Approach to the Pediatric Patient With Thyroid Nodules

Authors: Sin-Ting Tiffany Lai, Andrew J Bauer

Abstract

Thyroid nodules in pediatrics carry a higher risk for thyroid carcinoma, and the risk of malignancy in indeterminate thyroid nodules is higher when compared to adults. Despite a substantial increase in the incidence of pediatric thyroid carcinoma since the 1990s, the 30-year disease survival remains >98% in pediatrics. The goal of thyroid nodule management focuses on reducing comorbidity of care while maintaining remission from disease and low disease-specific mortality. In 2014, the Cancer Genome Atlas introduced the concept that molecular subtypes better predict the invasive behavior of papillary thyroid carcinoma compared to pathological classification alone. Recent advancements in the understanding of the genetic landscape of pediatric thyroid carcinoma support the potential utility of incorporating somatic oncogene analysis to evaluate and manage thyroid nodules and thyroid carcinoma. In this Approach to the Pediatric Patient, we review the risk factors of differentiated thyroid carcinoma and present an integrative approach to evaluating and managing thyroid nodules in pediatric patients, incorporating sonographic and cytopathologic findings with somatic oncogene analysis. We illustrate the potential clinical application of knowledge on somatic oncogenic drivers to improve the preoperative accuracy in diagnosing differentiated thyroid carcinoma and present an integrative analysis paradigm designed to individualize the surgical approach of pediatric patients with thyroid nodules and differentiated thyroid carcinoma.

The majority of pediatric patients diagnosed with thyroid nodules present with an asymptomatic neck mass (80%) discovered by patients, their families, or a health care provider. The remaining 20% are discovered incidentally on non-thyroid-related radiological imaging (1). Local obstructive symptoms such as hoarseness, dysphagia, and difficulty breathing are uncommon. Risk factors for thyroid nodules and differentiated thyroid carcinoma (DTC) include a positive family history, chronic iodine deficiency or excess, previous exposure to ionizing radiation (medical radiotherapy for a nonthyroidal malignancy more commonly than environmental), and a history of autoimmune thyroiditis (2). The method of discovery is unrelated to the risk of malignancy, with nonpalpable incidentalomas having the same risk for malignancy as palpable lesions of the same size. However, the likelihood of malignancy in a thyroid nodule is higher in pediatrics (19%-22%) than in adults (12%-14%) (3). In pediatrics and young adults, papillary thyroid carcinoma (PTC) may present as an infiltrative process associated with diffuse enlargement of a lobe or the entire thyroid gland. Thus, if a child or adolescent patient presents with diffuse thyroid enlargement and palpable lymphadenopathy, especially in the mid to lower lateral cervical levels (III or IV) of the neck, ultrasound of the neck should be performed prior to consideration for excisional biopsy of the lymph node (2).

Mirroring the epidemiological landscape of thyroid carcinoma in adults, a substantial increase in the incidence of pediatric thyroid carcinoma has been observed since 1990. Nonetheless, 30-year disease-specific survival remains >98% in pediatric patients (4, 5). The goal of thyroid nodule management is to reduce the comorbidity of care by improving the preoperative diagnostic accuracy for malignancy with resultant individualized stratification of the surgical approach. Similar to adults, 30% to 35% of pediatric patients have indeterminate cytology on fine needle aspiration (FNA), with preoperative somatic oncogene analysis providing the only objective data to help guide surgical management. Identification of a somatic oncogenic driver predicts the risk of a thyroid neoplasm as well as the likelihood of invasive behavior of DTC (6).

In the following article, we review the risk factors of DTC and present an integrative system that incorporates sonographic, cytopathologic, and somatic oncogene data to (i) improve the preoperative diagnostic accuracy and reliability in identifying thyroid carcinoma in thyroid nodules and (ii) optimize the management of thyroid nodules based on the risk for invasive disease. In pediatrics, the risk of malignancy associated with a thyroid nodule is higher, especially for nodules classified with Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) category IV cytology (7), and the implications of our management decisions are more significant compared to those in adults. With this in mind, while our proposed integrative paradigm is supported by the literature cited throughout this article, a prospective, multicenter study would be beneficial to validate and confirm its clinical utility and applicability within the pediatric population.

To illustrate our diagnostic approach, we present 3 cases exhibiting DTCs of low, intermediate, and high risk for invasive behavior.

Clinical Cases

Case 1

A 15-year-old female patient is found to have a right neck swelling during a routine physical examination. Ultrasound confirms a 3.5-cm solid, hypoechoic intrathyroidal nodule with a wider-than-tall shape, smooth margin, and no punctate echogenic foci (Fig. 1). No abnormal lymph nodes are identified in the central or right lateral neck. FNA of the left thyroid nodule is compatible with the TBSRTC category III (atypia of unknown significance). Somatic oncogene testing reveals an NRAS Q61 (c.182A>G) mutation.

Case 2

A 16-year-old previously healthy female patient is noted by her mother to have a left neck swelling. Thyroid ultrasound identifies a 3-cm solid, very hypoechoic infiltrative nodule, wider-than-tall shape, with an irregular margin and extrathyroidal extension. Multiple punctate echogenic foci and areas of calcifications are noted (Fig. 2A). Multiple enlarged, round hypoechoic cervical lymph nodes with peripheral blood flow are identified in the right central neck (levels VI and VII) (Fig. 2B). FNA of the thyroid nodule is compatible with TBSRTC VI (malignant) cytology.

Case 3

A 16-year-old previously healthy male patient is found to have right neck swelling during a routine physical examination. Thyroid ultrasound identifies an infiltrative lesion in the right thyroid lobe with mixed echogenicity, irregular margin, extrathyroidal extension, and diffuse punctate echogenic foci extending into the isthmus and the left lobe (Fig. 3A and 3B). Multiple rounded, echogenic lymph nodes with punctate echogenic foci and peripheral blood flow are noted in right levels VI and VII (central), as well as right levels II, III, and IV (lateral) (Fig. 3C and 3D). FNA of the thyroid lesion and lateral neck lymph nodes is consistent with TBSRTC category VI (malignant) cytology.

Approach to Diagnosis

The approach to evaluating and managing thyroid nodules in pediatrics follows a similar pattern as that for adults. A dedicated thyroid ultrasound should be performed on all children with a suspected or incidental thyroid nodule with ultrasound evaluation of the cervical lymph nodes (central/levels VI and VII and lateral/levels II, III, IV, and V) for nodules with sonographic characteristics suspicious for malignancy. The purpose of the ultrasound is to stratify which nodules warrant FNA as well as to assess for the presence and the extent of lymphadenopathy to optimize the surgical plan.

Hyperfunctioning Thyroid Nodules

All patients with a thyroid nodule should have a serum thyroid-stimulating hormone (TSH) assay performed to identify an autonomously functioning thyroid nodule (AFTN). The majority of AFTN nodules have mixed solid-cystic composition and smooth margins (Fig. 4A) (8). In patients with a low or suppressed TSH, thyroid scintigraphy should be pursued rather than FNA as indeterminate cytology is common (9) and the risk for malignancy is lower compared to nodules without autonomous function (10). Increased radioisotope uptake within the nodule compared to background thyroid parenchyma confirms the diagnosis of an AFTN (Fig. 4B).

Predisposing Risk Factors of Differentiated Thyroid Carcinoma

In pediatrics, the majority (90%) of DTCs are sporadic without an identifiable etiology. Radiation exposure and family history are the most commonly identified predisposing risk factors for DTC (2).

Ionizing Radiation

Within pediatric DTC, 5% have a history of radiation exposure, and 10% of pediatric cancer survivors develop thyroid carcinoma as a second malignancy (11). Radiation exposure could be environmental or, more commonly, associated with medical therapy, to include total body irradiation or radiation directed toward craniospinal lesions, or lesions in the chest or neck (12, 13). Low-level radiation exposure of 0.05 to 0.1 Gy is associated with an increased risk of thyroid malignancy, with the risk peaking at about 18 Gy and remaining above a baseline compared to nonexposure up to 50 Gy (14, 15). Exposure at a younger age (≤10 years) is associated with an increased risk and a shorter latency for developing DTC (12, 14).

At the molecular level, radiation-induced PTCs are enriched for gene fusion events (16, 17). Radiation induces DNA double-strand breaks and results in chromosomal recombination (16). The association between radiation exposure and an increased risk of developing follicular thyroid carcinoma (FTC) is less clear.

Long-term surveillance is recommended for pediatric cancer survivors at risk for PTC (11). Patients younger than or equal to 10 years at the time of radiation exposure have the most significant risk of developing thyroid carcinoma (12). Ultrasound surveillance in these individuals can help detect PTC earlier, before regional and distant metastasis occurs (12, 18).

Family History of Thyroid DTC

Familial non-medullary thyroid carcinoma (FNMTC) accounts for 3% to 9% of all DTC (19). It is divided into syndromic (95%) and nonsyndromic forms (20, 21).

Syndromic FNMTC are cancer predisposition syndromes caused by a pathogenic germline mutation of a known susceptibility gene with increased risk for DTC and other non-thyroid disease-specific neoplastic and non-neoplastic lesions (20). The most common syndromic FNMTC include PTEN hamartoma tumor syndrome (PTEN gene), DICER1 syndrome (DICER1 gene), and familial adenomatous polyposis (APC gene) syndrome (20, 22-28) (Table 1). Syndromic FNMTC transmission is autosomal dominant with germline mosaicism only reported in PTEN hamartoma tumor syndrome (29, 30). Patients with a somatic pathogenic variant involving a gene associated with syndromic FNMTC should proceed with germline testing to optimize surveillance for neoplastic and non-neoplastic manifestation in affected probands and their families. In general, the majority of patients with syndromic FNMTC have a low risk of developing extrathyroidal metastatic disease. Patients with DICER1 syndrome are the exception, where a small percentage of patients have been diagnosed with angio-invasive FTC or poorly differentiated thyroid carcinoma (PDTC) (24, 31).

The risk for DTC is increased by 4-fold among individuals with a family history of DTC in their first-degree relatives and up to 7-fold when their siblings are affected (19). A higher prevalence of DTC may be found in individuals with 3 or more affected family members (32). Nonsyndromic FNMTC is defined by having 3 or more first-degree relatives with DTC in the absence of identified gene mutations (PTEN, DICER1, APC) (21, 33). While transmission follows an autosomal dominant pattern, a unifying susceptibility gene or locus has not been identified with recent data suggesting that nonsyndromic FNMTC may be a polygenic disorder associated with multiple genes with variable penetrance (21, 34). Compared with sporadic DTC, some but not all studies report more invasive behavior with increased multifocality, lymph node metastasis, and poor outcomes in nonsyndromic FNMTC (34-38), and genetic anticipation may be observed in the subsequent affected generation with DTC presenting at a younger age and with more advanced disease (33, 37-39). Ultrasound surveillance of siblings of affected individuals is associated with the detection of DTC of smaller size and lower incidence of lymph node metastases and extrathyroidal extension (40). However, because the extension of ultrasound surveillance to include all unaffected relatives of individuals with FNMTC will identify a high prevalence of thyroid nodules but a low prevalence of DTC (32), surveillance strategies of family members of affected individuals remain controversial.

Clinical Behavior of Differentiated Thyroid Carcinoma

Thyroid carcinoma is the most common endocrine tumor and the eighth most common malignancy in pediatrics, with an incidence of 1.1 per 100 000 in the United States (41). The incidence of thyroid carcinoma increases with age, most commonly diagnosed between 15 to 19 years of age (41). In prepubertal patients, there is no difference in incidence based on sex. Conversely, adolescents have an increased female-to-male ratio of 3 to 1 (41).

Ninety-five percent of thyroid carcinomas are derived from follicular cells, and 5% from parafollicular C-cells, termed medullary thyroid carcinoma (20). Within follicular cell–derived thyroid carcinoma, the majority (95%) are differentiated thyroid carcinoma (DTC), with PTC being the most common histologic subtype (85%), followed by FTC (10%) (20) (Fig. 5). In PTC, somatic oncogenic alterations more closely predict clinical behavior than the histological type (6, 42, 43).

The disease-specific survival of pediatric DTC is >98%, with a higher disease-specific mortality of 10% in those with distant metastasis (4, 5). The overall excellent prognosis of pediatric thyroid carcinoma is reflected by the American Joint Committee on Cancer (AJCC) TNM nomenclature, where all thyroid carcinoma diagnosed before age 55 were either stage 1 (M0) or stage 2 disease (M1) regardless of tumor and nodal classification (2).

Papillary Thyroid Carcinoma

Within pediatric PTC, there are 13 histologic subtypes with variable clinical 50% classic (C-PTC), 5% to 25% infiltrative follicular variant (IFVPTC), 10% to 15% diffuse sclerosing (DS-PTC), 1% to 15% tall cell (TC-PTC) and 1% to 2% solid/ trabecular subtype (20) (Fig. 5). The tumor is typically 1 to 2 cm in diameter (20). PTC follows a predictable pattern of metastasis, locally via the lymphatic system and distally via a hematogenous route, most commonly to the lungs. Invasive forms of PTC present an increased risk for bilateral disease (30%-40%) and regional lymph node metastasis (60%-80%) (44, 45). In the presence of lateral neck lymph node metastasis, the risk for distant metastasis increases to 15% to 20% (45). The invasiveness of PTC differs among histologic subtypes. C-PTC, IFVPTC, and DS-PTC are associated with increased risk for multifocality, extrathyroidal extension, and a ≥ 50% higher risk for lymph node metastasis (20). IFVPTC and encapsulated angio-invasive FVPTC with extensive vascular invasion (≥4 vessels) are also associated with an increased risk for distant metastasis (20, 43, 46). Conversely, invasive encapsulated follicular variant PTC (IEFVPTC) is mostly unilateral and unifocal, with a low risk for lymph node metastasis (20, 43).

Follicular Thyroid Carcinoma

FTC is typically an encapsulated tumor often larger than 2 cm (20). The diagnosis of FTC is based on the histologic identification of capsular and/or vascular invasion, which is absent in follicular adenoma (20). FTC is classified into 3 groups according to invasiveness (Fig. 5) (20). The most common group is minimally invasive FTC, where the invasion is limited to the tumor capsule, followed by encapsulated angio-invasive FTC, where the tumor invades into blood vessels within or adjacent to the tumor capsule. In widely invasive FTC, invasion extends beyond the thyroid capsule into adjacent tissues in addition to extensive vascular invasion (20). Because FTC spreads hematogenously, lymph node metastasis is rare. In angio-invasive FTC, the risk for distant metastasis correlates with the number of capsular vessels with tumor invasion. In adults, FTC with vascular invasion into ≥4 vessels within the tumor capsule increases the risk of distant metastasis (most commonly to the bone and lung) and invasion into ≥2 vessels is associated with an increased risk of recurrence (20, 47, 48).

Sequential Analysis

The 2015 American Thyroid Association (ATA) management guidelines for children with thyroid nodules and DTC present an evaluation of thyroid nodules without any detail for stratification outside of screening for autonomous function (2). In patients presenting with a thyroid nodule and a normal TSH, thyroid ultrasound is followed by FNA (2). This approach does not take advantage of the ATA adult sonographic pattern to predict the risk of malignancy (49) or the American College of Radiology Thyroid Imaging, Reporting and Data System (TIRADS) (50), 2 models that are being used with increasing frequency in clinical practice since publication of the 2015 ATA pediatric guidelines.

Sonographic Characteristics

The composition, echogenicity, shape, and margin of the thyroid nodule and the presence or absence of punctate echogenic foci should be assessed during a thyroid ultrasound. Sonographic features of a malignant thyroid nodule include solid composition, hypoechogenicity, taller-than-wide shape (transverse view), irregular margins as well as the presence of microcalcification and abnormal cervical lymph nodes (51, 52) (Figs. 1-3). If a nodule has indeterminate or high-risk features, ultrasound of the central and lateral compartments of the neck should be performed to assess for abnormal cervical lymphadenopathy. Sonographic features of metastatic lymph nodes include a round-shaped lesion with increased echogenicity, the presence of cystic areas, punctate echogenic foci, and peripheral blood flow on Doppler imaging (52, 53) (Fig. 6). PTC, including C-PTC, DS-PTC, and IFVPTC, may present with an infiltrative pattern where a lobe or the entire gland is diffusely enlarged with scattered punctate echogenic foci throughout the gland. In these situations, a discrete thyroid nodule may not be identified (Fig. 3).

The 2015 ATA adult sonographic pattern system and subsequent 2017 TIRADS (TR) are adult-based ultrasound stratification paradigms designed to help stratify the selection of nodules for FNA. Comprehensive thyroid ultrasound and accurate interpretation of the sonographic findings are essential because the decision to proceed with FNA relies heavily on the quality and interpretation of sonographic findings. Because interpretation is subjective, the diagnostic performance in evaluating thyroid nodules is influenced by institutional practice patterns, operator performance, and interpretation with significant inter- and intra-observer variability (54, 55). ATA and TIRADS appear to reliably stratify nodules in pediatric patients ultimately found to have benign nodules (TR 2/3) as well as malignant nodules (TR 5) (43, 56, 57). However, there is reduced accuracy and reliability to distinguish benign from malignant nodules with intermediate sonographic features (TR 4) (54, 55, 57, 58).

Fine Needle Aspiration and Cytology

For patients with nodules having sonographic characteristics indeterminate to high risk for malignancy, an ultrasound-guided FNA is indicated. Cytologic confirmation of sample adequacy at bedside can decrease the rate of nondiagnostic results (59). FNA of suspicious lymph nodes in the lateral neck compartments should also be performed at the same time as the thyroid nodule to determine the need for and the extent of lateral neck dissection. Thyroglobulin measured from a FNA needle washout can improve the diagnostic accuracy for malignancy in a lymph node with indeterminate cytology (60).

Cytologic findings on FNA are classified into 6 categories according to TBSRTC, with a pediatric-specific implied risk of malignancy in the 2023 version (61) (Table 2). Pediatric nodules with indeterminate; cytology (TBSRTC III, IV, and V) have an overall higher risk for malignancy compared to adults within each category however, similar to ultrasound, there is high inter- and intra-observer variability in interpreting pediatric thyroid cytology reflected by a wide range of malignancy for each indeterminate category; 0% to 50% for TBSRTC III, 20% to 100% for TBSRTC IV, and 40% to 100% for TBSRTC V (7, 62, 63). Consequently, pediatric patients with indeterminate nodules (TBSRTC III, IV, and V) have a higher surgical resection rate than the adult counterparts, even for nodules with benign cytology (64, 65). With wide variability for both pediatric ultrasound and cytology, the incorporation of an objective, preoperative diagnostic test with high reliability and accuracy to detect thyroid carcinoma is likely to reduce overtreatment of benign nodules and optimize the surgical approach for nodules at high risk for ultimately being diagnosed as a thyroid malignancy.

Genetic Landscape of Pediatric DTC

In 2014, the Cancer Genome Atlas introduced the concept that molecular subtypes better predict invasive behavior of adult PTC (42). Although the oncogenic drivers of DTC are similar in pediatrics and adults, the genetic landscape in pediatrics differs from the two-tier molecular classification in adults where RAS-like mutant diseases are associated with a low risk for invasive disease and BRAF-like mutant diseases associated with an increased risk for invasive and metastatic disease (42).

Compared with adults, pediatric PTC have a higher prevalence of kinase fusions, especially in patients ≤10 years of age (6, 66). The presence of a fusion oncogene is associated with an increased risk for extrathyroidal invasion and distant metastasis compared with RAS-like mutant or BRAF V600E mutant disease (4, 6, 67). This clinical observation leads to the proposal that the somatic oncogenic alteration in pediatrics DTC follows a 3-tier molecular classification in regard to risk for invasive (Tier 1) RAS-like oncogenes associated with low risk for invasive behavior (RAS, DICER1, PTEN, non-V600E BRAF mutations or fusions involving PAX8::PPARG) (Tier 2) BRAF V600E associated with intermediate risk for invasive behavior and (Tier 3) oncogenes associated with high risk for invasive behavior (fusions involving RET, NTRK, ALK, or BRAF) (4, 6, 67).

Integrative Analysis

Since the publication of the 2015 ATA pediatric guidelines, there has been an accumulating body of data on the potential utility of integrating ultrasound features with cytology to increase the accuracy of predicting thyroid malignancy with the addition of somatic oncogene analysis to provide objective data to help stratify the surgical approach. The majority of nodules with TBSRTC III and IV cytology have low-risk to indeterminate sonographic features (solid or complex composition with smooth margins and no punctate echogenic foci) and harbor an oncogenic alteration associated with a low risk for invasive behavior (RAS, DICER1, PTEN, non-V600E BRAF mutations or fusions involving PAX8::PPARG) (43, 46, 68) (Fig. 7). In contrast, nodules with TBSRTC V or VI cytology typically have sonographic features suspicious for malignancy (solid composition, hypoechoic echogenicity, irregular, lobulated or infiltrative margin, and punctate echogenic foci) and harbor an oncogenic alteration associated with intermediate or high risk for invasive behavior (BRAF V600E mutation or kinase fusions) (69-71) (Fig. 8). While there are ongoing challenges in access to testing as well as reimbursement of cost, preoperative somatic oncogene data are the only objective data that may be used to predict malignant histology as well as clinical behavior.

Incorporating Molecular Data to Optimize Risk Stratification According to Disease Invasiveness to Allow Individualization of Surgical Treatment

Both the 2015 ATA pediatric guidelines and the 2022 European Thyroid Association pediatric guidelines suggest that total thyroidectomy is the favored surgical approach for nearly all pediatric patients with PTC secondary to an increased incidence of bilateral disease in pediatrics and that more extensive surgery is associated with a lower rate of recurrence (2, 72). After 2015, emerging studies have consistently shown that bilateral and multifocal diseases are not as prevalent as previously described and that lobectomy with/without central neck dissection may be adequate in selective cases of PTC to achieve remission (43, 46, 68, 73-75). An individualized surgical approach with the extent of surgery driven by objective preoperative assessment to minimize complications secondary to unnecessary extensive surgery is warranted. The goal of stratification of surgery is to reduce the risks for hypoparathyroidism as well as the need for lifelong levothyroxine replacement therapy, while at the same time achieving remission. All patients undergoing thyroidectomy have some risk for permanent hypoparathyroidism. This risk is reduced when the surgeon is one who performs a high volume of thyroidectomies however, access to high-volume thyroid surgeons continues to be an ongoing challenge (76). In patients undergoing hemithyroidectomy, the risk for hypoparathyroidism is essentially eliminated, and 80% or more of patients will not require levothyroxine therapy to achieve euthyroidism (77).

Integrative Paradigm for Evaluation and Management of Pediatric Thyroid Nodules

We herein present an integrative paradigm, incorporating sonographic and cytopathologic characteristics with somatic oncogene testing to optimize risk stratification and individualization of surgical treatment. Based on current data, the summative proposed individualized management for pediatric patients with thyroid nodules (i) lobectomy without prophylactic central neck lymph node dissection in patients with TBSRTC III and IV with a somatic oncogenic alteration of low risk for invasive behavior; and (ii) total thyroidectomy with prophylactic central neck lymph node dissection in patients with TBSRTC V or VI and a somatic oncogenic alteration with an intermediate or high-risk for invasive behavior. In the latter group, a lobectomy with prophylactic central neck lymph node dissection may be considered in select patients with no evidence of lymph node metastasis on preoperative ultrasound (AJCC N0b). In adults, prophylactic central neck dissection is not common, as there is no evidence that the additional dissection improves disease-specific mortality. In pediatrics, where 99% of patients survive, the purpose of prophylactic central neck lymph node dissection is to garner data on invasive behavior in an effort to determine if completion thyroidectomy would be beneficial as well as to stratify the use of radioiodine (per the ATA initial postoperative pediatric risk levels) (2) (Fig. 9).

While the majority of patients are likely to follow the proposed integrative paradigm, there are situations where the paradigm may not apply. However, in each of these instances, evidence suggests that management should be guided by the oncogene rather than relying solely on the cytology. Specifically, these situations (i) patients with indeterminate cytology (TBSRTC III or IV) and an oncogene associated with intermediate to high-risk for invasive behavior, typically a BRAF V600E mutation; (ii) patients with malignant cytology (TBSRTC V or VI) and an oncogene associated with low risk for invasive behavior, most commonly a RAS mutation; and (iii) patients with indeterminate cytology (TBSRTC III or IV) who do not have any identified somatic alterations using a comprehensive somatic oncogene panel. Considering each situation in (i) patients should be managed based on the presence of the BRAF V600E, an oncogene with nearly 100% specificity for PTC (78); (ii) patients should be managed based on the presence of the RAS-like alteration as the likelihood for invasive behavior is lower compared to BRAF V600E as well as fusion alterations, with post-lobectomy management dictated by the presence/absence of angio-invasion and/or high-grade histological features (79). For patients with nodules that have indeterminate cytology and no identified oncogene (iii), the benign call rate and negative predictive value are higher for nodules classified as TBSRTC III than those classified as IV, approximately 70% compared to 50%, and 99% compared to 95% respectively (80-83). In pediatrics, molecular testing has demonstrated a similarly high negative predictive value; however, due to limited data, surveillance should only be considered for nodules classified as TBSRTC III with negative molecular testing using a comprehensive somatic oncogene panel (84).

Back to the Cases

Case 1

The patient underwent right thyroid lobectomy with an uneventful recovery. Histopathologic examination revealed a unifocal, minimally invasive FTC measuring 3.6 cm. Two lymph nodes were incidentally removed from right level VI (central) and were negative for malignancy. TNM staging was T2N0aM0(clinical) placing the patient in the ATA pediatric low risk for persistent post-initial surgical disease category. Postoperative thyroid function testing 6 weeks after surgery revealed a TSH in the low-mid portion of the reference range. Thyroid ultrasound surveillance at postoperative 1 year showed normal left thyroid lobe without evidence of persistent/ recurrent disease in the right thyroid bed.

Supportive Data

Nodules harboring low risk for invasive behavior oncogenes may be unifocal or multifocal (germline DICER1 and PTEN) and typically have indeterminate features on ultrasound (solid, isoechoic or hypoechoic, wider-than-tall shape on transverse imaging, with smooth margin and without punctate echogenic foci or regional lymphadenopathy) as well as indeterminate cytology (TBSRTC III or IV) (43, 46, 68) (Fig. 7). Oncogenes associated with a low risk for invasive behavior include RAS, DICER1, PTEN, non-V600E BRAF, and PAX8::PPARG fusion. Tumor pathology associated with these alterations ranges from benign follicular neoplasms, including follicular adenoma (encapsulated follicular tumor without invasion) and noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP; noninvasive encapsulated tumor with follicular growth pattern and papillary-like nuclear features), to thyroid carcinoma, most commonly IEFVPTC and minimally invasive or angio-invasive FTC (6, 43, 66, 84-86) (Fig. 7). In contrast to adults, where most IEFVPTC and FTC are associated with RAS mutations, DICER1 is the most common somatic genetic alteration in pediatrics (6, 85). Notably, however, one must be aware that DICER1 mutations and PAX8::PPARG fusion have been identified in PDTC, somatic DICER1 with thyroblastoma, and somatic PTEN mutations in anaplastic carcinoma (only in adults) (31, 87-89).

In patients with unilateral nodules and no predisposing risk factors (no history of radiation exposure or germline tumor predisposition), thyroid lobectomy should be considered as the initial surgical approach (43, 46, 68, 90). A prophylactic central neck dissection is not indicated secondary to a low risk for lymph node metastasis (68). On review of the surgical pathology, if the lesion displays low or no mitotic activity and no extrathyroidal extension, lobectomy should be adequate to achieve remission (43, 46, 68, 90). Histological subtypes where lobectomy is adequate include follicular adenoma, NIFTP, IEFVPTC, and minimally invasive FTC. If histology reveals extensive angio-invasion or high-grade histological features, completion thyroidectomy should be considered to provide an opportunity to evaluate and surveil for distant metastasis that may be present at the time of initial surgery or present 5 to 10 years after initial diagnosis (Fig. 9) Because of the paucity of data for angio-invasive FTC in pediatrics, the flow diagram (Fig. 9) suggests > 2 vessels to raise consideration and discussion as to the benefit of completion thyroidectomy. High-grade histologic features include solid, trabecular, or insular growth patterns, mitotic index ≥3 per 10 high power fields, necrosis, and convoluted nuclei (20) (Fig. 9).

Case 2

The patient underwent a total thyroidectomy with ipsilateral central neck dissection. Pathology demonstrated a 3.2-cm C-PTC without extrathyroidal extension but with lymphatic invasion, including intrathyroidal psammomatous metastasis in the right lobe (primary tumor in the left lobe). PTC was present in 9/25 central lymph nodes. AJCC TNM staging was pT2N1aM0; ATA pediatric intermediate risk for persistent post-initial surgical disease. Molecular testing revealed a BRAF V600E mutation. A diagnostic whole-body scan revealed residual activity in the thyroid bed, and single-photon emission computed tomography (CT)/CT was not sensitive enough to determine if the activity was thyroid remnant or residual lymph node metastasis. The patient received ^131^I therapy (1 mCi/Kg) with an excellent response.

Case 3

The patient underwent total thyroidectomy with central and right modified radical neck dissection. Histopathologic examination revealed a 5-cm infiltrative DS-PTC of the right thyroid lobe without extrathyroidal extension but with extensive intrathyroidal lymphatic invasion. PTC was present in 12/37 central lymph nodes and 11/43 right lateral lymph nodes. AJCC TNM staging was T3aN1b; ATA pediatric high-risk for persistent post-initial surgical disease. A ^123^I-diagnostic whole-body scan revealed several foci of activity in the neck as well as diffuse lung activity (M1). The patient received ^131^I therapy (2.5 mCi/Kg; 150 mCi). A baseline non-contrast chest CT revealed diffuse, micronodular pulmonary metastasis. One year after initial therapy, the patient has incomplete biochemical and structural response. Somatic oncogene testing from the primary tumor revealed an ETV6::NTRK3 fusion.

Supportive Data

The sonographic features of DTC associated with BRAF V600E and kinase fusions oncogenic alteration are similar, including solid composition, hypoechoic to very hypoechoic echogenicity, irregular margin or infiltrative pattern, an increased rate of extrathyroidal extension, and punctate echogenic foci (69-71) (Fig. 8). Central and lateral neck lymphadenopathy are common (6, 67, 68, 85). Cytology is typically TBSRTC category V or VI (6, 67, 68, 85) (Fig. 8). Postoperative surgical pathology typically reveals invasive features, including multifocality, extrathyroidal extension, and lymphatic invasion (6, 67, 68, 85). BRAF V600E and kinase fusions oncogenic alternation are associated with different histologic subtypes of PTC. BRAF V600E oncogenic alteration is associated with C-PTC and TC-PTC, and kinase fusion oncogenic alterations are associated with C-PTC, DS-PTC, and IFVPTC (6, 43, 66, 67, 84-86) (Fig. 8).

In patients with BRAF V600E or kinase fusion oncogenic alterations, a preoperative ultrasound to assess for unilateral/bilateral disease and central/lateral neck involvement is of utmost importance to optimize the extent of surgical dissection. In the presence of bilateral disease and/or lateral neck involvement, total thyroidectomy with therapeutic central neck dissection is warranted, with the extent of the lateral neck dissection based on the preoperative ultrasound and FNA confirmation of lateral neck compartment metastasis (74) (Fig. 9). Conversely, a more conservative surgical approach with lobectomy and prophylactic central neck dissection may be sufficient to achieve remission in patients with unilateral disease and without lateral neck lymph node involvement on preoperative ultrasound (73, 74) (Fig. 9). A prophylactic central neck dissection is recommended for all patients harboring a BRAF V600E or fusion oncogenic alteration secondary to the increased risk of central neck lymph node metastasis (68). In patients who undergo lobectomy as the initial surgical approach, completion thyroidectomy should be considered if histology reveals multifocal disease or ≥5 positive central neck lymph nodes identified on prophylactic central neck lymph node dissection (73, 75) (Fig. 9).

Conclusions

We have reviewed the risk factors of DTC and presented an integrative approach to evaluating and managing thyroid nodules in pediatric patients, incorporating sonographic and cytopathologic findings with somatic oncogene analysis. The Cancer Genome Atlas introduced the concept that molecular subtypes better predict invasive behavior of PTC in adults to improve histological classification (42). In this Approach to a Pediatric Patient review, we illustrate the potential clinical application of knowledge on somatic oncogenic drivers to improve preoperative accuracy in diagnosing DTC in pediatrics, where the range of risk for malignancy as well as the overall risk of malignancy for thyroid nodules with indeterminate cytology is higher when compared to adults. In addition, the integrative inclusion of ultrasound features with cytology and somatic oncogene has the potential to optimize the surgical approach, stratifying less extensive surgery for nodules with a low risk for invasive behavior while pursuing more extensive surgery for patients with a high risk for invasive behavior. Within this paradigm, one must be cognizant that risk for a particular behavior does not equate to the presence of low- or high-risk features. Specifically, patients who undergo lobectomy based on low-risk features (ultrasound, cytology with or without oncogene confirmation) may still benefit from completion thyroidectomy if the histology identifies features associated with an increased risk for regional or distant metastasis. Similarly, while the majority of patients with high-risk features on preoperative evaluation will likely benefit from thyroidectomy and prophylactic central neck dissection, the extent of the lateral neck dissection needs to be individualized based on preoperative ultrasound and FNA confirmation of metastasis. In addition, in selected patients with high-risk features, lobectomy with ipsilateral central neck dissection may be considered if there is no evidence of bilateral intrathyroidal disease or lateral neck lymph node metastasis.

We recognize that access to somatic oncogene testing continues to be an obstacle in pediatrics in the United States and our international community, but the objective nature of preoperative somatic oncogenic data in predicting the risk and behavior of thyroid malignancy supports wider incorporation into clinical practice as efforts are made to decrease the inter- and intra-observer variability associated with the subjective interpretation of sonographic and cytologic analysis of thyroid nodules. Continuous efforts are underway to build and expand international pediatric consortia, including in the European Union as well as the Child and Adolescent Thyroid Consortium, in an effort to fill the gaps in the current molecular knowledge and to pursue multicenter prospective studies to confirm if changes in the approach to the evaluation and management of thyroid nodules and DTC in pediatric patients is associated with improved outcome.

Disclosures

S.T.L. has nothing to declare. A.J.B. is on consulting/ advisory boards for Egetis Therapeutics and IBSA Pharma.