Abstract
Background: The Bethesda System for Reporting Thyroid Cytopathology (BSRTC) is used to interpret fine-needle aspiration (FNA) cytology of thyroid nodules in children and adults. Nodule management is guided by the implied malignancy risk of each cytological category, which has been derived from adult populations. Whether these implied risks are applicable to pediatric thyroid nodules remains uncertain. We compared malignancy rates between pediatric and adult thyroid nodules within each cytological category.
Methods: We evaluated consecutive thyroid nodules ≥1 cm that underwent FNA at the Boston Children's Hospital and Brigham and Women's Hospital from 1998 to 2016. All cytology was interpreted by a single cytopathology group according to the BSRTC. Malignancy rates were compared between pediatric (<19 years) and adult (≥19 years) patients.
Results: Four hundred thirty pediatric thyroid nodules and 13,415 adult nodules were analyzed. Pediatric nodules were more likely to be malignant than adult nodules (19% vs. 12%, p = 0.0002). Within cytological categories, malignancy rates were higher in pediatric nodules than in adult nodules that were cytologically nondiagnostic (11% vs. 4%, p = 0.03), atypia of undetermined significance (AUS; 44% vs. 22%, p = 0.004), or suspicious for follicular neoplasm (SFN; 71% vs. 28%, p = 0.001). There were no significant differences between children and adults in the types of thyroid cancers diagnosed in these cytological categories. Among cytologically benign nodules, the difference in malignancy rates was statistically significant but clinically minimal (0.7% vs. 1%, p = 0.001). Malignancy rates did not differ between children and adults among nodules with cytology suspicious for papillary carcinoma (73% vs. 68%, p = 0.67) or positive for malignancy (97% vs. 95%, p = 1). Among the subset of nodules that were resected, the malignancy rate was higher in children than in adults only in nodules that were SFN (71% vs. 36%, p = 0.007).
Conclusions: Among thyroid nodules that are cytologically AUS, SFN, or nondiagnostic, malignancy rates are higher in children than in adults. These discrepancies likely represent true differences in malignancy risk between pediatric and adult patients, rather than differences in cytological interpretation. Our findings provide pediatric-specific data to inform the optimal management of thyroid nodules in children, which may differ from that of adult nodules with equivalent cytology.
Keywords: thyroid nodules, pediatric, indeterminate, cytology, Bethesda system, malignancy
Introduction
Interpretation of cytology from fine-needle aspiration (FNA) is a key element in the evaluation of thyroid nodules and provides an approximate risk of malignancy that guides decisions regarding surgical resection. The Bethesda System for Reporting Thyroid Cytopathology (BSRTC) was developed in adult patients to provide a standardized framework for interpreting and reporting FNA cytology (1), and its use is recommended for the evaluation of thyroid nodules in both adults and children (2,3). However, it remains unclear whether the performance of the BSRTC and its implied malignancy rates are similar in pediatric and adult patients (4), and therefore whether pediatric nodules should be managed similarly to adult nodules with the same cytology.
Prior studies have assessed the accuracy of FNA cytology in pediatric thyroid nodules and have reported malignancy rates in various cytological categories. However, many of these studies are limited by factors including small sample size, inclusion of nonpediatric patients (up to age 21 years), and use of non-BSRTC cytological categories that are less relevant to present practice (5–11). Moreover, the rarity of pediatric thyroid nodules may limit the frequency with which pediatric pathologists interpret thyroid cytology, leading to variability in cytological classification. This problem is particularly salient for samples with indeterminate cytology, for which there is significant inter- and intraobserver variation even among experienced cytopathologists (12). To address these issues, we sought to determine the malignancy rates of pediatric thyroid nodules in each BSRTC category and to compare them with malignancy rates of adult thyroid nodules, all evaluated by a single cytopathology group.
Methods
Subjects
We reviewed the medical records of all consecutive patients who underwent ultrasound-guided FNA of a thyroid nodule ≥1 cm in diameter in the Thyroid Nodule Clinic at Boston Children's Hospital (BCH) and Brigham and Women's Hospital (BWH) between January 1998 and November 2016. Thyroid ultrasound was performed by a radiologist with expertise in thyroid ultrasonography. FNA was performed under ultrasound guidance by an adult or pediatric thyroidologist using a 25-gauge 1.5-inch needle, and aspirates were processed using ThinPrep liquid cytology preparation (Hologic, Inc., Marlborough, MA). Cytology was interpreted and reported according to the BSRTC (1) by one of a team of BWH cytopathologists, all with special qualification in cytopathology of the American Board of Pathology. Although part of the study period predates the formal introduction of the BSRTC, all cytopathologists used diagnostic criteria and terminology that were later adopted by the BSRTC. On-site cytology evaluation for sample adequacy was not performed, except in the rare patient who required sedation for FNA or for nodules with repeated nondiagnostic cytology.
Nodule management after FNA was based primarily on cytology results. Observation was recommended for most nodules with benign cytology, but in some cases lobectomy was performed for large nodules (>4 cm), compressive symptoms, or based on patient or family preference. For patients whose initial sample was nondiagnostic or atypia of undetermined significance (AUS), repeat FNA was recommended. For nodules with repeated nondiagnostic cytology, lobectomy was recommended in children, whereas lobectomy or observation was pursued in adults based on individual patient factors. Patients with cytology suspicious for follicular neoplasm (SFN) or suspicious for papillary carcinoma (SUSP) were advised to undergo either lobectomy or total thyroidectomy. Total thyroidectomy was generally recommended for patients with cytology positive for malignancy.
Before 2012, nodules with AUS cytology underwent repeat FNA, and lobectomy was recommended for nodules with persistent AUS cytology. Patients with SFN cytology were advised to undergo lobectomy. Starting in 2012, a gene expression classifier (GEC; Afirma, Veracyte, Inc.) was used in some adult nodules with indeterminate cytology based on an individualized assessment of risk, physician discretion, and patient preference. If the GEC was benign, observation was recommended; if suspicious, surgical resection was recommended (lobectomy or total thyroidectomy based on individual patient factors). GEC was not used for pediatric nodules.
Data collection
Patient characteristics were abstracted from the medical record. For each nodule, sonographic features (including size, cystic content [solid, 1–25%, 25–50%, 50–75%, or >75%], and the presence of other nodules), cytopathology, and histopathology were collected from the medical record. Because noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) was not defined during the majority of the study period, nor in pediatric patients, nodules interpreted as NIFTP on histopathology were considered malignant. Although all nodules were ≥1 cm in diameter on ultrasound evaluation, malignancies measuring <1 cm on histological examination were classified as microcarcinomas. All nodules in each patient were analyzed individually. For nodules that underwent repeat FNA for any reason, the final cytology was defined as the result of the most recent FNA performed on that nodule. A nodule was considered to be in a multinodular gland if at least one other nodule ≥1 cm was present. Ultrasound images were reviewed by a radiologist (D.M.R.) if needed to clarify ultrasound reports. Study data were managed using REDCap Software Version 7.6.1 (13) and FileMaker Pro Version 11 (FileMaker, Inc.).
Statistical analysis
The prevalence of nodules in each cytological category was compared between pediatric (<19 years) and adult (≥19 years) nodules by χ2 or Fisher's exact test. Differences between groups in other categorical variables were similarly assessed. Pearson correlation was used to assess the relationship between the percentage of nodules with nondiagnostic cytology and the percentage of biopsied nodules that were ≥50% cystic in each calendar year. Malignancy rates in each cytological category were compared between pediatric and adult nodules. In the primary analysis, malignancy rates were defined as the number of histologically malignant nodules in each cytological category divided by the total number of nodules in that category. Malignancy rates were also calculated among the subgroup of nodules that were resected; however, this was performed as a secondary analysis because—given the bias toward resecting nodules judged more likely to be malignant for reasons other than cytology (4)—the comparison of malignancy rates only among resected nodules might be confounded by differences in nodule resection rates between children and adults. Two-tailed p-values <0.05 were considered significant, and p-values were not adjusted for multiple comparisons. This study was approved by the Institutional Review Boards of both institutions.
Results
Patient and nodule characteristics
Four hundred thirty nodules in 334 pediatric patients and 13,415 nodules in 9220 adult patients were analyzed. Baseline characteristics of patients and nodules are shown in Table 1. There were no significant differences in patient sex or nodule size between children and adults. Adult patients were more likely than pediatric patients to have multiple nodules (47% vs. 27%, p < 0.0001). A greater proportion of biopsied nodules were ≥50% cystic in the pediatric cohort than in the adult cohort (27% vs. 11%, p < 0.0001).
Table 1.
Characteristics of Patients and Thyroid Nodules
| Pediatric (<19 years) | Adult (≥19 years) | ||
|---|---|---|---|
| Patients | 334 | 9220 | p |
| Female sex, n (%) | 277 (83) | 7717 (84) | 0.71 |
| Age, years | |||
| Mean (SD) | 15.1 (2.8) | 52.7 (14.8) | |
| Median (IQR) | 15.5 (13.7–17.2) | 53 (42–64) | |
| 0–10 | 29 (9) | — | |
| 11–14 | 101 (30) | — | |
| 15–18 | 204 (61) | — | |
| 19–39 | — | 1839 (20) | |
| 40–59 | — | 4267 (46) | |
| ≥ 60 | — | 3114 (34) | |
| Multiple nodules present, n (%) | 89 (27) | 4385 (47) | <0.0001 |
| Nodules | 430 | 13,415 | p |
| Nodule size, mm | |||
| Mean (SD) | 23 (12) | 23 (14) | |
| Median (IQR) | 20 (13–29) | 19 (14–28) | |
| 10–19 | 204 (47) | 7040 (52) | |
| 20–29 | 128 (30) | 3445 (26) | 0.18 |
| 30–39 | 53 (12) | 1568 (12) | |
| ≥40 | 45 (10) | 1362 (10) | |
| Cystic content | |||
| Solid | 134 (31) | 5420 (40) | |
| <25% | 87 (20) | 4813 (35) | |
| 25–50% | 95 (22) | 1735 (14) | <0.0001 |
| 50–75% | 50 (12) | 689 (>5) | |
| >75% | 64 (15) | 758 (6) | |
IQR, interquartile range; SD, standard deviation.
Initial and final cytology in pediatric nodules
Initial FNA cytology was adequate for interpretation in 75% of pediatric nodules (324/430) and nondiagnostic in 25% (106/430; Table 2). Final FNA cytology was benign in 64% of pediatric nodules and positive for malignancy in 7%. Among nodules with indeterminate final cytology, most were AUS (7%) or SUSP (6%), and a small fraction of them were SFN (3%).
Table 2.
Initial and Final Cytological Diagnoses in Pediatric Thyroid Nodules with Diameter ≥1 cm
| No. of nodules (%) | ||
|---|---|---|
| Cytology | Initial diagnosis | Final diagnosis |
| ND | 106 (25) | 53 (12) |
| Benign | 226 (53) | 275 (64) |
| AUS | 39 (9) | 32 (7) |
| SFN | 9 (2) | 14 (3) |
| SUSP | 22 (5) | 26 (6) |
| Positive for malignancy | 28 (7) | 30 (7) |
| Total | 430 | 430 |
AUS, atypia of undetermined significance; ND, nondiagnostic; SFN, suspicious for follicular neoplasm; SUSP, suspicious for papillary carcinoma.
Among nodules with nondiagnostic initial cytology, 73/106 (69%) underwent repeat FNA, which yielded sufficient material for diagnosis in 79% of cases (58/73; Table 3). Cytology on reaspiration was benign in 70% (51/73), indeterminate in 10% (7/73), and malignant in no cases. Of nodules with initial AUS cytology, 29/39 (74%) underwent repeat FNA (Table 3). The most frequent result on repeat FNA was persistent AUS cytology (13/29, 45%) or a higher grade abnormality (8/29, 28%), while 28% (8/29) were cytologically benign.
Table 3.
Final Cytological Diagnoses on Repeat Fine-Needle Aspiration in Pediatric Thyroid Nodules Initially Diagnosed as Nondiagnostic or as Atypia of Undetermined Significance
| Initial cytological diagnosis | ||
|---|---|---|
| Final cytological diagnosis | ND | AUS |
| ND | 15 (21) | 0 (0) |
| Benign | 51 (70) | 8 (28) |
| AUS | 4 (5) | 13 (45) |
| SFN | 1 (1) | 3 (10) |
| SUSP | 2 (3) | 4 (14) |
| Positive for malignancy | 0 (0) | 1 (3) |
| Total | 73 | 29 |
Final cytology in adult and pediatric nodules
Final cytological diagnoses in pediatric and adult nodules are compared in Table 4. Nondiagnostic cytology was twice as common in pediatric compared with adult nodules (12% vs. 6%, p < 0.0001), whereas benign (64% vs. 71%, p < 0.0001) and SFN (3% vs. 6%, p = 0.01) cytology was less frequent in children. There were no differences in the prevalence of AUS, SUSP, or malignant cytology.
Table 4.
Correlation of Final Cytology with Histological Diagnosis in Pediatric and Adult Thyroid Nodules
| Pediatric (<19 years) | Adult (≥19 years) | Pediatric vs. adult malignancy rate | Bethesda implied riska | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Total | Resected | Malignant | Total | Resected | Malignant | Malignant | |||||||
| Final cytology | n (%) | n (%) | n | % of resected | % of total | n (%) | n (%) | n | % of resected | % of total | p % of resected |
p % of total |
% of total |
| ND | 53 (12) | 20 (38) | 6 | 30 | 11 | 835 (6) | 192 (23) | 34 | 18 | 4 | 0.23 | 0.03 | 5–10 |
| Benign | 275 (64) | 76 (28) | 2 | 3 | 0.7 | 9487 (71) | 1289 (14) | 96 | 7 | 1 | 0.11 | 0.001 | 0–3 |
| AUS | 32 (7) | 26 (81) | 14 | 54 | 44 | 907 (7) | 530 (58) | 199 | 38 | 22 | 0.10 | 0.004 | 10–30 |
| SFN | 14 (3) | 14 (100) | 10 | 71 | 71 | 829 (6) | 653 (79) | 236 | 36 | 28 | 0.007 | 0.001 | 25–40 |
| SUSP | 26 (6) | 25 (96) | 19 | 76 | 73 | 669 (5) | 626 (94) | 455 | 73 | 68 | 0.82 | 0.67 | 50–75 |
| Positive for malignancy | 30 (7) | 29 (97) | 29 | 100 | 97 | 688 (5) | 660 (96) | 651 | 99 | 95 | 1 | 1 | 97–99 |
| Total | 430 | 190 (44) | 80 | 43 | 19 | 13,415 | 3950 (29) | 1671 | 42 | 12 | 1 | 0.0002 | |
From Cibas and Ali (4).
To examine the reason for the higher rate of nondiagnostic cytology in the pediatric cohort, we analyzed the relationship between the rate of nondiagnostic cytology and the prevalence of predominantly cystic nodule content among nodules undergoing FNA. The rate of nondiagnostic cytology was significantly higher in nodules that were ≥50% cystic than in nodules that were <50% cystic among both children (25.4% vs. 7.6%, p < 0.0001) and adults (21.5% vs. 4.4%, p < 0.0001; Table 5). In addition, the annual proportion of nodules with nondiagnostic cytology correlated closely with the proportion of biopsied nodules that were ≥50% cystic in children (r = 0.74, R2 = 0.54, p < 0.001), as well as in adults (r = 0.48, R2 = 0.23, p = 0.04). Among predominantly cystic nodules, rates of nondiagnostic cytology did not differ between children and adults; however, nodules that were <50% cystic were more likely to be nondiagnostic in children (7.6% vs. 4.4%, p = 0.006). Over the 19 years of practice studied, the rate of initial nondiagnostic cytology decreased in the last 3 years compared with the preceding 16 years (13.1% vs. 27.1%, p = 0.02). In the last three years of this study, the rate of final nondiagnostic cytology among children was 7.8%, which is similar to the rate in our adult cohort (6%, p = 0.55) and to the nondiagnostic rate reported in other large adult studies (2).
Table 5.
Comparison of Nodule Composition and Nondiagnostic Cytology Rates in Children and Adults
| Pediatric (<19 years) | Adult (≥19 years) | p | |||
|---|---|---|---|---|---|
|
Nodules n (%) |
ND n (%) |
Nodules n (%) |
ND n (%) |
% ND | |
| <50% cystic | 316 (73) | 24 (7.6) | 11,968 (89) | 524 (4.4) | 0.006 |
| ≥50% cystic | 114 (27) | 29 (25.4) | 1447 (11) | 311 (21.5) | 0.32 |
| Total | 430 | 53 (12.3) | 13,415 | 835 (6.2) | <0.0001 |
Malignancy rates among adult and pediatric nodules
Malignancy was diagnosed in 80 pediatric and 1671 adult nodules (Table 4). The overall malignancy rate among all nodules was higher in children (19% [80/430] vs. 12% [1671/13,415], p = 0.0002). Within individual cytological categories, malignancy rates were higher in pediatric compared with adult nodules that were cytologically nondiagnostic (11% vs. 4%, p = 0.03), AUS (44% vs. 22%, p = 0.004), or SFN (71% vs. 28%, p = 0.001) (Fig. 1). Among nodules with benign cytology, there was a small but statistically significant difference between pediatric and adult malignancy rates (0.7% vs. 1%, p = 0.001). Malignancy rates did not differ between children and adults among nodules with SUSP or malignant cytology. Malignancy rates by cytological category in the adult cohort were similar to those recommended by the BSRTC (Table 4) (4).
FIG. 1.
Malignancy rates in pediatric and adult nodules, by cytological category. *p < 0.05. AUS, atypia of undetermined significance; SFN, suspicious for follicular neoplasm; SUSP, suspicious for papillary carcinoma.
Surgical resection was performed in 190/430 pediatric nodules (44%) and 3950/13,415 adult nodules (29%, p < 0.0001). Pediatric nodules were more likely to be resected than adult nodules in the cytological categories of nondiagnostic (38% vs. 23%, p = 0.01), benign (28% vs. 14%, p < 0.0001), and AUS (81% vs. 58%, p = 0.01). Rates of resection did not differ between children and adults for SFN, SUSP, or malignant nodules. The malignancy rate among resected nodules was similar in children and adults (43% vs. 42%, p = 1). Among nodules with SFN cytology that were resected, pediatric nodules had a higher rate of malignancy than adult nodules (71% vs. 36%, p = 0.007). In all other cytological categories, similar rates of malignancy were observed between pediatric and adult nodules that were resected.
Among nodules with a final cytological diagnosis of AUS, the difference in malignancy rates between pediatric and adult nodules was significant when calculated among all nodules (44% vs. 22%, p = 0.004) but not among the subset of resected nodules (54% vs. 38%, p = 0.1). We considered the possibility that this discrepancy may be due to the lower rate of resection of AUS nodules in adults—particularly after the introduction of a GEC with high negative predictive value (14)—leading to disproportionate inflation of the malignancy rate among resected adult relative to pediatric AUS nodules. Using GEC data available beginning in 2012, we performed a sensitivity analysis in which all unresected adult nodules with AUS cytology that were GEC-benign (n = 133) were considered histologically benign. This assumption was based on the actual malignancy rate of 0/18 (0%) among resected GEC-benign, cytologically AUS nodules in our cohort. In this sensitivity analysis, the malignancy rate among AUS nodules was significantly higher in children than in adults (54% vs. 30%, p = 0.01). This result was not significantly altered (54% vs. 31%, p = 0.02) by assuming a higher malignancy rate of 5% among GEC-benign, cytologically AUS nodules, based on published data that included our institution (14).
Among adult nodules with AUS cytology that were resected, nodules with cytological atypia had a twofold higher rate of malignancy than nodules without cytological atypia (45% [156/347] vs. 23% [43/183], p < 0.0001). Similar malignancy rates were observed in pediatric AUS nodules with cytological atypia and without cytological atypia, although the number of nodules was small and the difference was not statistically significant (60% [12/20] vs. 33% [2/6], p = 0.37). Among AUS nodules, there was no difference between children and adults in the prevalence of cytological atypia overall (72% [23/32] vs. 61% [555/907], p = 0.27), nor among the subset of nodules resected (77% [20/26] vs. 65% [347/530], p = 0.29).
The histological diagnoses in children and adults by cytological category are shown in Supplementary Table S1. Most cancers were papillary thyroid carcinomas (≥1 cm) or microcarcinomas (<1 cm). Poorly differentiated and anaplastic thyroid carcinomas were observed in adults but not in children. There was a slight excess of follicular thyroid carcinomas (FTC) among nodules with SUSP cytology in children compared with adults (5.2% [1/19] vs. 1.3% [6/455], p < 0.0001), which was accounted for by a single pediatric case of FTC. Otherwise, no differences were observed in the distribution of cancer types between pediatric and adult nodules, either overall or within cytological categories.
Discussion
The BSRTC is the recommended standard for the interpretation of FNA cytology, which guides the management of pediatric and adult thyroid nodules. However, it remains uncertain whether the implied malignancy risks of its cytological categories derived from adult populations can be extrapolated to children, and prior studies of FNA cytology in pediatric thyroid nodules have been limited by small size and variable cytological interpretations. In this large consecutive cohort of pediatric and adult thyroid nodules, we observed higher malignancy rates in pediatric nodules than in adult nodules within certain BSRTC cytological categories. Specifically, malignancy rates were 2–2.5 times higher in pediatric AUS or SFN nodules than in adult nodules with equivalent cytology. A similarly increased risk was likely also present in cytologically nondiagnostic pediatric nodules. The principal strength of our study is the uniform interpretation of all pediatric and adult cytology by the same group of thyroid cytopathologists, which makes it likely that the discrepancies we observe between children and adults reflect the true differences in malignancy risk rather than variation in cytological interpretation. In addition, the validity of the cytological interpretation in this study is supported by the fact that malignancy rates in our adult cohort were similar to the implied malignancy risks recommended by the BSRTC (4).
These observed differences in malignancy rates provide pediatric-specific data that support a different approach to the management of some pediatric thyroid nodules compared with adult nodules with similar cytology. For nodules with AUS cytology, surveillance without surgery, which may be considered in adults (2), appears inappropriate in children due to the high rate of malignancy. Thus, our findings support the present recommendation for resection of pediatric nodules with persistent AUS cytology (3). Unlike the recommendation in adults to consider repeat FNA of AUS nodules before resection—which provides a more definitive cytological diagnosis in 70–90% of cases (2,15)—present pediatric guidelines recommend resection over repeat FNA for AUS nodules. However, in our pediatric cohort, repeat FNA allowed reclassification to a more definitive cytological category in many AUS nodules (72%). Some nodules (17%) were reclassified to more abnormal cytological categories (SUSP or malignant) that might change the surgical approach from lobectomy to near-total thyroidectomy (3). Conversely, a significant minority (28%) of AUS nodules were reclassified as benign on repeat FNA. We have previously reported a rate of malignancy of 11% (1/9) in such nodules, which may be low enough to consider observation over resection, particularly since no adverse outcomes were observed even when diagnosis of such malignances was slightly delayed (16). Although further data are needed to clarify this point, overall our findings suggest that repeat FNA in pediatric AUS nodules may alter clinical management and may be considered before proceeding to surgery, particularly for patients in whom repeat FNA is straightforward to perform.
Among SFN nodules, we observed a markedly higher malignancy rate in children than in adults. In fact, in our pediatric cohort, the risk of malignancy was similar for SFN and SUSP nodules (71% vs. 73%, p = 1). This does not necessarily imply, however, that the surgical management of pediatric SFN and SUSP nodules should be similar, since the types and risk profiles of malignancies in each cytological category may differ. Until additional data are available to address this issue, the present recommendation for initial lobectomy in children with SFN nodules seems appropriate (3).
Within the most worrisome cytological categories (SUSP and malignant), malignancy rates did not differ between children and adults, supporting the approach that clinical management of these nodules should not differ by age. Similarly, for cytologically benign nodules, the rates of malignancy in children and adults—although statistically different—were clinically indistinguishable, and our previous analysis has shown that the initial management of cytologically benign nodules in children can be similar to that recommended for adults (16).
Among nondiagnostic nodules, we observed a higher rate of malignancy in children among all nodules, but not among resected nodules. This discrepancy may be due to the lower rate of resection in adult nondiagnostic nodules, leading to greater inflation of the apparent malignancy rate among resected nodules in adults compared with children, but we are unable confirm this. We therefore consider this finding to represent a likely increased risk of malignancy in pediatric nondiagnostic nodules. The reason for this is not clear, and the small number of pediatric malignancies in this cytological category makes further subgroup analysis unreliable. Nevertheless, cytologically nondiagnostic nodules in children appear to have a substantial risk of malignancy, and their resection should be considered. Moreover, all six malignant nodules in this group were either large (>4 cm in diameter; n = 3) or entirely solid (n = 3), so cytologically nondiagnostic pediatric nodules with these characteristics should be strongly considered for resection.
It is notable that we observed a twofold higher rate of nondiagnostic cytology in pediatric nodules compared with adult nodules. The most important cause appears to be the higher proportion of biopsied nodules in children that were predominantly cystic (27% vs. 11%), a characteristic that increased the likelihood of nondiagnostic cytology in our and other cohorts (17). However, the rate of nondiagnostic cytology was higher in children even among predominantly solid nodules, implying that other unknown factors besides cystic content are involved as well. Among pediatric nodules with initial nondiagnostic cytology, repeat FNA provided a cytological diagnosis in the majority of cases (79%), which supports the present recommendation for repeat FNA of these nodules before considering resection.
The design of this study, with uniform review of all pediatric and adult cytology by a single group of cytopathologists, makes it highly likely that the discrepant malignancy rates between children and adults are due to differences in biology rather than in cytological interpretation. However, the specific reason for these differences remains unclear. As has been previously reported (18,19), cytological atypia was associated with an increased risk of malignancy in our adult cohort, and a similar trend was observed in our pediatric cohort. However, cytological atypia was similar in prevalence between pediatric and adult nodules, and therefore cannot explain the discrepancy in malignancy rates. Similarly, the distribution of histological subtypes of thyroid cancer was similar between children and adults. We hypothesize that the higher malignancy rates in cytologically indeterminate pediatric compared with adult nodules may be caused by a greater prevalence in adults of one or more benign conditions capable of causing cytological abnormalities that may be interpreted as AUS or SFN. Autoimmune thyroiditis is a potential candidate for such a condition, as thyroid autoimmunity has been suggested to mimic abnormal cytology in patients with indeterminate thyroid nodules (20) and is more prevalent in adults than in children.
This study has several limitations. The small number of nodules with AUS or SFN cytology in our pediatric cohort may have limited our ability to accurately define the malignancy rates in children and adolescents with such indeterminate cytology, and therefore to compare these rates with those in adults. Our findings suggest the need for further investigation of larger series of pediatric patients with nodules of indeterminate cytology. Because BCH is a tertiary referral center, it is possible that referral bias may have caused overestimation of the prevalence of malignancy in our pediatric data cohort. However, only 6% of pediatric patients had an FNA attempted at another institution before presentation, so any such bias is likely minimal. In addition, although we did not exclude patients with genetic risk factors for thyroid neoplasia (mutations in PTEN, DICER1, or APC) or previous neck irradiation, these patients comprised only a small percentage of the pediatric cohort (4.5% and 6.6%, respectively). Although FNA cytology for all patients was interpreted by a single cytopathology group, histopathology of resected specimens was interpreted by different pathology groups for children (BCH) or adults (BWH), which could cause discrepancies in rates of malignancy by age. However, because most cases with difficult or borderline histopathology are co-reviewed by pathologists at both institutions, we expect such bias to be minimal.
Due to limitations of data available over the study period, we were unable to analyze specific histological subtypes of thyroid cancer such as encapsulated versus invasive follicular variant of papillary thyroid carcinoma. Moreover, NIFTP recently has been reclassified in adults as an indolent tumor (21). Because this classification did not exist for the majority of the study period and is not yet well defined in children (22,23), we chose to classify NIFTP as a malignancy for this analysis. Unless NIFTP is more common in children than adults—which seems unlikely—classifying NIFTP as a malignancy should bias toward a similar or higher observed malignancy rate in adults, and therefore should underestimate the true difference in malignancy rate between children and adults.
In conclusion, our findings demonstrate that, using the accepted standard of the BSRTC, malignancy rates are higher in pediatric thyroid nodules compared with adult thyroid nodules with equivalent cytological findings of AUS and SFN, and likely among cytologically nondiagnostic nodules. While the reasons for these discrepancies in malignancy risk remain unclear, they are not due to differences in cytological interpretation. These findings provide pediatric-specific data needed to support and refine present consensus recommendations for the optimal management of thyroid nodules in children, which may differ from that of adult nodules with similar cytology.
Supplementary Material
Acknowledgment
This work was supported by the NIH grant T32-DK007699 (C.E.C.).
Author Disclosure Statement
No competing financial interests exist.
Supplementary Material
References
- 1. Cibas ES, Ali SZ; NCI Thyroid FNA State of the Science Conference 2009. The Bethesda system for reporting thyroid cytopathology. Am J Clin Pathol 132:658–665 [DOI] [PubMed] [Google Scholar]
- 2. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, Pacini F, Randolph GW, Sawka AM, Schlumberger M, Schuff KG, Sherman SI, Sosa JA, Steward DL, Tuttle RM, Wartofsky L. 2016. 2015 American Thyroid Association Management Guidelines for adult patients with thyroid nodules and differentiated thyroid cancer: the American Thyroid Association Guidelines task force on thyroid nodules and differentiated thyroid cancer. Thyroid 26:1–133 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Francis GL, Waguespack SG, Bauer AJ, Angelos P, Benvenga S, Cerutti JM, Dinauer CA, Hamilton J, Hay ID, Luster M, Parisi MT, Rachmiel M, Thompson GB, Yamashita S; American Thyroid Association Guidelines Task Force 2015. Management guidelines for children with thyroid nodules and differentiated thyroid cancer. Thyroid 25:716–759 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Cibas ES, Ali SZ. 2017. The 2017 Bethesda system for reporting thyroid cytopathology. Thyroid 27:1341–1346 [DOI] [PubMed] [Google Scholar]
- 5. Amirazodi E, Propst EJ, Chung CT, Parra DA, Wasserman JD. 2016. Pediatric thyroid FNA biopsy: outcomes and impact on management over 24 years at a tertiary care center. Cancer Cytopathol 124:801–810 [DOI] [PubMed] [Google Scholar]
- 6. Partyka KL, Huang EC, Cramer HM, Chen S, Wu HH. 2016. Histologic and clinical follow-up of thyroid fine-needle aspirates in pediatric patients. Cancer Cytopathol 124:467–471 [DOI] [PubMed] [Google Scholar]
- 7. Monaco SE, Pantanowitz L, Khalbuss WE, Benkovich VA, Ozolek J, Nikiforova MN, Simons JP, Nikiforov YE. 2012. Cytomorphological and molecular genetic findings in pediatric thyroid fine-needle aspiration. Cancer Cytopathol 120:342–350 [DOI] [PubMed] [Google Scholar]
- 8. Norlen O, Charlton A, Sarkis LM, Henwood T, Shun A, Gill AJ, Delbridge L. 2015. Risk of malignancy for each Bethesda class in pediatric thyroid nodules. J Pediatr Surg 50:1147–1149 [DOI] [PubMed] [Google Scholar]
- 9. Rossi ED, Straccia P, Martini M, Revelli L, Lombardi CP, Pontecorvi A, Fadda G. 2014. The role of thyroid fine-needle aspiration cytology in the pediatric population: an institutional experience. Cancer Cytopathol 122:359–367 [DOI] [PubMed] [Google Scholar]
- 10. Lale SA, Morgenstern NN, Chiara S, Wasserman P. 2015. Fine needle aspiration of thyroid nodules in the pediatric population: a 12-year cyto-histological correlation experience at North Shore-Long Island Jewish Health System. Diagn Cytopathol 43:598–604 [DOI] [PubMed] [Google Scholar]
- 11. Stevens C, Lee JK, Sadatsafavi M, Blair GK. 2009. Pediatric thyroid fine-needle aspiration cytology: a meta-analysis. J Pediatr Surg 44:2184–2191 [DOI] [PubMed] [Google Scholar]
- 12. Cibas ES, Baloch ZW, Fellegara G, LiVolsi VA, Raab SS, Rosai J, Diggans J, Friedman L, Kennedy GC, Kloos RT, Lanman RB, Mandel SJ, Sindy N, Steward DL, Zeiger MA, Haugen BR, Alexander EK. 2013. A prospective assessment defining the limitations of thyroid nodule pathologic evaluation. Ann Intern Med 159:325–332 [DOI] [PubMed] [Google Scholar]
- 13. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. 2009. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 42:377–381 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Alexander EK, Kennedy GC, Baloch ZW, Cibas ES, Chudova D, Diggans J, Friedman L, Kloos RT, LiVolsi VA, Mandel SJ, Raab SS, Rosai J, Steward DL, Walsh PS, Wilde JI, Zeiger MA, Lanman RB, Haugen BR. 2012. Preoperative diagnosis of benign thyroid nodules with indeterminate cytology. N Engl J Med 367:705–715 [DOI] [PubMed] [Google Scholar]
- 15. Yassa L, Cibas ES, Benson CB, Frates MC, Doubilet PM, Gawande AA, Moore FD, Jr., Kim BW, Nose V, Marqusee E, Larsen PR, Alexander EK. 2007. Long-term assessment of a multidisciplinary approach to thyroid nodule diagnostic evaluation. Cancer 111:508–516 [DOI] [PubMed] [Google Scholar]
- 16. Cherella CE, Feldman HA, Hollowell M, Richman DM, Cibas ES, Smith JR, Angell TE, Wang Z, Alexander EK, Wassner AJ. 2018. Natural history and outcomes of cytologically benign thyroid nodules in children. J Clin Endocrinol Metab 103:3557–3565 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Alexander EK, Heering JP, Benson CB, Frates MC, Doubilet PM, Cibas ES, Marqusee E. 2002. Assessment of nondiagnostic ultrasound-guided fine needle aspirations of thyroid nodules. J Clin Endocrinol Metab 87:4924–4927 [DOI] [PubMed] [Google Scholar]
- 18. Gan TR, Nga ME, Lum JH, Wong WM, Tan WB, Parameswaran R, Ngiam KY. 2017. Thyroid cytology-nuclear versus architectural atypia within the “Atypia of undetermined significance/follicular lesion of undetermined significance” Bethesda category have significantly different rates of malignancy. Cancer Cytopathol 125:245–256 [DOI] [PubMed] [Google Scholar]
- 19. Shrestha RT, Hennessey JV. 2016. Cytologic subclassification of atypia of undetermined significance may predict thyroid nodules more likely to be malignant at surgery. Diagn Cytopathol 44:492–498 [DOI] [PubMed] [Google Scholar]
- 20. Rago T, Di Coscio G, Ugolini C, Scutari M, Basolo F, Latrofa F, Romani R, Berti P, Grasso L, Braverman LE, Pinchera A, Vitti P. 2007. Clinical features of thyroid autoimmunity are associated with thyroiditis on histology and are not predictive of malignancy in 570 patients with indeterminate nodules on cytology who had a thyroidectomy. Clin Endocrinol (Oxf) 67:363–369 [DOI] [PubMed] [Google Scholar]
- 21. Nikiforov YE, Seethala RR, Tallini G, Baloch ZW, Basolo F, Thompson LD, Barletta JA, Wenig BM, Al Ghuzlan A, Kakudo K, Giordano TJ, Alves VA, Khanafshar E, Asa SL, El-Naggar AK, Gooding WE, Hodak SP, Lloyd RV, Maytal G, Mete O, Nikiforova MN, Nose V, Papotti M, Poller DN, Sadow PM, Tischler AS, Tuttle RM, Wall KB, LiVolsi VA, Randolph GW, Ghossein RA. 2016. Nomenclature revision for encapsulated follicular variant of papillary thyroid carcinoma: a paradigm shift to reduce overtreatment of indolent tumors. JAMA Oncol 2:1023–1029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Mariani RA, Kadakia R, Arva NC. 2018. Noninvasive encapsulated follicular variant of papillary thyroid carcinoma: should it also be reclassified in children? Pediatr Blood Cancer 65:e26966. [DOI] [PubMed] [Google Scholar]
- 23. Rossi ED, Mehrotra S, Kilic AI, Toslak IE, Lim-Dunham J, Martini M, Fadda G, Lombardi CP, Larocca LM, Barkan GA. 2018. Noninvasive follicular thyroid neoplasm with papillary-like nuclear features in the pediatric age group. Cancer Cytopathol 126:27–35 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.