Follicular thyroid carcinoma (FTC) is the second most common type of thyroid cancer, accounting for 10% to 15% of all thyroid malignancies [1,2]. FTC is more common in areas of iodine-deficient endemic goiter and more prevalent in African Americans than in Asians or Caucasians [3]. Previously, FTC was considered a more aggressive type of cancer, as compared to papillary thyroid cancer (PTC), because more distant metastases occurred in FTC patients [4,5]. However, the prognosis of FTC and PTC patients is similar if their age and initial cancer staging is similar [6,7]. Traditionally, FTC was classified as either widely or minimally invasive [8]. The World Health Organization (WHO) has defined widely invasive FTC (WI-FTC) as a tumor with widespread infiltration of blood vessels and/or adjacent extrathyroid tissue and lacking complete encapsulation [9]. Minimally invasive FTC (MI-FTC) is a grossly encapsulated solitary tumor with unequivocal vascular invasion and/or invasion that penetrates the full thickness of the capsule [9]. Distant metastases are common in patients with WI-FTC, but rare in MI-FTC patients [10].

Hurthle cell carcinoma (HCC), also known as oncocytic or oxyphillic carcinoma is classified by the WHO as a variant of FTC [9,11]. However, recent genome-wide analysis using mutation genotyping suggested that HCC may be a unique thyroid cancer distinct from PTC and FTC [12]. Hurthle cells (oxyphilic cells) are large polygonal cells with distinct borders derived from thyroid follicular epithelium and contain abundant granular cytoplasm because of an excess of mitochondria and a large nucleus with a prominent nucleolus [13]. HCC is defined as a thyroid cancer consisting of at least 75% Hurthle cells and the presence of capsular and/or vascular invasion and can be classified into a minimally or widely invasive type [9].

Data on clinicopathological features and clinical outcomes in patients with FTC or HCC in iodine-sufficient areas are limited in Korea. In this multicenter study we evaluated the clinical behavior and prognostic characteristics of FTC and HCC in Korea. We compared initial pathological parameters and disease-free survival (DFS) between the minimally invasive and widely invasive cancers. We also described the preoperative cytological diagnosis using fine-needle aspiration (FNA) in patients with FTC or HCC.

In this study, we evaluated a total of 563 patients with FTC or HCC. This is the first large multicenter cohort study in an iodine-sufficient area in Korea. HCC patients were older and had more lymphovascular invasions than FTC patients at the initial surgery. However, FTC and HCC patients had similar initial clinicopathological features. There was no significant difference in distant metastasis and DFS between the two groups. Older age, widely invasive type and extrathyroid extension were independent risk factors for distant metastasis in our study cohort. The relative risk of distant metastasis was 17.8-fold in patients older than 60 years, as compared with patients younger than 40 years. Primary tumor size and the presence of cervical LN metastasis were independent risk factors for recurrent FTC and HCC. There was no significant difference in recurrent FTC or HCC between minimally invasive and widely invasive types in our study cohort.

FNAC is the main diagnostic modality for evaluating thyroid nodules due to its accuracy and cost-effectiveness [14]. However, diagnosing FTC or HCC using FNAC is difficult because the risk of malignancy after the result of either follicular or Hurthle cell neoplasm is only 15% to 25% [15,16,17]. Therefore, clinicians attempt to determine diagnostic surgery according to clinical factors other than FNAC results, such as tumor size, the change in tumor size and patient age. In our study cohort, approximately 45% and 33% of patients with confirmed FTC and HCC, respectively, did not have malignant FNAC results. Diagnostic or therapeutic thyroidectomy was performed based on clinical decisions in the majority of these patients. Currently, a diagnostic marker useful for diagnosis of FTC or HCC is not available.

Previously, HCC was thought to have more aggressive clinical features with multifocal tumors, LN metastases and distant metastases compared with other DTCs [11,18,19]. In these studies, the frequency of distant metastases was 15% to 34% [11,18,19]. In a recent and large study using the Surveillance, Epidemiology and End Results (SEER) database, 3,311 patients with HCC were older and had larger tumor size, higher stage and higher disease-specific mortality, as compared with patients with other DTCs [20]. In this study, only 4.7% of patients had distant metastases [20]. In our study cohort, HCC patients were significantly older and had more lymphvascular metastases than FTC patients. However, the frequency of distant metastases and DFS after the first treatment was similar. Distant metastases were present in only 3% of our HCC patients. The frequency of distant metastases was lower in the recent large study and similar to our series [20]. Possibly, thyroid cancer was detected earlier in recent studies, contributing to the smaller number of distant metastases, as compared to historical studies.

Age was an important survival prognostic factor in the DTC and TNM staging system that includes age as a risk factor [9,10,19,21,22,23]. In a population-based study, the mortality HR was 6.35 in DTC patients over 50 years of age, compared with younger patients [20]. Older age was an independent risk factor for distant metastasis in our study cohort, which is in agreement with a previous study. When deciding on an initial therapeutic strategy for FTC and HCC patients, age should be considered as a risk factor.

The MI-FTC and WI-FTC categories defined by WHO classification are generally accepted [9]. However, definition of the capsular invasion extent varies between pathologists. D'Avanzo et al. [22] proposed that MI-FTC should be categorized as FTC with capsular invasion only and moderately invasive FTC should include angioinvasion with or without capsular invasion to emphasize the prognostic importance of vascular invasion. In our study cohort, lymphovascular invasion was not a significant prognostic factor associated with distant metastases or recurrent disease in multivariate analysis and showed significance as a prognostic factor only in the univariate analysis. Therefore, we analyzed our study subjects within minimally invasive and widely invasive subclassifications. The extent of capsular invasion was an important risk factor for distant metastasis and survival in previous studies [18,19,21,22,23,24]. In our study subjects, the widely invasive type of tumor was an important risk factor for distant metastases in FTC and HCC patients. The extent of capsular invasion was a marginally significant risk factor when estimating the recurrence of FTC, even though our median follow-up period was only 6 years. However, therapeutic modalities such as bilateral thyroid surgery and radioiodine ablation therapy could possibly influence DFS between widely invasive and minimally invasive tumors.

This study had several limitations. First, the patient follow-up period after their initial treatment was relatively short and was limited to evaluating cancer-specific survival in FTC and HCC patients. Second, this retrospective study included patients from four tertiary referral centers and, therefore, the possibility of selection bias exists. Additionally, therapeutic strategies between centers may vary. Third, the clinical importance of cervical LN metastasis in FTC and HCC patients could be underestimated because we could not classify Nx from N0 in this retrospective study. However, this is the first multicenter cohort study describing the clinical behaviors of FTC and HCC in Korea. Although HCC patients were older and had more lymphovascular invasions, clinicopathological features of FTC and HCC were relatively comparable. The age of patients, wide invasiveness and extrathyroidal invasion were independent risk factors for predicting distant metastases in FTC and HCC patients in the present study.