The new 4th edition World Health Organization classification for thyroid tumors, Asian perspectives.

Abstract

The World Health Organization (WHO) classification of tumors serves as an international standard of histopathological diagnosis and the essential basis of clinical practice for neoplastic diseases for all organ systems. The 4th edition WHO Classification of Tumors of Endocrine Organs was published in 2017, in which the new thyroid tumor classification was included.1 Several important modifications to follicular cell tumors were made to the new WHO classification of thyroid tumors, as listed in Table 1. In this review, thyroid follicular cell tumors (follicular-patterned neoplasms, papillary thyroid neoplasms and de-differentiated carcinomas) were reviewed and important changes were highlighted. Thus, non-follicular cell tumors were excluded from this review. As a new borderline tumor entity, noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) was detailed separately in an editorial published in this journal, Pathology International, and was also excluded from this review.2 This review further introduces publications on Asian patient series treated by standard Asian thyroid practice. The author considers these studies to provide a guide to current Asian thyroid practice, which cannot be supplemented by data elucidated from Western practice. This is because there are significant and irreconcilable differences between Asian and Western practices for management of thyroid nodules and thyroid cancer. These differences include the following: (i) Asian patient populations live in mostly iodine sufficient countries and the histological type of the majority of thyroid carcinoma is papillary thyroid carcinoma (PTC); (ii) the prevalence of the BRAFV600E point mutation in Asian PTC cohorts is higher than that reported from Western countries;3, 4 (iii) it is well-known that significant numbers of patients with low-risk papillary microcarcinoma (PMC) are treated conservatively (active surveillance) in Asia instead of by immediate surgery;5-9 (iv) resection of indeterminate thyroid nodules is quickly selected in Western practice and diagnostic surgery is more common to prevent missing malignancy due to medicolegal reasons in Western practice;10-16 and (v) approaches for low-risk thyroid carcinoma (lobectomy instead of total thyroidectomy) are more conservative in Asia.17-23 Therefore, many important publications from Asia were not well acknowledged in the WHO classification or in most Western clinical guidelines, as those data obtained from Asian practice are not reproducible in current Western practice.21-23 A review of Asian reports is important for readers practicing in Asia, and the author believes that it is not only essential to Asian practice but also to the future of Western practice because clinical management in some Western practices resembles current Asian practice, termed a multidisciplinary approach,12, 13, 20-26 as well as help to bridge the gaps between practices.21-23 Furthermore, this review provides a pathology reporting check list that adapts the new WHO classification of thyroid tumors and risk stratification of differentiated thyroid carcinomas recommended by the American Thyroid Association (ATA).1, 12, 13 One of our special missions was making a bridge between histopathological classification and clinically-relevant risk stratification of thyroid tumors. There have been only a few review articles in English covering the new WHO classification of thyroid tumors.27-31 The topics these review articles were: (i) the introduction of borderline tumors (UMP and NIFTP) in thyroid tumor classification;32, 33 (ii) PTC comprising 15 variants, and a new (hobnail) histologic variant being included;34 (iii) follicular thyroid carcinoma (FTC) being subdivided into three prognostic groups: minimally invasive (capsule invasion only), encapsulated angioinvasive and widely invasive; (iv) Hürthle/oncocytic cell tumors being separated from follicular adenoma/carcinoma (FA/FTC) in an independent chapter as it has a different genetic profile from those of the other types of thyroid cancer;35 (v) the diagnostic criteria of poorly differentiated carcinoma (PDC) being more precisely defined and adopting the Turin consensus criteria;36 (vi) emphasizing the prognostic value of genetic markers, such as BRAFV600E and telomerase reverse transcriptase (TERT) promoter mutation, in thyroid carcinoma of follicular cell origin; and (vi) emphasizing examination of the entire capsule to identify or exclude invasiveness in encapsulated thyroid tumors37-41 (Table 1). After the introduction of borderline tumors, the definition of FA was altered and “without nuclear features of PTC (PTC-N)” was added.1 FA is a benign, encapsulated and noninvasive neoplasm demonstrating evidence of thyroid follicular cell differentiation without PTC-N. The differential diagnoses include benign hyperplastic nodule, encapsulated noninvasive PTC and minimally invasive FTC. Differential diagnosis between hyperplastic nodule and FA is not always possible without molecular studies (polyclonal proliferation vs. monoclonal neoplasm), but it is not important clinically because both lesions are benign. Differential diagnosis between FA and malignant encapsulated carcinomas (FTC and PTC) is difficult and has caused significant observer variability in the past.42-45 The borderline tumor category was proposed to solve this diagnostic difficulty for pathologists.32, 33, 46-48 According to the current ATA clinical guidelines, NIFTP is a surgical disease. It requires surgical removal and histological evaluation, because it is a precancer in definition and should be removed under the Western logic.13, 15 However, in Asian perspectives, accurate distinction between FA and borderline tumors (UMP and NIFTP) is not important clinically, and under-diagnosis of borderline tumors as FA does not harm to the patients because both can be treated by simple excision and cured. Most importantly, pathologists must consider a benign diagnosis first when an encapsulated thyroid nodule is limited to within the thyroid gland (intrathyroidal). Under-diagnosis of intrathyroid encapsulated carcinomas as borderline tumors or benign tumors does not seriously harm the patients. Even if they are found to be malignant tumors and adverse events develop, they can be managed curatively as long as they are encapsulated and intrathyroidal tumors (ex0, N0 and M0). In our Asian and Australasian experience, the vast majority of borderline tumors are diagnosed in the benign category (FA),42, 43, 49-54 and clinically significant cancers are rarely missed using this conservative approach.22, 54, 55 However, some intrathyroidal tumors are true malignant tumors and have a non-negligible risk for structural disease recurrence, such as non-encapsulated 2-4 cm PTC (5% of recurrence) and BRAF mutated <4 cm PTC (10% of recurrence),12 but they can be detected preoperatively by applying a multidisciplinary clinical approach by experienced clinicians and pathologists.18-24, 55 We propose here a new tumor entity, low-risk intrathyroidal neoplasia, which encompasses thyroid follicular cell tumors currently classified in different categories, benign (FA), borderline (UMP and NIFTP) and low-risk thyroid follicular cell carcinoma (PMC, minimally invasive FTC and encapsulated PTC), which have a very low risk of disease recurrence after a complete excision (curative with lobectomy alone). The authors consider this terminology (low-risk intrathyroidal neoplasia instead of cancer) helpful for clinical colleagues to more easily understand that limited surgery (lobectomy) is sufficient to cure these low-risk intrathyroidal neoplasms, and aggressive cancer treatment (total thyroidectomy followed by radioactive iodine treatment) is not necessary. We also believe it can alleviate the psychological burden of a cancer diagnosis in patients. The definition of PTC in the 2004 WHO classification (the 3rd edition) was “a malignant epithelial tumor demonstrating evidence of follicular cell differentiation and characterized by distinct nuclear features”,56 and it was changed to “PTC is a malignant epithelial tumor demonstrating evidence of follicular cell differentiation and a set of distinct nuclear features. PTC is usually invasive. Papillae, invasion or cytological features of papillary thyroid carcinoma are required” in the 2017 WHO classification (the 4th edition).1 These modifications were based on noninvasive encapsulated follicular variant PTC being reclassified into the borderline tumor (non-malignant) category (NIFTP)37 and encapsulated follicular variant PTC with incomplete capsular/lympho-vascular invasion being reclassified into the borderline tumor category [well differentiated thyroid tumor of uncertain malignant potential (WDT-UMP)].57 However, these tumors [invasive (PTC), incomplete invasive (WDT-UMP) and noninvasive (NIFTP)] were still classified in malignant tumors as intrathyroidal encapsulated follicular variant PTC in the 2015 ATA guidelines.12 This was caused by several previous publications by eminent thyroid experts who emphasized that ground-glass nuclei were diagnostic for PTC-type malignancy even in noninvasive encapsulated follicular-patterned (non-papillary) thyroid nodules.58-60 Encapsulated thyroid tumors with worrisome PTC-N would have been previously referred to as PTC even in noninvasive encapsulated follicular-patterned (non-papillary) thyroid nodules in cases without invasion and/or metastasis.10, 12, 38, 42-44, 49-51, 60 Significant numbers of patients with such tumors were treated aggressively, particularly in North American practice, because the 2009 ATA guidelines recommended that all PTC larger than 1 cm be treated by total thyroidectomy.10 However, many pathologists raised a concern regarding classifying noninvasive encapsulated follicular pattern nodules with worrisome PTC-N as cancer.11, 38, 42-44, 47-54 After publication of the 2017 WHO classification, the identification of definite invasion, as well as the demonstration of well-developed PTC-N and true papillae have become essential components in the diagnosis of PTC. Even for encapsulated thyroid nodules with unequivocal papillae, identification of invasion, fully developed PTC-N or BRAFV600E mutation are required for a diagnosis of encapsulated conventional PTC to exclude FA with papillary hyperplasia. These stricter diagnostic criteria for encapsulated PTCs were intended to reduce over-diagnosis of indolent thyroid tumors61 and proper identification of aggressive variants of PTC which develop high rates of disease recurrence (Table 2). In addition to conventional (classic, common or usual) type PTC, 14 variants (Table 2) were listed in the new WHO classification of thyroid tumors.1 The prevalence of these variants varies among patient populations, but PMC, encapsulated, follicular and tall cell variants are relatively more common than other variants. PMC comprised more than half of the surgically treated PTC in some patient series.1, 5, 56, 61 Follicular variant PTC is a prevalent histological variant in Western experience, and was reported to comprise 37.9% of all PTC in five Italian and American hospitals reported by Nikiforov et al.33 and 17.9% in a large international collaborative study (1126/6282 cases from 14 countries) by Shi.62 However, it was much rarer, 4.0% (2.2–9.8%), in a recent large multi-institutional Asian series.63 In the 2017 WHO classification, the prevalence of encapsulated variant PTC accounted for approximately 10% of all PTC and the solid variant constitutes 1–3% of adult PTC1 (Table 2). In a review by Fagin and Wells, 84% of all thyroid carcinomas were PTC, 33% were PMC, 32% were classic PTC, all follicular variants accounted for 27% (6% infiltrative, 4% encapsulated invasive and 17% encapsulated without invasion = NIFTP) and the tall cell variant accounted for 7% of PTC in the USA64 (Table 2). The definition of PMC is a PTC less than or equal to 10 mm in diameter, regardless of the presence of high-risk features such as BRAFV600E mutation, vocal cord paralysis, clinical lymph node metastasis and distant metastasis.1, 5, 13, 56, 65-70 Lo et al. suggested that a distinction should be made between clinically overt and occult PMC, despite a relatively good prognosis for PMC.67 Sugitani et al. classified PMC into three prognostic groups: symptomatic PMC with clinically apparent lymph node metastasis and/or vocal cord paralysis was malignant, and had a 30% recurrence rate and a 74.1% rate of cause-specific survival at 10 years, whereas asymptomatic PMC without these features had corresponding rates of 3% and 100%, respectively.8 More recently, a classification of PMCs into incidental and non-incidental obtained more adoption by clinicians and pathologists.71, 72 Comprehensive risk stratification of PMC considers a constellation of clinical, pathological, and molecular parameters.70, 73 Observation without surgical treatment is practiced for patients with low-risk PMC in several leading medical centers in Japan, Korea and the USA.4-8, 68, 74, 75 Less than 15% of low-risk PMC grow to more than 3 mm or develop lymph node metastasis during follow-up, and most importantly, no thyroid cancer death was reported among more than 1000 patients for more than 10 years of follow-up.4-8, 68, 74-76 Expert thyroid pathologists proposed calling it papillary microtumor instead of carcinoma,77 and a proposal to classify PMC in the borderline tumor category was reported by Kakudo et al.48 However, papillary microtumor has not become a popular diagnosis among pathologists and clinicians because a significant amount of PMC cases (15–35%) have lymph node metastasis at surgery and rare cases (<0.4%) develop distant metastasis.65-68, 71-80 Encapsulated variant is conventional PTC completely surrounded by a fibrous capsule that may be intact or only focally infiltrated by the carcinoma.1 Although regional lymph node metastases may be present, patients with encapsulated PTC have an excellent prognosis with a survival rate of 100% or almost 100%81-85 (Table 2). The differential diagnosis includes FA with papillary hyperplasia, WDT-UMP with minor papillary growth and FTC with incomplete PTC-N (well-differentiated carcinoma, not otherwise specified). Follicular variant PTC is a group of tumors with an exclusively (100%) or almost exclusively follicular pattern of growth, which was subclassified into encapsulated and non-encapsulated (infiltrative) variants. The encapsulated follicular variant PTC was further divided into invasive and noninvasive.86-88 The noninvasive encapsulate follicular variant of PTC was downgraded from carcinoma to NIFTP, and cases with incomplete invasion were downgraded from carcinoma to WDT-UMP in the 4th edition WHO classification,1, 32, 37, 57 whereas those with definite invasion remain in the malignant category (invasive encapsulated follicular variant PTC) (Fig. 1a).1 Follicular variant PTC comprises mostly low grade tumors and RAS-like tumors86-97 (Table 2). It can be classified into either borderline (WDT-UMP or NIFTP) or malignant (invasive encapsulated PTC or infiltrative PTC) according to the absence of, incomplete or definite capsular/lympho-vascular invasion.1 When any exclusion criteria for NIFTP, such as true papillae, psammoma bodies and/or BRAFV600E mutation, are present in noninvasive encapsulated follicular pattern thyroid tumors with PTC-N, they should be excluded from the borderline category and be placed in the malignant category (noninvasive encapsulated follicular variant PTC).1, 2, 33, 39-41, 93 This is a narrow definition of noninvasive encapsulated follicular variant PTC (BRAF-like), and both NIFTP (RAS-like) and encapsulated follicular variant PTC (BRAF-like) are often mixed up in discussion, even though they should be clearly separated because of differences in oncogenesis. Asian and Australian pathologists have stricter diagnostic criteria for PTC-N, and accept only BRAF-like tumors as PTC. NIFTP (RAS-like tumor) is often classified as benign FA, whereas nuclear features of NIFTP are accepted as PTC-N by Western pathologists and most NIFTP was previously classified in the malignant category (RAS type PTC).98 Diffuse sclerosing variant PTC is an uncommon variant that is more frequently observed in young females. It is characterized by diffuse involvement of a single lobe or entire thyroid gland. Histologically, squamous metaplasia, stromal fibrosis, lymphocytic infiltration and psammoma bodies are common in the tumor.99, 100 Marked lymphatic invasion by tumor cells is characterized by a micropapillary/discohesive growth pattern simulating the hobnail variant of PTC (Fig. 1b). Recent meta-analysis still considers diffuse sclerosing variant as a high-risk PTC because of a high incidence of extrathyroid invasion at surgery, advanced tumor stage at presentation and bilaterality, however there is an obvious trend in the modern institutional series that diffuse sclerosing variant PTC has no adverse effect on survival101-105 (Table 2). According to the definition by the 4th edition WHO classification, tall cell variant PTC is a PTC composed of cancer cells that are 2- to 3-times taller than wide (Fig. 2a).1 A high incidence of extrathyroid extension at surgery, advanced tumor stage at presentation, older patient age, BRAF mutation and TERT promoter mutation are noted in tall cell variant PTC.106-109 In an international collaborative study, Shi et al. examined 6282 PTC cases, and tall cell variant PTC (3.8%, n = 239) had a recurrence rate of 27.3% and mortality rate of 6.7%, whereas conventional PTC (74.8%, n = 4702) had a recurrence rate of 16.1% and mortality rate of 2.5%62 (Table 2). As the majority of PTC is histological variants in differing proportions, the most recent WHO classification defined that tall cells must account for more than 30% of all tumor cells for a diagnosis of the tall cell variant (revised from the previous definition of more than 50% by the 2004 WHO classification) (Table 3).1, 56, 110 However, Ito et al. evaluated 70 Japanese cases of tall cell variant PTC (more than 30% tall cells), and found that PTC with more than 50% tall cells significantly and independently affects the disease-free survival, whereas PTC with 30–49% tall cells did not. They concluded that the definition of more than 50% tall cells by the previous WHO classification is appropriate and that the prognostic impact of PTC with tall cell features (30–49%) is limited.111 In an Italian consensus, 10% was recommended as the cut-off for tall cell features.25 Therefore, the diagnostic criteria for tall cell variant PTC and other variants of PTC need further verification studies. In particular, what proportion of a certain feature is required for diagnosis, as estimation of % tumor area under a microscope has significant observer variation112 (Tables, and 3 2). Please refer to the section “How to handle multiple prognostic factors and which more greatly impact patient outcomes” below. Columnar cell variant PTC has tall cell morphology but lacks frequent nuclear pseudoinclusions, eosinophilic cytoplasm, and distinct cell borders, which are common in tall cell variant PTCs (Fig. 2a). Columnar cell variant PTC exhibits prominent nuclear pseudostratification (Fig. 2b) and a poor prognosis.113-116 It is a rare subtype of PTC accounting for only 0.15–0.2% of all PTC115 (Table 2). As typical PTC-type nuclear features are not seen in columnar cell variant PTC,117 it was excluded from the PTC lineage and classified in other tumors in the Japanese classification of thyroid tumors proposed by the Guidelines for Clinical Practice for the Management of Thyroid Nodules in Japan 2015.18 The columnar cell variant PTC is difficult to differentiate from metastatic adenocarcinoma, particularly from the endometrium and colon. Carcinoembryonic antigen and TTF1 immunohistochemistry are helpful in distinguishing columnar cell variant PTC from metastatic adenocarcinoma of the colon because columnar cell variant PTC is negative for carcinoembryonic antigen and variably positive for TTF1. However, immunocytochemistry for estrogen receptor and CDX2 (intestinal-type differentiation marker) are not useful because they have been found to be positive in some columnar cell variant PTC.118-121 The prognosis of columnar cell variant PTC was revised in the current WHO classification.1 Encapsulated columnar cell variant PTC has indolent biological behavior, but cases with extrathyroid extension demonstrate aggressive clinical behavior and have a less favorable prognosis118, 122 (Table 2). Cribriform-morular variant PTC can occur as a manifestation of a familial adenomatous polyposis coli or sporadic form.123-127 More than half of the cases of this variant occur in patients with familial adenomatous polyposis syndrome and APC gene alteration. Eleven patients (16% female and 0% male) among 129 Japanese familial adenomatous polyposis patients developed cribriform-morular variant PTC.127 This variant usually develops multiple nodules in the thyroid gland that are encapsulated. Histologically, it is characterized by round squamoid structures called morules and colloid production is absent in the follicle lumen (Fig. 3a).123, 124, 128, 129 Cribriform-morular variant of PTC had a lower frequency of lymph node metastases at presentation (12%) and distant metastases (3%) as well as lower recurrence rates (8.5%) and patients' mortality rates (2%) (Table 2).129 Hobnail variant PTC is a newly added histological variant of PTC and was first introduced in the 4th edition WHO classification of thyroid tumors.1 Histologically, the hobnail feature consists of complex papillary structures with a fibrovascular core and micropapillary structures lacking a true fibrovascular core. These papillary structures are covered with follicular cells containing eosinophilic voluminous cytoplasm and large apically located nuclei (hobnail, teardrop or comet-like) resulted from the loss of cellular polarity and cellular cohesion (Fig. 3b). As minor hobnail features are often observed in the invasive front of conventional PTC,130, 131 cystic PTC, diffuse sclerosing variant PTC (Fig. 1b) and even PMC, hobnail variant PTC was defined with a cut-off of greater than 30%, which made this variant rare, and its prevalence ranged from 0.3% (South Korea) to 2.7% (Southern Italy)132, 133 (Table 2). The hobnail variant is a moderately differentiated PTC with aggressive behavior and poor prognosis (Table 2), which may be related to the epithelial–mesenchymal transition.34, 130, 131, 134-137 As the hobnail features (>10%) were often associated with PDC (22%) and anaplastic thyroid carcinoma (ATC) (3.8%), Amatcher et al. suggested them to be a manifestation of high-grade transformation.138 Consequently, some authors suggest that minor hobnail micropapillary features (10–30%) should also be correctly identified and stated in pathology report due to potential aggressive behavior.81, 130, 131, 139, 140 Hobnail cytological features were first described in PTC by Tang et al. in 2003 as the loss of cellular cohesion/cellular polarity.140 Hobnail features were associated with a high Ki67 labeling index and loss of retinoid receptor expression in PTC.140 Later, it was termed micropapillary/discohesive-type PTC by Bai et al., and hobnail features (>10%) were characterized by a high risk of disease recurrence.81, 130, 131, 141, 142 These tumor cells are typically positive for TTF-1 and variably positive for thyroglobulin. The Ki67 proliferation index was reported to be as high as 10%.34 BRAFV600E mutation was common (72.2%), followed by TP53 mutation (55.6%) and TERT mutation (44.4%) in an Italian patient cohort.137, 143 Teng et al. reported 17 cases of hobnail variant PTC in a Chinese cohort, and the prevalence of BRAFV600E mutation was 16/17 (94%) cases, that of TP53 mutation was 3/17 (17.6%) and TERT mutation was noted in only 1/17 (5.9%).144 Solid/trabecular variant PTC was defined differently from the other PTC variants because a solid/trabecular growth pattern is common in conventional and other variants. The solid/trabecular variant is indicated when all (100%) or nearly all of a carcinoma not belonging to any of the other variants has a solid, trabecular or insular growth pattern with clear PTC-N (Fig. 4a).1 A solid/follicular pattern in PTC was often noted in pediatric patients following the Chernobyl nuclear power plant accident.145-149 It was referred to as non-aggressive solid/follicular variant PTC in children, whereas solid/trabecular variant PTC in adults has been reported to have a slightly higher risk of disease recurrence81 and a slightly increased cause-specific mortality (10% at 10 years).150 However, these studies require further validation because the diagnostic criteria for PDC were modified in the 2017 WHO classification and a significant amount of PDC was downgraded to solid/trabecular variant PTC1, 36 (please refer to the PDC section below). The differential diagnosis includes FA with solid growth (absence of both conventional malignant features and PTC-N), NIFTP (absence of conventional malignant features and less than 30% solid/trabecular/insular growth), WDT-UMP (no definite invasion), conventional PTC with solid/trabecular/insular growth (presence of papillae), FTC with solid/trabecular/insular growth (presence of follicular growth pattern and absence of PTC-N) and PDC (absence of PTC-N and presence of high-grade histology, such as increased mitoses, and tumor necrosis). Warthin-like variant PTC shares histological features with Warthin tumor of salivary gland origin, and the prognosis is similar with that of conventional PTC.1, 56 Oncocytic variant PTC was reported to have a higher tumor recurrence rate (28% vs 11%) and cause-specific mortality (17% vs 4%) than conventional PTC,151 and this was confirmed in a Korean patient cohort (recurrence rates, 30.8% vs 11.7%) by Hong et al.152 The Warthin-like variant was included in oncocytic variant PTC in the previous WHO classification, and was regarded as a distinctive sub-variant in the current 2017 WHO classification because of the different prognosis and histological features.1, 56 The PTC variants with fibromatosis/fasciitis-like stroma, spindle cell and clear cells are rare, and the prevalence of such variants was 0.06% (1/1521) for fibromatosis/fasciitis-like stroma, 0% for spindle cell and 0.06% (1/1521) for clear cell variant PTC in a Japanese patient cohort reported by Ito et al.153 (Table 2). They are low-risk PTC similar with conventional PTC, and no definite prognostic information is available.1 Therefore, the details were omitted in this review. Follicular neoplasm is a spectrum of thyroid tumors whose major driving mutations are in RAS family of genes (RAS-like tumors),89-97 which covers benign tumor (FA), borderline tumor (FT-UMP, NIFTP and WDT-UMP), low-risk (minimally invasive FTC), intermediate-risk (encapsulated angioinvasive FTC and widely invasive FTC) and high-risk carcinoma (PDC). The distinction between benign (FA) and malignant (FTC and PDC) is based on conventional malignancy criteria (presence of capsular/lympho-vascular invasion) (Fig. 5a, b). FTC was traditionally divided into two prognostic subgroups according to the invasion pattern, minimally invasive or widely invasive.56 However, Ito et al. demonstrated this prognostic difference between minimally and widely invasive FTC diminished when cases with PDC or distant metastasis were excluded.154 O'Neil et al. combined the invasive pattern and angioinvasion, and classified FTC into three groups: (i) minimally invasive (capsule invasion only) FTC, (ii) encapsulated angioinvasive FTC and (iii) widely invasive FTC. Disease-free survival rates at 40 months of the above three groups were 97%, 81% and 46%, respectively.155 This 3-tiered risk classification was incorporated into the 2017 WHO classification1 (Table 1). In the 2015 AFIP atlas, FTC was classified into four prognostic groups: (i) minimally invasive FTC with capsular invasion; (ii) minimally invasive FTC with limited (<4 vessels) vascular invasion; (iii) minimally invasive FTC with extensive (≥4 vessels) vascular invasion; and (iv) widely invasive FTC.156 These two classifications highlighted the importance of vascular invasion in risk stratification of thyroid carcinomas.157-165 As FT-UMP and WDT-UMP were introduced in the thyroid tumor classification,1, 32, 33 evaluation of capsular and vascular invasion has become much stricter, and only definite invasion (Fig. 5a, b) is accepted for the diagnosis of FTC.57 Pathologists must select borderline (FT-UMP and WDT-UMP) diagnosis when only questionable invasion is found, to minimize the over-diagnosis of FTC.1, 45, 46, 165, 166 Please refer to Figures 2.16 and 2.17 in the 4th edition WHO Blue Book for examples of incomplete invasion which are not qualified for malignant diagnosis.1 The WHO classification handled Hürthle cell tumors as a separate tumor entity acknowledging the peculiar biological and clinical features of this unique group of tumors,167 and unique genetic profiles different from non-Hürthle cell FTC.35 Hürthle cells are large cells, and have abundant eosinophilic granular cytoplasm and large centrally located nuclei with prominent nucleoli (Fig. 4b). Some earlier studies suggested that all Hürthle cell neoplasms have the propensity of recurrence/metastasis and should be regarded as carcinoma,168, 169 but some did not accept this hypothesis.170 The new WHO classification put an end to this fruitless argument by claiming that “Hürthle cell adenoma is benign and it will not recur”.1 A malignant diagnosis of Hürthle cell carcinoma is based on the presence of capsular and vascular invasion, similar to non-Hürthle cell FTC.167 The prognosis of patients with Hürthle cell carcinoma was believed to be poorer than that of patients w