Primary Evaluation of Appling HMGA2 in the Diagnosis of Thyroid Carcinoma


 Background

Molecular marker is on the hot spot for thyroid tumor research in the recent years. The purpose of this study is try to clarify the value of HMGA2 for differential diagnosis between benign and malignant thyroid nodules in order to further assessing the risk of simultaneous cervical lymph nodes metastasis.
Methods

With a total of 125 thyroid samples, including 85 papillary thyroid carcinomas, 20 follicular thyroid adenomas and 20 normal thyroid tissue were retrospectively analyzed, and we compare expression level of HMGA2 protein among these three groups in this study. PTC group was then subdivided according to the number of metastasis lymph nodes in order to explore the possibility of using HMGA2 protein to predict invasion of lymph nodes in PTC. We also compare the efficiency of BRAFV600E mutation and HMGA2 expression level in the prediagnosis of cervical lymph node metastasis during the experiment.
Results

Expression rate of HMGA2 protein in PTC was measured at 100%, which has significant statistical differences with HMGA2 in FTA and NT (P < 0.05). Moreover, there was no significant correlation between expression level of HMGA2 and the size of PTC; whether it accompanied by Hashimoto's thyroiditis; high expression level of HMGA2 could be more sensitive in the prediagnosis of cervical lymph node metastasis than BRAFV600E mutation.
Conclusion

①HMGA2 protein can be used as a molecular marker for differential diagnosis of benign and malignant thyroid nodules. ②The expression level of HMGA2 protein is correlated with cervical lymph node metastasis in papillary thyroid carcinoma, performing an auxiliary value for clinical decision.


Abstract
Background Molecular marker is on the hot spot for thyroid tumor research in the recent years. The purpose of this study is try to clarify the value of HMGA2 for differential diagnosis between benign and malignant thyroid nodules in order to further assessing the risk of simultaneous cervical lymph nodes metastasis.

Methods
With a total of 125 thyroid samples, including 85 papillary thyroid carcinomas, 20 follicular thyroid adenomas and 20 normal thyroid tissue were retrospectively analyzed, and we compare expression level of HMGA2 protein among these three groups in this study. PTC group was then subdivided according to the number of metastasis lymph nodes in order to explore the possibility of using HMGA2 protein to predict invasion of lymph nodes in PTC. We also compare the e ciency of BRAF V600E mutation and HMGA2 expression level in the prediagnosis of cervical lymph node metastasis during the experiment.

Results
Expression rate of HMGA2 protein in PTC was measured at 100%, which has signi cant statistical differences with HMGA2 in FTA and NT (P < 0.05). Moreover, there was no signi cant correlation between expression level of HMGA2 and the size of PTC; whether it accompanied by Hashimoto's thyroiditis; high expression level of HMGA2 could be more sensitive in the prediagnosis of cervical lymph node metastasis than BRAF V600E mutation.

Conclusion
HMGA2 protein can be used as a molecular marker for differential diagnosis of benign and malignant thyroid nodules. The expression level of HMGA2 protein is correlated with cervical lymph node metastasis in papillary thyroid carcinoma, performing an auxiliary value for clinical decision.

Background
Thyroid carcinoma, most common endocrine malignancies, is also the fastest growing malignant tumor in the world [1,2]. High-resolution ultrasound which considered as the most effective method for detecting diagnosing thyroid carcinoma [3]. Studies have shown about 19 to 67 percent of people were diagnosed with thyroid nodules by high-frequency ultrasound among 7 to 15 percent was malignant [4,5]. Currently, percutaneous ne needle aspiration (FNA) cytology, considered as the best practice for the management of thyroid nodules. However combining both two approaches together, they still fail to provide de nitive malignancy con rmation of a thyroid nodule in many cases [6].
Operation is the rst choice for thyroid carcinoma; however, due to limited preoperative diagnosis, some patients with cervical lymph nodes metastasis are being miss diagnosed. This is also the reason that the treatment of papillary thyroid micro-carcinoma(PTMC) has always been controversial. Some doctors believe patients with PTMC can be monitored on a regular basis; others point out that although some of the PTMC can reach as small as 5 mm, it could still occur with imperceptible lymph nodes metastasis, requiring early clinical intervention [7]. Therefore, preoperative evaluation posts great importance to both patients and surgeons.
Accurate molecular markers for preoperative diagnosis of thyroid carcinomas could provide patients with reasonable treatment options; high mobility group protein AT-hook 2 (HMGA2) gene plays an important role of fetal development and tumorigenesis. And overexpression of HMGA2 has been reported as a feature of many malignant tumors of epithelial origin, it also associated with poor prognosis, such as chemoresistance, metastases and low long-survival rate[8-10].
Herein, we choose HMGA2 to be the marker to evaluate potential of its expression level in the diagnosis and prognosis of thyroid nodules.

Subjects
Patients enrolled with thyroid nodules admitted to our center from August 2016 to August 2017. Total of 125 samples were randomly selected, 85 cases of papillary thyroid carcinoma(PTC), 20 cases of follicular thyroid adenoma(FTA) and 20 cases of normal thyroid tissue(NT). Enrollment criteria: (1) con rmed pathological diagnosis; (2) thyroid ultrasound examination and surgery were performed in our hospital and all the information were complete; (3) One patient corresponds to one pathological type. This study receive approval from the Ethical Committee of Shanghai Jiao Tong University A liated Sixth People's Hospital. All procedures performed in the study involving human participants were in accordance with the ethical standards of the institutional research committee and the declaration of Helsinki [11].

Intraoperative tissue acquisition
All samples for this study, including PTC, FTA and NT, were obtained intraoperatively, tissues of PTC/FTA were taken from the center part of the thyroid nodules. NT was taken from the part with a minimum of 1 cm away from the margin of tumor. Extreme cases such as tumor can occupy most of gland, forcing NT taken from the opposite side. After all tissues were put in the sterile tubes, then moved to -80℃ immediately for future use; sample tissue size taken from the experiment measured across 4 to 5 mm.

Immunohistochemical analysis of clinical samples
Process start with defrosting the tissue samples under room temperature, then adding formalin and setting for 24 hours for preparing the para n-embedded tissue. Six-µm-thick para n sections were serially cut from para n-embedded tissues; sections obtained mounted onto slides; and then depara nized. All slides were rinsed in PBS for three 5-minute cycles, and then methanol containing 0.3% hydrogen peroxide was added to the slides. After incubating for 20 min at 37℃, slides were rinsed in distilled water for another three 5-minute cycles.
For immunohistochemical detection of HMGA2, slides were incubated in a humid chamber at 37℃ with 1:100 dilution of rabbit anti-human HMGA2 polyclonal (Protein Group, Inc., Wuhan, China) antibody. Bound antibodies (Protein Group, Inc., Wuhan, China) were added subsequently follow by incubation in a humid chamber at 37℃ for 30 min.
We used diaminobenzidine as the chromogen in this experiment. After the expected stain intensity developed, we rinsed slides thoroughly by running water to stop color rendering; sections were lightly counterstained with hematoxylin in the experiment. Given negative controls, immunostaining was performed by incubating samples with PBS instead of the primary antibody; remaining steps can be duplicated from previous procedures. Data are presented with mean ± SD and analyzed with SPSS 24.0, T-test and ANOVA was used as appropriate. If there is no special explanation, p < 0.05 was considered statistically signi cant.

Population characteristics
Total of 125 thyroid tissues, including 85 PTC, 20 FTA and 20 NT, were analyzed in this study, among 125 patients, 58.4%(73/125) were females and 41.5%(52/125) were males; the average age of the patients was 44 years old (range from 15 to 80 yrs.

Expression of HMGA2 protein in different thyroid tissue
All tissues were con rmed by two pathologists respectively after H&E stained. 15(17.6%) PTC from 85 cases was selected randomly to compare the differential expression of HMGA2 protein with 20 FTA and 20 NT by IHC. Through the preliminary observation, result showed HMGA2 was negative in NT; optical microscope showed the nucleus and cytoplasm were not signi cantly stained (Fig. 1A). HMGA2 in FTA was weak positive; optical microscope showed the nucleus was light brown stained and the cytoplasm was not stained (Fig. 1B). HMGA2 in PTC was strongly positive; optical microscope showed the nucleus was deep brown stained, the cytoplasm was brownish yellow stained (Fig. 1C).
Then we compared the expression level of HMGA2 protein in PTC with or without HT, result showed there was no signi cant difference between the two groups(p = 0.564). Statistical results also showed that the expression of HMGA2 was independent of patients' gender(p 0.05).
Furthermore, we analyze the correlation between expression of HMGA2 and cervical lymph nodes metastasis in PTC patients and found out AOD of HMGA2 in patients with or without cervical lymph nodes metastasis was 0.52240 ± 0.00563(95% CI: 0.51112, 0.53368), 0.48582 ± 0.00649, (95% CI: 0.47242, 0.49923), respectively. Result showed there was statistical difference between above two groups (p = 0.000) (Fig. 2C); the area under the ROC curve (AUC) measured 0.737(p < 0.05), indicated the expression level of HMGA2 was statistically signi cant in determining whether PTC has cervical lymph node metastasis. Cutoff value of AOD was 0.51 (Fig. 3).
Comparison of HMGA2 expression with BRAF V600E mutation to judge the diagnostic e cacy of cervical lymph node metastasis in PTC Our study concluded 56.57% (48/85) PTC accompanied by BRAF V600E mutation. 38 out of 60 patients who suffered cervical lymph node metastasis had BRAF V600E mutation; 10 out of 25 patients without cervical lymph node metastasis had BRAF V600E mutation. Sensitivity of BRAF V600E mutation to indicate cervical lymph nodes was 79.17%, speci city was 40.54%, positive prediction was 63.33%, negative prediction was 60%, false positive rate was 59.46%, and false negative rate was 20.83%. We obtained Cutoff AOD =0.51 in our earlier study, so we hypothesized that AOD > 0.51 indicated cervical lymph node metastasis, while AOD < 0.51 indicated no lymph node metastasis. The sensitivity result came 89.47%, speci city was 44.68%, positive prediction was 56.67%, negative prediction was 84%, false positive rate was 55.32%, false negative rate was 10.53% (Table 2 and Table 3).  SEN-sensitivity,SPE-speci city PPV-positive prediction NPV-negative prediction FPR-false positive rate FNR-false negative rate.

Discussion
Most PTC are inert growth with good prognosis afterwards, but the probability of cervical lymph node metastasis, which was not detected by ultrasound before surgery, is still existed with the rate of 20%-30% [12,13]. Wang et al. con rmed that age (< 55 yrs.), male, tumor size (0.5-1.0 cm), multifocal are risk factors for predicting cervical lymph node metastasis in the central region of neck in PTMC patients [14]. However, lymph nodes metastasis in central region still can be easily missed during ultrasound scanning. Previous studies have found the speci city of preoperative ultrasound detection of cervical lymph node metastasis in PTC patients was 87%~97%, but the sensitivity measure was only 23%~38% [15]. Since the involvement of lymph nodes greatly affects the clinical treatment strategy, it is helpful to assess the risk of lymph node metastasis in patients by combining preoperative ultrasound and molecular biological indicators.
High mobility family protein A2 (HMGA2), previously known as HMGI-C, was rstly discovered by Giancotti et al as a nuclear protein associated with malignant phenotype of rat thyroid cells transformed with murine retroviruses in 1985 [16]. While the expression of HMGA2 could be tested in several benign mesenchymal tumors, most predominantly lipomas, was mainly found in malignant tumors; HMGA2 would be present at a much higher level in malignancy, such as prostate cancer, ovarian cancer, gastric cancer, gallbladder cancer, breast cancer, and etc. [8,9,[17][18][19]. Previous studies have demonstrated that malignant tumors expressing HMGA2 usually come up with a poor prognosis [10,[20][21][22][23][24], thus, HMGA2 possess the potential for diagnosis of thyroid malignancies.
After experiment, we con rmed that the expression level of HMGA2 in PTC is signi cantly higher than FTA and NT cases (85/85, 100%). In addition, we found that a small amount of HMGA2 protein expression in NT(AOD = 0.08084 ± 0.00472), of which the expression level could be negligible. We speculate that the reason may be associated with microcirculation, since the normal thyroid tissue we obtained essentially from the area 1 cm away from the nodules in the ipsilateral thyroid. We analyzed the correlation between PTC and expression of HMGA2 protein and found that HMGA2 expression did not show signi cantly correlation with neither the size of PTC nor the gender of patients, and there was no statistical difference for the expression of HMGA2 between the PTMC and non-PTMC. We also proved that Hashimoto's thyroiditis had no impact on the expression level of HMGA2 protein in PTC, which has not been reported in previous studies.
We found there is statistical difference between the two groups by comparing the expression of HMGA2

Conclusion
In summary, this study demonstrates the differential diagnostic value of HMGA2 protein expression for thyroid nodules, and we con rm that HMGA2 protein is associated with cervical lymph node metastasis in PTC. Therefore, we could assess the risk of cervical lymph node metastasis based on the expression level of HMGA2. Thus, HMGA2 could to be a new molecular marker for the diagnosis and prognosis of hypoechoic lesion with obscure boundary and calci cation. IHC SP×400 strongly positive, the nucleus was deep brown stained, the cytoplasm was brownish yellow stained.

Figure 2
A. Comparison of AOD among Normal group, FTA group and PTC group. There were statistical differences among the three groups by pairwise comparison (P<0.01). B. There was no statistical difference in the expression level of HMGA2 in PTMC and non-PTMC (P>0.05). C. The expression of HMGA2 in PTC with or without cervical lymph nodes metastasis. There was statistical difference between two groups, p(N0 vs N) 0.05. N0-no evidence of regional lymph nodes metastasis; N-cervical lymph nodes metastasis.