EWSR1 rearrangement in papillary thyroid microcarcinoma is related to classic morphology and the presence of small-cell phenotype

The Ewing sarcoma breakpoint region 1 (EWSR1) rearrangements with unknown genes were detected in a high percentage of classic variants of papillary thyroid carcinoma. The small-cell carcinoma of the thyroid with Ewing family tumor elements typically presents with EWSR1-FLI1 rearrangement suggesting the possible role of EWSR-FLI1 translocation in the loss of thyroid differentiation and acquisition of a small-cell phenotype. To determine the frequency and association of EWSR1 rearrangements, particularly the EWSR1-FLI1 fusion with clinicopathological features of papillary thyroid microcarcinoma (m-PTC) and the presence of small cells, we analyzed a series of 99 m-PTCs using the fluorescence in situ hybridization method. Ninety cases (90.9%) of m-PTC were positive for small cells. This group of m-PTC has shown more often invasive growth, lymphatics invasion, and moderate/extended intratumoral fibrosis. Three cases out of 99 were inconclusive for EWSR1 rearrangement. Eighty-nine (92.7%) and 27 (28.1%) out of 96 m-PTC cases were positive for EWSR1 rearrangement and EWSR1-FLI1 fusion, respectively. m-PTC with classical architectural pattern presented more frequently with EWSR1 rearrangement relative to m-PTC with other patterns (p = 0.005). Other clinicopathological features were not related to the presence of EWSR1 rearrangement or EWSR1-FLI1 fusion. The percentage of small cells present significantly correlated with the percentage of cells positive for EWSR1-FLI1 fusion (p = 0.05) and EWSR1 rearrangement (p < 0.001). EWSR1-FLI1 fusion is not rare in m-PTC and it is associated with the acquisition of small-cell phenotype. The EWSR1 gene rearrangement is a frequent event in m-PTC and is related to the classical pattern of m-PTC.


INTRODUCTION
The Ewing sarcoma breakpoint region 1 (EWSR1) gene encodes a protein that is involved in various cellular processes, including regulation of gene expression at the level of transcription initiation, signal transduction, and mRNA transport [1,2]. In addition to its role in gene expression, EWSR1 is also a player in preserving genome integrity and cellular senescence, as previously demonstrated [1,2]. Rearrangements between EWSR1 and other genes lead to the formation of hybrid proteins with oncogenic potential associated with neoplastic cell transformation, both in the mesenchymal neoplasms and in other cancer models [3,4]. EWSR1 rearrangements with different partner genes are related to different phenotypes of neoplasms, namely small-cell, spindle-cell, clear-cell phenotype or neoplasms with distinctive myxoid stroma [3]. The Ewing Sarcoma family of tumors (ESFT) encompasses tumors with common small-cell phenotype and molecular features that include balanced reciprocal translocations between EWSR1 and genes of the ETS family of transcription factors. The two most common rearrangements include fusion of EWSR1 with FLI1 and ERG genes, which were described in 85% and 10% of cases, respectively [3,4]. EWSR1 rearrangement is also present in a small number of non-mesenchymal tumors including benign and malignant myoepithelial tumors and hyalinizing clear cell carcinomas of the salivary glands [5][6][7], hidradenoma of the skin [8], clear cell odontogenic carcinoma [9] and primary hyalinizing clear cell carcinoma of the thymus [10]. Rare cases of EWSR1 translocation were also described in renal cell carcinoma and colon adenocarcinoma [11,12].
In the group of thyroid tumors, the small-cell carcinoma of the thyroid with Ewing family tumor elements (CEFTE) typically discloses EWSR1-FLI1 rearrangement [13][14][15]. This rearrangement was also described in rare cases of papillary www.bjbms.org extension, predominant architectural pattern of the tumor, growth pattern, multifocality, lymphatics invasion, vascular invasion, intratumoral fibrosis, stromal calcification, and psammoma bodies.
The presence of small-cell phenotype was also analyzed through the presence of small cells. The number of small cells was determined semi-quantitatively after analysis of the whole sections of the tumor and it was estimated as a percentage of the total number of tumor cells. Small cells were defined as cells smaller than surrounded tumor cells of PTC, not larger than 4x lymphocyte size, with scant or inconspicuous cytoplasm, hyperchromatic oval or irregular nuclei or nuclei with incomplete features of PTC (moderate nuclear enlargement with chromatin clearing, with or without nuclear grooves or pseudoinclusions). The cellular clusters which appear as small nests of highly hyperchromatic, deformed nuclei usually present at the edge of peripherally localized m-PTC near to surgical resection margins, were considered as an artifact and were not counted as small cells [23]. Slightly distorted cells present in the deeper parts of the tumor surrounded by well-preserved small cells and regular cells of m-PTC were considered as small cells, as small cells commonly undergo crushing artifact [24]. The cases of m-PTC with less than 1% of small cells were considered as negative for small cells, and cases with at least 1% of small cells were counted as m-PTC positive for small cells. Figure 1 is shown the morphology of small cells in m-PTC. Two pathologists (BK, SC) reviewed the Hematoxylin and eosin-stained (H&E) slides of all cases and histological parameters were analyzed according to the classification of the World Health Organization Classification of Tumors of Endocrine Organs [16]. The staging of disease was performed according to the AJCC/TNM Staging System for PTC (8 th Edition) [25].
Fluorescence in situ hybridization (FISH) testing Tissue microarrays (TMA) from the 99 cases were built and an H&E stained section from each TMA paraffin block were produced to evaluate the representativity of the m-PTCs in each core. FISH was performed according to the methodology previously described for the technique and for the establishment of the cutoff [13]. FISH technique was performed in 4µm tissue sections obtained from the TMAs paraffin blocks. The sections were extended on polysine slides, following deparaffinization with xylene and alcohol. Antigen retrieval was performed using citrate buffer pH6.0 at 95°C (Thermo Scientific, Waltham, MA USA) and enzymatic digestion with pepsin at 37°C (Sigma-Aldrich Quimica, Sintra, Portugal). Afterward, co-denaturation at 79°C and hybridization at 37°C were performed overnight in a SPOT-Light CISH Hybridizer with a ZytoLight SPEC EWSR1-FLI1 TrickCheck Probe thyroid carcinoma (PTC), suggesting that PTC may be a possible origin of CEFTE [13][14][15]. The same authors question the possible role of EWSR-FLI1 rearrangement in the loss of thyroid differentiation and acquisition of a small-cell phenotype. In PTC, the EWSR1 rearrangements with unknown genes were also detected in a high percentage of classic variants of PTC cases [13].
Papillary thyroid microcarcinoma (m-PTC) is a PTC with 10 mm or less in the highest dimension [16]. m-PTC is the most common variant of PTC and discloses an excellent prognosis. Despite its favorable general clinical behavior, a group of m-PTC includes tumors with a potentially aggressive course, such as those giving rise to loco-regional metastases, recurrence of the disease, rarely, and distant metastasis. m-PTC can disclose varied growth patterns and cellular features, as described in larger PTC [17][18][19]. The mitogen-activated protein kinase pathway is the most frequently altered signaling pathway in m-PTCs, being the presence of BRAF V600E mutation related to a distinctive phenotype (classic morphology, tall cell features, infiltrative borders, presence of stromal fibrosis and psammoma bodies, and well-developed nuclear features typical of PTC). The association of BRAF V600E mutation and aggressiveness of m-PTC is questionable [17,19,20]. Besides the BRAF V600E mutation, mutation of the RAS gene, like in larger PTC is related to the follicular pattern of m-PTC, as well as, RET/PTC rearrangements are associated with the classical pattern and/or the so-called "Bonsai phenotype" of m-PTC [16,21].
This study aimed to evaluate the frequency of EWSR1 rearrangements, particularly the EWSR1-FLI1 fusions, in m-PTC, their possible relation with clinicopathological features of m-PTC and their possible association with the presence of a small-cell phenotype.

Patient selection
A series of 99 m-PTC cases with a minimum size of 5 mm was selected from the archive of the Institute of Pathology and Forensic Medicine of Medical Military Academy (Belgrade, Serbia) according to the previously described methodology [22].

Clinicopathological features of m-PTC
General clinical information such as age, gender, and type of surgical procedure were obtained from the medical records of the patients. The following pathohistological parameters were reviewed for each case according to the previously described methodology [22]: Tumor size, presence of LN metastases, presence of distant metastases, extrathyroid www.bjbms.org EWSR1 and the blue signal of FLI1, the case was counted positive for EWSR1-FLI1 fusion -t (11;22). A normal nucleus was accepted when exhibiting 2 signals fusion of EWSR1 (5'-orange and 3'-green) and 2 blue FLI1 signals separated.
The cases were considered positive for EWSR1-FLI1 fusion if this fusion signal was present in ≥2.4% of neoplastic nuclei. Positive cases for EWSR1 rearrangement were considered in cases when break-apart signal was present in ≥2.4% of neoplastic nuclei. Polysomy was acknowledged when the number of signals per neoplastic nucleus was superior to those observed in the normal nuclei. Monosomy was acknowledged when the number of signals in each neoplastic nucleus was inferior to those observed in the normal nuclei.
(ZytoVision, Bremerhaven, Germany). Nuclei were counterstained with DAPI (ZytoVision, Bremerhaven, Germany). Adequate positive and negative controls were used in each set of slides. Fluorescence hybridization signals were analyzed and recorded with a D-Sight Fluo 2.0 (Menarini, Florence, Italy).
In each case, 50 nuclei were evaluated, and records were undertaken regarding EWSR1 rearrangement, EWSR1 monosomy, EWSR1 polysomy, and EWSR1-FLI1 fusion. Cases with least than 50 nuclei with preserved signals due to technical reasons were classified as inconclusive. EWSR1 rearrangement was contemplated if a signal separation of the 5'-orange and the 3'-green of the EWSR1 probe was present. If this signal separation was simultaneous with the fusion of the 5'-orange signal of The relationship between nuclear size and EWSR1 rearrangement Of the five selected cases of m-PTC with the highest percentage of nuclei with EWSR1-FLI1 fusion and EWSR1 rearrangement, the perimeter of the nucleus of cells was measured using the morphometric tools of the D-Sight Navigator viewer. The perimeter of the neoplastic nuclei was compared after its division into three groups: Group 1-neoplastic nuclei without EWSR1 rearrangement; Group 2 -neoplastic nuclei with EWSR1 rearrangement, and Group 3 -neoplastic nuclei with EWSR1-FLI1 fusion.

Ethical statement
This work was performed in archived paraffin tissue that remained from the diagnosis of the thyroid samples of patients that were anonymized. The materials used for the work are not needed for diagnosis in the present nor in the future. The Ethics Committee of the Military Medical Academy and University of Defense, Serbia, approved this study (Number: 16-71; 5/29/2019).

Statistical analysis
Statistical analysis of data was done with the statistical software package, SPSS Statistics 18. Most of the variables were presented as frequency of certain categories, while statistical significance of differences was tested with the Chisquare test (F test in case of expected frequencies <5). In case of continuous data, variables were presented as mean value ± standard deviation, minimal and maximal values. The Kolmogorov-Smirnov test was applied to assess the normality of distribution of continuous data. According to the results of this test, statistical significance between groups was tested by ANOVA test (post hoc Tukey test). Spearman correlation analysis was used to establish the relationship between parameters. All the analyses were estimated at minimal p < 0.05 level of statistical significance.

RESULTS
In the selected cohort of 99 m-PTCs, the patients had a mean age of 49 ± 14.4 years, ranging from 19 to 85 years, at the time of diagnosis. There were 22 males (22.2%) and 77 females (77.8%). The mean tumor size was 7.6 ± 1.5 mm. Cervical LNs metastases were present in 30 cases (30.3%). Fifteen patients had LNs metastases in the central neck compartment and 15 patients had lateral cervical LNs metastases. None of the patients had documented distant metastases. In the database of pathohistological reports, six cases (6.06%) had histological confirmation of disease recurrence in regional LNs. According to the medical records, these patients received radioiodine therapy 6-8 weeks after thyroidectomy was performed. Most of the patients (n = 92; 92.9%) were Stage I and 7 of patients (7.07%) were Stage II of disease. Only three cases (3%) showed gross ETE. The classical architectural pattern of the tumor was the most common and was present in 55 cases (55.6%), followed the follicular pattern in 38 cases (38.4%) and solid in six cases (6.1%). The infiltrative growth pattern was observed in 72 cases (72.7%) and 27 cases (27.3%) showed an expansive/ circumscribed growth pattern. Sixty-four cases (64.6%) disclosed lymphatics invasion and vascular invasion was present in 15 cases (15.2%). Tumor multifocality was observed in 81 cases (81.81%). Moderate/extended intratumoral fibrosis was observed in 57 cases (59.4%), and absent/mild intratumoral fibrosis was observed in 39 cases (40.6%). Stromal calcification was observed in 33 cases (33.3%) and psammoma bodies were observed in 46 cases (46.5%).
In the studied cases, the mean percentage of small cells was 12.3% ± 16.9%, ranging from 0% to 80%. Nine cases (9.1%) of m-PTC were considered negative for small cells and 90 cases (90.9%) of m-PTC were positive for small cells. The mean percentage of small cells in the small cells positive m-PTC cases was (13.6% ± 17.2%) ranging from 1% to 80%. In the group of m-PTC positive for small cells the presence of invasive growth, lymphatic invasion, and moderate/extended intratumoral fibrosis were statistically more common than in the group of m-PTC negative for small cells. In regard to intratumoral occurrence of small cells, it was not possible to establish the specificity of their topographic distribution. Most commonly, small cells were present as single or clusters of cells mixed or surrounded by cells with typical PTC nuclear features. It was possible to notice areas of small cells in the vicinity, within but also away from intratumoral fibrosis. The clinicopathological features of m-PTC with and without the presence of small cells are described in Table 1.
After FISH analysis, three cases (3.03%) without signals in less than 50 nuclei were considered inconclusive and excluded from further analysis. Inconclusive cases were in two females and one male patient and all cases were in group staging I. Two cases were classical and one was solid in morphology. Two of them had LNs metastases.
Eighty-nine out of 96 m-PTC cases (92.7%) were considered positive for EWSR1 rearrangement. The mean percentage of cells with EWSR1 rearrangement in the EWSR1 positive m-PTC cases was (16.7% ± 9.7%) ranging from 4% to 66%. Figure 2 displays FISH analysis of cases with EWSR1 rearrangement and EWSR1-FLI1 fusion.
Regarding the occurrence of small cells in m-PTC, the percentage of their presence significantly correlated with the percentage of cells with EWSR1-FLI1 fusion (ρ = 0.201; p = 0.05) and EWSR1 rearrangement (ρ = 0.349; p <0.001), and it was in a negative correlation with the percentage of neoplastic nuclei with monosomy (ρ = -0.202; p = 0.048). This association is further supported by the results of five m-PTC cases that have shown the statistically significant smaller size of neoplastic nuclei with EWSR1-FLI1 fusion and EWSR1 rearrangement in comparison to the size of neoplastic nuclei of m-PTC without EWSR1 rearrangement. Neoplastic nuclei with EWSR1-FLI1 fusion were also smaller than nuclei with EWSR1 rearrangement, but statistically, a significant difference was proven only in case number 5 (F = 5.889; p = 0.006; Tukey post hoc test; p = 0.047). Table 2 summarizes the results of the study of the nuclear perimeter of individual m-PTC cases and the total number of analyzed nuclei. Figure 3  The percentage of monosomic neoplastic nuclei in m-PTC ranged from 0% to 72% (mean 13.7% ± 12.7%) and was detected in 88 cases (91.7%). Eight cases were negative for monosomy and in six cases monosomy was present in 2.0% of neoplastic nuclei. An increasing percentage of neoplastic nuclei with monosomy of EWSR1 gene in more than 2.4% of neoplastic cells is in a statistically significant negative correlation with the percentage of EWSR1 rearrangement (ρ = -0.536; p < 0.001), the presence of LN metastasis (ρ = -0.319; p < 0.002) and lymphatics invasion (ρ = -0.252; p < 0.013).
Only one (1.0%) out of 96 cases disclosed the presence of polysomy of EWSR1 gene that was detected in 2% of nuclei.

DISCUSSION
EWSR1 is one of the most frequently altered genes in PTC [13]. The meaning of EWSR1 rearrangement is unknown, but per se, looks like it does not have any significant clinical meaning, as documented previously and in the present study, namely, by the lack of association between EWSR1 rearrangement and clinicopathological features such as the presence of LN metastasis and distant metastasis. This finding, regarding the high frequency of EWSR1 rearrangement in m-PTC, is in line with available evidence that there is no single genetic change that alone can predict the aggressiveness of m-PTC [17,19].
EWSR1 rearrangement represents one of the most frequent molecular events in mesenchymal neoplasm [3,4], and a rare genetic event, that has an increasing number of reported cases, in non-mesenchymal tumors [5][6][7][8][9][10][11][12][13]. In the present study, the rearrangement of the EWSR1 gene was detected in all cases of m-PTC with a classical pattern. The frequency of EWSR1 gene rearrangements in m-PTC with other patterns was significantly lower even absent as occurred in solid pattern m-PTCs supporting the phenotype-related significance of EWSR1 rearrangements. Previously, it has been demonstrated that the classic architectural pattern of m-PTC and the classic variant of PTC were also related to the presence of BRAF V600E mutation and RET-PTC rearrangement. As we did not analyze other molecular alterations and their association with EWSR1 rearrangement, we cannot exclude that different combinations of genetic alterations in association with EWSR1 rearrangement are responsible for the development of the classical pattern m-PTC.
Our results have been shown that the region of EWSR1 is highly prone to frequent alteration in m-PTC, not only as recurrent balanced aberrations but also as genes monosomy. The high frequency of EWSR1 monosomy has not been shown to be associated with clinicopathological features of m-PTC. Moreover, m-PTC cases without EWSR1 monosomy were those showing LNs metastases and lymphatics invasion than m-PTC with EWSR1 monosomy. This fact can suggest that the loss of the EWSR1 gene from neoplastic cells of m-PTC can be perceived as a secondary aberration that accumulates during tumor evolution, probably as a result of genome instability. Furthermore, all m-PTC cases without EWSR1 www.bjbms.org monosomy had a high percentage of EWSR1 rearrangement with unknown partner' s genes. Further studies for the identification of one or more partner genes are necessary and may lead to the identification of specific translocation related to the metastatic potential of m-PTC.
Considering the association of EWSR1-FLI1 rearrangement with the presence of small-cell phenotype in mesenchymal neoplasm, this study analyzed the presence of small cells in EWSR1-FLI1 positive m-PTC. The current study demonstrated a significant positive correlation between the presence of small cells and the percentage of cells with EWSR1-FLI1 rearrangement in m-PTC. These findings were also supported by the results of the nuclei perimeter measurements, being the smallest size of m-PTC nuclei are the nuclei positive for EWSR1-FLI1 fusion. This finding is in line with the idea that this translocation has a role in the acquisition of small-cell phenotype, not only in mesenchymal neoplasms but also in thyroid tumors, specifically CEFTE, and now, in m-PTC [13,14]. The small-cell phenotype can be present in a wide variety of secondary and rarely primary thyroid tumors, such as medullary thyroid carcinoma (MTC), poorly differentiated thyroid carcinoma (PDTC) and lymphoma [16,26]. In general, a designation of thyroid malignancy as small cell thyroid carcinomas is related to the aggressive clinical course of these tumors [27]. The small cell variant of MTC tends to be more aggressive than conventional MTC [16]. The situation is not so uniform in PDTC where the presence of cells with small size is associated with more aggressive behavior of oncocytic PDTC, at variance to its conventional variant when the occurrence of small cells does not influence prognosis [28].
In m-PTC, the most important diagnostic criterion is the presence of nuclear changes [16] regardless of the size of the cells. The presence of a small-cell phenotype in PTC was not documented in the literature until now, but in few studies about PTC morphology, the presence of cells with different appearances has been discussed. According to our methodology, at least in part, cells that we recognized as small, corresponding to previously described unusual cellular forms of PTC, such as cells with a dark nucleus [29], dormant cells [30] or cells with incomplete nuclear features of PTC [31,32].
The meaning of these unusual cellular forms previously described is unknown, but it was assumed that they have been sequestered from the normal cell cycle because they did not show proliferative activity [30]. In another study, it was reported that the presence of cells with dark nuclei were a more common finding in malignant thyroid tumors than in benign one, and their presence was assumed to be useful assisting the diagnosis of malignancy, specifically in difficult cases of follicular lesions suspected for a follicular variant of PTC [29].
In the present series, the association of small cells-positive m-PTC was observed between the presence of invasive growth, lymphatic invasion, and moderate/extended intratumoral fibrosis of m-PTC. These findings suggest a possible role of small cells in the aggressiveness of m-PTC like it was previously shown for lymphatics invasion and intratumoral fibrosis [22,33]. The development of intratumoral fibrosis required time, and since most m-PTC represents old lesions, an association of intratumoral fibrosis and small cells may indicate that acquisition of small-cell phenotype is a consequence of m-PTC aging. Future studies of small cells in the larger series of m-PTC with a long time follow and with a comparison of small cell significance in larger PTC, may be useful in determining their clinical significance.
In the group of mesenchymal neoplasms, the EWSR1-FLI1 rearrangement has been related to an aggressive clinical course [3,4] at variance with what is documented in the EWSR1-FLI1 positive m-PTC of the current series. This could be explained by the fact that EWSR1-FLI1 positive cells represent only a small cellular clone in m-PTC which was not sufficient to induce the change of tumor behavior with a consequent aggressive course. However, EWSR1-FLI1 rearrangement is present in all tumor cells of CEFTЕ and a small-cell phenotype is completely developed, but prognosis so far is  [13][14][15]. On the contrary, rare cases of pancreatic neuroendocrine tumors with EWSR1-FLI1 rearrangement were not showing small-cell phenotype and other clinicopathological features of ESFT [34]. Accordingly, the acquisition of specific tumor phenotype does not depend only on the  presence of EWSR1-FLI1 fusion. The same fusion transcript could induce phenotype change only in the right cellular context because there are significant phenotype differences in various tumors that arise from the different cell types with identical gene fusion [3,4]. As well, additionally, genetic and epigenetic events that can also influence of EWSR1-FLI1 protein activity are necessary as its continuous expression is required to maintain the oncogenic phenotype of ESFT cells, acquisition of specific tumor phenotype and clinical behavior [3,4,35,36].
CONCLUSION EWSR1-FLI1 rearrangement is not rare in m-PTC and it is related to the acquisition of small-cell phenotype. The molecular alterations of the EWSR1 gene, monosomy, and rearrangements, are frequent events in m-PTC, they have, apparently, no clinical significance and are related to the classic pattern of m-PTC.