September 2019 HSC Section 1 Congenital and Pediatric Problems

Otolaryngology–Head and Neck Surgery 157(4)

radiologic, and pathologic findings known to be associated with increased risk in adults. 12 There are currently no similar diagnostic tools available for the assessment of nodular disease in children and adoles- cents. As such, in the evaluation of patients presenting with thyroid nodules, the clinician is faced with the particular challenge of differentiating benign from malignant disease to weigh the benefits of observational management versus surgical intervention, which harbors the risk of hypopar- athyroidism and damage to the recurrent laryngeal nerve. 13 The American Thyroid Association (ATA) has recognized this difficult task and recently published guidelines for the management of pediatric thyroid nodules. 14 The validity of the MTNS in this age group remains unclear, and certain adult criteria may need to be revised for the scoring system to be an accurate predictor of malig- nancy in the pediatric population. Therefore, we have con- ducted a pilot study to investigate the clinical value of a comprehensive risk assessment tool within this population, and herein we evaluate the ability of a modified MTNS to stratify the risk of thyroid nodules in children. Materials and Methods Institutional Review Board approval was sought and obtained from Connecticut Children’s Medical Center for this study. A retrospective review was performed to identify patients 18 years of age presenting with solitary or domi- nant thyroid nodules who underwent thyroidectomy at Connecticut Children’s Medical Center from September 2008 to December 2015. Data necessary to complete the cumulative MTNS ( Table 1 ) were abstracted from medical charts and managed with REDCap (Research Electronic Data Capture) tools hosted at the University of Connecticut. 15 Patients with incomplete data to formulate the MTNS were excluded. All grayscale and color Doppler ultrasounds of the thyr- oid were reviewed and interpreted in 3 dimensions for con- sistency by a radiologist (D.M.) blinded to final pathology. The following guidelines, outlined by the ATA, 16 were used to classify radiographic criteria: Nodule echogenicity was assessed with respect to the surrounding thyroid tissue or strap muscles. Increased nodule vascularization was defined as an amplified intranodular flow, in comparison with normal thyroid tissue, during color Doppler examination. Shape was categorized as taller than wide when measured in the transverse plane. Microcalcifications were defined as multiple punc- tate bright echoes \ 2 mm, with or without acoustic shadowing. All calcifications not meeting these cri- teria were defined as macrocalcifications. Increasing nodule size was assessed and categorized as either . 10% growth in 2 dimensions or . 30% growth in 2 dimensions, representing 20% and 70% volume increase in an elliptical nodule, respectively.

Lymph nodes were categorized as abnormal if 1 of the following characteristics were present: increased size, increased Doppler flow, cystic appearance, rounded shape, loss of echogenic fatty hilum, irregular margins, or internal calcifications. The MTNS was modified for the population in our study and is displayed in Table 1 . The modified MTNS was developed by a multidisciplinary team from the McGill University Health Center and Connecticut Children’s Medical Center, consisting of 2 pediatric endo- crinologists, 1 pediatric radiologist, 2 otolaryngologists (head and neck and pediatric subspecialist), 1 pediatric general surgeon, and 1 thyroid pathologist. Age . 45 years was excluded as a risk factor. Given the lack of evidence in the pediatric population, a positive positron emission tomography scan and ethnicity of Filipino, Hawaiian, and Icelandic were excluded in the modified pediatric MTNS. The cytology portion of the MTNS was converted to a com- parable system utilizing the Bethesda System for Reporting Thyroid Cytopathology, which has been described. 17 Categorical diagnoses I through VI included nondiagnostic (I), benign (II), atypia of undetermined significance (AUS; III), suspicious for follicular neoplasm (SFN; IV), suspicious for malignancy (V), and malignant (VI). Nondiagnostic results were assigned a value of 0 points given the low-risk nature of these insufficient samples in adults. It was determined that 3 points would be subtracted for benign cytology (unpublished data, Hua et al) given the low false-negative rate. 18 AUS was assigned 3 points, coin- ciding with mild atypia in the adult MTNS. 12 SFN was assigned 4 points, based on the slightly higher risk for malignancy than AUS in the pediatric population. For the purposes of this study, nodules with AUS and SFN cytology were described as indeterminate. Given the high rates of malignancy in aspirates diagnosed as suspicious for malig- nancy and malignant, these were assigned point values of 7 and 8, respectively. Statistical analysis was performed with SPSS 16 for Windows (IBM Corp, Armonk, New York). The relation- ship between categorical variables, such as final pathology and sex, was evaluated with the Fischer’s exact test. For continuously distributed variables, such as the MTNS and age, groups were compared with the independent sample t test. All tests were 2-tailed, and statistical significance was defined as a P value \ .05. Results Of 46 patients 18 years of age who underwent resection of a dominant thyroid nodule, the mean age at surgery was 14.8 6 2.43 years, with no significant difference in age between those with benign and malignant disease. Within our sample, there was a female predominance (76.1%). Thirty-six patients (78.3%) had benign nodules, while 10 patients (21.7%) had well-differentiated thyroid cancer diag- nosed on final pathology. The histologic subtypes included papillary thyroid carcinoma (n = 8, 80%), follicular variant

156