Thyroid Cytology Diagnostic Challenges and Controversies

Transcription

1 Thyroid Cytology Diagnostic Challenges and Controversies Tarik M. Elsheikh, MD Professor and Medical Director Anatomic Pathology Cleveland Clinic

2 Outline Thyroid Cytology I and II Introduction The Bethesda system classification (BSRTC) Diagnostic challenges and controversies Management options Molecular testing

3 Thyroid nodules Clinically palpable thyroid nodules: 4-7% of the adult population Increased use of diagnostic imaging increased # of incidentally discovered thyroid nodules (up 50%) Head and neck ultrasound studies for carotid or parathyroid disease CT, MRI, or PET scans for work-up of metastatic disease unrelated to thyroid Cancer risk is similar in palpable and nonpalpable nodules: 5-8 %

4 FNA OF THYROID JUST THE FACTS Pre-FNA thyroid era 20% of thyroid surgery showed cancer Post-FNA thyroid era Cancer identified at surgery increased by 75% and benign disease decreased by 70% Unnecessary surgery reduced by 50%

5 Introduction FNA proven effective management tool in patients with thyroid nodules Thyroid FNA is diagnostic in most conditions It is primarily a screening test Main purpose is to provide for a rational approach to management, and determine extent of surgery when needed

6 Introduction 2 Pre-Bethesda classification No standards existed for reporting thyroid FNAs Different classification schemes based on personal/institutional experiences and biases Discord between pathologists and clinicians on the perceptions of terminology used in reporting thyroid FNAs

7 Diagnostic Terminology and Reporting Redman 2006 Surveyed 133 clinicians (Endocrinologists, Surgeons, Thyroid specialists) Implications of FNA Dx on management options Non-diagnostic 98% repeat FNA Suspicious for malig 96% surgery Indeterminate 58% repeat FNA, 32% surgery Atypical 37% repeat FNA, 52% surgery Indeterminate was confused with ND in 58% of cases. Atypical was too ambiguous and treated as Susp. malignant in over ½ of cases

8 Justifications for Bethesda System Classification FNA has become the standard of care for initial workup of thyroid nodules Most clinicians use FNA results in conjunction with clinical findings to guide treatment Clinicians generally utilize FNA to provide a relative risk of malignancy, from which they can base their management decisions The proposed diagnostic categories are important in providing a risk of malignancy to clinicians and patients Surgery vs. follow-up

9 Bethesda System Classification Diagnostic Categories Risk of Malignancy Benign < 2-3 % Atypia (Follicular lesion) of US 5-10 % Neoplasm (Follicular / Oncocytic) 20-30% Suspicious for Malignancy 50-75% Malignant 100 % Non-diagnostic

10 Major Challenges and Controversies Assessment of watery/thin colloid Conventional Paps and ThinPrep Criteria for diagnosis of Follicular Neoplasm Proportion of microfollicles Overall cellularity Suspicious for PTC How much atypia is enough? Minimal criteria for PTC Oncocytic lesions Neoplastic vs. non-neoplastic Malignant mimickers

11 Nodular Goitre/Hyperplastic Nodule 80% of aspirated thyroid nodules Most patients are euthyroid May become large with pressure symptoms

12 Nodular Goiter/Hyperplastic Nodule Abundant colloid Variable cellularity Oncocytic metaplasia Degenerative changes

13 Flat sheets- honeycomb Few microfollicles accepted Occasional balls and microtissue fragments

14 Benign follicular cells Uniform nuclei: Same size as RBC Minimal nuclear overlapping Anisonucleosis Finely granular chromatin Rare nucleoli

15 ThinPrep Dense colloid Easy to recognize Dark blue-violet-magenta (DQ) Dark green-orange (Pap)

16 Watery/Thin Colloid Blue-violet (DQ), light green-orange (Pap) Folds, thin-membrane, crazy pavement, ocean waves

17 Thin Colloid in Bloody Specimens Difficult to recognize Easily confused with serum in bloody specimens (Stelow 2005)

18 Conventional ThinPrep Thin Colloid May disappear completely on LB preps Tissue paper like appearance on TP ThinPrep

19 ThinPrep: 58 YOF, Right thyroid nodule

20 Cytology Dx: Nodular goiter Histology follow-up: FVPC

21 Not colloid colloid TP: Watery colloid has tissue paper appearance

22 Follicular Lesions Hyperplastic/adenomatoid nodule Follicular Neoplasm Follicular adenoma Follicular carcinoma Follicular variant of Papillary carcinoma

23 Differential Diagnosis of Follicular Lesions -Abundant colloid= benign - Marked cellularity= Neoplasm Colloid Hyperplastic Nodule I Indeterminate, Cellular Nodule, Atypical FL Grey Zone II Follicular Neoplasm Cells III Demay. Art & Science of Cytopathology, 1996 Overlapping cytologic features make it difficult, at times, to separate between HN & FN Indeterminate category accounts for 5-42% of FNA Dx s

24 Differential Diagnosis of Follicular Lesions 2 Hyperplastic Nodule Cellular Lesion Grey Zone Follicular Neoplasm Colloid Microfollicles Uniformity Architecture Cellularity Nuclear atypia Sheets Permissiveness in applying strict criteria to Dx of FN significant reduction of malignancy rate on FU

25 Follicular Lesions and Terminology 2 studies found no malignancies on F/U of FNAs diagnosed as FL and FN Authors advocated a less aggressive approach to management, i.e. clinical follow-up FN was defined as: Hypercellular smear Scant colloid Microfollicles present (Foppiani 2003, Piromalli 1992)

26 Bethesda System Approach to Grey Zone and Terminology Although Follicular Lesion & Follicular Neoplasm are used interchangeably by some authors We do not consider them synonymous Indeterminate cytologic category included FN, FL, Susp. for malignancy, Atypia NOS

27 Bethesda System Diagnostic Categories 1. Benign 2. AUS/FLUS 3. Follicular Neoplasm/SFN 4. Suspicious for Malignancy 5. Malignant 6. Non-diagnostic

28 Follicular Neoplasm FA Cytologic DDx: Follicular adenoma Follicular carcinoma FVPC Need histologic confirmation FC Follow-up: 70% neoplasm 30% cancer (FC, FVPC)

29 Follicular Neoplasm Cytologic Criteria High cellularity Scant colloid Prominent microfollicles and/or syncytial fragments (> 50-75% of cells) Significant nuclear overlapping and crowding Monotonous cell population

30 Microfollicles <15 cells arranged in circle that is at least 2/3 complete Nuclear crowding/overlap Microfollicles + no atypia : low cancer risk (6 %) Microfollicles + abundant colloid + absence of nuclear overlap : 0 % cancer Ersoz 2004, Kelman 2001, Yang 2003, Goldstein 2002, Barbaro 2001, Renshaw 2006

31 Follicular Neoplasm

32 Follicular Neoplasm

33 Follicular Neoplasm

34 Follicular Neoplasm Cytologic Criteria 2 Uniform enlargement >2X RBC Coarse and clumped chromatin ± Prominent nucleoli ± Severe nuclear pleomorphism Ersoz 2004, Kelman 2001, Yang 2003, Goldstein 2002, Barbaro 2001

35 FN Case study L thyroid nodule (1.3 cm) from a 32 year old man NG

36 Atypia of Undetermined Significance (AUS/FLUS) Cytologic Features Major differential diagnoses are HN vs. FN Reactive changes vs. PTC High cellularity, scant colloid Smears from different passes show a spectrum ranging from benign to possible FN Admixture of flat sheets and microfollicles/syncytia Minimal nuclear overlapping and crowding Low cellularity, but prominent microfollicles and nuclear overlap (highly vascular lesions)

37 AUS/FLUS

38 AUS/FLU S Specimen consisted predominately of blood Rare groups of follicular cells Clue: abundant blood with rare microfollicles or syncytia (Yang 2003, Lowhagen & Oertel)

39 AUS R thyroid nodule, 45 yof FU: FVPC

40 AUS NG Nuclear grooves in NG, mimicking PTC

41 LT NG Repair Nuclear irregularities and grooves may be associated with LT, NG, and repair

42 A,B: Benign C,D: AUS/FLUS

43 Nodular Goiter NOT Microfollicles Balls and micro-tissue fragments Few microfollicles allowed in NG

44 AUS/FLUS Cytology not convincingly benign, yet degree of nuclear or architectural atypia is not sufficient for diagnosis of FN or susp. for malignancy Some cases are due to a compromised specimen, i.e. low cellularity, poor fixation, obscuring blood Avoid overuse of this category Ideally < 12% of thyroid FNAs (BTS <7%)

45 Papillary Carcinoma 80% of all thyroid CA Variants, including follicular, tall cell, columnar, oncocytic Nuclear features: Crowding and overlapping of nuclei Nuclei usually enlarged Finely powdery chromatin Nuclear grooves Intranuclear holes

46 Classic PTC

47 Follicular Variant of PTC Second to sampling error as most common cause of false negative Dx s (Wu 2006)

48 FVPC Branching monolayered sheets: most significant low power discriminator from FN (Fulciniti 2001)

49 Bubble Gum Colloid and PTC

50 Squamoid cytoplasm Oval enlarged nuclei, powdery chromatin Grooves/irregular nuclear membranes Marginated nucleoli Intranuclear holes

51 Thyroid Cytology Diagnostic Challenges and Controversies Surrounding BSRTC Part II Tarik M. Elsheikh, MD Cleveland Clinic

52 Bethesda System Diagnostic Categories 1. Benign 2. AUS/FLUS 3. Follicular Neoplasm/SFN 4. Suspicious for Malignancy 5. Malignant 6. Non-diagnostic

53 Case Study 2.5 cm R thyroid nodule in a 56 year old woman

54 FVPC, false negative

55 FVPC may show: Paucity of nuclear features of PTC Abundant colloid Misdiagnosed as B9 or FN Should have been Dx d as susp. for PTC

56 Suspicious for PTC Strong suspicion for malignancy Cytologic criteria: 1. Quantitative: PTC features present but very sparse cellularity Patchy/focal nuclear changes of PTC 2. Qualitative Diffuse but incomplete nuclear changes of PTC i.e. generalized nuclear enlargement and pallor, but rare grooves or inclusions Hypervacuolated and atypical histiocytoid cells

57 Suspicious for PTC Sensitive cytologic criteria for detecting FVPC Flat syncytial sheets Most sensitive Nuclear enlargement Fine chromatin Nuclear grooves Most specific < ½ FVPC showed intra-nuclear holes (Wu 2003) Important NOT to lump these cases (70-75% cancer risk) with other indeterminate Dx s: AUS/FLUS (5-10% cancer risk) FN (20-30% cancer risk)

58 Suspicious for PTC Focal grooves, nuclear enlargement and powdery chromatin COMBINED in same nuclei

59 Generalized nuclear enlargement & pallor, but rare grooves Susp. for PTC

60 AUS Rare cells with distinct mild focal nuclear atypia More commonly associated with LT and cyst Occasionally with FVPC

61 Susp for PTC

62 Susp for PTC ThinPrep

63 FN ThinPrep

64 B9 Hypervacuolated and atypical histiocytoid cells

65 Atypical Histiocytoid cells in PTC An Under-recognized cytologic pattern Resemble histiocytes but larger Enlarged nuclei, abundant vacuolated cytoplasm No grooves, no prominent inclusions Cystic PTC may be dominated by these macrophage-looking cells Potential pitfall false negative Dx AE1/3+, TTF1+

66 AHC: Cytologic Features Single cells or loose clusters Large cells, abundant cytoplasm Histiocytic or epithelioid appearance Two cell types: 1. Hypervacuolated cytoplasm - Vesicular nuclei, prominent nucleoli - Dark smudgy nuclei 2. Dense cytoplasm - Dark atypical nuclei, eccentrically placed

67 36 yof, isthmus nodule Susp. for PTC

68 Thyroidectomy: FVPC, BRAF-

69 A 54 YOF, left thyroid nodule

70 Cytology Dx: Susp. for PTC Thyroidectomy: Lymphocytic thyroiditis - Atypia threshold should be increased in LT

71 PTC associated with LT

72 Oncocytic Lesions

73 Oncocytic (Hurthle Cell) Neoplasm Cytology can t separate adenoma from carcinoma Highly cellular specimens with scant to absent colloid or lymphocytes

74 Sheets, loosely cohesive clusters Uniform cell appearance

75 Abundant granular cytoplasm, well defined cell border Prominent nucleoli

76 Many isolated cells Variable atypia is allowed

77 Differential Diagnosis of Oncocytic Neoplasm Non-neoplastic lesions Oncocytic nodule/metaplasia Nodular goiter Lymphocytic thyroiditis Malignancies with oncocytic features Papillary carcinoma variants oncocytic, tall cell and Warthin-like Medullary carcinoma

78 Oncocytic metaplasia/nodule in NG

79 Admixture of B9 thyroid follicular cells and colloid favors NG

80 Lymphocytic thyroiditis

81 Anisonucleosis may be prominent in LT

82 Lymphocytic thyroiditis

83 Neoplastic vs. Nonneoplastic Numerous isolated cells Intra-cytoplasmic lumens Cytologic atypia Transgressing blood vessels Absence of Colloid Absence of lymphoid cells Nodule size > 3 cm Canberk 2013, Yang 2013, Wu 2008, Elliott 2006

84 Medullary Carcinoma

85 Medullary Carcinoma Variable cytologic appearance: - Carcinoid-like/microfollicular - Spindle - Epithelioid - Plasmacytoid - Oncocytic

86 Mostly isolated cells, occasional clusters Amyloid may be present (Congo red +)

87 Delicate lacy cytoplasm Anisonucleosis

88 Monomorphic-pleomorphic Round-oval nuclei, coarsely granular chromatin, small nucleoli Intranuclear inclusions

89 Ancillary Studies in MTC Calcitonin ICC Calcitonin + Chromogranin +, Synaptophysin + CEA + Thyroglobulin Serum calcitonin elevated Chromogranin

90 Immunocytochemistry Oncocytic neoplasm TTF1 TG PAX8 Calcitonin Chromo/ Synapto * - Medullary CA * Beware of Calcitonin false-positive staining in cell blocks!

91 Oncocytic PTC

92 BSRTC Diagnostic Categories 1. Benign 2. Follicular lesion of Undetermined Significance 3. Follicular/Hurthle cell Neoplasm 4. Suspicious for Malignancy 5. Malignant 6. Non-diagnostic

93 Specimen Adequacy Criteria General Principles Main purpose is to minimize # of false negative diagnoses Provide a meaningful interpretation that is clinically useful A diagnosis of rare/few benign follicular cells without qualification, is not considered meaningful An adequate sample should be representative of the lesion (appropriate cellularity) and technically well prepared, i.e. good fixation, thin smear, adequate staining

94 BSRTC Adequacy Criteria Specimen processed and examined, but Nondiagnostic due to: No follicular cells or limited cellularity Poor fixation and preservation Optimal # of passes: 2-5 Any significant cytologic atypia precludes the interpretation of ND Solid nodules: minimum of 6 groups (at least 10 cells/group), preferably on a single slide A repeat FNA can be recommended

95 Exceptions to Minimal Number Criteria Inflammatory process such as thyroiditis Abundant colloid C/W Colloid nodule Cyst fluid with rare benign follicular cells C/W benign cyst CONTROVERSY: Cyst fluid only (no follicular cells) Diagnostic vs. ND?

96 Thyroid Cysts and Malignancy Cysts most commonly due to cystic degeneration in NG Any residual mass after aspiration should be sampled Risk of malignancy Of all aspirated cysts, as low as 1% are malignant Simple, non-complex cysts = 1-4% cancer risk Mixed solid-cystic nodules, large cysts (>3cm) and recurring cysts = Up to 14% cancer risk

97 NCI Thyroid Conference Conclusions Cyst Fluid Only Cystic lesions, lacking follicular cells, that collapse completely following aspiration Cyst fluid only Include under Non-diagnostic (majority agreement) or Benign Options: Recommend correlation with cyst size, complexity and US features Disclaimer that cystic CA can not be entirely excluded

98 Clinical Implications and Management Benign < 3% cancer risk Clinical/periodic US 6-18 month intervals, for at least 3-5 years Repeat FNA if significant increase in nodule size Follicular/Oncocytic neoplasm 20-30% cancer risk Lobectomy Baloch 2008, Greaves 2000, Sidawy 1997, Hamburger 1998, LaRosa 1991

99 Clinical Implications and Management 2 AUS/FLUS Approximately 10% cancer risk Repeat FNA in 3-6 months, correlate with clinical and radiologic findings If repeat FNA is Atypical or worse consider surgery NOT equivalent to Susp. for malignancy Baloch 2008, Greaves 2000, Sidawy 1997, Hamburger 1998, LaRosa 1991

100 Clinical Implications and Management 3 Suspicious for PTC 60-75% cancer risk Options: 1. Lobectomy 2. Lobectomy + intra-operative consult Helpful in additional 30% of cases (Baloch 2002) 3. Total thyroidectomy

101 NIFT "Non-invasive follicular thyroid neoplasm with papillary-like nuclear features (NIFT) A new terminology for non-invasive encapsulated/well circumscribed FVPTC, recommended by Endocrine Pathology Society This terminology was agreed upon in a multidisciplinary meeting, March endocrine pathologists, 2 endocrinologists, 1 surgeon, 1 ethicist

102 NIFT Literature review and careful assessment of histologic criteria Also reviewed 200 cases submitted by working group Very low malignant potential Goal is to decrease overtreatment lobectomy only with no adjuvant RAI Consensus paper is being drafted

103 How will NIFT Terminology Impact Cytology and Rate of Malignancy on FU? Cytologic Diagnosis (cases) Malignant Risk % Revised Risk Malignancy % Difference % AUS/FLUS (97) FN/SFN (88) SFM (94) Malignant (156) Non-invasive FVPTCs accounted for 25% of malignancies Most significant drop seen in SFM category Strickland, USCAP poster 2015

104 Molecular Testing With increasing knowledge about molecular mechanisms employed in cancer, and Lack of definitive cytologic characterization of indeterminate lesions Utilization of molecular techniques has become an attractive option for clinicians Major purpose is to increase predictive power of thyroid FNA diagnosis

105 Currently, there is no single molecular marker that is sensitive or specific enough to justify its use alone as a predictor of benign or malignant disease

106 Best Available Commercial Tests 1. ThyGenX Confirms malignant diagnosis 2. Afirma (Veracyte) Confirms benign diagnosis 3. ThyroSeq v.2 (CBL Path) Targeted next-generation sequencing

107 1. ThyGenX (formerly mirinform) In August 2014, Interpace Diagnostics acquired mirinform Thyroid Test from Asuragen Panel of oncogene mutations BRAF V600E mutation RAS mutations RET/PTC re-arrangements PAX8/PPARγ fusion Identify lesions that are malignant, i.e. increase PPV of FNA Excellent specificity but up to 30% NPV

108 PTC FC FA BRAF 45% RAS 15% (FVPC) 40-50% 20-40% RET/PTC 20% (Adults) PAX8/PPAR 30-40% 2-10% BRAF : mostly classic and tall cell PTC PPV approx 100% Associated with more aggressive tumor behavior? Debatable prognostic value beyond that obtained from traditional pathology and radiology evaluation Possible risk of PTC recurrence, but not mortality (Xing 2014) RAS has no clear predictive role in tumor aggressiveness found in benign adenomas and PD carcinomas

109 ThyGenX Sens % Spec % PPV % Cancer prevl % FN % AUS FN Susp Largest prospective trial to date: 479 indeterminate FNA samples Yielded 87 positive mutations (18%) 19 BRAF, 62 RAS, 1 RET/PTC, 5 PAX8/PPARg All FP results (9) due to RAS+ follicular adenomas Suggested that the presence of any mutation, especially BRAF and RET/PTC, would be strong indication for total thyroidectomy Nikiforov 2011

110 2. Afirma (Veracyte) Gene expression classifier Measures expression of 167 RNA transcripts from indeterminate thyroid FNAs Identify lesions that are benign, i.e. increase NPV of FNA Generates a diagnosis of either benign or suspicious

111 Alexander Veracyte Validation Study Prospective multicenter study tested 265 indeterminate FNAs Sensitivity 92%, Specificity 52% Identified 78 of of 85 malignancies as suspicious NPV 93% among all indeterminate lesions NPV % Cancer prevalence % AUS FN Susp 85 (FN 15%) 62 NPV increases with lower cancer prevalence

112 Afirma Malignancy Classifiers 2 Kloos R, 2014 (Veracyte poster, not yet published) MTC classifier was identified in % of AUS/FN and 1-1.8% of SFM/Malig FNAs Identified all 39 histologically confirmed MTC Only 15 of those cases (38%) were suspected or diagnosed by FNA PPV for MTC was 98% (1 False Positive: paraganglioma)

113 Syn Calc Left thyroid 2.3 cm nodule, 54 YOM Initial & repeat FNA Dx: AUS- predominate oncocytic cells Afirma (Veracyte): Suspicious- MTC classifier identified Serum Calcitonin 3,300 (Norm <7.5) Revised DX: Medullary CA

114 Suspicious Afirma Results & Lastra R, 2014 Oncocytic Lesions 13 FN with oncocytic features: suspicious Afirma 11/13 (85%) benign, 2/13 (15%) malignant Compared to malignancy rates of 53% (9 cases) for FN, and 63% (10 cases) for AUS/FLUS Harrell R, 2014 Suspicious Afirma in 19/21 oncocytic indeterminate FNAs (AUS and FN) 9/13 benign on FU (30% malignant rate) Afirma adds little information to enhance surgical decision making in oncocytic-rich lesions

115 3. ThyroSeq v2 Targeted NGS, simultaneously detects >400 mutations and gene fusions in >60 thyroid cancer genes CBL Path recently offered test for indeterminate thyroid FNAs- developed by UPMC Prospectively evaluated 52 cytologies with DX of FN/SFN: 14 out of 52 cases were malignant on FU(27%) No published studies to date on AUS/FLUS Nikiforov, Cancer Dec 2014

116 Molecular testing ThyroSeq v.2 Mutation negative: 35/37 benign, 2 malignant Mutation positive: 12/15 malignant, 3 benign Accuracy 90% Sensitivity 86%, Specificity 92% PPV 80%, NPV 95% ThyroSeq v2 is currently the leading candidate for test of the future, but still needs validation Relatively small sample size- single institution, so it needs to be verified by other groups and bigger Nikiforov, Cancer Dec 2014

117 TCGA Study Until recently 25% of PTC (most common thyroid cancer) had no known oncogenic drivers (dark matter) Now with publication of most recent research by Cancer Genome Atlas project (October 2014), this number has shrunken to 3.5% This has direct bearing on molecular testing of thyroid nodules, as it can improve their performance

118 TCGA Study Analyzed nearly 500 PTC samples to identify all genetic mutations that play a role-largest cohort studied to date Clinically aggressive thyroid cancers (poorly and undifferentiated carcinomas) were excluded Overall, found the thyroid cancer genome to be relatively quiet (fewer genetic mutations compared to other common cancers) This may explain the indolent clinical behavior of PTC Discovered several new cancer genes as well as new variations of existing genes The Cancer Genome Atlas Research Network. Cell, Oct. 2014

119 TCGA Study Striking signaling differences in RAS- and BRAF V600E driven PTCs BVL-PTCs signal preferentially through MAPK while RL-PTCs signal through both MAPK and

120 TCGA Study RL-PTCs & BVL-PTCs are fundamentally different in their genomic, epigenomic, and proteomic profiles Significantly different biologies BRAF V600E PTC represents a diverse grp of tumors, consisting of at least 4 molecular subtypes, with variable degrees of cancer differentiation (BRAF-associated PTC was considered a homogeneous group in past)

121 TCGA Study Limitations: Focused on PTC- Indolent cancer types with 95% cure rate No long term follow-up data (need 20 years) Questions raised: Should follicular-patterned thyroid tumors be reclassified based on molecular signature?

122 How Can We Incorporate Molecular Tests Into Our Practice Today? Consider only if results will significantly change management Observe vs. surgery, or extent of surgery AUS/FLUS (5-10% cancer risk) Only if considering surgery, i.e. after repeat FNA is AUS/FLUS Afirma performs best (high NPV) to rule in benign Dx ThyroSeq v2- No data available to date for AUS

123 How Can We Incorporate Molecular Tests Into Our Practice Today? Follicular Neoplasm (20-30% cancer risk) Afirma if clinically low risk disease ThyGenX if clinically high risk disease ThyroSeq v2- limited research Perhaps combination? Cost prohibitive ($ per panel) Susp. For malignancy (60-75% cancer risk) Limited to no value? help confirm extent of surgery, i.e. BRAF+

124 Future of Molecular Testing As more genes are added, molecular testing will become more sensitive and accurate Preoperatively identifying more aggressive thyroid cancers can tailor extent of surgery Expansion of the somatic genetic landscape has potential to further enhance care of patients by expanding biological basis for targeted therapy

125 Summary FNA can provide a definitive DX in most instances Thyroid FNA, however, is primarily a screening tool, therefore a conclusive DX is not always required Pathologist s role: minimize # of indeterminate diagnoses without yielding an unacceptably high false negative rate

126 Summary 2 The use of the term Atypical or Indeterminate as a stand alone diagnosis is not recommended. Its meaning is not standardized and may be interpreted in different ways

127 Atypia in Thyroid FNA Nuclear Architectural PTC-like or Pleomorphism Predominant: FN Focal: AUS Diffuse: Malignant Focal: Susp. Malig Equivocal: AUS

128 Summary 2 Use of TBTS diagnostic categories is highly encouraged Uniformity and cancer risk Recommendations for follow-up may be included in the report, if acceptable to clinicians

129 Bethesda System Classification Diagnostic Categories Risk of Malignancy Management Benign <5% F/U AUS/FLUS 5-15 % Follicular/ Oncocytic Neoplasm 20-30% Susps for Malig 50-75% F/U- repeat FNA Molecular testing Lobectomy Molecular testing Lobectomy ± FS/TT Malignant 100% Total thyroidectomy Non-diagnostic Repeat FNA (US)

130 Summary 3 Potential beneficial role for molecular testing as an ancillary tool, in selected patients Molecular studies, however, should not be ordered indiscriminately as a reflex test, and their results should not outweigh clinical judgment As cost falls and discriminant capability of molecular testing improves, it is anticipated that it will become standard practice

131 Thank You!