ambrosinispaltro andrea tesi

ambrosinispaltro andrea tesi
Alma Mater Studiorum – Università di Bologna
DOTTORATO DI RICERCA IN
ONCOLOGIA E PATOLOGIA SPERIMENTALE
PROGETTO N. 1 ONCOLOGIA
Ciclo XXIV
Settore Concorsuale di afferenza: 06/A4
Settore Scientifico disciplinare: MED/08
MARCATORI IMMUNOISTOCHIMICI E MOLECOLARI AD IMPATTO
PROGNOSTICO E PREDITTIVO DI RISPOSTA ALLA TERAPIA IN
AMBITO NEOPLASTICO
IMMUNOHISTOCHEMICAL AND MOLECULAR
PROGNOSTIC/PREDICTIVE MARKERS IN NEOPLASTIC DISEASES
Presentata da:
Dott. Andrea Ambrosini Spaltro
Coordinatore Dottorato
Relatore
Prof. Sandro Grilli
Prof. Maria P. Foschini
Esame finale anno 2012
CONTENTS
INTRODUCTION ..................................................................................................................3
AIM S OF THE PROJECT ...................................................................................................10
IGFBP2 AS A DIAGNOSTIC MARKER IN PROSTATIC ADENOCARCINOMAS.......... 11
HEPARAN SURFACE PROTEOGLYCANS IN ORAL SQUAMOUS CELL
CARCINOMAS AS PROGNOSTIC MARKERS AND THEIR PREDICTIVE ROLE TO
ADJUVANT RADIOTHERAPY ......................................................................................... 22
EGFR MUTATION-SPECIFIC ANTIBODIES IN PULMONARY ADENOCARCINOMA:
A COMPARISON WITH DNA DIRECT SEQUENCING ................................................... 39
REFERENCES ..................................................................................................................... 49
2
INTRODUCTION
Biological markers or biomarkers are gaining increasing importance in clinical practice. They
can be used in many different conditions and for many different purposes. As defined by the
Biomarkers Definition Working Group in Bethesda, a biomarker is “a characteristic that is
objectively measured and evaluated as an indicator of normal biological processes, pathogenic
processes, or pharmacologic responses to a therapeutic intervention”(1). Biomarkers can be
analyzed in the serum, in the body fluids, and in tissue specimens, the last being of particular
importance in pathology since all this material is stored in the pathological Archives.
In pathology departments and laboratories, many traditional morphological examinations are
carried out, especially in neoplastic diseases. Nevertheless, many of the classical
morphological analyses, including staging, grading, vascular invasion and assessment of the
surgical margins, are not enough anymore to correctly report a tumor. Several subsequent
clinical decisions are made upon biomarkers that the pathological report should evaluate. A
paradigm example is represented by the mammary neoplasia, in which a complete evaluation
of estrogen and progesterone receptors is of mandatory importance in correctly addressing
patients to tamoxifen therapy (2). Many biomarkers are currently under study, and their
discovery and subsequent validation may help in better define their patients. Unfortunately,
although many markers are discovered, little of them are currently validated for clinical
practice. One of the reasons is that many of them showed tremendous variations from the
different studies and cannot be consistently applied. An international Committee from the
Statistics Subcommitee of the National Cancer Institute-European Organization for Research
and Treatment of Cancer has developed guidelines, referred to as REMARK, for the reporting
of tumor marker studies (3). In these guidelines, it is clearly expressed the importance of good
study design and data quality. Poor study reporting has many negative consequences over the
research community as a whole. Inappropriately analyzed studies may deserve
disproportionate clinical attention, just because the results are apparently dramatic; on the
contrary, carefully designed studies may not attract so much consideration, even if they were
appropriately conducted, but did not so produce so impressive results (3). Also technical
3
problems may obstacle the results in a validation process of a biomarker, and in particular
immunohistochemical markers may be affected by preanalytical and analytical problems (4).
Preanalytical problems include the time to collection, details of fixation, dehydration steps
and conditions for paraffin-embedding. Analytical issues are represented, most of all, by:
antigen retrieval techniques, type of detection system, choice of the antibody and material to
be used (5).
Biomarkers may be subdivided on the basis of their application:
As a diagnostic tool: they help in identifying patients with a disease (diagnostic markers)
As a tool for staging a disease or classification of the extent of disease (prostate specific
antigen, PSA in the blood)
As an indicator of disease prognosis (prognostic markers)
As an indicator of response to a specific target therapy (predictive markers)(1).
From a technical point of view, in pathology, biomarkers may be classified in:
Immunohistochemical markers
Molecular markers
Diagnostic markers
Biomarkers with diagnostic purposes are extremely important in correctly identifying patients
carrying the specific disease. Morphological examination may not always reach consistent
results, especially in small biopsies, with artifactual changes or in difficult cases. By
immunohistochemistry, alpha-methylacyl-CoA racemase (AMACR) expression is currently
used as a reliable immunohistochemical diagnostic marker for invasive prostatic carcinoma
(6). In the next sections, we will also explain our study upon the evaluation of a new
diagnostic immunohistochemical marker for prostatic carcinoma, which was compared to
AMACR. Laminin-5 γ2 chain is another important immunohistochemical marker, which may
4
help in identifying invasiveness in colorectal carcinomas (7), squamous cervical carcinomas
(8) and glandular cervical adenocarcinomas (9), especially when there is only focal invasion,
with small aggregates of detached neoplastic cells (budding).
At the molecular level, FISH (Fluorescence Is Situ Hybridization) may play an important role
in establishing monosomy in different chromosomes, by using multiple centromeric probes.
For example, the differential diagnosis between renal oncocytoma and chromophobe renal
cell carcinoma may be extremely difficult with the morphology, even in association with
immunohistochemistry (cytokeratin 7, S100A1) and histochemistry (Hale’s colloidal iron
stain). However, chromphobe renal cell carcinoma frequently exhibits multiple losses among
whole chromosomes 1, 2, 6, 10, and 17 by FISH (10) and by interphase FISH (11). The more
complex karyoptypic abnormalities may then address the diagnosis toward a more malignant
neoplasm, i.e. the chromopohe renal cell carcinoma.
Prognostic markers
Ki67/MIB1 is a typical prognostic marker and it helps in better identifying grade and
biological behavior in many malignancies, especially for those located in the central nervous
system (12).
By gene expression profiling, diffuse large B cell lymphomas (DLBCL) have been
extensively studied by Alizadeh et al, who identified two major groups: one group expressed
genes of germinal centre B cells ('germinal centre B-like DLBCL'), while the other group
expressed genes normally induced during in vitro activation of peripheral blood B cells
('activated B-like DLBCL'); patients with germinal centre B-like DLBCL had a significantly
better overall survival than those with activated B-like DLBCL (13). Similar subgrouping has
been proposed on the basis of the immunohistochemical results, with comparable findings.
5
Predictive markers
Paradigm examples are provided by mammary neoplasia, where high expression of estrogen
and progesterone receptors correlate with tumor responsiveness to the anti-estrogen tamoxifen
(2). Estrogen and progesterone receptors are also prognostic markers, since they are more
expressed in many well differentiated neoplasms, with a favorable prognosis. In mammary
neoplasias, also the amplification of c-erb-B2 is strictly associated with responsiveness to
trastuzumab. Its amplification status is currently determined firstly with an
immunohistochemical test, followed by a molecular test for suspicious cases. Molecular
analysis for amplification of c-cerb-B2 is usually performed with FISH, even if a new dual
ISH base in chromogenic assay has been recently approved by the Food and Drug
Administration (FDA).
Mutated EGFR pulmonary adenocarcinomas may benefit form tyrosine –kinase inhibitor
(TKI) therapy, such as gefitinib (14) and erlotinib (15).
Molecular pathogenesis: the basis to identify new markers (lesson form the colorectal
carcinoma)
The basis for identifying new biological markers is certainly represented by molecular
carcinogenesis models. Carcinogenesis is a multistep process in which many mutations occur:
activation of oncogenes, inactivation of oncosuppressor genes and altered expression of DNA
repair genes. All these molecular modifications determine the loss of growth control from the
neoplastic cells and, consequently, the neoplastic transformation. A classic example is
represented by colorectal carcinoma and its multistep carcinogenesis. In colorectal
carcinogenesis two main molecular pathways are identified (16).
The first is the classical APC/β-catenin pathway which is particularly involved in familial
adenomatous polyposis and activated in 80% of sporadic colorectal carcinomas; this pathway
is characterized by the activation of the Wnt pathway, determines k-ras mutations with EGFR
activation among the first events and causes p53 mutations in advanced stages. The
morphological counterpart is exemplified by the classic adenoma-carcinoma sequence. One of
6
the critical target in this first oncogenetic pathway is the possibility to block the EGFR
signaling induced by k-ras mutations by using anti-EGFR monoclonal antibodies (cetuximab)
(17); so the k-ras status represents an important predictive marker for responsiveness to this
drug. Other important factors extensively studied in this pathway are: p53 and VEGF. P53 has
been shown to be an important prognostic factor in many studies, summarized by metaanalysis reviews (18). VEGF has been described both as prognostic factor (19;20) and as
predictive factor to preoperative radiochemotherapy in rectal carcinomas (21).
The second pathway is the so-called microsatellite instability (MSI) pathway, in which the
DNA mismatch repair genes are damaged. Deficit in these DNA repair genes causes
expansion of microsatellite regions. The familial corresponding disease is the Lynch
syndrome. In sporadic forms, this pathway affects approximately 20% of the colorectal
carcinomas, mainly mucinous adenocarcinomas. The precursor lesion is frequently the sessile
serrated adenoma. These lesions are predominately located in the right colon. Molecularly,
this oncogenetic pathway is characterized by activation of TGFβ and BAX, BRAF mutations,
MLH1 methylation. In translating these basic concepts into clinical practice, several studies
have shown that microsatellite instability is both a prognostic and a predictive marker. A
meta-analysis upon 7642 cases has shown that MSI tumors are associated to a better
prognosis (22). Microsatellite status is also gaining more and more popularity because of its
role as predictive factor to response to adjuvant 5-fluorouracil treatment: a study with 570
cases have attributed to the MSI tumors a less responsiveness to 5-FU treatment (23). We
have previously reported the role of MUC2 as a predictive marker of responsiveness to
radiochemotherapy in rectal adenocarcinoma (24); MUC 2 is associated with mucinous
differentiation and, ultimately, to the MSI.
Colorectal carcinogenesis has been herein described as an example of a carcinogenetic
process. The same process may be applied in many other human malignancies, as in oral
squamous cell carcinoma (OSCC), the central and most important section in this thesis.
7
Statistical consideration
Statistical validation of a biological marker is one of the most crucial points in determining
the real reproducibility of the analyses and its effective application. In many cases, the
statistical analyses are not conducted properly and this may alter the impact of the results.
Altman et al. reviewed all papers which included analyses of survival data and were published
in British Journal of Cancer, European Journal of Cancer, Journal of Clinical Oncology and
American Journal of Clinical Oncology in a 3-month period; they found many problems in
representing statistical data, in reporting all the variables and not only the p-values, in
describing all the parameters in the survival curves and in establishing the cut-off points from
quantitative to qualitative variable transformation (25).
Survival curves are certainly the gold standard method in order to evaluate a prognostic
marker (26). They can be used to study survival, disease-free-survival or any time-dependent
event. With this method the main event on study is time and not the event per se. Graphically,
survival curves can also give an idea to what is the prognostic impact of any disease. One of
the most common applications is the comparison between two (or more) survival curves in
two groups of patients: the “log-rank test”. It is a simple and direct test which may be very
useful in comparing the survival curves between two groups. However, the best analysis
would be taken if the two groups would be randomized (27). Randomization means that the
two groups are almost equal, expect for the only one difference that we want to analyze. One
of the most common mistakes in the medical history is the evaluation of Salk vaccine against
polio in 1954 (28). The study was designed as follows (28).
The plan of procedure announced by the National Foundation for Infantile Paralysis
and its Advisory Committee was to administer vaccine to children in the second grade
of school; the corresponding first and third graders would not be inoculated but would
be kept under observation for the occurrence of poliomyelitis in comparison with the
inoculated second graders. This has been designated the "Observed Control" study.
In observed areas where only those second grade children whose parents requested
participation were vaccinated, the problem of establishing the control population was
more complex.
8
In this study, even if this was one of the biggest trials worldwide at that time, the concept of
randomization was totally lacking. Only the second year school children whose parents gave
consent were vaccinated and the first and third year of all children represented the control
group. The main problems were two. Firstly, the groups were composed of children of
different ages; secondly, children from a poor social background were more exposed to polio
antigen and their parents more favorable to assign them to the vaccine. So children who were
vaccinated were at higher risk to develop the disease. The results showed paradoxically that
Salk vaccine enhanced the risk of developing polio disease! This episode underlines the
importance of a correct randomization in comparing two groups. Ideally, the two groups
should be equal and differ only for the characteristic we would like to study. Unfortunately,
this situation is very difficult to realize in the clinical practice, but enormous problems may be
present when there are many differences between the two groups.
When comparing more variables at the same time, a common procedure is to evaluate
survival function with multivariate analysis; the most common method is the use of Cox
regression (26). The different variables used in the multivariate analysis should be
independent one form the other. Alternatively, many problems and confounding factors may
develop.
9
AIM S OF THE PROJECT
The aim of the project was to identify immunohistochemical and molecular markers which
may be useful in correctly identify neoplastic diseases.
We subdivide the work in three main sections:
1) IGFBP2 as a diagnostic marker in prostatic carcinoma.
2) Heparan-sulfate proteoglycans as prognostic markers and their predictive role to
responsiveness to adjuvant radiotherapy in oral squamous cell carcinomas.
3) EGFR as predictive markers for responsiveness to tyrosine-kinas inhibitors in
pulmonary adenocarcinomas.
10
IGFBP2 AS A DIAGNOSTIC MARKER IN PROSTATIC ADENOCARCINOMAS
Insulin-like growth factors (IGFs) and insulin-like growth factor binding proteins (IGFBPs)
play a central role in cellular growth, and in normal and neoplastic development.
IGFBP2 has been shown to be hyper-expressed in many human malignancies, including
ovarian carcinoma (29), colorectal carcinoma (30), hepatocellular carcinoma (31) and
neuroblastoma (32). In addition, IGFBP2 has been shown to be highly expressed in
glioblastoma, both genetically and immunohistochemically (33). IGFBP2 has even been
considered to be a therapeutic target in neoplastic cell lines derived from breast carcinoma,
both directly (34) and by modulation of the immune system (35). Furthermore, serum levels
of IGFBP2 are reduced in mammary carcinoma (36) and increased in ovarian carcinoma (37),
suggesting its possible role as a serological marker for early diagnosis.
In prostatic tissues, genetic profiling studies documented that IGFBP2 was among the genes
overexpressed in malignant lesions in comparison to normal cases (38-40).
Immunohistochemically, IGFBP2 has been found to be highly reactive in prostatic
adenocarcinoma (PAc) (41-43). IGFBP2 has been proposed as a serum prognostic marker for
patients affected by PAc (44). A significant association between elevated serum levels of
IGFBP2 and the presence of PAc, especially when it is in advanced stages (45-47), has been
documented. Yu et al. have shown that serum levels of IGFBP2 were higher in patients with
remission than in patients with relapse (48). However, Roddam and colleagues (49) and a
meta-analysis conducted by Rowlands et al. (50) did not find a strong association between
IGFBP2 serum levels and prostate cancer risk.
Nevertheless, presently IGFBP-2 is rarely applied in routine diagnoses. Alpha-methylacylCoA racemase (AMACR) is currently used as an immunohistochemical marker for PAc,
especially in biopsies with small acinar lesions suspicious for malignancy (51). However,
AMACR is frequently expressed in high-grade prostatic intraepithelial neoplasia (HG-PIN)
and, to some extent, also in some benign lesions (52), thus sometimes making the
11
interpretation of the immunohistochemical results difficult. Therefore, it would be important
to find other possible diagnostic markers which could help in the diagnosis of malignancy.
The aim of the study was to evaluate the diagnostic value of IGFBP2 expression in normal
epithelium, HG-PIN and PAc, both in patients hormonally untreated and in patients having
undergone complete androgen ablation. Results were compared with Alpha-methylacyl-CoA
racemase (AMACR).
Materials and methods
Sixty prostatectomy specimens were utilized in this study. The specimens represented the
following three groups:

Group 1: 10 consecutive simple prostatectomy specimens from patients with bladder
outlet obstruction due to benign prostatic hyperplasia;

Group 2: 40 consecutive radical prostatectomy specimens with prostatic carcinoma
from patients hormonally untreated before surgery. Preoperative biopsies were
available for all the cases;

Group 3: 10 consecutive radical prostatectomies with prostatic carcinoma from
patients who underwent complete androgen ablation three months before surgery.
Preoperative biopsies were available in all cases.
The cases of Groups 1 and 2 were retrieved from the files of the Section of Anatomic
Pathology “Marcello Malpighi” of the University of Bologna, whereas those of Group 3 were
from the United Hospitals-Polytechnic University of the Marche Region, Ancona. All cases
had been fixed in 10% buffered formalin and paraffin embedded. Five mm thick sections
were stained with hematoxylin and eosin (H&E). For the purpose of this study, the slides of
all cases of the three groups were re-examined by two of the authors (AAS and MPF). The
cases of Group 1 did show neither PAc nor HG-PIN. For the cases of Groups 2 and 3, the
samples with PAc were from the peripheral zone. The pathological stage was based on the 7 th
2009 revision of the TNM (53). The Gleason score of the cancers of Group 2 was based on
the ISUP 2005 modification (54). Due to the neoadjuvant therapy, the Gleason grading
12
system was not applied to the cases of Group 3. The presence of HG-PIN was recorded for the
cases of Groups 2 and 3.
Immunohistochemistry
Immunohistochemical analysis for IGFBP2 (Santa Cruz Biotechnology, Santa Cruz, CA,
USA, dilution 1:100) and for AMACR (Cell Marque, Rocklin, CA, U.S.A., dilution 1:100)
was performed. Antigen retrieval was obtained by pre-treatment in W-CAP citrate buffer pH
6.0 for IGFBP2 and in W-CAP TEC buffer pH 8.0 for AMACR (Bio-Optica Milano SpA,
Milan, Italy) at 98°C for 25 minutes. Inhibition of endogenous peroxidases was performed in
3% H2O2 solution. After rinsing the slides in phosphate-buffered saline (PBS) 1x solution
with 0.1% solution of detergent Tween 20 (phosphate-buffered saline–Tween 20; BioOptica), the sections were incubated in a humid chamber at room temperature for 5 minutes
with Ultra V Block solution (Ultravision LP, LabVision Corporation, Thermo Fisher
Scientific Inc. Fremont, CA, U.S.A.). They were subsequently incubated in the humid
chamber at 4°C for 6 hours with primary antibody for IGFBP2 and at room temperature for 1
hour with primary antibody for AMACR. Sections were then washed in buffered solution and
incubated in the humid chamber at room temperature for 20 minutes with primary antibody
enhancer solution (Ultravision LP, LabVision Corporation). After several washes in buffered
solution, the sections were incubated for 30 minutes in horseradish peroxide (HRP)
(Ultravision LP, LabVision Corporation) polymer solution. Reaction was revealed with
diaminobenzidin (DAB) solution for 3 minutes and counterstained with hematoxylin.
Evaluation of immunohistochemistry
For each immunohistochemical marker, the percentage of positive cells was calculated in
prostate cancer, HG-PIN and normal looking epithelium in Groups 2 and 3 as well as in
normal tissue in Group 3. In order to consider not only the percentage of positive neoplastic
cells but also their staining intensity, we also estimated the immunohistochemical score,
according to McCarthy’s scoring system, originally performed on breast neoplasia (55) and
13
subsequently also applied to prostate carcinoma (56). The immunohistochemical score was
calculated as follows: percentage of positive neoplastic cells multiplied by the staining
intensity (0: none; 1: weak; 2: moderate; 3: strong). It ranges from 0 to 300. The mean and
standard deviation (SD) were determined for the percentages of positive cells and their
immunohistochemical scores for both IGFBP2 and AMACR in the three groups.
Statistics
Statistical analysis was carried out with the statistical package SPSS 13.0 for Windows (SPSS
Inc., Chicago, IL, U.S.A.). It included ROC (receiver-operator characteristic) curves, the
Wilcoxon signed rank test and the Spearman rank test. The differences between the groups
were considered statistically significant at a value of p<0.05.
Results
1) Group 1 (patients with bladder outlet obstruction)
The ducts and acini of all cases of Group 1 with the exception of one were negative for
IGFBP2. A weak positivity for IGFBP2 was seen in urothelial metaplasia and periurethral
glands. Stromal and endothelial cells were negative. An identical staining pattern was seen in
the normal looking ducts and acini of Groups 2 and 3. Among normal ducts and acini,
scattered cells were intensely stained (data not shown). In deeper sections, the cells in the
same location were Chromogranin A positive and thus interpreted as neuroendocrine (NE)
cells.
As far as AMACR was concerned, normal tissue in the three groups was negative, both in
biopsies and in surgical specimens.
14
2) Group 2: Patients hormonally untreated before surgery
Prostate cancer
IGFBP2 was positive in the cytoplasm of all 40 cases of PAc, both in the preoperative
biopsies and in the surgical specimens. The percentage of neoplastic cells immunoreactive for
IGFBP2 ranged from 10 to 90% in the biopsies (mean 66%, SD 22%) and from 10 to 80% in
the surgical specimens (mean 59%, SD 21%). The immunohistochemical score for IGFBP2
ranged from 10 to 270 in the biopsies (mean 152, SD 82) and from 20 to 240 in the surgical
specimens (mean 142, SD 68).
AMACR was positive in all cases (40/40) of PAc. The percentage of neoplastic cells positive
for AMACR ranged from 20 to 90% in the biopsies (mean 80%, SD 14%) and from 50 to
90% in the surgical specimens (mean 79%, SD 12%), while the immunohistochemical score
ranged from 20 to 270 in the biopsies (mean 210, SD 63) and from 50 to 270 in the surgical
specimens (mean 205, SD 56).
No significant correlation was found between IGFBP2/AMACR immunohistochemical scores
and the Gleason score, neither in the biopsies (ρ=-0.52 for the Gleason score and p=0.750 for
the IGFBP2 immunohistochemical score; ρ=-0.64 for the Gleason score and p=0.693 for the
AMACR immunohistochemical score) nor in the surgical specimens (ρ=-0.89 for the Gleason
score and p=0.585 for the IGFBP2 immunohistochemical score; ρ=0.001 for the Gleason
score and p=0.994 for the AMACR immunohistochemical score). In the surgical specimens,
no correlation was found between IGFBP2/AMACR immunohistochemical scores and the
corresponding pathological stage (ρ=-0.245 for the stage and p=0.312 for the IGFBP2
immunohistochemical score; ρ= -0.156 for the stage and p=0.523 for the AMACR
immunohistochemical score).
HG-PIN
In HG-PIN, a subtle positivity for IGFBP2 was detected in all but 1 biopsy and in all but 4
surgical specimens. The percentage of positive HG-PIN cells for IGFBP2 ranged from 10 to
15
70% (mean: 25%; SD: 25%) in the biopsies and from 5 to 80% (mean: 18%; SD: 20%) in the
surgical specimens. The immunohistochemical score ranged from 5 to 100 (mean: 47; SD: 52)
in the biopsies and from 5 to 160 (mean: 39; SD: 47) in the surgical specimens.
The positivity for AMACR in HGPIN was observed in all but 1 biopsy and in all but 1
surgical specimen. The percentage of positive HG-PIN cells for AMACR ranged from 10 to
90% (mean: 55%; SD: 29%) in the biopsies and from 10 to 80% (mean: 34%; SD: 26%) in
the surgical specimens, while the immunohistochemical score for AMACR ranged from 10 to
270 (mean: 114; SD: 77) in the biopsies and from 10 to 240 (mean: 76; SD: 71) in the surgical
specimens.
Statistical analysis
According to the ROC curve analysis, by examining the percentages of positive neoplastic
cells, the overall accuracy (as expressed by the area under each curve) in detecting invasive
PAc vs. HG-PIN was higher for IGFBP2 than for AMACR. The area under the ROC curve
was higher for IGFBP2 than for AMACR, both in the biopsies (0.914 for IGFBP2 and 0.787
for AMACR) and in the surgical specimens (0.906 for IGFBP2 and 0.887 for AMACR).
Cut-off values for IGFBP2 positivity in discriminating PAc vs. HG-PIN were determined if
greater than or equal to 25% of the lesional cells (sensitivity: 0.950, 1-specificity: 0.421 in the
biopsies; sensitivity: 0.875, 1-specificity: 0.211 in the surgical specimens). A lesion can be
then considered highly suspicious for HG-PIN when less than 25% of the lesional cells are
positive for IGFBP2. Cut-off values for AMACR were not identified since they were more
difficult to determine and generally higher, the overall performance of the test also being
lower than for IGFBP2.
According to the Wilcoxon signed rank test, by examining both the percentages of positive
neoplastic cells and their corresponding immunohistochemical scores, IGFBP2 in comparison
to AMACR revealed fewer neoplastic cells not only in invasive PAc, but also in HG-PIN. In
invasive PAc, IGFBP2 detected fewer neoplastic cells than AMACR, both in the biopsies
(Z=-3.213, p=0.001 by percentage of positive neoplastic cells; Z=-3.006, p=0.003 by their
16
immunohistochemical scores) and in the surgical specimens (Z=-4.127, p<0.001 by
examining the percentage of positive neoplastic cells; Z=-4.015, p<0.001 by
immunohistochemical score). A similar reduction was also observed in HG-PIN, both in the
biopsies (Z=-3.595, p<0.001 by percentage of positive neoplastic cells; Z=-3.600, p<0.001 by
immunohistochemical score) and in the surgical specimens (Z=-3.001, p=0.003 by percentage
of positive neoplastic cells; Z=-2.760, p=0.006 by immunohistochemical score).
3) Group 3: Patients who underwent androgen ablation before surgery
Prostate cancer
Immunohistochemical expression of IGFBP2 in PAc was detected in all (10/10) cases in the
biopsies and in 9 out of 10 in the surgical specimens (Figure 1). In particular, the percentage
of neoplastic cells immunoreactive for IGFBP2 ranged from 10 to 90% (mean 41%, SD 26%)
in the biopsies and from 5 to 70% in the surgical specimens (mean 24%, SD 25%). The
immunohistochemical score ranged from 10 to 180 in the biopsies (mean 76, SD 69) and from
10 to 210 in the surgical specimens (mean 52, SD 69).
Immunoreactivity for AMACR was observed in PAc in all (10/10) cases, in the biopsies and
the surgical specimens. The percentage of neoplastic cells positive for AMACR ranged from
50 to 90% in the biopsies (mean 79%, SD 12%) and from 10 to 90% in the surgical specimens
(mean 48%, SD 28%), while the immunohistochemical scores varied from 100 to 270 in the
biopsies (mean 184, SD 57) and from 10 to 270 in the surgical specimens (mean 105, SD 75).
HG-PIN
HG-PIN was seen in 2 biopsies and in 5 surgical specimens.
17
Discussion
The results obtained in the present study showed that IGFBP2 is consistently positive in PAc,
while it is negative in benign prostatic tissues.
Three other studies have documented consistent immunoreactivity for IGFBP2 in invasive
PAc (57-59). However, differences have been noted in examining IGFBP2 expression in HGPIN. The current investigation shows that IGFBP-2 is expressed in HG-PIN, but at a lower
level than in PAc. Tennant and coworkers reached a similar conclusion by examining 28
prostatectomy specimens with adenocarcinoma; they identified weak immunoreactivity for
IGFBP2 in normal epithelium, moderate staining in PIN and strong or very strong expression
in adenocarcinoma (60). In another study conducted by the same group on 24 prostatic
specimens (with 20 adenocarcinomas), comparable results were obtained since IGFBP-2
immunoreactivity increased from the normal through the premalignant (i.e., HG-PIN) and into
the malignant states (i.e., PAc) (61). Richardsen et al. confirmed negativity or very weak
expression in normal epithelium and in benign prostatic hyperplasia (62); however, they did
not find any differences between HG-PIN and carcinoma, and they described the
overexpression of IGFBP2 in both HG-PIN and carcinoma; in different cases, the intensity
varied from weak to moderate to strong, and the pattern varied from diffuse granular staining
to strong cytoplasmic staining (63).
The present data failed to find any correlation between IGFBP2 expression and Gleason grade
or tumor stage. This is comparable to the three above-mentioned studies (64-66). However,
considering total IGFBP2 RNA expression in neoplastic tissue, Figueroa et al. detected
significantly higher RNA expression of IGFBP-2 in tumors with a high Gleason score in
comparison to tumors with a low score and benign tissue (67).
The present study shows that IGFBP-2 is strongly expressed in the NE cells present in benign
prostatic glands. Cells in a similar location do not express AMACR. This finding is similar to
that seen previously by Richardsen et al (68). Furthermore, Tennant et al. described very
strong immunoreactivity in scattered stromal cells (69), which may represent what we
interpreted as NE cells. Previous studies suggested a role of IGFBP2 in stimulating
proliferation of prostatic cells (70;71). It is not easy to understand the meaning of IGFBP2
18
expression in NE prostatic cells, but it may represent a further proof of IGFBP2 involvement
in growth stimulation.
In cases of Group 3 (i.e. patients following complete androgen ablation), the expression of
IGFBP2 was consistently detected in the initial diagnostic biopsies, but markedly reduced in
the surgical specimens after hormonal treatment. The expression of AMACR was also
reduced, but at a lower level than that of IGFBP2. Interestingly, androgen ablation did not
affect IGFBP2 expression in the NE cells, while AMACR remained negative in the NE cells.
This is among the first studies to document the immunoreactivity for IGFBP2 of PAc
following complete androgen ablation. Bubendorf et al. demonstrated that IGFBP2 was
genetically overexpressed in a hormone refracting cell line of prostate cancer; on a tissue
microarray, they also observed consistent immunoreactivity for IGFBP2 in all tumors which
had developed a recurrent tumor during androgen deprivation therapy (72). Similar
conclusions were noted by Kyiama et al., who found that IGFBP2 mRNA and protein levels
increased 2–3-folds after androgen withdrawal in LNCaP (an androgen-sensitive human
prostatic carcinoma cell line); they also identified increased IGFBP-2 immunohistochemical
levels after castration using a human prostate tissue microarray of untreated and posthormone
therapy treated prostatectomy specimens (73). Other experimental studies on cancer cell lines
showed that stimulation by IGFBP2 had a potent stimulatory effect on the growth of LAPC-4
(an androgen-dependent cell line) prostate cancer cells, this effect being more pronounced in
the absence of androgens (74). Inman et al. investigated the serum level of IGFBP2 following
androgen ablation (75) and they found that high serum levels of IGFBP2 were associated with
a better prognosis in patients who received hormonal neoadjuvant therapy and with a worse
prognosis in patients who did not receive any preoperative treatment (76). All these findings
may support a possible interaction between IGFBP2 and the androgen receptor system.
All these results lead to two considerations. One is that IGFBP2 is a good
immunohistochemical marker in the identification of PAc. This concept is also sustained by
the role played by IGFBP2 as a serum marker (both diagnostic and prognostic) for PAc (7783). The other is the fact that IGFBP2 could be involved not only in the transition from HG-
19
PIN to PAc but also, through the action of the NE cells, in the progression of PAc towards an
androgen-independent phase, as demonstrated in those cases with NE differentiation (84-86).
In conclusions the principal findings in the current study were:
IGFBP2 is not expressed in the normal-looking tissue of the transition and peripheral
zones;
IGFBP2 is expressed in the cytoplasm of untreated PAc and, to a lesser extent, in HGPIN
IGFBP-2 is also expressed in PAc and HG-PIN following complete androgen ablation,
but to a lesser extent than in the untreated neoplasms;
IGFBP2 expression in the untreated specimens is lower in HG-PIN than in invasive
PAc. This finding can be of diagnostic help especially in thin pre-operative needle
biopsies, when a small amount of tissue is available for immunohistochemical
stainings.
Neuroendocrine cells present in prostatic glands are immunoreactive for IGFBP-2
(87).
20
Figure 1. IGFBP2 as a diagnostic marker for prostatic adenocarcinoma (PAc).
Group 3: patients who underwent androgen ablation before surgery.
In biopsies, invasive PAc (a) highly expresses both IGFBP2 (b) and AMACR (c). In surgical
specimens, invasive PAc after androgen ablation (d) shows a markedly lowered expression of
IGFBP2 (e) while AMACR immunoreactivity is still detectable (f).
21
HEPARAN SURFACE PROTEOGLYCANS IN ORAL SQUAMOUS CELL
CARCINOMAS AS PROGNOSTIC MARKERS AND THEIR PREDICTIVE ROLE
TO ADJUVANT RADIOTHERAPY
Oral squamous cell carcinoma (OSCC) is a major disease, with estimated 263,000 new cases
worldwide per year (88). In pathological evaluation of OSCC, two of the most critical
prognostic points are stage (89;90) and Brodman’s grade (91). Other important morphological
factors that may play an important role at the morphological evaluation are: tumor thickness
and depth of invasion (92;93), desmoplastic reaction of the surrounding stroma (94), tumor
associated tissue eosinophilia (95), vascular and/or perineural invasion.
Molecular carcinogenesis in oral squamous cell carcinoma is a step by step process, which
involves numerous factors (96;97), similarly to the colorectal carcinogenesis described in the
introduction section. From this step-by-step process, many factors have been identified, and
some of them deserve special attention because of their prognostic value. Prognostic factors
extensively studied are: p16 (98), p53 and p63 (99), cyclin D1 (100). Lymph node metastasis
is certainly one of the histopathological parameters that primarily affect prognosis, but many
cases and especially small tumors still have undetectable nodal disease (101). Despite the
recent diagnostic and therapeutic improvements, factors determining metastatic disease are
mostly unknown.
Proteoglycans (PGs) are glycoproteins with one or more covalently attached heparin sulfate
chains (102). On the basis of their core protein primary structure, they are classified in cell
surface PGs, extracellular PGs, and intracellular PGs (103). Cell surface PGs are either
integral membrane PGs or are linked to the membrane via a phosphatidylinositol moiety;
they appear to serve as receptors for growth factors and other components of the extracellular
matrix, for cell-matrix and cell-cell interactions and as receptors for other cell-cell interaction
molecules (104). The major cell surface PGs are syndecans and glypicans (105;106). The
syndecan (SYN) family comprises four integral membrane proteins, named SYN-1 to SYN-4
(107). In the glypican (GPC) family there are six family members, known as GPC-1 to GPC-6
(108). Another important cell surface proteoglycan is NG2 (109;110).
22
Several studies have demonstrated the important role played by cell surface PGs in promoting
cell growth and development in human development and in neoplastic events (111;112). In
OSCC, immunohistochemical investigations have shown that SYN-1 staining intensity is
associated with keratinocyte differentiation and clinical outcome, both in epithelial neoplastic
cells (113;114), and in the stroma (115). SYN-2 may function as a cell surface receptor in
highly migratory tumor cells (116). Several studies suggested that GPC-3 could act as a tumor
suppressor gene (117;118). In culture cell studies, NG2 expression is related to tumor
initiations and growth rate, predisposing to poorer prognosis (119).
Aim of the present study is to define the molecular and immunohistochemical expression of
cell surface PGs in OSCC, both in epithelial neoplastic cells and in the accompanying stroma,
and to establish possible relationship with the clinical outcome.
Materials and methods
Patient selection
A total of 150 cases of OSCC were obtained from the files of the Departments of Surgical
Pathology of the Universities of Bologna and Parma (Italy). Patients were surgically treated
by three groups of Maxillo-Facial surgeons from Bologna (University of Bologna at
Polyclinic S.Orsola-Malpighi and Bellaria Hospital) and Parma (University of Parma),
applying the same surgical procedures.
Criteria of selections were the follows: a) all cases were primary OSCC (assessed by preoperative biopsy), not previously treated by radio or chemotherapy; b) fresh frozen tissue was
available for molecular studies.
Follow up information was available in 93 patients for a period ranging from 6 to 34 months
(mean: 19, SD: 7).
23
Adjuvant therapies
In selected patients, especially in those affected by advanced disease, adjuvant postoperative
treatments were performed. Chemotherapy consisted of 5-fluorouracil and cisplatinum. Radio
therapy was composed of 60 Gy administered in 30-33 sections.
Tissue microarray (TMA) construction
From each case, the block containing representative neoplastic tissue was selected. In order to
have uniform immunohistochemical characterization of all cases, tissue micro-arrays (TMAs)
were constructed. TMA construction was performed following a previously described
procedure (120-122). Briefly, a new cut haematoxylin and eosin-stained section was obtained
from each paraffin block and used to define diagnostic areas. Subsequently representative 0.6
mm cores were obtained from each case, and inserted in a grid pattern into a recipient paraffin
block using a tissue arrayer. Cases were considered representative when at least 50% of the
section was composed of neoplastic tissue. For each case, the core with the highest percentage
of tumor cells stained was used for analysis.
Immunohistochemistry
Sections (4 µm) were cut from TMA and stained with the antibodies listed in Table 1.
Immunohistochemistry was performed as follows: dewaxing and antigens unmasking
occurred simultaneously with the solution W-Cap TEC buffer pH 6 or W-Cap TEC buffer pH
8 (Bio-Optica, Milan, Italy) for 25 minutes at 98° C. The endogenous peroxidases inhibition
was performed by 10 minutes incubation with H2O2 (3% in H2O), washing in distilled water
and incubating for 5 minutes with Blocking Solution (LabVision, Fremont, CA, USA) to
induce the non-specific binding sites saturation; both steps occurred at room temperature.
Primary monoclonal antibodies, listed in Table 1, were applied on sections for 60 minutes at
room temperature. Following, chromogenic detection was performed using the UltraVision
Detection System (LabVision, Fremont, CA, USA), which provided incubation with Antibody
Enhancer for 20 minutes followed by HRP-Polymer for 30 minutes. Finally, DAB chromogen
24
(Dako, Carpenteria, CA, USA) was applied for 3-5 minutes and sections were counterstained
with hematoxylin after washing with water.
25
Manufacturer
Dilution
Antigen
Antibody
Clone
SYN-1
MI15
Dako
1:100
W-CAP pH 6
SYN-2
1F10/B8
Santa Cruz Biotechnology
1:50
W-CAP pH 8
1:50
W-CAP pH 8
Sigma Aldrich (Powered by
SYN-3
AtlasProtein)
Retrieval
SYN-4
H-140
Santa Cruz Biotechnology
1:100
W-CAP pH 8
GPC-1
4D1
Millipore
1:50
W-CAP pH 6
GPC-3
1G-12
Biomosaics
1:250
W-CAP pH 6
1:500
W-CAP pH 8
1:20
W-CAP pH 8
1:50
W-CAP pH 6
1:5
W-CAP pH 8
Immundiagnostik
GPC-4 (aa
54-66)
Sigma Aldrich (Powered by
GPC-6
NG2
4D1 surnatant
AtlasProtein)
132.38
Sigma Aldrich
Handle
Made
Table 1. List of all antidodies used. Legend: SYN syndecan, GPC glypican.
26
Immunohistochemical evaluation
The evaluation and scoring of the immunohistochemical results were performed with a light
microscope (Nikon) at a magnification of 40Χ. Each antigen expression was
semiquantitatively evaluated and scored as follows:
Score 0 (-) = no positive cells were detected
Score 1 (+\-) = <10% of cells were positive
Score 2 (+) = 10-50% of cells were positive
Score 3 (++) = >50% of cells were positive
Score 4 (+++) = >90% of cells were positive
Statistical analysis
For each case, mean and standard deviation (SD) of all variables was determined. Spearman’s
rank order correlation (ρ) was calculated among the different immunohistochemical scores
and differentiation grade, in order to detect possible relationships among them.
Overall survival and disease-free specific survival were analyzed in both univariate and
multivariate analyses. Single factors were analyzed with univariate analysis and the statistical
significance was calculated with log rank test. Group of factors were studied together in
multivariate analysis and their statistical significance was calculated with Cox regression
analysis. In multivariate analyses, many different models were tested, with different
combinations. The main models were: 12 independent variables (SYN-1, SYN-2, SYN-3,
SYN-4, GPC-1, GPC-3, GPC-4, GPC-6, 4D1 surnatant in the epithelial neoplastic cells, and
SYN-1, SYN-2 , GPC-1 in the stroma), 6 independent syndecan variables (SYN-1, SYN-2,
SYN-3, SYN-4, in the epithelial neoplastic cells, and SYN-1, SYN-2 in the stroma), 4
independent syndecan variables (SYN-1, SYN-2, in the epithelial neoplastic cells, and SYN1, SYN-2 in the stroma) . For the validity of the regression models, GPC-4 expression in the
stroma was removed from the variables, because it was correlated with GPC expression in the
epithelial cells. NG2 was also not included, because too few measurements were performed.
27
All data were analyzed with SPSS 13.0 for Windows (SPSS Inc., Chicago, IL, USA).
Statistical significance was taken at p value (two-tailed) less than 0.05.
Results
Clinical and pathological features
Patients were 62 females and 88 males, with age varying from 27 to 93 (mean 62; SD 14).
Smoke and alcohol consumption was known in 86 patients, among which 54 were habitual
smokers and 44 referred regular alcohol consumption.
With the 2010 TNM staging system (53), cases were classified as follows: 41 pT1, 54 pT2,
13 pT3, and 42 pT4; 88 pN0, 27 pN1, 31 pN2, 2 pNx; 35 stage I, 28 stage II, 27 stage III, 57
stage IV.
Lymph node metastases were detected in 60 cases at presentation, and in 4 cases in the FU; in
2 cases no lymph node was resected.
Follow up revealed local recurrence in 10 cases, 3 cases with lymph node metastasis, 1 case
with local recurrence and lymph node metastasis; among all of them, 11 patients were
deceased (1 with esophageal carcinoma). In 71 patients no recurrence and/or metastasis was
documented.
In the constructed TMA, 3 cases were composed of carcinoma in situ, among them 1 case was
composed only of in situ carcinoma; 1 case did not contained sufficient material.
Representative sections for all markers under study were available in 148/148 invasive
OSCCs.
Immunohistochemical features in invasive carcinoma (148 cases)
SYN-1 (CD138). The staining was predominantly observed with a cell membrane pattern.
The epithelial neoplastic cells were positive for SYN-1 in 133/148 cases (Figure 1a), while
28
the stromal component was positive in 31/148 cases (Figure 1b). Staining was mainly
localized in keratinizing neoplastic cells, located at the center of the neoplastic nests.
SYN-2. The staining was predominantly observed with a cell membrane pattern. SYN-2 was
positive in the epithelial neoplastic cells in 21/148 cases (Figure 1c). The stromal cells were
positive in 108/148 cases (Figure 1d). Staining increased when desmoplastic stroma appeared.
Furthermore, SYN-2 marked the thin stromal vessel walls in 45/148 cases.
SYN-3. The staining was predominantly observed with a diffuse cytoplasmic pattern and focal
membrane reinforcement. This marker was positive in the epithelial neoplastic cells in 54/148
cases. No reactivity was detected in the stroma.
SYN-4. The staining was predominantly observed with a diffuse cytoplasmic pattern and focal
membrane reinforcement. In the epithelial neoplastic cells, SYN-4 was positive in 31/148
cases. No reactivity was detected in the stroma.
GPC-1. The staining was predominantly observed with a diffuse cytoplasmic pattern and
membrane reinforcement. In the epithelial neoplastic cells, GPC-1 was positive in 111/148
cases (Figure 2a). Stromal reactivity was strongly seen only in 8/148 cases.
GPC-3. The staining was predominantly observed with a diffuse cytoplasmic pattern and
membrane reinforcement (Figure 2b). In the epithelial neoplastic cells, GPC-3 was positive in
27/148 cases. No reactivity was detected in the stroma.
GPC-4. The staining was predominantly observed with a diffuse cytoplasmic pattern. In the
epithelial neoplastic cells, GPC -4 was positive in 55/148 cases (Figure 2c). In the stroma,
27/148 cases exhibited diffuse immunoreactivity.
GPC -6. The staining was predominantly observed with a granular cytoplasmic pattern. In the
epithelial neoplastic cells, GPC-6 was positive in 54/148 cases. No reactivity was detected in
the stroma.
NG 2. The staining was predominantly observed with a nuclear pattern. In the epithelial
neoplastic cells, NG2 was positive in 35/36 cases. No reactivity was detected in the stroma.
29
4 D1 surnatant. The staining was predominantly observed with a granular cytoplasmic
pattern. The epithelial component was negative in all 98/98 cases. In the stromal component,
4 D1 was positive only in stromal cells in 18/148 cases (Figure 2d).
Immunohistochemical features in carcinoma in situ
The 3 cases composed of carcinoma in situ showed moderate/diffuse staining for SYN-1 and
GPC-1 in the epithelial neoplastic cell. The other markers were negative, both in the epithelial
neoplastic cells and in the stroma.
Figure 2. Immunohistochemical expression of syndecans.
Syndecan-1 (SYN-1) immunoreactivity in epithelial neoplastic cells (a) and in stromal cells
(b); syndecan-2 (SYN-2) immunoreactivity in epithelial neoplastic cells (c) and in stromal
cells (d).
30
Figure 3. Immunohistochemical expression of glypicans and 4 D1.
Immunohistochemical expression of glypican-1 (GPC-1) (a), glypican-3 (GPC-3) (b) and
glypican-4 (GPC-4) (c) in epithelial neoplastic cells; 4D1 surnatant is expressed only in
stromal cells (d).
31
Statistical analysis
Statistical analysis was carried out only in invasive carcinoma (148 cases). The strongest
correlation among the immunohistochemical scores was found between GPC-4 expression in
the epithelial neoplastic cells and GPC-4 expression in the stroma (ρ= 0.775, p < 0.0001).
In cases analyzed with all data and available follow-up, survival multivariate analysis with
Cox regression model revealed a statistical significance between syndecan 1 expression and
reduction of overall survival. In such association we can appreciate 0.046 p-value with 1.967
odds radio. This is the summary of the analysis (Table 2).
B
SE
Wald
df
P
Exp(B)
Syndecan 1
(epitelial
neoplastic
cells)
0.676
0.339
3.973
1
0.046
1.967
Syndecan 1
(stroma)
0.473
0.415
0.011
1
0.917
1.044
Syndecan 2
(epitelial
neoplastic
cells)
-0.616
0.898
0.471
1
0.493
0.540
Syndecan 2
(stroma)
0.485
0.249
3.787
1
0.052
1.624
Table 2. Summary of the Cox regression model in all patients with OSCC.
32
The same analysis, repeated on patients with received subsequent adjuvant radiotherapy (total
44 cases) revealed a stronger association between SYN-1 expression and reduction of overall
survival. In such analysis we can appreciate 0.023 p-value with 3.479 odds radio: herein the
statistical association is much stronger than the previous one. Then we can reasonably
conclude that, in patients submitted to adjuvant radiotherapy, the immunohistochemical
expression of SYN-1 in neoplastic cells, significantly decreases the overall survival. This is
the summary of the analysis (Table 3).
Syndecan 1
B
SE
Wald
df
P
Exp(B)
1.247
0.548
5.167
1
0.023
3.479
0.319
0.635
0.253
1
0.615
1.376
-0.964
1.049
0.845
1
0.358
0.381
0.050
0.325
0.024
1
0.877
1.052
(epitelial
neoplastic
cells)
Syndecan 1
(stroma)
Syndecan 2
(epitelial
neoplastic
cells
Syndecan 2
(stroma)
Table 3. Summary of the Cox regression model in patients with OSCC who received
radiotherapy.
In order to better clarify the role of SYN-1 alone in patients who received radiotherapy, a
univariate analysis of survival was performed. Log-rank test revealed a significant association
between SYN-1 score and survival (p=0.001).
33
The scoring system in 5 levels, as described before, was important in correctly evaluating the
exact percentage of neoplastic cells which were positively stained for SYN-1. However, just
for practical purposes, we tried to subdivide the scoring system for SYN-1 immunoreactivity
in two major groups: expression more or less than 50% of the neoplastic cell population. With
this cut-off, the log-rank test confirmed the statistical significance (p=0.019). Herein we can
appreciate this statistical significance with the two overall survival curves.
In this graph we could graphically evaluate that, if SYN-1 expression was more than 50%
(green line), the survival was significantly reduced. On the contrary, if SYN-1
immunoreactivity was less than 50% (blue line), the overall survival was significantly
increased.
34
In summary, such association between SYN-1 and overall survival became stronger after
repeating the same analysis only in patients who received radiotherapy. In this view, SYN-1
may be considered not only as a prognostic factor, but also as a predictive factor for
responsiveness to adjuvant radiotherapy.
Other interest results were obtained in considering the disease-free survival curves, in
examining the role of radiotherapy. We noted a significant reduction in disease-free survival
for patients who received adjuvant radiotherapy (p=0.021).
In this graph the green line represented patients who received radiotherapy. This group of
patients showed a significant reduction in disease-free survival. This apparently paradoxical
35
result is explained by the fact that radiotherapy is usually administered in patients affected by
OSCC at advanced stage; these patients would certainly display an adverse prognosis in
comparison to the other group.
Unfortunately, the role of chemotherapy was not examined, due to the fact that a very small
number of patients received this type of treatment. In fact, only 6 patients were treated with
this adjuvant chemotherapy; this number was too few for any type of statistical consideration.
Discussion
Cell adhesion molecules, such as integrins, cadherins and cell-surface glycoproteins, are
involved in many phases of the cell cycle, in differentiation, migration and in many different
stages of neoplastic development (123-126).
The major cell surface PGs are syndecans and glypicans (105;106). The syndecan (SYN)
family comprises four integral membrane proteins, named SYN-1, SYN-2, SYN-4 and SYN-4
(107). In the glypican (GPC) family there are six family members, known as GPC-1, GPC-2,
GPC-3, GPC4, GPC-5 and GPC-6 (108). Another important cell surface proteoglycan is NG2
(109;110).
Cell surface PGs play an important role, nor only in cellular development, but also in many
neoplastic diseases (111;112). In OSCC, some researchers have focused on SYN-1 (also
known as CD138) immunoreactivity as a prognostic factor. Inki et al. have studied 29 patients
affected by squamous cell carcinoma of the head and neck and they found significant
association between SYN-1 immunohistochemical expression and grade of differentiation;
overall survival and recurrent-free survival were reduced in patients with low SYN-1
expression (113). Anttonen et al. have analyzed 175 patients affected by squamous cell
carcinoma of the head and neck; they have noted important associations between SYN-1
expression and histological grade, stage, tumor size; they also have described a significant
reduction of overall survival in patients with reduced (<80%) expression of SYN-1.
36
Our results are apparently in contrast with these two previous studies, because our data
suggested that an increased SYN-1 expression was associated with reduced overall survival.
Nevertheless, these two studies considered all squamous cell carcinomas of the head and neck
region. In our study, we collected only the tumors of the oral cavity. Furthermore, other
studies performed upon tumors in the head and neck area and in other districts showed similar
conclusion to ours. For example, Chen et al. revised 157 nasopharingeal carcinomas and
found that tumors with SYN-1 expression showed reduction of overall survival in comparison
to tumors negative for SYN-1. Barbareschi et al. found an association between high SYN-1
expression and a a more aggressive phenotype in breast neoplasias (127). Other studies found
high SYN-1 expression in association with poorer prognosis or aggressive phenotypes, in
prostate neoplasias (128), and in thyroid carcinomas (129).
All these previous studies reported conflicting results on the role played by SYN-1 as a
prognostic marker. In this view, our data may help in understanding future studies.
Furthermore, this association between SYN-1 and prognosis was not only found in all cases
with available follow up, but it was much more increased when calculated in patients which
underwent radiotherapeutic treatment (p-value 0.023; odds ratio 3.479). This increased
association not only confirmed the role of SYN-1 as a prognostic factors, but it introduces
also the possibility to consider SYN-1 as a predictive factor for establishing the
responsiveness to radiotherapy. To be a confirmed as a predictive role as responsiveness to
radiotherapy, a randomized study may be performed.
SYN-1 expression was also studied in the stroma by Mukunyadzi (115), who found an
increased expression in invasiveness foci, but without any association with prognosis. We
also evaluated the stroma, and found no association with prognosis. Other cell surface PGs
have been studied in the literature. SYN-2 has been evaluated in neoplastic cell culture studies
and may function as a cell surface receptor in highly migratory tumor cells (116). Some
authors have demonstrated that GPC-3 could act as a tumor suppressor gene (117;118) . In
culture cell studies, NG2 expression is related to tumor initiations and growth rate,
predisposing to poorer prognosis (119). In our study, with all the other markers (except SYN1) we did not found any statistical association.
37
The last analysis conducted was the disease-free survival curves upon patients with
radiotherapy against patients which received no adjuvant treatment. This analysis was carried
out in order to better clarify the two groups of patients; if different results had been
discovered, further analyses would have been performed and possible association with PG
expression investigated. Nevertheless, even if statistical associations were found, they totally
lacked any clinical significance. As explained before, this apparently paradoxical result is
explained by the fact that radiotherapy is usually administered in patients affected by OSCCs
at advanced stage; these patients would certainly display an adverse prognosis in comparison
to the other group. This is an example of a statistically but not clinically significant result.
The last consideration is for the recent perspectives upon the role of HPV infection in the
development of OSCC. It is now clear that HPV, and in particular the high risk type 16 (HPV16), play a central role in the oncogenetic process (130). The OSCCs associated with HPV are
predominantly localized in the palatine tonsil and lingual tonsils of the oropharynx and are
usually not-keratinizing, mainly of basaloid type. OSCCs associated with HPV affect
predominantly young patients, with no other known risk factor for OSCC (alcohol and
tobacco) and carry a significant better prognosis. In our data, HPV infection was not analyzed
and is currently under investigation in the University of Pavia for another research project.
Nevertheless, the cases we selected were mainly localized on the oral cavity and not in the
oropharynx. Furthermore, it is becoming more and more difficult to discriminate which is the
real role played by HPV in OSCC. HPV is a ubiquitous infection and even its detection may
represent only a superimposed infection, especially for the low risk types.
In conclusion our study demonstrated that:
1. SYN-1 is a prognostic factor, since it is significantly increased in OSCCs with poor
prognosis, with reduction of overall survival.
2. This association between SYN-1 and overall survival is much stronger in patients with
subsequent adjuvant radiotherapy
3. SYN-1 may be a predictive factor of responsiveness to adjuvant radiotherapy.
38
EGFR MUTATION-SPECIFIC ANTIBODIES IN PULMONARY
ADENOCARCINOMA: A COMPARISON WITH DNA DIRECT SEQUENCING
Epidermal Growth Factor Receptor (EGFR) is a trans-membrane receptor with tyrosine
kinase activity that plays a central role in regulating cell growth and differentiation, both in
normal and neoplastic cells (131). In patients affected by non-small cell lung cell cancer
(NSCLC), specific mutations of the EGFR gene correlate with pathological features and
responsiveness to tyrosine kinase inhibitors (TKIs) such as gefitinib (14) and erlotinib (15). In
EGFR-mutated lung adenocarcinoma, gefitinib is superior to carboplatin-paclitaxel therapy
and EGFR mutations strongly predict outcome after therapy (132). Mutational status of the
EGFR gene is of thus of primary importance in defining therapeutic decisions (133).
The mutational spectrum of EGFR in lung adenocarcinomas is variable including in-frame
deletions, in-frame insertions/duplications and point mutations. The two most common
genetic changes are an in-frame deletion in exon 19 at codons 746 to 750 (E746-A750
deletion) and the substitution of leucine 858 by arginine (point mutation L858R) in exon 21
(134). Together, these two mutations account for approximately 90% of the cases and are
termed “classical” mutations (135). The gold standard technique for identifying EGFR
mutation is direct DNA sequencing of PCR-amplified regions of exons 18, 19, 20, 21 (133)
but its clinical application is limited due to problems of tissue conservation and sampling,
costs and technical difficulties.
Recently, two novel antibodies that specifically recognize the E746-A750 deletion in exon 19
and the L858R point mutation in exon 21 have been described (136). As
immunohistochemistry (IHC) may represent a faster, more economic and more widely
applicable alternative to DNA sequencing, the aim of the present study was to compare the
two procedures in a series of pulmonary adenocarcinomas.
Materials and Methods
We retrieved from our files all 18 cases of advanced pulmonary adenocarcinomas with EGFR
mutations and 15 cases with wild-type EGFR. Specimens were routinely fixed in buffered
39
formalin and embedded in paraffin. Sections (4 µm) were stained with haematoxylin and
eosin for histological diagnosis. All cases were morphologically reviewed and classified
according to the most recent IASLC guidelines for lung adenocarcinomas; in each case the
prevalent pattern of growth was recorded (137). Immunostaining using a panel of antibodies
(138) was performed in selected cases, especially in the metastatic lesions and in the small
endoscopic biopsies, to confirm histotype and pulmonary origin of the lesion.
Representative blocks were selected for each case, additional unstained sections were
obtained and both molecular and immunohistochemical analyses were performed.
Molecular analysis
Tumor tissue was micro-dissected from formalin-fixed paraffin-embedded (FFPE) sections to
obtain samples with at least 50% of neoplastic cells and genomic DNA was extracted. PCR
amplification of exons 18, 19, 20, 21 of EGFR was performed using previously described
primers (14). Amplicons were sequenced and analyzed on both forward and reverse strands;
mutations were verified in two independent experiments.
Immunohistochemical analysis
IHC sections (4 µm) were stained with the following primary antibodies: EGF Receptor,
E746-A750del Specific (6B6) at a working dilution of 1/200 and EGF Receptor, L858R
Mutant Specific (43B2) at a working dilution of 1/200 (Cell Signaling Technology, Inc,
Danvers, MA, USA). Each case was tested with both antibodies. For antigen retrieval,
sections were treated with pH 9 Tris-EDTA buffer for 30 minutes in water-bath at 98°C. The
slides were developed in diaminobenzidine (DAB) using the HRP Polymer (Ultravision LP
Large Volume Detection System; Lab Vision, Fremont, CA, USA) in accordance with the
manufacturer’s specifications and were counterstained with hematoxylin.
Immunoreactivity was determined with the following scoring system, as previously described
(139): 0=no staining or faint staining intensity in < 10% of tumor cells; 1+=faint staining in
40
>10% of tumor cells; 2+=moderate staining, 3+=strong staining. In differentiating score 1+
from score 2+, we found useful the presence of membrane reinforcement. Cases with faint,
diffuse cytoplasmic staining were classified as 1+, while cases with moderate staining and
focal membrane reinforcement were classified as 2+. Distinction between 1+ and 2+ is
crucial, because the subsequent statistical evaluation considered 1+ as negative and 2+, 3+ as
positive for EGFR mutational status. The IHC scoring system is summarized in this Table 4.
Cases were further classified on the basis of the pattern of immunoreactivity: patchy or
diffuse, comparing different areas of the same slide.
IHC score
Reactivity
Membrane
reinforcement
Mutational status
0
no staining or faint
staining intensity in
< 10% of tumor cells
no
not consistent with
mutation
1
faint staining intensity no
in >10% of tumor
cells
2
moderate staining
intensity
focal
3
strong staining
intensity
focal or diffuse
Table 4. Immunohistochemical scoring system.
41
consistent with
mutation
Statistical analysis
Sensitivity and specificity of IHC were calculated using the molecular status as reference. The
agreement between the two techniques was calculated with Coehn’s kappa. In comparing the
two techniques, IHC scores 0 and 1+ were considered as negative for mutational status, while
IHC scores 2+ and 3+ were interpreted as positive, as already suggested by Kawahara et al.
(140). All data were analyzed with SPSS 13.0 for Windows.
Results
We evaluated 33 cases of lung tumours that had been previously analyzed for EGFR
mutations. There were 21 females and 12 males, with age ranging from 48 to 78 (mean 62.0 ±
10.6). All cases were lung adenocarcinomas, 23 primaries (5 endobronchial biopsies, 18
pulmonary surgical resections) and 10 metastatic (7 pleural biopsies, 1 hepatic biopsy, 1
axillary lymph node resection, 1 cerebral metastasis resection). The prevalent pattern of
growth was acinar in 17 cases (in 3 of them with extensive mucinous features), lepidic in 2,
papillary in 3, solid in 11. In 1 case, a small cell neuroendocrine component was associated;
another case exhibited focal squamous differentiation.
By conventional DNA sequencing, 12 cases had EGFR mutations in exon 19 and 6 in exon
21. In exon 19, we considered 9 cases with E746-A750 deletion and 3 cases with alternative
in-frame deletions: 1 with L747-T751del in homozigotic status, 1 with L747-P753del, 1 with
E747-S752del. In exon 21, we examined 5 cases with point mutation L858R, and 1 case with
alternative point mutation L861Q+L862L. The tumor with small cell neuroendocrine
component exhibited the L858R point mutation in the glandular component. The tumor with
squamous cell differentiation was examined in the glandular component and was EGFR wildtype. Furthermore, 15 EGFR wild-type cases were evaluated as negative controls.
Overall, there were 11 cases with strong staining (2+ and 3+ scores), 4 with weak staining
(score 1+) and 18 cases that showed no immunoreactivity. Staining was diffuse in 13 cases
and patchy in 2 cases. Patchy staining was observed in 1+ cases, whereas in 2+ and 3+ cases
the percentage of stained cells was always more than 70%.
42
The E746-A750del specific antibody detected 6 of 9 cases with E746-A750del mutation (5
with score 3+, 1 with score 2+) whereas it was negative in three cases (2 with score 1+, 1 with
score 0) (kappa=0.744, sensitivity: 66.7%, specificity: 100%). The three cases with alternative
mutations on exon 19 were negative. Overall, 6 of 12 of cases with exon 19 mutations were
identified (kappa=0.560, sensitivity: 50%, specificity: 100%).
The L858R specific antibody correctly classified all five cases with the corresponding gene
mutation (4 with score 3+, 1 with score 2+) (kappa=1, sensitivity: 100%, specificity: 100%)
(Figure 4). The case with the alternative exon 21 mutation L861Q+L862L was negative.
Overall, 5/6 cases with mutations in exon 21 were detected (kappa=0.891, sensitivity: 83.3%,
specificity: 100%).
All immunoreactive cases were negative when tested with the second antibody. All 15 EGFR
wild-type control cases were negative with both antibodies (100% specificity). Among them,
two cases exhibited score 1+ immunoreactivity with the E746-A750del specific antibody,
whereas all remaining tumors scored 0.
The overall performance of the two mutation-specific antibodies in the 33 tested cases gave a
kappa value of 0.588, with 61.1% sensitivity and 100% specificity (Table 5).
43
Sensitivity
Specificity
kappa
E746A750del
(6B6)
L858R
(43B2)
E746A750del
(6B6)
L858R
(43B2)
E746A750del
(6B6)
L858R
(43B2)
Detection of
specific
mutation
6/9 (66.7%)
5/5 (100%)
100%
100%
0.744
1
Detection of
all mutations
in the same
exon
6/12 (50%)
5/6 (83.3%)
100%
100%
0.560
0.891
Overall
11/18 (61.1%)
100%
0.588
Table 5.
Summary of sensitivity, specificity and kappa values of EGFR mutation-specific antibodies
compared with molecular detection.
44
Figure 4. Immunohistochemical expression of EGFR, detecting mutation L858R on exon
21. In case N. 16 the L858R specific antibody strongly and intensely stains the neoplastic
glands; membrane reinforcement is detectable (bottom) (score 3+, consistent with mutation).
45
Discussion
The comparison between molecular and immunohistochemical methods of EGFR mutations
detection in lung adenocarcinomas demonstrated two main points:
1. the L858R antibody had higher sensitivity than the E746-A750del antibody
2. IHC showed very high specificity (100% in each comparison), but lower sensitivity
(ranging from 61.1% to 100% in the different comparisons).
The antibody L858R had 100% sensitivity in detecting L858R mutation on exon 21 and was
completely negative in the case with the alternative point mutation L861Q+L862L on the
same exon. The antibody E746-A750del was less sensitive, as it detected 6 of 9 cases with the
specific mutation and 6 of 12 cases with all mutations in exon 19.
Previous studies with the same antibodies reported conflicting results (Table 6). Kawahara
(140) described lower sensitivity for the anti E746-A750del antibody (75%), whereas Brevet
(139) , Kato (141) and Kitamura (142) found lower sensitivity for anti-L858R antibody (94%,
75%, and 32%, respectively). Overall, the sensitivity values obtained in the different studies
are comparable to ours. Nevertheless, Kitamura et al reported 32% sensitivity for L858R
(142); this study has been performed using tissue microarrays (TMA), whereas we used whole
sections, probably allowing us a more complete evaluation of the tumoral immunoreactivity.
In our series, immunoreactivity was sometimes variable in different neoplastic areas;
therefore, use of whole sections is advisable to avoid false negative results. However, we did
not consider patchy staining and, in positive cases, immunoreactivity was always more than
70%. All previous studies are concordant in showing higher specificity (from 92% to 100%)
than sensitivity (from 39% to 100%) for both antibodies (139-144). Similarly, in our series we
described very high specificity (100% in each comparison), but lower sensitivity (ranging
from 61.1% to 100% in the different comparisons).
In determining EGFR immunoreactivity, one of the crucial points was to differentiate score
2+ from 1+, because 2+ was considered positive for mutational status and 1+ negative, as
46
previously suggested by Kawahara et al. (140). However, the frequency of 2+ staining is low
as it was detected only in one case with the antibody L858R and in one case with the antibody
E746-A750del. In these cases, 2+ and 1+ immunoreactivities could be reliably distinguished
by staining intensity, percentage of positive cells and membrane reinforcement.
IHC has distinct advantages over standard sequencing methods. First of all, it is less
expensive and is more widely available. Secondly, IHC is a rapid procedure and time is
critical in treating advanced pulmonary neoplasms. Thirdly, IHC may provide reliable results
even on limited amount of material, i.e. small biopsies or cytological samples. Kawahara et al.
(145), have reported 100% sensitivity and 100% specificity in a series of 24 patients with
cytological samples composed of pleural effusions and cerebrospinal fluids. Finally, IHC
allows to detect the tissue distribution of the mutated cells. This might be useful to evaluate
cases with combined histology and to improve correlation of mutational status with
pathological features (8).
Considering the high specificity of the test, IHC may be used for up-front selection of patients
which could benefit from TKI therapy, reserving DNA sequencing for negative and/or
suspicious cases. A similar strategy is currently applied in breast carcinomas for Cerb-B2
testing. Cerb-B2 is initially evaluated by IHC that can provide negative, positive or
ambiguous results; in the latter case (score 2+) further molecular studies, i.e. fluorescence in
situ hybridization (FISH), are performed (2). Other mutation-specific antibodies are currently
being evaluated for clinical use, such as antibodies detecting EML4-ALK gene fusion
products (146).
In conclusion, mutation-specific EGFR antibodies are sufficiently accurate to be used in
routine practice to perform a first-line screening of patients candidate to TKI-therapy, as they
are less expensive and time-consuming than traditional DNA sequencing. DNA sequencing
analyses should be always performed in negative or suspicious cases.
47
Ref. (N.)
N. of
cases
studied
IHC
methodology
Genetic/molecular
test used
IHC
scoring
criteria
Brevet et
al (139)
218
TMA
PCR-RFLP assays
and sequencing for
selected cases
Kawahara
et al (140)
45
Individual
slides
Exons 19 and 21
sequencing
Kato et al
(141)
70
TMA
Exons 18 to 21
sequencing
Kitamura
et al (142)
238
TMA
Exons 19 and 21
sequencing
Simonetti
et al (143)
78
Individual
slides
Exons 19 and 21
sequencing
Ilie et al
(144)
61
TMA
Exons 19, 20 and 21
sequencing
Current
study
33
Individual
slides
Exons 18, 19, 20 and
21 sequencing
4
grades,
visual
scoring
4
grades,
visual
scoring
H score,
cut off
values
at 20
4
grades,
digital
scoring
4
grades,
visual
scoring
4
grades,
visual
scoring
4
grades,
visual
scoring
Sensitivity in detecting EGFR mutations
E746E746L858R
A750del
A750del
antibody in
antibody in
antibody in
identifying
identifying
identifying
the specific
the specific
mutations
mutation
mutation
of exon 19
20/20
17/35 (49%) 17/18 (94%)
(100%)
L858R
antibody in
identifying
mutations
of exon 19
NS
9/12 (75%)
10/19 (53%)
15/19 (79%)
No
alternative
mutations
9/11
(81.8%)
9/18 (50%)
9/12 (75%)
No
alternative
mutations
NS
16/41 (39%)
NS
12/37 (32%)
17/17
(100%)
17/29 (59%)
29 (69%)
25/25
(100%)
25/27 (93%)
8/8 (100%)
8/9 (89%)
No
mutations
in exon 21
No
mutations in
exon 21
6/9 (66.7%)
6/12 (50%)
5/5 (100%)
5/6 (83.3%)
Table 6. Summary of all previous reported studies. Legend: TMA tissue micro array, RFLP
restriction fragment length polymorphism, NS not specified.
48
REFERENCES
(1) Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clin
Pharmacol Ther 2001; 69(3):89-95.
(2) Tavassoli FA, Eusebi V. Staging of breast carcinoma and prognostic and predictive
indicators. In: Tavassoli F.A., Eusebi V., editors. AFIP Atlas of Tumor Pathology. Tumors of
the Mammary Gland. Washington DC: American Registry of Pathology, 2009: 123-148.
(3) McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM. Reporting
recommendations for tumor marker prognostic studies. J Clin Oncol 2005; 23(36):90679072.
(4) Marchio C, Dowsett M, Reis-Filho JS. Revisiting the technical validation of tumour
biomarker assays: how to open a Pandora's box. BMC Med 2011; 9:41.
(5) Bussolati G, Leonardo E. Technical pitfalls potentially affecting diagnoses in
immunohistochemistry. J Clin Pathol 2008; 61(11):1184-1192.
(6) Jiang Z, Wu CL, Woda BA, Dresser K, Xu J, Fanger GR et al. P504S/alpha-methylacyl-CoA
racemase: a useful marker for diagnosis of small foci of prostatic carcinoma on needle
biopsy. Am J Surg Pathol 2002; 26(9):1169-1174.
(7) Shinto E, Tsuda H, Ueno H, Hashiguchi Y, Hase K, Tamai S et al. Prognostic implication of
laminin-5 gamma 2 chain expression in the invasive front of colorectal cancers, disclosed
by area-specific four-point tissue microarrays. Lab Invest 2005; 85(2):257-266.
(8) Skyldberg B, Salo S, Eriksson E, Aspenblad U, Moberger B, Tryggvason K et al. Laminin-5 as
a marker of invasiveness in cervical lesions. J Natl Cancer Inst 1999; 91(21):1882-1887.
(9) Negri G, Romano F, Vittadello F, Kasal A, Mazzoleni G, Colombetti V et al. Laminin-5
gamma2 chain immunohistochemistry facilitates the assessment of invasiveness and
improves the diagnostic reproducibility of glandular lesions of the cervix uteri. Hum Pathol
2006; 37(6):704-710.
(10) Brunelli M, Eble JN, Zhang S, Martignoni G, Delahunt B, Cheng L. Eosinophilic and classic
chromophobe renal cell carcinomas have similar frequent losses of multiple chromosomes
from among chromosomes 1, 2, 6, 10, and 17, and this pattern of genetic abnormality is
not present in renal oncocytoma. Mod Pathol 2005; 18(2):161-169.
(11) Brunelli M, Delahunt B, Gobbo S, Tardanico R, Eccher A, Bersani S et al. Diagnostic
usefulness of fluorescent cytogenetics in differentiating chromophobe renal cell carcinoma
from renal oncocytoma: a validation study combining metaphase and interphase analyses.
Am J Clin Pathol 2010; 133(1):116-126.
49
(12) Louis DN, Ohgaki H, Wiestler OD, Cavenee WK. WHO Classification of Tumours of the
Central Nervous System. Fourth Edition. Lyon: IARC, 2007.
(13) Alizadeh AA, Eisen MB, Davis RE, Ma C, Lossos IS, Rosenwald A et al. Distinct types of
diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000;
403(6769):503-511.
(14) Lynch TJ, Bell DW, Sordella R, Gurubhagavatula S, Okimoto RA, Brannigan BW et al.
Activating mutations in the epidermal growth factor receptor underlying responsiveness of
non-small-cell lung cancer to gefitinib. N Engl J Med 2004; 350(21):2129-2139.
(15) Zhou C, Wu YL, Chen G, Feng J, Liu XQ, Wang C et al. Erlotinib versus chemotherapy as firstline treatment for patients with advanced EGFR mutation-positive non-small-cell lung
cancer (OPTIMAL, CTONG-0802): a multicentre, open-label, randomised, phase 3 study.
Lancet Oncol 2011; 12(8):735-742.
(16) Narayan S, Roy D. Role of APC and DNA mismatch repair genes in the development of
colorectal cancers. Mol Cancer 2003; 2:41.
(17) Karapetis CS, Khambata-Ford S, Jonker DJ, O'Callaghan CJ, Tu D, Tebbutt NC et al. K-ras
mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med 2008;
359(17):1757-1765.
(18) Munro AJ, Lain S, Lane DP. P53 abnormalities and outcomes in colorectal cancer: a
systematic review. Br J Cancer 2005; 92(3):434-444.
(19) Ishigami SI, Arii S, Furutani M, Niwano M, Harada T, Mizumoto M et al. Predictive value of
vascular endothelial growth factor (VEGF) in metastasis and prognosis of human colorectal
cancer. Br J Cancer 1998; 78(10):1379-1384.
(20) Des GG, Uzzan B, Nicolas P, Cucherat M, Morere JF, Benamouzig R et al. Microvessel
density and VEGF expression are prognostic factors in colorectal cancer. Meta-analysis of
the literature. Br J Cancer 2006; 94(12):1823-1832.
(21) Zlobec I, Steele R, Compton CC. VEGF as a predictive marker of rectal tumor response to
preoperative radiotherapy. Cancer 2005; 104(11):2517-2521.
(22) Popat S, Hubner R, Houlston RS. Systematic review of microsatellite instability and
colorectal cancer prognosis. J Clin Oncol 2005; 23(3):609-618.
(23) Ribic CM, Sargent DJ, Moore MJ, Thibodeau SN, French AJ, Goldberg RM et al. Tumor
microsatellite-instability status as a predictor of benefit from fluorouracil-based adjuvant
chemotherapy for colon cancer. N Engl J Med 2003; 349(3):247-257.
(24) Ambrosini-Spaltro A, Salvi F, Miglio R, Liguori L, Fabbri C, Piemontese A et al. High MUC2
immunohistochemical expression is a predictor of poor response to preoperative
radiochemotherapy (RCT) in rectal adenocarcinoma. Appl Immunohistochem Mol Morphol
2008; 16(3):227-231.
50
(25) Altman DG, De Stavola BL, Love SB, Stepniewska KA. Review of survival analyses published
in cancer journals. Br J Cancer 1995; 72(2):511-518.
(26) Machin D, Cheung YB, Parmar MK. Survival analysis. A practical approach. Second Edition.
ed. John Wiley & Sons Ltd, 2006.
(27) Bland M. An introduction to medical statistics. 3rd edition ed. Oxford University Press,
2000.
(28) Brownlee K.A. Statistics of the 1954 Polio Vaccine Trials. J Am Stat Assoc 1955;
50(272):1005-1013.
(29) Flyvbjerg A, Mogensen O, Mogensen B, Nielsen OS. Elevated serum insulin-like growth
factor-binding protein 2 (IGFBP-2) and decreased IGFBP-3 in epithelial ovarian cancer:
correlation with cancer antigen 125 and tumor-associated trypsin inhibitor. J Clin
Endocrinol Metab 1997; 82(7):2308-2313.
(30) el Atiq F, Garrouste F, Remacle-Bonnet M, Sastre B, Pommier G. Alterations in serum levels
of insulin-like growth factors and insulin-like growth-factor-binding proteins in patients
with colorectal cancer. Int J Cancer 1994; 57(4):491-497.
(31) Ranke MB, Maier KP, Schweizer R, Stadler B, Schleicher S, Elmlinger MW et al. Pilot study of
elevated levels of insulin-like growth factor-binding protein-2 as indicators of
hepatocellular carcinoma. Horm Res 2003; 60(4):174-180.
(32) Russo VC, Schutt BS, Andaloro E, Ymer SI, Hoeflich A, Ranke MB et al. Insulin-like growth
factor binding protein-2 binding to extracellular matrix plays a critical role in
neuroblastoma cell proliferation, migration, and invasion. Endocrinology 2005;
146(10):4445-4455.
(33) Marucci G, Morandi L, Magrini E, Farnedi A, Franceschi E, Miglio R et al. Gene expression
profiling in glioblastoma and immunohistochemical evaluation of IGFBP-2 and CDC20.
Virchows Arch 2008; 453(6):599-609.
(34) So AI, Levitt RJ, Eigl B, Fazli L, Muramaki M, Leung S et al. Insulin-like growth factor binding
protein-2 is a novel therapeutic target associated with breast cancer. Clin Cancer Res 2008;
14(21):6944-6954.
(35) Park KH, Gad E, Goodell V, Dang Y, Wild T, Higgins D et al. Insulin-like growth factor-binding
protein-2 is a target for the immunomodulation of breast cancer. Cancer Res 2008;
68(20):8400-8409.
(36) Espelund U, Cold S, Frystyk J, Orskov H, Flyvbjerg A. Elevated free IGF2 levels in localized,
early-stage breast cancer in women. Eur J Endocrinol 2008; 159(5):595-601.
(37) Flyvbjerg A, Mogensen O, Mogensen B, Nielsen OS. Elevated serum insulin-like growth
factor-binding protein 2 (IGFBP-2) and decreased IGFBP-3 in epithelial ovarian cancer:
51
correlation with cancer antigen 125 and tumor-associated trypsin inhibitor. J Clin
Endocrinol Metab 1997; 82(7):2308-2313.
(38) Chaib H, Cockrell EK, Rubin MA, Macoska JA. Profiling and verification of gene expression
patterns in normal and malignant human prostate tissues by cDNA microarray analysis.
Neoplasia 2001; 3(1):43-52.
(39) Bubendorf L, Kolmer M, Kononen J, Koivisto P, Mousses S, Chen Y et al. Hormone therapy
failure in human prostate cancer: analysis by complementary DNA and tissue microarrays. J
Natl Cancer Inst 1999; 91(20):1758-1764.
(40) Kiyama S, Morrison K, Zellweger T, Akbari M, Cox M, Yu D et al. Castration-induced
increases in insulin-like growth factor-binding protein 2 promotes proliferation of
androgen-independent human prostate LNCaP tumors. Cancer Res 2003; 63(13):35753584.
(41) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(42) Tennant MK, Thrasher JB, Twomey PA, Birnbaum RS, Plymate SR. Insulin-like growth factorbinding protein-2 and -3 expression in benign human prostate epithelium, prostate
intraepithelial neoplasia, and adenocarcinoma of the prostate. J Clin Endocrinol Metab
1996; 81(1):411-420.
(43) Thrasher JB, Tennant MK, Twomey PA, Hansberry KL, Wettlaufer JN, Plymate SR.
Immunohistochemical localization of insulin-like growth factor binding proteins 2 and 3 in
prostate tissue: clinical correlations. J Urol 1996; 155(3):999-1003.
(44) Inman BA, Harel F, Audet JF, Meyer F, Douville P, Fradet Y et al. Insulin-like growth factor
binding protein 2: an androgen-dependent predictor of prostate cancer survival. Eur Urol
2005; 47(5):695-702.
(45) Ho PJ, Baxter RC. Insulin-like growth factor-binding protein-2 in patients with prostate
carcinoma and benign prostatic hyperplasia. Clin Endocrinol (Oxf) 1997; 46(3):333-342.
(46) Cohen P, Peehl DM, Stamey TA, Wilson KF, Clemmons DR, Rosenfeld RG. Elevated levels of
insulin-like growth factor-binding protein-2 in the serum of prostate cancer patients. J Clin
Endocrinol Metab 1993; 76(4):1031-1035.
(47) Kanety H, Madjar Y, Dagan Y, Levi J, Papa MZ, Pariente C et al. Serum insulin-like growth
factor-binding protein-2 (IGFBP-2) is increased and IGFBP-3 is decreased in patients with
prostate cancer: correlation with serum prostate-specific antigen. J Clin Endocrinol Metab
1993; 77(1):229-233.
(48) Yu H, Nicar MR, Shi R, Berkel HJ, Nam R, Trachtenberg J et al. Levels of insulin-like growth
factor I (IGF-I) and IGF binding proteins 2 and 3 in serial postoperative serum samples and
risk of prostate cancer recurrence. Urology 2001; 57(3):471-475.
52
(49) Roddam AW, Allen NE, Appleby P, Key TJ, Ferrucci L, Carter HB et al. Insulin-like growth
factors, their binding proteins, and prostate cancer risk: analysis of individual patient data
from 12 prospective studies. Ann Intern Med 2008; 149(7):461-468.
(50) Rowlands MA, Gunnell D, Harris R, Vatten LJ, Holly JM, Martin RM. Circulating insulin-like
growth factor peptides and prostate cancer risk: a systematic review and meta-analysis. Int
J Cancer 2009; 124(10):2416-2429.
(51) Jiang Z, Wu CL, Woda BA, Dresser K, Xu J, Fanger GR et al. P504S/alpha-methylacyl-CoA
racemase: a useful marker for diagnosis of small foci of prostatic carcinoma on needle
biopsy. Am J Surg Pathol 2002; 26(9):1169-1174.
(52) Beach R, Gown AM, Peralta-Venturina MN, Folpe AL, Yaziji H, Salles PG et al. P504S
immunohistochemical detection in 405 prostatic specimens including 376 18-gauge needle
biopsies. Am J Surg Pathol 2002; 26(12):1588-1596.
(53) Edge S, Byrd D, Compton C, Fritz A, Greene F, Trotti A. AJCC Cancer Staging Manual. 7th
edition ed. Springer, 2009.
(54) Epstein JI, Allsbrook WC, Jr., Amin MB, Egevad LL. The 2005 International Society of
Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic
Carcinoma. Am J Surg Pathol 2005; 29(9):1228-1242.
(55) McCarty KS, Jr., Miller LS, Cox EB, Konrath J, McCarty KS, Sr. Estrogen receptor analyses.
Correlation of biochemical and immunohistochemical methods using monoclonal
antireceptor antibodies. Arch Pathol Lab Med 1985; 109(8):716-721.
(56) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(57) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(58) Tennant MK, Thrasher JB, Twomey PA, Birnbaum RS, Plymate SR. Insulin-like growth factorbinding protein-2 and -3 expression in benign human prostate epithelium, prostate
intraepithelial neoplasia, and adenocarcinoma of the prostate. J Clin Endocrinol Metab
1996; 81(1):411-420.
(59) Thrasher JB, Tennant MK, Twomey PA, Hansberry KL, Wettlaufer JN, Plymate SR.
Immunohistochemical localization of insulin-like growth factor binding proteins 2 and 3 in
prostate tissue: clinical correlations. J Urol 1996; 155(3):999-1003.
(60) Tennant MK, Thrasher JB, Twomey PA, Birnbaum RS, Plymate SR. Insulin-like growth factorbinding protein-2 and -3 expression in benign human prostate epithelium, prostate
intraepithelial neoplasia, and adenocarcinoma of the prostate. J Clin Endocrinol Metab
1996; 81(1):411-420.
53
(61) Thrasher JB, Tennant MK, Twomey PA, Hansberry KL, Wettlaufer JN, Plymate SR.
Immunohistochemical localization of insulin-like growth factor binding proteins 2 and 3 in
prostate tissue: clinical correlations. J Urol 1996; 155(3):999-1003.
(62) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(63) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(64) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(65) Tennant MK, Thrasher JB, Twomey PA, Birnbaum RS, Plymate SR. Insulin-like growth factorbinding protein-2 and -3 expression in benign human prostate epithelium, prostate
intraepithelial neoplasia, and adenocarcinoma of the prostate. J Clin Endocrinol Metab
1996; 81(1):411-420.
(66) Thrasher JB, Tennant MK, Twomey PA, Hansberry KL, Wettlaufer JN, Plymate SR.
Immunohistochemical localization of insulin-like growth factor binding proteins 2 and 3 in
prostate tissue: clinical correlations. J Urol 1996; 155(3):999-1003.
(67) Figueroa JA, De Raad S, Tadlock L, Speights VO, Rinehart JJ. Differential expression of
insulin-like growth factor binding proteins in high versus low Gleason score prostate
cancer. J Urol 1998; 159(4):1379-1383.
(68) Richardsen E, Ukkonen T, Bjornsen T, Mortensen E, Egevad L, Busch C. Overexpression of
IGBFB2 is a marker for malignant transformation in prostate epithelium. Virchows Arch
2003; 442(4):329-335.
(69) Tennant MK, Thrasher JB, Twomey PA, Birnbaum RS, Plymate SR. Insulin-like growth factorbinding protein-2 and -3 expression in benign human prostate epithelium, prostate
intraepithelial neoplasia, and adenocarcinoma of the prostate. J Clin Endocrinol Metab
1996; 81(1):411-420.
(70) Degraff DJ, Aguiar AA, Sikes RA. Disease evidence for IGFBP-2 as a key player in prostate
cancer progression and development of osteosclerotic lesions. Am J Transl Res 2009;
1(2):115-130.
(71) Kiyama S, Morrison K, Zellweger T, Akbari M, Cox M, Yu D et al. Castration-induced
increases in insulin-like growth factor-binding protein 2 promotes proliferation of
androgen-independent human prostate LNCaP tumors. Cancer Res 2003; 63(13):35753584.
54
(72) Bubendorf L, Kolmer M, Kononen J, Koivisto P, Mousses S, Chen Y et al. Hormone therapy
failure in human prostate cancer: analysis by complementary DNA and tissue microarrays. J
Natl Cancer Inst 1999; 91(20):1758-1764.
(73) Kiyama S, Morrison K, Zellweger T, Akbari M, Cox M, Yu D et al. Castration-induced
increases in insulin-like growth factor-binding protein 2 promotes proliferation of
androgen-independent human prostate LNCaP tumors. Cancer Res 2003; 63(13):35753584.
(74) Moore MG, Wetterau LA, Francis MJ, Peehl DM, Cohen P. Novel stimulatory role for
insulin-like growth factor binding protein-2 in prostate cancer cells. Int J Cancer 2003;
105(1):14-19.
(75) Inman BA, Harel F, Audet JF, Meyer F, Douville P, Fradet Y et al. Insulin-like growth factor
binding protein 2: an androgen-dependent predictor of prostate cancer survival. Eur Urol
2005; 47(5):695-702.
(76) Inman BA, Harel F, Audet JF, Meyer F, Douville P, Fradet Y et al. Insulin-like growth factor
binding protein 2: an androgen-dependent predictor of prostate cancer survival. Eur Urol
2005; 47(5):695-702.
(77) Ho PJ, Baxter RC. Insulin-like growth factor-binding protein-2 in patients with prostate
carcinoma and benign prostatic hyperplasia. Clin Endocrinol (Oxf) 1997; 46(3):333-342.
(78) Cohen P, Peehl DM, Stamey TA, Wilson KF, Clemmons DR, Rosenfeld RG. Elevated levels of
insulin-like growth factor-binding protein-2 in the serum of prostate cancer patients. J Clin
Endocrinol Metab 1993; 76(4):1031-1035.
(79) Kanety H, Madjar Y, Dagan Y, Levi J, Papa MZ, Pariente C et al. Serum insulin-like growth
factor-binding protein-2 (IGFBP-2) is increased and IGFBP-3 is decreased in patients with
prostate cancer: correlation with serum prostate-specific antigen. J Clin Endocrinol Metab
1993; 77(1):229-233.
(80) Yu H, Nicar MR, Shi R, Berkel HJ, Nam R, Trachtenberg J et al. Levels of insulin-like growth
factor I (IGF-I) and IGF binding proteins 2 and 3 in serial postoperative serum samples and
risk of prostate cancer recurrence. Urology 2001; 57(3):471-475.
(81) Roddam AW, Allen NE, Appleby P, Key TJ, Ferrucci L, Carter HB et al. Insulin-like growth
factors, their binding proteins, and prostate cancer risk: analysis of individual patient data
from 12 prospective studies. Ann Intern Med 2008; 149(7):461-468.
(82) Rowlands MA, Gunnell D, Harris R, Vatten LJ, Holly JM, Martin RM. Circulating insulin-like
growth factor peptides and prostate cancer risk: a systematic review and meta-analysis. Int
J Cancer 2009; 124(10):2416-2429.
(83) Inman BA, Harel F, Audet JF, Meyer F, Douville P, Fradet Y et al. Insulin-like growth factor
binding protein 2: an androgen-dependent predictor of prostate cancer survival. Eur Urol
2005; 47(5):695-702.
55
(84) Burchardt T, Burchardt M, Chen MW, Cao Y, de la TA, Shabsigh A et al. Transdifferentiation
of prostate cancer cells to a neuroendocrine cell phenotype in vitro and in vivo. J Urol
1999; 162(5):1800-1805.
(85) Abrahamsson PA. Neuroendocrine differentiation in prostatic carcinoma. Prostate 1999;
39(2):135-148.
(86) Mazzucchelli R, Morichetti D, Lopez-Beltran A, Cheng L, Scarpelli M, Kirkali Z et al.
Neuroendocrine tumours of the urinary system and male genital organs: clinical
significance. BJU Int 2009; 103(11):1464-1470.
(87) Ambrosini-Spaltro A, Farnedi A, Montironi R, Foschini MP. IGFBP2 as an
immunohistochemical marker for prostatic adenocarcinoma. Appl Immunohistochem Mol
Morphol 2011; 19(4):318-328.
(88) Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden
of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010.
(89) Kreppel M, Drebber U, Rothamel D, Eich HT, Kubler A, Scheer M et al. Prognostic impact of
different TNM-based stage groupings for oral squamous cell carcinoma. Head Neck 2010.
(90) Kreppel M, Eich HT, Kubler A, Zoller JE, Scheer M. Prognostic value of the sixth edition of
the UICC's TNM classification and stage grouping for oral cancer. J Surg Oncol 2010;
102(5):443-449.
(91) Arduino PG, Carrozzo M, Chiecchio A, Broccoletti R, Tirone F, Borra E et al. Clinical and
histopathologic independent prognostic factors in oral squamous cell carcinoma: a
retrospective study of 334 cases. J Oral Maxillofac Surg 2008; 66(8):1570-1579.
(92) Pentenero M, Gandolfo S, Carrozzo M. Importance of tumor thickness and depth of
invasion in nodal involvement and prognosis of oral squamous cell carcinoma: a review of
the literature. Head Neck 2005; 27(12):1080-1091.
(93) Fukano H, Matsuura H, Hasegawa Y, Nakamura S. Depth of invasion as a predictive factor
for cervical lymph node metastasis in tongue carcinoma. Head Neck 1997; 19(3):205-210.
(94) Breuninger H, Schaumburg-Lever G, Holzschuh J, Horny HP. Desmoplastic squamous cell
carcinoma of skin and vermilion surface: a highly malignant subtype of skin cancer. Cancer
1997; 79(5):915-919.
(95) Dorta RG, Landman G, Kowalski LP, Lauris JR, Latorre MR, Oliveira DT. Tumour-associated
tissue eosinophilia as a prognostic factor in oral squamous cell carcinomas. Histopathology
2002; 41(2):152-157.
(96) Forastiere A, Koch W, Trotti A, Sidransky D. Head and neck cancer. N Engl J Med 2001;
345(26):1890-1900.
56
(97) Williams HK. Molecular pathogenesis of oral squamous carcinoma. Mol Pathol 2000;
53(4):165-172.
(98) Suzuki H, Sugimura H, Hashimoto K. p16INK4A in oral squamous cell carcinomas--a
correlation with biological behaviors: immunohistochemical and FISH analysis. J Oral
Maxillofac Surg 2006; 64(11):1617-1623.
(99) de Oliveira LR, Ribeiro-Silva A, Zucoloto S. Prognostic impact of p53 and p63
immunoexpression in oral squamous cell carcinoma. J Oral Pathol Med 2007; 36(4):191197.
(100) Michalides RJ, van Veelen NM, Kristel PM, Hart AA, Loftus BM, Hilgers FJ et al.
Overexpression of cyclin D1 indicates a poor prognosis in squamous cell carcinoma of the
head and neck. Arch Otolaryngol Head Neck Surg 1997; 123(5):497-502.
(101) Coughlin A, Resto VA. Oral cavity squamous cell carcinoma and the clinically n0 neck: the
past, present, and future of sentinel lymph node biopsy. Curr Oncol Rep 2010; 12(2):129135.
(102) Carlsson P, Presto J, Spillmann D, Lindahl U, Kjellen L. Heparin/heparan sulfate
biosynthesis: processive formation of N-sulfated domains. J Biol Chem 2008;
283(29):20008-20014.
(103) Kolset SO, Salmivirta M. Cell surface heparan sulfate proteoglycans and lipoprotein
metabolism. Cell Mol Life Sci 1999; 56(9-10):857-870.
(104) Montreuil J, Vliegenthart JFG, Schachter H. Glycoproteins II. New York: Elsevier, 1997.
(105) David G. Integral membrane heparan sulfate proteoglycans. FASEB J 1993; 7(11):10231030.
(106) Kjellen L, Lindahl U. Proteoglycans: structures and interactions. Annu Rev Biochem 1991;
60:443-475.
(107) Couchman JR. Syndecans: proteoglycan regulators of cell-surface microdomains? Nat Rev
Mol Cell Biol 2003; 4(12):926-937.
(108) Filmus J, Capurro M, Rast J. Glypicans. Genome Biol 2008; 9(5):224.
(109) Stallcup WB. The NG2 proteoglycan: past insights and future prospects. J Neurocytol 2002;
31(6-7):423-435.
(110) Trotter J, Karram K, Nishiyama A. NG2 cells: Properties, progeny and origin. Brain Res Rev
2010; 63(1-2):72-82.
(111) Blackhall FH, Merry CL, Davies EJ, Jayson GC. Heparan sulfate proteoglycans and cancer. Br
J Cancer 2001; 85(8):1094-1098.
57
(112) Iozzo RV. Proteoglycans and the intercellular tumor matrix. Curr Top Pathol 1987; 77:207221.
(113) Inki P, Joensuu H, Grenman R, Klemi P, Jalkanen M. Association between syndecan-1
expression and clinical outcome in squamous cell carcinoma of the head and neck. Br J
Cancer 1994; 70(2):319-323.
(114) Anttonen A, Kajanti M, Heikkila P, Jalkanen M, Joensuu H. Syndecan-1 expression has
prognostic significance in head and neck carcinoma. Br J Cancer 1999; 79(3-4):558-564.
(115) Mukunyadzi P, Liu K, Hanna EY, Suen JY, Fan CY. Induced expression of syndecan-1 in the
stroma of head and neck squamous cell carcinoma. Mod Pathol 2003; 16(8):796-801.
(116) Han I, Park H, Oh ES. New insights into syndecan-2 expression and tumourigenic activity in
colon carcinoma cells. J Mol Histol 2004; 35(3):319-326.
(117) Li M, Shuman C, Fei YL, Cutiongco E, Bender HA, Stevens C et al. GPC3 mutation analysis in
a spectrum of patients with overgrowth expands the phenotype of Simpson-Golabi-Behmel
syndrome. Am J Med Genet 2001; 102(2):161-168.
(118) Kim H, Xu GL, Borczuk AC, Busch S, Filmus J, Capurro M et al. The heparan sulfate
proteoglycan GPC3 is a potential lung tumor suppressor. Am J Respir Cell Mol Biol 2003;
29(6):694-701.
(119) Chekenya M, Hjelstuen M, Enger PO, Thorsen F, Jacob AL, Probst B et al. NG2 proteoglycan
promotes angiogenesis-dependent tumor growth in CNS by sequestering angiostatin.
FASEB J 2002; 16(6):586-588.
(120) Hans CP, Weisenburger DD, Greiner TC, Gascoyne RD, Delabie J, Ott G et al. Confirmation
of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry
using a tissue microarray. Blood 2004; 103(1):275-282.
(121) Zimpfer A, Schonberg S, Lugli A, Agostinelli C, Pileri SA, Went P et al. Construction and
validation of a bone marrow tissue microarray. J Clin Pathol 2007; 60(1):57-61.
(122) Monteiro LS, Diniz-Freitas M, Garcia-Caballero T, Forteza J, Fraga M. EGFR and Ki-67
expression in oral squamous cell carcinoma using tissue microarray technology. J Oral
Pathol Med 2010; 39(7):571-578.
(123) Gallagher JT. The extended family of proteoglycans: social residents of the pericellular
zone. Curr Opin Cell Biol 1989; 1(6):1201-1218.
(124) Ruoslahti E. Integrins. J Clin Invest 1991; 87(1):1-5.
(125) Takeichi M. Cadherin cell adhesion receptors as a morphogenetic regulator. Science 1991;
251(5000):1451-1455.
58
(126) Bernfield M, Kokenyesi R, Kato M, Hinkes MT, Spring J, Gallo RL et al. Biology of the
syndecans: a family of transmembrane heparan sulfate proteoglycans. Annu Rev Cell Biol
1992; 8:365-393.
(127) Barbareschi M, Maisonneuve P, Aldovini D, Cangi MG, Pecciarini L, Angelo MF et al. High
syndecan-1 expression in breast carcinoma is related to an aggressive phenotype and to
poorer prognosis. Cancer 2003; 98(3):474-483.
(128) Zellweger T, Ninck C, Mirlacher M, Annefeld M, Glass AG, Gasser TC et al. Tissue microarray
analysis reveals prognostic significance of syndecan-1 expression in prostate cancer.
Prostate 2003; 55(1):20-29.
(129) Ito Y, Yoshida H, Nakano K, Takamura Y, Miya A, Kobayashi K et al. Syndecan-1 expression
in thyroid carcinoma: stromal expression followed by epithelial expression is significantly
correlated with dedifferentiation. Histopathology 2003; 43(2):157-164.
(130) Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tan PF et al. Human
papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010;
363(1):24-35.
(131) Hynes NE, Lane HA. ERBB receptors and cancer: the complexity of targeted inhibitors. Nat
Rev Cancer 2005; 5(5):341-354.
(132) Mok TS, Wu YL, Thongprasert S, Yang CH, Chu DT, Saijo N et al. Gefitinib or carboplatinpaclitaxel in pulmonary adenocarcinoma. N Engl J Med 2009; 361(10):947-957.
(133) Marchetti A, Normanno N, Pinto C, Taddei GL, Adamo V, Ardizzoni A et al.
Recommendations for mutational analysis of EGFR in lung carcinoma. Pathologica 2010;
102(3):119-126.
(134) Kosaka T, Yatabe Y, Endoh H, Kuwano H, Takahashi T, Mitsudomi T. Mutations of the
epidermal growth factor receptor gene in lung cancer: biological and clinical implications.
Cancer Res 2004; 64(24):8919-8923.
(135) Yatabe Y, Mitsudomi T. Epidermal growth factor receptor mutations in lung cancers. Pathol
Int 2007; 57(5):233-244.
(136) Yu J, Kane S, Wu J, Benedettini E, Li D, Reeves C et al. Mutation-specific antibodies for the
detection of EGFR mutations in non-small-cell lung cancer. Clin Cancer Res 2009;
15(9):3023-3028.
(137) Travis WD, Brambilla E, Noguchi M, Nicholson AG, Geisinger KR, Yatabe Y et al.
International association for the study of lung cancer/american thoracic society/european
respiratory society international multidisciplinary classification of lung adenocarcinoma. J
Thorac Oncol 2011; 6(2):244-285.
59
(138) Rossi G, Pelosi G, Graziano P, Barbareschi M, Papotti M. A reevaluation of the clinical
significance of histological subtyping of non--small-cell lung carcinoma: diagnostic
algorithms in the era of personalized treatments. Int J Surg Pathol 2009; 17(3):206-218.
(139) Brevet M, Arcila M, Ladanyi M. Assessment of EGFR mutation status in lung
adenocarcinoma by immunohistochemistry using antibodies specific to the two major
forms of mutant EGFR. J Mol Diagn 2010; 12(2):169-176.
(140) Kawahara A, Yamamoto C, Nakashima K, Azuma K, Hattori S, Kashihara M et al. Molecular
diagnosis of activating EGFR mutations in non-small cell lung cancer using mutation-specific
antibodies for immunohistochemical analysis. Clin Cancer Res 2010; 16(12):3163-3170.
(141) Kato Y, Peled N, Wynes MW, Yoshida K, Pardo M, Mascaux C et al. Novel epidermal growth
factor receptor mutation-specific antibodies for non-small cell lung cancer:
immunohistochemistry as a possible screening method for epidermal growth factor
receptor mutations. J Thorac Oncol 2010; 5(10):1551-1558.
(142) Kitamura A, Hosoda W, Sasaki E, Mitsudomi T, Yatabe Y. Immunohistochemical detection
of EGFR mutation using mutation-specific antibodies in lung cancer. Clin Cancer Res 2010;
16(13):3349-3355.
(143) Simonetti S, Molina MA, Queralt C, de A, I, Mayo C, Bertran-Alamillo J et al. Detection of
EGFR mutations with mutation-specific antibodies in stage IV non-small-cell lung cancer. J
Transl Med 2010; 8:135.
(144) Ilie MI, Hofman V, Bonnetaud C, Havet K, Lespinet-Fabre V, Coelle C et al. Usefulness of
tissue microarrays for assessment of protein expression, gene copy number and mutational
status of EGFR in lung adenocarcinoma. Virchows Arch 2010; 457(4):483-495.
(145) Kawahara A, Azuma K, Sumi A, Taira T, Nakashima K, Aikawa E et al. Identification of nonsmall-cell lung cancer with activating EGFR mutations in malignant effusion and
cerebrospinal fluid: Rapid and sensitive detection of exon 19 deletion E746-A750 and exon
21 L858R mutation by immunocytochemistry. Lung Cancer 2011.
(146) Mino-Kenudson M, Chirieac LR, Law K, Hornick JL, Lindeman N, Mark EJ et al. A novel,
highly sensitive antibody allows for the routine detection of ALK-rearranged lung
adenocarcinomas by standard immunohistochemistry. Clin Cancer Res 2010; 16(5):15611571.
60
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement