleic acid probe kit. For enumeration of gene amplifica-
tion in paraffin-embedded specimens: a multicenter clinical valida-
tion study. Ann Clin Lab Sci 1998;28:215-23.
 Detailed guide: breast cancer. What are the key statistics for breast
cancer? American Cancer Society, http://www.cancer.org/docroot/
breast_cancer_5.asp?sitearea= (Revised September 18, 2006).
 Nofech-Mozes S, Spayne J, Rakovitch E, Hanna W. Prognostic and
predictive molecular markers in DCIS: a review. Adv Anat Pathol
 McGuire WL. Prognostic factors for recurrence and survival in
human breast cancer. Breast Cancer Res Treat 1987;10:5-9.
 Schechter AL, Stern DF, Vaidyanathan L, et al. The neu oncogene: an
erb-B–related gene encoding a 185,000-Mr tumour antigen. Nature
 Chen L, Zhang W, Fregien N, Pierce M. The her-2/neu oncogene
stimulates the transcription of N-acetylglucosaminyltransferase Vand
expression of its cell surface oligosaccharide products. Oncogene
 Slamon DJ, Godolphin W, Jones LA, et al. Studies of the HER-2/neu
proto-oncogene in human breast and ovarian cancer. Science 1989;
 Toikkanen S, Helin H, Isola J, Joensuu H. Prognostic significance of
HER-2 oncoprotein expression in breast cancer: a 30-year follow-up.
J Clin Oncol 1992;10:1044-8.
 Descotes F, Pavy JJ, Adessi GL. Human breast cancer: correlation
study between HER-2/neu amplification and prognostic factors in an
unselected population. Anticancer Res 1993;13:119-24.
 Makar AP, Desmedt EJ, De Potter CR, Vanderheyden JS, Schatteman
EA. Neu (C-erbB-2) oncogene in breast cancer and its possible
association with the risk of distant metastases. A retrospective study
and review of literature. Acta Oncol 1990;29:931.
 Tiwari RK, Borgen PI, Wong GY, Cordon-Cardo C, Osborne MP.
HER-2/neu amplification and overexpression in primary human
breast cancer is associated with early metastasis. Anticancer Res
 Seshadri R, Horsfall DJ, Firgaira F, et al. The relative prognostic
significance of total cathepsin D and HER-2/neu oncogene
amplification in breast cancer. The South Australian Breast Cancer
Study Group. Int J Cancer 1994;56:61-5.
 Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire
WL. Human breast cancer: correlation of relapse and survival with
amplification of the HER-2/neu oncogene. Science 1987;235:
 Allred DC, Clark GM, Tandon AK, et al. HER-2/neu in node-
negative breast cancer: prognostic significance of overexpression
influenced by the presence of in situ carcinoma. J Clin Oncol 1992;
 Chang J, Clark GM, Allred DC, Mohsin S, Chamness G, Elledge
RM. Survival of patients with metastatic breast carcinoma:
importance of prognostic markers of the primary tumor. Cancer
 Zeillinger R, Kury F, Czerwenka K, et al. HER-2 amplification,
steroid receptors and epidermal growth factor receptor in primary
breast cancer. Oncogene 1989;4:109-14.
 Bacus SS, Zelnick CR, Plowman G, Yarden Y. Expression of the
erbB-2 family of growth factor receptors and their ligands in breast
cancers. Implication for tumor biology and clinical behavior. Am J
Clin Pathol 1994;102:S13-24.
 Pegram MD, Finn RS, Arzoo K, Beryt M, Pietras RJ, Slamon DJ. The
effect of HER-2/neu overexpression on chemotherapeutic drug
sensitivity in human breast and ovarian cancer cells. Oncogene
 Fornier M, Risio M, Van Poznak C, Seidman A. HER2
testing and correlation with efficacy of trastuzumab therapy.
Oncology (Huntingt). 2002;16:1340-8,
trastuzumab therapy. Oncology (Huntingt). 2002;16:1576, 1578
 Pegram MD, Slamon DJ. Combination therapy with trastuzumab
(Herceptin) and cisplatin for chemoresistant metastatic breast cancer:
evidence for receptor-enhanced chemosensitivity. Semin Oncol 1999;
26(4 Suppl 12):89-95.
 Fornier MN, Seidman AD, Schwartz MK, et al. Serum HER2
extracellular domain in metastatic breast cancer patients treated with
weekly trastuzumab and paclitaxel: association with HER2 status by
immunohistochemistry and fluorescence in situ hybridization and
with response rate. Ann Oncol 2005;16:234-9.
 Feldman AM, Lorell BH, Reis SE. Trastuzumab in the treatment of
metastatic breast cancer: anticancer therapy versus cardiotoxicity.
 Nissenblatt MJ, Karp GI. Bleeding risk with trastuzumab (Herceptin)
treatment. JAMA 1999;282:2299-301.
 Chen Y, Dong J, Lu Y, McGee JO. Quantitative detection of
amplification of proto-oncogenes in breast cancer. Chin Med J (Eng)
 Donovan-Peluso M, Contento AM, Tobon H, Ripepi B, Locker J.
Oncogene amplification in breast cancer. Am J Pathol 1991;138:
 Valeron PF, Chirino R, Fernandez L, et al. Validation of a differential
PCR and an ELISA procedure in studying HER-2/neu status in breast
cancer. Int J Cancer 1996;65:129-33.
 van de Vijver MJ, Mooi WJ, Wisman P, Peterse JL, Nusse R.
Immunohistochemical detection of the neu protein in tissue sections
of human breast tumors with amplified neu DNA. Oncogene 1988;2:
 Molina R, Ciocca DR, Tandon AK, et al. Expression of HER-2/neu
oncoprotein in human breast cancer: a comparison of immunohisto-
chemical and western blot techniques. Anticancer Res 1992;12:
 Press MF, Pike MC, Chazin VR, et al. Her-2/neu expression in node-
negative breast cancer: direct tissue quantitation by computerized
image analysis and association of overexpression with increased risk
of recurrent disease. Cancer Res 1993;53:4960-70.
 Birner P, Oberhuber G, Stani J, et al. Austrian Breast & Colorectal
Cancer Study Group. Evaluation of the United States Food and
Drug Administration–approved scoring and test system of HER-2
protein expression in breast cancer. Clin Cancer Res 2001;7:
 O'Malley FP, Parkes R, Latta E, et al. Comparison of HER2/neu
status assessed by quantitative polymerase chain reaction and
immunohistochemistry. Am J Clin Pathol 2001;115:504-11.
 Pauletti G, Godolphin W, Press MF, Slamon DJ. Detection and
quantitation of HER-2/neu gene amplification in human breast cancer
archival material using fluorescence in situ hybridization. Oncogene
 Fernandez JL, Goyanes V, Lopez-Fernandez C, Buno I, Gosalvez J.
Quantification of C-ERB-B2 gene amplification in breast cancer cells
using fluorescence in situ hybridization and digital image analysis.
Cancer Genet Cytogenet 1996;86:18-21.
 Press MF, Bernstein L, Thomas PA, et al. HER-2/neu gene
amplification characterized by fluorescence in situ hybridization:
poor prognosis in node-negative breast carcinomas. J Clin Oncol
1157Metallographic in situ hybridization
 Xing WR, Gilchrist KW, Harris CP, Samson W, Meisner LF. FISH
detection of HER-2/neu oncogene amplification in early onset breast
cancer. Breast Cancer Res Treat 1996;39:203-12.
 Gonzalez-Angulo AM, Hortobagyi GN, Esteva FJ. Adjuvant therapy
with trastuzumab for HER-2/neu–positive breast cancer. Oncologist
 Romond EH, Perez EA, Bryant J, et al. Trastuzumab plus adjuvant
chemotherapy for operable HER2-positive breast cancer. N Engl J
 Joensuu H, Kellokumpu-Lehtinen PL, Bono P, et al. Adjuvant
docetaxel or vinorelbine with or without trastuzumab for breast
cancer. N Engl J Med 2006;354:809-20.
 Coudert BP, Arnould L, Moreau L, et al. Pre-operative systemic (neo-
adjuvant) therapy with trastuzumab and docetaxel for HER2-
overexpressing stage II or III breast cancer: results of a multicenter
phase II trial. Ann Oncol 2006;17:409-14.
study of neoadjuvant docetaxel, vinorelbine, and trastuzumab
followed by surgery and adjuvant doxorubicin plus cyclopho-
sphamide in women with human epidermal growth factor receptor
2–overexpressing locally advanced breast cancer. J Clin Oncol 2007;
 Gouvea AP, Milanezi F, Olson SJ, Leitao D, Schmitt FC, Gobbi H.
Selecting antibodies to detect HER2 overexpression by immunohis-
tochemistry in invasive mammary carcinomas. Appl Immunohisto-
chem Mol Morphol 2006;14:103-8.
 Baak JP, Chin D, van Diest PJ, Ortiz R, Matze-Cok P, Bacus SS.
Comparative long-term prognostic value of quantitative HER-2/neu
protein expression, DNA ploidy, and morphometric and clinical
features in paraffin-embedded invasive breast cancer. Lab Invest
 Press MF, Hung G, Godolphin W, Slamon DJ. Sensitivity of HER-2/
neu antibodies in archival tissue samples: potential source of error in
immunohistochemical studies of oncogene expression. Cancer Res
 Johnson RC, Ricci Jr A, Cartun RW, Ackroyd R, Tsongalis GJ.
p185HER2 overexpression in human breast cancer using molecular
and immunohistochemical methods. Cancer Invest 2000;18:336-42.
 Wang S, Saboorian MH, Frenkel E, Hynan L, Gokaslan ST, Ashfaq
R. Laboratory assessment of the status of Her-2/neu protein and
oncogene in breast cancer specimens: comparison of immunohis-
tochemistry assay with fluorescence in situ hybridisation assays. J
Clin Pathol 2000;53:374-81.
 Tubbs RR, Pettay JD, Roche PC, Stoler MH, Jenkins RB, Grogan
TM. Discrepancies in clinical laboratory testing of eligibility for
trastuzumab therapy: apparent immunohistochemical false-positives
do not get the message. J Clin Oncol 2001;19:2714-21.
 Seidman AD, Fornier MN, Esteva FJ, et al. Weekly trastuzumab and
paclitaxel therapy for metastatic breast cancer with analysis of
efficacy by HER2 immunophenotype and gene amplification. J Clin
 Tedesco KL, Thor AD, Johnson DH, et al. Docetaxel combined with
trastuzumab is an active regimen in HER-2 3+ overexpressing and
fluorescent in situ hybridization–positive metastatic breast cancer: a
multi-institutional phase II trial. J Clin Oncol 2004;22:1071-7.
 Wolff AC, Hammond ME, Schwartz JN, et al. American Society of
Clinical Oncology; College of American Pathologists. American
Society of Clinical Oncology/College of American Pathologists
guideline recommendations for human epidermal growth factor
receptor 2 testing in breast cancer. J Clin Oncol 2007;25:118-45.
 Press MF, Slamon DJ, Flom KJ, Park J, Zhou JY, Bernstein L.
Evaluation of HER-2/neu gene amplification and overexpression:
comparison of frequently used assay methods in a molecularly
characterized cohort of breast cancer specimens. J Clin Oncol 2002;
 Lal P, Salazar PA, Hudis CA, Ladanyi M, Chen B. HER-2 testing
in breast cancer using immunohistochemical analysis and
fluorescence in situ hybridization: a single-institution experience
of 2,279 cases and comparison of dual-color and single-color
scoring. Am J Clin Pathol 2004;121:631-6 [Erratum in: Am J
Clin Pathol, 2004;6:58.
 Ma Y, Lespagnard L, Durbecq V, et al. Polysomy 17 in HER-2/neu
status elaboration in breast cancer: effect on daily practice. Clin
Cancer Res 2005;11:4393-9.
 Persons DL, Tubbs RR, Cooley LD, et al. HER-2 fluorescence in situ
hybridization: results from the survey program of the College of
American Pathologists. Arch Pathol Lab Med 2006;130:325-31.
 Dal Lago L, Durbecq V, Desmedt C, et al. Correction for chro-
mosome-17 is critical for the determination of true Her-2/neu gene
 Vera-Roman JM, Rubio-Martinez LA. Comparative assays for the
HER-2/neu oncogene status in breast cancer. Arch Pathol Lab Med
 Tanner M, Gancberg D, Di Leo A, et al. Chromogenic in situ
hybridization: a practical alternative for fluorescence in situ
hybridization to detect HER-2/neu oncogene amplification inarchival
breast cancer samples. Am J Pathol 2000;157:1467-72.
 Dandachi N, Dietze O, Hauser-Kronberger C. Chromogenic in situ
hybridization: a novel approach to a practical and sensitive method
for the detection of HER2 oncogene in archival human breast
carcinoma. Lab Invest 2002;82:1007-14.
 Gupta D, Middleton LP, Whitaker MJ, Abrams J. Comparison of
fluorescence and chromogenic in situ hybridization for detection of
HER-2/neu oncogene in breast cancer. Am J Clin Pathol 2003;119:
 Isola J, Tanner M, Forsyth A, Cooke TG, Watters AD, Bartlett JM.
Interlaboratory comparison of HER-2 oncogene amplification as
detected by chromogenic and fluorescence in situ hybridization. Clin
Cancer Res 2004;10:4793-8.
 Peiro G, Mayr D, Hillemanns P, Lohrs U, Diebold J. Analysis of
HER-2/neu amplification in endometrial carcinoma by chromogenic
in situ hybridization. Correlation with fluorescence in situ hybridiza-
tion, HER-2/neu, p53 and Ki-67 protein expression, and outcome.
Mod Pathol 2004;17:227-87.
 Gong Y, Gilcrease M, Sneige N. Reliability of chromogenic in situ
hybridization for detecting HER-2 gene status in breast cancer:
comparison with fluorescence in situ hybridization and assessment of
interobserver reproducibility. Mod Pathol 2005;18:1015-21.
carcinoma: a comparison of chromogenic in situ hybridization with
fluorescence in situ hybridization in tissue microarrays, with targeted
Immunohistochem Mol Morphol 2005;13:194-200.
 Bhargava R, Lal P, Chen B. Chromogenic in situ hybridization for the
detection of HER-2/neu gene amplification in breast cancer with an
emphasis on tumors with borderline and low-level amplification:
does it measure up to fluorescence in situ hybridization? Am J Clin
 Nitta H, Christensen K, Grogan TM. Automated colormetric
tyramide signal amplification in situ hybridization (TISH) detection
of single HER-2/neu oncogene copies. Acta Histochem Cytochem
 Laakso M, Tanner M, Isola J. Dual-colour chromogenic in situ
hybridization for testing of HER-2 oncogene amplification in archival
breast tumours. J Pathol 2006;210:3-9.
 Faulk WP, Taylor GM. An immunocolloid method for the electron
microscope. Immunochemistry 1971;8:1081-3.
 Handley DA. The development and application of colloidal gold as a
microscopic probe; methods for synthesis of colloidal gold. In: Hayat
MA, editor. Colloidal gold: principles, methods and applications. San
Diego, CA: Academic Press; 1989. p. 1-33.
 Wu M, Davidson N. Transmission electron microscopic method for
gene mapping on polytene chromosomes by in situ hybridization.
Proc Natl Acad USA Sci 1981;78:7059-63.
1158R. D. Powell et al.
 Hutchison NJ, Langer-Safer PR, Ward DC, Hamkalo BA. In situ Download full-text
hybridization at the electron microscope level: hybrid detection by
autoradiography and colloidal gold. J Cell Biol 1982;95:609-18.
 Ferguson DJ, Burns J, Harrison D, Jonasson JA, McGee JO.
Chromosomal localization of genes by scanning electron microscopy
using in situ hybridization with biotinylated probes: Y chromosome
repetitive sequences. J Histochem 1986;18:266-70.
 Danscher G. Localization of gold in biological tissue. A photo-
chemical method for light and electronmicroscopy. Histochemistry
 Danscher G, Norgaard JO. Light microscopic visualization of
colloidal gold on resin-embedded tissue. J Histochem Cytochem
 Hacker GW, Springall DR, Van Noorden S, Bishop AE, Grimelius L,
Polak JM. The immunogold-silver staining method. A powerful tool
in histopathology. Virchows Arch A Pathol Anat Histopathol 1985;
 Hacker GW, Graf AH, Hauser-Kronberger C, et al. Application of
silver acetate autometallography and gold-silver staining methods for
in situ DNA hybridization. Chin Med J (Engl) 1993;106:83-92.
 Hainfeld JF, Powell RD. New frontiers in gold labeling. J Histochem
 Hainfeld JF. STEM analysis of Janssen Auroprobe One. In: Bailey
GW, editor. Proc. XII Int. Cong. Elec. Microsc. San Francisco, CA:
San Francisco Press; 1990. p. 954-5.
 Behnke O, Ammitzboll T, Jessen H, et al. Non-specific binding of
protein-stabilized gold sols as a source of error in immunocytochem-
istry. Eur J Cell Biol 1986;41:326-38.
 Kramarcy NR, Sealock R. Commercial preparations of colloidal
gold–antibody complexes frequently contain free active antibody. J
Histochem Cytochem 1990;39:37-9.
 Sun XJ, Tolbert LP, Hildebrand JG. Using laser scanning confocal
microscopy as a guide for electron microscopic study: a simple
method for correlation of light and electron microscopy. J Histochem
 Hainfeld JF, Furuya FR. Silver enhancement of Nanogold and
Undecagold. In: Hayat MA, editor. Immunogold-silver staining:
principles, methods and applications. Boca Raton, FL: CRC Press;
1995. p. 71-96.
 Hacker GW, Zehbe I, Hainfeld J, et al. High-performance Nanogold
in situ hybridization and in situ PCR. Cell Vision 1996;3:209-14.
 Zehbe I, Hacker GW, Su H, Hauser-Kronberger C, Hainfeld JF,
Tubbs R. Sensitive in situ hybridization with catalyzed reporter
deposition, streptavidin-Nanogold, and silver acetate autometallo-
graphy: detection of single-copy human papillomavirus. Am J Pathol
 Hacker GW, Hauser-Kronberger C, Zehbe I, et al. In situ localization
of DNA and RNA sequences: super-sensitive in situ hybridization
using streptavidin-Nanogold-silver staining: minireview, protocols
and possible applications. Cell Vision 1997;4:54-65.
 Hainfeld JF, Powell RD. Silver- and gold-based autometallography of
Nanogold. In: Hacker GW, Gu J, editors. Gold and silver staining:
techniques in molecular morphology. Boca Raton: CRC Press; 2002.
 Owen GR, Meredith DO, Ap Gwynn I, Richards RG. Enhancement of
immunogold-labelled focal adhesion sites in fibroblasts cultured on
metal substrates: problems and solutions. Cell Biol Int 2001;25:1251-9.
 Cheung AL, Graf AH, Hauser-Kronberger C, Dietze O, Tubbs RR,
Hacker GW. Detection of human papillomavirus in cervical
carcinoma: comparison of peroxidase, Nanogold, and catalyzed
reporter deposition (CARD)–Nanogold in situ hybridization. Mod
 Powell RD, Joshi VN, Halsey CMR, et al. Combined Cy3/Nanogold
conjugates for immunocytochemistry and in situ hybridization. In:
BaileyGW, Jerome WG, McKernan S, Mansfield JF, Price RL, editors.
Microsc Microanal, 5 (Suppl. 2: Proceedings); 1999. p. 478-9.
 Tubbs R, Pettay J, Skacel M, et al. Gold-facilitated in situ
hybridization: a bright-field autometallographic alternative to
fluorescence in situ hybridization for detection of Her-2/neu gene
amplification. Am J Pathol 2002;160:1589-95.
 Tubbs R, Skacel M, Pettay J, et al. Interobserver interpretative
reproducibility of GOLDFISH, a first generation gold-facilitated
autometallographic bright field in situ hybridization assay for HER-2/
neu amplification in invasive mammary carcinoma. Am J Surg Pathol
 Hicks D, Ruddy T, Downs-Kelly E, et al. Quantitative image analysis
of GOLDFISH, (a first generation gold-facilitated autometallographic
bright field in-situ hybridization assay) for HER2 gene amplification
in invasive breast cancer. 26th San Antonio Breast Cancer
Symposium; 2003; poster #611.
 Hainfeld JF, Eisen RN, Tubbs RR, Powell RD. Enzymatic
metallography: a simple new staining method. In: Lyman CE,
Albrecht RM, Carter CB, Dravid VP, Herman B, Schatten H, editors.
Microsc Microanal, 8 (Suppl. 2: Proceedings); 2002. p. 916 CD.
 Hainfeld JF, Powell RD, Hacker GW. Nanoparticle molecular labels.
In: Mirkin CA, Niemeyer CM, editors. Nanobiotechnology. Ger-
many: Wiley-VCH, Weinheim; 2004. p. 353-86. [ch. 23].
 Tubbs R, Pettay J, Hicks D, et al. Novel bright field molecular
morphology methods for detection of HER2 gene amplification. J
Mol Histol 2004;35:589-94.
 Tubbs RR, Pettay J, Skacel M, et al. Gold and silver-facilitated
metallographic in situ hybridization procedures for detection of
HER2 gene amplification. In: Hacker GW, Tubbs RR, editors.
Molecular morphology in human tissues. Boca Raton, FL: CRC
Press; 2005. p. 101-6. [ch. 5].
 Tubbs R, Pettay J, Powell R, et al. Enzyme metallography (EnzMet):
a novel second generation bright field assay for detection of HER2
gene amplification in breast carcinoma. J Mol Diagn (AMP 2003
Meeting Abstracts) 2003;5:279.
 Bonin S, Petrera F, Rosai J, Stanta G. DNA and RNA obtained from
Bouin's fixed tissues. J Clin Pathol 2005;58:313-6.
 Sato T, Abe K, Kurose A, Uesugi N, Todoroki T, Sasaki K.
Amplification of the c-erbB-2 gene detected by FISH in gastric
cancers. Pathol Int 1997;47:179-82.
 Ishikawa T, Kobayashi M, Mai M, Suzuki T, Ooi A. Amplification
of the c-erbB-2 (HER-2/neu) gene in gastric cancer cells. Detection
by fluorescence in situ hybridization. Am J Pathol 1997;151:761-8.
of subcellular compartments in tissue microarrays by enzyme
metallography. Appl Immunohistochem Mol Morphol 2005;13:371-5.
 Tubbs RR, Pettay J, Roche P, et al. Concomitant oncoprotein
detection with fluorescence in situ hybridization (CODFISH): a
fluorescence-based assay enabling simultaneous visualization of gene
amplification and encoded protein expression. J Mol Diagn 2000;2:
 Downs-Kelly E, Pettay J, Hicks D, et al. Analytical validation and
interobserver reproducibility of EnzMet GenePro: a second-genera-
tion bright-field metallography assay for concomitant detection of
HER2 gene status andprotein expression in invasive carcinoma of the
breast. Am J Surg Pathol 2005;29:1505-11.
 Skacel M, Skilton B, Pettay JD, Tubbs RR. Tissue microarrays: a
powerful tool for high-throughput analysis of clinical specimens: a
review of the method with validation data. Appl Immunohistochem
Mol Morphol 2002;10:1-6.
 Myers J. Automated slide stainers for SS, IHC, and ISH: A review of
current technologies and commercially available systems. MLO Med
Lab Obs 2004;36:28-30.
 Rimsza LM, Vela EE, Frutiger YM, et al. Rapid automated combined
in situ hybridization and immunohistochemistry for sensitive
detection of cytomegalovirus in paraffin-embedded tissue biopsies.
Am J Clin Pathol 1996;106:544-8.
 Lloyd RV. Practical markers used in the diagnosis of neuroendocrine
tumors. Endocr Pathol 2003;14:293-301.
 Lloyd RV. Advances in pituitary pathology: use of novel techniques.
Front Horm Res 2004;32:146-74.
1159Metallographic in situ hybridization