| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinological Oncology |
Istituto di Medicina Interna e di Malattie Endocrine e del Metabolismo, Cattedra di Endocrinologia, University of Catania; Servizio di Anatomia Patologica, Ospedale V. Emanuele II (P.G., G.R., A.F.); and Cattedra di Chirurgia, University of Catania (O.I.), Catania; and Dipartimento di Scienze Biomediche ed Oncologia, University of Torino (M.F.D.R., P.C.), Torino, Italy
Address all correspondence and requests for reprints to: Antonino Belfiore, M.D., Istituto di Medicina Interna e di Malattie Endocrine e del Metabolismo, Cattedra di Endocrinologia, Ospedale Garibaldi, Piazza S. Maria di Gesù 1, 95123 Catania, Italy. E-mail: segmeint{at}mbox.unict.it
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Met/HGF-R was absent in the normal thyroid tissue, absent or focally
expressed in follicular and anaplastic tumors, and expressed at various
levels in most papillary carcinomas, including microcarcinomas.
Papillary carcinomas were thus categorized as having negative/low
Met/HGF-R (n = 50; total score, 
5) or high Met/HGF-R expression
(n = 70; total score, >5). High Met/HGF-R was inversely
associated with vascular invasion (P = 0.0308), but
not with other prognostic factors. Negative/low Met/HGF-R expression
was the most effective predictor by multivariate Cox analysis of
distant metastases (hazard ratio = 9.71; P =
0.0036), higher than extrathyroid invasion (hazard ratio = 4.25;
P = 0.0181), age (45 vs. >45 yr;
hazard ratio = 3.99; P = 0.0099), and vascular
invasion (hazard ratio = 3.19; P = 0.0358).
These findings suggest a role for Met/HGF-R in papillary thyroid cancer
and its clinical use to select patients with a high risk of distant
metastases.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We previously found, by Western blot analysis, that the Met/hepatocyte
growth factor receptor (Met/HGF-R) is aberrantly expressed in
approximately 70% of papillary and one third of follicular carcinomas.
We, therefore, hypothesized that Met/HGF-R has a role in thyroid cancer
(13). Met/HGF-R is the receptor for the HGF/scatter factor (HGF/SF), a
factor secreted by cells of mesenchymal origin that stimulates
proliferation, motility, and morphogenesis in epithelial cells (14, 15). The Met/HGF-R is a 190,000 molecular mass heterodimeric
glycoprotein composed of an extracellular 
-subunit of 50,000 and a
transmembrane ß-subunit of 145,000 linked by disulfide bonds. The
cytoplasmic domain of the ß-subunit has ligand-dependent tyrosine
kinase activity. Met/HGF-R is expressed in most epithelial cells and is
overexpressed in a proportion of colonic (16), pancreatic (17), ovarian
(18), prostatic (19), bone (20), and gastric carcinomas (21). In
gastric carcinomas, the c-met protooncogene has been found
to be amplified, and Met/HGF-R constitutively activated.
As Western blot analysis requires fresh tissue and, therefore, is not suitable to study a consecutive series of carcinomas, in the attempt to clarify the possible clinical usefulness of Met/HGF-R measurement in thyroid cancer, we evaluated Met/HGF-R expression in a large series of archival benign and malignant thyroid tumors by immunohistochemistry. High Met/HGF-R expression was observed exclusively in papillary carcinomas, suggesting a role for the HGF-Met/HGF-R system in the biology of this thyroid cancer. Furthermore, in papillary tumors, negative/low Met/HGF-R expression was the most effective predictor of hematogenous metastatic spread.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This retrospective study was carried out in 212 thyroid specimens: 163 from thyroid cancer of follicular origin (129 papillary, 21 follicular, and 13 anaplastic) and 49 from thyroid adenomas. Patients had undergone thyroidectomy between 1984–1991 by the same surgical team and were followed-up at our Thyroid Clinic. Patients with differentiated thyroid cancer underwent total thyroidectomy plus paratracheal lymph node dissection. Primary total thyroidectomy was carried out in 141 patients on the basis of a cytological diagnosis of malignancy and/or as determined by frozen sections (22). Lobectomy with isthmectomy was carried out in the remaining 9 patients and was followed by completion total thyroidectomy after the histological diagnosis was available. Patients with anaplastic carcinomas underwent palliative thyroidectomy. The 49 patients with thyroid benign neoplasms included 9 patients with an autonomously functioning follicular adenoma and 40 with a hypofunctioning (cold at scintiscan) follicular adenoma.
Postoperative patient follow-up
Patients operated upon for differentiated thyroid cancer were evaluated for residual and/or metastatic tumor tissue 6–8 weeks after thyroidectomy and 2 weeks after withdrawal of hormone treatment (L-T3) to produce a TSH serum level greater than 30 mU/L. The serum thyroglobulin level was also measured, and a tracer dose of 37 megabecquerels (MBq; 100 µCi) 131I was administered to evaluate the 24-h cervical uptake. Patients with an uptake of 2% or less were given 185 MBq (5 mCi) 131I and evaluated 72 h later by total body scan (TBS). Patients with a cervical uptake of greater than 2% were given 1110 MBq (30 mCi) 131I to ablate the thyroid remnant and underwent TBS 5 days later. Residual or metastatic tumoral tissue was considered present when TBS was positive and also when, in the absence of a thyroid remnant, the serum thyroglobulin concentration in the hypothyroid patient was more than 10 µg/L (up to 1989, as measured by RIA) or more than 5 µg/L (after 1989, as measured by immunoradiometric method). In most cases distant metastases were localized by additional imaging tests (standard x-rays, computed axial tomographic scan, nuclear magnetic resonance imaging, or bone scan). Nonsurgically removable distant metastases were treated with 131I (3700 to 5550 MBq; 100–150 mCi) every 6–12 months. Diagnosis of metastatic lymph nodes was made on the basis of neck ultrasounds and radioiodine uptake at TBS or by fine needle aspiration cytology and/or high thyroglobulin levels in the aspirates.
Tissues
Archival thyroid tissues. Standard hematoxylin and eosin sections were obtained from the same block used for immunohistochemistry and reviewed by the same pathologist. Tumor size, histological type and variant (23), presence of multiple foci, extrathyroid tumor extension, blood vessel invasion, and nodal involvement were all recorded.
Frozen thyroid tissues and cryostat sections. Fresh tissue specimens were collected at surgery from 20 thyroid tumors, immediately placed in liquid nitrogen, and stored until processing. Cryostat sections (6-µm thick) were obtained.
Met/HGF-R evaluation
Anti-Met/HGF-R antibodies. Antibody sc161 (Santa Cruz Biotechnology, Santa Cruz, CA) is an affinity-purified rabbit IgG raised against a synthetic peptide corresponding to the last 12 amino acids at the carboxy-terminal of human Met/HGF-R. This antibody is specific for Met/HGF-R, as indicated by immunoprecipitation and immunoblotting (24). An optimal IgG concentration of 0.5 µg/mL, as found by serial dilutions, was used for immunostaining and Western blot.
Immunohistochemistry of frozen sections was carried out using, in addition to the sc-161 polyclonal antibody, the monoclonal antibody DO-24, directed against the extracellular domain of Met/HGF-R (25). It was used at a dilution of 1:100 (ascitic fluid). Antibody DQ-13, directed against a peptide corresponding to the last 19 amino acids (Ser1372 to Ser1390) of human Met/HGF-R carboxy-terminal (26), was used for Western blot at a dilution of 1:100 (ascitic fluid).
Immunohistochemical staining procedure. Cryostat sections were air-dried and fixed in a chloroform-acetone mixture (1:1). Paraffin-embedded sections were deparaffinized, treated with 3% hydrogen peroxide (H2O2), and processed using the biotin-streptavidin-peroxidase method; after incubation with approximately 100 µL primary antibody at room temperature for 60 min, sections were sequentially incubated with the secondary antibody (biotinylated IgGs) and the streptavidin-biotin complex. 3-Amino-9-ethylcarbazole in N,N-dimethylformamide was used as substrate. Counterstaining was carried out with Mayer’s hematoxylin. Negative controls included omission of the primary antibody and, for antibody sc-161, preincubation with the Met/HGF-R oligopeptide used for raising the antibody (30 µg/mL for 30 min).
Evaluation of immunohistochemical staining. The proportion of stained cells was determined by examining the entire slide and scored as 0 (none), 1 (<1%), 2 (1–10%), 3 (10–33%), 4 (33–66%), and 5 (>66%). The intensity of positively stained cells was indicated as 0 (negative), 1 (very weak), 2 (weak), 3 (moderate), 4 (strong), and 5 (very strong). The proportion and the intensity scores were then combined in the total score (range, 0–10). The staining was considered negative if the total score was 0 and positive if the total score was in the range of 2–10.
To minimize variations in staining intensity among different experiments several steps were taken: 1) a positive and a negative control were routinely included to check the staining procedure; 2) as the smooth muscle cells of vessels are weakly reactive whereas fibroblasts are consistently negative, these cells also were taken as internal controls; 3) the same batch of antibody was used for all slides; and 4) preincubation with the Met/HGF-R oligopeptide used as immunogen was used to check staining specificity.
Immunostaining was independently evaluated by two investigators (A.F. and A.B.) who had no previous knowledge of the clinical data. In case of different evaluations, the lower score was adopted. Interobserver variability of the scoring system, checked by blind evaluation of 50 cases, was less than 3%.
Western blot analysis. Frozen tissues were pulverized in the presence of liquid nitrogen (by Mikro-Dismembrator, B-Braun) and solubilized in boiling Laemmli buffer (13). Forty micrograms of proteins were then loaded in each lane, separated by SDS-PAGE, and transferred to nitrocellulose membranes. The membranes were incubated with DQ-13 monoclonal antibody and then with a second antibody conjugated with horseradish peroxidase and developed by an enhanced chemiluminescence detection system (Amersham International, Aylesbury, UK). To compare results obtained with different antibodies, filters were stripped and then incubated with antibody sc-161. The signal was quantitated by laser densitometry.
Statistical analysis
The distribution of clinical and pathological variables in the
two groups with different levels of Met/HGF-R expression was compared
with the use of contingency tables and 
2 test. The
prognostic impact of Met/HGF-R and of clinical and pathological
variables was assessed by univariate and multivariate analysis
according to the Cox proportional hazards regression model (27) using
the occurrence of distant metastases as end point. All important
prognostic factors, as shown by univariate analysis, were included in
the multivariate model. Kaplan-Meier plots were used to demonstrate the
difference in the probability to develop distant metastases between
groups, and the log rank test was used to evaluate the differences
between curves.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In preliminary experiments, antibodies sc-161, DQ-13, and DO-24
were compared for optimal staining. Best results were obtained with
antibody DO-24 in frozen sections and with antibody sc-161 in
paraffin-embedded sections. The pattern of immunostaining, the
proportion of positive cells, and staining intensity were, thereafter,
compared in frozen and in paraffin-embedded sections obtained from 20
thyroid carcinomas. In frozen sections, immunoreactivity was
exclusively located at the cell membrane (Fig. 1A
), as
previously reported (28), whereas in formalin-fixed paraffin-embedded
sections, immunoreactivity was found both in the cytoplasm and, more
intensely, at the cell membrane (Fig. 1B). Results obtained in parallel
experiments carried out in paraffin-embedded sections and in frozen
sections were, however, very similar (total scores were highly
correlated by linear regression analysis: r = 0.85;
P = 0.0001). Met/HGF-R immunostaining was almost
exclusively localized at the level of epithelial neoplastic cells,
except for a weak staining of smooth muscle cells of small vessels. In
the paranodular thyroid tissue, Met/HGF-R was generally undetectable,
except for occasional and weakly stained follicles (Fig. 1C).
|
).
|
Differentiated carcinomas.
Papillary carcinomas:
One hundred and seventeen of the 129 papillary carcinomas examined
showed Met/HGF-R immunoreactivity; 74 (57.4%) were intensely and
homogeneously stained, 43 (33.3%) were positive focally and at a low
level, and 12 (9.3%) were negative (Table 1 and Fig. 1
, A–C).
|
).
Undifferentiated carcinomas.
Anaplastic carcinomas were either
Met/HGF-R negative 6 (46.1%) or positive focally and at a low level
(53.8%, Table 1).
Benign tumors.
Met/HGF-R expression was intense in 1 of 49
thyroid adenomas, focally present in 32, and absent in 16 cases.
Met/HGF-R expression was more frequent in autonomously functioning
adenomas than in cold adenomas (P = 0.011, by
2 test). No significant difference was observed in the
proportion of Met/HGF-R-positive cases between typical and atypical
cold adenomas (Table 1).
Patient follow-up
Of the 129 patients with papillary carcinoma, 9 were lost to follow-up (after 8–18 months) and were, therefore, excluded from further analysis. Five of them were in the negative/low Met/HGF-R group, and 4 were in the high Met/HGF-R group (see below). None of them presented distant metastases at the time of the last examination. The overall follow-up time for patients with differentiated thyroid carcinomas (120 papillary and 21 follicular) ranged from 25–147 months, with a median follow-up of 67 months.
Relationship between Met/HGF-R expression and tumor histopathological characteristics, and clinical evolution
Patients with papillary cancer were subdivided into 2 subgroups, 1
with negative/low Met/HGF-R (n = 50; total score, 5) and 1 with
high Met/HGF-R expression (n = 70; total score, >5). The
distribution of clinical and histopathological characteristics in the 2
groups is shown in Table 2. Papillary carcinomas in the
2 groups of patients of similar gender and age did not differ
significantly in terms of tumor diameter, histological grade, presence
of multiple or bilateral tumor foci, extrathyroid invasion, or
metastatic lymph nodes. Papillary carcinomas with negative/low
Met/HGF-R expression were, however, associated with vascular invasion
(18.0% vs. 4.3%; P = 0.0308) and distant
metastases at tumor presentation (10.0% vs. 0%;
P = 0.0251). The follow-up periods were similar in the
2 groups (67.3 \ 25.1 vs. 73.3 \ 21.7 months),
but the patient outcome was different. Tumor relapses at distant sites
were more frequent in the group with negative/low Met/HGF-R than in the
group with high Met/HGF-R (P = 0.0248; Table 3). The cumulative frequency of both lymph node
metastases (88% vs. 55.7%) and distant metastases (28.0%
vs. 2.2%) was significantly higher in the group with
negative/low Met/HGF-R than in the group with high Met/HGF-R
(P = 0.0002; Table 3). Eighteen of 50 (36%) patients
who had died of cancer or had persistent disease at the last control
evaluation were in the negative/low Met/HGF-R group vs. 7 of
70 (10%) in the high Met/HGF-R group (P = 0.0012;
Table 3).
|
|
Univariate analysis.
As Met/HGF-R was absent or only focally
expressed in follicular carcinomas, only papillary carcinomas were
analyzed. By univariate Cox analysis, the risk of developing distant
metastases was higher in patients with tumors with negative/low
Met/HGF-R than in those with high Met/HGF-R (P =
0.0011). Vascular invasion (P = 
0.0000), age over 45
yr (P = 0.0140), extrathyroidal extension
(P = 0.0073), and tumor diameter more than 4 cm
(P = 0.01352) were also associated with an increased
risk of distant metastases. Hazard ratios and 95% confidence intervals
are given in Table 4. Other variables, including sex,
capsule infiltration, multifocality, bilaterality, and presence of
metastatic lymph nodes, were not significantly related to the
development of distant metastases. Kaplan-Meier plots demonstrating the
cumulative risk of distant metastases in papillary carcinomas with
either negative/low or high Met/HGF-R expression are shown in Fig. 3.
|
|
45 vs. >45 yr;
P = 0.0099), and vascular invasion (P =
0.0358). Met/HGF-R expression was the most effective predictor (hazard
ratio = 9.71). Hazard ratios and 95% confidence intervals are
given in Table 5.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In our hands, Met/HGF-R staining evaluation was reproducible and highly correlated to Western blot analysis. However, given the inherent difficulties of quantitative immunostaining, the method needs adequate standardization in the various laboratories. As Met/HGF-R expression can be measured not only in tissue sections but also in cytological aspirates, this measurement may become a useful parameter for planning patient management at both thyroidectomy and postsurgical follow-up.
Although several abnormalities of receptor tyrosine kinases have been recently described in papillary thyroid carcinomas, this is the first report suggesting that abnormal expression of a tyrosine kinase receptor may have an impact on clinical practice. In concert with our present findings, others have reported that negative/focal Met/HGF-R expression is related to a significantly shorter survival in pancreatic cancer (29). Furthermore, loss of heterozygosity on chromosome 7q, affecting the Met/HGF-R gene locus, is an independent predictor of poor survival in human breast cancer (30).
The biological basis of the association between negative/low Met/HGF-R expression and distant metastases in papillary thyroid cancer is not addressed by the present descriptive study. However, this association is not surprising in view of the pleiotropic effects of HGF/SF, the Met/HGF-R ligand. HGF/SF stimulates cell proliferation, cell motility, invasion of reconstituted basement membranes, and morphogenetic processes of varying complexity, such as tubulogenesis and angiogenesis (31, 32, 33). In some cancer cell lines, provided they are differentiated, HGF induces a morphogenetic response that follows a tissue-specific and genetically determined program (34). It is possible, therefore, that well differentiated thyroid cancer cells with high expression of Met/HGF-R respond to HGF/SF present in serum or produced locally (35, 36) mainly with changes in cell polarization resulting in morphogenetic epithelial-mesenchymal interactions, as observed during the ontogeny of parenchymal organs and epithelia regeneration (31). Papillae formation may be related to these morphogenetic changes, as suggested by the negative/low Met/HGF-R expression in normal thyroid tissue, benign thyroid tumors, and follicular carcinomas.
As Met/HGF-R is expressed at high levels in nearly 70% of papillary microcarcinomas, it is an early event in the multistep process of papillary cancer development, suggesting a selective advantage in favor of cells overexpressing this membrane tyrosine kinase. As described for estrogen or insulin-like growth factor I receptors in breast cancer, loss of Met/HGF-R may indicate cancer evolution to growth factor independence (37, 38).
In conclusion, our study indicates that 1) overexpression of Met/HGF-R is an early phenomenon in thyroid cancerogenesis and is restricted to the papillary histotype; and 2) in papillary thyroid cancer, negative/low Met/HGF-R expression is the most important predictor of hematogenous metastases development.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received January 27, 1997.
Revised April 9, 1997.
Accepted April 17, 1997.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Fujarewicz, M. Jarzab, M. Eszlinger, K. Krohn, R. Paschke, M. Oczko-Wojciechowska, M. Wiench, A. Kukulska, B. Jarzab, and A. Swierniak A multi-gene approach to differentiate papillary thyroid carcinoma from benign lesions: gene selection using support vector machines with bootstrapping Endocr. Relat. Cancer, September 1, 2007; 14(3): 809 - 826. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Ghoneim, M. Soula-Rothhut, C. Blanchevoye, L. Martiny, F. Antonicelli, and B. Rothhut Activating Transcription Factor-1-mediated Hepatocyte Growth Factor-induced Down-regulation of Thrombospondin-1 Expression Leads to Thyroid Cancer Cell Invasion J. Biol. Chem., May 25, 2007; 282(21): 15490 - 15497. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Bardella, D. Dettori, M. Olivero, N. Coltella, M. Mazzone, and M. F. Di Renzo The Therapeutic Potential of Hepatocyte Growth Factor to Sensitize Ovarian Cancer Cells to Cisplatin and Paclitaxel In vivo Clin. Cancer Res., April 1, 2007; 13(7): 2191 - 2198. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. L. Griffith, A. Melck, S. J.M. Jones, and S. M. Wiseman Meta-Analysis and Meta-Review of Thyroid Cancer Gene Expression Profiling Studies Identifies Important Diagnostic Biomarkers J. Clin. Oncol., November 1, 2006; 24(31): 5043 - 5051. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Mineo, A. Costantino, F. Frasca, L. Sciacca, S. Russo, R. Vigneri, and A. Belfiore Activation of the Hepatocyte Growth Factor (HGF)-Met System in Papillary Thyroid Cancer: Biological Effects of HGF in Thyroid Cancer Cells Depend on Met Expression Levels Endocrinology, September 1, 2004; 145(9): 4355 - 4365. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Pellegriti, C. Scollo, G. Lumera, C. Regalbuto, R. Vigneri, and A. Belfiore Clinical Behavior and Outcome of Papillary Thyroid Cancers Smaller than 1.5 cm in Diameter: Study of 299 Cases J. Clin. Endocrinol. Metab., August 1, 2004; 89(8): 3713 - 3720. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. E. Turner, A. L. Harris, S. Melmed, and J. A. H. Wass Angiogenesis in Endocrine Tumors Endocr. Rev., October 1, 2003; 24(5): 600 - 632. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Vella, G. Pandini, L. Sciacca, R. Mineo, R. Vigneri, V. Pezzino, and A. Belfiore A Novel Autocrine Loop Involving IGF-II and the Insulin Receptor Isoform-A Stimulates Growth of Thyroid Cancer J. Clin. Endocrinol. Metab., January 1, 2002; 87(1): 245 - 254. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V Vella, L Sciacca, G Pandini, R Mineo, S Squatrito, R Vigneri, and A Belfiore The IGF system in thyroid cancer: new concepts Mol. Pathol., June 1, 2001; 54(3): 121 - 124. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. E. Gardner, R. M. Tuttle, K. D. Burman, S. Haddady, C. Truman, Y. H. Sparling, L. Wartofsky, R. B. Sessions, and M. D. Ringel Prognostic Importance of Vascular Invasion in Papillary Thyroid Carcinoma Arch Otolaryngol Head Neck Surg, March 1, 2000; 126(3): 309 - 312. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Loss of Heterozygosity of the Long Arm of Chromosome 7 in Follicular and Anaplastic Thyroid Cancer, but Not in Papillary Thyroid Cancer J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3235 - 3240. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Pellegriti, A. Belfiore, D. Giuffrida, L. Lupo, and R. Vigneri Outcome of Differentiated Thyroid Cancer in Graves' Patients J. Clin. Endocrinol. Metab., August 1, 1998; 83(8): 2805 - 2809. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
