| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Pathology, University Health Network (M.L.C.K., S.L.A.); Departments of Otolaryngology (M.L.C.K., N.J.P.B., J.L.F.) and Medicine (Endocrinology) (S.E.), Mount Sinai Hospital; and Departments of Laboratory Medicine and Pathobiology (S.L.A.), Otolaryngology (J.L.F.), and Medicine (S.E.), University of Toronto, Toronto, Ontario, Canada M5G 2M9
Address all correspondence and requests for reprints to: Sylvia L. Asa, M.D., Ph.D., Department of Pathology, University Health Network, 610 University Avenue, Suite 4-302, Toronto, Ontario M5G 2 M9 Canada. E-mail: . sylvia.asa{at}uhn.on.ca
Abstract
Lymph node metastasis in papillary thyroid carcinoma increases the morbidity of treatment and the risk of local regional relapse and may also affect cure rates and survival. Factors that predict lymph node metastasis are, however, unclear. We analyzed 125 patients with papillary thyroid carcinoma for factors that predict lymph node metastasis. On univariate analysis, age, extrathyroidal extension, tumor focality, overexpression of cyclin D1, and underexpression of p27 predicted lymph node metastasis, whereas patient gender and tumor size did not. On multivariate analysis, extrathyroidal extension, overexpression of cyclin D1, and underexpression of p27 proved to be strong independent predictors of lymph node metastasis. We suggest that immunohistochemistry for cyclin D1 and p27 will prove valuable in identifying papillary thyroid carcinomas with metastatic potential.
PAPILLARY THYROID CARCINOMA has a strong propensity to metastasize to regional lymph nodes. Lymph node involvement generally begins with the lymph nodes of the central compartment of the neck, followed by the nodes of the lateral neck and superior mediastinum. The primary mode of treatment for lymph node metastases is surgery, usually followed by thyroxine suppression and radioactive iodine therapy. The presence of lymph node metastases not only increases the morbidity of treatment, but also increases the risk of loco-regional relapse (1, 2). Although controversial, some reports in the literature also suggest that lymph node involvement in papillary thyroid carcinoma may compromise cure rates and survival (3, 4, 5).
Several clinical and histopathologic factors have been shown to be important prognosticators for survival in papillary thyroid carcinoma. These include patient age, patient gender, tumor size, extrathyroidal extension, and the presence of distant metastases (6, 7, 8). However, with the exception of extrathyroidal extension, it is unclear whether these prognostic factors predict lymph node metastases (9).
The cyclin D1 gene is a positive regulator of the cell cycle located on chromosome 11q23. It encodes a nuclear protein that forms complexes with cyclin-dependent kinases (cdk) 4 and 6, which phosphorylate and inactivate the retinoblastoma protein (pRb). Inactivation of pRb allows cell cycle progression from G1 to S phase (10). Although little is known about the role cyclin D1 plays in the pathogenesis of thyroid carcinoma, we know that normal thyroid follicular cells do not show immunoreactivity for cyclin D1 on immunohistochemistry, but cyclin D1 is overexpressed in up to 60% of papillary thyroid carcinomas (11, 12, 13, 14, 15).
p27 is a tumor suppressor gene located on chromosome 12p13. It encodes a nuclear protein that is a cdk inhibitor and inhibits the formation of cyclin D1/cdk complexes during G0 and early G1 phases of the cell cycle. This inhibits inactivation of pRb and prevents G1 to S phase transition. Normal thyroid follicular cells show strong immunoreactivity for p27 on immunohistochemistry, whereas p27 expression is reduced in both hyperplastic and neoplastic thyroid nodules. p27 levels have been reported to be lower in neoplastic nodules compared with hyperplastic nodules, in malignant tumors compared with benign tumors, and in poorly differentiated carcinomas compared with well differentiated carcinomas (16, 17, 18, 19, 20).
Overexpression of cyclin D1 and underexpression of p27 have been shown to predict lymph node metastases in several malignancies (21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32). However, their role in predicting lymph node metastases in thyroid carcinoma has not been investigated. We hypothesized that changes in expression of cyclin D1 and p27 might also predict lymph node metastases in papillary carcinomas of the thyroid.
Subjects and Methods
Subjects
This analysis included 125 patients with papillary thyroid carcinoma who underwent surgery at our institution between 1994–2000. All patients provided informed consent. Of these patients, 36 had evidence of gross lymph node metastases at the time of surgery and underwent total thyroidectomy and regional lymphadenectomy (central neck dissection, lateral neck dissection, superior mediastinal dissection, or a combination of the above). The remaining 89 patients had no evidence of lymph node involvement at the time of surgery and underwent total thyroidectomy alone. As is the policy at our institution, all 125 patients received postoperative radioactive iodine therapy followed by T4 suppression.
All of the clinical charts and histopathology reports were reviewed for data regarding patient age, gender, tumor size, tumor focality, presence or absence of extrathyroidal extension, and presence or absence of lymph node involvement. All 36 patients who underwent lymphadenectomy were confirmed to have gross (>2 cm) lymph node involvement on histopathological examination.
All 125 patients selected for analysis had typical papillary thyroid carcinomas on histopathology. Patients whose tumors showed tall cell or columnar cell differentiation or foci of insular or anaplastic dedifferentiation were not included. Patients with Hurthle cell predominant papillary thyroid carcinomas were also excluded.
Immunohistochemistry
Archival paraffin-embedded tumor tissue was analyzed by immunohistochemistry for cyclin D1 and p27 expression. Formalin-fixed paraffin-embedded sections (3 µm thick) were dewaxed in toluene and rehydrated through graded alcohols to water. Endogenous peroxidase activity was blocked in 3% hydrogen peroxide. Antigen retrieval was performed in 10 mM citrate buffer (pH 6.0) inside a microwave pressure cooker, and endogenous biotin detection was blocked with the Avidin-Biotin blocking kit (Vector Laboratories, Inc., Burlingame, CA). Primary antibody incubations were carried out at room temperature as follows: cyclin D1, mouse monoclonal (DCS-6) (DAKO Corp. A/S, Glostrup, Denmark), 1:50 dilution, overnight incubation; and p27/Kip1, mouse monoclonal (57) (BD PharMingen, Mississauga, Canada), 1:1000 dilution, 1 h incubation. Then, slides were washed in PBS, and secondary incubations were carried out with Biotin anti-mouse/anti-rabbit IgG followed by Streptavidin-HRP (Signet Pathology System, Dedham, Massachusetts) for 30 min. Immunoreactivity was revealed by incubation in 3-amino-9-ethylcarbazol. Slides were counterstained in hematoxylin and mounted with Crystal Mount. Positive controls for cyclin D1 included tumors with known cyclin D1 overexpression due to amplification; normal tissues were used for p27. Negative controls included normal tissue for cyclin D1 and cases of known down-regulation of p27 protein for p27. Nonspecific staining was evaluated by replacing primary antibody with normal mouse ascites.
Quantitation
Assessment of immunoreactivity was done jointly by two of the authors (M.L.K. and S.L.A.). Cyclin D1 expression was assessed on both the intensity of nuclear staining within the tumor cells and the percentage of cells that were positive. Cytoplasmic staining was not regarded as positive expression. Tumors were divided into two groups, namely expressers of cyclin D1 and nonexpressers of cyclin D1. In general, tumors that showed strong nuclear cyclin D1 expression also showed diffuse distribution of positive cells within the tumor (Fig. 1a
). These tumors were regarded as positive expressers of cyclin D1. Tumors that were negative expressers of cyclin D1 generally showed no cyclin D1 immunoreactivity at all (Fig. 1b). However, a small number of tumors showed very faint and very focal nuclear staining, the significance of which was unclear. As this pattern of staining was so different from the typical positive staining tumors, we elected to regard these tumors as negative expressers of cyclin D1.
|
); grade 2, moderate nuclear staining (easily appreciated) (Fig. 2b); grade 3, strong nuclear staining (but noticeably less than normal) (Fig. 2c); and grade 4, staining as strong as adjacent normal thyroid tissue. Underexpression was defined as a grade 0–1 staining. Quantitation of p27 immunoreactivity was based on the prevalent intensity of staining within the tumor. Most tumors displayed a uniform intensity of staining throughout the tumor, and in such cases the quantitation was straightforward. For the occasional tumor with heterogenous expression, we graded the p27 immunoreactivity according to the intensity displayed by the majority of tumor cells. Internal control within each slide was provided by adjacent normal thyroid or lymphoid tissue that invariably showed intense nuclear staining for p27 and served as a benchmark for assessment of intensity.
|
Seven parameters were assessed as predictors of lymph node metastases, namely age, gender, tumor size, tumor focality, extrathyroidal extension, cyclin D1 expression, and p27 expression. For the purposes of statistical analysis, patients were divided into two groups for each of the parameters, except for p27 expression in which the patients were divided into three groups. Univariate analysis was performed on all seven parameters using the 
2 test. Those parameters showing statistical significance on univariate analysis were then analyzed by multivariate analysis using cox regression. Statistical significance was ascribed at P < 0.05.
Results
At the time of surgery, the patients ranged in age from 17 to 85 yr (mean 44.8 yr, median 43 yr), with 66 (52.8%) patients less than 45 yr old and 59 (47.2%) patients 45 yr or older. We used the age of 45 yr as the cut off point between the two groups because this figure is widely used in clinical prognostic scores (6, 7, 8). There were 96 females and 29 males, giving a female to male ratio of 3.3:1.
For tumor size, patients were divided into two groups based on a size of 4 cm. Again, we chose this figure because it is widely used in clinical prognostic scores (6, 7, 8). One hundred three patients (82.4%) had tumors less than 4 cm in maximum dimension, whereas 22 patients (17.6%) had tumors greater or equal to 4 cm in size. After total thyroidectomy, histopathological analysis of the entire thyroid showed a single focus of tumor in 66 patients (52.8%). The remaining 59 patients (47.2%) were found to have not only a main tumor nodule, but between 1–7 separate smaller tumor nodules within either the ipsilateral or contralateral thyroid lobe. These patients were considered to have multifocal disease.
Gross extrathyroidal extension of tumor was noted in 25 patients (20%) either at the time of surgery or on histopathological examination. Isolated microscopic foci of extrathyroidal extension seen only on microscopic examination was not regarded as extrathyroidal involvement for the purposes of this analysis. The remaining 100 patients (80%) had tumors confined to the thyroid gland with no extrathyroidal uptake on postradioactive iodine scanning.
For cyclin D1 immunoreactivity, tumors were divided into expressers and nonexpressers as described above. There were 74 tumors (63.2%) that showed cyclin D1 immunoreactivity and 43 tumors (36.8%) that did not. Data were not available for eight tumors. For p27 immunoreactivity, owing to the very small number of tumors that showed either grade 0 or grade 4 immunostaining, we combined grades 0 and 1 and grades 3 and 4, giving 3 groups, namely grades 0–1, grade 2, and grade 3–4. There were 18 tumors (15.5%) with grade 0–1 expression, 68 tumors (58.6%) with grade 2 expression, and 30 tumors (25.9%) with grade 3–4 expression. Data were not available for nine tumors.
The breakdown of the 125 patients according to clinical and histopathologic parameters is summarized in Table 1.
|
The selected seven clinical and histopathological parameters were included in a univariate analysis for predictors of lymph node metastases in papillary thyroid carcinoma. The results are summarized in Table 1
. Five of the parameters tested proved to be significant predictors of lymph node involvement, namely patient age, extrathyroidal extension, tumor focality, cyclin D1 immunoreactivity, and p27 immunoreactivity. Patient gender and tumor size were not found to be predictors of lymph node metastases. Based on the 2 results, the presence of extrathyroidal extension (P < 0.001), underexpression of p27 (P < 0.001), tumor multifocality (P = 0.001), and overexpression of cyclin D1 (P = 0.002) were all very strong predictors of lymph node metastases.
Multivariate analysis
To identify independent predictors of lymph node metastases in papillary thyroid carcinoma, multivariate analysis using cox regression was performed on the five parameters shown to be significant on univariate analysis. These results are summarized in Table 2. After regression analysis, patient age lost its significance as a predictor of lymph node metastases, and tumor focality proved to be of only borderline significance (P = 0.04). However, extrathyroidal extension (P < 0.001), underexpression of p27 (P = 0.001), and overexpression of cyclin D1 (P = 0.011) retained strong independent predictive value for lymph node metastases.
|
Papillary thyroid carcinoma is the commonest endocrine malignancy, but it has a low mortality rate. Distinction of aggressive disease that bears significant morbidity and mortality requires definition. Several clinical and histopathologic parameters have been shown to be important predictors of prognosis and survival in well differentiated thyroid carcinoma (6). Although controversial, some reports also implicate lymph node metastases as a modifier of prognosis (3, 4, 5). This suggests that the prediction of lymph node metastasis is important. It is unclear whether the same factors that influence survival also influence metastatic potential.
In selecting patients for this study, we sought to keep the patient population homogenous with respect to tumor histology. It is well known that even within the broad group of papillary thyroid carcinomas, tumors with certain histologic features can have markedly different behavior compared with the more typical lesions (33). We therefore excluded patients whose tumors showed either tall cell or columnar cell differentiation, or foci of poorly differentiated, insular, or anaplastic dedifferentiation. Hurthle cell tumors that are variants of papillary thyroid carcinoma have recently been recognized (34). However, because little is known about their clinical behavior, Hurthle cell papillary thyroid carcinomas were also excluded from this analysis.
We included seven clinical and histopathological parameters in this assessment of predictors of lymph node metastases. Four of the factors, namely patient age, patient gender, tumor size, and extrathyroidal extension, have been shown to be important predictors for survival in well differentiated thyroid carcinoma and are widely used in prognostic scores. We hypothesized that these factors might also predict the metastatic potential of these tumors. Tumor focality was included in this analysis because factors that influence tumor focality might likewise influence lymph node metastases. Finally, we included tumor immunoreactivity to cyclin D1 and p27 based on the findings in other cancers.
After univariate analysis, age, extrathyroidal extension, tumor focality, cyclin D1 expression, and p27 expression were found to be significant predictors of lymph node metastases, whereas gender and tumor size were not. To identify independent predictors within this group of five, we analyzed these parameters by cox regression. After multivariate analysis, age lost its significance as a predictor, and tumor focality was only borderline significant (P = 0.04). However, extrathyroidal extension (P < 0.001), p27 expression (P = 0.001), and cyclin D1 expression (P = 0.011) retained strong independent predictive value for lymph node metastases.
The results of both the univariate and multivariate analyses are not surprising. With the exception of distant metastases, age and extrathyroidal extension have been found to be the most important predictors of survival in prognostic scores for papillary thyroid carcinoma (6, 7, 8). Although age did not retain independent predictive value for lymph node metastases, extrathyroidal extension remained the strongest independent predictor of lymph node metastases.
Our results with regard to tumor focality are intriguing. Tumor multifocality in papillary thyroid carcinoma has long been recognized, and it is debated whether this finding represents intraglandular spread from the index lesion, or de novo occurrence of separate carcinomas. Given the high incidence of asymptomatic papillary microcarcinomas in the general population (35) and the identification of discrete ret/PTC rearrangements in multifocal cancers (36), many people favor the latter explanation. However, if multifocality represents intraglandular spread from the index tumor, it might be associated with lymph node metastases. In all likelihood, the multifocality reported in the literature as well as in our study represents a combination of both groups (36). In our study, multifocality was a borderline predictor of lymph node metastases. This suggests that at least some of the multifocality identified by histopathology represents intraglandular metastases.
Our results with respect to cyclin D1 and p27 expression agree with data reported for several other malignancies (21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32). To our knowledge, this is the first report linking cyclin D1 and p27 to lymph node metastases in thyroid cancer. In our series, 63.2% of the papillary thyroid carcinomas expressed cyclin D1. This agrees with other reports in the literature (11, 12, 13, 14, 15). Our results show that papillary carcinomas that overexpress cyclin D1 have a significantly higher incidence of lymph node metastases compared with tumors that do not. The mechanism of cyclin D1 overexpression in thyroid is not known. Several studies have examined thyroid tumors with cyclin D1 overexpression but have found no evidence of gene amplification or rearrangement to explain this finding (12, 14, 37).
Most papillary thyroid carcinomas underexpress p27, compared with normal thyroid tissue; however, the degree of underexpression varies. In this series, 74.1% of the papillary thyroid carcinomas showed at least moderate underexpression of p27 (grade 0–2 staining) compared with normal thyroid tissue. Our results demonstrate that tumors which underexpress p27 also have a significantly higher incidence of lymph node metastases compared with tumors that do not. Again, the mechanisms underlying p27 down-regulation are not clear.
In conclusion, this study identifies three strong independent predictors of lymph node metastases in papillary thyroid carcinoma, namely extrathyroidal extension, underexpression of p27, and overexpression of cyclin D1. Although extrathyroidal extension has previously been identified as an important predictor of lymph node metastases, this is the first report linking overexpression of cyclin D1 and underexpression of p27 to lymph node metastases in thyroid cancer. Immunohistochemistry for cyclin D1 and p27 may prove to be valuable ancillary tests for identifying papillary carcinomas of the thyroid with metastatic potential.
Acknowledgments
We gratefully acknowledge the technical assistance of James Ho and Kelvin So.
Footnotes
This work was supported in part by Temmy Latner/Dynacare and by the Rita Banach Thyroid Cancer Research Fund.
Abbreviations: cdk, Cyclin-dependent kinase; pRb, retinoblastoma protein.
Received June 28, 2001.
Accepted November 19, 2001.
References
This article has been cited by other articles:
 |
|
 |
|
  S. Libertini, I. Iacuzzo, G. Perruolo, S. Scala, C. Ierano, R. Franco, G. Hallden, and G. Portella Bevacizumab Increases Viral Distribution in Human Anaplastic Thyroid Carcinoma Xenografts and Enhances the Effects of E1A-Defective Adenovirus dl922-947 Clin. Cancer Res., October 15, 2008; 14(20): 6505 - 6514. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Melck, H. Masoudi, O. L. Griffith, A. Rajput, G. Wilkins, S. Bugis, S. J. M. Jones, and S. M. Wiseman Cell Cycle Regulators Show Diagnostic and Prognostic Utility for Differentiated Thyroid Cancer. Ann. Surg. Oncol., December 1, 2007; 14(12): 3403 - 3411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Brzezinski, A. Migodzinski, A. Toczek, J. Tazbir, and M. Dedecjus Patterns of Cyclin E, Retinoblastoma Protein, and p21Cip1/WAF1 Immunostaining in the Oncogenesis of Papillary Thyroid Carcinoma Clin. Cancer Res., February 1, 2005; 11(3): 1037 - 1043. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. B. Dackiw, S. Ezzat, P. Huang, W. Liu, and S. L. Asa Vitamin D3 Administration Induces Nuclear p27 Accumulation, Restores Differentiation, and Reduces Tumor Burden in a Mouse Model of Metastatic Follicular Thyroid Cancer Endocrinology, December 1, 2004; 145(12): 5840 - 5846. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Chevillard, N. Ugolin, P. Vielh, K. Ory, C. Levalois, D. Elliott, G. L. Clayman, and A. K. El-Naggar Gene Expression Profiling of Differentiated Thyroid Neoplasms: Diagnostic and Clinical Implications Clin. Cancer Res., October 1, 2004; 10(19): 6586 - 6597. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. E. A. LaRue, M. Khalil, and J. P. Freyer Microenvironmental Regulation of Proliferation in Multicellular Spheroids Is Mediated through Differential Expression of Cyclin-Dependent Kinase Inhibitors Cancer Res., March 1, 2004; 64(5): 1621 - 1631. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V Vasko, M Saji, E Hardy, M Kruhlak, A Larin, V Savchenko, M Miyakawa, O Isozaki, H Murakami, T Tsushima, et al. Akt activation and localisation correlate with tumour invasion and oncogene expression in thyroid cancer J. Med. Genet., March 1, 2004; 41(3): 161 - 170. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Ying, H. Suzuki, H. Furumoto, R. Walker, P. Meltzer, M. C. Willingham, and S.-Y. Cheng Alterations in genomic profiles during tumor progression in a mouse model of follicular thyroid carcinoma Carcinogenesis, September 1, 2003; 24(9): 1467 - 1479. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Robbins Statins Sentence Thyroid Cancer Cells to Death Rho J. Clin. Endocrinol. Metab., July 1, 2003; 88(7): 3019 - 3020. [Full Text] [PDF]
|
|
|
|
 |
|
 G. Cheng, A. E. Lewis, and J. L. Meinkoth Ras Stimulates Aberrant Cell Cycle Progression and Apoptosis in Rat Thyroid Cells Mol. Endocrinol., March 1, 2003; 17(3): 450 - 459. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. C. Khoo, S. Ezzat, J. L. Freeman, and S. L. Asa Cyclin D1 Protein Expression Predicts Metastatic Behavior in Thyroid Papillary Microcarcinomas But Is Not Associated with Gene Amplification J. Clin. Endocrinol. Metab., April 1, 2002; 87(4): 1810 - 1813. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
