Introduction

Thyroid cancer is the most common endocrine malignant tumor and its incidence has markedly increased worldwide in recent decades (Siegel et al. 2019; La Vecchia et al. 2015). Although papillary thyroid carcinoma (PTC), the most common type of thyroid carcinoma, has a high overall survival, approximately 10% of PTC cases are characterized as aggressive disease and high mortality rates (McLeod et al. 2013; Roman et al. 2017; Durante et al. 2013), such as locally advanced PTC (LAPTC), which in here refers to tumor infiltrating the perithyroid organs, including the muscle, skin, larynx, trachea, esophagus, important nerve and vascular tissues. Recently, various studies have emphasized the importance of risk stratification to design individualized treatments for patients with aggressive PTC (Xing et al. 2013a, b; Nikiforov et al. 2011). The molecular-based risk stratification can provide insight into the accuracy of earlier identification of LAPTC. So, it is important to identify novel molecular biomarkers to improve the accurate identification of high recurrence risk and poor prognosis in patients with thyroid carcinoma.

Genetic mutations may be crucial in the progress of thyroid cancer. Many studies reported a role of BRAFV600E mutation in aggressive PTCs (Xing et al. 2005, 2015; Elisei et al. 2008; Xing et al. 2013a). BRAFV600E mutation occurs in about 50% of patients with PTC and has been reported to be associated with the high-risk features of PTC, including lymph node metastasis (LNM), larger tumor size and advanced tumor stages (Xing et al. 2005; Xing et al. 2013a ; Li et al. 2012). However, other studies reported no significant associations between BRAFV600E mutations and high‑risk PTC characteristics (Brzezianska et al. 2007; Nasirden et al. 2016; Gandolfi et al. 2015; Gouveia et al. 2013). Hence, BRAFV600E mutation alone may have limited utility as a prognostic biomarker of PTC.

More recently, human telomerase reverse transcriptase (TERT) promoter mutations (C228T/ C250T) were found in about 10% of PTC patients and have been associated with disease severity (Vinagre et al. 2013; Liu et al. 2013; Landa et al. 2013). TERT promoter mutations were also shown to be associated with aggressiveness of other human cancers, such as melanoma, brain tumor and bladder cancer (Griewank et al. 2014; Killela et al. 2013; Rachakonda et al. 2013). TERT is an important gene involved in the maintenance of chromosomal integrity and genome stability (Alzahrani et al. 2016). Mutations in the TERT promoter lead to TERT overexpression, thus facilitating cancer growth (Vinagre et al. 2013; Huang et al. 2013). Previous studies have shown that TERT promoter mutation is involved in the pathogenesis of tumors and is associated with tumor aggressiveness (Alzahrani et al. 2016). However, the role of TERT promoter mutations in the aggressiveness of PTC remains to be further investigated.

The co-existence of gene mutation has been focused. However, the clinicopathological significance of synergetic BRAFV600E and TERT promoter mutations (B&T) on LAPTC was unclear. The relationship of coexisting genetic mutations and aggressiveness of thyroid cancers requires further investigation. In this study, we focused on the prevalence of genetic mutations using ThyroSeq v2 next-generation sequencing (57-gene panel) to identify molecular characteristics of LAPTC and analyze the genomic features of LAPTC. The correlation between morphological features and genomic features was further investigated. Then the pathological subtypes and clinicopathological features between tumors with or without B&T mutations were compared. Additionally, the prognosis of LAPTC patients with different genetic alterations was also combined to analyze.

Materials and methods

Study participant

An observational cohort study was performed to identify molecular profile in LAPTC and the correlation with clinicopathological characteristics. This prospective study planned to enroll 200 patients with aggressive thyroid cancer, who underwent surgery in Shanghai Jiao Tong University Affiliated Sixth People’s Hospital between January 2018 and June 2021. All patients will be screened and suggested for the detection of 57-gene panel sequencing, which was conducted on post-surgical fresh tumor tissue in accordance with the inclusion criteria. After obtaining informed consent, we collected information on age at diagnosis, gender, pathology, tumor diameter, multifocality, extrathyroidal invasion, LNM, and tumor stage. All patients will receive at least 6 months follow-up. The inclusion flowchart can be seen in Fig. 1. The research was approved by the Ethics Committees from Shanghai Jiao Tong University Affiliated Sixth People’s Hospital. All patients provided written informed consent.

Fig. 1
figure 1

A flowchart of the study selection. PTC papillary thyroid carcinoma; ETE extrathyroidal extension; LNM lymph node metastasis

Full size image

Eligibility criteria

Inclusion criteria

(1) Patients were preoperatively considered as locally advanced thyroid carcinoma; (2) patients underwent initial surgery and was morphologically confirmed to be papillary thyroid carcinoma; (3) patients did not receive other treatments, such as radioactive iodine-131, radiofrequency ablation (except neoadjuvant targeted therapy); and (4) patients were willing to undergo genetic testing.

Exclusion criteria

(1) Reoperation; (2) non papillary thyroid carcinoma; (3) age > 80 years old or patients with serious medical illness; and (4) patients who declined genetic testing or lacked clinicopathological data.

Genomic DNA isolation and next-generation sequencing

Genomic DNA was isolated from fresh tumor tissues using QIAamp DNA Kit (Qiagen) according to manufacturer’s instructions. The concentration of the DNA was determined using a Qubit fluorometer 3.0 (Life Technologies). A 57-gene panel for primary thyroid cancer (thyroid cancer panel, Thcp) was designed, specially based on the data in the Cancer Genome Atlas (TCGA), National Comprehensive Cancer Network (NCCN) and literatures. Driver genes associated with thyroid cancer and other solid tumors and targets related with targeted therapy, chemotherapy and immunotherapy of thyroid cancer were included in this panel. 50–100 ng of sheared genomic DNAs were subjected to library construction with an MGIEasy universal DNA library kit (MGI, China), then followed by hybrid capture using an xGen Hybridization and Wash Kit (IDT, USA). Libraries’ quality and concentration were determined using a LabChip® GX Touch™ nucleic acid analyzer (PerkinElmer, USA) and a Qubit fluorometer 3.0 (Life Technologies, USA), respectively. The qualified libraries were sequenced with 2 × 100 bp paired-end reads on a MGISEQ-2000 (MGI, China) platform. ThyroSeq v2 next-generation sequencing was performed on available tissue for mutation detection and genotype-clinicopathological correlation of the tumor was analyzed.

Statistical analysis

Continuous data were expressed by means ± standard deviations (mean ± SD) or medians (range) and categorical data were summarized by frequencies and percentages (%). Mann–Whitney U test was used for comparison of continuous variables. Pearson chi-square test or Fisher’s exact test was used for comparison the clinicopathological features among LAPTC subgroups with or without B&T mutations. The associations between mutation status and various characteristics, such as the pathological subtypes, were determined by calculation of odds ratios. The correlation between the probability of B&T co-mutation and age was analyzed by multivariate logistic regression. Patient recurrence and survival status with different genetic alterations were also combined to analyze.

A P value of < 0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using R3.6.3 software and SPSS statistical software (version 22.0, SPSS Inc., Chicago, IL).

Results

Demographic and clinicopathological characteristics of the study cohort

A total of 137 patients were enrolled and 65 LAPTC patients with initial surgery met criteria for inclusion were included in the final analysis. Among them, 41.5% (27/65) of these patients were women; the median age at diagnosis was 41 (16–80) years old and 29.2% (19/65) of them were more than 55 years old; the median max-diameter of nodes is 17 mm (4–45 mm) and 61.5% (40/65) of cases were multifocality; histopathologic analysis of these tumors revealed 43 PTCs with BRAFV600E mutation and 17 PTC cases with B&T co-mutation; about 55.4% (36/65) cases exhibited characteristic of extrathyroidal invasion and 73.9% (48/65) cases had lateral lymphatic metastases; 27.7% (18/65) of patients were classified as advanced TNM stage (III/IV). After follow-up for 6–48 months (mean 24.6 months), with regard to mortality, three cases died in LAPTC with co-existence of B&T mutations and no patient died in LAPTC without B&T co-mutation. Demographic and clinicopathological characteristics of the study cohort are presented in Table 1.

Table 1 Baseline characteristics of study population of locally advanced papillary thyroid carcinoma
Full size table

Genomic and morphological features of LAPTC

The results showed that most LAPTC (93.8%, 61/65) harbored at least one genetic alteration and BRAFV600E was the most commonly mutated gene, with a mutation prevalence of 66.2% (43/65), and all the BRAF mutations were identified as BRAFV600E mutation (exon15:c.T1799A). RAS mutation, including KRAS, NRAS and HRAS, was detected in 3.1% (2/65) cases of this cohort. In additional, 38.5% (25/65) PTC cases were found to harbor TERT promoter mutations (23 cases: C228T, 2 cases: C250T) that were the second most frequently identified mutation. Histologically, 70.8% (46/65) of the cases were tall-cell variant PTCs. Among them, 12 cases showed 5–10% of carcinoma cells with hobnail feature. 73.9% (34/46) cases harbored BRAFV600E mutations and 41.3% (19/46) cases harbored TERT promoter mutations.

Co-existent mutation of B&T was identified in 26.2% (17/65) LAPTC cases, and six cases of them were observed to be mixed variants with high proportion of tall cells and small proportion of hobnail cells (Fig. 2). Interestingly, both two cases with TERT promoter mutation (C250T) harbored BRAF mutation, one case was typical hobnail variant PTC with more than 90% hobnail cells (Fig. 3), the other was tall-cell variant PTC with 100% of tall cells. However, only one LAPTC with KRAS (exon3:c.A182G:p.Q61R) and TERT promoter mutation (C228T) was found in our series, which showed subtle nuclear features with slightly enlarged nuclear size and a few nuclear grooves but not nuclear pseudoinclusions (Fig. 4). The patient underwent neither recurrence nor metastasis.

Fig. 2
figure 2

A 52 year-old female patient harbored mutations of BRAFV600E and TERT promoter (C228T). A and B High proportion of carcinoma cells with tall-cell features infiltrates into the surrounding thyroid and fibrous tissue. C Necrosis is seen in the center of the photo. D Tumor thrombi are seen histologically in the capillary vessel

Full size image
Fig. 3
figure 3

A 60 year-old male patient harbored mutations of BRAFV600E and TERT promoter (C250T). A and B The hobnail carcinoma cells arranged in small papillary pattern invade into the surrounding thyroid tissue. C Foci of carcinoma show markedly detached carcinoma cells loss of cellular cohesiveness and polarity. D Foci of carcinoma show warthin-like features with marked inflammatory cells in the stroma of the papillae. The patient developed distant metastasis including bone and spine, and was critically ill after post-surgical 5 months

Full size image
Fig. 4
figure 4

A 58 year-old female, LAPTC with mutation of KRAS and TERT promoter. A and B The carcinoma cells are mainly arranged in follicular and solid pattern, infiltrating into the surrounding thyroid tissue. C and D LAPTC showed subtle nuclear features with slightly enlarged nuclear size and a few nuclear grooves but not nuclear pseudoinclusions. C Solid growth pattern with foci of necrosis. No mitotic figure is found. D Nuclear grooves are marked by red arrow

Full size image

The most commonly fused gene was RET, which accounted for 7.7% (5/65) and comprised NCOA4-RET fusion (Fig. 5) and CCDC6-RET fusion. ALK fusions did not exceed 1–3% of PTC, and the most common characterized fusions included STRN/ALK and EML4/ALK. Furthermore, we rarely detected mutations in PI3K/AKT pathway genes, including PTEN (1.5%, 1/65) and PIK3CA (1.5%, 1/65). In additional, we found few genetic mutations of unknown significance in thyroid cancer, including CHEK2 mutation (4.6%, 3/65), RBM10 mutation (1.5%, 1/65), ATM mutation (1.5%, 1/65), GNAS mutation (1.5%, 1/65), EGFR mutation (1.5%, 1/65), and TSHR mutation (1.5%, 1/65). As shown in Fig. 6.

Fig. 5
figure 5

A 23 year-old male patient with NCOA4-RET gene fusions and TSHR mutation. A–F The carcinoma cells mainly arranged in solid lobular and glomerular-like small papillary pattern (D yellow triangle) with pushing invasive front into the surrounding muscular tissue (A yellow triangle). The carcinoma cells are smaller than conventional variant PTC with increased number of pseudoinclusions (E red arrow). Mitosis could be found (blue arrow)

Full size image
Fig. 6
figure 6

The mutation status of the gene-panel in the study sample

Full size image

Relationship of BRAF and TERT mutations with clinicopathological features of LAPTC

To demonstrate the relationship between genetic mutations and clinicopathological features and determine the significance of genetic mutations in risk stratification. The entire cohort of the LAPTC cases was divided into three genotype subgroups, that were wild type in both BRAF and TERT (n = 14), only BRAF mutation (n = 26) and coexistent mutation of BRAF and TERT (n = 17) groups. As shown in Table 2, a significant difference in patient age at diagnosis was detected among above-mentioned three groups, which demonstrated older average age in patients with B&T co-mutation than that in patients without B&T co-mutation (58.9 ± 12.5 vs. 34.9 ± 13.7, P < 0.001). Interestingly, the probability of coexisting mutation in B&T was increased as age growth (P = 0.004, Fig. 7). Furthermore, there were significant differences on extrathyroidal invasion (P = 0.005) and advanced TNM stage (P < 0.001) in three genotype subgroups. However, we have not observed significant differences among three groups on sex (P = 0.68), max-diameter (P = 0.24), multifocality (P = 0.52), and LNM (P = 0.17).

Table 2 Relationship between genetic mutations and clinicopathological features of locally advanced papillary thyroid carcinoma
Full size table
Fig. 7
figure 7

Correlation between the probability of BRAF and TERT co-mutation and age (P = 0.0036). Risk of BRAF and TERT co-mutation is associated with increasing age. Analyses were adjusted for sex, multifocality, max-diameter, extrathyroidal extension(ETE), and central lymph node metastasis

Full size image

Comparison of clinicopathological characteristics between wild-type tumors and those tumors with BRAF and TERT co-mutations

In this study, synergetic mutations of B&T were particularly associated with high-risk clinicopathological features. As indicated in Table 3, compared with the group harboring neither BRAFV600E nor TERT promoter mutation, the group harboring B&T co-mutation had a significant association with high-risk clinicopathological characteristics. Specifically, synergetic mutations of B&T were associated with aggressive pathological subtypes including tall-cell pathological subtype (P = 0.02) and hobnail variant pathological subtype (P = 0.04), as well as extrathyroidal invasion (P = 0.008) and advanced TNM stage (P < 0.001). However, there were no significant differences in venous invasion, lymphoid vessel invasion and lateral lymphatic metastasis.

Table 3 Comparison of clinical pathological features between wild-type tumors versus those with BRAF and TERT co-mutations
Full size table

Additionally, we found the incremental effect of coexisting B&T mutations on PTC-related patient prognosis. The synergetic mutations of B&T increased the risk of tumor recurrence than BRAF mutation alone (hazard ratio [HR], 6.0; 95% confidence interval, CI 1.26–28.55, P = 0.02). Moreover, the mortality rate in B&T co-mutation group was significantly higher than that in wild-type group (3/17 versus 0/14).

Exploratory analyses of survival

After follow-up for 6–48 months (mean 24.6 months), with regard to mortality in LAPTC, the overall mortality was 17.65% (3 of 17 patients) for patients with B&T co-mutations and no patient died in LAPTC without B&T co-mutation. Patients with B&T synergetic mutations generally had a higher chance of death.

Discussion

This prospective study shows the genetic landscape of LAPTC in a single institution of China, and the relationship between genetic mutations and clinicopathological features and prognosis. We found that the incidence of common genetic mutations in LAPTC is high, with 93.8% tumors harboring at least one mutation, which providing important insights into the etiology of LAPTC. Of them, around 66.2% of the cases carried BRAFV600E mutations and 38.5% cases harbored TERT mutations, which was higher than that reported by Liu et al. (2014). It is probably due to the difference in patient selection, we aimed to study LAPTC, which was different from the non-specified PTC in Liu’s study. Specifically, the prevalence of the B&T co-mutation in the present study was 26.2%, which was higher than 12% reported by Colombo et al. (2019). Furthermore, compared with patients without B&T mutations or the BRAFV600E mutation alone, synergetic mutations in the B&T were significantly associated with aggressive clinicopathological features, such as the presence of tall-cell pathological subtype, extrathyroidal invasion and advanced TNM stage. Additionally, we found the incremental effect of coexisting B&T mutations on patient prognosis.

As we know that BRAFV600E is the most frequently genetic alteration in PTC patient. Previous studies demonstrated a significant role of BRAFV600E mutation in the development of aggressive features of PTC. Interestingly, some studies showed an association between BRAFV600E and TERT promoter mutations (Liu et al. 2014; Xing et al. 2014). The BRAFV600E mutation may up-regulate the expression of TERT by activating the MAPK pathway (Liu and Xing 2016).

TERT promoter mutations are generally found in aggressive thyroid cancer, with the frequency ranging from 4.7 to 25.5% (Liu and Xing 2016). Previous study demonstrated that co-existence of B&T mutations was strongly associated with high recurrence and poor clinicopathological outcomes of PTC (Xing et al. 2014). Xing MZ et al. claimed that the coexisting of BRAFV600E and TERT C228T mutations presented the worst clinicopathologic outcomes (Xing et al. 2014). Vuong HG et al. claimed that the combination of B&T mutations indicated increasing risk of aggressiveness of PTC than TERT or BRAFV600E mutation alone (Vuong et al. 2017). Liu et al. proposed that TERT promoter mutations in thyroid cancer are particularly prevalent in BRAFV600E mutated PTC, having a role in the dedifferentiation, progression and aggressiveness of the tumor (Liu et al. 2013). Therefore, it is essential to explore the function of genetic events as prognostic markers for risk stratification.

In the present study, we found that 18 (27.7%) LAPTCs harbored TERT promoter mutations. Among them, 17 LAPTCs harbored co-mutations of B&T and only one case harbored KRAS mutation (exon3:c.A182G:p.Q61R) and TERT promoter mutation (C228T), which showed subtle nuclear features with enlarged nuclear size and a few nuclear grooves but not nuclear pseudoinclusions. Our study suggested that synergetic mutations in the B&T are significantly associated with aggressive clinicopathological features of LAPTC compared with patients without B&T co-mutations or the BRAFV600E mutation alone. On the other hand, co-existence with RAS-like mutation and TERT promoter mutation is also molecular feature of high- grade follicular cell originated tumor, although it is very rare. LAPTC in the current study should be a group of high-grade follicular cell originated aggressive thyroid carcinoma, which has more cytological features of tall cell and hobnail cell, and more aggressive clinical behaviors.

We first found that co-mutations of B&T were detected mainly in tall-cell variant and hobnail variant pathological subtype of LAPTC. Those sub-variants were suggested to classify as aggressive PTC by the WHO classification of thyroid tumors in 2017, which indicated more aggressive clinical features than common classical PTCs, and more frequency of recurrence, distant metastasis and mortality rate (Bai et al. 2020; Lloyd et al. 2017). This also indicated that patients harboring B&T mutations were significantly associated with recurrence, stage III/IV and disease free survival. When B&T mutations were analyzed together, double-mutants compared to double wild types had a 5.85-fold higher risk of extrathyroidal invasion and a 31.2-fold higher risk of advanced TNM stage.

Additionally, we also found synergetic mutations of the B&T were significantly associated with an older age at diagnosis compared with the wild-type LAPTC group. In this study, 29.2% patients were more than 55 years old, which also validates the fact in American Joint Committee on Cancer (AJCC, 8th edition) that the age was raised from 45 to 55 years for risk stratification. This highlights the specific role of the age in the mutational event. Hence, our study also provides further support for the role of age in advanced stage of LAPTC.

Interestingly, nearly all TERT mutations were identified co-existence with BRAF mutations. The molecular mechanisms that generate co-mutation of B&T are still unclear. TERT is the catalytic component of the telomerase complex, which plays a key role in maintaining the telomere length of chromosomes and cell immortality and in controlling cellular activities (Low and Tergaonkar 2013). Study showed that mutations in the TERT promoter lead to TERT overexpression by creating an extra E26 binding motif, thus facilitating tumor growth (Vinagre et al. 2013; Huang et al. 2013). According to this study, the coexisting of B&T mutations have a synergistic effect on aggressive clinicopathological characteristics, which is likely due to the acquisition of a TERT mutation could extend the lifespan of BRAF-driven clones and enable accumulation of additional genetic defects, leading to more advanced disease (Estrada-Flórez et al. 2019).

One of the strength in this study is that we performed a prospective observational study and performed 57-gene panel next-generation sequencing to expect new discovering in landscape of genetic mutations. Importantly, we demonstrated a significance of synergetic mutations of B&T in aggressive clinicopathological features of LAPTC. On the other hand, we first found the correlation between synergetic mutants of B&T in LAPTC and aggressive histology (tall-cell pathological subtype and hobnail variant pathological subtype), while these pathological subtypes are associated with more aggressive clinical features. However, several limitations should be considered. One of the main limitations of the present study is that the small sample size and observational nature. Fortunately, LAPTC is not common, so it is not easy to obtain a big sample. Also, we could not perform the subgroup analysis of the relationship between the different mutational status and clinicopathology variables because of the small number of other genetic mutation. In addition, the mean follow-up time of 24.6 months was not long either. Thus, the findings of the present study need to be further confirmed with a larger sample size and longer follow-up.

Conclusions

This study indicated that synergetic mutations of B&T were associated with poor clinicopathological outcomes of LAPTC, including invasive and aggressive phenotypes and increased disease recurrence and mortality. The present data suggest that the detection of B&T co-mutations may be promising molecular biomarkers for predicting and identifying high-risk of patient with LAPTC, which may be also helpful to improve the risk stratification system and guide clinicians to select appropriate treatments for high‑risk PTC patients preoperatively. Further studies are required to strengthen the role of B&T co-mutations in improving the accurate identification and risk stratification in the early stage of aggressive thyroid carcinoma.