RET fusion genes in pediatric and adult thyroid carcinomas: cohort characteristics and prognosis

Thyroid cancer is associated with a broad range of different mutations, including RET (rearranged during transfection) fusion genes. The importance of characterizing RET fusion-positive tumors has recently increased due to the possibility of targeted treatment. The aim of this study was to identify RET fusion-positive thyroid tumors, correlate them with clinicopathological features, compare them with other mutated carcinomas, and evaluate long-term follow-up of patients. The cohort consisted of 1564 different thyroid tissue samples (including 1164 thyroid carcinoma samples) from pediatric and adult patients. Samples were analyzed for known driver mutations occurring in thyroid cancer. Negative samples were subjected to extensive RET fusion gene analyses using next-generation sequencing and real-time PCR. RET fusion genes were not detected in any low-risk neoplasm or benign thyroid tissue and were detected only in papillary thyroid carcinomas (PTCs), in 113/993 (11.4%) patients, three times more frequently in pediatric and adolescent patients (29.8%) than in adult patients (8.7%). A total of 20 types of RET fusions were identified. RET fusion-positive carcinomas were associated with aggressive tumor behavior, including high rates of lymph node (75.2%) and distant metastases (18.6%), significantly higher than in NTRK fusion, BRAF V600E and RAS-positive carcinomas. Local and distant metastases were also frequently found in patients with microcarcinomas positive for the RET fusions. ’True recurrences’ occurred rarely (2.4%) and only in adult patients. The 2-, 5-, 10-year disease-specific survival rates were 99%, 96%, and 95%, respectively. RET fusion-positive carcinomas were associated with high invasiveness and metastatic activity, but probably due to intensive treatment with low patient mortality.


Introduction
The incidence of thyroid cancer is increasing rapidly worldwide.Uncovering the genetic background of thyroid carcinomas can help determine the diagnosis, prognosis and appropriate treatment of patients.The broad spectrum of genetic alterations found in thyroid cancer includes different types of mutations: somatic point mutations, indels, copy number alterations, and fusion genes (Nikiforov 2011).
Among the fusion genes, the most common and best known are RET (rearranged during transfection) fusion genes (also known as RET/PTC rearrangements) arising from a chromosomal rearrangement between the RET gene and its partner gene.Expression of the fusion gene is driven by the promoter of the partner gene, leading to constitutive activation of the kinase domain of the RET gene.This leads to stimulation of the MAPK and PI3K pathways and oncogenesis (Arighi et al. 2005).
The CCDC6/RET (also known as RET/PTC1) fusion gene was discovered in 1987 and described 3 years later (Fusco et al. 1987, Grieco et al. 1990).It was one of the first reported genetic causes of thyroid cancer.A few years later, PRKAR1A/RET (RET/PTC2) and NCOA4/ RET (RET/PTC3) fusions were identified (Bongarzone et al. 1993, Santoro et al. 1994).RET fusion genes have been studied in association with exposure to ionizing radiation during childhood and the development of thyroid cancer.The genetic landscape of radiationinduced papillary thyroid carcinomas (PTCs) was compared with sporadic PTCs in pediatric patients.RET fusions were highly prevalent in both pediatric PTC cohorts but occurred more frequently in radiationinduced PTCs with a predominance of NCOA4/RET fusion.In sporadic cases, CCDC6/RET fusion was the most common genetic alteration in pediatric thyroid cancer (Nikiforov et al. 1997, Ricarte-Filho et al. 2013).In adult cohorts, RET fusion genes were detected less frequently than in pediatric cohorts (Santoro et al. 2020).Studies presenting long-term follow-up of patients with RET fusion-positive carcinoma are almost absent.Interest in RET fusion genes has recently increased due to the novel possibility of targeted treatment with highly potent and selective RET inhibitorsselpercatinib and pralsetinib (Markham 2020a,b).
The objectives of this study were as follows: (i) analysis of genetic alterations with emphasis on RET fusion genes in a large series of local malignant and benign thyroid diseases, (ii) characterization of RET fusion-positive cases based on clinical and pathological data, and (iii) evaluation of long-term follow-up of patients with RET fusion-positive thyroid carcinoma.

The cohort
The cohort consisted of a total of 1164 thyroid carcinomas (998 PTCs, 98 medullary thyroid carcinomas (MTCs), 24 follicular thyroid carcinomas (FTCs), 16 anaplastic thyroid carcinomas (ATCs), 11 poorly differentiated thyroid carcinomas (PDTCs), and 17 oncocytic carcinomas (OCAs)), 47 low-risk neoplasms (26 follicular tumors of uncertain malignant potential, 18 noninvasive follicular tumors with papillary-like nuclear features, three hyalinizing trabecular tumors), and 353 benign thyroid tissues (adenomas, hyperplastic nodules, chronic lymphocytic thyroiditis (CLT)) (Fig. 1).Thyroid tissue samples were collected from three Prague hospitals: the Motol University Hospital (between years 2003 and 2022), the University Hospital Kralovske Vinohrady (between years 2016 and 2022), and the Military University Hospital (between years 2019 and 2022).One thousand four hundred eightyone samples were fresh frozen thyroid tissues and 83 samples were formalin-fixed, paraffin-embedded (FFPE) thyroid tissues.Informed consent was obtained from all patients or parents/legal guardians of pediatric patients.All molecular genetic analyses were performed in the laboratory of the Institute of Endocrinology (Department of Molecular Endocrinology).The study was approved by the Ethics Committee of the Institute of Endocrinology in Prague (EK-EÚ/F192/08062020).

RET fusion gene analysis of pediatric samples
The majority of pediatric PTC samples (n = 93) was analyzed for the presence of RET fusion genes using the QIASeq Targeted RNAscan panel (Qiagen) or FusionPlex Comprehensive Thyroid and Lung (CTL) panel (ArcherDx, Boulder, CO, USA) as reported in our previous study (Pekova et al. 2020).Pediatric samples (n = 31) collected between 2020 and 2022 were analyzed using the same procedure as adult samples described below (Fig. 1).

Analysis of selected fusion genes
Carcinoma samples that were negative for a driver point mutation in the BRAF, HRAS, KRAS, NRAS, or RET genes were further analyzed for the CCDC6/ RET, NCOA4/RET, PAX8/PPARγ, and NTRK fusions as described in our previous study (Pekova et al. 2021) (Fig. 1).This selective analysis approach was chosen because these driver mutations are typically mutually exclusive with other driver mutations (Agrawal et al. 2014, Pekova et al. 2020, Parimi et al. 2023).
Libraries were prepared by the FusionPlex CTL panel (ArcherDx) using Anchored Multiplex PCR technology, which allowed detection of novel fusion genes.Briefly, total RNA (250 ng) was reverse-transcribed into cDNA using random primers.The cDNA ends were modified to allow ligation of Archer molecular barcode adapters (including i5 indexes).Two rounds of PCR followed.During the second round PCR, i7 indexes were added.Purified libraries were quantified and pooled to equimolar concentrations.The resulting libraries were subjected to paired-end sequencing on the MiSeq (Illumina).Bioinformatic analysis was performed using the Archer Analysis software version

30:12
Endocrine-Related Cancer e230117 6.2.7 (ArcherDx).For a gene fusion to be considered valid, the breakpoint had to span at least five high-quality unique reads and at least three reads had to have a unique start site.

Real-time PCR analysis
All RET fusion genes identified by the FusionPlex CTL panel were verified by real-time PCR analysis.Total RNA (2 µg) was reverse transcribed to cDNA using random primers, dNTPs, RNase inhibitor and AMV reverse transcriptase (Promega) as described in our previous study (Pekova et al. 2020).The cDNA was diluted five times and amplified using the TaqMan Fast Advanced Master Mix (Applied Biosystems) and gene-specific primers and hydrolysis probes that were synthesized based on our design (Supplementary Table 1, see section on supplementary materials given at the end of this article).The ACTB gene was used as the reference gene.Each experiment included a positive control and a negative control where RNasefree water was used instead of template cDNA.Realtime PCR was performed as follows on the Light Cycler® 480 (Roche): 50°C for 2 min, 95°C for 20 s followed by 40 cycles of 95°C for 3 s and 60°C for 30 s.
A total of 400 low-risk neoplasms and benign thyroid tissue samples were screened for all 20 types of RET fusion genes identified in carcinoma samples in this study (Fig. 1).Testing of the samples was performed using real-time PCR analysis as described above.

Statistical analysis
Categorical data are summarized as n (%).Normally distributed data are summarized as the mean (standard deviation) and non-normally distributed data as the median (range).Categorical variables were compared using the Fisher's exact test or Pearson's chi-squared test, and continuous variables were compared using the t-test.Statistical analyses were performed using the Simple Interactive Statistical Analysis and the GraphPad tools (GraphPad Software).A P < 0.05 was considered statistically significant.

Molecular characterization
RET fusion genes were found only in PTCs.No RET fusion genes were identified in other types of thyroid carcinomas, MTCs, FTCs, ATCs, PDTCs, OCAs, or in any low-risk neoplasms or benign thyroid lesions.RET fusions were detected in 113/993 (11.4%) patients with PTC.
The TERT promoter mutation was detected along with the CCDC6/RET fusion in two male patients.In two multifocal cases, the CCDC6/RET fusion gene was detected in one nodule and the BRAF V600E mutation in the other.In one unique case of a female, different multiple mutations were found in her three thyroid nodules: CCDC6/RET + NRAS Q61R in the first nodule, CCDC6/RET + KRAS G13D in the second nodule, and NRAS Q61R + PTEN Q245* in the third nodule.Her lymph node metastases (LNM) were positive only for the CCDC6/RET fusion gene.

Clinicopathological characteristics
Clinicopathological data of patients with RET fusionpositive carcinoma are summarized in Table 1 and Fig. 3.
The cohort consisted of 113 patients, 85 (75.2%) females and 28 (24.8%)males, with the mean age at diagnosis of 32.6 ± 17.4 years.Four patients had malignant disease (invasive adenocarcinoma of the lung, prostate cancer, dermatofibrosarcoma protuberans, and non-Hodgkin lymphoma) before the diagnosis of PTC.The mean tumor size was 21.8 ± 12.6 mm.Microcarcinoma (≤10 mm) had 16% of patients.In one case, only PTC metastases in lymph nodes were found in a patient without an identified primary thyroid tumor.A combination of papillary and follicular growth patterns on histology was predominant, found in 37 cases, and in a further nine cases the specimens also had a component of a solid growth pattern.Less frequent histological subtypes were diffuse sclerosing variant identified in five cases and solid and clear cell variant both in two cases.Multifocality was found in 48.2% of cases and 66.7% of these cases had numerous, in some cases up to hundreds, intrathyroidal Endocrine-Related Cancer e230117 microcarcinomas.Tumors usually showed infiltrative growth in the thyroid parenchyma and were unencapsulated.The majority (75.2%) of patients had LNM, of which 21.2% were located only in the central compartment and 78.8% in the lateral compartment, and 18.6% had evidence of distant metastases (DM), which in all cases affected the lungs, additionally in two cases the bones.The frequency of LNM and DM in association with primary tumor size is shown in Table 2. Tumor size was divided into four categories: ≤ 10 mm, 11-20 mm, 21-30 mm, and ≥31 mm, and the data were analyzed separately for adult and pediatric patients.LNM and DM also occurred frequently in patients with microcarcinoma.
Of the 113 patients with PTC positive for the RET fusion gene, 76/869 (8.7%) were adults and 37/124 (29.8%) were pediatric and adolescent patients aged 5 to 20 years.A comparison of the two cohorts is shown in Table 1.Pediatric patients had a significantly larger mean tumor size compared to adults, 26.8 mm vs 19.2 mm (P = 0.002).Other features in which pediatric patients differed significantly from adults were extrathyroidal extension (62.2% vs 33.8%; P = 0.004) and LNM (89.2% vs 68.4%; P = 0.012).

NCOA4/RET
fusion-positive carcinomas were compared with carcinomas harboring other RET fusions (Table 3).PTCs positive for NCOA4/RET were significantly associated with larger tumor size (P = 0.021) and more frequently had extrathyroidal extension (P = 0.049), in contrast to carcinomas positive for other RET fusions, which were significantly associated with a predominant papillary growth pattern (P = 0.001) and CLT (P = 0.005).Patients with PTC positive for NCOA4/RET had more frequent LNM and DM than patients with other RET fusions; however, the differences only approached the threshold of statistical significance (P = 0.074 and P = 0.053, respectively).
Comparison of clinicopathological features between PTCs positive for RET fusions and individual groups of PTCs with NTRK fusions, BRAF V600E and RAS mutations was performed (Table 4).Only patients with available clinicopathological data were included.Patients with RET fusion-positive PTCs were significantly younger than patients with BRAF V600E and RAS-mutated carcinomas (P < 0.001) and had significantly larger tumor than patients with the BRAF V600E-mutated carcinomas (P < 0.001).In RET fusion-positive PTCs, features associated with higher

Treatment and follow-up
Total thyroidectomy (TTE) was performed in all patients except one who underwent hemithyroidectomy (HTE) and whose follow-up data were not available.Two patients underwent first HTE followed by completion to TTE.Lymph node dissection for suspected LNM was performed in 88 (77.9%) patients.Twenty (17.7%) patients had reoperation, including 19 patients for LNMand one patient for elimination of thyroid remnants.
One hundred and four (92%) patients received radioactive iodine (RAI) treatment.Six patients (five adults, one pediatric patient) were not indicated for RAI therapy due to small tumor size (≤12 mm) and the absence of metastases.One pediatric patient died due to late surgical treatment and massive lung metastases before RAI treatment.Follow-up data of two patients were not available.Approximately half of the patients (55% of adult and 46% of pediatric patients) received one therapeutic dose of RAI, 20% of adults and 19% of pediatric patients received two doses, and 18% of adult and 30% of pediatric patients received three or more doses.The median activity administrated per therapeutic session was 120 mCi (range: 50-200 mCi) equally in adult and pediatric patients.In addition to RAI therapy, two patients with RAI-refractory PTC were treated with radiotherapy and one of them also with chemotherapy.One patient was on targeted treatment with the RET inhibitor selpercatinib.The patient's tumor infiltrated the respiratory tract, the lesion was inoperable and the patient also had lung metastases.The patient responded to treatment; he had a partial remission.Median follow-up was 75 months (range: 4-228 months).Response to treatment was determined based on the definitions and criteria in the 2015 American Thyroid Association Guidelines as an excellent response (ER), an indeterminate response (IR), a biochemical incomplete response (BIR), a structural incomplete response (SIR) (Haugen et al. 2015) or death.Response to treatment was assessed in the following time periods after surgery: 0.5-2 years, 2-5 years, 5-10 years, years, and 15-19 years.Detailed follow-up data describing the course of treatment for individual patients are shown in Fig. 3. Follow-up data were unknown for two patients and four patients were not evaluated due to short-term follow-up (less than 6 months after surgery).
During the first 2 years after surgery, the number of patients who had ER and SIR to treatment was similar.Later, the patient's response to treatment improved, although only slightly from the fifth year after surgery.Forty-one (38.3%) patients had ER to treatment during the entire follow-up.Two adult patients (2.4%) had a 'true recurrence', a relapse after achieving structural and biochemical disease-free status.LNM were identified in the first case 4 years after surgery and in the second case 7 years after surgery.Four adult patients (4.8%) had SIR after IR/BIR.In three cases, the evidence of metastases (2× LNM, 1× DM) was found around 10 years after surgery and in one case of DM 5 years after surgery.Twelve patients (14.6%, an equal proportion of pediatric and adult patients) had ongoing SIR more than 2 years after surgery.Three patients (2.7%) died and all of them due to PTC.All were males, two adults and one pediatric patient.Two patients had PTC positive for NCOA4/RET and one patient for CCDC6/RET along with the TERT C228T mutation.The 2-, 5-, 10-year disease-specific survival rates were 99%, 96%, and 95%, respectively, for RET fusion-positive patients.
According to published definitions, nine (8.7%) patients had RAI-refractory PTC (Tuttle et al. 2019).Among these patients were both adults and children, and most were males.Six carcinomas were RAIrefractory from the start of RAI treatment, and in three cases the refractoriness developed over time.RAI-refractory LNM were in seven cases, of which two patients also had RAI-refractory lung metastases and two patients had only RAI-refractory lung metastases.
Response to treatment was compared between pediatric and adult patients (Fig. 4A).In the first 2 years after surgery, a significantly lower number of pediatric patients than adult patients had an ER to

30:12
Endocrine-Related Cancer e230117 treatment (P = 0.006), probably due to more advanced disease in pediatric patients.Later, improvement was demonstrable in both cohorts and response to treatment became similar in groups.Ten years after surgery, more pediatric than adult patients had an ER to treatment due to recurrences in the adult group.
Response to treatment was also compared between RET fusion and NTRK fusion-positive patients (Fig. 4B).Data on NTRK fusion-positive patients were used from our previous study (Pekova et al. 2021).In the first 2 years after surgery, a significantly lower percentage of RET fusion-positive patients than NTRK fusion-positive patients had an ER to treatment (P = 0.023).Ten years after surgery, no NTRK fusion-positive patient had SIR to treatment, in contrast to 19% of RET fusionpositive patients.

Discussion
Thyroid cancer is associated with a broad range of different mutations, including RET fusion genes.Previously, the frequency of RET fusion genes included only the most common CCDC6/RET and NCOA4/ RET fusions, and other RET fusions were rarely tested.To our knowledge, this is one of the most comprehensive studies of a RET fusion-positive cohort of thyroid carcinomas that has been analyzed and correlated with clinicopathological and follow-up data.
The frequency of RET fusion genes in our adult PTC cohort was 8.7%, which is in the range of 4.3-9.1% as previously reported (Agrawal et al. 2014, Lu et al. 2017, Liang et al. 2018, Yang et al. 2021, Shi et al. 2022, Ullmann et al. 2022, Wang et al. 2022, Parimi et al. 2023).Pediatric and adolescent PTCs harbored RET rearrangements more than three times more frequently than adult PTCs, in 29.8% of cases.This result is within the range of 10.4-57.7%previously published results (Ricarte-Filho et al. 2013, Galuppini et al. 2019, Pekova et al. 2020, Macerola et al. 2021, Rogounovitch et al. 2021, Franco et al. 2022, Satapathy & Bal 2022).The variability is likely due to the range of methods used, the size and composition of cohorts and geographic location.
Next-generation sequencing (NGS) is currently the most reliable method for detection of RET fusion genes.The use of a comprehensive panel for the detection of fusion genes should be the gold standard, especially in pediatric thyroid carcinomas, in which the occurrence of fusions prevails.A more cost-effective approach using mutational exclusivity of RET fusions with other driver mutations could be applied in adult thyroid carcinomas.First, detection of point driver mutations in the BRAF and RAS genes would be performed.Negative samples would then be analyzed for the most common CCDC6/RET and NCOA4/RET fusions, representing approx.Eighty percent of all RET fusions by real-time PCR.Negative samples would undergo NGS testing, which would also allow the detection of novel fusion partner genes.To date, at least 50 RET partner genes have been described in thyroid cancer, and the number has been rising steeply in recent years (Agrawal et al. 2014, Grubbs et al. 2015, Yakushina et al. 2017, Pekova et al. 2020, Lee et al. 2021, Rogounovitch et al. 2021, Yang et al. 2021, Franco et al. 2022, Wang et al. 2022).
In this study, RET fusion genes were detected only in PTCs.A few rare cases of RET rearrangements in PDTCs, in eight ATCs, and in two MTCs have been described in the literature (Grubbs et al. 2015, Landa et al. 2016, Yang et al. 2021, Parimi et al. 2023).There are also studies in which RET fusions have been found in benign samples, particularly in Hashimoto's thyroiditis (Wirtschafter et al. 1997, Sheils et al. 2000, Elisei et al. 2001).Most of these studies were performed around year 2000, when the methods used had substantial technical limitations leading to a high risk of falsepositive results (Nikiforov 2006).Although no recent case of RET fusion in a benign thyroid sample has been reported, their presence in these tissues has been questioned (Nikiforova et al. 2002), the occurrence of RET fusions in Hashimoto's thyroiditis has been reported in recent review articles (Rangel-Pozzo et al. 2020, Nacchio et al. 2022).In our study, no RET fusion was detected in any of the benign samples.Based on these results we conclude that the identification of an oncogenic RET fusion is associated with a 100% probability of malignancy.
Coexistence of CLT with RET fusion-positive PTC was found in 65.2% of our cases, similarly in pediatric and adult patients.The frequency of this coexistence is higher than the range of 15-40% reported in the literature for PTC cohorts with unknown genetic background (Kim et al. 2009, Babli et al. 2018, Lee et al. 2020).In another study, the authors found a negative correlation between the coexistence of CLT with PTC and the BRAF mutation (Lee et al. 2020).In addition, the presence of numerous psammoma bodies was found in our study.The association between them and RET fusions has been previously described (Adeniran et al. 2006).Thus, the presence of coexisting CLT as well https://doi.org/10.1530/ERC-23-0117

30:12
Endocrine-Related Cancer e230117 as numerous psammoma bodies could indicate a higher probability of RET fusion gene in PTC.
Other examined clinicopathological features that were abundantly identified in RET fusion-positive PTCs multifocality, extrathyroidal extension, intravascular invasion, LNM, and DM.RET fusionpositive carcinomas were associated with aggressive and invasive behavior.This correlation was also noticed by the authors of studies on small RET fusionpositive cohorts (Franco et al. 2022, Ullmann et al. 2022, Wang et al. 2022).
Multifocality was detected in 48.2% of RET fusion-positive carcinomas, which was significantly higher than in our BRAF V600E and RAS-mutated PTCs.In another study, multifocality was even found in 86% o f RET fusion-positive PTCs (Ullmann et al. 2022).Most of our patients with multifocal PTC had numerous millimeter-sized intrathyroidal microcarcinomas.The frequency of these cases may have been even higher due to several cases, in which the lobes were filled with carcinomas and may have been formed by the union of multiple foci.The sensitivity of ultrasound for these clinically occult lesions is low.Thus, TTE should be considered in the case of preoperative detection of an oncogenic RET fusion gene in a thyroid nodule.
RET-fusion positive patients had a significantly higher incidence of LNM and DM compared to patients with NTRK fusions, BRAF V600E or RAS mutations.In RETfusion positive patients, LNM were identified in 68.4% of adult and 89.2% of pediatric patients.Other studies reached similar results, in which 91-94% of pediatric patients with RET fusion-positive PTCs had LNM (Lee et al. 2021, Franco et al. 2022).In our cohort, the majority of patients had evident multiple LNM in the lateral compartment.The correlation of RET fusion-positive PTCs with LNM in the lateral compartment was also noted by the authors of other studies (Rogounovitch et al. 2021, Wang et al. 2022).Therefore, sonographic evaluation of the lymph nodes in the central and lateral compartments should be performed conscientiously due to the high probability of LNM regardless of tumor size.
Differences were also observed within the RET fusion gene cohort.PTCs positive for NCOA4/RET had significantly larger tumors and more frequent extrathyroidal extension.In addition, PTCs positive for the NCOA4/RET fusion had a tendency to be associated with more LNM and DM, although this finding was not statistically significant.The correlation between the NCOA4/RET fusion and higher aggressiveness has been mentioned in other studies (Galuppini et al. 2019, Rogounovitch et al. 2021).
Follow-up data on patients with RET fusionpositive PTC are very scarce in the literature.In one study, 50% of RET fusion-positive pediatric patients achieved remission one year after surgery (Franco et al. 2022), a high percentage considering that 94% of patients had LNM and 35% had DM.In another study, 24% and 19% of RET fusion-positive pediatric patients had ER and IR/BIR to treatment, respectively, throughout follow-up.SIR to treatment had 57% of patients, of whom 14% achieved ER or biochemical disease without structural evidence of disease at last follow-up (Lee et al. 2021).Compared to our study, a similar proportion of patients had SIR to treatment, but significantly more of our pediatric patients achieved clinical improvement.An explanation could be more advanced disease of their patients, 38% vs our 25% patients with DM, and the shorter follow-up in their study (Lee et al. 2021).Follow-up data on adult RET fusion-positive patients are even scarcer.In one study, none of the RET fusion-positive patients relapsed and 2/14 (14%) patients had persistent disease (Ullmann et al. 2022).
Only 8.7% of our patients had RAI-refractory carcinoma, which is within the reported range of 5-15% for differentiated thyroid carcinomas (Worden 2014).Only one patient in our cohort had treatment with selpercatinib.Promising results of this type of treatment from other studies have already been published (Lee et al. 2021).
Despite more aggressive disease of the RET-fusion positive cohort, the 2.4% 'true recurrence', 2.7% cancerspecific mortality, 96% and 95% survival at 5 and 10 years were comparable to data (approximately 1% 'true recurrence', 2-2.5% mortality, 99% and 95% survival at 5 and 10 years) for PTC cohorts with unknown genetic background (Sciuto et al. 2009, Ito et al. 2018).It is important to note that the majority of patients in our cohort had intensive treatment, relatively radical surgery and many patients underwent repeated RAI therapy.In addition, the patients' condition was regularly monitored over a long period of time, which is essential for a good outcome, as recurrence can occur many years after surgery.It is questionable what the response to treatment would have been with a milder treatment (e.g.lobectomy and no RAI treatment), whether or not the recurrence and mortality rates would have been significantly higher.

Figure 3
Figure 3 Tile plot of patients with RET fusion-positive carcinoma, their clinicopathological and follow-up data.Clinical and pathological data such as gender, age at diagnosis, tumor size, lymph node metastases, and distant metastases are shown.Follow-up data display patient´s response to treatment and cases of radioiodine-refractory carcinomas.LNM, lymph node metastases; DM, distant metastases; RAI-R, radioactive iodine-refractory.

Figure 4
Figure 4Graphical representation of the responses to treatment during follow-up periods (A) in adult and pediatric and adolescent patients with RET fusion-positive carcinomas (B) in all RET fusion-positive carcinomas and NTRK fusionpositive carcinomas.The numbers in the columns indicate the number of patients.ER, excellent response; IR, indeterminate response; BIR, biochemical incomplete response; SIR, structural incomplete response.

Table 1
Clinical and pathological features of patients harboring RET fusion-positive carcinomas.

Table 2
Frequency of LNM and DM in relation to primary tumor size in adult and pediatric and adolescent patients with RET fusion-positive carcinoma.

Table 3
Comparison of clinical and pathological features between NCOA4/RET and other RET fusion-positive PTCs.

Table 4
Comparison of clinicopathological features between PTCs positive for RET and NTRK fusions, BRAF V600E and RAS mutations.