Challenges in Diagnostic Cytomorphology

Size: px
Start display at page:

Download "Challenges in Diagnostic Cytomorphology"


1 Challenges in Diagnostic Cytomorphology Jeffrey Krane, MD, PhD Kristen Atkins, MD Richard DeMay, MD Syed Ali, MD Dina Mody, MD Disclosure information The speakers have no relationship that represents a possible conflict of interest with respect to the content of this presentation.

2 Challenges in Diagnostic Cytomorphology: Thyroid Jeffrey F. Krane, MD, PhD Associate Director, Cytopathology Chief, Head and Neck Pathology Brigham and Women s Hospital Boston, MA DISCLOSURES: I have no financial disclosures. The Bethesda System for Reporting Thyroid Cytopathology TBS Reporting in Practice: Summary of 9 Published Studies Diagnostic Category Risk of Malignancy Usual Management I. Nondiagnostic / Unsatisfactory 1 4% Repeat FNA II. Benign 0 3% Clinical follow up III. Atypia of Undetermined Significance or Follicular Lesion of Undetermined Significance 5 15% Repeat FNA IV. Follicular neoplasm or Suspicious for a Follicular Neoplasm 15 30% Lobectomy V. Suspicious for Malignancy 60 75% N T Thyroidectomy or Lobectomy VI. Malignant 97 99% N T Thyroidectomy Adapted from Ali and Cibas, TBSRTC, ND B AUS FOL SUS M Median (%) Minimum (%) Maximum (%) Fold Var Adapted from Krane et al Cancer Cytopathol (2011) plus Wu et al Diagn Cytopathol (2011) Questions to address TBS has given us uniform terminology, but not uniform use Greatest variation is in ND, AUS, and FOL What are the problematic patterns? What do we know about these patterns? How can we try to resolve these diagnostic problems? Why do we use the Bethesda categories so differently? 1

3 AUS/FLUS Scenarios 1. Atypia hindered by preparation artifact 2. Hürthle cells only, in a patient with Hashimoto s Multinodular goiter 3. Hürthle cells only, but sparsely cellular 4. Focal architectural features of FOL 5. Focal cytologic features of PTC 6. Atypical cyst lining cells 7. Focal marked nuclear atypia 8. Atypical lymphoid infiltrate 9. Not otherwise specified TBS Reporting in Practice Study ND B AUS FOL SUS M A Median Scenario A: Trading ND for AUS/FLUS Adapted from Krane et al Cancer Cytopathol (2011) plus Wu et al Diagn Cytopathol (2011) AUS and ND Rates in TBS are Inversely Related Non diagnostic Literature review, by institution BWH, by cytologist 25 y = -3.22x R 2 = 0.91, p < Non-Diagnostic Rate (%) Adequacy criteria (quality and quantity): 6 groups, each with at least 10 benignappearing, well visualized follicular cells Exceptions: Solid nodules (U/S) with cytologic atypia Solid nodules with inflammation (Benign: thyroiditis, abscess, etc...) Colloid nodule: abundant colloid (Benign) VanderLaan et al Am J Clin Pathol In press Management: Repeat FNA (with U/S guidance). Non diagnostic (<6 groups of 10 follicular cells) AUS/FLUS: Atypia with air drying artifact Non diagnostic Cyst fluid only Ali and Cibas, TBSRTC,

4 AUS/FLUS: Atypia with clotting artifact The AUS vs ND Dilemma Per TBS, both should prompt repeat aspirate AUS starts the clock toward surgery A single AUS may prompt surgery (see for example Poster #69 [Archuletta et al, >90% of patients went to surgery after a single FLUS]) Conclusion: Generally better to opt for ND rather than have a compromised specimen result in overtreatment TBS Reporting in Practice Study ND B AUS FOL SUS M B Median Scenario B: Trading AUS/FLUS for FOL Adapted from Krane et al Cancer Cytopathol (2011) plus Wu et al Diagn Cytopathol (2011) AUS/FLUS Scenarios (versus FOL) 1. Atypia hindered by preparation artifact 2. Hürthle cells only, in a patient with Hashimoto s Multinodular goiter 3. Hürthle cells only, but sparsely cellular 4. Focal architectural features of follicular neoplasm 5. Focal cytologic features of PTC 6. Atypical cyst lining cells 7. Focal marked nuclear atypia 8. Atypical lymphoid infiltrate 9. Not otherwise specified (Suspicious for a) Follicular Neoplasm Diagnostic Category IV. Follicular neoplasm or Suspicious for a Follicular Neoplasm Identification of architectural changes associated with follicular carcinoma Predominance of microfollicles / trabeculae Absence of nuclear changes of PTC Scant colloid Cellular Management: Lobectomy Seems to be a challenging pattern [see poster #74, Eilers et al; 22% FN in ASCP NonGYN Assessment Program] Risk of Malignancy Usual Management 15 30% Lobectomy Enlarged overlapping nuclei Central cytoplasmic core and cytoplasm does not extend beyond follicle Abele and Levine Cancer Cytopathol (2010) 3

5 AUS/FLUS: Microfollicles Abele and Levine Cancer Cytopathol (2010) Atrophic follicles AUS/FLUS: Microfollicles Renshaw Cancer Cytopathol (2011) Scant atrophic pattern Sparsely cellular, predominantly microfollicular Small nuclei Fragile cytoplasm Alternatively, distinct cytoplasmic perimeter AUS/FLUS: Microfollicles Renshaw Cancer Cytopathol (2011) Cellular atrophic pattern Mixed pattern, often focally microfollicular Complex macrofollicles ( Chinese characters ) AUS/FLUS: Microfollicles Renshaw Cancer Cytopathol (2011) Scant atrophic pattern: 21% malignancy risk Cellular atrophic pattern: 34% malignancy risk Conclusion: Atrophic microfollicular pattern does not warrant a benign diagnosis. AUS/FLUS: Microfollicles VanderLaan et al Am J Clin Pathol (2011) Liquid based preparations Architectural atypia (vs cytologic, both, unspecified) Sparsely cellular, predominantly microfollicular AUS/FLUS: Microfollicles VanderLaan et al Am J Clin Pathol (2011) Architectural atypia 16% malignancy risk All other patterns 33% malignancy risk Conclusion: Isolated architectural atypia has half the risk of other patterns of malignancy in liquid based preparations. Crowding 4

6 AUS/FLUS: Microfollicles Outstanding issues Risk of malignancy comparable to FOL Is AUS justified? Mixed architectural patterns TBS says predominance of microfollicles for AUS How to handle lower percentage mixed patterns? Conclusion: Isolated architectural atypia poses a low, but significant, risk of malignancy. AUS use is justified. Follicular Neoplasm, Hürthle cell type Hypercellular with scant colloid Renshaw Cancer Cytopathol (2002) Carcinoma also has one of the following: Small cell dysplasia Large cell dysplasia Crowding Dyshesion AUS/FLUS: Hürthle cell atypia Cellular Hürthle cell aspirate in MNG or Hash Special case in TBS Either AUS or FOL is considered acceptable Repeat FNA not helpful AUS/FLUS: Hürthle cells Some guidance for Hash Roh, Jo et al Am J Clin Pathol (2011) PPV of FOL in Hash 9.5% (vs 25.2% without Hash) Conclusion: AUS justified in Hürthle cell rich aspirate with clinical Hash AUS/FLUS: Hürthle cells Renshaw Cancer Cytopathol (2010) Atypia, rule out Hürthle cell neoplasm Includes all AUS Hürthle cell patterns 7% malignancy risk vs 25% for AUS overall TBS Reporting in Practice Study ND B AUS FOL SUS M C Median Conclusion: Hürthle cell type AUS is low risk, falling between B and FOL. AUS is justified. Scenario C: Trading Benign for AUS/FLUS Adapted from Krane et al Cancer Cytopathol (2011) plus Wu et al Diagn Cytopathol (2011) 5

7 TBS Reporting in Practice Study ND B AUS All abnormals (AUS+FOL+SUS+M) A B C Median Both scenarios A and C: AUS to increase sensitivity Adapted from Krane et al Cancer Cytopathol (2011) plus Wu et al Diagn Cytopathol (2011) Percent malignant outcome AUS Rate vs. Malignant Outcome: BWH Individual Practitioner y = 2.6x R 2 = 0.46, p < AUS diagnosis percentage VanderLaan et al. Acta Cytol; In press. Malignancy Rate (%) % 10 Literature review: AUS Rate vs. Malignant Outcome y = x R² = , P<0.05 What malignancies are we diagnosing with AUS? Diagnosis Number of cases % Papillary carcinoma Follicular carcinoma Other % AUS Rate (%) Adapted from VanderLaan et al. Acta Cytol; in press and Wu et al. Diagn Cytopathol Data pooled from published studies. AUS/FLUS Scenarios (versus SUS) 1. Atypia hindered by preparation artifact 2. Hürthle cells only, in a patient with Hashimoto s Multinodular goiter 3. Hürthle cells only, but sparsely cellular 4. Predominantly microfollicles, sparsely cellular 5. Focal cytologic features of PTC 6. Atypical cyst lining cells 7. Focal marked nuclear atypia 8. Atypical lymphoid infiltrate 9. Not otherwise specified AUS/FLUS with cytologic atypia or focal features of papillary carcinoma 6

8 Papillary carcinoma Predictive value of AUS/FLUS with focal papillary features Weber et al Acta Cytol (2008) 44% malignancy risk Renshaw Cancer Cytopathol (2010) 38% malignancy risk VanderLaan Am J Clin Pathol (2011) 28%/48% malignancy risk Olson et al (Poster 82) 47% malignancy rate for AUS with nuclear atypia Conclusion: AUS/FLUS with focal features of papillary carcinoma has a consistently higher risk of malignancy than other patterns of AUS/FLUS AUS/FLUS with focal papillary features Weber et al Acta Cytol (2008) Nuclear inclusions, psammoma bodies, and squamoid cytoplasm associated with malignancy Walts et al (Poster 84) Nuclear crowding, nuclear grooves, nuclear inclusions (but not nuclear enlargement) 61% malignancy rate when 1 present Conclusion: AUS/FLUS with focal features of papillary carcinoma is better called SUS when well developed features are present What malignancies are we diagnosing with AUS? Diagnosis Number of cases % Papillary carcinoma Follicular carcinoma Other 11 4 Papillary Number % Subtype Follicular variant Other Data pooled from published studies. The Problem of Follicular Variant of PTC Encapsulated follicular lesions with borderline cytologic features Extremely poorly reproducible Indolent tumors posing minimal risk Alternative terminology to carcinoma has been proposed but not widely accepted What are the goals of thyroid FNA? To detect malignancy Yes, but 7

9 Thyroid Cancer Cancer Cytopathol (2010) Since 1990, incidence has doubled Death rate unchanged >95% 5/25yr survival American Cancer Society. Cancer Facts & Figures What are the goals of thyroid FNA? The aim of thyroid cancer diagnosis is less the reduction of mortality and more the prevention of morbidity from local recurrence and metastases What are the goals of thyroid FNA? To detect malignancy that will benefit from early intervention Medullary, poorly diff, etc. Papillary carcinoma (aggressive variants) Abele and Levine, Cancer Cytopathol (2010) Papillary Carcinoma by Subtype M preferentially high risk Malignant Suspicious for Malignancy AUS Fol Var Conv Var High-risk Vanderlaan and Krane (Poster 76) AUS preferentially FVPTC Percent of Nodules Percent of Nodules High risk Features of PTC are Predicted by TBS Category Malignant Tumor Stage Suspicious for Malignancy AUS T1 T2 T3 T4 Extrathyroidal Extension Malignant Suspicious for Malignancy AUS Present Absent Lymphovascular Invasion Malignant Malignant Suspicious for Malignancy AUS Present Absent Lymph Node Metastasis Suspicious for Malignancy AUS N1 N0 NX 8

10 TBS Reporting in Practice Why do we use the Bethesda categories so differently? Study ND B AUS All abnormals (AUS+FOL+SUS+M) A B C Median Both scenarios A and C: AUS to increase sensitivity Scenario B: Avoids AUS to increase specificity Adapted from Krane et al Cancer Cytopathol (2011) plus Wu et al Diagn Cytopathol (2011) Different Philosophies of Thyroid FNA My opinion Maximize specificity Eliminate FPs Content to detect clinically significant carcinomas Not concerned about borderline/indolent lesions (like marginal FVPTC) Minimize unnecessary treatment Maximize sensitivity Eliminate FNs Believe it is important to detect all carcinomas Important to detect borderline lesions (or perhaps afraid to miss borderline lesions?) May result in overtreatment We should strive to diagnose entities with clearly defined criteria (eg, classical papillary, tall cell papillary, medullary, anaplastic, follicular neoplasms) Limit AUS/FLUS use Follow most AUS/FLUS conservatively with serial U/S, exam, and/or repeat FNA Consider alternate terminology for borderline FVPTCs to reduce overtreatment Consider limited molecular testing for indeterminate lesions Cytomorphologic Conclusions AUS/FLUS is the most problematic diagnosis in TBS Avoid AUS/FLUS as alternative to ND Architectural atypia patterns are low risk Hürthle cell atypia is low risk in the absence of features associated with carcinoma Hürthle cell lesions with Hash are low risk Focal, but well developed, features of papillary carcinoma should call SUS rather than AUS/FLUS Future of Thyroid FNA TBS provides framework for uniform terminology Need for greater consensus and reappraisal of the diagnosis and treatment of borderline lesions Molecular testing will likely be part of the picture 9

11 What is the Role of Molecular Testing in Resolving Indeterminate Diagnoses? Can molecular testing effectively triage AUS FNAs for surgery? Common differentiated thyroid cancer mutations BRAF 35 75% PTCs V600E active More aggressive behavior RET/PTC 10 20% PTCs RAS 15 50% PTC, FTCs, FAs PAX8/PPAR 30 40% FTCs, 2 10% FAs, <5% FVPTCs Although mutations found in ~70% of DTC, no one is completely sensitive/specific for malignancy. Nikiforov YE. Mod Pathol (2011) Robbins and Cotran Pathologic Basis of Disease, 8 th ed Molecular Testing for Malignancy Revised ATA guidelines: Recommend use of molecular panel for indeterminate FNAs Nikiforov et al. J Clin Endocrinol Metab (2011) Molecular Testing for Benignity Afirma Gene Expression Classifier by Veracyte Monroe et al (Poster 132) has the potential to reduce the number of diagnostic thyroidectomies on nodules, with indeterminate FNA cytology, by more than 50%. 10

12 Diagnostic challenges in Lung Kristen Atkins, MD Epithelial tumors A diagnosis of non-small cell carcinoma, small cell carcinoma and carcinoid used to be sufficient cytologic interpretations for management. With increased knowledge on mutations and genetic alterations specific to types of cancer as well as treatment protocols, this distinction is sometimes not enough. When the patient is low stage, usually meaning maximally only hilar lymph nodes are involved, surgery is the mainstay for treatment. o In this scenario, NSCLC is sufficient cytologically as the exact subtype can be discerned on the excision specimen. o However, when the FNA or bronchoscopic specimen is the only diagnostic tissue due to advanced stage (usually contralateral mediastinal lymph nodes and above), more precise classification is becoming imperative for treatment options. EGFR expression Epidermal growth factor receptor is a surface receptor that is activated by epidermal growth factor and transforming growth factor. EGFR dimerization activates intracellular tyrosine kinase, which eventually stimulates several signal transduction cascades important in DNA synthesis, proliferation, cell adhesion and migration. Mutations that result in upregulation of EGFR promote constant cell division, making a cell more prone to mutations and cancer development. Some lung cancers have been found to have EGFR mutations, often resulting in increased gene copy number (similar to Her 2 neu amplification in breast carcinoma). EGFR mutations can be detected by polymerase chain reaction (PCR) and amplification of the gene can be seen by fluorescence in situ hybridization (FISH). EGFR-targeted drugs One study used EGFR Tyrosine Kinase Inhibitors (TKI) gefitinib and erlotinib in an unselected NSCLC population and found a minority of patients had a marked response. This was most prominent in E. Asians, women, and those with adenocarcinomas (particularly bronchoalveolar carcinoma). EGFR TKI responsive tumors contain somatic mutations in the EGFR TK domain (usually a point mutation or a substitution mutation). One can screen for these mutations to predict which tumors will respond to EGFR TKIs. These tumors may inherently be more indolent. IHC for EGFR does not correlate with the presence of the mutations.

13 In a study by Miller et al (J Clin Oncology, 2008), EGFR mutation and copy number correlated to response rate and overall survival with erlotinib. This was not true of EGFR IHC positivity. Conversely, the KRAS mutation (another signal transduction protein, GTPase, commonly found in pancreatic cancers) is an indicator of a poor prognosis and predicts failure to EGFR TKI therapy, regardless of whether there is an increased copy number of EGFR. 1. Differentiating non-small cell carcinomas, when cytology works, when it doesn t: Challenges come with high grade cancers. Differentiation of poorly differentiated squamous cell carcinoma vs large cell neuroendocrine carcinoma, vs high grade adenocarcinoma can be tough. Helpful IHC p63 TTF1 CK 5/6 Chromogranin, synaptophysin, CD56 Sq cell carcinoma Adenocarcinoma -/+ (30%) Small cell carcinoma -/+ (20%) + (75) + (85) Large cell neuroendocrine -/+ (20%) Small cell vs basalooid squamous cell carcinoma Spindle cell tumors and Clear cells Differential Diagnosis for spindle and clear cell tumors of the lung PRIMARY Pulmonary hamartoma Clear cell BAC Acinic cell carcinoma PEComa METASTATIC Melanoma Renal cell carcinoma Leiomyosarcoma Hepatocellular carcinoma Adrenal cortical carcinoma

14 Gynecologic clear cell carcinomas PEComa Clear cell tumors (sugar tumors) of the lung are synonymous with PEComas They are rare, and thus far benign The cytomorphology of clear cell tumor of the lung overlaps with primary and metastatic processes (melanoma, conventional renal cell carcinoma). Most metastases will show atypia, which is usually absent from PEComa. Spindle cell melanoma and renal cell carcinoma are cytomorphologically probably the two closest neoplasms to PEComa. Thinking about the diagnosis paramount! Eosinophilic granuloma/langerhans cell histiocytosis Mesothelial cells vs malignancy The general rules of malignancy versus reactive mesothelial cells applies to both sites. Perhaps the simplest and most helpful points to recognizing malignancy are: In the setting of a malignant history, be suspicious before you have even looked at the slide, particularly for breast, pancreas, melanoma and stomach (which notoriously look like mesothelial cells. In the setting of ascites of unknown etiology, look carefully (see next point). When the slide is cellular, be cautious o If the cellularity is due to numerous mesothelial cells, are they atypical, are there inclusions, targetoid bodies o If the cellularity is due to histiocytes, are you sure they are histiocytes? Are the nuclei consistently eccentric (signet ring carcinoma). In the setting of ascites and liver disease, uremia, chemotherapy, be very careful and expect reactive mesothelial cells. When the slide is cellular but mixed, be cautious (not to over-interpret). This is a great setting for reactive mesothelial cells. Is there a spectrum of atypia? Don t be afraid to grab an immunochemical stain for equivocal cases- make a cell block if material is left over. Differential Immunostaining Immunostain Mesothelial cells Mesothelioma Adenocarcinoma mcea* Negative Negative Positive LeuM1 Negative Negative Positive (76-94%) B72.3* Negative Negative Positive Ber-EP4 Negative Negative Positive EMA** Negative (Few weak positive) cells) Positive (strong membrane) staining) Positive CK 5/6 Positive Positive Negative Calretinin Positive Positive Negative

15 HMBE-1 Positive membranous Positive membranous Positive (lung) cytoplasmic WT-1 *** Negative Negative Positive (Ovary) Mammoglobin/BRST- 2*** Negative Negative Positive (breast) TTF-1*** Negative Negative Positive (lung, thyroid) * Although most adenocarcinomas are CEA positive, renal cell carcinomas and serous ovarian carcinomas are exceptions. B72.3 is typically positive in serous adenocarcinomas. **EMA appears to be most useful in distinguishing reactive from malignant effusions. *** When these stains are positive, they aid in a precise diagnosis (high specificity), although a negative stain does not exclude the diagnosis(ie lower sensitivity). First: NEVER make a diagnosis of mesothelioma unless you know the clinical and radiographic findings. Are there peritoneal masses, caking of the omentum, or is this ascites without masses? No peritoneal thickening should make a cytologist very hesitant about a diagnosis of mesothelioma. According to the US-Canadian Reference Panel, the true stromal invasion is the most reliable criterion of mesothelial malignancy. This statement alone highlights the difficulty in making the diagnosis cytologically. Although cytologic atypia in the face of peritoneal thickening or masses suggest mesothelioma, the absence of atypia is common. In fact monotony is often described on surgical specimens. IHC that might be helpful: mesotheliomas are almost always EMA and p53 positive, reactive mesothelial cells can be positive. So, a negative staining pattern for these two antibodies would be very unusual for mesothelioma. However, relying on negative staining is always a daunting task, particularly in cytologic specimens. Well differentiated mesotheliomas are usually incidental small lesions on the surface of the peritoneum. It is an unfortunate nomenclature, for most of these behave in a benign fashion. Atypical mesothelial proliferation is recommended for equivocal cases. When the cells are not immunoreactive for mesothelial or adenocarcinoma, think about melanoma and epithelioid hemangioendothelioma.

16 10/25/2011 Conflict of interest Challenges in Diagnostic Cytomorphology: Lung No conflict of interest. Kristen Atkins, M.D. University of Virginia Challenges Non small cell carcinoma and molecular testing Start with the clear and move to the tricky Epithelial Adenocarcinoma vs squamous cell carcinoma Small cell carcinoma vs basaloid squamous cell carcinoma vs lymphoid Spindle cell Solitary fibrous tumor, hamartoma, metastasis Clear cells/vacuolated cells Histiocytes and Langerhans cell histiocytosis PEComa/ Sugar tumor Metastasis Mesothelial vs other Benign or malignant Epithelioid hemangioendothelioma So much talk Comes down to necessity of EGFR and ALK mutation assessment If low stage, probably doesn t matter EGFR mutations more likely in: BAC Adenocarcinoma BAC features but other adenos as well Women but can occur in men Asian but can be other ethnicities Many adenocarcinomas have mixed histology 1

17 10/25/2011 BAC VOLUME Columnar cells Inclusions, some atypia Not an outright diagnosis for cytology BAC cytologic features Overlap with well differentiated adenocarcinoma Columnar Enlarged nuclei (compared to internal control) Pseudonuclear inclusions, grooves (Like thyroid papillar carcinoma) Psammoma bodies 2

18 10/25/2011 A notable exception Small Cell Carcinoma Mucinous BAC Usually not EGFR mutated Often has k ras mutation Cytologically has BAC features (inlusions, grooves, columnar) Stains like colon might be TTF, is often CK 20 + Very blue Nuclear streaking artifact Variability Single cell apoptosis Nuclear molding, chromatin pattern Small Cell Carcinoma Basaloid Squamous Cell Carcinoma 3

19 10/25/2011 Basaloid Squamous Cell Carcinoma Basaloid Squamous Cell Carinoma Small Cell Carcinoma High N:C DNA Streaking artifact Nucelar molding Single cell necrosis Tissue Fragments No Psuedoglandular spaces No Single keratinized cells usually No Beware Intrapulmonary lymph nodes Well circumscribed Can look like small cell! Spindle cells Differential is extensive Practical is limited Reactive fibroblasts: hypocellular Solitary fibrous tumor Hamartoma Sugar tumor aka PEComa Spindle cell tumor Smooth muscle, neural, variety of sarcoma metastasis Spindle cells smooth muscle Smooth muscle cells 55 year old woman with a lung mass Can be primary or metastatic or benign metastasizing In a woman, think about uterine source Can be part of a hamartoma Lymphangioleiomyomatosis is usually radiographically clear and the lesions are small 4

20 10/25/2011 Solitary Fibrous Tumor Well circumscribed and near pleura Usually benign but can be malignant and metastasize CD 34 + Hamartomas also well circumscribed and peripheral Chondroid Hamartoma Vacuolated cells Bubbly cytoplasm Clear cell tumors Histiocytes Langerhans cell histiocytosis Spindle and epithelioid cells Eosinophilic granuloma Can present as a stellate mass In smokers Young lesions may have numerous eosinophils Old lesions may be scarred 5

21 10/25/2011 Diff Dx Cytoplasmic vacuoles: renal cell carcinoma, adrenal cortical tumors, melanoma, clear cell bronchogenic tumors Spindle cell neoplasm smooth muscle, neural, solitary fibrous tumor, melanoma Spindle cell melanoma Renal cell carcinoma SMA HMB45 PEComa (Perivascular epithelioid cell) oma AKA sugar tumor Sporadic and associated with tuberous sclerosis 6

22 10/25/2011 PEComa Melanoma SMA + HMB Melan A + + S100 /+ + Peripheral based lesions PEComa RCC SMA + /+ RCC +/ Hamartomas Solitary Fibrous tumor Epithelioid hemangioendothelioma Mesothelioma Melan A/ HMB 45 + Mesothelial cells Mesothelioma When bland can be impossible to diagnose When anaplastic can be impossible to distinguish from adenocarcinoma on cytology alone 7

23 10/25/2011 Epithelioid Hemangioendothelioma Metastasis When known primary comparing to prior is usually the most effective step Suspecting unknown primary Multiple lesions, doesn t fit usual pattern of lung carcinoma Let s try a few 8

24 Body Cavity Fluids Body Cavity Fluids Effusions: Can be benign or malignant, but are always pathologic Richard M. DeMay,MD Professor of Pathology Director of Cytopathology The University of Chicago Transudates: Usually benign Exudates: Can be malignant Mesothelial Cells Mesothelial Cells Mesodermally derived epithelial cells Line body cavities: Pleura, pericardium, peritoneum Normal: Single layer, flat (squamous) cells Effusions: Reactive Cells: plump, cuboidal Nuclei: active DDx: Cancer Common dilemma Mostly single Small lobulated groups Histiocytes Windows Hugging Mesothelial Cells Vacuoles Common, often nonspecific Cells: Variable size Nuclei: Reactive changes common Cytoplasm: Dense center, lacy skirts Degenerative Lipid Glycogen Clear, glassy Mucin => adenocarcinoma Degenerative Vacuoles Page 1

25 Marked Reactive Changes Pulmonary embolism or infarct Active cirrhosis or hepatitis Uremia Note blebs (degenerative) Pancreatitis Long term dialysis ChemoRx Radiation and chemotherapy Repeat tap can be helpful! Hepatitis Uremia Pancreatitis Histiocytes Practically ubiquitous Nuclei: beans, mitoses! Cytoplasm: lacy throughout, phagocytosis typical DDx: Mesothelial Cells Hemosiderin No: Molding, hugging, windows, tight groups Other Inflammatory Cells Eosinophils Usually => prior tap Infx, infarct, allergy, Not: Cancer, Hodgkin Charcot Leyden crystal Lymphocytes Very common, usually nonspecific Neutrophils Nonspecific, but often infectious > markedly reactive meso s Rarely numerous in malignant effusions Psammoma Bodies Not diagnostic of cancer! Mesothelial hyperplasia Endosalpingiosis Cancer: Ovary most common in women Collagen Balls Odds & Ends Usually peritoneal washes: Associated w/ mesothelial cells Benign proliferation Pick ups Curschmann spiral Megakaryocyte Malignant mesothelioma Rare in adenocarcinoma Hematoxylin Body LE Cell Sickle cell Ciliocytophthoria Page 2

26 Tuberculosis Turbid, yellow or metallic green (pseudochylous effusion) Classic cytology: Lymphs abundant Mesothelials rare Histiocytes present, giant cells rare TB, T cells; Lymphoma, B cells => flow cytometry (TGFF) Clusters: Benign clue! Rheumatoid Effusion Pseudochylous (metallic) classic Cyto: Rheumatoid Granuloma Necrotic debris: ~diagnostic Epithelioid histiocytes Multinucleated giant cells Mesothelial cells rare Bizarre cells, yet benign! Note flocculent debris Malignant Effusions Diagnosis of Malignant Effusions Any cancer can cause an effusion Common: Lung, Breast, Ovary, GI tract In oncology pts, most are malignant Can be benign (eg, CHF, Cirrhosis, Rx) Abdominal/pelvic malignancy ascites before pleural effusion Key: Foreign cells Diagnosis of Malignant Effusions Mesothelial Pattern Carcinoma When cytology is easy, it is very, very easy Page 3

27 When it s difficult Two helpful suggestions Mucicarmine stain Submit more material, if fluid reaccumulates If doesn t, probably benign Improves cellular preservation Benign atypia tends to resolve General Features Malignant Effusions Key: Foreign cells with foreign features ie, discrete or extra population of cells Some possible exceptions: Mesothelioma: Native cells, can be bland Tumor mimics reactive mesothelial cells eg, breast cancer, serous ovarian tumors Tumor Diagnosis: No single feature Cell Groups Cell Groups Group Contours Cell Shape Nuclei Cytoplasm Cell Surface Background + clues to primary Large groups: Suspicious Cell balls, papillae, glands Single tumor cells: Lymphoma (classic) Breast or Gastric CA [Sarcoma, Melanoma, Mesothelioma] Carcinoma Lymphoma Cannonballs Papillae Suggest Breast CA in a woman Pleural: Lung, Breast Ascites: Ovary, GI tract [Mesothelial hyper/neoplasia] Page 4

28 Caterpillars Single file chains Cell Shape All cells tend to round up in fluids (makes differential dx difficult) Breast, Pancreas, Gastric, Uterus [Lung (small cell), Mesothelioma] Bizarre shapes => malignancy except rheumatoid effusion Nuclei Intracytoplasmic Lumens (ICLs) Malignant nuclei = sine qua non Classic cytologic criteria: Enlarged, pleomorphic High N/C ratios Irregular membranes Abnormal chromatin Large/abn nucleoli Hyperchromasia variable May not be prominent due to proliferation Mitoses not helpful: Unless clearly abnormal Suggest breast cancer in women Targetoid Structure Sharp outline Clear Space Bull s eye (mucin) Signet Ring Cells Extreme Vacuolization Large, glassy, empty looking vacuoles Gastric, GI tract CAs Other Adenocarcinomas Ovary, Pancreas Page 5

29 Cell Surface Cilia Colorectal Carcinoma Well defined cell borders: Favor malignancy Ill defined cell borders: Favor benign (Mesos or histiocytes) Endosalpingiosis Microvilli Dirty necrosis Terminal bar Fat stogy Cilia: Strongly favor benign Exclude prominent microvilli Ciliated carcinoma very rare Cigar shaped nuclei Terminal bars Dirty necrosis Ciliated CA (very rare!) Small Cell Carcinoma Pseudomyxoma Peritonei Caterpillars Caterpillars: Usually easy diagnosis Clusters: Can be difficult to detect in busy inflammatory background Clusters Thick mucus ( jelly belly ) Endocervical like cells in mucous lakes Other cells sparse / absent (ie, meso s, histiocytes, WBCs) Hepatocellular Carcinoma Ascites common (cirrhosis) Tumor cells rare Malignant nuclei INCIs common Granular cytoplasm Hyaline globules Bile production (possible) Squamous Cell Carcinoma Common malignancy, but rarely sheds Thick, dense cytoplasm Distinct cell borders Clues: Bizarre shapes Keratin rings, pearls DDx: Sarcomas, rheumatoid effusion Also benign squames: cyst, fistula, contamination Page 6

30 Lymphoma/Leukemia Mercury droplets Primary Effusion Lymphoma Single cell pattern no tissue aggregates Larger cells, easier dx Clues to look for: Irregular membranes (eg, cleaves, knobs) Massive karyorrhexis (apoptosis) Flow cytometry for difficult cases (TGFF!) Mostly homosexual men with AIDS Also, HIV ( ) immunocompromised pts usually elderly men Particularly big & ugly lymphoma cells Bridges immunoblastic & anaplastic NHL HHV 8 & EBV frequently positive Hodgkin Lymphoma Rare diagnosis in fluids! Reed Sternberg cells Hodgkin cells Eosinophils rare DDx: Large Cell NHL; Metastatic Malignancy Clinical history! Melanoma Dark brown/black fluid Big, ugly tumor cells Binucleation common Macronucleoli; INCI Melanin often sparse Problems: Mimics many tumors, late recurrence, unknown 1 Sarcoma Effusions: Primary usually well known Pleomorphic (eg, TTFKA MFH) Round (eg, Neuroblastoma) Spindle (eg, Leiomyosarcoma) DDx: Mesothelioma, SCC, Rheumatoid All cells round up in fluid => DDx wide Germ Cell Tumors Suspect diagnosis: Big Ugly Tumor cells in young patient Important diagnosis, may respond to Rx Page 7

31 Mesothelioma Epithelioid Sarcomatoid Biphasic Other epithelial subtypes: clear cell, oncocytoid or granular cell, tubulopapillary, large polygonal cell, polyhedral stromal mucin producing, medullary, small cell, lymphohistiocytoid Mesothelioma: Fluid Cytology Often 1 st diagnostic specimen Fluid highly viscous (hyaluronic acid) Exudate with high cell count remains high after repeated taps Hyaluronic acid >8 mg/dl CEA (ng/dl): Meso <30; AdCA >150 Atypical mesothelial cells Cytodiagnosis of Mesothelioma In theory, easy Malignant cells ~ mesothelial Morphologic kinship No foreign cells In practice, difficult Two edged sword: WD looks benign PD difficult to classify Cytodiagnostic Features Cellularity Cell groups Cells Nucleus Cytoplasm Vacuoles Cell borders Background Not every feature is present in every case! Cellularity Cell Groups Important diagnostic clue: More and bigger (mesothelial) cells in more and bigger clusters Knobby Single Cells Papillae Complex, irregular Caterpillars Page 8

32 Cell Arrangements Cells Hugging Collagen cores Windows Tumor cells look mesothelial Key diagnostic feature! Cells Nucleus Spindle + Epithelioid Histiocytoid Spindle N/C ratio maintained > Uniformity Mesothelioid Gigantocyte Macronucleoli Mitoses Cytoplasm Vacuoles Lipid Glycogen Hyaluronic Acid Page 9

33 Cell Borders Blebs Background Microvilli Skirts Mesothelioma: Ultrastructure Two Part Mesothelioma Panel Long, slender branching microvilli (AdCA: short, stubby microvilli; glands) TTF 1 1st Part: Mesothelial Markers (+) Calretinin, WT 1 (front line) CK 5/6, CAM 5.2* (secondary) *Does not exclude carcinoma, but helps detect Calretinin, CK 5/6 ( ) cases, often sarcomatoid 2nd Part: Epithelial Markers ( ) CEA, Ber Ep4, BG 8, MOC 31 to exclude adenocarcinoma Strongly recommend cell blocks!!! Metastases Two take home points Foreign cells with foreign features: One of these things is not like the other Page 10

34 Mesothelioma g{tç~ çéâ Malignant cells that look mesothelial in origin More & bigger cells, in more & bigger clusters Page 11

35 10/21/2011 Challenges in Diagnostic Cytomorphology - PANCREAS Syed Z. Ali, M.D. Professor of Pathology and Radiology Director of Cytopathology The Johns Hopkins Hospital Diagnostic Challenge Pancreas can be difficult to target/interpret Difficult Anatomic Location Path of the traversing needle Diagnostic Challenge Pancreas is the 3 rd most common organ where misinterpretation of normal cellular constituents lead to an erroneous neoplasm diagnosis # # NA Young, et al. Arch Pathol Lab Med

36 10/21/2011 Diagnostic Challenge Diagnostic Challenge ~ 10% of Whipple surgeries performed for presumed malignancy (clinical and/or FNA findings) reveal benign disease on histopathology ~ 25% of these have autoimmune pancreatitis (AIP) or lymphoplasmacytic sclerosing pancreatitis Diagnostic Challenge (JHH, n=442, ) 47 (10.6%) were negative for neoplastic disease 40 - Clinical suspicion of malignancy (40) 5 - False-positive cytology 12.5% (or ~1% of all Whipples) Follow-up 8 (17%) alcohol-associated chronic pancreatitis 4 (8.5%) gallstone-associated pancreatitis 6 (12.8%) "ordinary" chronic pancreatitis of unknown etiology 11 (23.4%) AIP # Abraham SC, et al. Am J Surg Pathol

37 10/21/2011 FNA Of The Pancreas - A Multi Disciplinary Approach Surgeon Radiologist Clinical Findings Radiologic Features Chemical Analysis Cytomorphologic Interpretation Gastroenterologist Cytopathologist 3

38 10/21/2011 Diagnostic Challenge Non-neoplastic Cysts Pseudo Cyst Lack of epithelium High amylase Low CEA Diagnostic Challenge Lymphoepithelial Cyst 4

39 10/21/2011 Diagnostic Challenge Solid Neoplasms Of The Pancreas [Exocrine Pancreas] Ductal Adenocarcinoma Variants Mucinous Non-cystic Adenoca (Colloid) Medullary Carcinoma Adenosquamous Carcinoma Anaplastic Carcinoma Undiff Ca With Osteoclast-like Giant Cells Acinar Cell Carcinoma Pancreatoblastoma Diagnostic Challenge 5

40 10/21/2011 6

41 10/21/2011 Immuno Profile CK 7 +, CK 20 - CEA +, CA MUC1 +, MUC4 + MUC2- (Cf IPMN And Mucinous Cystic Neo) Chromogranin+ (Focal, Scattered Cells) DPC4 (SMAD4)- (Loss/Inactivation-55%) Other Markers Mesothelin Fascin 7

42 10/21/2011 8

43 10/21/2011 9

44 10/21/

45 10/21/2011 Diagnostic Challenge Acinar Cell Carcinoma 11

46 10/21/

47 10/21/2011 Immuno Profile Positive For CAM5.2, CK8, CK18, AE1/AE3 (65%) Positive For Trypsin (100%), Lipase (77%) Chymotrypsin (38%) And Amylase (31%) Chromogranin Isolated Cells (Upto 43%) Diagnostic Challenge Cystic Neoplasms Of The Pancreas Serous Cystic Neoplasm Mucinous Cystic Neoplasm Intraductal Papillary Mucinous Neoplasm (IPMN) Solid-pseudopapillary Neoplasm Mucinous Cystic Neoplasm Intraductal papillary Mucinous Neoplasm Solidpseudopapillary Neoplasm Serous Cystic Neoplasm Gender 20:1 1:1.5 10:1 7:3 (F:M) Head/Tail Tail Head Tail=Head Tail=Head Relation to Duct Central Scar None Always None None None None None Often Cyst Contents Mucinous Mucinous Necrotic/ Hemorrhagic Serous 13

48 10/21/2011 Diagnostic Challenge Serous Cystic Neoplasm Diagnostic Challenge Mucin-producing Neoplastic Cysts 14

49 10/21/2011 Intraductal Papillary Mucinous Neoplasm (IPMN) Glistening white mucin (grossly) Smooth outlined monolayered sheets or papillary structures Abundant background mucin Mucinous M i epithelium with apically-cleared l cytoplasm, Nuclear grooves Variable degree of atypia (mild to severe) Differential Diagnoses GI tract contamination Mucinous cystic neoplasm Ductal adenocarcinoma with mucinous features 15

50 10/21/

51 10/21/

52 10/21/2011 Diagnostic Challenge Pancreatic Endocrine Neoplasms (PEN) Well-Differentiated Endocrine Neoplasm (Islet Cell Tumor) Poorly-Differentiated NE Carcinoma (Small Cell Carcinoma) 18

53 10/21/

54 10/21/2011? Role of TTF-1 Immunostain Metastatic Tumors Adenocarcinomas Lung, Stomach, Colon, Breast Renal Cell Carcinoma Hepatocellular Carcinoma Malignant Melanoma Sarcomas Leiomyosarcoma MFH MPNST 20

55 10/21/

56 10/21/

57 Mody Spectrum and Pitfalls in Cervicovaginal Cytology: Conventional, ThinPrep and SurePath Dina R. Mody, MD Director of Cytology The Methodist Hospital, Houston, TX Professor of Pathology and Laboratory Medicine Weill Medical College of Cornell University Disclosure Dr Mody has no conflict to disclose You will learn to: HPV Type Specific Distributions in Adeno & Squamous Cell Carcinomas Describe and discuss the spectrum of ECA in cervicovaginal cytology Demonstrate the pitfalls and mimickers when diagnosing ECA Illustrate the appearances of ECA in the different specimen types with side-by-side comparisons HPV33 HPV35 HPV59 HPV45 HPV18 HPV Squamous Ca Adenoca Castellsague et al. JNCI. 2006; Epithelial Cell Abnormalities Squamous Cell Atypical Squamous cells of undetermined significance(asc-us) Cannot exclude High grade SIL(ASC-H) Low Grade Squamous Intraepithelial Lesion HPV and CIN I High Grade Squamous Intraepithelial Lesion CINII, CIN III, CIS, Susp for invasion Squamous Cell carcinoma Atypical Squamous Cells ASC refers to cytologic changes suggestive of SIL, which are qualitatively and quantitatively insufficient for a definitive interpretation. 3 essential features Squamous differentiation Increased N:C ratio Minimal hyperchromasia, ch clumping, irregularity, smudging or multinucleation Note: Applies to entire specimen not individual cells 1

58 Mody Atypical Squamous Cells- of Undetermined Significance (ASC- US) Mature Cell type (superficial or intermediate) Nuclei 2.5-3X the area of normal intermediate cell nucleus Slightly increased N:C ratio Minimal nuclear hyperchromasia, irregularity in chromatin distribution or shape Nu abnormality with dense orangeophilic cytoplasm (atypical parakeratosis) Note: Applies to entire specimen not individual cells ALTS Trial and HPV testing HPV testing at least as sensitive as immed. colposcopy in detecting CIN3+ (about 95%) 55% referred to colposcopy About 8-9% of ASC-US have CIN3+ after 2 yr. F/U; HPV arm detects most earliest Parakeratosis Vs ASC-US ASC-US, HR-HPV Neg cases from study 2

59 Mody ASC-US, HR-HPV + Koilocytes, LSIL A t y p ic a l S q u a m o u s C e lls, C a n n o t e x c lu d e HSIL(ASC-H) Immature Cell types Single cells or small fragments of <10 cells Small cells with high N:C ratios(atypical immature metaplasia) Metaplastic cells with nu X normal N:C ratio closer to HSIL but other nuclear abnormalities fall short In liquid based, cells small and 2-3X neutrophil nuclei ASC-US, HR-HPV Negative Almost Koilocytes, HR-HPV + PAP interpretations preceding HSIL Biopsies ASCUS 39% HSIL 31% LSIL 20% AGUS 10% Epithelial Cell Abnormalities Squamous Cell..continued Low Grade Squamous Intraepithelial lesion(lsil) High grade Squamous Intraepithelial Lesion(HSIL)(includes CIN II,III, CIS), with features suspicious for invasion Squamous Cell Carcinoma 3

60 Mody The Bethesda System and Other Classifications Bethesda CIN Dysplasia Low grade SIL CINI Mild High Grade SIL HSIL CIN II Moderate HSIL CIN III Severe HSIL CIN III CIS Sq CA Microinv Microinv Sq Ca Sq Ca Sq Ca LSIL Criteria Changes limited to Mature cells Nuclear enlargement >3X nl int nucleus Variable hyperchromasia, (exception in liquid based) nu size, number, shape Slight nuclear membrane irregularity Koilocytosis Must have nuclear abnormalities to qualify Note differences in liquid based LSIL LB NOT HPV! Hyperkeratosis Tight halos Navicular cells LSIL+HSIL/CIN II 4

61 Mody HSIL Criteria Small less mature cells affected Single, sheets or syncytial-like aggregates Nuclear hyperchromasia, irregularity, variation in size and shape, occasional prominent folds Nucleoli generally absent except gland extension Cytoplasm may be immature/lacy, dense or rarely densely keratinized TP TP TP TP SP TP 5

62 Mody Hyperchromatic Crowded Groups (HCGs) AIS HSIL ACA Cx Sq ca cx ACA endometrium Other Carcinomas Exodus ball Aggressive endocervical sampl Follicular cervicitis LUS endometrium Tubal metaplasia MGH Atrophy Squamous Cell Carcinoma Non Keratinizing and Keratinizing types Features and diathesis vary by preparation type Cellularity also variable Diathesis usually subtle in liquid based Keratinizing Squamous Cell Carcinoma Isolated cells or in aggregates Variable size, shape, tadpoles, spindles Variation in nuclear size, shape, hyperchromasia, granularity Macronucleoli uncommon Diathesis less than in non keratinizing types, clinging diathesis in liquid based Non Keratinizing Squamous Cell Carcinoma Syncytia with ill defined cell borders Features of HSIL but cells usually smaller Variation in nuclear size, shape, hyperchromasia, granularity Macronucleoli and basophilic cytoplasm in large cell variant Diathesis more obvious, clinging diathesis in liquid based 6

63 Mody Squamous Cell Carcinoma ThinPrep Small Cell Sq cell Ca Vs Small Cell Carcinoma Endocervical Adenocarcinoma in Situ (AIS) Precursor lesion of endocervical Adenocarcinoma Nearly 100% HPV + Type 18>16 Arise in endocx or TZ glands Surface and glandular epithelium involved Endocervical AIS Extent: Single focus or all quadrants 50% have associated SIL or Invasive Sqca Colposcopy usually normal or SIL related changes Histologically 5 types Cytologic diagnosis based on architecture and nuclear features WD AIS 7

64 Mody ThinPrep Endocervical AIS/HGIL Cellular..Crowded sheets, strips,gld Configuration openings, torn glands, rosettes, palisades. Nuclei..Oval, elongated, hyperchromatic SurePath Conventional Nucleoli...Inconspicuous Cytoplasm..Granular Background...Clear/intact blood AIS 20X Conventional ThinPrep AIS strips Surepath Helpful Cytologic Criteria HSIL Vs AIS Features Strips & Rosettes Gland forms Feathering Polarity Nu Shape Chromatin Cytoplasm Background HSIL Absent Absent Absent Lost Round/irreg Coarse Dense Isolated cells AIS Present Present Present Maintain Oval/cigar Even Even Rare/abs Reactive AIS Inv Adca Strips Feath Rosettes Nu Over DiTomasso, Ramzy, Mody. Acta Cytol 1996;

65 Mody Significant Criteria, ThinPrep Nu Crow Strips Feath Rosettes Reactive HSIL AIS Ozkan F, Ramzy I and Mody DR: Glandular lesions of the Cervix on Thin-Layer Pap Tests. Acta Cytol. 48; , 2004 AIS Vs Invasive Endocervical Adenocarcinoma (ACA) Many features of AIS in early invasive ACAs Nuclear pleomorphism and irregularity Chromatinic clearing Nucleoli Loss of polarization Three dimensional/acinar groupings Single intact malignant cells Mitosis Tumor Diathesis High Risk HPV and Cervical Adenocarcinomas Author Positive Method Andersson 71% PCR L1&E6 Casellsague 87.4% PCR G5+/6+ Andersson S et al. European J of Cancer 2001; Casrellsague et al. JNCI 2006; Other HPV Related Cancers of the Female Genital Tract Small Cell carcinoma (Type 18 or 16) Papillary Villoglandular Adenocarcinoma Clear cell Carcinoma +/- Adenoma Malignum+/- Adenoid basal carcinoma (HPV16, good) Adenoid cystic carcinoma (HPV16, mets) Mimics of Endocervical Adenocarcinoma and AIS Tubal Metaplasia High grade Squamous Intraepithelial Lesion Microglandular Hyperplasia LUS endometrium Aggressive endobrush sampling Repair, Polyps, Hormonal effects Radiation ATYPICAL GLANDULAR CELLS OF UNDETERMINED SIGNIFICANCE Definition Cells showing either endometrial or endocervical differentiation displaying nuclear atypia that exceeds obvious reactive or reparative changes but lacks unequivocal features of invasive adenocarcinoma 9

66 Mody ATYPICAL ENDOCERVICAL CELLS (Probably Neoplastic) Sheets, strips, rosettes Nu crowding, overlap, Incr N/C ratio Ill-defined cell borders Palisading, Feathering, stratification Hyperchromasia with even chromatin Nucleoli Inconspicuous, Mitosis Clean or slightly bloody background Atypical Endocervicals TP Atypical Endometrials TP AGC (Endometrial) Small groups of 5-10 cells Nu slight enlarged, small nucleoli Slight hyperchromasia Ill-defined cell borders, Scant cytoplasm, vacuoles+/_ Normal endocervicals TP Normal Endometrials TP Endometrial Carcinoma Age Cellularity Configuration Nuclei Nucleoli Cytoplasm Background Peri & post menopausal Low Loose cell groups, acini, papillae Round, vesicular Multi/macro Scant, cyanophilic Diathesis 10