Data Sources and Study Subject s
The Surveillance, Epidemiology, and End Results (SEER) database was used to identify individuals in the United States who were diagnosed with papillary thyroid cancer (International Classification of Diseases for Oncology [ICD-O] histology codes 8050/3, 8340/3, 8341/3, 8342/3, 8343/3, and 8344/3), follicular thyroid cancer (ICD-O histology codes 8330/3, 8331/3, 8332/3, 8335/3, and 8337/3), and Hurthle cell thyroid cancer (ICD-O histology code 8290/3) in 17 SEER regions between 1988 and 2003. The 17 SEER regions are Atlanta, Connecticut, Detroit, Hawaii, Iowa, New Mexico, San Francisco-Oakland, Seattle-Puget Sound, Utah, Los Angeles, San Jose-Monterey, Alaska, rural Georgia, greater California (excluding San Francisco/Los Angeles/San Jose), Kentucky, Louisiana, and New Jersey. Collectively, these regions represent approximately 26% of the total US population. The SEER population is comparable to the general US population with regard to poverty and education level, but it is more urban and has a greater proportion of foreign-born individuals.9
In the current retrospective cohort study, all patients who had histologically confirmed DTC were identified. Patients aged <45 years who were missing data regarding surgery and American Joint Committee on Cancer (AJCC) stage, and whose DTC diagnosis was established by “autopsy only” or ”death certificate only” were excluded. The registries collect data regarding patient demographics, primary tumor site, histology, stage at diagnosis, and first course of treatment (surgery and/or radiation therapy); all of these data were included in our analyses.
Associations between the demographic, clinical, and pathologic characteristics of patients and the likelihood of receiving surgical care or RAI were assessed. Independent demographic variables included age at diagnosis, sex, race/ethnicity (white, black, Asian/Pacific Islander, other), Hispanic origin, marital status at diagnosis, year of diagnosis (1988-1991, 1992-1995, 1996-1999, or 2000-2003), and geography (Northeast, Midwest, South, or West). Clinical variables included the number of primary tumors, surgical therapy (none, lobectomy/isthmusectomy, or total/near-total thyroidectomy), reason for no surgery (not recommended, contraindicated, or patient/guardian refusal), lymph node dissection (none, limited, modified radical, or radical lymph node dissection), and radiation therapy (none, radioactive isotopes, or other radiation). Data regarding lymph node dissection were available only from 1998 through 2002. We assumed that radioactive isotopes were RAI for all patients with thyroid cancer. Pathologic variables included disease stage at diagnosis, tumor size, tumor extension/capsular invasion, lymph node metastases (none, ipsilateral cervical, bilateral/contralateral/midline cervical, mediastinal, or distant), histology (papillary, follicular, or Hurthle cell), the number of regional lymph nodes examined, and the number of metastatic lymph nodes removed. Tumor stage was determined according to the TNM classification system sponsored by the AJCC (AJCC Staging Manual, sixth edition).10 For patients with DTC, different staging criteria are used for patients aged <45 years and patients aged ≥45 years, so patients aged <45 years were excluded from the current analysis. The 2006 ATA management guidelines for patients with thyroid nodules and DTC state that only nodules that measure >1 cm should be evaluated, because they have potential clinical significance.8 Therefore, we excluded patients who had DTC <1 cm.
With regard to appropriate surgery, the ATA guidelines state in Recommendation 26 that most patients who have thyroid cancer should undergo initial near-total or total thyroidectomy; whereas thyroid lobectomy alone may be sufficient for patients who have small, low-risk, isolated, intrathyroid papillary carcinomas without cervical lymph node metastases. According to Recommendation 27, patients who have papillary thyroid carcinoma and suspected Hurthle cell carcinoma should be considered for prophylactic central-compartment (level VI) lymph node dissection. With regard to RAI, Recommendation 32 recommends radioiodine ablation for patients with stage III and IV disease, for all patients with stage II disease aged <45 years, for most patients with stage II disease aged ≥45 years, and for selected patients with stage I disease, especially those who have with multifocal disease, lymph node metastases, extrathyroid or vascular invasion, and/or more aggressive histologies.8
Bivariate analyses of the independent variables with our outcomes of interest were performed by using chi-square statistical analyses for categorical variables and analyses of variance for continuous variables. Multivariate logistic regression models with a backward elimination procedure were used to adjust for patient demographic, clinical, and pathologic characteristics. Survival was calculated as the time (in years) from diagnosis until death, last follow-up, or December 31, 2003, whichever came first. Because SEER does not report the cause of death for individual patients, overall mortality was used for the survival analysis. For the bivariate analysis, the Kaplan-Meier method and the log-rank test were used to identify significant prognostic factors. For the multivariate analysis, a Cox proportional hazards regression model was used for variables that were identified as significant in the bivariate analysis. The strength of association between each predictor variable and the outcome of interest was expressed as an odds ratio (OR) for the likelihood of receiving treatment or as a hazard ratio (HR) for survival, which is presented along with a 95% confidence interval. Data analyses and management were performed using SPSS version 16.0 (SPSS, Inc., Chicago, Ill). All tests were 2-sided, and statistical significance was set at a probability value (P) of ≤.05. This study was granted an exemption from Institutional Review Board approval at our institution, because SEER is a public database with no personal identifying information.