Management Guidelines for Children with Thyroid Nodules and Differentiated Thyroid Cancer

[B2] How Common are Thyroid Nodules in Children and What is the Risk for Malignancy?

Thyroid nodules are less common among children than adults but are more likely to be malignant in children referred for evaluation of nodular thyroid disease (22%–26% versus approximately 5%) (27,55,56). Estimates from ultrasound (US) and postmortem examination suggest that 1%–1.5% of children and up to 13% of older adolescents or young adults have thyroid nodules (73,74), although it is unclear how many of these would have become clinically apparent. Recent data from a large Japanese series using high-resolution US confirm the incidence of solid nodules at 1.65% but also identified cystic lesions in 57% of children and adolescents (75). Such high-resolution US data have not yet been replicated in other pediatric populations, and it remains unclear if thyroid nodules are this prevalent in other regions. Nevertheless, it appears from multiple studies that the prevalence of thyroid nodules is much greater in children than is generally appreciated. It also remains unclear how many of these nodules would reach a clinical threshold during childhood.

[B3] Are there High-Risk Groups Who Might Benefit from Prospective Screening for Thyroid Nodules and Thyroid Cancer?

Several risk factors are associated with the development of thyroid nodules in children, including iodine deficiency, prior radiation exposure, a history of antecedent thyroid disease, and several genetic syndromes (Table 4). One high-risk population is that of childhood cancer survivors who were treated for their primary malignancy with radiation therapy, especially survivors of Hodgkin lymphoma, leukemia, and central nervous system tumors (76,77). Thyroid nodules, many of which can only be detected by US, develop in cancer survivors at a rate of about 2% annually and reach a peak incidence 15–25 years after exposure (78 –80). In general the risk is greatest among those who received radiation therapy at a younger age and with doses up to 20–29 Gy (77,81,82). High resolution US may identify small subclinical thyroid tumors (83,84). However, insufficient data exist to determine if early detection of nonpalpable tumors will significantly improve the quality and or longevity of life in patients screened by a standardized protocol using US and fine-needle aspiration (FNA). Furthermore, routine US screening may also identify incidental findings, such as ectopic thymus, that may confuse the clinical picture and potentially lead to unnecessary testing (75).

A variety of genetic disorders predispose to thyroid neoplasia (85,86) (Table 4). Benign and malignant thyroid tumors can occur in patients with APC-associated polyposis (87), the Carney complex (88), the DICER1 syndrome (89,90), the PTEN hamartoma tumor syndrome (91 –93), and Werner syndrome (94). Cases of DTC have also been reported in Beckwith–Wiedemann syndrome (95), the familial paraganglioma syndromes (96), Li–Fraumeni Syndrome (97), McCune–Albright syndrome (98), and Peutz–Jeghers syndrome (99).

Furthermore, children from kindreds with familial nonmedullary thyroid cancer (FNMTC) may have a predisposition to tumor development (100 –105). No clear recommendations exist for prospective screening (outside of routine physical examination) in most cases. However, updated recommendations for US screening have been put forth for both the PTEN hamartoma tumor syndrome and APC-associated polyposis (91,92,106). In addition, in nonsyndromic FNMTC, US surveillance of family members has been shown to detect earlier stages of disease as reflected by smaller tumor size (0.8 vs. 2.85 cm; p<0.001), a lower incidence of lymph node metastasis (23.2% vs. 65.6%; p<0.001) as well as a lower incidence of ETE (20.9% vs. 56.2%; p=0.002) compared to the proband (107).

Limited data exist on children with autoimmune thyroiditis. However, one report shows an increased prevalence of thyroid nodules perhaps as high as 30% with 7 of 11 PTC only detected by US examination (108). It is unclear how many of these would have achieved clinical importance, however. The presence of a palpable thyroid nodule or asymmetry with or without palpable cervical lymphadenopathy warrants referral to an experienced thyroid ultrasonographer and consideration of FNA as indicated based on suspicious sonographic features (see Section B4) or growth over time. There are increasing data to suggest that patients with a nodule and thyrotropin (TSH) levels in the upper tertiles of the reference range may be at increased risk for malignancy (109).

From these data we conclude that thyroid nodules are common in childhood cancer survivors who received radiation therapy, and they are associated with a modest risk of malignancy. Other groups of children with tumor syndromes, as well as those born into a kindred with FNMTC, have an increased risk for thyroid nodules and/or cancers. Some of these cancers are small and not likely to be detected without US. Although this task force could not recommend thyroid US as a routine screening tool in all of these patients, we do encounter children who have incidental nodules identified via screening thyroid US. Similar to palpable nodules, nodules detected in this setting should be interrogated by US performed by an experienced ultrasonographer, and FNA should be performed if the nodule has concerning sonographic features or growth over time.

[B4] What is the Optimal Evaluation of Children with Thyroid Nodules?

The 2009 ATA adult guidelines indicate that the evaluation and treatment of thyroid nodules in children should be the same as in adults (Recommendation 18). In general, this task force agrees with that sentiment, but there are specific areas in which we feel the approach should differ (Fig. 1).

OPEN IN VIEWER

FIG. 1.

Initial evaluation, treatment, and follow-up of the pediatric thyroid nodule. ¹Assumes a solid or partially cystic nodule ≥1 cm or a nodule with concerning ultrasonographic features in a patient without personal risk factors for thyroid malignancy (see Sections B3 and B4). ²A suppressed TSH indicates a value below the lower limits of normal. ³Refer to PTC management guidelines (Section C1) or MTC management guidelines. ⁴Surgery can always be considered based upon suspicious ultrasound findings, concerning clinical presentation, nodule size >4 cm, compressive symptoms, and/or patient/family preference. ⁵Surgery implies lobectomy plus isthmusectomy in most cases. Surgery may be deferred in patients with an autonomous nodule and subclinical hyperthyroidism, but FNA should be considered if the nodule has features suspicious for PTC. (See Section B10.) Consider intraoperative frozen section for indeterminate and suspicious lesions. Can consider total thyroidectomy for nodules suspicious for malignancy on FNA. ⁶Consider completion thyroidectomy ± RAI versus observation ± TSH suppression based upon final pathology (see Section E1).

The 2009 adult guidelines indicate that FNA is not warranted for the evaluation of a nodule <1 cm in size unless the patient is considered high-risk, most commonly with a history of exposure to ionizing radiation, or the nodule is associated with pathologic regional lymph nodes. A size criterion is more problematic in children because thyroid volume changes with age and the size of the nodule alone does not predict malignant histology (110 –112). Therefore, US characteristics and clinical context should be used more preferentially to identify nodules that warrant FNA. US features such as hypoechogenicity, irregular margins, and increased intranodular blood flow are more common in malignant lesions (110,113). In addition, the presence of microcalcifications and abnormal cervical lymph nodes increase the likelihood of malignancy (110,113). In all children with a suspicious nodule, US evaluation of the cervical lymph nodes should be performed.

The 2009 adult guidelines indicate that FNA is not warranted for the evaluation of a hyperfunctioning nodule in the adult. Although we concur that preoperative FNA of a hyperfunctioning nodule in a child is similarly not warranted, this is based on the understanding that all hyperfunctioning nodules in children will be surgically removed (see Section B10).

The 2009 adult guidelines indicate that calcitonin screening for MTC in adults with thyroid nodules may be cost effective, but it was neither recommended for nor against. In children and adolescents, the prevalence of sporadic MTC is extremely low. In addition, calcitonin reference ranges in children have not yet been widely validated, especially in children who have background thyroid disease such as thyroiditis. Further studies are needed to determine the cost-effectiveness of adding calcitonin to the evaluation of thyroid nodules in children.

The 2009 adult guidelines indicate that US-guided FNA is preferred for lesions with a higher likelihood of nondiagnostic cytology or sampling error. The sensitivity, specificity, and overall accuracy of FNA in children are similar to that of adults (114 –119). However, based on the higher proportion of malignant nodules in children and the potential difficulty in obtaining repeat samples from children, this task force recommends that all FNA in children should be performed with US guidance. This is particularly relevant for complex cystic lesions, which require FNA of the solid portion, and it may also reduce the need for repeat FNA. The latter is important since FNA may alter the ultrasonographic features of thyroid nodules (120), thus making short-term follow-up more difficult.

A unique but very important difference in children is that PTC may present as diffusely infiltrating disease that results in diffuse enlargement of a lobe or the entire gland. For this reason, diffuse thyroid enlargement, especially if associated with palpable cervical lymph nodes, should prompt imaging. With rare exception, the diffuse infiltrating form of PTC is associated with microcalcifications that warrant FNA.

Finally, for both children and adults, cytopathology findings on FNA are categorized according to The Bethesda System for Reporting Thyroid Cytopathology (121). In this six-tier system, FNA results are reported as (a) nondiagnostic or unsatisfactory, (b) benign, (c) atypia or follicular lesion of undetermined significance (AUS/FLUS), (d) follicular/Hürthle neoplasm or suspicious for follicular/Hürthle neoplasm, (e) suggestive of malignancy, or (f) malignant. Insufficient or nondiagnostic cytopathology refers to a specimen with limited cellularity (fewer than six follicular cell groups each containing 10–15 cells per group from at least two separate aspirates), absence of follicular cells or poor fixation and preservation (122). There is a 1%–4% risk of malignancy in insufficient samples from adults (121), but very few data in children. Repeat FNA is an option in children but should be delayed for a minimum of 3 months in order to decrease the potential for atypical cellular features that may arise during the reparative process (123). In adults, the risk of malignancy in indeterminate nodules ranges from ∼5% to 15% in the AUS/FLUS category to 15%–30% in the follicular neoplasm or suggestive of neoplasm group (122). The limited data available suggest these indeterminate FNA categories account for ∼35% of pediatric FNA and that, in children, 28% of AUS/FLUS lesions and 58% of suggestive of follicular or Hürthle cell neoplasm are malignant (26,124). The 2009 adult guidelines suggested that repeat FNA was an option for adults with indeterminate cytopathology. However, due to the apparent increased probability of malignancy among these indeterminate categories in children, the task force recommends definitive surgery (lobectomy plus isthmusectomy) for indeterminate FNA findings in children (see Fig. 1).

[B5] Are There Molecular Signatures That Complement FNA and Improve the Diagnostic Utility of FNA in Children?

Studies in adults have shown that molecular testing aids in the management of thyroid nodules with indeterminate cytopathology (125 –130). However, these diagnostic approaches have not yet been validated in pediatric patients. Mutational analysis has been used to examine thyroid nodules in children in limited single institution studies (26,131). Approximately 17% of pediatric FNAs may be positive for a mutation or rearrangement, the presence of which correlated with malignancy in 100% (26). However, the cytopathologic classification for these malignant tumors were AUS/FLUS, suggestive of follicular or Hürthle neoplasm, suggestive of malignancy, or malignant, all of which would have led to surgical removal regardless of the mutational analysis. Although a proprietary multigene expression classifier has been validated to corroborate a benign diagnosis in adults with indeterminate nodules (126), there are no studies determining its usefulness in the evaluation of the indeterminate pediatric thyroid nodule. Therefore, although current molecular diagnostics might improve the diagnostic acumen for indeterminate cytopathology in children, additional studies are required before a formal recommendation can be proffered.

[B6] How Should Thyroid Nodules Be Treated in Children?

The surgical approach to the child with a thyroid nodule is dictated by the FNA results (see Fig. 1). Every effort should be made to ensure the FNA is performed in a controlled setting designed to accommodate age-appropriate anesthesia and pediatric advanced life support monitoring and intervention. In an effort to provide clarity, the proposed classification scheme from the National Cancer Institute Thyroid FNA State of Science conference is used as a guide to stratify surgical intervention (122).

[B7] What Is the Recommended Approach for Children with Benign Thyroid Cytopathology?

A key element in this question is whether or not “benign” lesions in children as defined by absence of suspicious US findings and benign FNA are ever subsequently found to be malignant. There are insufficient data to answer this question in children, but there are studies that have included both children and adults (132 –134). The false-negative rate appears to be quite low, in the range of 3%–5% (114); however, the false negative rate may be higher in larger lesions secondary to an increased risk of sampling error (27,135 –137).

[B8] Is There a Role for Levothyroxine Suppression Therapy?

The literature in this area is conflicting. Not all studies have used the same methodology nor have they always separated spontaneous thyroid nodules from radiation-induced thyroid nodules. Furthermore, some but not all benign thyroid nodules regress spontaneously, and this might be more common in small cystic lesions (138,139). Levothyroxine (LT₄) suppression therapy has been evaluated for its efficacy to reduce nodule size or to reduce the risk of subsequent nodule formation. However, there are only minimal data regarding long-term safety and potential side-effects of LT₄ therapy (140,141).

LT₄ therapy has been prescribed to reduce the size of benign thyroid nodules, but the clinical benefit of a small to modest reduction in size is not clear (142 –147). About a third (30.6%) of euthyroid children had a ≥50% reduction in nodule size, which was directly correlated with TSH levels (r=0.640, p<0.001) and inversely with LT₄ dose (r=−0.389, p=0.009) (140).

Thyroid hormone has also been used in pediatric patients with radiation-induced thyroid nodules in which the formation of subsequent nodules has been shown to be reduced (148,149). It is not clear if this data can be extrapolated to pediatric patients with spontaneous nodules, and LT₄ therapy had no effect on the incidence of thyroid cancer (148).

Whether LT₄ therapy is used or not, an increase in nodule size is more commonly associated with malignant disease and should prompt re-evaluation and/or surgical resection (see Section B9). Alternatives to surgery have been evaluated in adults, but they have not yet been evaluated in children and their use cannot be recommended.

[B9] Is There a Role for Surgery in Children with Benign Nodules?

For the subset of children who have benign cytopathology, surgery may be considered due to increasing size, compressive symptoms, cosmetic reasons, or patient/parent choice. For growing nodules (defined in adults as a ≥50% increase in volume or ≥20% increase in at least two dimensions) or nodules that have developed suspicious US characteristics, repeat FNA should be performed prior to surgery to assist with surgical planning and preoperative staging. FNA of nodules >4 cm appears to have decreased sensitivity for the diagnosis of malignancy (27,135 –137). Given the high false-negative rate of FNA in large lesions, and also to simplify long-term follow-up, surgery should be considered for FNA-documented benign nodules >4 cm, especially if they are solid. If surgery is undertaken, lobectomy is preferred to minimize the risk for complications.

[B10] What Is the Optimal Management of the Child with an Autonomous Thyroid Nodule?

Pediatric patients are occasionally found to have an autonomously functioning nodule (toxic adenoma) diagnosed by a suppressed TSH and increased, nodule-specific uptake on nuclear medicine radioisotope scan (^99mTc pertechnetate or iodine-123 [¹²³I]) (150,151). These lesions are most frequently associated with somatic activating mutations within the genes encoding the TSH receptor or the G_s-alpha subunit (151). On examination, children are either euthyroid or may have mild signs or symptoms of hyperthyroidism.

In adults, the treatment options for autonomous nodules include ¹³¹I ablation, surgical resection, or ethanol injection. Because of concerns of the mutagenic effect of low-activity radioiodine on the normal thyroid tissue, and reports that up to one third of patients may be found to have an incidentally discovered DTC associated with autonomous nodules (150), surgical resection is the usual recommendation for most pediatric patients because the safety of observation or alternative treatments is unstudied in children. However, in asymptomatic patients with an autonomous nodule and subclinical hyperthyroidism, surgery may be deferred, but FNA should be considered if the nodule has features suggestive of PTC.

[C2] What Is the Optimal Preoperative Evaluation for the Child with Newly Diagnosed PTC?

The preoperative evaluation of the newly diagnosed pediatric PTC patient is critical for optimizing surgical outcome and medical therapy. In all cases, a comprehensive neck US using a high-resolution probe (7.5 MHz or higher) and Doppler technique should be obtained by an experienced ultrasonographer. All regions of the neck should be interrogated, recognizing that US has decreased sensitivity to identify malignant lymphadenopathy in the central neck (level VI) (152,153). A complete US examination should be performed prior to surgery if it was not performed with the FNA. The goal is to identify locoregional metastatic disease otherwise not appreciated on physical examination (154 –157).

Given the very high rate of cervical lymph node metastases in children with PTC, the preoperative identification of suspicious lymph nodes affords the surgeon an opportunity to more thoughtfully plan comprehensive, compartment-oriented, lymph node dissection during the initial surgery with the intent to decrease recurrence rates and the need for additional surgery (154,157). In patients with large or fixed thyroid masses or bulky metastatic lymphadenopathy, US may be less sensitive at detecting metastatic disease to deep tissue regions, such as the superior mediastinum (level VII), the retropharyngeal, parapharyngeal, and subclavicular spaces (152,153). The addition of cross-sectional imaging using contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI), depending on local expertise and preference, should be considered, especially if there is any concern for invasion of the aero-digestive tract (158 –161). If iodinated contrast agents are used, further evaluation and treatment with RAI may need to be delayed for 2–3 months until total body iodine burden decreases. The advantage of CT over MRI is that CT has shorter image acquisition times, making scheduling more accessible and reducing the need for conscious sedation in younger patients.

A chest x-ray and/or chest CT may be considered in those with substantial cervical lymph node disease, in whom the prevalence of lung metastases is increased (9,10,24,65,162,163). Pediatric-specific protocols should be used to minimize ionizing radiation exposure. Routine chest CT is not recommended for patients with minimal neck disease because pulmonary metastases are likely to be identified when the child is subsequently staged with a stimulated Tg and diagnostic whole-body scan (DxWBS) (see Section C8) (24,49,164 –166).

Thyroid nuclear scintigraphy for the evaluation of newly diagnosed PTC should only be pursued if the patient presents with a suppressed TSH. Decreased uptake on thyroid scintigraphy is nonspecific for thyroid malignancy (119). Additionally, the task force does not recommend the routine use of additional imaging (e.g., bone scan or [¹⁸F]-fluoro-deoxyglucose positron emission tomography/computed tomography [¹⁸FDG-PET/CT]) in the evaluation of children for PTC. These studies are not validated in this setting and the likelihood of finding disease in an otherwise asymptomatic patient is low and nonspecific (see Section D5).

[C3] What Is the Recommended Surgical Approach for the Patient with a Diagnosis of PTC?

For the majority of patients with PTC, total thyroidectomy (TT) is the recommended initial surgical approach. In this procedure, the left and right thyroid lobes, the pyramidal lobe (when present), and the isthmus are resected. Alternatively, in patients with a small unilateral tumor confined to the thyroid gland, a near-TT, whereby a small amount of thyroid tissue (<1%–2%) is left in place at the entry point of the recurrent laryngeal nerve (RLN) and/or superior parathyroid glands, might be considered in an effort to decrease the risk of permanent damage to these structures. This recommendation for more comprehensive thyroid surgery in pediatric patients is based on data showing an increased incidence of bilateral and multifocal disease (30% and 65%, respectively) (11,14,47,52,167), as well as an increased risk for recurrence and subsequent second surgical procedures when less than a near-TT or TT is performed (5,14,15,44,47,51,168). In long-term analysis of 215 pediatric patients with PTC, bilateral lobar resection compared with lobectomy was shown to decrease the incidence of local recurrence from 35% to 6% over 40 years of follow-up (5). Bilateral thyroid surgery also optimizes the use of RAI for imaging and/or treatment and Tg as a marker to detect persistent/recurrent disease (8,169 –171). Using an intracapsular approach, the superior parathyroid glands may be most easily preserved by maintaining arterial inflow and venous drainage (172 –174).

[C4] Should Central Neck Dissection Be Performed?

In patients with preoperative evidence of central and/or lateral neck metastasis, a therapeutic central neck dissection (CND) should be performed. For this subgroup of patients, who are also at increased risk of pulmonary metastases (10,14,65), CND is associated with a decreased risk of persistent/ recurrent locoregional disease as well as the potential to increase the efficacy of ¹³¹I treatment for distant metastases (14,15,22,47,48).

The increased incidence of cervical metastasis in children suggests that prophylactic CND, as defined in the 2009 ATA consensus statement on the terminology and classification of CND for thyroid cancer (175), should be considered at the time of initial surgery for pediatric patients with PTC. This is particularly relevant given that decreased disease-free survival (DFS) is most strongly correlated with the presence of persistent or recurrent locoregional disease (5,13 –15,22,47,52).

Unfortunately, there are no data that reliably predict which subgroup of patients is at increased risk for locoregional metastasis. Larger tumor size (>4 cm) has been shown to correlate with an increased risk of lymph node metastases (10,11,176). However, up to 36% of tumors ≤4 cm have cervical lymph node metastasis (10). In addition, several of the panel experts have cared for children with regional metastasis found in children with primary tumors ≤1 cm in size. In adults, these tumors are labeled papillary thyroid microcarcinoma (PTMC) and scoring systems have been described to predict the likelihood of metastasis (177). However, the thyroid volume is smaller in young children so that the size criteria used for tumor staging (see Section B4), as well as the diagnosis of PTMC, may not apply to children (178).

While data suggest that pediatric patients with thyroid cancer typically have 100% 10-year disease-specific survival (5,8,162,179), the extent of initial surgery appears to have the greatest impact on improving long-term DFS (5,47). However, without long-term, prospective data and a reliable set of criteria to stratify which patients would benefit from more aggressive surgical resection, one must weigh the risks of more aggressive surgery with the potential benefit of decreasing the incidence of persistent/recurrent disease.

The limited data suggest that, in children, TT with prophylactic CND is associated with increased DFS, as high as 95% at 5 and 10 years (46,163). However, the data are mixed and possibly related to the use of adjunctive RAI remnant ablation. In a retrospective study examining 75 children with PTC, 80% of whom underwent TT with ¹³¹I remnant ablation, the type and extent of neck dissection did not impact the risk for locoregional or distant metastasis (15). Conversely, another study suggested that TT with prophylactic CND may reduce the risk for reoperation that was as high as 77% in those without CND (44). Some groups suggest routinely considering a prophylactic CND, particularly for larger tumors (1,180,181), whereas others suggest making this decision based upon intraoperative findings (182).

If and when performed, CND should only be performed by a surgeon highly experienced in the procedure. To reduce the risk of recurrence, a comprehensive and compartment-based lymph node dissection should be pursued rather than “berry picking” (183). In patients with unifocal disease, data from adult patients suggest that ipsilateral, prophylactic CND may provide the same potential benefit while decreasing the higher complication rate associated with bilateral CND (184). During ipsilateral CND, the incorporation of frozen section to stratify which patients should undergo contralateral (complete) prophylactic CND may achieve a balance between the potential risks and benefits of this procedure (185).

With these considerations in mind, the following recommendations are made in an attempt to balance the goal of achieving surgical remission with the potential increased risk of complications that may be unnecessary for patients with minimal or no locoregional metastasis.

[C5] What Are the Indications for Lateral Neck Dissection?

Pediatric patients occasionally present with bulky disease to the lateral neck and may have suspicious lymph nodes in the lateral neck on preoperative US imaging. US findings suggestive of metastasis to a lymph node include increased size, rounded shape, loss of central hilum, cystic appearance, peripheral vascularity on Doppler imaging, and microcalcifications (186), with the latter two features having the highest specificity for malignancy (113). The US appearance of the lymph nodes may be considered sufficient evidence to pursue lateral lymph node dissection; however, in patients undergoing surgery, FNA to confirm metastasis to the lateral neck lymph nodes should be performed prior to lateral neck dissection. The addition of a Tg measurement in the FNA washout fluid may be used to confirm equivocal cytological evidence of metastatic disease, even in the presence of serum anti-Tg antibodies (TgAb) (187 –191) (see Section D2). When indicated, compartment-oriented lateral neck dissection (levels III, IV, anterior V, and II) is associated with a reduction in persistent/recurrent disease and improved DFS (10,14,47).

[C6] What Are the Possible Complications of Surgery and What Should Be Done to Minimize the Risks of Surgery?

The lower incidence of thyroid disease requiring surgical intervention in children combined with a higher incidence of locoregional lymph node metastasis has been associated with an increased risk of complications for pediatric patients undergoing TT. Utilizing high-volume thyroid surgeons, defined as a surgeon who performs 30 or more cervical endocrine procedures annually, can reduce the rate of complications (70,71). In a cross-sectional analysis of over 600 pediatric patients undergoing thyroid surgery, there were fewer general complications (8.7% vs. 13.4%) and endocrine complications (5.6% vs. 11%) when the procedures were performed by high-volume surgeons (71). In addition, the duration of stay and cost were significantly lower when the procedure was performed by a high-volume surgeon (71).

The most common complications after thyroidectomy are endocrine related and include transient or permanent hypoparathyroidism, with an average rate of approximately 5%–15%. In a high-volume tertiary endocrine surgical practice, the risk of permanent hypoparathyroidism is <2.5% (72). Surgery specific, non–endocrine-related complications include RLN damage, spinal accessory nerve injury, and Horner syndrome, with an average rate of 1%–6% (10,13,46,47,70,72). In patients younger than 10 years of age, there is an increased risk of complications associated with the presence of ETE, lymph node dissection, and repeat surgery (10,70,168).

The risk of hypoparathyroidism correlates with the extent of surgery. Even in patients in whom the parathyroid glands are identified and viability of gland function is likely, manipulation of the parathyroid glands may lead to transient or permanent hypoparathyroidism. Autotransplantation of parathyroid tissue after frozen-section confirmation is utilized if there is any concern of devitalization, and it is associated with a decreased risk of permanent hypoparathyroidism (192,193). Postoperatively, several approaches can predict which patients are at an increased risk of developing hypocalcemia, including serial measurements of serum calcium (194) as well as measurement of a peri-operative intact parathyroid hormone (iPTH) level. The utility of postoperative iPTH is fairly well established with a level of <10–15 pg/mL correlating with an increased risk to develop clinically significant hypocalcemia (195,196). An elevated postoperative serum phosphorous may also be predictive (197). The use of peri-operative iPTH and/or phosphorus monitoring may decrease morbidity and allow for stratification of patients who would benefit from more intensive monitoring and treatment with calcium and calcitriol. An alternative to this approach is to place all patients who have undergone TT, especially those who undergo concomitant CND, on empiric calcium with or without calcitriol replacement therapy.

No monitoring devices have been shown to decrease the rate of non-endocrine surgical complications. The use of intraoperative RLN monitoring may be considered as an adjunct monitoring device and may be considered for younger patients (<10 years of age), in patients undergoing CND, and in patients undergoing repeat surgical procedures. However, the use of RLN monitoring has not been clearly shown to lower the incidence of RLN damage (198).

[C7] What Tumor Classification Systems Can Be Used for Pediatric PTC?

No single postoperative staging system has been validated in children with PTC, and the utility of extrapolating adult staging systems into the pediatric setting is limited by the observed clinical disparity between the two age groups. Specifically, the age-metastasis-extent of disease-size of tumor (AMES) and metastasis-age-completeness of resection-invasion-size (MACIS) have been examined, but the data are limited and the utility of these staging systems in pediatric patients with PTC remains unclear (176, 199). The AJCC TNM classification system (Table 5) is the most widely used system for describing the extent of disease and prognosis in the adult population (69). However, due to the extremely low disease-specific mortality in children with PTC and the fact that all patients aged<45 years have either stage I (no distant metastases) or stage II disease (with distant metastases), the TNM classification system remains limited in terms of determining prognosis in children. Despite this, the TNM classification is an excellent system with which to describe the extent of disease as well as to stratify an approach to evaluation and management. Especially useful to risk-stratify the pediatric PTC patient is knowledge regarding lymph node status. Children with PTC who have gross cervical lymph node disease at diagnosis are more likely to have multifocal disease (89% vs. 16%), an increased incidence of pulmonary metastasis (20% versus none), and increased persistent (30% versus none) and/or recurrent (53% versus none) disease compared with children without palpable nodal disease (53,65).

Therefore, using the TNM classification system, specifically regional lymph node and distant metastasis staging, one can categorize pediatric patients into one of three risk groups. This categorization strategy does not define the risk of mortality (which is low for both stage I and II patients) but identifies patients at risk of persistent cervical disease and helps to determine which patients should undergo postoperative staging to screen for the presence of distant metastasis (Table 6 and Section C8). These three groups are

[C8] What Postoperative Staging Is Recommended?

For most patients, initial staging (Fig. 2) is typically performed within 12 weeks postoperatively. This affords the patient and family time to recover from surgery, while at the same time avoiding delay in additional therapy, if needed. The purpose of postoperative staging is to assess for evidence of persistent locoregional disease and to identify patients who are likely to benefit from additional therapy with ¹³¹I, such as those suspected or known to have distant metastases. The individual patient's risk level (Table 6) helps to determine the extent of postoperative testing. While the committee recognizes that no prospective studies have been performed to validate a stratified risk-based approach in children with PTC, an individualized approach incorporating pathologic findings and postoperative clinical data is founded on well-accepted approaches to therapy in adults (1,3) as well as personal experience in certain pediatric practices (200). The foundation of this stratification system for pediatric patients, however, assumes complete and accurate preoperative staging for regional disease (see Section C2) and appropriate surgery that is performed by a high-volume thyroid cancer surgeon.

OPEN IN VIEWER

FIG. 2.

Initial postoperative staging for American Thyroid Association pediatric intermediate- and high-risk pediatric thyroid carcinoma. ¹Assumes a negative TgAb (see Section D2) and a TSH >30 mIU/L; in TgAb-positive patients, consideration can be given (except in patients with T4 tumors or clinical M1 disease) to deferred evaluation to allow time for TgAb clearance (“delayed” staging). ²Imaging includes neck ultrasonography ± SPECT/CT at the time of the diagnostic thyroid scan. ³Consider ¹³¹I in patients with thyroid bed uptake and T4 tumors or known residual microscopic cervical disease. ⁴While there are no prospective studies in patients ≤18 years of age, the use of ¹³¹I remnant ablation may not decrease the risk for persistent or recurrent disease. Consider surveillance rather than ¹³¹I with further therapy determined by surveillance data. ⁵Repeat postoperative staging 3–6 months after surgery. ⁶See Table 6 and Figures 3 and 4.

For ATA Pediatric Low-risk patients, initial postoperative staging includes a TSH-suppressed Tg. The interpretation of serum Tg and most importantly, interpretation of the trend in serum Tg over time are summarized in Section D2.

In contrast, for ATA Pediatric Intermediate- and High-Risk patients, a TSH-stimulated Tg and DxWBS are generally recommended for further risk stratification and determination of treatment with ¹³¹I (Fig. 2). Children who fall into the ATA Pediatric Intermediate- and High-Risk categories are prepared following standard guidelines for ¹³¹I therapy (see Section C12), and the TSH-stimulated Tg and DxWBS data are used to assess for evidence of residual disease (Fig. 2). In patients without evidence of TgAb, the TSH-stimulated Tg is a reliable marker for evaluating for the presence or absence of residual disease (see Section D2). In a recent study examining 218 consecutive adult DTC patients across all ATA risk stratification levels, a TSH-stimulated Tg <2 ng/mL had a 94.9% predictive value for the absence of disease (201).

Whether a DxWBS might image disease that is not identified through neck US is a matter of debate and few data in children address this. In one pediatric study, US and DxWBS equally identified lymph node metastases in the majority of patients (35/45); however, in six patients, lymph node metastases were found only with a posttreatment RAI WBS (202). Two of the patients were TgAb positive, reinforcing the potential benefit of DxWBS in patients who are TgAb positive (202,203). DxWBS may also visualize disease in the lungs or mediastinum that would not otherwise be shown by neck US or other cross-sectional imaging (49).

For patients with cervical iodine uptake, the addition of hybrid imaging using single photon emission computed tomography with integrated conventional CT (SPECT/CT) offers improved anatomic imaging to determine whether cervical uptake is secondary to remnant thyroid tissue or metastasis to regional lymph nodes (204 –206).

A potential drawback of DxWBS imaging is that if ¹³¹I is used, the small diagnostic activity may theoretically “stun” the iodine-avid tissue and reduce subsequent ¹³¹I uptake if high-activity ¹³¹I treatment is then used (207,208). This issue can be reduced by selecting the lowest possible activity of ¹³¹I (2.7–4.0 mCi=100–148 MBq) (209) or by using ¹²³I (207,210). Due to its lower cost, ¹³¹I is more commonly used, but ¹²³I provides superior imaging quality and generates lower absorbed doses of radiation to the tissues (207,211), which favors its use in children.

Taken together, postoperative staging is used to further stratify which children may or may not benefit from additional treatment with surgery and/or RAI therapy. Irrespective of initial risk stratification, all patients will enter surveillance, ensuring that appropriate therapy will be given if evidence of disease is ultimately identified. As long as the patient is maintained on tailored LT₄ suppression and a proper surveillance plan is followed (Table 6, Fig. 2), delayed treatment is not expected to alter the already low disease-specific mortality due to the indolent nature of PTC in children. Furthermore, a more individualized and conservative approach to postoperative staging and treatment will decrease unnecessary exposure to ¹³¹I in children without evidence of disease, in whom the risks of routine ¹³¹I therapy likely outweigh any benefit. Because of the lack of high-level evidence to help guide these difficult medical decisions, families should be fully informed about the options and their risks/benefits as the surveillance and treatment plans are being formulated.

[C9] What Are the Goals of ¹³¹I Treatment?

The traditional approach to managing pediatric patients with DTC included reflexive postsurgical ¹³¹I therapy, which was prescribed in an effort to eliminate residual thyroid tissue in order to increase the sensitivity for using serum Tg as a biomarker for recurrent disease. In addition, ¹³¹I was prescribed in an effort to decrease the risk of recurrent disease (see Section C10).

The goal of ¹³¹I therapy is to decrease the risks of thyroid cancer recurrence and theoretically to improve mortality by eliminating iodine-avid disease. RAI was proposed as a specific treatment for DTC in 1946 after an adult with functional thyroid cancer metastases responded to multiple RAI treatments (212), and ¹³¹I therapy has since been broadly incorporated into treatment protocols for both adults and children (1,213). A recent survey indicates that use of therapeutic ¹³¹I for all thyroid cancers, regardless of tumor size, has increased from 1990 to 2008 (214).

With increased awareness of the potential long-term side effects of ¹³¹I treatment (see Section C16), there are increased efforts to identify patients who have a high likelihood of benefit from therapy. Unfortunately, the majority of available data are based on a nonstratified approach in which all children underwent TT and variable extent of lymph node dissection, and the majority received therapeutic ¹³¹I. The challenge is to reduce or eliminate unnecessary ¹³¹I exposure for children who may not benefit without increasing disease-specific morbidity and mortality. The following sections address various aspects of this question.

[C10] What Is the Impact of ¹³¹I Therapy on Recurrence and Survival for Children with PTC?

Adjunctive ¹³¹I therapy may improve DFS in young adults (including some adolescents) but this has not been universally shown for those with small, stage I lesions (215). Reflective of this, the 2009 ATA guidelines and the current NCCN guidelines support the selective rather than universal administration of ¹³¹I, especially for young patients (<45 years of age) with intrathyroidal PTC and either no or limited lymph node disease (1,3).

Studies specifically examining the potential benefits of ¹³¹I therapy in children have been difficult to perform because the number of patients is small and the prognosis is favorable, regardless of adjunctive therapy (11,169,216,217). Arguments in favor of universal therapeutic ¹³¹I for children have been based on the observation that retention of the normal thyroid remnant may decrease the sensitivity for detecting metastases or recurrent disease by serum Tg and/or DxWBS (218,219). Arguments against the universal prescription of therapeutic ¹³¹I are based on the known short- and long-term toxicities (220), lack of data showing conclusive benefit from routine ¹³¹I therapy (5,13), a possible increase in the risk of secondary malignancies (5 –7,221,222), and studies showing that Tg can remain useful and become undetectable in patients post TT despite not having received ¹³¹I (223 –225).

Most of the data regarding ¹³¹I use in children have examined treatment of known residual disease rather than ablation of the normal thyroid remnant only (9). In patients with known residual disease, ¹³¹I therapy appears to decrease recurrence (9,43,51,162,226). However, large retrospective series in children show conflicting results regarding the potential for benefit from adjunctive ¹³¹I. In one study, ¹³¹I remnant ablation did not clearly decrease the risk for locoregional recurrence, distant metastases, or all-sites recurrence compared with surgery alone, but there was a trend toward reduction in the risk of distant metastases (p=0.06) (5). Unfortunately, overall survival was reduced in patients who had received external beam therapy, radium implants, or ¹³¹I, primarily secondary to an increase in nonthyroid, second malignancies. Another study also found no improvement in DFS following remnant ablation (13), but additional studies revealed a significant improvement in DFS for children with PTC treated with ¹³¹I and no clear increase in the risk of second primary malignancies (169).

[C11] Which Children Might Benefit from Therapeutic ¹³¹I?

¹³¹I is indicated for treatment of nodal or other locoregional disease that is not amenable to surgery as well as distant metastases that are known or presumed to be iodine-avid (169). In addition, some experts also advocate routine ¹³¹I therapy for children with T3 tumors or extensive regional nodal involvement (extensive N1a or N1b disease) (9,169).

Published studies show that children with iodine-avid pulmonary metastases benefit from ¹³¹I treatment, and complete remission is achievable for many patients, particularly those with microscopic and small-volume lung disease (9,57,58,162,227). Thus, for patients with pulmonary metastases, ¹³¹I is considered therapeutic, with the understanding that increasing burden of disease may ultimately require administration of multiple activities (57) (see Section D8).

■  RECOMMENDATION 17

¹³¹I is indicated for treatment of iodine-avid persistent locoregional or nodal disease that cannot be resected as well as known or presumed iodine-avid distant metastases. For patients with persistent disease following ¹³¹I administration, the decision to pursue additional ¹³¹I therapy should be individualized according to clinical data and previous response (see Fig. 3 and 4). The potential risks and benefits must be weighed on an individual basis.

OPEN IN VIEWER

FIG. 3.

Management of the pediatric patient with known or suspected residual/recurrent disease (no known distant metastases). This algorithm is intended to be used in children who are known or suspected to have residual or recurrent disease based upon the suppressed Tg level and knowledge of previous disease extent 6–12 months after all primary therapies have been completed. ¹Assumes a negative TgAb (see Section D2); in TgAb-positive patients, the presence of TgAb alone cannot be interpreted as a sign of disease unless the titer is clearly rising. ²Imaging includes SPECT/CT at the time of the diagnostic thyroid scan and/or contrast-enhanced CT/MRI neck. ³Repeat ¹³¹I therapy in patients previously treated with high-dose ¹³¹I should generally be undertaken only if iodine-avid disease is suspected and a response to previous ¹³¹I therapy was observed (see Sections D7 and D8).

OPEN IN VIEWER

FIG. 4.

Management of the pediatric patient with known distant metastases. ¹Assumes a negative TgAb (see Section D2); in TgAb-positive patients, the presence of TgAb alone cannot be interpreted as a sign of disease unless the titer is clearly rising; a declining TgAb titer would suggest continued response to treatment. ²Tg can transiently rise after ¹³¹I therapy and should not be misinterpreted as evidence for progression. ³Repeat ¹³¹I therapy in patients previously treated with high-dose ¹³¹I should be undertaken only if iodine-avid disease is suspected and if there was a previous response to therapy (see Section D7 and D8).

Recommendation rating: B

[C12] How Should a Child Be Prepared for ¹³¹I?

If ¹³¹I is prescribed, the TSH should be above 30 mIU/L to facilitate uptake (21,228,229). The majority of children will achieve this level of TSH by ≥14 days of LT₄ withdrawal (230). For that reason, triiodothyronine supplementation during LT₄ withdrawal is not usually required but can be considered for children who are especially sensitive to hypothyroid symptoms or if the withdrawal period extends beyond 3 weeks.

Recombinant human TSH (rhTSH) has been used for remnant ablation as well as for treatment of intermediate- and high-risk DTC in adults (231 –233) and may result in a lower absorbed radiation activity to the blood (as much as one third lower) (234). However, data regarding the use of rhTSH in children are limited (235,236). Experience in children would suggest that the typical adult dose of rhTSH (two doses of 0.9 mg given 24 hours apart) appears to be safe and generates sufficient TSH levels (169,236,237). In particular, rhTSH might have a role in situations in which endogenous hypothyroidism should be avoided (e.g., significant medical comorbidities) or is impossible (e.g., TSH deficiency) (169,236).

To facilitate RAI uptake, a low-iodine diet is generally prescribed for 2 weeks prior to therapy, but the efficacy of this practice in children has not been specifically demonstrated. Nevertheless, a low-iodine diet has been shown to increase the effective radiation dose to the thyroid by 50%–150% in adults (238). For that reason, a low-iodine diet is commonly recommended. In children who received intravenous contrast during preoperative staging, it is advisable to wait approximately 2–3 months or to confirm normal (median normal 24-hour urine iodine excretion=143 μg/24 hour, 5%–95% range=75–297 μg/24 hour) (239) or low 24-hour urine iodine values before performing either a DxWBS or administering therapeutic ¹³¹I.

[C13] What Should Be Considered for Administration of ¹³¹I?

General guidelines for safety in the administration of ¹³¹I were reviewed by the ATA Task Force on Radiation Safety in 2011 (240). There are few specific references to children, but the overall document pertains to children as well as adults. Once the decision to administer ¹³¹I is made, the safety of family members and classmates will help guide the decision for inpatient or outpatient therapy. This will be largely based on patient age and ability to comprehend the tasks required for outpatient therapy. Other factors to consider are the amount of radiation retained by the patient and the potential exposure time and distance between patient and others (240). In general, children and adolescents with PTC are primarily a radiation risk to others during the first 1–2 days after ¹³¹I therapy. For young children, this may be especially problematic, if they are not yet toilet trained or are frightened to sleep alone. Detailed instructions for the daily care of children who have received ¹³¹I are provided in the ATA guidelines on radiation safety and abbreviated in appendix 1 of that document (240).

Adjunctive therapies to minimize the risk of ¹³¹I to the treated child have not been well studied. Adequate hydration is essential to enhance ¹³¹I clearance and should be encouraged. Regular bowel evacuation is also important, so stool softeners or laxatives may be considered. Nausea and/or vomiting are common following ¹³¹I therapy, particularly in young children and those receiving higher ¹³¹I activities. In such cases, anti-emetics like the serotonin 5-HT₃ receptor antagonists can be considered. Accelerating ¹³¹I clearance from the salivary glands may reduce the risk of sialadenitis, but the use of sialagogues such as sour candy or lemon juice is poorly studied in children. Some studies in adults found benefit by starting lemon drops 24 hours after ¹³¹I dosing (241), but some experts do not recommend this practice. Similarly, the use of the radioprotectant amifostine has not been validated in children, and a recent review of randomized control trials in adults found no benefit from amifostine therapy (242).

Adjunctive treatments to increase the efficacy of ¹³¹I therapy have also not been well studied in children. In adults with PTC, co-treatment with lithium has been suggested to increase ¹³¹I retention and improve the efficacy of treatment of metastatic tumors (243). To our knowledge, no study of children with PTC has evaluated the safety and efficacy of lithium co-treatment. Because the expression of NIS is more common and more robust in pediatric PTC, the effect of lithium on ¹³¹I retention might be less than that found in adults (169).

[C14] How Is the Activity of Therapeutic ¹³¹I Determined?

Therapeutic ¹³¹I administration is commonly based on either empiric dosing or whole body dosimetry. There are no standardized activities of ¹³¹I for children and, to our knowledge, there are no data that compare the efficacy, safety, or long-term outcomes from ¹³¹I administration in children using these different approaches.

Empiric dosing offers the advantage of simplicity. Some adjust ¹³¹I activity according to weight or body surface area and give a fraction (e.g., child's weight in kilograms/70 kg) based on the typical adult activity used to treat similar disease extent (1,3,21,219,229). Others suggest that ¹³¹I activities to treat residual disease should be based on body weight alone (1.0–1.5 mCi/kg; 37–56 MBq/kg), while still others feel this may not be as reliable as dosing based on body surface area. In general terms, a 15-year-old may require five sixths of the adult activity, a 10-year-old may require one half of the adult activity, and a 5-year-old may only require one third of the adult activity for similar extent of disease (169).

For children with diffuse lung uptake or significant distant metastases, those undergoing multiple ¹³¹I treatments, or children who may have limited bone marrow reserve due to prior chemotherapy or radiation therapy, whole-body dosimetry can be used to calculate the largest activity of ¹³¹I that could theoretically be administered so that the absorbed activity to the blood does not exceed 200 rads (cGy) and that the whole-body retention 48 hours after administration does not exceed 4.44 GBq (120 mCi) in the absence, or 2.96 GBq (80 mCi) in the presence, of iodine-avid diffuse lung metastases, respectively (244 –246). Lesional dosimetry can also be performed to select effective activities of ¹³¹I for children with substantial lung involvement or large tumor burden at distant sites such as bone (209,210,245,247,248). One must keep in mind that these toxicity constraints have not been validated in pediatrics and may be associated with significant toxicity in young children (249,250). Furthermore, these protocols are time consuming and not routinely available at all referral centers.

[C15] Should a Posttreatment Whole-Body Scan Be Obtained?

Approximately 4–7 days after ¹³¹I therapy, a posttreatment whole-body scan should be performed to take advantage of the increased sensitivity associated with administration of the larger activity of ¹³¹I used for therapy (208). Newer gamma-camera systems allow scanning as early as 72 hours after ¹³¹I therapy (251). On occasion, the posttreatment WBS (RxWBS) may reveal metastatic disease (regional or pulmonary) that was not apparent on the DxWBS (218), but it remains uncertain if this knowledge informs future treatment or outcomes. If new lesions are identified on the RxWBS, the addition of SPECT/CT to the RxWBS may afford greater definition of residual disease during postoperative restaging (252).

In addition, the clearance of RAI from thyroid cancers has been shown to vary substantially with biological half-life ranging from 3 to 12 days (253). Rapid turnover of iodine might clear ¹³¹I by the time standard imaging protocols are performed, while other lesions might be better revealed with delayed imaging (254). For the rare child with elevated serum Tg and negative RxWBS, serial acquisition times may be beneficial in documenting disease and iodine avidity.

[C16] What Are the Acute and Long-Term Risks of ¹³¹I Therapy in Children?

There are both acute and long-term side effects and complications associated with exposure to therapeutic ¹³¹I. The side effects can be organized by organ system, and the majority are explained by mechanistic exposure based on the method of delivery, absorption, distribution, and clearance.

The short-term side effects of ¹³¹I are well known and include damage to tissues that incorporate iodine, resulting in sialadenitis, xerostomia, dental caries, stomatitis, ocular dryness, and nasolacrimal duct obstruction (255,256). Strategies exist to help treat or prevent ¹³¹I-related side effects (257,258); however, even a single activity of ¹³¹I may lead to permanent salivary gland dysfunction with life-long xerostomia, an increased incidence of dental caries, and an increase in the risk for salivary gland malignancy (258,259). The use of sour candy or lemon juice, starting 24 hours after ¹³¹I dosing, with vigorous hydration for 3–5 days may protect salivary gland function (241). The use of rhTSH has not been shown to decrease salivary gland toxicity compared to thyroid hormone withdrawal (260); however, lacrimal dysfunction (watery eyes) was more frequent in patients undergoing thyroid hormone withdrawal (261). None of these prophylactic measures or other sialogogues have been formally studied in the pediatric population.

Gonadal damage has been reported in both women and men (262,263). In postpubertal males, transient rise in follicle-stimulating hormone is common and may persist for up to 18 months after ¹³¹I exposure (263,264). Increasing cumulative activities of ¹³¹I may lead to decreased spermatogenesis generally without an effect on testosterone production (263,265,266). Current guidelines recommend that males avoid attempts at conception for at least 4 months post ¹³¹I therapy. Postpubertal testes appear to be more vulnerable than prepubertal testes to the toxic effects of ionizing radiation (267). Therefore, postpubertal males should be counseled and sperm banking should be considered for those receiving cumulative activities ≥400 mCi (14.8 GBq) (268).

Transient amenorrhea and menstrual irregularities are reported in up to 17% of females under the age of 40 years. This is true despite the fact that 65% of young women received a single low activity of ¹³¹I (average=81 mCi; 3 GBq) (269). Other studies have not shown an increase in infertility, miscarriage, or birth defects following ¹³¹I (262,270,271). Collectively, these data have led to the recommendation that conception should be avoided during the year immediately following ¹³¹I administration (272).

Acute suppression of bone marrow may occur but hematologic parameters usually normalize within 60 days after ¹³¹I exposure. Commonly a decline in leukocyte (∼57%) and platelet counts (∼44%) occurs within the first month after treatment. This is followed by a less pronounced decline in erythrocyte count (∼10%) by the second month after treatment, but usually all parameters normalize ∼3 months post therapy (273). Long-term bone marrow suppression is rare; however, there are reported cases of leukemia after multiple high activities of ¹³¹I administered over a short span of time (274). Therefore, it is important to allow for recovery of bone marrow between ¹³¹I treatments.

In support of these clinical observations, ¹³¹I has been shown in peripheral lymphocytes to induce a significant increase in the number of dicentric chromosomes (275 –277), and the aberrations of chromosomes 1, 4, and 10 are not only more prevalent but are still apparent after 4 years (276). A recent report comparing thyroid hormone withdrawal to rhTSH suggests a lower frequency of lymphocyte chromosomal rearrangements after ¹³¹I dosing using rhTSH preparation (278).

A few studies combining patients of all ages have shown that ¹³¹I therapy is associated with an increased risk for second malignancies and an increase in overall mortality for patients with DTC (7). A large study by Brown et al. (6) reviewed data from over 30,000 subjects and found a significant increase in second malignancies among patients treated with ¹³¹I (relative risk 1.16, p<0.05). They also noted that the risk was greater for younger patients. In a study that exclusively evaluated children, Hay et al. (5) found that children who were treated with radiation (external beam radiation, radium implants, or ¹³¹I) developed a variety of second cancers (leukemia, stomach, bladder, colon, salivary gland, and breast) and had increased mortality compared with the general population. However, only 4 of the 15 patients that died from nonthyroid second primary malignancies were associated with the sole administration of ¹³¹I (one acute myelogenous leukemia, one lung, one adenocarcinoma, and one breast cancer) (5). Whether this resulted from aggressive treatment, an underlying predisposition to cancer, or from a direct effect of ¹³¹I is unknown, but the increase in overall mortality following ¹³¹I treatment of a disease with low disease-specific mortality is of growing concern.

It is difficult to determine from these data if there is a “safe” cumulative exposure to ¹³¹I or if the increase in second malignancies occurs following any amount of ¹³¹I. Further complicating this question is the fact that the effects of RAI may be amplified in children because a given activity of ¹³¹I is distributed over shorter distances, taken up by smaller organs, and accumulated by cells with increased growth and proliferation potential. Despite these limitations, Rubino et al. (7) proposed an activity–response relationship in which the relative risk for second malignancy appears to increase above a cumulative activity of 200 mCi (7.4 GBq) ¹³¹I, and Rivkees et al. (220) suggested an increased risk above a cumulative exposure to 300 mCi (11.1 GBq) ¹³¹I. However, there are anecdotal reports of acute myelogenous leukemia after 85 mCi (3.1 GBq) ¹³¹I, lung cancer after 150 mCi (5.6 GBq) ¹³¹I, and adenocarcinoma after 200 mCi (7.4 GBq) ¹³¹I (5). Unfortunately, there is a lack of long-term data to define a “safe” activity of ¹³¹I, and additional study is clearly warranted.

Lastly, for pediatric patients with lung metastases, a significant risk exists for ¹³¹I-induced pulmonary fibrosis when the retained ¹³¹I activity exceeds 80 mCi (3 GBq) (228,279). For that reason, patients with significant uptake on DxWBS are candidates for dosimetry or reduced ¹³¹I dosing.

In summary, there are clear benefits and risks, both acute and chronic, following administration of ¹³¹I. The challenge is to define the subgroup of patients who will not experience an increase in morbidity or disease-specific mortality if ¹³¹I is deferred or withheld.

[D2] What Is the Role of Tg Testing in the Follow-Up of PTC in Children?

Tg is a thyroid-specific glycoprotein that is synthesized and secreted by the normal thyroid and by differentiated thyroid carcinomas. Following surgery and ¹³¹I therapy, serum Tg levels serve as a sensitive marker of residual or recurrent disease (280 –283); the magnitude of serum Tg elevation appears to correlate with the site(s) of metastatic disease and with the tumor subtype (280).

Highly sensitive and specific immunometric analyses for serum Tg have been shown to be more sensitive for detecting residual thyroid cancer in adults compared with the DxWBS (281,284 –286). Currently, most laboratories use immunometric methods for Tg measurement, and these should be calibrated against the CRM-457 international standard (1). However, even with standardization, there can be significant differences in Tg results between various assays (171). For that reason, serial Tg measurements should ideally be performed in the same laboratory using the same assay.

In the absence of TgAb (see following section), serum Tg has a high degree of sensitivity and specificity to detect residual/recurrent DTC, with the highest sensitivity noted following TSH stimulation (TSH-stimulated Tg) (282). Serum Tg levels rise with TSH stimulation, and the duration of stimulation is generally longer in the hypothyroid state, resulting in higher serum Tg levels than occur after rhTSH in the euthyroid state (281). In adults, a TSH-stimulated serum Tg level >2 ng/mL has a high predictive value for disease (287).

Previous studies of children with DTC have focused on DxWBS as the “gold standard” for disease status (11,219,288), and there are few data regarding the interpretation of Tg levels in children with DTC. Because data suggest that serum Tg levels might be higher in children compared with adults with a similar extent of disease (58,289), application of data from adult studies to children is difficult. Therefore, it is not yet clear if elevated Tg levels have the same prognostic value for children, who may have a different Tg threshold for what would be considered clinically relevant or “actionable” disease.

[D3] What Is the Role of Ultrasound in the Follow-Up of PTC in Children?

Children with PTC who have residual/recurrent disease are most likely to have cervical lymph node disease (5,9,11,14,163,217,319). US, in conjunction with Tg levels, has proven highly effective in identifying and localizing regional nodal metastases in both adults and children with PTC (113,190,191,202,283,286) and appears even more sensitive than a TSH-stimulated Tg to identify disease (190,202,297). US has also proven useful for directing FNA of suspicious lesions/lymph nodes in the thyroid bed or lateral neck that can then be evaluated by routine cytology and Tg immunoassay of the needle washout, especially if cytopathology is equivocal or uninformative (187 –191). Therefore, US is the most important clinical tool for identifying cervical disease and is recommended at routine intervals based upon the patient's ATA Pediatric Risk level and clinical concern for persistent or recurrent disease (see Table 6, Fig. 2 and 3).

[D4] How Are Diagnostic RAI Scans Best Used in the Follow-Up of PTC in Children?

A DxWBS is usually performed as part of the postoperative staging following initial surgery in ATA Pediatric Intermediate- and High-Risk patients and can be considered in ATA Pediatric Low-Risk patients who have evidence of residual disease after short-term follow-up (see Table 6, Fig. 2 and 3, and Sections C8/C12).

Routine surveillance for persistent or recurrent disease in children with DTC has historically relied on sequential DxWBS (11,219,288). However, serial neck US and measures of serum Tg appear to be sensitive indicators of disease status in the vast majority of pediatric patients. For that reason, there is no role for serial thyroid scintigraphy in a child who has not previously been treated with ¹³¹I, unless evidence exists for persistent or recurrent disease (see Fig. 3).

For the child who has received therapeutic ¹³¹I, there may be a role for a follow-up DxWBS, typically 1–2 years following the initial treatment with ¹³¹I (Table 6). Children with known iodine-avid metastases based upon a prior posttreatment scan are the most likely to benefit from subsequent staging with a DxWBS. Ideally, the DxWBS should be performed only after a significant period of time has elapsed to assess the response from the last dose of therapeutic ¹³¹I, recognizing that clinical response can continue for years (58) (see Fig. 4 and Section D8). Finally, once a DxWBS is negative, repeating the procedure has no utility unless disease is clinically suspected.

[D5] What Imaging Studies Should Be Considered in the Pediatric PTC Patient Who Is Tg Positive but Who Has No Evidence of Disease on Cervical Ultrasound or DxWBS?

The child previously treated with surgery and ¹³¹I who has a serum Tg suggestive of residual/recurrent disease but no other evidence of disease presents a particularly challenging clinical situation. In this setting, one should first ensure that cervical US has been performed by an experienced radiologist and also confirm that iodinated contrast agents were not given to the patient within the 3 months prior to an RAI scan. Treatment algorithms have been proposed for adults who are “Tg-positive, scan-negative” and generally focus on anatomic imaging of the neck and chest, ¹⁸FDG-PET/CT, and additional therapeutic ¹³¹I with a posttreatment scan (1,3,320).

In children, the neck and chest are the most likely sites of persistent disease, and contrast-enhanced imaging of these areas with CT or MRI is favored when US cannot identify disease. ¹⁸FDG-PET/CT has become a commonly used tool in the evaluation of adults with persistent non–iodine-avid thyroid cancer (3,321 –326) and appears to offer prognostic information that might change clinical management (327 –330). However, there are extremely limited data regarding the use of ¹⁸FDG-PET/CT in children except for a case report (331), isolated pediatric subjects embedded within adult studies, and unpublished data that suggest low sensitivity of ¹⁸FDG-PET/CT to identify residual disease in children that otherwise cannot be identified via cervical US and cross-sectional imaging of the neck and chest (personal communication, SGW). Whether or not the use of ¹⁸F-FDG PET has similar prognostic value or will change disease management in children with thyroid cancer remains to be determined. Finally, empiric treatment with ¹³¹I does not appear to be effective in adults who have a negative DxWBS (332,333). Although children are more likely to have RAI-responsive disease compared with adults, empiric treatment with high-activity ¹³¹I is not generally advocated to identify disease unless there is evidence for clinical progression (e.g., a rising Tg level) and a documented clinical response to previous ¹³¹I therapy (see Fig. 3).

[D6] What Are the Goals and Potential Risks of TSH Suppression Therapy?

DTCs in children are well-differentiated tumors that may respond to TSH stimulation with increased growth and Tg production. For that reason, TSH suppression has been an important cornerstone of treatment, especially for high-risk groups (217,334 –336). However, there are no data in children with which to compare the outcomes, risks, and benefits of various TSH suppression strategies. Some experts recommend initial TSH suppression to <0.1 mIU/L followed by relaxation to 0.5 mIU/L following remission of DTC (337). The ATA guidelines for adults stratify target TSH levels based on the risk of recurrence (1). Recognizing the paucity of data regarding TSH suppression in children with DTC, the panel has concluded that the initial TSH goal should be tied to ATA Pediatric Risk level and current disease status (Table 6). In children without evidence of disease, the TSH can be normalized to the low-normal range after an appropriate period of surveillance.

The actual risks of TSH suppression in children with DTC have been poorly studied. Extrapolating from patients with Graves' disease, the potential risks of TSH suppression include growth acceleration, advanced bone age, early onset puberty, reduced bone mineral content, poor academic performance, tachyarrhythmia, and others (338,339). It should be emphasized, however, that patients with Graves' disease generally have much greater elevations in thyroxine levels than do patients on TSH-suppressive therapy for DTC. Thus, the applicability of these data to long-term DTC management is currently unknown.

[D7] What Is the Optimal Approach to the Patient with Persistent/Recurrent Cervical Disease?

The majority of residual/recurrent PTC in children will be identified in cervical lymph nodes (5,9,11,14,163,217,319), and the optimal management depends on several factors, including the location and size of disease, the previous surgical and ¹³¹I treatment history, the presence of distant metastases, and whether or not the disease is iodine-avid (see Fig. 3). Patients with cervical RAI uptake due to disease that is small (i.e., <1 cm) or that cannot be visualized via cross-sectional imaging can be considered for treatment with therapeutic ¹³¹I, which may reduce future recurrence risk but is unlikely to improve mortality (9,14,169). Although repeat surgery may also be an option, finding a small recurrence in the neck intraoperatively may be difficult. In most cases, children with small volume residual disease <1 cm can be safely observed while continuing TSH suppression. Given the overall excellent prognosis, and the low risk for clinically significant progression, the risk to benefit ratio for the treatment of small-volume disease in a child who has already undergone surgery and ¹³¹I is unfavorable.

On the other hand, in patients with structural disease >1 cm in size that is visualized by US and/or anatomic imaging (CT or MRI) and confirmed via FNA, surgical resection is preferable to ¹³¹I and can result in safe and effective long-term control of disease, especially when surgery is performed by a high-volume surgeon (340,341).

[D8] How Should Children with Pulmonary Metastases Be Managed?

The majority of children with pulmonary metastases have micronodular disease that typically demonstrates excellent RAI uptake. Because of this, distant metastases in children are more amenable and responsive to ¹³¹I therapy compared with adults. Serial ¹³¹I treatments can result in remission in many, but not all, children with pulmonary metastases from PTC, the vast majority of whom will demonstrate stable metastatic disease and low disease-specific mortality (10,57,58,162,176,179,227,289,319,342 –344). The optimal frequency of ¹³¹I treatment has not been determined. The maximal clinical and biochemical response from an administered activity of ¹³¹I may not be reached for up to 15–18 months (302), and recent studies have also demonstrated a continuous improvement in serum Tg levels for years following discontinuation of ¹³¹I therapy in children with pulmonary metastases (58), all of which suggests that the effects of therapy can be seen well beyond the first years after treatment. Because children with pulmonary metastases have historically been treated aggressively with repeated activities of ¹³¹I, it remains unknown how these pediatric patients would respond to the less-aggressive use of ¹³¹I. Given that a majority of children with pulmonary metastases will not have a complete response to therapy and because it may take years to see the full response of ¹³¹I, an undetectable Tg level should no longer be the sole goal of treatment of children with pulmonary metastases. Furthermore, longer intervals between ¹³¹I therapy would seem prudent in the child who does not demonstrate progressive disease.

For patients with persistent pulmonary metastases who have already received treatment with high-activity ¹³¹I, the decision to re-treat should be individualized (see Fig. 4). Because of our improved understanding regarding prognosis and duration of response in children with pulmonary metastases, and to minimize the long-term risks associated with high cumulative activities of ¹³¹I (see Section C16), it is logical to monitor the TSH-suppressed Tg and imaging studies in these children, deferring repeat evaluation and treatment with RAI until the full response to previous ¹³¹I is demonstrated. In the rare event of disease progression, further evaluation and treatment would be warranted, as long as it has been >12 months from the previous activity of ¹³¹I (see Fig. 4). For serologic progression, waiting at least 12 months would better establish a trend to ensure that the rise in the Tg or TgAb levels is not spurious or due to previous ¹³¹I-induced tumor destruction. Furthermore, a longer interval between treatments may minimize the risk of late effects of ¹³¹I (see Section C16). In all cases, therapeutic ¹³¹I should be considered only if the patient is known or presumed to have RAI-responsive disease and has not already received high cumulative activities of ¹³¹I. If the child did not have previous RAI uptake on a posttreatment scan or if their disease continued to progress despite high-activity ¹³¹I, further ¹³¹I therapy is unlikely to be helpful and should not be given. In these cases, continued observation and TSH suppression are indicated, with alternative therapies considered if progression of iodine-refractory disease becomes clinically significant (Section D10).

As children with pulmonary metastases may have diffuse RAI uptake in the lungs, there is a real concern about treatment-induced pulmonary fibrosis (57,258,345 –348). In these cases, administering lower ¹³¹I activities and employing dosimetry should be considered to limit radiation exposure to the nontarget normal lung parenchyma (244 –246,347,349) (Section C14). The utility and optimal intervals at which to perform pulmonary function testing in children with lung metastases have not been studied, but many experts recommend that pulmonary function testing be done intermittently in children with pulmonary metastases, especially if multiple ¹³¹I treatments are planned.

[D9] How Does One Approach the Child with an Incidental PTC Identified After Surgery for Another Thyroid Condition?

Small foci of PTC may be incidentally discovered on histological examination of thyroid tissue resected for other benign diseases such as Graves' disease, autonomous nodule(s), or multinodular goiter. No consensus exists regarding the benefit of completion thyroidectomy (assuming lobectomy was initially performed) or ¹³¹I therapy for children with incidental PTC. However, these children should undergo neck US, if not already performed, and be managed similar to other children with ATA Pediatric Low-Risk PTC (Table 6).

[D10] What Are the Optimal Approaches to the Pediatric Patient Who Develops Progressive Thyroid Cancer That No Longer Concentrates or Responds to ¹³¹I?

Very rarely, children with thyroid cancer may develop progressive symptomatic and/or life-threatening disease that is not amenable to further surgery or ¹³¹I. In such cases, systemic therapy should be considered. Clinical trials would be preferred, but there has not yet been a clinical trial developed for children with ¹³¹I-refractory DTC. Some drugs with potential efficacy may be available through phase 1 pediatric studies. Doxorubicin remains the only United States Food and Drug Administration (FDA)-approved cytotoxic chemotherapy for this indication and has been used either as a single agent or in combination with cisplatin or interferon-α (39,350,351), but it is generally ineffective in treating advanced DTC.

Molecularly targeted therapies using oral small molecule kinase inhibitors have brought newer options to the management of ¹³¹I-refractory thyroid cancer in adults (39,352). The pediatric experience has been limited to published case reports and anecdotal clinical experience, primarily with sorafenib (353,354), which is FDA-approved for the treatment of advanced iodine-refractory DTC in adults based upon the results of a phase III trial (355). [Subsequent to the completion of these guidelines, lenvatinib was also approved by the FDA based upon a pivotal phase III clinical trial in adults (356).] Although more study is required regarding the use of these agents in children, particularly with respect to dosing and toxicity, the use of molecularly targeted therapies may be contemplated in the rare situation in which a child warrants systemic treatment. However, it is difficult to define iodine-refractory disease in pediatric DTC, and iodine-refractory DTC can remain stable over years of follow-up. For that reason, all children being considered for anti-neoplastic therapy should be referred to centers familiar with the use of these novel therapeutic agents in thyroid cancer. In all cases, a systematic approach to care and toxicity evaluation should be undertaken (357).