Abstract
Background: The expression of melanoma-associated antigen (MAGE) gene has been studied in many types of cancer. In the present study we evaluated the correlation between MAGE expression and the clinical features and oncologic outcomes of patients with papillary thyroid cancer (PTC). Materials and Methods: We performed a retrospective review of 85 patients who underwent surgery for PTC and analysis of their tumor tissue by nested reverse transcription-polymerase chain reaction (RT-PCR) with the MAGE common primer to detect the MAGE A1-6 gene. The associations between MAGE expression and clinical characteristics were analyzed. Results: Expression of MAGE A1-6 in PTC was identified in 31 patients (36.5%). Only papillary thyroid microcarcinoma (PTMC) was significantly related to MAGE expression in our univariate analysis (p=0.002) and multivariate analysis (p=0.006). MAGE had no significant impact on survival. Conclusion: Expression of MAGE A1-6 in PTC is significantly correlated with the presence of PTMC. Our study suggests that MAGE expression may be related to early-stage PTC.
- Papillary thyroid carcinoma
- melanoma-associated antigen (MAGE)
- reverse transcriptase-polymerase chain reaction
With the worldwide rising incidence of thyroid cancer (1), studies on molecular markers of genetic mutations in papillary thyroid carcinoma (PTC) are rapidly increasing (2, 3). Molecular markers are expected to have the potential to improve the diagnostic accuracy of pre-operative fine-needle aspiration cytology. Moreover, they are expected to provide prognostic information that could help decide between conservative and aggressive treatment. The decision for optimal treatment (extent of thyroidectomy, necessity or extent of neck dissection, indications for additional treatment) based on clinical characteristics of PTC remains a controversial issue. In this aspect, the expectations and need for reliable molecular markers of PTC seem to be obvious.
Given these developments, we sought to evaluate the significance of melanoma antigen (MAGE) gene as a molecular marker of PTC. MAGE is a tumor-specific antigen first reported in melanoma cells (4) which is specifically expressed in cancer cells. Various types of cancers, including head and neck (5), stomach (6), esophagus (7), colorectal (3), lung (8), breast (9), ovary (10) and liver (11), have been reported to express this gene. The MAGE gene is classified as MAGE A, B, C or D based on its chromosomal locus (12). Among them, the MAGE A family consists of several subtypes including MAGE A1 through MAGE A12. Park et al. (13) developed a common MAGE primer, which can detect all subtypes of MAGE A1-6 by a single examination of nested reverse transcription-polymerase chain reaction (RT-PCR). This strategy can detect at least one of the MAGE genes tested in tissues and has been verified in studies on lung (8), colorectal (3) and head and neck (14) cancer.
In this study, we applied this technique to detect expression of MAGE A1-6 in PTC. We investigated the incidence of MAGE A1-6 expression and analyzed the association between gene expression and clinical characteristics of PTC, including clinicopathological factors and oncological outcomes. There have been few studies investigating the expression and clinical significance of MAGE in thyroid carcinoma (15-17). Among them, only a few subtypes of the gene family were investigated, and none of them showed any correlation with oncological outcomes. To the best of our knowledge, this is the first study to evaluate the correlations between MAGE expression and PTC clinical characteristics with relatively long-term follow-up after treatment.
Materials and Methods
Patients. We conducted a retrospective review of 85 patients who underwent thyroidectomy as initial treatment for PTC from 2002 to 2007 at our center. These patients subsequently agreed to genetic analysis of their tumor tissue to identify MAGE expression. Patients who underwent completion thyroidectomy or neck dissection for recurrence in the cervical lymph nodes were excluded. The clinicopathological factors and oncological outcome of each patient, as well as types of treatment, including extent of thyroidectomy, neck dissection and radioiodine therapy, were reviewed. Isolation of mRNA was performed after pathological confirmation of PTC by frozen sections. In cases of multifocal lesions, the largest tumor was evaluated. The study was approved by the Institutional Review Board of our center (12-125).
RT-PCR and nested PCR. Fresh-frozen tissue blocks were cut into 4-μm sections using a cryostat microtome at −20°C. The first and last sections were immediately stained with methylene blue and were examined under microscopy to confirm that the tissue consisted of at least 80% tumor cells. Total RNA was isolated from each of the 85 samples after lysis in guanidinium isothiocyanate and phenolic extraction using a commercial kit (Trizol; Invitrogen Laboratories, San Diego, CA, USA). cDNA was synthesized from 4 μg of RNA using ImProm-II® enzyme (Promega, Madison, WI, USA) after evaluating the RNA purity and concentration by spectrophotometry (DU530; Beckman, CA, USA). Thirty cycles of nested PCR were conducted to amplify MAGE A1-6 using MAGE A1-6 common primers designed and described in detail by Park et al. (13). The amplified MAGE products were run for 30 min in a Mupid electrophoresis chamber (Advance, Tokyo, Japan) using 1.2% agarose gel. A MAGE-positive cell line (SNU 484) was used as positive control.
Statistical analysis. The survival rate was evaluated using Kaplan-Meier survival analysis. Univariate analysis to evaluate the correlations between MAGE expression and clinicopathological factors was conducted using Chi-square test or Fisher's exact test as appropriate. The relationships between gene expression and oncological outcomes were analyzed by the log-rank test. Multivariate analysis was conducted with a binary logistic regression model considering p<0.05 as significant. Statistical analysis was performed using the PASW 18 software program (SPSS Inc, Chicago, IL, USA).
Univariate analysis of clinicopathologic factors in relation to melanoma-associated antigen (MAGE) expression.
Electrophoretic analysis of melanoma-associated antigen (MAGE) A1-6 gene (490 bp) amplification from papillary thyroid carcinoma tissues. M, Size marker; N, negative control; P, positive control.
Multivariate analysis of clinicopathological factors in relation to melanoma-associated antigen (MAGE) expression.
Results
The mean patient age was 49.5±15.1 (range=22-79) years, and the cohort included 15 males and 70 females. Expression of MAGE in PTC was detected in 31 out of 85 patients (36.5%) (Figure 1). Table I summarizes the relationships between clinicopathological factors and MAGE expression. Age, sex, pT stage, pN stage, presence of lymph node metastasis, extrathyroidal extension, multifocality and variant pathology were not significantly related to MAGE expression. There were no differences in treatments for patients with and those without MAGE expression, including extent of thyroidectomy (lobectomy vs. total thyroidectomy), central neck dissection, lateral neck dissection or post-operative radioiodine therapy.
During the mean follow-up period of 59.9±33.5 (range=9-122) months, there were nine cases of recurrence (lateral neck recurrence: eight cases, both central and lateral neck recurrence: one case). There was no recurrence in patients with papillary thyroid microcarcinoma (PTMC). One patient with follicular variant who showed MAGE expression died with multiple lung metastasis 12 months after surgery. The incidence of MAGE expression was lower in patients with lymph node metastasis (27.8% vs. 40.8%), but the correlation was not statistically significant (p=0.214). The expression of MAGE was not significantly related to recurrence rate (5.6% vs. 14.3%, p=0.483), 7-year disease-free survival (88.3% vs. 83.0%, p=0.342), nor to 7-year disease-specific survival (96.6% vs. 100%, p=0.172). The only factor that was significantly related to MAGE expression was PTMC with tumor size ≤1 cm (p=0.002). In multivariate logistic regression analysis, only PTMC was significantly related to MAGE expression (p=0.006) (Table II).
Discussion
MAGE genes are considered to be potential tumor markers that can be used for cancer diagnosis and the development of an antigen-specific vaccine against cancer in various organs. Given that MAGE genes are expressed only in malignant tumor tissues and not in normal adult tissues (except for testicular germ cells and occasionally the placenta), an anti-cancer vaccine could be developed (3, 5-11). Since the first report (4) of MAGE expression in melanoma in 1991, these genes have been observed in various other types of cancer in many organs (3, 5-11). However, the role of the gene in carcinogenesis, specifically in PTC, has not been clearly identified.
Interaction between MAGE and p53 in thyroid cancer has been recognized to be negatively correlated by several authors (16, 18). In fact, the negative correlation has been also identified in cell lines for other types of cancer. In a study of a melanoma cell line, MAGE A2 induced a novel p53 inhibitory loop, forming a MAGE A2-p53 complex that interacted with epigenetic controllers, such as histone deacetylase-3 (19). The growth-promoting role of MAGE A2 through its interaction with the p53 pathway via an increase in cellular proliferation and decrease of cell-cycle arrest has been well-identified in head and neck squamous cell carcinoma by Glazer et al. (20). They also demonstrated that transfection of MAGE A2 induced a decrease in downstream p53 targets, including cyclin-dependent kinase inhibitor-1A (CDKN1A) and BCL-2–associated X protein (BAX). Moreover, knockdown of MAGE A2 reduced cell growth.
MAGE A3 was identified as a candidate mediator in fibronectin-mediated thyroid cancer progression. The interaction between MAGE A3 and fibroblast growth factor receptor related to the development of thyroid cancer has also been demonstrated (18, 21). Liu et al. showed that down-regulation of fibronectin, which is related to thyroid cancer cell proliferation, invasion, and even lung metastasis, led to significant increase in MAGE A3 levels (18). Additionally, they demonstrated that forced expression of MAGE A3 in human thyroid carcinoma-derived cells resulted in accelerated cell cycle progression with decreased G0-G1 residence and enhanced S-phase entry, accompanied by reduced p21 levels and enhanced Rb phosphorylation.
Currently, studies of MAGE expression in thyroid cancer tissue are rare and have shown variable results related to the incidence of gene expression and correlations to clinical characteristics. Lee et al. conducted RT-PCR in a limited number of cases including 6 tissue specimens of PTC and showed that expression of MAGE A-1, -2, -3 and -6 occured in a heterogeneous pattern (22). Melo et al. undertook immunohistochemical (IHC) analysis of MAGE A4 and MAGE C1 expression in thyroid cancer and benign thyroid tumors, which revealed no immunolabeling in either (17). Cheng et al. also analyzed tissues of thyroid cancer using IHC to localize MAGE and demonstrated that an increase in cytoplasmic MAGE score was correlated with the size of tumor and number of lymph node metastases, while normal thyroid tissue had a weak cytoplasmic MAGE expression (16). They concluded that the MAGE family plays a role in the progression of thyroid cancer. They also showed that the inverse relationship of MAGE and p53 was more pronounced in tumors without extrathyroidal extension and suggested that uncoupling of MAGE p53 may be related to disease progression. On the other hand, Milkovic et al. performed IHC analysis of MAGE 3 expression in thyroid cancer, showing an expression rate of 65% as a whole and 80% in PTC (15). Higher expression of MAGE 3 was identified in classical types of PTC and small PTC sized up to 1 cm in diameter. They concluded that MAGE 3 expression may be related to early steps of thyroid cancer progression. Although these two studies showed conflicting results regarding the correlation between MAGE expression and tumor size of PTC, both suggest the role of MAGE coupled with p53 in early stages of thyroid cancer. Although the expression of p53 was not evaluated in our study, our result showing a significantly increased expression of MAGE A1-6 in the PTMC group may support such a hypothesis. However, contrary to the findings of Cheng et al. (16), our results showed a decreased tendency for lymph node metastasis in patients with MAGE expression. Thus, our study suggests that expression of MAGE may not be related to lymphatic invasiveness of PTC. Nonetheless, different methods used to detect MAGE expression between the studies (IHC vs. RT-PCR; scoring system versus detection of expression; MAGE 3 vs. MAGE A1-6) should be considered.
The limitations of this study include its retrospective design with possible selection bias and the absence of gene analysis on normal thyroid tissue. Since the RT-PCR undertaken in this study was conducted with a common primer, specific detection of subtypes A1-6 of the MAGE family could not be performed. Studies to detect MAGE subtypes should be performed to identify whether there are distinct patterns of thyroid cancer progression for different subtypes. Since death due to PTC is relatively rare, the impact of MAGE expression on disease-specific survival requires further investigation with a larger number of patients and a longer follow-up period. Additionally, studies to identify the impact of MAGE A1-6 on the p53 pathway and the correlation between these factors on thyroid cancer progression should be performed.
PTC has a wide spectrum of clinical progression, which may be related to a pattern of cumulative genetic defects correlating with tumor differentiation, metastatic potential, and aggressiveness (23). In this aspect, MAGE may be involved in one of the genetic defects linked to PTC. The role of MAGE in carcinogenesis and progression of PTC should be evaluated in relation to other genetic defects and tumor markers.
In this study, MAGE expression was significantly related to PTMC compared to PTC larger than 1 cm. However, MAGE expression was not significantly correlated to other clinicopathological factors or oncological outcome of patients with PTC. Our study suggests that MAGE expression may be related to early stages of PTC.
Footnotes
-
Conflicts of Interest
None.
- Received January 24, 2013.
- Revision received March 5, 2013.
- Accepted March 7, 2013.
- Copyright© 2013 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
