2017 Sec 1 Green Book

S. Dermody et al. / International Journal of Pediatric Otorhinolaryngology 89 (2016) 121 e 126

2. Materials and methods

most common head and neck malignancy in the pediatric popu- lation, accounting for 21% of all head and neck cancer in pediatric patients [3] . Differentiated thyroid carcinomas in children include papillary, follicular, and variant subtypes. Speci fi cally, the papillary subtype accounts for approximately 90% of pediatric thyroid car- cinomas and maintains an excellent prognosis, which parallels fi ndings in the adult literature [3] . The overall prognosis of pediatric thyroid cancer may be skewed due to the large number of papillary thyroid cancer diagnoses, which inherently yield excellent out- comes [4] . Very few studies have strati fi ed each variant subtype by disease speci fi c survival to determine if this statistical fi nding is, in fact, true [6] . Of the differentiated thyroid carcinomas, follicular subtype is less common in children, but is considered more aggressive and maintains a poorer prognosis if vascular invasion is present [3] . Also of note, medullary thyroid carcinoma in pediatric patients is asso- ciated with markedly shorter mean survival even if individual risk factors such as metastatic spread or lymphovascular invasion are absent [6] . Among those diagnosed with thyroid carcinomas, pe- diatric patients present with more advanced manifestation of dis- ease than adult patients, yet mortality is comparatively infrequent [4,6] . In the pediatric population, thyroid carcinomas typically present in teenage years, especially in Caucasian females, with a mean age of diagnosis of 16 years [6] . Previous studies have char- acterized the female predominance of pediatric thyroid carcinomas and have identi fi ed papillary carcinoma as the most common his- tological subtype within this population [3,6] . Currently, the American Thyroid Association Guidelines dictate that there is no indicated disease screening for thyroid cancer in the pediatric population aside from genetic testing and periodic ultrasound evaluation for patients with mutated genes such as BRAF, RET oncogene mutations (MEN2A/2B), Cowden Syndrome, Werner Syndrome and PTEN Related Syndromes. Further consid- erations include patients who received radiation for Hodgkin Lymphoma, Leukemias and CNS Tumors especially with radiation doses between 20 and 29 Gy [7 e 9,11] . The American Thyroid As- sociation Task Force on Pediatric Thyroid Cancer reports insuf fi - cient evidence for utilizing ultrasound to screen for non-palpable thyroid nodules in patient populations not already described due to the likelihood of false positives associated with an enlarged thymus or simple cysts [9,11] . As such, pediatricians and otolaryngologists are forced to rely on early identi fi cation of neck masses by physical examination in order to yield the excellent prognosis for pediatric patients currently described in the literature [9] . While mortality is rare, past studies have identi fi ed factors associated with poorer prognosis of well-differentiated pediatric thyroid carcinomas, such as presence of distant metastases, large primary tumor size, lym- phovascular invasion, and male sex [5,6] . Interestingly, one recent study has demonstrated that the V600 BRAF mutation implicated in adult papillary thyroid carcinoma does not signi fi cantly contribute to the development of pediatric thyroid cancer at the same rate as currently described in older cohorts [10] . The fi ndings of this study are signi fi cant because they provide primary care providers with a more re fi ned risk factor approach when evalu- ating patients [8] . This population-based analysis is imperative since children maintain a greater chance of recurrence of differentiated thyroid cancer compared to the adult population [9] . Thus, our group analyzed the new SEER Database updates provided up until the year 2012 and compared them to the current literature to deter- mine if there have been any signi fi cant changes in incidence and disease speci fi c survival based on the various risk factors reported in previous studies [10] . This study provides the most recent analysis of the SEER database with respect to pediatric thyroid carcinoma.

2.1. Description of source database

This manuscript was deemed exempt from Georgetown Uni- versity IRB review due to the use of de-identi fi ed data and was approved for data collection. The Surveillance, Epidemiology and End Results (SEER) Database was queried to identify pediatric pa- tients with pathologically con fi rmed “ thyroid carcinoma ” between 0 and 19 years of age from the years 2007 e 2012. The SEER Data- base, available at: “ seer.cancer.gov ” , provides the public with population-based data regarding cancer incidence, frequency, and disease speci fi c survival data from 1973 to 2012. At this time, the SEER database includes approximately 10% of the United States cancer population and is updated annually by the National Cancer Institute coding technicians. The SEER registry includes patient information from Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco, Louisiana, Seattle, Utah, Los Angeles, Alaska, and San Jose in order to provide its users with a normalized dis- tribution of patient cohorts in terms of geographic location and age groups. The SEERstat analysis program was downloaded from “ seer.- cancer.gov ” as previously described. Our study group utilized the “ Incidence - SEER 18 Regs Research Data þ Hurricane Katrina Impacted Louisiana Cases 1973 e 2012 ” database and performed a multivariable frequency analysis to determine the total number of patients with the below described pediatric thyroid subtypes. Each speci fi c thyroid cancer subtype was queried using the following criteria: “ malignant behavior ” , “ known age ” , “ cases in research database ” for the initial query. Additional clinicopathologic factors including age, sex, ethnicity, extent of disease, and lymphovascular invasion were also included within the aforementioned multivari- able outcomes query. Within the Case Selection Statement, the following search lim- itations were applied to obtain frequency data regarding our pa- tient cohorts: Pediatric Thyroid Subtypes, “ Papillary ” , “ Follicular ” , “ Medullary ” , and “ Papillary Follicular Variant. ” Coding for each of the described thyroid carcinoma subtypes were collected from the SEER Database Manual and inputted into the Case Selection Com- mand Feature [9] . Tumor recurrence events were not available in the SEER database at the time the query was performed and consequently were not included. Patients that expired within the fi rst four months after surgery were excluded since these in- dividuals were likely unable to complete a full course of radioactive iodine uptake or external beam therapy. Patients with uncoded primary tumor subsite were excluded from this study as well as those with anaplastic thyroid carcinomas and non-epithelial can- cers, such as lymphomas, due to paucity of data. The cohort was then further subdivided according to treatment modality into a “ surgery alone ” group and “ surgery with adjuvant radiation ” group, which included individuals who received radioactive iodine uptake as well as those who received external beam therapy. 2.3.1. Fifteen-year disease speci fi c survival curves Fifteen-year disease speci fi c curves were generated for the surgery group and surgery with adjuvant radiation group utilizing the SEERstat, “ Survival Session ” search query. Within the survival session, the Observed Survival “ Method ” was employed to include fi fteen years with intervals of twelve months. Exclusion criteria included: “ Alive with No Survival Time ” and “ All Death Certi fi cate 2.2. Patient cohort selection 2.3. Statistical analysis

166