Well differentiated thyroid carcinoma (WDTC) is the most common endocrine malignancy regardless of ethnicity or of geographic location. Incidence of WDTC has been increasing in most areas over the world during the recent 3 decades [1,2]. Many doctors believe that such increase is mainly due to increased detection which might have been caused by widespread use of medical imagings including thyroid ultrasonography, by increase in thyroid surgery revealing occult cancers, or by more deliberate examination of surgical specimen obtained by thyroid surgery and so on [1,2,3]. In United States, incidence of thyroid cancer has been increasing continuously from 1980's, and such increase was observed regardless of ethnicity [3]. In Korea, incidence of thyroid cancer increased rapidly from year 2000. By the year of 2010, age-adjusted rate of thyroid cancer was over 80 per 100,000 of population in women, and it was around 20 per 100,000 in men showing the highest incidence in the world [4]. The main histological subtype which has been contributing to surge in thyroid cancer is papillary thyroid carcinoma (PTC) both in Korea and Unites States [5,6].

One of the main reasons for the increased incidence of PTCs seems to be increased detection because PTCs are more easily detected by ultrasonography and subsequent aspiration cytology examinations compared to follicular carcinomas. Such phenomenon promoted a debate on over-diagnosis issue recently. However, the fact that incidence of follicular thyroid carcinoma has been increasing also in United States [7] and the fact that incidence of thyroid cancer in childhood (under age of 20) in which screening of thyroid cancer is not being performed [8] suggest that there is also a true increase in incidence of thyroid cancer in many areas over the world including Korea. However, the reason for such increase in incidence is not clear yet. I describe some aspects of thyroid cancer, especially of papillary cancer which is the most common type of thyroid cancers, based on recent research and clinical papers including those from our thyroid team in Asan Medical Center. It had been presented at annual meeting of Korean Endocrine Society in 2014 in Gwangju, Korea as "Namgok Award Lecture" with a title of "A closer look at PTC."

A recent report based on whole exome sequencing of various human cancers showed that thyroid cancer tissues have very low mean somatic mutation frequency across the genome when compared to cancers from other organs [9]. Cancers which have been considered to be clearly associated with environmental carcinogens such as lung cancers (smoking, etc.) or melanomas (ultraviolet lights, etc.) have very high mean somatic mutation frequencies. It suggests that thyroid epithelial cell has a low threshold for cancer development whatever the causes are. Thyrocytes have high oxidative load as they are continuously producing H₂O₂, a key element for thyroid hormone synthesis [10]. As chronic exposure of cells to pro-oxidant may result in slow accumulation of mutations in genome [11], high oxidative load in thyrocyte may predispose them to cancer development. Actually, serum level of oxidative stress markers is higher in patients with thyroid cancer, and the level of total antioxidant status is lower in them compared to control subjects [12]. Low mutational threshold of thyrocyte for development of cancer and high oxidative load of thyrocytes even in physiologic condition may predispose the thyrocytes to cancer development.

Potential etiologic factors for thyroid cancer are exogenous factors such as radiation exposures, high iodine intake (PTC), nitrate, westernized life style or unknown environmental pollutants and endogenous factors such as high thyroid stimulating hormone (TSH), the presence of Hashimoto's thyroiditis or obesity [13]. However, it is still unclear which factor(s) are more important and which of them might have mainly contributed to recent increase in thyroid cancer incidence. Recently, we reported that obesity is an independent risk factor for thyroid cancer in women when evaluated in routine health checkup setting. This association between risk factor and disease was unrelated to serum insulin and TSH levels [14]. Additional studies are required to understand the mechanism(s) behind the association of obesity with thyroid cancer risk. However, body mass index was not associated with the aggressive clinicopathological features or recurrence-free survivals in patients with PTC [15]. It suggests that obesity may be involved in earlier stage of carcinogenesis rather than affecting progression of thyroid cancer.

We evaluated STAT3 activities which are known to play important roles in progression of many cancers in PTC tissues, and found that STAT3 activities are low when compared to normal surrounding normal thyroid tissue. It was an unexpected, and a very unusual finding. Tumor size negatively correlated with STAT3 activities and such negative correlation was present only in BRAF mutation-positive cases [16]. It suggests that growth-stimulating signal (including BRAF mutation) is associated with decreased STAT3 activities in PTC, which may act as a breaking mechanism to slow down cell growth. Such low STAT3 activities in tumor tissues could be one of mechanisms for slow growth rates and for indolent courses of PTC.

We had found factors predicting Korean PTC patients' clinical course after initial standard treatment (complete surgery and radioiodine therapy [RAI Tx]) by observation of patients over long period [17]. Independent factors predicting cancer specific survival were age of patient (>45 years old), presence of extrathyroidal extension of tumor (microscopic or gross) and the presence of distant metastasis. Presence of node metastasis was not an independent factor for cancer specific survival. Independent factors predicting recurrence free survival were gender (male), presence of extrathyroidal extension of tumor and presence of lymph node metastasis (central or lateral compartment). Among various staging systems, American Thyroid Association risk stratification was the best for prediction of recurrence of PTC.

Additionally, we found that metastatic lymph node ratio (ratio of number of metastatic nodes/resected nodes) and the maximal size of metastatic lesion in lymph node are independent prognostic factor for recurrence. When the ratio is higher than 0.4 or when the maximal size of metastatic lesion is greater than 0.2 cm, the risk of recurrence increases [18]. Presence of BRAF V600E mutation in primary tumor was a poor prognostic marker for Korean PTC patients in terms of recurrence, but was not an independent factor when multivariate analysis using known clinicopathological parameters was done [19].

As the frequency of the BRAF mutation is very high up to 80% in Korean PTC patients, it may not be useful for stratification of risk for recurrences. So, we evaluated the potential role of the X-linked inhibitor of apoptosis protein (XIAP) expression as a novel prognostic marker to predict recurrence, in combination with the BRAF V600E mutational status. The presence of BRAF V600E mutation was significantly associated with higher recurrence rate in subjects with PTCs (hazard ratio [HR], 2.98; P=0.039). PTCs with BRAF V600E mutation, but negative for XIAP expression had a significantly higher rate of recurrent PTC (HR, 4.53; P=0.012). So, we reported that the evaluation of the XIAP expression and BRAF mutational analysis was more useful for the prediction of cancer recurrence in patients with PTC than BRAF genotype alone [20].

Coexistence of chronic lymphocytic thyroiditis was associated with lower recurrence rates in patients with PTC [21]. However, basic mechanism(s) for this association and clinical meaning of this finding is not clear at this point yet.

Initial therapy for PTC is surgery and/or RAI Tx. But, there had been continuing debates on optimal surgical extent or optimal radioiodine dose in individual patient setting.

For initial surgery, rationales for total thyroidectomy are (1) bilateral cancers are common (30% to 85%); (2) recurrent and contralateral lobe disease are not unusual (4.7% to 24.0%), and half of the patients who develop recurrence die from thyroid cancer; (3) survival is improved for lesions of >1.5 cm if total thyroidectomy is performed; (4) efficacy of RAI Tx is greater for both diagnosis and postoperative treatment; (5) serum thyroglobulin (Tg) value becomes a useful postoperative marker only in patient who underwent total thyroidectomy; (6) re-operative thyroid surgery has increased risk of complications. Rationales for less than total thyroidectomy (lobectomy & isthmusectomy) are (1) occult multicentricity is not clinically significant; (2) fewer than 5% of recurrences occur in the thyroid bed which show higher risk; (3) half of the local recurrences can be cured with second surgery; (4) total thyroidectomy does not address micrometastasis to lymph nodes; (5) there is an increased risk of complications with total thyroidectomy.

Only randomized clinical trials will answer the question on optimal extent of surgery in patients with PTCs. However, it has been estimated that randomization of 3,000 to 5,000 patients are required to see any difference in cause-specific mortality, which is not feasible. So, the current debates are unavoidable because we are discussing based on retrospective data only.

We recently evaluated retrospectively 1,700 patients treated for PTC with total thyroidectomy and RAI Tx and followed up for more than 6 years in terms of development of distant metastasis [22]. Among 1,700 patients, 40 patients (2%) were found to have distant metastasis. The cumulative risk of distant metastasis increased with the increase of tumor size and was significantly different according to the N (lymph nose metastasis) staging. The cumulative risk in N1b group showed the steepest increasing pattern according to the increase in tumor size compared to N1a and N0/Nx patients. The threshold of primary tumor size for develoment of distant metastasis was 1 cm in N1b group and was 2 cm in N1a group. So, optimal surgery in individual patient could be determined accordingly based on tumor size and extent of neck node metastasis.

Optimal dose of RAI Tx for individual patient with PTC after initial surgery is also not clear. It also needs prospective studies. However, there is a clear trend toward a more selective use of RAI Tx in PTC patients because (1) not all patients with metastatic diseases are likely to benefit from RAI Tx; (2) I-131 administration is by no means a risk-free procedure; (3) RAI Tx is no longer the only option for metastatic disease.

The current follow-up strategy for the patient with PTC in our institution is as follows. After initial surgery and remnant ablation, patients take suppressive dose of thyroxine. Usually 1 year after the initial treatment, intermediate to high risk group of patients undergo whole body scan along with neck ultrasound (US). Simultaneously, serum stimulated thyroglobulin (sTg) and anti-thyroglobulin antibody (anti-TgAb) are measured. Low risk patient does not undergo diagnostic whole body scan. If abnormal structural disease is found and is considered surgically amenable, reoperation is recommended. If there are vivid iodine uptake on diagnostic WBS with elevated sTg, high dose RAI treatment may be performed. In case of highly elevated sTg level of greater than 10 ng/mL, various imaging studies such as chest computed tomography, 18-fluoro-deoxyglucose positron emission tomography are applied, but mildly elevated sTg which is between 1 and 10 ng/mL warrants regular neck US only with watchful observation. If spontaneous biochemical remission (BR) occurs, then periodic Tg measurements under thyroxine along with neck US are done. If Tg rises, then appropriate imging studies are done [23].

In case of positive TgAb, Tg may be undetectable or falsely low, but changes of TgAb titer itself may be of prognostic significance [24]. In the absence of TgAb, if sTg level at 1 year after the initial treatment is undetectable, there is a very low risk of recurrence and patients may be followed without any further measurement of sTg [25]. Tg level determination under T4 suppression and neck ultrasonography may be enough for such patients. Although sTg level at 1 year after the initial treatment is positive, it may decline spontaneously and some patient may obtain spontaneous BR [26]. The probability and time to spontaneous BR is dependent on the magnitude of elevation of sTg level at 1 year after the initial treatment. Blind high dose radioiodine treatment for patients with elevated sTg level without demonstrable iodine uptake is not recommended [27].

Recurrences occur in 5% to 20% as locoregional form and as distant metastasis in 10% to 20% in long-term follow-up after initial therapy. Soft tissue recurrences as a form of local recurrence require aggressive therapy including wide excision and postoperative adjuvant therapy as they have dismal prognosis.

There are controversies in proper management of locoregional recurrences in neck lymph node, because improvement in clinical outcome of those patients through randomized, prospective study had never been documented and because it is not clear if lymph node recurrences could be a focus of further metastasis of cancer cells. Management includes surgery (compartment-oriented lymph node dissection), alcohol injection or radiofrequency ablation and simple observation. Adjuvant RAI Tx is not useful after re-operation, especially high dose radioiodine had been done as initial therapy [28,29].

Recurrences as distant metastasis require thorough evaluation and proper management according to site and progression of each lesion. Palliative surgery if critical structure is endangered, RAI Tx in 'RAI-avid' lesions, external beam radiation therapy or intravenous bisphosphonate, embolization should be considered in bone metastasis according to clinical setting. RAI-avid lung metastasis can be managed with radioiodine, but there is no available therapeutic modality in 'non-RAI-avid' lung metastatic lesions.

DECISION is phase III study of sorafenib in locally advanced or metastatic patients with radioactive iodine refractory thyroid cancers [30]. It was to compare the effect of sorafenib versus placebo on progression-free survival (PFS) in the patients with RAI refractory progressive differentiated thyroid carcinoma. It enrolled 380 eligible patients and the result has been recently published. In case of progression in patients on placebo, they could be switched to sorafenib treatment according to physicians' decision. The primary end-point, PFS was significantly prolonged with sorafenib use. Patients in sorafenib arm showed median 10.8 months of PFS compared to 5.8 months of PFS in placebo arm. And the difference was highly significant. Partial remission occurred in 12% of the patients who received sorafenib and 73% of the patients in sorafenib arm showed stabilization or partial remission, in contrast to 27% in placebo arm.

Like sorafenib, lenvatinib is a tyrosine kinase inhibitor. It blocks a range of molecular targets, from vascular endothlial growth factor receptor 1 to 3 and fibroblast growth factor receptor 1 to 4 to platelet-derived growth factor receptor β, KIT, and rearranged during transfection. In SELECT study which is an ongoing phase 3 study for levatinib in radioiodine-refractory thyroid cancer, 392 patients randomly received either lenvatinib or placebo [31]. Primary end-point was PFS. If progression was documented in placebo arm, the patient was allowed to switch to lenvatinib treatment. The primary end-point, PFS was significantly prolonged with lenvatinib use. Patients in lenvatinib arm showed median 18.3 months of PFS compared to 3.6 months of PFS in placebo arm. And the difference was highly significant. This result suggests that lenvatinib has a better efficacy than that of sorafenib. This study is ongoing and will be finished by February, 2015.

However, management of radioiodine-refractory metastatic/locally invasive PTC is a challenging task. We are trying to find a novel therapeutic target based on peculiar change(s) in cancer cell metabolism in thyroid cancer.

PTC is one of the most common cancers in human. Recent increase in incidence seems to be mainly due to increased detection, but true increase might have contributed to such increase. High prevalence and low virulence of disease have elicited debates on optimal management in individual patient. Although most cases are curable by traditional treatment modalities, but radioiodine refractory metastatic/recurrent PTC is still a clinical challenge. Only comprehensive understanding of pathogenesis and pathophysiology would answer the question how we should provide optimal management in individual patient based on individual risk for disease progression.