Correlation between low expression of protein disulfide isomerase A3 and lymph node metastasis in papillary thyroid carcinoma and poor prognosis: a clinicopathological study of 1,139 cases with long-term follow-up

Abstract

The incidence of papillary thyroid carcinoma (PTC) is increasing worldwide. The biomarkers to identify aggressive types of PTC are limited, illustrating the need to establish reliable novel biomarkers. Protein disulfide isomerase A3 (PDIA3) is a chaperone protein that modulates the folding of newly synthesized glycoproteins and stress-responsive proteins in the endoplasmic reticulum. Although the role of PDIA3 in various cancers such as breast, uterine cervix, head and neck, and gastrointestinal tract has been examined, its expression in thyroid cancer has not been reported. We retrospectively reviewed accumulated data with long-term follow-up of 1,139 PTC patients, and investigated the correlation between immunohistochemical expression of PDIA3 in PTC patients and clinicopathological features and prognosis. PDIA3 expression was significantly lower in PTCs compared to normal thyroid tissues (NTT; n = 80, p = 0.002). In PTCs, correlation between low PDIA3 expression and lymph node metastasis (p = 0.018) and the number of positive nodes (p = 0.004) was observed. Patients with low PDIA3 expression exhibited worse cause-specific survival compared to those with high PDIA3 expression (p = 0.013). Our findings indicate that low PDIA3 expression is related to poor clinical outcome in PTC patients, and that PDIA3 may potentially be a novel ancillary biomarker. Further clarification of the biological role of PDIA3 in PTC is warranted for the future clinical application.

THYROID CANCER is one of the most common malignancies of the endocrine system and its incidence is rapidly increasing worldwide [1]. Among thyroid cancers, papillary thyroid carcinoma (PTC) is the most common histological type. Although PTC generally has an indolent and slow, progressive behavior, 5–10% of PTC patients experience a more aggressive clinical course, characterized by recurrent disease and early metastases, and resistance to radioactive iodine, resulting in increased mortality [2]. Aggressive clinicopathologic features of PTC such as large tumor size, multifocality, massive extrathyroidal extension, lymph vascular invasion, large or extranodal invasive lymph node metastasis (LNM), distant metastasis, histological subtypes, and TERT promoter mutation have been reported [3-6]. Despite these facts, biomarkers to identify aggressive types of PTC are limited, illustrating the need to establish reliable novel biomarkers.

Protein disulfide isomerase A3 (PDIA3), also known as ERp57/GRP58, is of the protein disulfide isomerase family. The functions of PDIA3 in the endoplasmic reticulum (ER), including the proper folding and quality control of glycoproteins as well as participation in the assembly of major histocompatibility complexes (MHC) class I are well known [7]. The roles of PDIA3 in tumors vary depending on the tumor type. To date, PDIA3 has been reported to be involved in the following tumors: breast [8], uterine cervical cancer [9], gastric [10], hepatocellular [11, 12], colorectal [13], lung [14], laryngeal carcinoma [15], and glioma [16]. In some of these tumors, high expression is correlated with poor prognosis [11, 12], while in others it is correlated with good prognosis [9, 10]. The involvement of PDIA3 in thyroid cancer is still unknown.

Here, we investigated the immunohistochemical expression of PDIA3 in 1,139 PTC patients with long-term follow-up (mean, 12 years) at a single institution in order to identify its correlation with other clinicopathological features and prognosis.

Materials and Methods

Case selection

For this retrospective study, patients with PTC who underwent primary thyroid surgery between January 1993 and December 2012 at Cancer Institute Hospital, Japanese Foundation for Cancer Research (JFCR), Tokyo, Japan were targeted. Patients with papillary microcarcinoma (maximal diameter, ≤1.0 cm) were excluded. We designed our own risk group classification for predicting cause-specific death from PTC [5]. That is, patients with distant metastasis and older patients (≥50 years) with massive extrathyroidal invasion or large lymph node metastasis (LNM) (maximal diameter, ≥3 cm) were defined as high risk and all other patients were classified as low risk. According to the risk adapted management treatment approach, we provided treatment options as described in previous reports [4]. Briefly, we basically conducted lobectomy for patients with unilateral low-risk PTC. For patients with high-risk PTC, total thyroidectomy followed by radioactive iodine treatment was recommended. Extent of lymph node dissection was determined as follows [17]: a) Dissection of the central compartment (level VI) alone for patients with LNM only in the central zone or with no LNM; and b) lateral neck dissection (basically, levels II, III, IV, and VI) when the patient was diagnosed with lateral neck LNM. Bilateral neck dissection was performed only when preoperative imaging showed bilateral neck LNM. Patients were observed until December 26, 2018, or death, whichever came first. Cases with small tumor components, and/or severe calcification, were excluded from the study. In addition, 80 cases of normal thyroid tissue (NTT) array samples were purchased from Funakoshi Co., Ltd. (Tokyo, Japan). Patient demographics and clinicopathological data were obtained from medical records. This study was conducted under the principles embodied in the Declaration of Helsinki (revised in Brazil 2013), and was approved by the institutional review boards of Nippon Medical School Hospital (#30-03-1111, June 6, 2020) and of the JFCR (IRB number 2013–1128, on January 24, 2014, last updated on March 23, 2021). Written informed consent for the use of clinical samples for research purposes was obtained from the patients.

Immunohistochemistry

Expression of PDIA3 was evaluated by immunohistochemistry of tissue microarrays (TMAs). We constructed TMA as previously described [4]. In brief, for each tumor, the experienced, boarded pathologists selected one-to several representative histological areas for tissue microarray. We punched the selected these points of the donor paraffin blocks with a 2 mm diameter coring needle, and transferred the tissue to the array in the recipient block using a manual tissue arrayer (KIN-1; Azumaya, Tokyo, Japan). Tissue sections with a 4 μm thickness were used for immunostaining. Following deparaffinization, sections were pretreated at 95℃ for 40 min in 10 mM citrate buffer solution (pH 6.0). Endogenous peroxidase was blocked in 100% methanol (Wako Pure Chemical Industries, Ltd.) containing 0.3% hydrogen peroxide (Wako Pure Chemical Industries, Ltd.) for 30 min at room temperature. The sections were then incubated with an anti-PDIA3 mouse monoclonal antibody (CL2444, Invitrogen/Thermo Fisher Scientific K.K., Tokyo, 1:10,000 dilution), using the Nichirei Histostainer (Nichirei Biosciences INC., Tokyo, Japan). The sections were further incubated with Histofine Simple Stain MAX PO (mouse; Nichirei Biosciences, Inc., Tokyo, Japan) for 30 min at room temperature, and peroxidase activity was visualized by 3,3' diaminobenzidine. The sections were then counterstained with hematoxylin at room temperature for 1 min.

Immunohistochemical analysis

Intensity and ratio of stained carcinoma cells were scored using a semi-quantitative method. The slides were scanned using Nano Zoomer 2.0-HT (Hamamatsu Photonics K.K., Tokyo, Japan) and analyzed using QuPath software v0.2.3 [18]. Cytoplasmic staining was considered to indicate a positive reaction. The expression of PDIA3 was determined using the H-score. H-score was determined by the sum of (each category of cytoplasmic intensity from 0 to 3) × (percentage of tumor cells of each category) (i.e. 1 × 50% + 2 × 50% = 150). The results ranged from 0 to 300, where 0 indicates all cells are negative, and 300 indicates all cells are strongly positive. The QuPath software analysis was performed twice, and the average score was calculated. Then, the average score for several cores from each case was determined as H-score of the case. The cases were then divided into high and low groups depending on whether they were above and below the median score, respectively.

Statistical analyses

All categorical variables are presented as numbers and percentages, while continuous variables are presented as mean ± standard deviation. H-scores in the immunohistochemical analysis were compared using Mann Whitney’s U test. Clinicopathological parameters were analyzed using the Mann-Whitney’s U test, and Chi-square test. Cause-specific survival (CSS) time was measured from the date of diagnosis to the date of cause-specific death or the date of the last follow-up. Disease-free survival (DFS) time was measured from the date of diagnosis to the date of the first relapse or the date of the last follow-up. Cox proportional hazards model was used to identify independent factors which significantly influenced survival. The probability of CSS and DFS were calculated using the Kaplan-Meier method and compared using the log-rank test. Statistically significant difference was set at p < 0.05. All statistical analyses were performed using SPSS (version 26; IBM Corp., Armonk, NY, USA).

Results

Clinicopathological characteristics of PTCs

Of the 1,435 patients in our archives, clinical data and tissue of 1,139 PTC patients were available for analysis. Baseline clinicopathological characteristics of PTC patients are summarized in Table 1. Targeted patients were comprised of 873 women and 266 men with a mean age of 52.3 ± 13.9 years. Mean tumor size was 24.8 ± 14.9 mm. According to the AMES (age, metastasis, extrathyroidal extension) [19] risk classification systems, 947 patients (83%) were classified as low-risk, and 192 patients (17%) as high-risk. When applying the Cancer Institute Hospital risk classification group [5], 944 patients (83%) were classified as high-risk, and 195 patients (17%) were classified at low risk. The mean follow-up period was 12 years (range 154 days–25 years). During the follow-up period, 957 patients (84%) were alive without recurrent cancer, 98 patients (9%) were alive with residual or recurrent cancer, and 84 patients (8%) died of recurrent cancer. Mean CSS time was 23.8 years (95% CI 23.5–24.3 years). DFS time was 21.8 years (95% CI 21.3–22.3).

Table 1 Baseline characteristics of the patients

Factors	Cases (n = 1,139)
Age
Mean ± S.D.	52.3 ± 13.9
Sex
Female	873 (77%)
Male	266 (23%)
Mean tumor size ± S.D. (mm) (range)	24.8 ± 14.9 (11–130)
Tumor size
≤40 mm	1,010 (89%)
>41 mm	129 (11%)
Histological subtypes
Conventional type	985 (87%)
Follicular variant	152 (13%)
Diffuse sclerosing variant	2 (0.2%)
Extrathyroidal extension (pathological)
Absent	530 (47%)
Present	609 (53%)
LNM (pathological)
Absent	643 (56%)
Present	496 (44%)
Maximum size of LNM
Mean ± S.D. mm (range)	27.6 ± 14.7 (9–105)
LNM size
≤30 mm	1,000 (88%)
>31 mm	139 (12%)
Number of LNM
Mean ± S.D. (range)	2.01 ± 4.2 (1–34)
Distant metastasis
Absent	1,067 (94%)
Present	72 (6%)
AMES risk classification
Low-risk	947 (83%)
High-risk	192 (17%)
CIH risk classification group [5]
Low-risk	944 (83%)
High-risk	195 (17%)

AMES, age, metastasis, extent and size; CIH, Cancer Institute Hospital; LNM, lymph node metastasis; S.D., standard deviation.

The expression of PDIA3 in NTT and PTC

Significant expression of PDIA3 was diffusely found in the cytoplasm of follicular epithelial cells in NTT, whereas its expression was relatively weak in the cytoplasm of tumor cells in PTC. The H-score of PDIA3 expression ranged between 0 and 300, and the median H-scores of PDIA3 in NTT and PTC were 148.6 and 131.1, respectively, indicative of a decreased PDIA3 expression in PTC (p = 0.002) (Fig. 1A). Representative immunohistochemical stainings of NTT and PTC are shown in Fig. 1B–D.

Fig. 1

Immunohistochemical stainings of PDIA3. H-scores between NTT and PTC were compared, and H-scores of PTC were significantly lower than those of NTT (p = 0.002) (A). In NTT, follicular epithelial cells exhibited diffuse positive staining for PDIA3 in the cell cytoplasm (B). In PTC, PDIA3 was variably expressed in the cytoplasm of the tumor cells. Representative findings of PTC with the PDIA3-low (C) and -high (D) expressions.

NTT, normal thyroid tissue; PTC, papillary thyroid carcinoma. Original magnification, ×400. Bars, 50 μm.

Correlation between PDIA3 expression in PTC and clinicopathological features

PTC cases were divided into PDIA3-low (570 cases) groups and -high (569 cases) depending on whether they were above and below the median score of 131.1. In order to assess the correlation between PDIA3 expression and clinicopathological features (Table 2). The PDIA3-low expression group showed a higher incidence of LNM (p = 0.018), and an increased number of positive lymph nodes (p = 0.004), compared to the PDIA3-high expression group. The LNM is categorized as central and lateral LNM. The PDIA3-low group showed a higher incidence of lateral LNM, compared to PDIA3-high group (211/268 cases (82%) vs. 171/228 (75%), p = 0.038). Nine patients showed anaplastic transformation during follow-up; but no significant correlation between PDIA3 expression and the frequency of anaplastic transformation was observed (p = 1.000).

Table 2 Correlation between PDIA3 expression and clinicopathological features of PTCs

Factors	PDIA3 expression		p value
	low n = 570	high n = 569
Age	52.7 ± 13.3	52 ± 14.6	0.609
Sex
Female	434 (76%)	439 (77%)	0.686
Male	136 (24%)	130 (23%)
Mean tumor size (mm) (range)	25.0 ± 15.5 (11–115)	24.6 ± 14.4 (11–130)	0.865
Tumor size
≤40 mm	505 (89%)	505 (89%)	0.934
>41 mm	65 (11%)	64 (11%)
Histological subtypes
Conventional type	489 (86%)	496 (87%)	0.778
Follicular variant	80 (14%)	72 (13%)
Diffuse sclerosing variant	1 (0.2%)	1 (0.2%)
Extrathyroidal extension (pathological)
Absent	260 (46%)	270 (47%)	0.534
Present	310 (54%)	299 (53%)
LNM (pathological)
Absent	302 (53%)	341 (60%)	0.018*
Present	268 (47%)	228 (40%)
Maximum size of LNM
Mean ± S.D. (mm) (range)	27.3 ± 14.3 (9–90)	27.9 ± 15.2 (10–105)	0.745
LNM size
≤30 mm	492 (86%)	508 (89%)	0.127
>31 mm	78 (14%)	61 (11%)
Number of LNM
Mean ± S.D. (mm) (range)	2.3 ± 4.4 (1–29)	1.8 ± 3.8 (1–34)	0.004*
Distant metastasis
Absent	532 (93%)	535 (94%)	0.632
Present	38 (7%)	34 (6%)

LNM, lymph node metastasis; PTC, papillary thyroid carcinoma; S.D., standard deviation.

Correlation between PDIA3 expression and clinical outcome

During the follow-up, 56 patients (10%) in the PDIA3-low group and 28 patients (5%) in the PDIA3-high group died of the disease. In the PDIA3-low group, seven patients died of the local lesions/recurrence, and 49 patients died of metastatic lesions. In the PDIA3-high group, eight patients died of local lesions/recurrence, and 20 patients died of metastatic lesions. Anaplastic transformation was detected in five and four of the dead PDIA3-low and -high groups, respectively.

For patients with no distant metastasis (n = 1067), Cox proportional hazard regression analysis of the correlation between clinicopathological features and CSS was performed. Hazard ratios for age, extrathyroidal extension, tumor size >41 mm, LNM >31 mm, and low PDIA3 expression significantly increased according to univariate analysis (Table 3). According to multivariate analysis, age, LNM >31 mm, extrathyroidal extension, and low PDIA3 expression were determined to be independent factors for CSS (Table 3). According to log-rank test, CSS of the PDIA3-low group was significantly worse than that of the PDIA3-high group (23.9 years, 95% CI 23.4–24.4 vs. 24.9 years, 95% CI 24.4–25.3 years, p = 0.013) (Fig. 2A).

Table 3 Cox proportional hazard regression analysis of the correlation between clinicopathological features and cause-specific survival

Factors	Univariate analysis		Multivariate analysis
	HR (95%CI)	p value	HR (95%CI)	p value
Age	1.105 (1.076–1.135)	<0.001	1.095 (1.065–1.126)	<0.001
Sex (male vs. female)	0.677 (0.509–0.902)	0.008	0.953 (0.710–1.279)	0.750
Tumor size (≤40 mm vs. >41 mm)	8.524 (4.959–14.652)	<0.001	5.571 (3.163–9.810)	<0.001
LNM (≤30 mm vs. >31 mm)	6.838 (3.937–11.877)	<0.001	3.866 (2.168–6.894)	<0.001
Extrathyroidal extension	12.972 (4.679–35.962)	<0.001	4.918 (1.702–14.211)	0.003
PDIA3 (low vs. high)	0.482 (0.268–0.868)	0.015	0.453 (0.247–0.833)	0.011

LNM, lymph node metastasis.

Fig. 2

Comparison of CSS (A) and DFS (B) according to PDIA3 expression.

(A) PDIA3-low group exhibited a shorter CSS than that for the PDIA3-high group.

(B) PDIA3 expression was not an independent factor of DFS. DFS did not significantly differ between the PDIA3-low and -high groups.

CSS, cause-specific survival; DFS, disease-free survival.

Mean DFS in the PDIA3-low and -high groups were 23.3 years and 20.1 years, respectively. According to univariate and multivariate analyses, age, tumor size >41 mm, LNM >31 mm and extrathyroidal extension were independent factors, but PDIA3 expression was not an independent factor (Table 4, Fig. 2B).

Table 4 Cox proportional hazard regression analysis of the correlation between clinicopathological features and disease-free survival

Factors	Univariate analysis		Multivariate analysis
	HR (95%CI)	p value	HR (95%CI)	p value
Age	1.052 (1.036–1.068)	<0.001	1.043 (1.028–1.059)	<0.001
Sex (male vs. female)	0.804 (0.653–0.990)	0.040	0.964 (0.780–1.191)	0.733
Tumor size (≤40 mm vs. >41 mm)	4.953 (3.318–7.395)	<0.001	3.440 (2.277–5.198)	<0.001
LNM (≤30 mm vs. >31 mm)	1.034 (1.026–1.043)	<0.001	3.106 (2.012–4.795)	<0.001
Extrathyroidal extension	5.606 (3.383–9.289)	<0.001	3.233 (1.907–5.480)	<0.001
PDIA3 (low vs. high)	0.766 (0.525–1.116)	0.165	0.824 (0.563–1.205)	0.317

LNM, lymph node metastasis.

Discussion

The accumulated data of 1,139 patients with PTC were retrospectively reviewed, and expression of PDIA3 assessed. In the current large-scale population with a long-term follow-up study, correlation between low expression of PDIA3 in PTC and incidence of LNM, the number of positive nodes, and poor CSS were observed. The role of PDIA3 is complex and disease dependent, and, to date, no study has investigated PDIA3 expression in thyroid cancer. This is the first report to demonstrate a correlation between a low expression of PDIA3 in PTC and its clinicopathological features.

PDIA3 is a member of the PDI family and was first described as a stress-responsive protein, which is up-regulated under the glucose depletion-induced cellular stress [20]. Unlike other members, PDIA3 is not only located in the ER lumen but also in many subcellular locations, as it does not contain a C-terminal ER retention motif [21]. PDIA3 is expressed ubiquitously at the systemic level, but is present in various types of tissues at varying degrees. In this study, PDIA3 was identified in the cytoplasm of normal thyroid follicular epithelial cells as well as in tumor cells of PTC. These results suggest that possible localizations within thyroid cells should be in the ER, mitochondria, and cytosol. In the ER, biological functions of PDIA3 are reported to be protein folding [22], stabilization of MHC class I [23], and calcium homeostasis [24]. PDIA3 modulates apoptotic signaling [25, 26] and calcium uptake in the mitochondria, actively modulates mTOR complexes [27] and promotes monocyte/macrophage differentiation in the cytosol [28].

PDIA3 has been reported to be related to tumor proliferation and prognosis, but the degree of expression and its functions greatly vary among carcinomas of different sites. Our research group previously reported a higher PDIA3 expression in hepatocellular carcinoma in comparison with normal liver tissue [11]. A higher PDIA3 expression has also been reported for laryngeal cancer [15] and non-small cell lung cancer [14]. On the other hand, expression of PDIA3 has been reported to be inversely correlated with the degree of differentiation in some types of tumors. In uterine cervical cancer, PDIA3 expression was down-regulated compared to cervical intraepithelial neoplasia or normal tissue [9]. In gastric cancer, PDIA3 expression was lower in poorly differentiated components than that in well-differentiated components and normal gastric epithelium [10]. In this study, PDIA3 expression was significantly decreased in PTC cells compared to NTT. The roles of PDIA3 in PTC may be due to its multifunction, but the underlying mechanism of a low PDIA3 expression in PTC still remains unsolved.

In this study, a low expression of PDIA3 was significantly correlated with poor CSS in PTC patients. This finding agrees with previous reports on early-stage gastric cancer [10, 29], cervical cancer [9], head and neck cancer [30], and non-small cell lung cancer prognosis [14]. While the underlying mechanism has not been fully elucidated, this may be due to dysfunction of the MHC class I complex, which is stabilized by PDIA3 under the normal conditions [9, 10]. In cancer immunity, PDIA3 forms a complex with MHC class I to facilitate antigen processing, thus activating an immune response against cancer cells [10]. In contrast, when MHC class I expression is insufficient, cancer cells can avoid the cytotoxic effects of immune cells as seen in gastric cancer [31]. Similarly, down-regulation of the MHC class I complex and subsequent acceleration of tumor progression and metastasis has been reported for various tumors, including thyroid cancer [29, 30, 32-34]. Experiments which can clarify the correlation between the molecular mechanism of PDIA3 and MHC class I in thyroid cancer are needed in the future.

The treatment approach including the range of lymph node dissection varies depending on traditional policies and the guidelines. Lobectomy has been widely performed in Japan. On the contrary, treatment teams in Western countries are still apt to choose total thyroidectomy and radioactive iodine therapy. In Japan, According to the Japanese guideline established by Japan Association of Endocrine Surgeons (JAES)/Japanese Society of Thyroid Surgery (JSTS) (the current Japan Association of Endocrine Surgery), lobectomy with prophylactic central lymph node dissection is recommended for low- and intermediate-risk PTC patients [35, 36]. The prophylactic lateral node dissection is only recommended for intermediate- and high-risk patients, with a comprehensive evaluation of risk factors, such as tumor size, distant metastasis and patients’ background and preference. In the guideline of the American Thyroid Association, therapeutic lymph node dissection is well accepted, but prophylactic lymph node dissection is not recommended [2]. In our current study, PDIA3-low was associated with LNM status, especially lateral LNM, and numbers of LNM. Although the p-value is statistically significant, it should be noted that sensitivity and specificity of low PDIA3 in predicting LNM were 54% (95%CI: 49–58%) and 53% (95%CI: 49–56%), respectively, and therefore, low PDIA3 in predicting LNM is associated with 46% false negatives and 47% false positives. Considering the different treatment approaches, the impact of subclinical LNM on prognosis [37], and the potential complications associated with dissection (Chyle leakage, phrenic nerve injury, and accessory nerve paralysis [38]), prophylactic lateral lymph node dissection is not always a required approach. Still, the diagnosis of “cancer recurrence” gives the patients psychological impact [39], and PDIA3 can be one of the ancillary markers for a comprehensive assessment of the lymph node status. Since LNM itself is not considered a strong prognostic factor [37], the question remains that PDIA3 expression in PTC was associated with lymph node status and CSS, but not with DFS. There should be an underlying mechanism to explain this discrepancy, and future experiment is essential.

Regarding limitations to this study, first, all evaluations were carried out using TMA slides. Ideally, evaluation by conventional whole section analysis is desirable. Second, despite that the strength of our study lies in the large cohort treated under a uniform strategy accompanied by long-term, precise outcome information described in the Method section and previous report [5], this study was a retrospective analysis of data prospectively accumulated from a single tertiary oncology referral center in Japan. Both global and prospective verification, in order to confirm these results would be advantageous. Third, we assessed PDIA3 expression only in PTCs of mainly classical type in this study. Further analysis including other histological subtypes—follicular thyroid carcinoma, poorly differentiated thyroid carcinoma, and anaplastic thyroid carcinoma—may help to understand the biological function of PDIA3 in thyroid cancer.

In conclusion, this is the first study to demonstrate that PDIA3 expression is significantly lower in PTC compared to NTT, and that a lower PDIA3 expression is correlated with nodal status, and a poor CSS. These findings suggest that PDIA3 may influence the biological behavior of PTC, although further study to elucidate the underlying mechanism is warranted.

Abbreviations

ER, endoplasmic reticulum; CSS, cause-specific survival; LNM, lymph node metastasis; MHC, major histocompatibility; NTT, normal thyroid tissue; PDIA3, protein disulfide isomerase A3; PTC, papillary thyroid carcinoma; TMA, tissue microarray

Acknowledgments

We sincerely thank Dr. Toshiaki Otsuka (Department of Hygiene and Public Health, Nippon Medical School), Dr. Tomohiro Hayakawa (Department of Gastrointestinal and Hepato-biliary-pancreatic Surgery, Nippon Medical School), Ms. Satoko Baba (Division of Pathology, The Cancer Institute Hospital, Japanese Foundation for Cancer Research), Ms. Ikue Hiraga (Department of Pathology, Nippon Medical School), Ms. Yoko Kawamoto (Department of Pathology, Nippon Medical School), Mr. Kiyoshi Teduka (Department of Pathology, Nippon Medical School), and Mr. Takenori Fujii (Department of Pathology, Nippon Medical School) for their support.

Disclosure Statement

None of the authors have any potential conflicts of interest associated with this research.

Conflict of Interest

The authors declare no conflict of interest.

Ethics Approval and Consent to Participate

This study was conducted in accordance with the principles embodied in the Declaration of Helsinki (revised in Brazil 2013). Patients’ consents were obtained for the use of clinical samples for research purposes according to the regulations defined by the Ethics Committee of Nippon Medical School Hospital (#30-03-1111, June 3, 2020) and institutional review board of the JFCR (IRB number 2013–1128, 24 January 2014).

Author Contributions

S.K. designed the study. A.E., K. Toda, H.M., and I.S. collected the clinical data. S.K., T.C., A.E., I.S., and K. Toda, reviewed the histopathology of all cases. S.K. and T.C. reviewed the histopathology and immunohistochemistry of all cases and interpreted the results. S.K., T.C., A.E., K. Toda, T.J., N.M., H.M., I.S., K. Takeuchi, and R.O. provided acquisition, analysis and interpretation of data, and statistical analysis. S.K. wrote the initial draft of the manuscript. I.S., K.Takeuchi, and R.O. supervised the study. All authors read and approved the final paper.

Funding

This study is supported by JSPS KAKENHI Grant Number 18H02635 (K.Takeuchi), and 20K17567(S.K.).

Data Availability Statement

The datasets analyzed during the current study are available from the corresponding author on reasonable request.

References

• 1   Udelsman  R,  Zhang  Y (2014) The epidemic of thyroid cancer in the United States: the role of endocrinologists and ultrasounds. Thyroid 24: 472–479.
• 2   Haugen  BR,  Alexander  EK,  Bible  KC,  Doherty  GM,  Mandel  SJ, et al. (2016) 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer. Thyroid 26: 1–133.
• 3   Costa  AM,  Herrero  A,  Fresno  MF,  Heymann  J,  Alvarez  JA, et al. (2008) BRAF mutation associated with other genetic events identifies a subset of aggressive papillary thyroid carcinoma. Clin Endocrinol (Oxf) 68: 618–634.
• 4   Ebina  A,  Togashi  Y,  Baba  S,  Sato  Y,  Sakata  S, et al. (2020) TERT promoter mutation and extent of thyroidectomy in patients with 1–4 cm intrathyroidal papillary carcinoma. Cancers (Basel) 12: 2115.
• 5   Sugitani  I,  Kasai  N,  Fujimoto  Y,  Yanagisawa  A (2004) A novel classification system for patients with PTC: addition of the new variables of large (3 cm or greater) nodal metastases and reclassification during the follow-up period. Surgery 135: 139–148.
• 6   Liu  X,  Bishop  J,  Shan  Y,  Pai  S,  Liu  D, et al. (2013) Highly prevalent TERT promoter mutations in aggressive thyroid cancers. Endocr Relat Cancer 20: 603–610.
• 7   Turano  C,  Gaucci  E,  Grillo  C,  Chichiarelli  S (2011) ERp57/GRP58: a protein with multiple functions. Cell Mol Biol Lett 16: 539–563.
• 8   Ramos  FS,  Serino  LT,  Carvalho  CM,  Lima  RS,  Urban  CA, et al. (2015) PDIA3 and PDIA6 gene expression as an aggressiveness marker in primary ductal breast cancer. Genet Mol Res 14: 6960–6967.
• 9   Chung  H,  Cho  H,  Perry  C,  Song  J,  Ylaya  K, et al. (2013) Downregulation of ERp57 expression is associated with poor prognosis in early-stage cervical cancer. Biomarkers 18: 573–579.
• 10   Shimoda  T,  Wada  R,  Kure  S,  Ishino  K,  Kudo  M, et al. (2019) Expression of protein disulfide isomerase A3 and its clinicopathological association in gastric cancer. Oncol Rep 41: 2265–2272.
• 11   Takata  H,  Kudo  M,  Yamamoto  T,  Ueda  J,  Ishino  K, et al. (2016) Increased expression of PDIA3 and its association with cancer cell proliferation and poor prognosis in hepatocellular carcinoma. Oncol Lett 12: 4896–4904.
• 12   Kondo  R,  Ishino  K,  Wada  R,  Takata  H,  Peng  WX, et al. (2019) Downregulation of protein disulfideisomerase A3 expression inhibits cell proliferation and induces apoptosis through STAT3 signaling in hepatocellular carcinoma. Int J Oncol 54: 1409–1421.
• 13   Caorsi  C,  Niccolai  E,  Capello  M,  Vallone  R,  Chattaragada  MS, et al. (2016) Protein disulfide isomerase A3-specific Th1 effector cells infiltrate colon cancer tissue of patients with circulating anti-protein disulfide isomerase A3 autoantibodies. Transl Res 171: 17–28. E2.
• 14   Wang  K,  Li  H,  Chen  R,  Zhang  Y,  Sun  XX, et al. (2017) Combination of CALR and PDIA3 is a potential prognostic biomarker for non-small cell lung cancer. Oncotarget 8: 96945–96957.
• 15   Choe  MH,  Min  JW,  Jeon  HB,  Cho  DH,  Oh  JS, et al. (2015) ERp57 modulates STAT3 activity in radioresistant laryngeal cancer cells and serves as a prognostic marker for laryngeal cancer. Oncotarget v6: 2654–2666.
• 16   Zhang  H,  Zhou  Y,  Cheng  Q,  Dai  Z,  Wang  Z, et al. (2020) PDIA3 correlates with clinical malignant features and immune signature in human gliomas. Aging (Albany NY) 12: 15392–15413.
• 17   Sugitani  I,  Fujimoto  Y,  Yamada  K,  Yamamoto  N (2008) Prospective outcomes of selective lymph node dissection for papillary thyroid carcinoma based on preoperative ultrasonography. World J Surg 32: 2494–2502.
• 18   Bankhead  P,  Loughrey  MB,  Fernandez  JA,  Dombrowski  Y,  McArt  DG, et al. (2017) QuPath: Open source software for digital pathology image analysis. Sci Rep 7: 16878.
• 19   Cady  B,  Rossi  R (1988) An expanded view of risk-group definition in differentiated thyroid carcinoma. Surgery 104: 947–953.
• 20   Lee  AS (1981) The accumulation of three specific proteins related to glucose-regulated proteins in a temperature-sensitive hamster mutant cell line K12. J Cell Physiol 106: 119–125.
• 21   Soderberg  A,  Hossain  A,  Rosen  A (2013) A protein disulfide isomerase/thioredoxin-1 complex is physically attached to exofacial membrane tumor necrosis factor receptors: overexpression in chronic lymphocytic leukemia cells. Antioxid Redox Signal 18: 363–375.
• 22   Bourdi  M,  Demady  D,  Martin  JL,  Jabbour  SK,  Martin  BM, et al. (1995) cDNA Cloning and baculovirus expression of the human liver endoplasmic reticulum p58: characterization as a protein disulfide isomerase isoform, but not as a protease or a carnitine acyltransferase. Arch Biochem Biophys 323: 397–403.
• 23   Busch  DH,  Pamer  EG (1998) MHC class I/peptide stability: implications for immunodominance, in vitro proliferation, and diversity of responding CTL. J Immunol 160: 4441–4448.
• 24   Li  Y,  Camacho  P (2004) Ca2+-dependent redox modulation of SERCA 2b by ERp57. J Cell Biol 164: 35–46.
• 25   Obeid  M (2008) ERP57 membrane translocation dictates the immunogenicity of tumor cell death by controlling the membrane translocation of calreticulin. J Immunol 181: 2533–2543.
• 26   Panaretakis  T,  Joza  N,  Modjtahedi  N,  Tesniere  A,  Vitale  I, et al. (2008) The co-translocation of ERp57 and calreticulin determines the immunogenicity of cell death. Cell Death Differ 15: 1499–1509.
• 27   Ramirez-Rangel  I,  Bracho-Valdes  I,  Vazquez-Macias  A,  Carretero-Ortega  J,  Reyes-Cruz  G, et al. (2011) Regulation of mTORC1 complex assembly and signaling by GRp58/ERp57. Mol Cell Biol 31: 1657–1671.
• 28   Grindel  BJ,  Rohe  B,  Safford  SE,  Bennett  JJ,  Farach-Carson  MC (2011) Tumor necrosis factor-alpha treatment of HepG2 cells mobilizes a cytoplasmic pool of ERp57/1,25D(3)-MARRS to the nucleus. J Cell Biochem 112: 2606–2615.
• 29   Leys  CM,  Nomura  S,  LaFleur  BJ,  Ferrone  S,  Kaminishi  M, et al. (2007) Expression and prognostic significance of prothymosin-alpha and ERp57 in human gastric cancer. Surgery 141: 41–50.
• 30   Ogino  T,  Bandoh  N,  Hayashi  T,  Miyokawa  N,  Harabuchi  Y, et al. (2003) Association of tapasin and HLA class I antigen down-regulation in primary maxillary sinus squamous cell carcinoma lesions with reduced survival of patients. Clin Cancer Res 9: 4043–4051.
• 31   Chen  CN,  Chang  CC,  Su  TE,  Hsu  WM,  Jeng  YM, et al. (2009) Identification of calreticulin as a prognosis marker and angiogenic regulator in human gastric cancer. Ann Surg Oncol 16: 524–533.
• 32   Atkins  D,  Ferrone  S,  Schmahl  GE,  Storkel  S,  Seliger  B (2004) Down-regulation of HLA class I antigen processing molecules: an immune escape mechanism of renal cell carcinoma? J Urol 171: 885–889.
• 33   Mehta  AM,  Jordanova  ES,  Kenter  GG,  Ferrone  S,  Fleuren  GJ (2008) Association of antigen processing machinery and HLA class I defects with clinicopathological outcome in cervical carcinoma. Cancer Immunol Immunother 57: 197–206.
• 34   Angell  TE,  Lechner  MG,  Jang  JK,  LoPresti  JS,  Epstein  AL (2014) MHC class I loss is a frequent mechanism of immune escape in papillary thyroid cancer that is reversed by interferon and selumetinib treatment in vitro. Clin Cancer Res 20: 6034–6044.
• 35   Takami  H,  Ito  Y,  Noguchi  H,  Yoshida  A,  Okamoto  T (eds) (2013) Treatment of Thyroid Tumor, Japanese Clinical Guidelines. Springer Japan, Tokyo, Japan.
• 36   Ito  Y,  Onoda  N,  Okamoto  T (2020) The revised clinical practice guidelines on the management of thyroid tumors by the Japan Associations of Endocrine Surgeons: core questions and recommendations for treatments of thyroid cancer. Endocr J 67: 669–717.
• 37   Ito  Y,  Higashiyama  T,  Takamura  Y,  Kobayashi  K,  Miya  A, et al. (2011) Prognosis of patients with papillary thyroid carcinoma showing postoperative recurrence to the central neck. World J Surg 35: 767–772.
• 38   Ito  Y,  Higashiyama  T,  Takamura  Y,  Miya  A,  Kobayashi  K, et al. (2007) Risk factors for recurrence to the lymph node in papillary thyroid carcinoma patients without preoperatively detectable lateral node metastasis: validity of prophylactic modified radical neck dissection. World J Surg 31: 2085–2091.
• 39   Misra  S,  Meiyappan  S,  Heus  L,  Freeman  J,  Rotstein  L, et al. (2013) Patients’ experiences following local-regional recurrence of thyroid cancer: a qualitative study. J Surg Oncol 108: 47–51.