Patients with PTC who underwent thyroid excision surgery during the period 2011–2013 at The Second Affiliated Hospital of Harbin Medical University (Harbin, China) were recruited for the present study. The standard pathological diagnosis of PTC was based on the World Health Organization criteria (32) and two pathologists independently reviewed histological specimens in a blinded manner. A total of 271 patients with PTC (24 males and 247 females) were diagnosed and included in the current study. All patients met the inclusion criteria, which were: i) Underwent thyroid excision surgery; ii) confirmed as PTC by intraoperative rapid pathology and postoperative pathological detection; and iii) no history of thyroid disease and thyroid-related medication use. Informed consent was obtained from all participants and the study was performed according to the guidelines of the Ethics Committee of the Harbin Medical University.

Relevant patient data were collected, including demographic features (gender and age), clinical features (thyroid-stimulating hormone levels, tumor size and Hashimoto's disease) and pathological features (multifocality, T stage, extrathyroid invasion, intact capsule, prophylactic central compartment lymph-node (neck) dissection and node metastases). Tumor-node-metastasis (TNM) classification was performed according to the American Joint Committee on Cancer (33). T1 was defined as a tumor ≤2 cm in diameter and limited to the thyroid gland. T2 was defined as a tumor that was >2 and ≤4 cm, and limited to the thyroid gland. T3 was defined as a tumor >4 cm and limited to the thyroid gland, or any tumor with minimal glandular infiltration, including infiltration of the sternothyroid muscle or perithyroid soft tissues.

PTC tissue specimens were sectioned (4-µm), deparaffinized in xylene and rehydrated in a graded series of ethanol. The slides were then incubated in 3% hydrogen peroxide in distilled water at room temperature for 10 min to inactivate endogenous peroxidase activity, and subsequently underwent antigen retrieval in a sodium citrate solution in a microwave oven. The tissue sections were blocked with 5% bovine serum albumin (Boster Bio-Engineering, Wuhan, China) for 30 min at room temperature, and then incubated overnight at 4°C with a rabbit anti-human PEDF primary polyclonal antibody at 1:100 dilution (#BA1348-1; Boster Bio-Engineering) in a humidified chamber. The tissue sections were then incubated with a ready-to-use biotinylated horseradish-peroxidase-conjugated secondary antibody (#SA1022; Boster Bio-Engineering) and a streptavidin biotin complex (#SA1022; Boster Bio-Engineering) at 37°C for 30 min. The staining procedures were performed according to the manufacturer's instructions: Visualization with a 3,3′-diaminobenzidine solution and counterstaining with hematoxylin. The distribution and expression levels of PEDF were examined from the images obtained using the Olympus Imaging system (DP73: Olympus Corporation, Tokyo, Japan).

PEDF expression levels were semi quantitatively categorized into three groups, as follows: Negative (0 points), ≤5% positive cells; weakly positive (1 point), 6–30% positive cells; moderately positive (2 points), 31–60% positive cells.

The frozen tissues specimens obtained from PTC patients with LNM (n=15) and without LNM (n=10) were used for laser capture microdissection (ArcturusXT™; Thermo Fisher Scientific, Inc., Waltham, MA, USA) to obtain target thyroid epithelial cells, as described previously (34). The age and gender of the participants were matched for each group.

Total RNA was extracted from the microdissected cells using RNAiso Plus (9108; Takara Biotechnology Co., Ltd., Dalian, China). The extracted RNA was subsequently incubated with Recombinant DNase I (D2270; Takara Biotechnology Co., Ltd.) to erase the genomic DNA. cDNA was then obtained from mRNA using the PrimeScript™ RT reagent kit (DRR025A; Takara Biotechnology Co., Ltd.), which was then amplified using PCR on the ABI 7500 Real-Time PCR system (Thermo Fisher Scientific, Inc.) with SYBR Green I dye (DRR041S; Takara Biotechnology Co., Ltd.), in accordance with the manufacturer's protocol. The cycling conditions were as follows: 1 cycle as an initial denaturation at 95°C for 10 min; 40 cycles of 95°C for 5 sec, 60°C for 30 sec and 72°C for 15 sec; and a final extension step at 72°C for 5 min. The relative expression levels of PEDF, VEGF and HIF1α were determined using the comparative Cq method (35), following normalization to the endogenous GAPDH control gene. The primer sequences used in the RT-qPCR are presented in Table I.

All statistical analyses in the current study were performed with SPSS software, version 13.01S (SPSS, Inc., Chicago, IL, USA). An independent-sample t-test was used to compare the means between two groups, and a χ² or Fisher's exact test was used to compare frequencies between the groups. The data are presented as the mean ± standard deviation, or as percentages where appropriate. A Pearson correlation analysis was used to analyze the associations among the expression levels of PEDF, VEGF and HIF1α in PTC cells. All t-tests were two-tailed; P<0.05 was considered to indicate a statistically significant difference.

The clinicopathological characteristics of the patients are shown in Table II. Between 2011 and 2013, 271 patients with PTC, including 24 male and 247 female patients, were enrolled in the present study. The mean age of the patients was 43.1±10.6 years (age range, 19–73 years) and the mean size of the tumor was 10.9±7.8 mm (range, 2–50 mm). Multifocal tumors were identified in 12 patients; 14 patients (5.2%) had extrathyroid invasion and 86 patients (31.7%) had LNM. Hashimoto's disease was observed in 45 patients. A total of 245 patients had T1 tumors, 26 had T3 tumors and none had tumors in another stage. The BRAF^V600E mutation was present in 70.1% of patients. Immunohistochemistry with an anti-PEDF antibody detected PEDF expression in 74.5% of the PTC tissues. PEDF expression was determined to be negative in 69 patients, weakly positive in 131 patients and moderately positive in 71 patients.

In order to determine whether PEDF contributes to the progression of PTC, the associations between PEDF and aggressive clinicopathological features were investigated. As a result of the importance of metastasis to the progression of PTC, the association between PEDF expression levels and LNM was initially evaluated. As presented in Fig. 1 and Table III, PEDF expression levels were significantly decreased in PTC tissues with LNM, as compared with PTC tissues without LNM (P=0.006). Similarly, PEDF expression levels were significantly decreased in PTC tissues with extrathyroid invasion (P=0.035), a high TNM stage (P=0.013), BRAF^V600E mutation positivity (P<0.001) and a tumor size of >10 mm (P=0.018). However, PEDF expression levels were not significantly associated with multifocality (P=0.152; Table III).

To elucidate the potential mechanisms underlying the role of PEDF in metastasis by affecting angiogenesis, the mRNA expression profiles of PEDF, VEGF and HIF1α in stored patient tissue samples were determined using RT-qPCR. As indicated in Fig. 2, PEDF mRNA expression levels were significantly decreased in PTC tissues with LNM (0.2766±0.0910), as compared with PTC tissues without LNM (0.6251±0.2034; P<0.01). By contrast, the mRNA expression levels of HIF1α and VEGF were significantly increased in PTC tissues with LNM (1.6646±0.5533 and 3.9321±0.9235, respectively), as compared with PTC tissues without LNM (0.6847±0.2240 and 0.7537±0.1988, respectively; P<0.01). In addition, as shown in Table IV, through the analysis of the mRNA expression levels of PEDF, VEGF and HIF1α in the thyroid tissues specimens obtained from PTC patients with LNM (n=15) and without LNM (n=10), the present study identified a significant inverse correlation between PEDF and VEGF expression levels (r=−0.514; P=0.009), an inverse association between PEDF and HIF1α expression levels (r=−0.287; P=0.164) and a significant positive correlation between VEGF and HIF1α expression levels (r=0.489; P=0.013).